Body and Machine in Classical Antiquity

BODY AND MACHINE IN CLASS ICAL AN TIQUI TY
This innovative and wide-ranging volume is the first systematic

exploration of the multifaceted relationship between human bodies
and machines in classical antiquity. It examines the conception of the
body and bodily processes in mechanical terms in ancient medical
writings and looks into how artificial bodies and automata were
equally configured in human terms; it also investigates how this
knowledge applied to the treatment of the disabled and the diseased
in the ancient world. The volume examines the prehistory of what
develops at a later stage – and more specifically during the early
modern period – into the full science of iatromechanics, in the
context of which the human body was treated as a machine and
medical treatments were devised accordingly. The volume facilitates
future dialogue between scholars working on different areas, from
classics, history and archaeology to history of science, philosophy and
technology.
maria gerolemou is currently a research fellow at the Center for

Hellenic Studies, Washington. She has published widely on ancient
Greek drama, specifically on gender and madness, on Wunderkultur
and on ancient science and technology. She is the author of Technical
Automation in Classical Antiquity (2023).
george kazantzidis is Assistant Professor of Latin Literature in
the Department of Philology at the University of Patras. He is
particularly interested in the history of mental illness and the history
of emotions in antiquity. His book Lucretius on Disease: The Poetics of
Morbidity in De rerum natura was published in 2021. He is currently
working on a monograph provisionally entitled ‘Greek and Roman
Wonders: Medicine, Horror, the Sublime’.
Published online by Cambridge University Press

BODY AND MACHINE
IN CLASSICAL ANTIQUITY
edited by
MARIA GEROLEMOU
Center for Hellenic Studies, Washington
GEORGE KAZANTZIDIS
University of Patras

Shaftesbury Road, Cambridge cb2 8ea, United Kingdom
One Liberty Plaza, 20th Floor, New York, ny 10006, USA
477 Williamstown Road, Port Melbourne, vic 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India
103 Penang Road, #05–06/07, Visioncrest Commercial, Singapore 238467
Cambridge University Press is part of Cambridge University Press & Assessment,

a department of the University of Cambridge.
We share the University’s mission to contribute to society through the pursuit of
education, learning and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781316514665
doi: 10.1017/9781009085786
© Cambridge University Press & Assessment 2023
This publication is in copyright. Subject to statutory exception and to the provisions
of relevant collective licensing agreements, no reproduction of any part may take
place without the written permission of Cambridge University Press & Assessment.
First published 2023
A catalogue record for this publication is available from the British Library.
A Cataloging-in-Publication data record for this book is available from the Library of Congress.
isbn 978-1-316-51466-5 Hardback
Cambridge University Press & Assessment has no responsibility for the persistence
or accuracy of URLs for external or third-party internet websites referred to in this
publication and does not guarantee that any content on such websites is, or will
remain, accurate or appropriate.

Contents
List of Figures page vii

List of Contributors x
An Introduction to Body–Machine Intersections 1

Maria Gerolemou and George Kazantzidis
part i blended bodies 17

1 More Than a Thing: Figuring Hybridity in Archaic Poetry
and Art 19
Deborah Steiner
2 Automata, Cyborgs, and Hybrids: Bodies and Machines
in Antiquity 48
Jane Draycott
3 Not Yet the Android: The Limits of Wonder in Ancient
Automata 70
Isabel A. Ruffell
part ii the technological body 105

4 Technical Physicians and Medical Machines
in the Hippocratic Corpus 107
Maria Gerolemou
5 The Empirical, Art and Science in Hippocrates’ On Joints 126
Jean De Groot
6 Hippocrates’ Diseases 4 and the Technological Body 155
Colin Webster

vi Contents
part iii towards the mechanization of the human
body 179
7 Aristotle on the Lung and the Bellows–Lungs Analogy 181
Giouli Korobili
8 The Ill Effect of South Winds on the Joints
in the Human Body: Theophrastus, De ventis 56
and Pseudo-Aristotle, Problemata 1.24 203
Robert Mayhew
9 The Beauty That Lies Within: Anatomy, Mechanics
and Thauma in Hellenistic Medicine 218
George Kazantzidis
10 The Mechanics of the Heart in Antiquity 245
Matteo Valleriani
11 The Mechanics of Galen’s Theory of Nutrition 262
Orly Lewis
12 Iatromechanism and Antiquarianism in Morgagni’s Studies
on Celsus, 1720–1761 296
Marquis Berrey
Conclusions or From Antiquity to the Early Modern 318
Index of Passages 321

General Index 327

Figures
1.1a–b Eleusis amphora by the Polyphemus Painter, dated page 26

to ca. 670–650. Eleusis Archaeological Museum 2630
1.2 A lekythos-oinochoe of ca. 700. New York, Metropolitan 27
Museum of Art 23.160.18
1.3 A red-figure krater by the Kleophrades Painter of ca. 29
500–490. Cambridge, MA, Arthur M. Sackler Museum,
Harvard University Art Museums 1960.236. Attic Red-
Figure Vase Painters2 185.31
1.4 A red-figure cup attributed to the Kiss Painter of ca. 500. 35
Baltimore, Johns Hopkins University Archaeological
Collection B 5, Beazley Archive Pottery Database 201626
1.5 A red-figure cup from ca. 490, the name vase of the 36
Foundry Painter. Berlin, Antikensammlung, Staatliche
Museen zu Berlin F 229, Beazley Archive Pottery Database
204340. Sides A and B.
1.6 The tondo of the red-figure cup by the Foundry Painter. 38
Berlin, Antikensammlung, Staatliche Museen zu Berlin F
229, Beazley Archive Pottery Database 204340
5.1 Simple lever, FCIT (Florida Center for Instructional 128
Technology). J. L. Comstock, A System of Natural
Philosophy: Principles of Mechanics (Pratt, Woodford and
Company, 1850), 69
5.2 Moving radius, concentric circles 132
5.3a A beam with equal weights at equal distances from a 133
fulcrum.
5.3b Equal weights at unequal distances from a fulcrum. 133
5.3c Disequilibrium induced by the addition of weight to one 133
end of a balance of equal beams at equal distances.
5.3d Unequal weights in equilibrium at unequal distances from 133
the fulcrum.
vii

viii List of Figures
5.4a A healthy shoulder joint 136
5.4b A dislocation such as Hippocrates describes 136
5.5 A bone-reduction method with a plank (xulon) that has 145
a lip or projection on one side of the wooden piece
6.1 Flow chart of humours in Morb. 4, with normal flow lines 161
marked in red, pathological movements marked in grey
6.2 Peytel aryballos, 480–470 bce. Louvre CA 2183 166
6.3 Footbath with stand, late fifth or early fourth century bce. 171
Metropolitan Museum of Art, New York, 38.11.5a, b
6.4 Theagenes spring/fountainhead at Megara, ca. 600 bce 174
7.1 Detail in red-figure cup by Douris (c. 475 bce). Single 196
bellows with two tubes, after Stefou
7.2 A double bellows. Detail from the Siphnian Treasury 197
(Delphi, c. 525 bce), after Stefou
7.3 The right bag of a bellows as it is stretched out, after 197
Stefou. Black-figure kantharos (c. 550 bce)
10.1 Visual display of the twin-cylinder force pump as 249
described by Vitruvius. The visual scheme is based on early
modern reconstructions. From Vitruvius, Frontinus, and
Giocondo 1513, 183 v. Courtesy of the Library of the Max
Planck Institute for the History of Science
10.2 Visual display of the twin-cylinder force pump as 250
described by Hero of Alexandria. The visual scheme is
based on early modern reconstructions. From Hero of
Alexandria and Woodcroft 1851, 44
10.3 Bronze force pump of Sotiel coronada (Valverde del 253
Camino, Portugal), first century ce. Museo Arqueológico
Nacional of Madrid. Photo by Elena Paulino Montero,
2018
10.4 Pump proposal. Philo of Byzantium, Pneumatics. A.S. 257
3713, f. 82a, Ayasofia Museum, fourteenth century. From
Philo of Byzantium and Prager 1974, 229
11.1 An overview of the digestive-nutritive system and the 269
course of food according to Galen. Image by Yotam Giladi
and Project ATLOMY
11.2 Schematic depiction of the two layers of fibres in the 274
oesophagus and three layers in the stomach. Image by
Yotam Giladi and Project ATLOMY
11.3 The stomach during digestion. Image by Yotam Giladi 276
and Project ATLOMY

List of Figures ix
11.4 The three layers of the stomach fibres. Image by Yotam 277
Giladi and Project ATLOMY
11.5 Expelling digested matter from the stomach into the 278
intestines through the widened pylorus. Image by Yotam
11.6 The veins from the intestines through which the digested 281
matter moves into the liver. Image by Yotam Giladi and
Project ATLOMY
11.7 The heart with its ventricles revealed. Image by Yotam 286

Contributors
marquis berrey is currently a second-year medical student at the

University of Maryland School of Medicine. He was formerly
Associate Professor of Classics at the University of Iowa, where he was
editor of Syllecta Classica. He is the author of Hellenistic Science at Court
(2017), as well as book chapters and journal articles on ancient medicine.
His current historical research combines historical and biomedical
knowledge in studies of early modern neo-Latin medical writers.
jean de groot received her PhD in history of science from Harvard
University and is Ordinary Professor of Philosophy at the Catholic
University of America, where she specializes in ancient science and
philosophy. Her books include Aristotle’s Empiricism: Experience and
Mechanics in the Fourth Century bc (2014) and Aristotle and Philoponus
on Light (2015). She is writing a book on the history of ancient mechanics
and was in 2021 to 2022 a fellow in Byzantine studies at Dumbarton
Oaks Libraries and Collections in Washington, DC.
jane draycott is Lecturer in Ancient History at the University of
Glasgow. Her research investigates science, technology and medicine
in the ancient world. She has published extensively on the history and
archaeology of medicine, impairment and disability. Her publications
include the monographs Prosthetics and Assistive Technology in Ancient
Greece and Rome (2022), Roman Domestic Medical Practice in Central
Italy from the Middle Republic to the Early Empire (2019) and Approaches
to Healing in Roman Egypt (2012) and the edited volumes Women in
Classical Video Games (2022, co-edited with Kate Cook), Women
in Historical and Archaeological Video Games (2022), Prostheses in
Antiquity (2019) and Bodies of Evidence: Ancient Anatomical Votives
Past, Present and Future (2017, co-edited with Emma-Jayne Graham).
Her research has been funded by the Royal Society of Edinburgh, the
Wellcome Trust and the Arts and Humanities Research Council.
x

List of Contributors xi
maria gerolemou is a research fellow at the Center for Hellenic Studies,
Washington. She has published widely on ancient Greek drama,
Wunderkultur and ancient science and technology. She is the author of
two monographs: Bad Women, Mad Women: Gender und Wahnsinn in der
Griechischen Tragödie (2011) on drama and Technical Automation in
Classical Antiquity (2022) on ancient technology. She is the editor and co-
editor of several contributory volumes, on miracles (Recognizing Miracles
in Antiquity and Beyond, 2018), mirrors (Mirrors and Mirroring from
Antiquity to the Early Modern Period, 2020) and technology (Body
Technologies in the Greco-Roman World: Technosoma, Gender and Sex
and Technical Animation in Classical Antiquity, both forthcoming). Her
new book project, Missing Persons in the Greco-Roman World, explores the
experience of being missing and recognition in antiquity through con-
cepts such as body signs, mapping and transportation technology.
george kazantzidis is Assistant Professor of Latin Literature at the
University of Patras. He is especially interested in the history of ancient
medicine, with an emphasis on mental illness and the history of emo-
tions. His book Lucretius on Disease: The Poetics of Morbidity in De
Rerum Natura was published in 2021. He is co-editor of the Ancient
Emotions series and is currently working on a monograph provisionally
entitled ‘Greek and Roman Wonders: Science, Horror, the Sublime’.
giouli korobili is a Marie Skłodowska-Curie postdoctoral fellow at the
University of Utrecht. She studied classical philology and philosophy at
the National and Kapodistrian University of Athens (BA), at the
University of Ioannina (MA) and at Humboldt University of Berlin
(PhD). Her research interests include Aristotle and the Aristotelian
tradition, the history of science, especially biology, medicine and
meteorology, Lucretius and Seneca. She has contributed to a number
of edited volumes on Aristotle, ancient medicine and Byzantine
Aristotelian commentators and has completed a project on the first
edition of Theodorus Metochites’ Paraphrasis of Aristotle’s PA i at the
Berlin-Brandenburg Academy of Sciences and Humanities. She is cur-
rently working on a book dealing with medical analogies employed in
ancient Greek and Roman meteorological accounts. She is the author of
Aristotle: On Youth and Old Age, Life and Death, and Respiration 1–6.
With Translation, Introduction and Interpretation (2022).
orly lewis obtained her PhD in 2014 from Humboldt University of
Berlin and is Assistant Professor in the Department of Classics at the

xii List of Contributors
Hebrew University of Jerusalem. She is Principal Investigator of
Anatomy in Ancient Greece and Rome: An Interactive Visual and
Textual Atlas (ATLOMY), a unique interdisciplinary research team
studying the history of Greco-Roman anatomy and developing an
interactive research atlas of ancient anatomy. The atlas includes three-
dimensional models of the body and its parts as described by Greek and
Roman authors, as well as terminological and textual information and
research tools. She has published widely on Greco-Roman medicine
and ancient scientific method, particularly in the Hellenistic and Early
Imperial periods and with a focus on anatomy, physiology and
diagnostics.
robert mayhew is Professor of Philosophy at Seton Hall University. His
specialization is ancient Greek philosophy and science and especially the
thought of Aristotle and other early Peripatetics. His most recent publi-
cations include Aristotle’s Lost Homeric Problems (2019) and Theophrastus
of Eresus: On Winds (2018). He is the editor of The Aristotelian
Problemata Physica: Philosophical and Scientific Investigations (2015) and
co-editor of Clearchus of Soli: Text, Translation, and Discussion (2022). He
prepared the Loeb Classical Library edition of the Problemata Physica
attributed to Aristotle (2011), has recently completed a text and translation
of the On Marvelous Things Heard for a new Loeb edition of the opuscula
attributed to Aristotle (of which he is general editor) and is working on
a new Loeb edition of Aristotle’s Eudemian Ethics. He is in the early stages
of work on a volume on the fragments of Aristotle’s lost Zoïka (texts,
translation and commentary).
isabel a. ruffell is Professor of Greek Drama and Culture at the
University of Glasgow. She has published on comedy, tragedy, Roman
satire and history of technology, where her main focus is on Hero of
Alexandria’s automata.
deborah steiner is the John Jay Professor of Greek at Columbia
University. Her most recent book, Choral Constructions in Greek
Culture: The Idea of the Chorus in the Poetry, Art and Social Practices of
the Archaic and Early Classical Period (2021), follows a series of books and
articles treating Greek myth, poetry and the interface of visual and
poetic works in early Greece; she is also the author of a commentary
on two books of the Odyssey (2010). She is currently working on a book
on figural hybridity in Archaic Greece.

List of Contributors xiii
matteo valleriani is Research Group Leader at the Max Planck
Institute for the History of Science, Honorary Professor at the Berlin
Institute of Technology, Professor by Special Appointment at Tel Aviv
University and Principal Investigator at the Berlin Institute for the
Foundations of Learning and Data (BIFOLD). In his research, he
investigates processes of emergence, transformation, circulation, hom-
ogenization and oblivion of scientific knowledge in relation to its
practical, social and institutional dimensions. His current research
takes place in the framework of computational history (https://sphaera
.mpiwg-berlin.mpg.de). A further focus of his research is on the epi-
stemic function of visual material in scientific research. Within this
context, he co-develops and applies machine-learning technologies.
Among his major publications are Publishing Sacrobosco’s De sphaera
in Early Modern Europe: Modes of Material and Scientific Exchange (with
A. Ottone, 2022), De sphaera of Johannes de Sacrobosco in the Early
Modern Period: The Authors of the Commentaries (2020), The Structures of
Practical Knowledge (2017), Metallurgy, Ballistics and Epistemic
Instruments: The Nova Scientia of Nicolò Tartaglia. A New Edition
(2013) and Galileo Engineer (2010).
colin webster is Assistant Professor at the University of California–
Davis. His primary research focuses on ancient science, technology and
medicine, and he has written on Greek optics, Methodist physicians,
Hippocratic voice pathologies and the sonics of ancient healing prac-
tices. His first book, Tools and the Organism: Technologies and the Body in
Greek and Roman Medicine (forthcoming) examines when medical
theorists began to think about the body as an object with tool-like
parts (organa), and it does so while also tracking the particular material
technologies that helped inform ideas about this new epistemic object. It
emphasizes that changing technological realities shaped different
notions about the human body and its components. For his work, he
has received fellowships from the Hellman Fellows Program, Stanford
Humanities Center and the Mellon Foundation, and he is currently the
president of the Society for Ancient Medicine.

An Introduction to Body–Machine Intersections
Explanations of the human body through technological artefacts and machines

are evidence of a need that exists to bring together nature’s creations on the one
hand1 and artificial products on the other by tracing and establishing resem-
blances and correspondences between the two domains.2 A machine or, more
generally, a technological artefact is something that we can see and touch,
construct and dismantle; we can delve into the properties and functions of its
individual parts and acquire a thorough knowledge – one that comes with an
unfailing sense of power3 – of the principles which underpin its construction
and operation. When this knowledge is transferred, through metaphors based
on analogies between the artificial and the natural domain,4 to the inquiry into
the human body, it provides us with a solid point of reference on the basis of
which body parts,5 specific organs and bodily processes, as well as the purposes
of their function, can be scrutinized in more detail.6
In December 2018, a group of archaeologists, philosophers, historians and classicists came together at
the University of Cyprus to discuss body–machine interactions in Greek and Roman antiquity. The
idea for this volume has grown from that conversation. We would like to thank the University of
Cyprus and the Department of Classics and Philosophy there for their support.
1
On the concept of the ‘machine’ in Greek and Roman antiquity, see Gerolemou in this volume,
Chapter 4. See also Berryman 2009, ch. 6; 2020; and De Groot 2014, 15–19.
2
On the artificial and the natural, see Bensaude-Vincent and Newman 2007. On the application of this
polarity to ancient Greek medicine and ancient mechanics, see von Staden 2007 and Schiefsky 2007
respectively.
3
Such sense of power is most evident in cases of automata construction and derives from the realization
that a human can bring life to the lifeless, as it were. But this power also entails, naturally, a profound
sense of powerlessness: as soon as an automaton becomes independent, it inspires the fear that it is no
longer under our control. See Kang 2011, 24–5, and Gerolemou 2022, especially ch. 2.
4
On metaphor and analogy, see Aristotle, Rhetoric 1411a, with Derrida 1974; Ricoeur 1978, 9–43; Kirby
1997; Gentner, Bowdie, Wolff, and Boronat 2001; White 2010, 27–72. On the heuristic use of a wide
range of analogies in natural sciences and medicine in antiquity, see Taub 2012 and the collections of
essays in Wee 2017; compare Althoff 1997. Lloyd 1966 remains essential reading.
5
Such solid point of reference is quite often a product of ‘invention’ – a product, that is, of the metaphor
itself rather than a pre-existing comparandum that was lying hidden, waiting to be discovered. See Ricoeur
1978, 283: ‘It would seem that the enigma of metaphorical discourse is that it “invents” in both senses of
the word: what it creates, it discovers; and what it finds, it invents.’
6
The notion of purpose needs, of course, to be qualified and modified in each case, depending on
the kind of sources we are looking at. Thus, while its application is fitting in the case of, for instance,
https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

2 maria gerolemou and george kazantzidis
As has been shown by a number of modern scholars, classical antiquity
often understood and explained the human body specifically through ana-
logy – or by means of its close interaction – with a variety of technological
artefacts. This volume is the first to provide a comprehensive and systematic
review of the evolution of the body–machine concept from its first appear-
ances in archaic thought down to the time of Galen; in doing so, it builds on
the insights offered by the important work that has been done on the subject
by Geoffrey Lloyd, Heinrich von Staden, Sylvia Berryman and others.7 The
book explores the interaction – literal as well as metaphorical – of the human
body with artefacts and mechanical devices in two directions: on the one
hand, it looks into cases where man-made devices of varied complexity come
into close contact with the body in the context of medical praxis, in the form
of tools, surgical devices, instruments of traction or prosthetic limbs; on the
other hand, it examine scientific and, more broadly, cultural paradigms
according to which the properties and functions of machines are invoked
as a heuristic tool providing concrete information for understanding the
design and function of specific body parts and organs.
Interestingly enough, such paradigms begin to make their presence felt
at a period when the body’s interior remains, more or less, an uncharted
territory – while it still lies ‘beyond the threshold of perception’, as Brooke
Holmes puts it.8 Aristotle conceives of respiration as a process in which the
lungs operate in a way similar to the bellows in a smith’s forge;9 likewise,
Erasistratus’ account of the heart closely parallels the mechanics of the
water pump.10 There is a crucial difference, though, between the two, and
this relates to the fact that, while Aristotle’s knowledge of what lies inside
the human body draws, for the most part, from comparative anatomy (i.e.
the dissection of animals), Erasistratus’ observations are based on the actual
autopsy of that body’s interior;11 human dissection, during the Hellenistic
period, breaks new ground on the level of perceiving, sensing and codifying
what up to that point had remained virtually hidden and unseen. Thus,
while the analogy with the smith’s forge links a product of human craft
with the ‘imaginary’ that is developed around the physical body,12 the
Aristotle and Galen, one should be far more cautious when discussing evidence from the
Hippocratic corpus; see Gundert 1992, 465, and Holmes 2018, 73.
7
Lloyd 1966, 272–94; von Staden 1996, 1997, 1998; Berryman 2009. See also Grmek 1972, Lonie 1981
and Mayor 2018.
8 9
Holmes 2010, 129. See Korobili in this volume, Chapter 7.
10
See Valleriani in this volume, Chapter 10.
11
For this difference, see Kazantzidis in this volume, Chapter 9.
12
We borrow the term ‘imaginary’ from Holmes 2018, 73. Holmes speaks in this case of the
Hippocratic body. Compared to the latter’s obscure interior, Aristotle’s anatomical insights are, of

correspondences drawn by the Hellenistic anatomist between the heart and
the water pump are solidly rooted in empirical, tangible evidence.
This crucial difference notwithstanding, what these thinkers have in com-
mon is that they construe their analogies on the belief – sometimes spelled out
explicitly, at others subtly implied – that the human body works ‘like
a machine’, that is, seamlessly and effectively, and according to
a predetermined plan that can be mapped in all its details. Naturally, this
approach runs the risk of leaving out all those constituent qualities and aspects
of the human body that render it intrinsically dependent on tuchê and to
automaton, the ‘accidental’ and the ‘spontaneous’.13 But then again it may be
that the concept of a body part resembling a machine was driven, at least partly,
precisely by the need to exercise some kind of control on the unruly forces
inside of us. Compared to a (humoral) body in constant flux – as conceived by
the Hippocratics14 – there is something reassuring in the Hellenistic anatom-
ists’ reconfiguration of that body, amongst others, on the basis of its assumed
resemblances with crafted objects made of solid and enduring materials.15
Understanding parts of the body as machines consisting of tiny bits and
pieces that work ceaselessly does not imply a process of objectification that
turns the body into a ‘passive’, impersonal thing devoid of agency.16 Recent
studies inspired by the currents of post-humanism and the so-called ‘new
materialist’ turn have helped to shift the boundaries of the embodied self
beyond the skin that envelops it17 and to extend them so as to include, in a deep
ontological sense, the objects and the technology with which that self is
interacting.18 Drawing from these insights, the present volume is not interested
course, far more detailed and precise; however, considering that there is no conclusive evidence that
Aristotle carried out human dissection, we need to remain open to the possibility that he too is
operating on a largely speculative basis.
13
For the ‘automatic body, see Holmes 2010, 142–7, and Gerolemou 2022; Compare Lo Presti 2008.
14
See King 2013.
15
It is certainly not a coincidence, in this respect, that Erasistratus was not a great fan of humoral
theory; see von Staden 1989, 243.
16
For the reductivist effects of iatromechanics and the risk of depriving the body and the self of their
agency, see Riskin 2016, 4: ‘The classical mechanist approach to science, with its attendant
mechanical model of nature and of living creatures, relied crucially as it was developing from
around the mid-seventeenth century upon an accompanying theology . . . A purely passive artifact
devoid of agency would not have been a plausible account of living nature on its own and it won no
converts on its own. This mode of science, call it theological mechanism, relied upon a divine
Designer to whom it outsourced perception, will and purposeful action.’ Contra: Wolfe 2016, 72,
who argues that the very notion of a mechanistic, ‘dead’ materialism ‘misses the vital character of the
unique Radical Enlightenment formation’, going so far as to suggest that it may be ‘that there was no
such thing’ (p. 51); compare Wolfe 2017.
17
On dermal boundaries and the limits of the body, see Cohen 2003.
18
For post-humanism and classical literature, see the recent volume by Chesi and Spiegel 2020;
compare the collection of essays in Bianchi, Brill and Holmes 2019. For the new materialist turn,

in establishing any hierarchies between the ‘human’ and the ‘technological’;19
rather, our intention is to explore the interdependence between the two and, in
some cases, their actual fusion.20 Some of the volume’s authors, for instance,
argue for the human body’s inherent capacity to absorb external objects in
a way that puts together the natural and the artificial in a properly symbiotic
relationship.21 Accordingly, artefacts and machines are not discussed as though
they were some kind of inert matter but as dynamic products of human
ingenuity, as sites of projected agency which involve natural forces of various
sorts and imply, at points, the presence of purposeful action. A lot of post-
human and new materialist discussion in recent years has called attention to
notions such as the ‘extension’ of the body, its opening up to the surrounding
environment, natural and artificial, and its meshing with objects and construc-
tions that lie beyond the body’s limited physical boundaries. Our volume helps
to show that such an extension can be simultaneously imagined as having also
an ‘inward’, as it were, turn: the human–technology interface which we are
discussing, focused as it is in many cases on the body’s interior, confirms that
body’s capacity to extend itself by virtue of the mimetic qualities that turn its
inner landscape into a reflection of what lies out there. To put it otherwise:
a smith’s forge or a water pump are not just constructions with which a human
‘blends’ while operating them; that blending is also taking place inside the
physical body figurally understood to contain parts and organs operating just
like machines do.22 Needless to say, we are still a long way from androids and
cyborgs, that is, from robots that resemble a human being or from organisms
that are part-organic and part-machine. Still, the authors in this volume test,
time and again, the established boundaries between the living and the artificial,
in ways that help to bring out the porous conceptual environment that these
two categories inhabit, constantly feeding into each other and defining each
other’s essence.
see Canevaro, 2019. Related issues are also raised in the context of ‘distributed cognition’ theories;
see Anderson, Cairns and Sprevak 2018.
19
For the non-hierarchical material continuum which blends together human technology and moral
agency (in an eco-critical context), see Morton 2010.
20
On ‘fusion’ in the context of the human–technology interface, see Sobchack 2006, 19. Compare
Vaccari 2012.
21
See, especially, Webster and Draycott in this volume, Chapters 6 and 2; compare Gerolemou in this
volume, Chapter 4.
22
Compare Vernant 2006, 301: ‘the tool, when directly manipulated by man, is still an extension of his
own organs’. See also Aho and Aho 2008, 105, on the dual meaning of organon in ancient Greek
thought (‘bodily organ’ and ‘instrument’/‘tool’): ‘Like the other equipment that surrounds us, our
organs are always busy at work: digesting, breathing, seeing, hearing, walking, and so forth. And just
as I do not normally notice the keyboard, the desk or the chair as I type my notes on the computer,
when my organs are laboring smoothly, my body “hides” itself.’ This idea is also explored in detail in
Webster forthcoming.

The volume is divided into three sections, the first of which is entitled
‘Blended Bodies’. The authors in this section explore the beginnings of
a conceptual blending between organic and fabricated elements, and the
interaction between body and technological artefacts from archaic poetry
and Pindar down to Hero of Alexandria. In its simplest form, this inter-
action is understood as being generated at the point where man-made
constructions – for instance, an artificial limb or device – blend with the
human body, thus helping it recover its physical integrity or enhance its
functionality. On another level, a fusion can be seen as taking place within
the body’s interior space: the implements, objects and substances applied
externally to the body often pass through the surface of the skin, merging
with the patient’s inner matter. What is at stake in this case is the body’s
capacity to absorb and enfold adjacent technologies. But this section is not
concerned only with how living bodies are construed with reference to
machines and artefacts. It also explores how this interaction works the
other way around, namely by investigating cases in which an artificial
product – for example, an anthropomorphic piece of art or an automaton –
plays out similar notions of hybridity, by blurring the boundaries between
made-up objects and natural elements. Just as a body made of flesh and
bones can be understood by virtue of its interaction with technological
devices, so do certain types of technology invite an association with a living
body, through which they are appreciated aesthetically qua art objects.
This first section starts with a chapter which highlights the constant
interplay between human bodies, art and artefacts; and it ends with
a discussion considering how art and artefacts, even in the form of the
most sophisticated mechanical devices, have only a restricted mimetic
potential in regard to the representation of the human body.
Deborah Steiner’s chapter, entitled ‘More Than a Thing: Figuring
Hybridity in Archaic Poetry and Art’, examines cases where technological
and artistic attitudes and methods overlap and complement each other.
Her survey covers a wide range of texts and objects, from Hephaestus’ self-
moving tripods and animated bellows in Homer to Pelops’ composite body
and the Delphic temple of Apollo in Pindar. A close reading of these
sources reveals a constant interplay between human bodies and inanimate
matter. So, for example, Hephaestus’ tripods are designed so as to mimic
features of human anatomy; the god’s bellows look like lungs performing
respiration; Pelops’ living body would have been anatomically incomplete
had it not been equipped with an artificial element; the Delphic temple
displays a quasi-animated structure and has a life rhythm of its own. These
cases, according to Steiner, present us with figural hybrids of riddling

qualities, each one offering ‘its own amalgam of organic and fabricated
elements’ (p. 21).
Jane Draycott’s chapter, entitled ‘Automata, Cyborgs and Hybrids:
Bodies and Machines in Antiquity’, explores the use of assistive technology
as a means by which the impaired human body recovers its wholeness –
through the use of technological implements – and transforms into some-
thing new: a hybrid form of life which operates on the premise of a fusion
between art and nature. Starting from Homer and moving down to
Hippocratic and later sources, Draycott explores a wide number of mater-
ial objects which fall under the category of ‘assistive technology’, from
staffs, sticks, canes and crutches to corrective footwear and extremity
prostheses. These objects, according to the author, do not simply contrib-
ute to the recovery of the human body’s ‘wholeness’ – by allowing it to
acquire its previous, ‘complete’ form – but, crucially, they augment it and
transform it into something new: a hybrid form of life which, by virtue of
an emerging fusion between art and nature, provides us with a first glimpse
of what will in a later period evolve into the idea of the body as machine.
Similar to Hephaestus’ golden maidens and automatic tripods – serving as
assistants for the god’s disabled body while being simultaneously conceived
as a recuperative extension of that body – a prosthetic limb reinstates
completeness and enhances the human body’s function.
In the final chapter of this section, ‘Not Yet the Android: The Limits of
Wonder in Ancient Automata’, Isabel Ruffell explores the extent to which
mechanics can create, as it were, the human body from scratch, by produ-
cing its artificial replica. The answer Ruffell gives is a negative one:
machines in antiquity appealed to the mimesis of human forms only to
a limited degree. Starting from the notion of automation in Homer and
moving down through Aristotle to Hero of Alexandria, Ruffell argues that,
overall, those involved in the construction and production of ancient
automata were not particularly interested ‘in representing self-powered,
programmable versions of the human body’ (p. 70); on the contrary, the
role of the human body, or animal bodies, is strictly limited, and those
bodies that do appear are strictly limited in what they can do. According to
the author, some of these limitations are due to technical reasons and
follow from the restricted nature of available power sources and mechan-
isms; on another level, the absence of human-like forms in automatic
culture is explained as a reaction by the engineers to ‘cheap’ wonder-
working: the wonder that is sought in the technical treatises has less to
do with the ability to recreate or mimic the human body and is more
oriented towards producing complex devices of mechanical virtuosity.

Thus, while in modern culture fantasies of the automated human prevail
(consider, for instance, the penetrating presence of cyborgs in sci-fi
literature), machines in antiquity appealed to the mimesis of human
forms only circumstantially and to a limited degree.
The chapters in Part ii, entitled ‘The Technological Body’, focus on the
Hippocratic body and on its relationship to technical artefacts and instru-
ments. This section opens with two chapters that deal with medical
machines in Hippocratic practice and address questions concerning the
theoretical understanding that may – or may have not – underpinned their
use. In ‘Technical Physicians and Medical Machines in the Hippocratic
Corpus’, Maria Gerolemou offers a detailed review of the technological
implements used by Hippocratic physicians, by distinguishing them into
three categories: (a) simple instruments made of natural materials, such as
wood, wool and linen; under this category, Gerolemou argues, we can
include the doctor’s own hands, feet and heels, as we often see them being
used – in a way which simulates the activities of a real mechanism – for the
sake of exercising pressure upon the patient; (b) advanced tools such as
pestles, windlasses, levers, axles, cupping vessels and wineskins; (c) compli-
cated devices such as the ladder, the board and the bench, which come in
handy when the exertion of a greater force upon the patient’s body is
needed. The use of these types of implements in the Hippocratic corpus is
not based, according to the author, on a theoretical understanding of
machines but is anchored instead to needs of a practical nature. In
Gerolemou’s words, ‘by examining the degree to which physicians transfer
knowledge and skills from the technical to the medical domain, we can
conclude that Hippocratic physicians, while they apply technical methods
of technical knowledge to medicine, that is, by integrating the two domains
when designing medical solutions, do not act like engineers’ (p. 108). At
the same time, and although these medical devices do not claim a high
degree of complexity on a scientific level, they become quite essential for
the doctor’s understanding of his art in an ethical context. Provided that
machines can unleash energy and produce results which go beyond the
control and power of the physician, their presence in medical practice
comes with a certain awareness that they should be operated wisely and
with caution.
While Gerolemou looks at how technological implements mediate and
define – practically as well as ethically – the relationship between
Hippocratic physicians and patients, the next chapter by Jean De Groot,
‘The Empirical, Art and Science in Hippocrates’ On Joints’, focuses its
attention on asking specifically what kind of theoretical knowledge

underlies the use of those implements. In agreement with Gerolemou –
who draws attention to the banausic basis on which Hippocratic ‘tech-
nology’ operates – De Groot reaches the conclusion that the doctors’
familiarity with a number of technical devices does not necessarily
indicate an understanding of the theoretical principles that explain
their function. What is mainly at play is an experiential/intuitive
grasp of mechanical phenomena. In this context, Jean De Groot’s
chapter discusses in detail Hippocratic accounts of how to reduce
shoulder dislocation in a patient. While these accounts, it is argued,
offer evidence of uses of mechanical tools and tropes, ‘it can be difficult
to assess what sort of understanding accompanies their use’ (p. 126). De
Groot’s aim is to show that Hippocratic doctors display an ‘intuitive’
grasp of physical forces, which is significantly different from the phys-
ical science of mechanics that emerges a century or more later. Through
a close reading of the language used in medical passages which describe
the best possible treatment for dislocations, the author arrives at the
conclusion that the Hippocratics have their own understanding of
combining forced movements, which is not anchored in equilibrium
as later scientific accounts of leverage are. This empirical understanding
involves oppositions of movements and oppositions of movement and
stability; these oppositions are meant to provoke some other movement
remotely; the remote movement intended – to slide a bone back to its
joint – is highly calibrated in its effect. When examined closely, this
evidence shows that while some of these methods for applying force can
be construed as falling under the category of leverage, in other cases they
cannot. Hippocratic doctors, according to De Groot, could skilfully use
a number of technical devices and employ them successfully in the
healing process; this, however, should not lead us to believe that their
practical familiarity with these devices was also oriented towards dis-
covering the hidden causes and theoretical principles that explained
their function.
With Colin Webster’s chapter, ‘Hippocrates’ Diseases 4 and the
Technological Body’, we move a step further, from the body’s surface to
its hidden interior and from the application of a medical implement to its
metaphorical use for understanding the function of the human organism.
Webster agrees with Gerolemou and De Groot that we should be cautious
when using the term ‘mechanics’ when it comes to Hippocratic medicine,
since we are still missing the existence of a corresponding scientific field (in
the sense that we first find it during the early Hellenistic period). His

chapter looks at how technical instruments are put not only into practical
use, but how they are also employed metaphorically by the doctors as
comparable analogues to the functions of certain organs in the body.
Thus, while Hippocratic medicine does not yet cast the (interior of the)
body as a machine, it is nonetheless fair to say that it signals a first attempt
to involve tools and technologies in the conceptual construction of that
body as a smoothly operating entity. By putting forward the notion of
organic sympathy, Webster bypasses the old and important opposition
between mechanism and vitalism. He specifically illustrates how medical
tools (including plants/pharmaka) can operate in sympathy with the body
and become absorbed into that body’s ‘conceptual interior’. Through an
examination of the conceptual interface of medical implements and the
bodies that they touch, Webster arrives at the conclusion that the tools
used by a physician in the latter’s attempt to intervene in the body often
merge into explanations of the processes which are imagined as taking
place beneath its surface. A prime example is that of the motion of fluids
around the body, which is explained in Diseases 4 through a series of
comparisons assimilating, among others, the head to a cupping vessel and
the body’s organs to interconnected bronze bowls. Thus, as Webster
argues, while there is no doubt that Hippocratic authors do not cast the
body as a type of machine (certainly not in the sense that Hellenistic
anatomists do at a later stage with direct reference to pneumatics and
mechanics), it is nonetheless legitimate to speak of an emerging relation-
ship between medicalized bodies and tools in terms of ‘iatrotechnics’ –
instead, that is, of iatromechanics. In his words, ‘bodies have not yet
become organized into machines, but this does not mean that tools and
technologies do not participate in the conceptual construction of the
body’ (p. 159).
The six chapters of the third and final section, entitled ‘Towards the
Mechanization of the Body’, move down chronologically to Aristotle and
the Peripatetic school, Hellenistic medicine and Galen and reach up to the
iatromechanist debates of the eighteenth century. The main focus here is
on a tradition of texts and authors, which appears to be more intensely
focused on figuring out and establishing the machine-like nature of the
human body. Human artefacts and technological devices, of varying
structures and complexity, are systematically invoked as precise analogues
for the design and function of specific bodily organs. At the same time,
mechanics becomes increasingly consolidated as an autonomous scientific
field, and its theoretical principles provide a firm point of reference for
understanding ‘scientifically’ the properties of the human body. As noted

above (p. 2–3), this process is crucially linked to developments in the field
of anatomy: as the body is dissected and becomes accessible to the naked
eye, its internal structure lends itself open to close scrutiny and, through it,
to an enhanced degree of assimilation with the world of machines.
Giouli Korobili, in ‘Aristotle on the Lung and the Bellows–Lungs
Analogy’, moves on to a period during which the comparative findings
of animal dissection become systematized and allow for a more solid
context of inference as to what lies inside the human body and what it
‘looks’ like. More specifically, Korobili discusses in detail the bellows–
lungs analogy, and the way in which this is established through a long list of
correspondences pertaining to shape and function in Aristotle’s De juven-
tute. This analogy – probably of Aristotle’s own inspiration, since it is not
attested before in any surviving Greek text – is believed by the philosopher
to be a better heuristic tool for understanding respiration when compared
to Empedocles’ earlier use of the clepsydra in the same context. Aristotle
not only finds the clepsydra – a technological device used for transferring
and perhaps measuring small amounts of liquid – to provide an imprecise
point of reference, he also proceeds to criticize Empedocles for failing to
speak of the purpose of respiration. The bellows–lungs analogy is thus
shown by Korobili to be firmly situated within Aristotle’s teleological
framework: the close similarities between a bodily organ and a smoothly
functioning product of human inventiveness and craft help essentially to
bring out the purpose-oriented structure of our bodies. In other words, the
firm background of empirical observation underlying Aristotle’s close
scrutiny of the body gives birth to an analogy which is not only more
carefully tailored to the function and properties of human artefacts but also
establishes, in rather straightforward terms, the principle that nature has
‘designed’ everything in wisdom.
In the next chapter, entitled ‘The Ill Effect of South Winds on the Joints
in the Human Body: Theophrastus, De ventis 56 and Pseudo-Aristotle,
Problemata 1.24’, Robert Mayhew continues to explore the conceptual
merging between human artefacts and the body in a Peripatetic context.
The effects of winds on humans and other living things, according to
Theophrastus, are explained through processes of moisture and dissol-
ution, density and solidification; to understand these pairs of opposites,
Theophrastus argues, we can compare them to what happens to inanimate
objects when they become moistened and loosened, for instance, beds,
chairs, tables and chests; these wooden objects are presented as constituent
parts of a house, in the same way that individual joints contribute to the
wholeness of the human body. Although the human body is not treated as

a machine or artefact per se, Theophrastus, according to Mayhew, invites us to
treat an artefact as a model or analogy of the human body on the basis of
commonly shared, fundamental material principles. This parts-of-
furniture–joints-in-the-body analogy is present also in [Arist.] Pr. 1.24, this
time being applied to furniture and sections of the house with moving parts,
for example, looms and doors. Although none of this evidence contributes to
a solid ‘mechanistic conception of the human body’, it represents a significant
step towards displaying how the same basic principles of natural philosophy
are at work in explaining the functioning of both inanimate objects and the
human body. Korobili’s and Mayhew’s chapters deal with texts that represent
the Peripatetic school of thought. While Aristotle and Theophrastus expand
on and refine the assimilation between the interior of the human body and
crafted objects – as this can already be traced in Hippocratic medicine – at the
same time they establish a more robust analogical reasoning in regard to the
function and purposiveness that is shared between certain bodily parts and
specific artificial objects and devices.
The next three chapters discuss the body–machine analogy in light of
the establishment of mechanics as a scientific field in Hellenistic Alexandria
and beyond. George Kazantzidis, in his chapter ‘The Beauty That Lies
Within: Anatomy, Mechanics and Thauma in Hellenistic Medicine’,
examines evidence from a period in which the body–machine analogy
acquires, for the first time in history, a concrete form. It is only with
Herophilus and Erasistratus, during the third century bc, that we move
directly to human dissection and vivisection, and, through them, to
unprecedented anatomical discoveries about various bodily organs.
Mechanics consolidates as a formal discipline in exactly the same period,
allowing doctors for the first time to draw concrete analogies to mechanical
devices as a guide for the investigation of the human body which had been
now laid open to the naked eye as never before. The emphasis in scholar-
ship, as Kazantzidis argues, has been usually placed on the ways in which
the close interaction between medicine and mechanics in Ptolemaic
Alexandria helped physicians to understand the body better; but, according
to the author, there is also an interesting aesthetic angle to this interaction.
Aristotle famously claims that the interior of the human body looks messy
and repulsive but can nonetheless yield immeasurable pleasures as
a spectacle so long as the specific purpose of each individual organ has
been fully comprehended. By taking as his main example Erasistratus, who
is consistently reported by Galen as ‘marvelling’ at the artful design of
nature, Kazantzidis argues that Hellenistic medicine nourishes a different
aesthetic model: the human body, understood now through a system of

close correspondences with mechanical devices, instils feelings of wonder
and admiration mainly by virtue of the fact that parts of it look almost
identical to products of human ingenuity and craft. Unlike Aristotle, who
attributes the artistry of nature to an ultimate authority of divine propor-
tions, the body–machine analogy thus puts human inventiveness and its
beauty centre stage. The close conceptual similarities between bodily
organs and mechanical devices in Hellenistic medicine have usually been
explored in relation to the question of ‘who influenced whom’. Some
scholars support the view that it was the doctors who influenced the
engineers; others maintain that the interaction worked the other way
around, and that it was primarily the field of mechanics which gave rise
to a corpus of practical knowledge that was subsequently applied to the
human body. A third approach speaks of a cross-disciplinary environment,
choosing to avoid the model of one-directional influence and speaking
instead of an ongoing dialogue between different epistemic fields which
constantly feed each other with new findings and insights.
While Kazantzidis explores how the dissection of the human body by
the Hellenistic anatomists, in combination with the contemporary con-
solidation of mechanics as a discipline, allows doctors, for the first time in
history, to draw concrete analogies with mechanical devices and, effect-
ively, to understand but also to appreciate aesthetically in new ways the
machine-like artistry and qualities of the human interior, Matteo
Valleriani, in a chapter entitled ‘The Mechanics of the Heart in
Antiquity’, tackles the intensely interactive, cross-disciplinary environment
of early Hellenistic Alexandria. Valleriani proposes that, instead of adopt-
ing a one-directional approach (which would lead us to assume either that
medicine draws from mechanical knowledge or, inversely, that mechanical
knowledge builds on medical discoveries), we should instead view the
unprecedented contact between medicine and mechanics in Alexandria
as the result of an ongoing, lively dialogue between the two epistemic fields.
He specifically sets out to contest the established belief according to which
Erasistratus’ conception of the heart was modelled on textual and perhaps
diagrammatic codification of the construction and function of the
hydraulic hand pump by the engineer Ctesibius. According to Valleriani,
our knowledge of the design and technicalities of Ctesibius’ water pump
derives from substantially later sources, such as Vitruvius and Hero of
Alexandria; this should make us, in the first place, cautious as to the degree
to which the pump described in these sources – on the basis of which
scholars argue for similarities with Erasistratus’ heart – corresponds with
Ctesibius’ original device. As Valleriani argues, a close look at the

archaeological evidence suggests that twin-cylinder force pumps made of
bronze, as described by Vitruvius and Hero, were quite limited in the
ancient world; what is more, there is no conclusive evidence that they were
in use at the time of Erasistratus. This leads Valleriani to the conclusion
that the details of Ctesibius’ water pump have been reconstructed by the
above-mentioned technical authors with a view to later developments in
the field of mechanics, which post-dated Ctesibius. This being so, it would
be risky, the author argues, to claim that Erasistratus modelled his descrip-
tion of the heart on mechanical knowledge.
In the next chapter, ‘The Mechanics of Galen’s Theory of Nutrition’,
Orly Lewis discusses a fundamental but generally overlooked aspect of the
digestive-nutritive process in Galen, namely the physical motions of the
bodily parts involved in the process, such as the oesophagus, the stomach
and the intestines. The established view in scholarship is that Galen
explains digestion by invoking the natural faculties which reside in each
one of the different organs through which food passes, and not as
a mechanical process. As Lewis argues, while it is true that Galen focuses
his attention on the organs’ teleologically framed capacity to exercise
attraction, retention, assimilation and expulsion, the ‘mechanics’ of the
motions involved should not be underestimated either. In this context,
according to Lewis, teleology and mechanical explanations go together; the
meeting point, so to speak, is the body’s material structure. By focusing on
the anatomical path which the nutritive matter takes once it enters the
mouth, Lewis illustrates how the purposiveness of each bodily organ – the
function, in other words, that has been assigned to it by nature – material-
izes through motion and how this motion can in fact be explained in
mechanical terms. Galen’s theory of nutrition turns out, in this respect, to
be a fitting case study exemplifying how a mechanistic perspective can go
hand in hand with a teleological one.
In the volume’s final chapter, ‘Iatromechanism and Antiquarianism in
Morgagni’s Studies on Celsus, 1720–1761’, Marquis Berrey discusses the
reception of Celsus’ De medicina in the work of Giovanni Battista
Morgagni (1682–1771), a professor of anatomy in Padua. Berrey explores
in detail Morgagni’s interpretation of (the lacuna in) Celsus 4.27, a section
on bladder stones, and 7.26, a part of the text which provides advice on
how bladder stones should be removed surgically. According to the author,
Morgagni advanced a learned and iatromechanist reading of Celsus;
Morgagni’s understanding of the Roman author’s text, by means of
remaining attentive to assigning the proximate causes of disease to the
dynamic processes of force, pressure and motion within bodily anatomy,

corresponds to the increasing importance of iatromechanist theories in the
Age of Enlightenment and marks a step away from a Galenist vision of
humoralism.
The case studies discussed in this volume are not meant to offer an
exhaustive discussion of body–machine-intersections in antiquity. Rather,
we are interested in presenting a collection of essays that trace the archae-
ology of what, during the early modern period, acquires the concrete
conceptual form of a mechanized human body. And we do this by collect-
ing significant evidence from the ancient world, which is focused on the
close interaction of the human body with technological artefacts and
machines, in a wide range of periods and genres, and a wide range of fields
and disciplines: literary and scientific, textual, archaeological and artistic.
In doing so, we hope that this volume will facilitate future dialogue
between scholars workings on different areas, from classics, history and
archaeology to history of science, philosophy and technology.
RE F E RE N CE S
Aho, J. and Aho, K. 2008. Body Matters: A Phenomenology of Sickness, Disease, and
Illness (Lanham, MD).
Althoff, J. 1997. ‘Vom Schicksal einer Metapher: Die Erde als Organismus in
Senecas Naturales Quaestiones’, in K. Döring, B. Herzhoff, and G. Wöhrle,
eds., Antike Naturwissenschaft und ihre Rezeption, vol. 7 (Trier), 95–110.
Anderson, M., D. Cairns and M. Sprevak, eds. 2018. Distributed Cognition in
Classical Antiquity (Edinburgh).
Bensaude-Vincent, B. and W. R. Newman. 2007. ‘Introduction: The Artificial
and the Natural: State of the Problem’, in B. Bensaude and W. R. Newman,
eds., The Artificial and the Natural: An Evolving Polarity (Cambridge,
MA), 1–19.
Berryman, S. 2009. The Mechanical Hypothesis in Ancient Greek Natural Philosophy
(Cambridge).
2020. ‘Ancient Greek Mechanics and the Mechanical Hypothesis’, in L. Taub
(ed.), The Cambridge Companion to Ancient Greek and Roman Science
(Cambridge), 229–47.
Bianchi, E., S. Brill and B. Holmes, eds. 2019. Antiquities Beyond Humanism
(Oxford).
Canevaro, L. 2019. ‘Materiality and Classics: (Re-)Turning to the Material’,
Journal of Hellenic Studies, 139: 222–32.
Chesi, G. M. and F. Spiegel, eds. 2020. Classical Literature and Posthumanism
(London).
Cohen, W. A. 2003. ‘Deep Skin’, in J. J. Cohen and G. Weiss, eds., Thinking the
Limits of the Body (Albany, NY), 63–82.

De Groot, J. 2014. Aristotle’s Empiricism: Experience and Mechanics in the Fourth
Century bc (Las Vegas).
Derrida, J. 1974. ‘White Mythology: Metaphor in the Text of Philosophy’, New
Literary History, 6: 5–74.
Gerolemou. M. 2022. Technical Automation in Classical Antiquity (London).
Gentner, D., B. Bowdie, P. Wolff, and C. Boronat. 2001. ‘Metaphor Is Like
Analogy’ in D. Gentner, K. J. Holyoak, and B. N. Kokinov, eds., The
Analogical Mind: Perspectives from Cognitive Science (Cambridge, MA),
99–253.
Grmek, M. D. 1972. ‘A Survey of the Mechanical Interpretations of Life from the
Greek Atomists to the Followers of Descartes’ in A. D. Breck and
W. Yourgrau, eds., Biology, History, and Natural Philosophy (Boston), 181–95.
Gundert, B. 1992. ‘Parts and Their Roles in Hippocratic Medicine’, Isis, 83:
453–65.
Holmes, B. 2010. The Symptom and the Subject: The Emergence of the Physical Body
in Ancient Greece (Princeton, NJ).
2018. ‘Body’, in P. E. Pormann, ed., The Cambridge Companion to Hippocrates
Kang, M. 2011. Sublime Dreams of Living Machines: The Automaton in the
European Imagination (Cambridge, MA).
King, H. 2013. ‘Female Fluids in the Hippocratic Corpus: How Solid Was the
Humoral Body?’, in P. Horden and E. Hsu, eds., The Body in Balance:
Humoral Medicines in Practice (New York), 25–52.
Kirby, J. T. 1997. ‘Aristotle on Metaphor’, American Journal of Philology, 118:
517–54.
Lloyd, G. E. R. 1966. Polarity and Analogy: Two Types of Argumentation in Early
Greek Thought (Cambridge).
Lonie, I. M. 1981. ‘Hippocrates the Iatromechanist’, Medical History, 25: 113–50.
Lo Presti, R. 2008. ‘La notion d’automaton dans les textes médicaux (Hippocrate
et Galien) et la Physique d’Aristote: hasard, spontanéité de la nature, et
téléologie du comme si’, in K. D. Fischer, ed., 30 Jahre Arbeitskreis Alte
Medizin in Mainz: Beiträge der Tagung 2010, special issue of Les Etudes
Classiques, 80: 25–54.
Mayor, A. 2018. Gods and Robots: Myths, Machines, and Ancient Dreams of
Technology (Princeton, NJ).
Morton, T. 2010. ‘Thinking Ecology: The Mesh, the Strange Stranger, and the
Beautiful Soul’, Collapse, 6: 265–93.
Ricoeur, P. 1978. The Rule of Metaphor: The Creation of Meaning in Language
(London).
Riskin, J. 2016. The Restless Clock: A History of the Centuries-Long Argument over
What Makes Living Things Tick (Chicago).
Schiefsky, M. J. 2007. ‘Art and Nature in Ancient Mechanics’, in B. Bensaude and
W. R. Newman, eds., The Artificial and the Natural: An Evolving Polarity
(Cambridge, MA), 67–108.

Sobchack, V. 2006. ‘A Leg to Stand On: Prosthetics, Metaphor, and Materiality’,
in M. Smith and J. Morra, eds., The Prosthetic Impulse: From a Posthuman
Presence to a Biocultural Future (Cambridge, MA), 17–41.
Taub, L. 2012. ‘Physiological Analogies and Metaphors in Explanations of the
Earth and the Cosmos’, in M. Horstmanshoff, H. King and C. Zittel, eds.,
Blood, Sweat and Tears: The Changing Concepts of Physiology from Antiquity
into Early Modern Europe (Leiden), 41–63.
Vaccari, A. 2012. ‘Dissolving Nature: How Descartes Made Us Posthuman’,
Techné: Research in Philosophy and Technology, 16: 138–86.
Vernant, J.-P. 2006. Myth and Thought among the Greeks, transl. J. Lloyd and
J. Fort (New York).
Von Staden, H. 1989. Herophilus: The Art of Medicine in Early Alexandria
(Cambridge).
1996. ‘Body and Machine: Interactions between Medicine, Mechanics, and
Philosophy in Early Alexandria’, in J. Walsh and T. F. Reese, eds.,
Alexandria and Alexandrianism (Malibu, CA), 85–106.
1997. ‘Teleology and Mechanism: Aristotelian Biology and Early Hellenistic
Anatomy’, in W. Kullmann and S. Föllinger, eds., Aristotelische Biologie:
Intentionen, Methoden, Ergebnisse (Stuttgart), 183–208.
1998. ‘Andréas de Caryste et Philon de Byzance: médecine et mécanique à
Alexandrie’, in G. Argoud and J.-Y. Guillaumin, eds., Sciences exactes et
sciences appliquées à Alexandrie (Saint-Étienne), 147–72.
2007. ‘Physis and Technê in Greek Medicine’, in B. Bensaude and
Wee, J. Z., ed. 2017. The Comparable Body: Analogy and Metaphor in Ancient
Mesopotamian, Egyptian and Greco-Roman Medicine (Leiden).
Webster, C. (forthcoming). Organa and the Organism (Chicago).
White, R. 2010. Talking About God: The Concept of Analogy and the Problem of
Religious Language (Farnham).
Wolfe, C. T. 2016. Materialism: A Historico-Philosophical Introduction
(Dordrecht).
2017. ‘Varieties of Vital Materialism’, in J. Zammito and S. Ellenzweig, eds.,
The New Politics of Materialism: History, Philosophy, Science (London), 44–65.

part i
Blended Bodies

chapter 1
More Than a Thing

Figuring Hybridity in Archaic Poetry and Art
Deborah Steiner
Introduction
In a discussion of the place of tools (organa) in the larger business of
household management, Aristotle distinguishes between implements
that are lifeless (apsucha) and those that are endowed with psuchê and
goes on to cite slaves, here styled ‘living pieces of property’, as exemplary
of the latter category (Pol. 1253b27–32). As he then observes, there is only
one condition in which we could imagine masters not needing slaves; this
would be that each instrument ‘could do its own work, being so com-
manded or by intelligent anticipation, like they say of the statues of
Daedalus or the tripods of Hephaestus, of which the poet relates that
“of their own motion they entered the divine agôn”,1 as if a shuttle should
weave of itself and a plectrum should do its own lyre playing’. A fourth-
century reader would readily identify the source of Aristotle’s citation
and the larger episode framing it: the well-known scene in Iliad 18, where
Thetis, come to petition Hephaestus to forge a new set of armour for
Achilles, finds the craftsman deity engaged in fashioning a set of self-
moving tripods (373–9, cited below, p. 22). Where Aristotle focuses on
the utility of such objects, their ability to carry out their master’s will of
their own accord, Homer’s account dwells more on the artistry and
properties of these remarkable vessels. As my close reading of this
Iliadic passage and of a series of other descriptions of automata and
variously ‘ensouled’ artefacts in the archaic Greek sources aims to dem-
onstrate, for early poets, artists and their audiences, the fascination of
these objects seems principally to derive from their composite or hybrid
character, their capacity to combine disparate and contrasting spheres
and so to breach seemingly inviolable boundaries, uniting in one the
1
Translations are my own unless otherwise indicated.
19

20 deborah steiner
sentient and inert, the divine and human, the organic and factured, the
original and its counterfeit.2
A cursory review of Homeric and Hesiodic poetry alone yields a large
and diverse range of vivified objects, their multiplicity and heterogeneity
proof in and of themselves of the enduring hold of these products – some
made by the gods, others by supremely skilled mortal craftsmen – on the
Greek imagination. Shortly after the description of the tripods that
Hephaestus is accessorizing, Homer introduces not only the golden serving
girls whom the god has fashioned to help him in his labours, metallic
objects styled ‘like to living maidens’ (ζωῇσι νεήνισιν εἰοικυῖαι, Il. 18.418)
who can speak, think and move,3 but also the metal-forged shield of
Achilles, whose depictions of divinities, mortals, plants, animals and the
natural world prove all but indistinguishable from their living counter-
parts. Arriving at the palace of the Phaeacian king Alcinous in Odyssey 7,
Odysseus encounters a further set of Hephaestus’ animated metal goods:
keeping guard on the threshold are gold-and-silver watchdogs, whose
‘immortal and unageing’ (7.94) character (the doublet is elsewhere exclu-
sively reserved for the eternally young and ever-living gods) stands in
poignant contrast to Argos, the faithful, flea-bitten hound whose death
will coincide with the hero’s belated return to his home. Scheria, home of
the Phaeacians, is also the land where dockyard-built ships propel them-
selves with self-rowing oars and where golden youths who fulfil the role of
lamps are at once living beings and static images mounted on their plinths.
Hesiod twice details another of Hephaestus’ more infamous creations
(made with the aid of other gods), the vivified Pandora;4 moulded from
both earth and water in the manner of a pot which is then stocked with
a variety of properties, she also closely resembles both a cult image and
a bride, decked out in finely woven textiles and ornaments.5 A third set of
early hexameter compositions, the Homeric Hymns, includes more vivified
objects to add to the list. Among these belongs the lyre in the Homeric
Hymn to Hermes, where the god celebrated by the song and himself skilled
in bricolage devises an intricate medley of once-living creatures (a tortoise
supplies the ‘shell’ or instrument’s body while sheep contribute the sinews
2
For an exploration of a number of these questions with regard to the ontology of statues in archaic
and classical Greek imagination, see Steiner 2001.
3
See this volume, Chapter 2, for these.
4
Ruffell in Chapter 3, this volume, cites both Alcinous’ dogs and Pandora in her discussion.
5
Particularly apposite to my topic is the diadem forged by Hephaestus with which the figure is
crowned and whose decoration consists of wild beasts ‘like to living creatures that speak’ (ζωοῖσιν
ἐοικότα φωνήεσσιν, Th. 584). For a discussion of Pandora, see Steiner 2013.

More Than a Thing 21
used for the strings) and man-made objects and then endows this now
seemingly inanimate assemblage with the capacity to converse, sing, dance
and engage in amorous dalliance with its player-boyfriend.6
Turning to the early lyric corpus, no poet offers a richer compendium of
automata, figural hybrids and crafted bodies than Pindar, whose elaborate
and rococo conceits often suggest parallels between his poetic creations –
compositions that grow, travel, speak and dazzle – and works of superlative
artistry that transcend the limits of their material production. Among these
belong the aulos fashioned (or ‘interwoven’, διαπλέξαισ’, in the poet’s
words) by Athena in Pythian 12 that channels the Gorgons’ shrieks and
the hissing of their serpentine locks through a tube of hammered bronze
and reeds (8–10, 25); the sequence of mnemonic and partially animated
products that enunciate the victor’s glory in Olympian 7,7 a work that also
makes a place for the statues that line the streets of Rhodes, ‘like to beings
that live and move’ (52), which would put an audience in mind of the
animated statues attributed to the wizardly blacksmiths, the Rhodian
Telchines; and a musical message stick that doubles for a chorus trainer
at Olympian 6.91–2. Beyond his epinicians, Pindar populates other pieces
with still more whimsical creations, few more fanciful than the columns
that rise up from the earth of their own volition, wearing sandals of
adamant, in the poet’s so-called Hymn to Zeus (fr. 33d.5–9).
Confronted with this extensive and varied material, I have selected just
four objects from the archaic sources, two from Homer and two from
Pindar, each of which offers its own amalgam of organic and fabricated
elements; each of them, while sharing certain features and properties, also
poses different questions of their audiences. In each discussion, I aim both
to offer detailed readings that explore the riddling quality of these bound-
ary-blurring embodied artefacts and to draw attention to the dense rela-
tions between the figural hybrids and the narrative frames in which they
appear, whose themes and motifs, I propose, they regularly emblematize
and comment on; on several occasions too, near-contemporary visual
accounts variously reinforce and complicate the poetic accounts. My
discussion, moreover, serves as ‘preliminary’ to this volume on two princi-
pal counts: first, because the material it treats antedates the sources on
which subsequent chapters largely focus;8 and second, because in archaic
6
The chief passages are HHMer. 38–54 and 475–88.
7
For rich discussion of what she calls the ode’s ‘thingliness’, see Kurke 2015.
8
As already noted, Chapters 2 and 3 in this volume do include some of the Homeric material treated
here, insofar as they also discuss the cauldrons, bellows and serving girls, albeit with a focus very
different from mine.

22 deborah steiner
times, the heuristic device used to understand how bodies function is not
the machine, but the products of human craft that can merge with,
supplement and enhance the living organism.
In Hephaestus’ Forge: The Twenty Tripods and Bellows

On her arrival at Hephaestus’ Olympian home in Iliad 18, Thetis comes
upon the craftsman adding the finishing touches to the first among the
several metallic and other crafted products that the episode includes
(372–80):
τὸν δ’ εὗρ’ ἱδρώοντα ἑλισσόμενον περὶ φύσας
σπεύδοντα· τρίποδας γὰρ ἐείκοσι πάντας ἔτευχεν
ἑστάμεναι περὶ τοῖχον ἐϋσταθέος μεγάροιο,
χρύσεα δέ σφ’ ὑπὸ κύκλα ἑκάστῳ πυθμένι θῆκεν,
ὄφρά οἱ αὐτόματοι θεῖον δυσαίατ’ ἀγῶνα
ἠδ’ αὖτις πρὸς δῶμα νεοίατο, θαῦμα ἰδέσθαι.
οἳ δ’ ἤτοι τόσσον μὲν ἔχον τέλος, οὔατα δ’ οὔ πω
δαιδάλεα προσέκειτο· τά ῥ’ ἤρτυε, κόπτε δὲ δεσμούς.
ὄφρ’ ὅ γε ταῦτ’ ἐπονεῖτο ἰδυίῃσι πραπίδεσσι
[Thetis] found him sweating and circling around the bellows as he made
haste. He was fashioning twenty tripods which were to stand along the wall
of his strong-founded palace, and he had set golden wheels beneath the base
of each one in order that of their own motion they could enter into the
immortal agôn, and come back to his house again, a wonder to behold.
These were so far finished, but he had not yet added on the cunningly
wrought ear handles. He was arranging these, and he was beating out the
chains. So he was expending labour with his knowing diaphragm.
In detailing the powers and properties of these remarkable cauldrons –
which sit on top of the twenty tripods – I propose situating them within
the larger teleological sequence of objects that anticipate what the Homeric
poet presents as the most remarkable and capstone manifestation of the
divinity’s artistry and manufacturing skill: the next-to-last metallic tableau
on the shield of Achilles, where Hephaestus depicts a chorus of youths and
maidens, decked out in gleaming accessories, performing an intricately
choreographed dance (18.590–605). As I have argued elsewhere, not only
does the narrator signal the superlative quality of this penultimate metal-
forged representation,9 but here he imagines an idealized choral ensemble
that explicitly operates with all the precision of a mechanistic and
9
Steiner 2021, 25–63.

manufactured device: comparing the dancers’ spins to the unceasing, self-
generating turns of the potter’s wheel (600–601), the poet invites us to see
the choristers as possessing all the dynamism and unvarying, unending
quality of this essential tool of the coroplast’s skill.10 In his fabrication of
the tripods, Hephaestus, I suggest, already undertakes the creation of an
embodied, incipiently choral object whose several elements rehearse the
essence of the ‘objectified’ bodies performing on the shield.
Postponing discussion of the tripods’ projected position (‘along the
wall’) and the wheel-propelled mobility that is the prime marker of their
vivification, I begin with the handles and chains. The first, styled ‘richly
worked’ and perhaps reminiscent of the ornately decorated and neatly
rounded handles on still-extant bronze tripod cauldrons, already give to
these artefacts the anthropomorphism that becomes more pronounced
with the golden serving girls. Just like pots endowed by Greek potters
and painters with human features, whether outsized eyes, a projecting foot
or phallus by way of a base11 or a nipple (the so-called mastoi jugs), so the
tripods’ οὔατα (‘ears’) span the realms of object and body, inviting
a viewer/user to engage with and deploy the vessels as though they were
interacting with sentient beings, even those endowed with the faculty of
hearing and able to respond to the directives issued by their master and
maker.
The second features to be fastened to the objects, the ‘bands’ that
Hephaestus hammers out, contribute to the tripods’ high visual appeal
and promote their likeness to living bodies in a different respect. Not so
much chains that would allow the transport of the vessels (self-moving
objects would have no need of these, and no other Homeric cauldrons
possess such additions), these instead aim at the products’ enhanced
embellishment.12 Exactly in this manner, the ropes formerly used to
move large-scale storage jars from place to place would evolve into the
decorative incised, stamped and stippled bands that encircled relief pithoi,
10
Note too Frontisi-Ducroux 2002, 482, who comments on the reasons why a choral performance
should figure as the supreme expression of manufacturing virtuosity and the artisan’s capacity to
blur representation with the thing itself: ‘la danse est un modèle de l’art total, à la fois visuel,
figurative, cinétique, vivant et musical’. As commentators have observed, the way in which Homer
envisages the construction of the shield recalls the making of the vase, built up in successive coils. We
might also note that Daedalus, mentioned uniquely in the poem at the outset of this passage, was
also credited with discovering the potter’s wheel.
11
The most notorious example of this design is the so-called Bomford cup, an Attic red-figure kylix in
the manner of the Lysippides Painter. Oxford, Ashmolean Museum 1974.344.
12
Edwards 1991 ad loc. interprets these fastenings as ‘rivets’ to fasten the handles to the tripod body;
however, the term is never used in that sense elsewhere, and when the identical phrase recurs at Od.
8.274, δεσμοί are unmistakably ‘chains’.

24 deborah steiner
another set of valuable, high-status objects featuring a wealth of ornamen-
tation that date from the early archaic age on. Insofar as Hephaestus’
beaten-out bands are supplementary and non-functional, they invite the
audience again to envisage the embodied quality of his tripods, whose
likeness to accessorized, bedecked performers I demonstrate below (pp.
000–000).
Beyond these features, much of the remainder of the description is given
over to the wheels that permit the tripods, first positioned in a linear
formation against the wall, to perform their circuit from Hephaestus’
home into the immortals’ agôn. Following an observation made by
Nassos Papalexandrou,13 who notes the spatial meaning that agôn can
carry and proposes that we imagine the objects travelling to a location set
aside for the staging of a spectacle, I would suggest that Homer intends us
to read the tripods as proto-performers. On two occasions in the Odyssey,
agôn refers to the venue where choreia (choral song-dance) takes place, the
first when the Phaeacian stewards prepare a dancing area within an agôn,
where Demodocus will sing while around him youths ‘beat the wonderful
dance/dance floor with their feet’ (8.258–64). This performance done, the
two premier Phaeacian dancers then entertain the company with their
skilled display while the other chorus members ‘standing about the agôn
stamped out the time’ (379–80). Another archaic usage supports
Papalexandrou’s reading: in Alcman’s third Partheneion, the choral ego
declares itself ‘eager to hear the voice of girls singing a beautiful melody . . .
and bids me to go to the agôn, where I shall rapidly shake my yellow hair . . .
soft feet’ (4–10).
With this paradigm in mind, the features and actions picked out for
special mention by the Homeric poet make good sense. Quitting their
initial alignment and moving into a circle, the tripods exactly replicate the
two chief formations attested for Greek choruses of the archaic and classical
period, first processing, then ring-dancing, even as Hephaestus, the one
who lines them up, assumes the role of chorêgos, the chorus leader who
institutes and marshals his ensemble, stations them in their chorus line
and, on occasion, decks out the typically richly dressed, adorned and
visually compelling singer-dancers. In just this manner, the Libyan
Antaios, reenacting a courtship dance first staged by the mythical king
Danaus, which conflates a choral performance with a running race, ‘orna-
ments’ (the verb is κοσμέω) his daughter, both a participant and the prize
13
Papalexandrou 2005, 32 and n. 73.

in the event, prior to positioning her in the choral ensemble and directing
the choristers to begin their dance (Pind. Pyth. 9.117–19).
The golden wheels that the craftsman fastens on can no less readily be
matched up with one of the prime features that archaic and classical poets
repeatedly privilege in their descriptions of choral dancers: their sandals,
the special footwear that, where divinities are concerned, may take
metallic or other fantastical form. Inasmuch as these attachments
would lend extra sparkle to the tripods as they moved, they would be
the source of the same ‘twinkling’ and ‘flashing’ that emanates from
choral dancers’ rapid and interchanging footwork in hexameter poetry,
an incandescence rendered by the formulaic phrase reserved for it that
occurs in several of the poetic accounts, μαρμαρυγαί . . . ποδῶν (e.g.,
HHAp. 203, Od. 8.265). Indeed, absent dancing shoes, archaic epic and
lyric poets all but unfailingly include a mention of the choral dancers’ feet
in visualizations of choral performances, those body parts written into the
tripods’ very name (τρίποδες) that stands at the outset of the Homeric
passage. Read this way, Hephaestus’ choice to fashion the triple-footed,
self-moving tripod cauldrons could even be viewed as a compensatory
device: hobbled and with his own movements impaired, he constructs
vessels whose properties supply and exceed what this defective body
lacks.14
A glance at the visual corpus suggests that Homer was not alone among
our early sources in combining tripod cauldrons with a dancing chorus.
The same seemingly unlikely combination occurs on the so-called Eleusis
amphora, an oversized vase by the Polyphemus Painter dated to ca. 670 to
650 (Figures 1.1a and 1.1b). On the pot’s central field, two Gorgons,
presented frontally as they move away from the body of their just-
decapitated sister, who appears as though floating horizontally in mid-
air, pursue Perseus as he flees around the damaged curve of the vase. Two
features in the representation of the monstrous duo are particularly arrest-
ing and unusual: first, as several discussions have noted, the sisters are
shown less as the runners of later visual renditions of the episode than as
a dancing pair, the position of their feet, their coordinated motions and
style of dress all broadly conforming to the visual typology used for maiden
dancers of the period;15 and second, their heads are imagined on the model
of tripod cauldrons. Not only do the beasts sprouting from the head and
14
For detailed discussion of Hephaestus’ impairment, see Chapter 2, this volume.
15
See Langdon 2008, 6–7, with her fig. 0.4, and 112–13 for other examples of Gorgons in the early visual
record depicted as choral dancers. Topper 2007 and 2010 more broadly demonstrate how texts and
images imagine the sisters as members of a parthenaic ensemble.

26 deborah steiner
(a) (b)
Figure 1.1a–b Eleusis amphora by the Polyphemus Painter, dated to ca. 670–650.
Eleusis Archaeological Museum 2630
shoulders of the Gorgon on the right curve around as though forming one
half of the handles on the bronze vessels contemporary with the
amphora,16 but the structural composition of the monsters themselves
exactly reproduces the design of cauldrons in the artistic record in more
wholesale manner. To cite one example, a Protocorinthian lekythos-
oinochoe of ca. 700 in New York includes an Oriental-style cauldron
that, like the Gorgons, is tripartite in form (Figure 1.2); while its topmost
portion, complete with snake-like protomes and zigzags at the rim much
like the ‘growth’ topping the Eleusis amphora Gorgons, matches the two
sisters’ heads, the middle unit presents a bulbous diamond equivalent to
the Gorgons’ torsos; the vessel’s triangular base in turn forms the coun-
terpart to their skirts.
Bypassing the question of why the Polyphemus Painter chose to
depict the monsters in this fashion and only noting the multiple
relations between tripods and choral dancers in the poetic, visual
and archaeological records of early Greece,17 I would instead focus
on the ways in which the Eleusis amphora both realizes what is
already latent in the Iliadic representation of Hephaestus’ first set of
products – that these vessels can be envisaged as animated partici-
pants in a performance and members of a choral troupe – and adds
to the epic description an additional level of complication. Like the
16 17
See Payne 1971, 214. For these, see Steiner 2014 and Steiner 2021, 78–95.

Figure 1.2 A lekythos-oinochoe of ca. 700. New York, Metropolitan Museum of Art
23.160.18
Homeric poet before him, the artist has effected a combination of

living bodies and metal-forged goods, here affixing one to the other
with no discernible break, and like his predecessor too, he challenges
the viewer to discern the ontological status of the hybrids he imagines
and to identify the dividing line between the different orders of the
fabricated and the corporeal, the static and mobile, metallic (and
ceramic insofar as the tripods are rendered in clay) and organic. But
exclusive to the visual account is the presence of a second manufac-
tured object, whose shape and function the Gorgons’ own morph-
ology recalls, namely, the painted pot on which the monsters appear.
By virtue of showing vivified vessels on the amphora that replicates
some of the features of the painted tripods and the monsters whom
they top, and which is itself a highly worked and costly receptacle
designed as much for display and dedication as for household use, the
artist/potter prompts us to see in his creation something more than an
inert artefact.
Before leaving the animated products in Hephaestus’ forge behind, one
further set of objects equipping the divine foundry, tools essential to the
metallurgist’s craft, deserves a mention, not least because the poet lingers

28 deborah steiner
on them, granting them powers and attributes that make evident their
more than quotidian character (468–73):
βῆ δ’ ἐπὶ φύσας·
τὰς δ’ ἐς πῦρ ἔτρεψε κέλευσέ τε ἐργάζεσθαι.
φῦσαι δ’ ἐν χοάνοισιν ἐείκοσι πᾶσαι ἐφύσων
παντοίην εὔπρηστον ἀϋτμὴν ἐξανιεῖσαι,
ἄλλοτε μὲν σπεύδοντι παρέμμεναι, ἄλλοτε δ’ αὖτε,
ὅππως Ἥφαιστός τ’ ἐθέλοι καὶ ἔργον ἄνοιτο.
And then [Hephaestus] went to the bellows, which he turned towards the
fire and gave them orders to work. And they, all twenty of them, blew on the
hollow melting places, sending forth their well-blowing breath this way and
that, as he hurried to be in one place, and then in another, wherever
Hephaestus might wish them to blow, and the work might be completed.
This depiction of these implements neatly recapitulates properties of both the
tripods and the golden girls introduced in the preceding passages, establishing
their correspondence with the earlier automata. Like the tripods, the bellows
number twenty and seem self-moving too; there is no clear demarcation of
subject and agent here, and once the god has ‘turned’ the tools in the direction
he desires, they seem to perform their tasks without his intervention, following
their master in his passage from place to place. Featured here, as in the account
of the metal maidens, are both the bellows’ role as workers (ἐργάζεσθαι recalls
ἔργα, used at 420 of the tasks that the serving maids are charged with
fulfilling) and their possession of an attribute that distinguishes living beings
from the inanimate, supplying the stuff of life and permitting the production
of voice. ‘Breath’, here underscored by the addition of the adjective
εὔπρηστον modifying ἀϋτμή (the prefix gives the term an almost aesthetic
cast), is the vital force that Achilles retains until the moment of his death (so Il.
9.609–10), while the verb used of the bellows’ exhalations, ἐξανίημι, offers
a compound of the term repeatedly chosen by hexameter poets for the
emission of words, songs and the music sounded on an instrument.
Reading these bellows alongside the other ‘para-performative’ objects featured
in the larger episode and noting that they issue their well-regulated breath and
move together with their master, thereby attaining the vocal and kinetic
synchronicity that choral singer-dancers should ideally achieve under the
leadership of their chorêgos, we might see in the tools a fresh iteration of
a moving and now vocalizing chorus and a further anticipation of the
penultimate tableau on Achilles’ shield.
The latent musicality of the bellows, and even their connection to the
dance in their back-and-forth movements, becomes visible in the later

Figure 1.3 A red-figure krater by the Kleophrades Painter of ca. 500–490.

Cambridge, MA, Arthur M. Sackler Museum, Harvard University Art Museums
1960.236. Attic Red-Figure Vase Painters2 185.31.
visual repertoire where, on a red-figure krater by the Kleophrades Painter

of ca. 500 to 490 (Figure 1.3),18 an aulos-playing satyr in the retinue
accompanying Hephaestus back to Olympus carries a supersized set of
leather bellows over his left shoulder as he supplies the music to which the
two satyrs behind him dance. There is no mistaking the visual homology
between the pipes and the metallurgical device – the painter takes care to
show the bellows’ tube which, hanging down in front of the satyr’s left leg,
mirrors the shape of the double-reeded instrument above (as well, of
course, as inverting the satyr’s phallus, one upward pointing and erect,
the other downward and flaccid) – but apparent here too is a verbal as well
as visual pun. As François Lissarrague points out,19 the term aulos can
describe both the pipes and the bellows’ tube, investing the implement
with the musicianship and sonic range assigned to the instrument.
Their vocal powers and musicianship aside, the breath with which these
bellows are endowed grants the objects a more particularized corporeal
character and an animating dimension that together reinforce the symbi-
osis between the tools and their master. Anticipating Aristotle, who cri-
tiques Empedocles’ comparison of the lungs to the clepsydra and proposes
that the bellows supply a better analogy for the workings of the organs
18 19
For discussion, see Lissarrague 1990, 41–4. For the pun, see Lissarrague 1990, 43–4.

30 deborah steiner
instead (De respiratione 473a15–474a34),20 here Homer prompts us to
understand Hephaestus’ utensils as a supplementary and externalized set
of lungs, a reduplication of one among the god’s own signature anatomical
features. As Frontisi-Ducroux has observed,21 Hephaestus’ pulmonary
equipment, which the poet terms his prapides,22 occupies an over-
determined role in the larger episode in the forge, twice mentioned in
the sequence of scenes in the formulaic phrase ἰδυίῃσι πραπίδεσσι, ‘know-
ing wits’ (18.380 and 482) and on each occasion in the context of the god’s
fabrication of one of his wondrous metallic objects, first the tripods and
then the shield. Sharing this same critical breathing equipment with their
master, and perhaps his organs’ ‘knowingness’ too,23 the bellows, whose
very shape recalls that of a pair of lungs and whose function, as several late
archaic visual depictions make clear, is to keep the furnace fire hot through
their emission of breath,24 can double for Hephaestus, furnishing him with
the means of multiplying his efforts, so, as the closing phrase of the passage
cited above (p. 28) indicates, enabling him to complete the task at hand.
The Pindaric Pelops

Third in this quartet of hybrids, the figure of Pelops in Pindar’s first
Olympian re-sounds several of the issues rehearsed in my previous
examples, among them both the challenge of disentangling the
manufactured from the corporeal and the supplementary or pros-
thetic role of crafted goods. But Pelops also stands distinct from the
several objects considered so far: in place of the fabricated goods
that incorporate the animating and ‘embodying’ properties featured
in Homeric poetry, Pindar’s composite is very much a living being
who happens to be equipped with an artificial and crafted element,
an inanimate device that renders whole what would otherwise be
anatomically incomplete.
A familiar character in Greek myth, Pelops enters the Pindaric compos-
ition in a passage that would immediately alert an audience to the poet’s
revisionary telling of what seems an already at least partially codified story.
20
This passage is analysed in Chapter 7, this volume. 21 Frontisi-Ducroux 2002.
22
See Onians 1988, 27–115, for detailed discussion of the πραπίδες which, in his view, describe the
thoracic area in Homeric poetry.
23
For Frontisi-Ducroux 2002, the epithet describes the lungs’ visionary powers, and she imagines
them as the site of Hephaestus’ creative powers.
24
For depictions of bellows in other foundry scenes, these of a more realistic kind, see the cup of the
Foundry Painter (Figures 1.5 and 1.6) and a black-figure oinochoe with another foundry scene of ca.
500, London, British Museum B507.

According to the standard sources,25 Pelops’ father Tantalus, the recipient
of the signal benefaction of divine hospitality, returns the favour by
inviting the gods to a feast at his home, serving up by way of savoury
course his dismembered son; none of the gods tastes the dish except for
Demeter who, distraught with grief at the loss of Persephone, takes a bite
from the shoulder. The gods then place the remaining body parts in
a cauldron, boil them up and reconstitute Pelops, replacing the missing
feature with an ivory substitute. But Pindar’s introductory lines, posi-
tioned immediately after the proem’s end, already feature several depar-
tures from the expected scenario, whose familiar objects and motifs they
recast in novel but still recognizable form (25–7):
τοῦ μεγασθενὴς ἐράσσατο Γαιάοχος
Ποσειδάν, ἐπεί νιν καθαροῦ λέβη-
τος ἔξελε Κλωθώ,
ἐλέφαντι φαίδιμον ὦμον κεκαδμένον.
[Pelops] with whom the great-hearted Earth-Holder Poseidon fell in love
when/since Klotho drew him out of the pure cauldron, excelling as to his
shoulder shining with ivory.
In this altered narrative, Pelops already possesses the ivory shoulder at his
birth (hence the presence of one of the Fates playing the role of birth
attendant, and the ‘pure’ cauldron, there for the bathing of the newborn)
and additionally becomes the object of Poseidon’s erotic passion (possibly
on account of the gleaming body part).
Leaving aside the vexed question that turns on the syntax of line 25 and
the problem of determining whether Poseidon’s love was already integral
to the story or a Pindaric novelty,26 I would highlight instead the way in
which, as Thomas Hubbard acutely argues,27 Pelops as figured here exactly
tropes the poet’s narrative, itself an amalgam of disparate parts and whose
own meld of ‘original’ and new and supplementary material finds its
analogue in the hero’s gleaming body. Like the newly restored individual
who emerges from the cauldron, the tale is itself a composite that fastens
crafted and ‘foreign’ bodies onto what existed before, and in so doing
engages in a complex procedure that embellishes even as it claims to strip
away what it configures as lying accretions – the calumnies devised by the
25
For the various sources, almost all of which post-date Pindar, see n. 28.
26
Among the several detailed discussions of these questions, see Köhnken 1974 and 1983, Gerber 1982,
55–56, Howie 1983 and Hubbard 1987; several of these authors survey the different sources for the
myth.
27
Hubbard 1987, 14–15.

32 deborah steiner
malice-bearing neighbour who spreads the specious story of the gods’
gourmandise. The manner in which Pindar reports these slanders, with
the choice of the punning term μέλη, at once body parts and songs,28 makes
evident his metapoetic design (47–51):
ἔννεπε κρυφᾷ τις αὐτίκα φθονερῶν γειτόνων,
ὕδατος ὅτι τε πυρὶ ζέοισαν εἰς ἀκμάν
μαχαίρᾳ τάμον κατὰ μέλη,
τραπέζαισί τ’ ἀμφὶ δεύτατα κρεῶν
σέθεν διεδάσαντο καὶ φάγον.
Someone of the spiteful neighbours straight off said in secret that into the
water boiling rapidly on the fire they cut you up limb from limb with a knife
and for the secondary course distributed your flesh around the table and
ate it.
The revised tale already previewed by the Pindaric song not only involves
disassembling and reconstituting these narrative pieces but also engages in
complex temporal revisionism and reversals too. Just as Pindar, introducing
his story at the very start of the mythical portion of his ode, recasts the ivory
shoulder as something there already at the moment of the hero’s birth, ‘a
peculiar birthmark and element of distinction’, as Köhnken describes it,29
and so as chronologically prior (a precedence mirrored by the poem’s own
design) to what the falsifying but pre-existing account presents as an ‘add-
on’, so the poet positions his reworked story as the originary tale, subse-
quently vitiated by the gainsayer who has altered its parts, removing some-
thing integral and then reattaching it in its later supplementary role.
Looking back at the lines introducing Pelops with these subsequent
complexities in mind, their ambiguous diction and syntax seem already
designed to flag the slippages that Pindar’s act of narrative revisionism
involves and to forestall the possibility of distinguishing the prosthetic
feature from the body to which it belongs, making the figure of Pelops the
product of both biological and technical processes and an assemblage of the
inborn and constructed, the organic and material/artificial. Suggestive of
the natural act of birth, as noted above (p. 31), is the presence of one of the
three Fates playing midwife here, as well as the mention of the ‘pure’
cauldron, a vessel that, as the accompanying adjective indicates, serves both
to cleanse the newborn baby and, again in more metapoetic fashion, to
sanitize an impious story, a trope for Pindar’s ongoing compositional act.
28
As Hubbard 1987, n. 60 notes, the pun is again used by Pindar at Nem. 11.15–18.
29
Köhnken 1983, 71.

But insofar as Pelops is ‘boiled’, a detail postponed until line 48, where the
poet describes the water bubbling up in the heated cauldron, his body also
becomes the object of two technological processes.30 While the first of these
looks to the culinary art, which transforms the raw into the cooked through
the application of fire (itself a cultural product), the second, as the phrase
χρυσὸς ἑψόμενος in Nem. 4.82 indicates, suggests an act of additional refine-
ment and points us to the sphere of the artisan and metallurgist: to boil gold is
to temper or smelt the metal, a treatment aimed at ridding the substance of
any impurities prior to its casting in the form of a manufactured good. Like
the hammer wielded by Hephaestus on the occasion of Athena’s genesis from
the head of Zeus on a red-figure pelike of ca. 450 (the goddess is full-grown
and appears equipped with a metal-forged helmet and spear),31 or Tyndareus’
axe in the scene of Helen’s birth from an egg on an Apulian krater of ca.
375–50 attributed to the Dijon Painter,32 so the narrative details supplied by
Pindar position Pelops’ emergence midway between a natural birth and an act
of artisanship. Add to this the semantic difficulties embedded in the syntax of
the phrase ἐλέφαντι φαίδιμον ὦμον κεκαδμένον (27). Reading the participle as
an indication that Pelops’ distinction inheres in his ivory shoulder (‘excelling
as to his shoulder’), the verb exactly suits Pindar’s version of events; but if that
shoulder is a feature with which the hero is ‘equipped’, as καίνυμαι also allows,
then it appears more as the supplement of the rejected account.
Finally, there is the substance from which that shoulder is fashioned.
Granted, Pindar was not at liberty to emend this element in his story,
a motif de rigueur insofar as the ivory shoulder was, relic-like, on display at
Olympia, and likely already visible during the poet’s lifetime.33 But the
material also suits his thematic purpose very well: from the Homeric poems
on, ivory comes surrounded by a teasing set of associations which look
again, at one and the same time, to the organic and ‘artified’. Used on one
occasion of Penelope’s skin as it is purified and embellished at the hands of
Athena, here acting as divine cosmetician (Od. 18.196, ‘whiter than sawn
ivory’), it also appears in the context of a work of human artistry, where
ivory serves to fashion a cheek piece for a horse (styled an agalma), which is
being stained with red dye (Il. 4.141–7). In this Iliadic instance, Homer
cites the technical intervention by way of comparandum for the blood
30
See Hubbard 1987, 8 and 14 for several of the points that follow.
31
London, British Museum E410, Beazley Archive Pottery Database 205560, with Uhlig 2020, 63–4, for
discussion. The hammer typically figures in foundry scenes (see below, p. 36–37).
32
Bari, Museo Archeologico Provinziale 3899; for discussion, see Walsh 2009, 135–7, and Uhlig
2020, 62.
33
See Paus. 5.13.4–6 and Pliny NH 28.34 for this.

34 deborah steiner
saturating the wounded Menelaus’ thighs, devising a verbal figure that
newly combines a natural process with an act of skilled craftsmanship
featuring hard matter and a man-made compound. Most notorious is the
Odyssean passage where Penelope, exploiting the homonyms ἐλέφας and
ἐλεφαίρω, punningly observes that deceptive dreams are the ones that issue
from the gates of ‘sawn ivory’ (19.564–5). Pelops’ shoulder very much
retains the enigmatic quality of the substance in these Homeric precedents:
does the ivory element refer in more imagistic fashion to the heightened
sheen and dazzle of the hero’s ultra-white skin (an instigator of divine
desire), or does it signify the presence of inorganic matter, something that
must be shaped and worked before it can fulfil its function?
Without resolving the confusions in his account, Pindar presents the
‘finished’ human body as a medley of organic and hard matter, flesh and
ivory, something that only realizes its most complete and compelling form
through acts of refined artistry and the poet’s corresponding verbal craft, whose
capacity for deception he does not deny even as he claims its positive valence.34
The equivalence between the human body, its genesis and formation, and
goods that result from artisanal processes seems furthermore to be written into
the Greek language and patterns of thought: in the compelling argument
proposed by Maria Karvouni, ‘in the Greek mind, “coming into being” stands
parallel to a “tektonic process”, resulting from processes of both division and
adding on’.35 As she further observes, the human body, demas, is cognate with
the verb δέμω, a term that indicates not just building, but ‘suggests complete-
ness and totality’ while the tektôn and his technê belong to the same root as
τίκτω, the verb for giving birth. Much as Pelops exemplifies, a body is
‘understood as a unified construct’ or totality, a built-up or skilfully assembled
composite made up of different parts joined into one so as to form a seamless
whole.36
Although the extant visual repertoire includes no representation of
Pelops with his ivory shoulder, two images that blur the boundaries
between the man-made and animate offer fresh considerations of the
quandaries that surround the body qua tectonic artefact. A red-figure
cup attributed to the Kiss Painter of ca. 500 (Figure 1.4), produced about
34
Pindar explores these ideas in the lines that immediately follow Pelops’ first appearance in the poem
(28–32). Note too the reuse of δαιδάλλω at 105, the verb that initially described stories ‘decked out
with lies’ (29) but now refers to Pindar’s wholly positive intention of glorifying his patron through
the embellishments of his song.
35
Karvouni 1999, 107. The citations that follow come from her pp. 10 and 11.
36
For this argument, see Zeitlin 1996, 41–2, drawing on Starobinski 1975. For the analogy between the
beautified Odysseus and a work of art, see Od. 23.159–62 and, for Penelope, the lines cited earlier
from book 18.

Figure 1.4 A red-figure cup attributed to the Kiss Painter of ca. 500. Baltimore,
Johns Hopkins University Archaeological Collection B 5, Beazley Archive Pottery
Database 201626.
a quarter-century before the composition of Olympian 1, shows in its tondo

a familiar scene: a youthful athlete appears together with his older trainer,
who stands before him, inclining his stick towards the boy’s body as
though to correct his stance. The fillet around the athlete’s head indicates
his recent victory in an agonistic event, while the downward glance of the
older bearded man, together with the kalos inscription acclaiming this (or
another) youth, gives the representation the erotic charge that these images
so regularly possess. But for all the ‘by-the-book’ nature of the scene
(familiar elements include the sponge, pick and aryballos suspended on
the wall), the presence of the pedestal on which the athlete stands would
give a viewer pause. Well accustomed to the iconography of victory images,
in which a winner was portrayed as an ephebic youth mounted on a plinth,
typically wearing his crown, the artist poses the question of whether this is
an actual athlete or an image of the same. Since the comparison between
beautiful bodies and sculpted images is a poetic commonplace, particularly
in the case of athletes,37 has the painter realized that conceit and addition-
ally suggested that, under the amorous gaze of the older man who views the
37
See Steiner 1998 for this.

36 deborah steiner
juvenile victor as an object of delectation (much as Poseidon does on
witnessing the body of his erômenos), the youth has been literally calcified?
A second image on a red-figure cup from ca. 490, the name vase of the
Foundry Painter, shows a more complex scene on each of its two external
faces, both of particular pertinence to the dismemberment and reassem-
blage of the Pindaric Pelops and to his interstitial status. On side
A (Figure 1. 5), we see a bronze foundry with several artisans, one of
whom hammers at a statue shown semi-prone with its arms raised up
high. In the preface to his illuminating discussion of the cup, Richard Neer
identifies the ‘rhetorical strategy’ at work in its several scenes, which is ‘to
confound the distinction between the depiction of a thing and the thing
itself, between depiction and depicted’.38 Among the confusions that he
goes on to treat belongs the visual echo between the head of one of the
workers (he is crouching behind the furnace where he plies the bellows),
Figure 1.5 A red-figure cup from ca. 490, the name vase of the Foundry Painter.
Berlin, Antikensammlung, Staatliche Museen zu Berlin F 229, Beazley Archive
Pottery Database 204340. Sides A and B.
38
Neer 2002, 79, from whom the subsequent discussion is almost wholly drawn.

shown frontally as it peeps out, and the two disembodied also forward-
looking faces hanging in the form of plaques on the wall to his upper left.
Surrounded by other such pinakes, these are marked as painted images and
‘artisanal products’.39 Exactly midway between the living and manufac-
tured is another bodiless head, this time shown in profile in between the
legs of one of a second pair of workers, who bends over it as he hammers at
the statue he is in the process of assembling, fixing on its still-missing body
parts. As it looks upwards and towards the right, the head positioned on the
ground offers a virtual mirror image of the downward-gazing, left-facing
countenance of the ‘real’ artisan resting on his mallet to its left, even as it is
clearly destined to be joined onto the headless statue lying prone on the
ground almost adjacent to it.
Additional blurring occurs with the presence of two feet that seem to
dangle from the wall and visually recapitulate the out-turned feet of the
individual who, seated on the stool at the centre of the scene, stokes the
furnace. Noting that this figure’s legs also resemble greaves more than
appearing to belong to the living body, Neer connects them with these
same manufactured artefacts as configured in the tondo, now unequivocally
defined as objects that are placed in a storage rack (Figure 1.6). Fresh
exchanges and overlaps between manufactured goods and the animate
occur on side B (Figure 1. 5), which includes two diminutive individuals
working at the oversized sculpted body of a warrior, whose pose points the
viewer back to the central youth, seemingly one of the living artisans, shown
leaning on his mallet with the same torsion of his body, on the opposite face.
On several counts, these several scenes offer their own reflections on one of
the central concerns voiced by the Pindaric ode: how do we determine reality
from representation, and to what degree can crafted bodies, the products of
the affective engagement between artisans, their tools and inorganic matter, be
distinguished from the living and animate? Particularly apposite to Pelops is
an earlier reading of the scene that Neer alerts us to. As he comments,
although without reference to the Pindaric Pelops story where the same
complications abound, ‘no surprise, therefore, that nineteenth-century viewers
of the cup thought that it represented a cannibal feast’.40
Animated Architecture in Pindar’s Paean 8

The final object treated here, the Delphic temple of Apollo described in
Pindar’s eighth Paean, both points us back to the performing troupes of
39 40
Neer 2002, 81. Neer 2002, 84.

38 deborah steiner
Figure 1.6 The tondo of the red-figure cup by the Foundry Painter. Berlin,
Antikensammlung, Staatliche Museen zu Berlin F 229, Beazley Archive Pottery
Database 204340.
tripods and bellows with which my discussion began and reintroduces the
craftsman god from that earlier scene. But unique to Pindar’s account of
this quasi-animated structure is the suggestion seemingly written into the
story’s dénouement:41 that manufactured goods that, in one way or
another, supersede their materiality and take on the attributes of living
things violate a divinely sanctioned demarcation and must be stripped of
their transgressive properties and returned to the inanimate realm to which
they properly belong.
In what remains of this highly lacunose composition, most probably
produced for performance at or in proximity to the recently completed
Alcmaeonid-financed temple at Delphi, the poet describes the sequence of
buildings that previously occupied the site, cobbling together his account
41
As noted below, the very fragmentary nature of the passage at this point in the papyrus allows for
a number of different interpretations.

from what seems like a patchwork of motifs from Homer and Hesiod, pre-
existing traditional tales and elements drawn from Pindar’s own richly
imaginative store.42 The best preserved portion of the song introduces the
third of the Delphic structures, the work of Hephaestus and Athena, and
details first the materials from which it was fashioned and then one of its
decorative features (fr. 52i.102–24 S.–M.).
ὦ Μοῖσαι, το<ῦ> δὲ παντέχ[νοις
Ἁφαίστου παλάμαις καὶ Ἀθά[νας
τίς ὁ ῥυθμὸς ἐφαίνετο;
χάλκεοι μὲν τοῖχοι χάλκ[εαί
θ’ ὑπὸ κίονες ἕστασαν,
χρύσεαι δ’ ἓξ ὑπὲρ αἰετοῦ
ἄειδον Κηληδόνες.
ἀλλά μιν Κ̣ρόνο̣υ̣ π̣α̣ῖ̣ [δες
κεραυνῷ χθόν’ ἀνοιξάμ[ε]ν̣ο̣[ι
ἔκρυψαν τὸ [π]άντων ἔργων ἱερώτ[ατον
γλυκείας ὀπὸς ἀγασ̣[θ]έντες,
ὅτι ξένοι ἔφ[θ]<ι>νον
ἄτερθεν τεκέων
ἀλόχων τε μελ[ί]φρονι
αὐδ[ᾷ θυμὸν ἀνακρίμναντες· επε̣ [
λυσίμβροτον παρθενίᾳ κε̣ [
ἀκηράτων δαίδαλμα [
ἐ̣ νέθηκε δὲ Παλλὰς ἀμ[
φωνᾷ τά τ'έ όντα τε κα[ΐ
πρόσθεν γεγενημένα ]ται Μναμοσύνα[
]παντα σφιν έφρα[σ.ν ant. D?
“ ]αιον δόλον άπνευ[-
But what, O Muses 102–24 rhuthmos was shown by the all-fashioning skills of
Hephaestus and Athena? Bronze were the walls, bronze columns stood
beneath, and six golden Kêlêdones (Charmers) sang above the gable. But
the sons of Kronos opened the ground with a thunderbolt and hid it, the
most holy of all works . . . astonished at the sweet voice, that visitors wasted
away apart from children and wives, hanging up their spirits as dedication to
the voice that is like honey to the mind . . . contrivance that causes mortals to
fall into fatal dissolution, of pure (words: ἐπέ[ων [Snell]?) in the maiden’s . . .
and Pallas put in . . . the voice and (the daughters of?) Memory told them
everything that is and that was before (and that will be?). . .(making)
breathless
42
For recent discussions of the work’s performance context and the myth it narrates, see Rutherford
2001, 210–32, Power 2011 and Weiss 2016.

40 deborah steiner
Emphatic in this description is the materiality of the temple, whose
anaphoric bronze walls and pillars might put the audience in mind of the
historical shrine of Athena Chalkoikos (‘of the Bronze House’) built on the
Spartan acropolis and cited in fifth-century and later sources; the temple,
dated to the sixth century, was apparently sheathed in bronze plates, some of
which are still extant. But the noun with which Pindar frames his inquiry
would from the first condition an audience’s understanding of this notional
building and offers an initial pointer towards its capacity to overcome the stasis
that should properly constrain a metal dwelling: rhuthmos, a Pindaric hapax,
would not appear again in an explicitly architectural context until the work of
Philo Mechanicus (ca. 200 bc), when, at Syntaxis 4.4, this writer on mechan-
ics and technology remarks, ‘it was not possible at the very beginning to
establish the forms (ῥυθμoί) of works of architecture without engaging in prior
experimentation’. Philo uses the term in slightly altered fashion later in the
passage, now invoking the notion of eurhuthmia, in what may be a reference to
the widespread practice, apparently already in use in the sixth century, of
altering column proportions by means of entasis and other structural refine-
ments so as to factor in the distortions created by the viewer’s eye: ‘by a process
of trial and error, adding to [the parts of the buildings’] bulk and again
subtracting from them . . . architectural forms are produced which are suited
to the vision and have the appearance of being well-shaped (εὔρυθμα)’. In
Strabo, eurhuthmia is the overall impression created by the temple of Artemis
Leukophryene at Magnesia-on-the-Maeander (14.1.40), whose sacred enclos-
ure surpasses that of the Artemision at Ephesus owing to ‘the well-designed
quality (εὑρυθμία) of its appearance’.
But does the Pindaric use of rhuthmos already have the technical ring
that these later architectural discussions give to it? Commentators have
puzzled over the Paean’s introduction of the term, variously used in
seventh- and sixth-century sources for the ‘disposition’ of an individual,
the ‘pattern’ of human life or perhaps man’s alternation between one
condition and another. Post-Pindaric fifth-century authors give the noun
the meaning ‘shape’, ‘form’ and ‘configuration’ and also apply it to regular,
repeated movement, whether that observed in marching or, at
Aristophanes Thesm. 985, in choral dancing (πάλλ’, ἀνάστρεφ’ εὐρύθμῳ
ποδί). Although the term’s etymology remains unresolved, some deriving
it from ῥέω, ‘to flow’, others from ἐρύω, ‘to pull, draw’, Barbara Kowalzig’s
recent discussion follows, in somewhat modified form, Emile Benveniste
in placing motion and fluidity at the concept’s centre – ‘rhuthmos desig-
nates the form when it is moving’ – and refers to, in Benveniste’s phrase, ‘a

configuration of movements organized in time’;43 so too, according to
Jerome Pollitt, ‘a single well-chosen rhuthmos could, in fact, convey the
whole nature of movement’.44 In this ‘dynamic’ interpretation, the refer-
ence to ‘flowing movement’ in Paean 8 may describe the impression of
a mobile structure generated by the play of light on the metal surfaces.45
But, as Timothy Power proposes, Pindar most likely chooses rhuthmos
because it succinctly unites two different spheres, the architectural, as
visible in later accounts, and the musico-choral with which the poet will
chiefly be concerned, and already invests the temple with the ‘kinetic and
sonoric’ dimensions more fully realized in the choral ensemble of the
Kêlêdones (‘Charmers’) topping the structure, whom the very next lines
introduce.46 Siren-like beguilers, their irresistible song would supply the
music and set the cadence to which the structural components just detailed
move and even dance. Eugen Petersen’s much earlier discussion of rhuth-
mos observes that link with dancing, also visible in the expression used by
Aristophanes cited above, and suggests that rhuthmoi would originally –
and in a usage on which Pindar may be drawing – have described the
‘positions’ or schêmata repeatedly assumed by the body in the course of the
dance, whose intervals it marked.47 Commenting on this view, Pollitt
notes that the synchronization of dance and music would mean that ‘the
recurrent positions taken by the dancer in the course of his movements also
marked distinct intervals in the music; the rhuthmoi of the dancer thus
became the rhuthmoi of the music’.48 In Plato, rhuthmos expressly describes
the measured motions of the chorus, these marked out by the voices
performing the words and melody, and by the strike of the choristers’
feet. What the Pindaric audience would witness, as it watched the singer-
dancers perform the Paean with the current temple in their line of sight,
would be nothing less than the choristers’ embodiment of the building(s)
their words describe.49
Beyond their representation as a chorus that not only sings but by
definition dances too, the introduction of the Kêlêdones would reinforce
the poem’s opening suggestion of a building in motion. In devising the
Charmers, wholly unattested in earlier sources, Pindar seems to have had
several models in mind as he recasts the Homeric Sirens in the likeness of
the akroteria familiar from Greek archaic and classical temple architecture,
43
Kowalzig 2013, 182 and 184, citing Benveniste 1971, 281–2. 44 Pollitt 1974, 139.
45
This is the explanation of Rutherford 2001, 219. 46 Power 2011, 79.
47
Petersen 1917; Kowalzig 2013, 182, without referring to the piece, modifies that view, pointing out
that schêmata describe static and rigid poses, and not the movements between them.
48
Pollitt 1974, 141. 49 This is the reading proposed at greater length in Weiss 2016.

42 deborah steiner
where they stand at the ridge and corners of the gables (so αἰετοῦ at 107, if
the text is sound) and most typically take the form of Sphinxes, although
Nikai are commonplace subjects too. Wings are the chief point of overlap
between these several figures, and many readers, both ancient and modern,
propose that the Pindaric Charmers share the feature with their architectural
templates; although the Odyssean Sirens are conspicuously wingless, archaic
lyric poets and artists regularly depict the songstresses as a winged ensemble,
and Philostratus, who would identify the Kêlêdones as wrynecks (VAp. 6.11),
assumes the same of Pindar’s maidens. Wings have already made a passing
appearance in Paean 8 where, at the start of the third triad, the poet described
the second temple, apparently built by bees from wax and feathers, as winged
(πετ[ην- or πετ[ειν-),50 a characterization that may look to a tradition of
‘winged’ architecture also visible in the terms ἀειτός, derived, according to
some, from the gable’s resemblance to the spread wings of the eagle,51 and
περίπτερος, used of the design of temples with a single row of columns.52
Where the powerful wind that carries off this second temple allows it to
realize its avian character as it literally takes flight, the implicitly winged
Kêlêdones signal the third building’s matching capacity to become airborne,
this time through its own kinetic properties.
But for Pindar, the Charmers’ choral (and avian) character most
patently declares itself in their supremely alluring voices, which receive
no fewer than three mentions in the extant block of lines, twice denoted by
different nouns, as though to suggest that no single term can adequately
express the marvel of these singers’ vocal powers. Topped by its anthropo-
morphized sextet, the Pindaric temple becomes sonorous as well as kinetic,
emitting not the speech that is quite regularly assigned to inanimate
objects, but, more unusually, song.53 Characterized by adjectives that
regularly describe the singing voice – ‘sweet’ and ‘honeyed’ – the golden
matter from which the Charmers are made seemingly stands at odds with
the qualities assigned to their vocalizations, as Pindar unites in one two
disparate and even antithetical categories, the metallic and the mellifluous,
the rigid and unchanging and the flowing and volatile.54 In Power’s
50
See Rutherford 2001, 218, for the reading.
51
Rutherford 2001, 219, suggests a derivation from the Near Eastern practice of decorating gables with
winged solar discs instead.
52
See Rutherford 2001, 225–6, and Weiss 2016, 243.
53
In some of the inscriptions recorded on the colossal statue of Memnon at Egyptian Thebes, the
image not only emits sounds and speech, but also sings. For these, see nos. 14 and 54 in Rosenmeyer
2018, appendix ii.
54
This is not the only occasion on which Pindar suggests that voices, particularly those of maidens,
may combine sweetness with the hardness of metal; he elsewhere describes the parthenaic choruses

attractive reading, the Charmers’ golden bodies, indices of their inhumanly
durable and unbreakable nature, in combination with their powers of song,
turns them into ‘indefatigable singers’, whose ‘automatic voice never stops
for breath’.55 Like a machine with ever-moving metal parts that never give
out, they sing ceaselessly, allowing auditors no respite, a scenario earlier
proposed by Homer but ruled out of court, when he wished that he might
acquire ‘ten tongues and ten mouths, and an unbreakable voice and heart
of bronze’ (φωνὴ δ’ ἄρρηκτος, χάλκεον δέ μοι ἦτορ ἐνείη, Il. 2.490) so as to
be able to recite the ‘multitude’ that came to Troy; declaring this a task that
would exceed the powers of the mortal bard, the poet imagines the
reification of his internal organs, their transformation into enduring metal-
lic matter.56
Distinctive too, and in a marked departure from the Homeric prece-
dent, is the impact that the Charmers’ singing has on its audience and that,
I suggest, alerts us to the transgressive nature of these divas’ powers and the
reasons for their (perhaps) ultimate suppression. Like the Odyssean Sirens
before them, so compelling are their voices that visitors who listen and
yield to the enchantment of their song forsake all chance of a return home.
But where Homer imagines the Sirens’ victims ending up as a heap of
rotting bones, in Pindar’s revisionary account the Kêlêdones’ spellbound
audience, quite literally transfixed by the performance, ‘hang up their
souls’ on the choristers’ ‘voice like honey to the mind’ (μελ[ί]φˈρονι αὐδ
[ᾷ θυ/ μὸν ἀνακρίμναντες, fr. 52i.116). As Ian Rutherford’s translation, cited
here, of the verb brings out, it evokes the dedication of a votive good,57
figuring the internal organ – the locus of auditory pleasure in other
accounts too – as one of the familiar plaques that, equipped with holes,
were regularly hung from a tree in proximity of the sanctuary, or affixed to
a wall, ceiling beam or temple column, or even suspended from the cult
image itself (cf. Herodotus 5.77 and 95). Even as Pindar vivifies the
Charmers by granting them voice and motion, so he suggests that their
confusion of the living and inanimate causes their auditors to undergo
a corresponding but inverse change of state: from sentient beings, they
that perform at Delphi and Delos as ‘singing a sweet strain with brazen voice’ (χαλ/κέᾳ] κ̣ελαδ
[<έον>]τι γλυκὺν αὐδᾷ, fr. 52b.100–1, with discussion in Power 2011, 105). Since these are real-world
performers, and not the singing automata of Paean 8, the metallic conceit has become literal.
55
Power 2011, 91.
56
Also cited in Power 2011, 107. The epithet ἄρρηκτος is frequently used in Homeric and later sources
of metal objects, whether a bronze wall or fetters.
57
Rutherford 2001, 220, noting the two facets to the experience described, comments that the
expression suggests ‘both a religious dedication and psychological dependency’.

44 deborah steiner
become objectified offerings to this ‘adhesive’ voice to which they are
lastingly in thrall.
Much like magicians confronted with their products run amok, the
Pindaric gods seem to have no choice but to repress these objects of their
own manufacturing arts. In the lines where the poet introduces the
Kêlêdones, he already anticipates their story’s end with their dispatch
beneath the earth by the ‘sons of Kronos’ (Zeus and Poseidon?), who
bury the entire temple complex, consigning it and its chorus to the realm
where other immortals who dare to challenge the Olympian order undergo
punishment. For all that the papyrus becomes increasingly fragmentary at
the point where Pindar seems to circle back to the Charmers’ fate, enough
remains to suggest that the gods undertake their punitive action at least in
part because of the artefacts’ multiple violations of the proper relations that
should exist among men, and between men and gods.58 As commentators
variously propose, the singers’ baneful impact on their listeners, whom
Pindar styles xenoi (76), is not only, in the broadest terms, an abuse of the
powers that gods regularly give to human singers, but more particularly
constitutes an offence against Zeus Xenios, among the enactors of their
burial; so too the seemingly prophetic powers that the Charmers enjoy
positions them, in keeping with the Homeric Sirens, as rivals to the Muses,
even anti-Muses, who must be replaced by the divinely sanctioned and
vocally bounded Pythia who comes to occupy the Delphic shrine. Also
patent here is a perversion of the proper relations between a lyric chorus
and its audience, and the symbiotic relations that structure exchanges
between the singer-dancers and those attending their performance:59
votives should be offered to the gods, not to those who celebrate them in
tribute-rendering spectacles.60
But I would add to these interpretations the suggestion that the gods’
punishment is more narrowly targeted and trains its sights on the very source
of Kêlêdones’ maleficent impact. There is little consensus concerning the
meaning of line 124, which indicates that someone, most likely either Athena
or her agent Mnemosyne, ‘made breathless a cunning device of old’. Where
Rutherford prefers to understand the phrase as part of a ‘flash-back’ to the
58
Rutherford 2001, 222, following a review of the various nterpretations, offers a different reading. In
his view, the Kêlêdones’ voice goes on sounding from beneath the earth, an attractive possibility that
finds an analogue in Pindar and other sources’ representation of the punishment of the also-‘pan-
vocal’ Typhoeus.
59
The clearest expression of this reciprocity appears at HHAp. 161–4.
60
For several of these different accounts of the Charmers’ ‘crimes’, see Rutherford 2001, 220–22, and
Power 2011, 91–101.

original construction of the temple and to the Charmers’ manufacture,61
Pindar may instead maintain the ring-compositional design observed so far
and present this action as part of the penalty exacted by the gods; in a move
to strip the enchanters of the source of their deleterious powers, they deprive
them of the breath that, as earlier noted, supplies the wherewithal of speech
and song.62 Indeed, the first extant usage of the closely cognate term
ἄπνευστος describes exactly this speechlessness: barely sentient as he
emerges from his buffeting by the waves, Odysseus washes up on the
shore of Scheria ‘breathless and speechless’ (ἄπνευστος καὶ ἄναυδος, Od.
5.456). Since voice is, with motion, among the prime demarcators of a living
being, and the incapacity to speak the hallmark of objects fashioned from
stone, wood, metal and other inert matter, the Charmers are returned to the
sphere of silent and immobile artefacts from which they have, misguidedly,
been permitted to emerge.
RE F E RE N CE S
Benveniste, E. 1971. ‘The Notion of “Rhythm” in Its Linguistic Expression’, in
E. Benveniste, Problems in General Linguistics (Florida), 281–9.
Clarke, M. 1997/8. ‘πινύσκω and Its Cognates: A Note on Simonides, Fr. 508
Page’, Glotta, 74: 135–42.
Edwards, M. 1991. The Iliad: A Commentary. Volume v. Books 17–20 (Cambridge).
Frontisi-Ducroux, F. 2002. ‘“Avec son diaphragme visionnaire: Ἰδυίῃσι
πραπίδεσσι”, Iliade xviii, 481. À propos du bouclier d’Achille’, Revue des
études grecques, 115: 463–84.
Gerber, D. E. 1982. Pindar’s Olympian One: A Commentary (Toronto).
Howie, J. G. 1983. ‘The Revision of Myth in Pindar Olympian One: The Death
and Revival of Pelops (25–27, 36–66)’, Papers of the Liverpool Latin Seminar,
4: 277–313.
Hubbard, T. 1987. ‘The Cooking of Pelops: Pindar and the Process of
Mythological Revisionism’, Helios, 11: 3–21.
Karvouni, M. 1999. ‘Demas: The Human Body as a Tectonic Construct’, in
A. Pérez-Gómez and S. Parcell, eds., Intervals in the Philosophy of
Architecture (Montreal), 103–24.
Köhnken, A. 1974. ‘Pindar as Innovator: Poseidon Hippios and the Relevance of
the Pelops Story in Olympian 1’, Classical Quarterly, 24: 199–206.
Köhnken, A. 1983. ‘Time and Event in Pindar O. 1.25–33’, Classical Antiquity, 2:
66–76.
Kowalzig, B. 2013. ‘Broken Rhythms in Plato’s Laws: Materialising Social Time in
the Chorus’, in A.-E. Peponi (ed.), Performance and Culture in Plato’s Laws
61 62
Rutherford 2001, 222. For discussion, see Clarke 1997/1998.

46 deborah steiner
Kurke, L. 2015. ‘Pindar’s Material Imaginary: Dedications and Politics in
Olympian 7’. UCL Housman Lecture.
Langdon, S. 2008. Art and Identity in Dark Age Greece (Cambridge).
Lissarrague, F. 1990. The Aesthetics of the Greek Banquet: Images of Wine and Ritual,
transl. A. Szegedy-Maszak (Princeton, NJ).
Neer, R. 2002. Style and Politics in Athenian Vase-Painting (Cambridge).
Onians, R. B. 1988. The Origins of European Thought: About the Body, the Mind, the
Soul, the World, Time and Fate (Cambridge).
Papalexandrou, N. 2005. The Visual Poetics of Power: Warriors, Youths, and Tripods
in Early Greece (Lanham, MD).
Payne, H. 1971. Necrocorinthia: A Study of Corinthian Art in the Archaic Period
(College Park, PA).
Petersen, E. A. H. 1917. Rhythmus. Abhandlungen der Königlichen Gesellschaft
der Wissenschaften zu Göttingen. Philologisch-historische Klasse 16.5
(Göttingen).
Pollitt, J. J. 1974. The Ancient View of Greek Art: Criticism, History, and
Terminology (New Haven, CT).
Power, T. 2011. ‘Cyberchorus: Pindar’s Κηληδόνες and the Aura of the Artificial’,
in L. Athanassaki and E. Bowie, eds., Archaic and Classical Choral Song:
Performance, Politics and Dissemination (Berlin), 67–114.
Rosenmeyer, P. A. 2018. The Language of Ruins: Greek and Latin Inscriptions on the
Memnon Colossus (Oxford).
Rutherford, I. 2001. Pindar’s Paeans: A Reading of the Fragments with a Survey of the
Genre (Oxford).
Starobinski, J. 1975. ‘The Inside and the Outside’, The Hudson Review, 28: 333–51.
Steiner, D. T. 1998. ‘Moving Images: Fifth-Century Victory Monuments and the
Athlete’s Allure’, Classical Antiquity, 17: 123–49.
Steiner, D. T. 2001. Images in Mind: Statues in Archaic and Classical Greek
Literature and Thought (Princeton, NJ).
Steiner, D. T. 2013. ‘The Priority of Pots: Pandora’s Pithos Re-viewed’, Mètis, 11:
207–34.
Steiner, D. T. 2014. ‘From the Demonic to the Divine: Cauldrons, Choral
Dancers and Encounters with the Gods’, in S. Estienne, V. Huet,
F. Lissarrague and F. Prost, eds., Figures de dieux: Construire le divin en
images (Rennes), 155–74.
Steiner, D. T. 2021. Choral Constructions in Greek Culture: The Idea of the Chorus
in the Poetry, Art and Social Practices of the Archaic and Early Classical Periods
(Cambridge).
Topper, K. 2007. ‘Perseus, the Maiden Medusa and the Imagery of Abduction’,
Hesperia, 76: 73–105.
Topper, K. 2010. ‘Maidens, Fillies and the Death of Medusa on a
Seventh-Century Pithos’, Journal of Hellenic Studies, 130: 1–19.
Uhlig, A. 2020. ‘Birth by Hammer: Pandora and Somatic Construction’, in
J. Dyer and A. Surtees, eds., Exploring Gender Diversity in the Ancient
World (Edinburgh), 54–66.

Walsh, D. 2009. Distorted Ideals in Greek Vase-Painting: The World of Mythological
Burlesque (Cambridge).
Weiss, N. A. 2016. ‘The Choral Architecture of Pindar’s Eight Paean’, Transactions
of the American Philological Association, 146: 237–55.
Zeitlin, F. 1996. Playing the Other: Gender and Society in Classical Greek Literature
(Chicago).

chapter 2
Automata, Cyborgs, and Hybrids

Bodies and Machines in Antiquity
Jane Draycott
Introduction
Impairment and disability in classical antiquity have formed the basis of
a variety of scholarly enquiries over the course of the last two decades.1
Increasingly, the theoretical approaches of the new academic discipline of
Disability Studies are being applied to ancient material such as literary,
documentary, archaeological, and bioarchaeological evidence for impair-
ment and disability in classical antiquity as a matter of course. Yet despite
this, since the lived experience of the impaired and disabled in ancient
Greece and Rome is not readily described, discussed, or even depicted in
detail by ancient authors or artists, it has not been a consistent focus of
study.2 Perhaps this is due to the fact that any attempt to access this
experience requires piecing together fragmentary information drawn
from the wide range of evidence previously adduced, then comparing
these findings with those from neighbouring ancient civilisations (e.g.
Egypt, India, China).3 It is notable that one particular aspect of the lived
experience of impairment and disability in classical antiquity that has
received surprisingly little attention to date, despite increasing amounts
of attention having been paid to ancient technology over the last decade, is
the assistive technology that could and would have been utilised by
1
On impairment and disability in classical antiquity, see Garland 1995 (2010); Rose 2003 (2013);
Breitwieser 2012; Laes, Goodey, and Rose 2013; Laes 2014; Krötzl, Mustakallio, and Kuuliala 2015;
Laes 2017; Laes 2018; articles surveying the period include Kelley 2007. On impairment and disability
in Judaism, Christianity, and the Bible, see Ohry and Dolev 1982; Abrams 1998; Avalos, Melcher, and
Schipper 2007; Fishbane 2007; Olyan 2008. For studies on specific conditions, see for example Dasen
1993; Harris 2013; Trentin 2015.
2
On attempting to reconstruct the lived experience of an impaired individual in Roman Egypt, see
Draycott 2015. On impairment and disability in ancient art, see Grmek and Gourevitch 1998; also
individual chapters in some of the above volumes, such as Mitchell 2013; Mitchell 2017; Trentin 2017.
3
See for example a comparison between a Roman prosthetic leg and a Chinese prosthetic leg, both
dating from the third century bc, at Draycott 2018b.
48

Automata, Cyborgs, and Hybrids 49
individuals with impairments that affected their physical mobility in
classical antiquity.4
Yet from the earliest surviving classical literature, when individuals
experiencing physical impairments such as lameness or the loss of
a limb are attested, they are depicted utilising a variety of different
types of technology to assist them in their everyday activities, and it is
clear from these accounts that these technologies are envisaged as
having a significant impact on their quality of life. Such individuals
might create automata to act as their assistants.5 Such individuals might
replace their lost body parts with man-made substitutes, thereby ren-
dering themselves cyborgs.6 Such individuals might requisition and
utilise the bodies of other humans or even animals, thereby rendering
themselves hybrids.7 The Greek god Hephaestus is firmly situated
within the history of technology by both ancient and modern com-
mentators, and there seems to have been a strong connection between
impairment and technology, and between technology and impairment,
in the minds of Greeks and Romans, with impaired individuals
described as undertaking technical trades and warnings issued that
those who practise technical trades risk becoming impaired as a result.8
This chapter will examine the use of technology as a means of supple-
menting, augmenting, and even transforming the impaired body in
ancient Greece and Rome. It will start by surveying mythological
examples of impaired individuals using assistive technology and assess-
ing the insights they can offer into the ways in which the impaired
body was scrutinised, before investigating the extent to which any
insights gained from these examples can be applied to real examples
4
To date, examinations of the use of technology for medical purposes in antiquity have concentrated
on medical instruments – see the publications of Lawrence Bliquez, Ralph Jackson, and Ernst
Künzl – although medical machines have been discussed in the context of the history of technology,
on which see the publications of John Peter Oleson.
5
The English words automaton (singular) and automata (plural) are taken from the Greek word
αὐτόματος, ‘acting by one’s own will, of oneself’, and indicate a self-moving object.
6
The English word cyborg was coined in 1960 by Manfred E. Clynes and Nathan S. Kline,
a portmanteau of ‘cybernetic organism’, used to denote a being with a combination of organic and
biomechatronic body parts; for explication, see Clynes and Kline 1960. A cyborg might be
a restorative cyborg, in which a lost body part is replaced with an equivalent, or an enhanced cyborg,
in which a lost body part is replaced with a superior one. For the purposes of this chapter, I shall be
considering ancient individuals who can potentially be viewed as cyborgs as restorative cyborgs rather
than enhanced cyborgs.
7
The English word hybrid refers to a thing made by combining two different elements. Numerous
hybrids are present in ancient Greek and Roman culture; for an overview, see Hughes 2010.
8
Other mythological figures such as Prometheus and Daedalus are likewise firmly situated within the
history of technology.

50 jane draycott
of impaired individuals. To what extent could impaired and disabled
individuals in ancient Greece and Rome attempt to utilise technology
as an aid or even a cure for their conditions? And if they could, to what
extent did they? At what point does technology, when used in this
manner, cease to be simply assistive and in fact become regenerative?
The Greek God Hephaestus

This chapter will utilise the Greek god Hephaestus as a starting point
for an examination not only of physical impairment but also assistive
technology in classical antiquity. Hephaestus is the god of fire, smiths,
artisans, metalworking, stonemasonry, and sculpture, yet despite this
he was not often mentioned in ancient literature and appears to have
been ‘not a very important god in historical times’.9 He is the only
physically impaired Olympian deity.10 He is consistently depicted in
ancient literature as being physically impaired; his legs and feet are
variously described as being lame, crooked, clubfooted, and so on.11
While earlier sources focus on the fact of his impairment, later sources
are more concerned with its implications, particularly the way that it
affects how he moves.12 Explanations for his physical impairment vary:
in some versions, he was lame and weak from birth; in others, he was
injured when Zeus threw him from Mount Olympus and he crash
landed on Lemnos.13 It is possible that the rationale for Hephaestus
being born lame was that Hera, competing with Zeus, who had
generated Athena, attempted to generate Hephaestus herself, contrary
to the natural order.14 His physical impairment is one of his defining
characteristics; two of his sons, Periphetes (also known as Korynetes)
and Palaimonios, were said to have inherited it from him, which
9
Brennan 2016, 163; Bremmer 2010, 193.
10
Hephaestus is one impaired and disabled individual who has been frequently and comprehensively
discussed by scholars. See for example Hermary and Jacquemin 1988; Malten 1912 and 1913; Delcourt
1957; Brommer 1978; on Hephaestus’ Roman counterpart Vulcan, see Capdeville 1995.
11
Brennan 2016.
12
Brennan 2016, 178–9. This is in line with the World Health Organization’s definitions of the terms
‘impairment’ and ‘disability’: ‘Disabilities is an umbrella term, covering impairments, activity
limitations, and participation restrictions. An impairment is a problem in body function or
structure; an activity limitation is a difficulty encountered by an individual in executing a task or
action; while a participation restriction is a problem experienced by an individual in involvement in
life situations’: www.who.int/topics/disabilities/en/ (accessed October 2018).
13
For Hephaestus being lame from birth, see Homer, Iliad 8.136, 8.267; Homeric Hymn 3 to Pythian
Apollo 310. For Hephaestus being injured as a result of his fall from Olympus, see Homer, Iliad 1.568,
1.590; Valerius Flaccus, Argonautica 2.8.5; Pseudo-Apollodorus, Library 1.3, 1.19.
14
De Ciantis 2005, 185.

accords with ancient ideas about the hereditary nature of impairments in
classical antiquity.15 It was through their respective impairments that
their divine parentage was made clear, which was particularly useful for
the latter when his paternity was contested.16 It has notable physical
affects; he is described as being disproportionate, with legs slenderer than
one would expect them to be considering the size of the rest of him, and
he moves slowly and awkwardly, leaving him open to public mockery and
humiliation.17 The most famous example of this occurs in Homer’s Iliad,
when Hephaestus, serving the other Olympian deities drinks, is com-
pared to Zeus’ cupbearer Ganymede and found wanting.18 Much hilarity
ensues as a result of this juxtaposition.19
Considering Hephaestus’ position and his impairment, it is not surpris-
ing, then, that he is depicted in both ancient literature and ancient art as
utilising the sort of assistive technology that would have been readily
accessible to those suffering from impairments that affected their physical
mobility, such as the staff, the stick, the cane, or even the crutch; he is
described as using a stout staff (skêptron) to support himself as he moves
around his forge.20 However, he is also depicted in ancient literature and art
as utilising a variety of different types of assistive technology that would not
have been readily accessible to those suffering from impairments that
affected their physical mobility: he owns a pair of winged sandals and
a winged chariot, and his forge is filled with self-directing technology,
most notably automata in the form of a pair of maidens. Thus,
Hephaestus can potentially be seen as an early example of an individual
with the means to do so replacing impaired or lost body parts with man-
made substitutes and thereby rendering himself a cyborg. He can also,
15
Apollonius, Argonautica 1.202–6, 3.217. On belief in the hereditary nature of impairment and
disability in classical antiquity, see Hippocratic Corpus, Sacred Disease 3; Aristotle, History of
Animals 585b, 586a; Pliny the Elder, Natural History 7.11.50.
16
Apollonius, Argonautica 1.203.
17
For Hephaestus’ disproportionate physique, see Homer, Iliad 18.415, 20.36–7. For Hephaestus’ slow
and awkward gait, see Homer, Iliad 18.410; Homer, Odyssey 8.311, 8.330.
18
Homer, Iliad 1.584–600. In view of this, perhaps it is not surprising that Hephaestus is believed to
have created twenty tripod automata that attended the banquets of the gods, and whose purpose may
have been to serve the drinks in his stead; see Homer, Iliad 18.373–9. This connection seems to have
been made in antiquity: see Philostratus, Life of Apollonius of Tyana 3.27.2–3, in which Apollonius
travels to India and while there attends a banquet where he is served by four tripod automata and
bronze humanoid automata that resemble the cupbearers Ganymede and Pelops.
19
On the public mockery and humiliation of disfigured, deformed, or impaired individuals occurring
as a matter of course in classical antiquity, see Garland 1994. Yet see Halliwell 2008, 63, for the
suggestion that such individuals were not necessarily automatically laughed at, and if they were
laughed at, it does not necessarily follow that their differences were the reason for this laughter.
20
Homer, Iliad 18.416.

52 jane draycott
somewhat more prosaically, potentially be seen as an individual with the
means to do so requisitioning the bodies of other humans and animals to
compensate for his own deficiencies and thereby rendering himself a hybrid.
Finally, and most impressively, he creates automata to act as his assistants.
These technologies have a clear impact upon him, his life, and his lived
experience. For our purposes here, it does not necessarily matter whether the
creators of these depictions intended members of their audiences to envision
Hephaestus as something akin to a real-life smith, using his knowledge, skill,
and experience in conjunction with his actual tools (he is usually depicted
with the hammer, anvil, tongs, and bellows) and materials (metal, wood,
etc.).21 We could see these depictions as simply ‘embellishing a kind of
mystique surrounding craft activity’ and ‘[exaggerating] a kind of awe at
those who make devices’.22 Certainly, it would potentially have been awe-
inspiring to see someone how had been previously unable to walk without
physical assistance from a family member or friend now locomoting inde-
pendently thanks to the use of a staff, stick, cane, or crutch, or an extremity
prosthesis, or even some sort of frame.
Ancient Assistive Technology
Cyborgs: Staffs, Sticks, Canes, and Crutches; Corrective Footwear;

Extremity Prostheses
As mentioned above (p. 000), Hephaestus is originally depicted as walking
with the assistance of a stout staff. The staff, walking stick, or crutch
(usually ὁ σκίπων, τὸ ξύλον, or τὸ βάκτρον in Greek; baculum, bacillum,
or baccillum in Latin) is the piece of assistive technology most commonly
depicted in both literature and art, although examples are not commonly
found in the archaeological record due to the fact that organic materials
such as wood are only preserved in very wet or very dry contexts. Ancient
literature particularly associates the staff with the elderly, but it could be
utilised by anyone in need of support and stabilisation.23 The staff (or
walking stick, or cane, or crutch) has not been comprehensively studied.24
21
Berryman 2009, 24–5. However, see an alternative view presented in Devecka 2013, which notes the
passage of time between the archaic and the classical periods and the concomitant development of
ancient Greek technology and the ancient Greek understanding of mechanics.
22
Berryman 2009, 25.
23
For the staff as one of the standard attributes of old age, see Emery 1999, 23; Cokayne 2003, 54; for
physical weakness of the elderly generally, see Cokayne 2003, 53–6.
24
See Loebl and Nunn 1997 for a cursory overview; for a more detailed study, see Draycott (in
preparation). For studies of the iconography, see Couvret 1994; Brule 2006.

However, even a cursory glance at the scattered ancient literary and
material evidence for it indicates that it was possible for a considerable
amount of thought to go into the creation of one. Theophrastus recom-
mended mallow (Malva sylvestris) as a good material for a staff because the
plant could grow to great heights quickly and was strong and recom-
mended bay (Laurus nobilis) specifically for the creation of walking sticks
for the elderly because it was relatively lightweight.25 However, there are
references to other materials being utilised; various other types of wood
such as olive and fig, as well as other materials such as ivory, whalebone,
and gold. There are references to plain ones and elaborately decorated ones.
While someone might be fortunate enough to find a piece of wood just the
right size and shape, or someone with a basic knowledge of carpentry could
carve their own, there are indications that there could be much more to the
process of creating a staff than that.26 The author of the Hippocratic
treatise Joints views the provision of a crutch as part of the physician’s
duty of care to their patient and recommends a variety of different types of
crutch, the specific type of crutch recommended depending on the condi-
tion diagnosed and the treatment prescribed. Judging by the explanation
and justification given at the outset, this seems to be something of
a departure from common practice:
One might say that such matters [i.e. processes of rehabilitation] are outside
the healing art. Why, forsooth, trouble one’s mind further about cases
which have become incurable? This is far from the right attitude. The
investigation of these matters too belongs to the same science; it is
impossible to separate them from one another. In curable cases we
must contrive ways to prevent their becoming incurable, studying the best
means for hindering their advance to incurability; while one must study
incurable cases so as to avoid doing harm by useless efforts.27
So, for the author of Joints, at least, the crutch and how it would affect the
patient is viewed as an integral part of the treatment. For untreated
dislocations that had occurred in utero or in early childhood, one might
use one or two crutches.28 The difference in the quality of life for those who
25
Theophrastus, Enquiry into Plants 1.3.2, 5.7.7.
26
See Theocritus, Idyll 9.23–4 for an example of a natural staff given as a gift and reckoned as fine as
any produced by an artisan. See Acton 2014, 200, for carpentry at home in Classical Athens.
27
Hippocratic Corpus, Joints 58 (trans. E. T. Withington): φαίη μὲν οὖν ἄν τις, ἔξω ἰητρικῆς τὰ
τοιαῦτα εἶναι· τί γὰρ δῆθεν δεῖ περὶ τῶν ἤδη ἀνηκέστων γεγονότων ἔτι προσσυνιέναι; πολλοῦ δὲ
δὲ οὕτως ἔχειν· τῆς γὰρ αὐτῆς γνώμης καὶ ταῦτα συνιέναι· οὐ γὰρ οἷόν τε ἀπαλλοτριωθῆναι ἀπ᾿
ἀλλήλων. δεῖ μὲν γὰρ ἐς τὰ ἀκεστὰ μηχανάασθαι, ὅπως μὴ ἀνήκεστα ἔσται, συνιέντα ὅπῃ ἂν
μάλιστα κωλυτέα ἐς τὸ ἀνήκεστον ἐλθεῖν· δεῖ δὲ τὰ ἀνήκεστα συνιέναι, ὡς μὴ μάτην λυμαίνηται.
28
Hippocratic Corpus, Joints 53.

54 jane draycott
utilise one crutch or two compared to those who do not is made clear, to
the point where we can interpret this as an example of the physician
expressing sympathy for their patient and dismay at their quality of life:
In the case of adults, their mode of walking has been described; but when
this accident occurs in those who are very young, for the most part they lack
energy to keep the body up, but they crawl about [miserably] on the
sound leg, supporting themselves with the hand on the sound side on
the ground. Some even among those to whom this accident happens when
adult lack the energy to walk standing up; but when persons are afflicted by
this accident in early childhood and are properly trained, they use the sound
leg to stand up on, but carry a crutch under the armpit on that side, and
some of them under both arms. As for the injured leg, they keep it off the
ground, and do so the more easily, because in them the injured leg is smaller;
but their sound leg is as strong as if both were sound.29
For dislocations that occurred in adulthood and were not successfully
reduced, leading to one leg being significantly shorter than the other,
they could use a long crutch if they were capable of walking erect but
either could not or did not want to place their foot on the ground, or
a shorter crutch if they could or wanted to place their foot on the ground.30
The author of Instruments of Reduction likewise made recommendations
regarding the use of mobility aids, similar but not identical to those of the
author of Joints, stating that a crutch should be short rather than long, since
if it were long the user would not use the foot.31
Hephaestus is frequently depicted in ancient works of art as wearing
winged shoes. These depictions usually but not always occur in conjunc-
tion with him assisting with the birth of Athena, and so in this context the
winged shoes could be interpreted as lending him speed, making the
normally slow god as swift as Hermes and Perseus who are generally
endowed with those attributes, something very necessary in such an
extreme situation. Considering the emphasis placed upon Hephaestus’
slow and uneven gait in ancient literature, in increasing Hephaestus’
29
Hippocratic Corpus, Joints 52 (trans. E. T. Withington): τοῖσι μὲν οὖν τετελειωμένοισιν εἴρηται οἵη
τις ἡ ὁδοιπορίη γίνεται· οἷσι δ᾿ ἂν νηπίοισιν ἐοῦσιν ἡ συμφορὴ αὕτη γένηται, οἱ μὲν πλεῖστοι
καταβλακεύουσι τὴν διόρθωσιν τοῦ σώματος, ἀλλὰ [κακῶς] εἰλέονται ἐπὶ τὸ ὑγιὲς σκέλος, τῇ χειρὶ
πρὸς τὴν γῆν ἀπερειδόμενοι τῇ κατὰ τὸ ὑγιὲς σκέλος. καταβλακεύουσι δὲ ἔνιοι τὴν ἐς ὀρθὸν
ὁδοιπορίην καὶ οἷσιν ἂν τετελειωμένοισι αὕτη ἡ συμφορὴ γένηται. ὁπόσοι δ᾿ ἂν νήπιοι ἐόντες
ταύτῃ τῇ συμφορῇ χρησάμενοι ὀρθῶς παιδαγωγηθέωσι, τῷ μὲν ὑγιέϊ σκέλει χρέονται ἐς ὀρθόν,
ὑπὸ δὲ τὴν μασχάλην τὴν κατὰ τὸ ὑγιὲς σκέλος σκίπωνα περιφέρουσι, μετεξέτεροι δὲ καὶ ὑπ᾿
ἀμφοτέρας τὰς χεῖρας· τὸ δὲ σιναρὸν σκέλος μετέωρον ἔχουσι, καὶ τοσούτῳ ῥηΐους εἰσίν, ὅσῳ ἂν
αὐτοῖσιν ἔλασσον τὸ σκέλος τὸ σιναρὸν ᾖ.
30
Hippocratic Corpus, Joints 58. 31 Hippocratic Corpus, Instruments of Reduction 23.

speed, the winged shoes can be viewed as akin to corrective footwear. Since
Hephaestus is described as making shoes for the gods, it could have been
intended that he made them.32 The winged shoes have a counterpart in the
corrective footwear that was recommended for the treatment of congenital
conditions such as clubfoot, Talipes equinovarus.33 The author of Joints
advised treatment as early as possible, dressing and bandaging the foot in
a very particular way and then adding a sole made from a firm substance
such as stiff leather or lead.34 It is recognised that there will be variations in
the condition and that the dressings and bandages should be likewise
varied.35 If the manual adjustment, dressing, and bandaging are not suffi-
cient, one can go a step further and utilise corrective footwear. The author
refers to contemporary styles of footwear:
A leaden shoe shaped as the Chian boots used to be might be made, and
fastened on outside the dressing; but this is quite unnecessary if the manual
adjustment, the dressing with bandages, and the contrivance for drawing up
are properly done. This then is the treatment, and there is no need for
incision, cautery, or complicated methods; for such cases yield to treatment
more rapidly than one would think. Still, time is required for complete
success, till the part has acquired growth in its proper position. When the
time has come for footwear, the most suitable are the so-called ‘mud-shoes’,
for this kind of boot yields least to the foot; indeed, the foot rather yields to
it. The Cretan form of footwear is also suitable.36
The treatment recommended here is notable because it is neither surgical
nor mechanical. Potentially, a treatment such as this – which admittedly
might initially cause the patient a degree of discomfort but which they
would hopefully grow accustomed to over the extended period of time
during which they were undertaking it – was considered preferable to
32
For Hephaestus as a shoemaker, see Hyginus, Fables 166.
33
On congenital deformities of the legs and feet, see Roberts and Manchester 2010, 57‒9; for an
example of clubfoot that was left untreated from a Romano-British cemetery, see Roberts, Knusel,
and Race 2004. The modern incidence of Talipes equinovarus is 1 in 800 to 1,000 births; it is more
common in males and runs in families. See above, p. 000, for the theory that impairment and
disability were inherited. For depictions in ancient art, see Grmek and Gourevitch 1998, 151f., 282‒
287; Ziskowski 2012.
34
Hippocratic Corpus, Joints 62. For commentary, see Michler 1963.
35
Hippocratic Corpus, Joints 62; see also Hippocratic Corpus, Instruments of Reduction 32.
36
Hippocratic Corpus, Joints 62 (trans. E. T. Withington): οἷον αἱ Χῖαι [κρηπῖδες] ῥυθμὸν εἶχον· ἀλλ᾿
οὐδὲν αὐτοῦ δεῖ, ἤν τις ὀρθῶς μὲν τῇσι χερσὶ διορθώσῃ, ὀρθῶς δὲ τοῖσιν ὀθονίοισιν ἐπιδέῃ, ὀρθῶς
δὲ καὶ τὰς ἀναλήψιας ποιοῖτο. ἡ μὲν οὖν ἴησις αὕτη, καὶ οὔτε τομῆς οὔτε καύσιος οὐδὲν δεῖ, οὔτ᾿
ἄλλης ποικιλίης· θᾶσσον γὰρ ἐνακούει τὰ τοιαῦτα τῆς ἰητρείης ἢ ὡς ἄν τις οἴοιτο. προσνικᾶν μέντοι
χρὴ τῷ χρόνῳ, ἕως ἂν αὐξηθῇ τὸ σῶμα ἐν τοῖσι δικαίοισι σχήμασιν. ὅταν δὲ ἐς ὑποδήματος λόγον
ἴῃ, ἀρβύλαι ἐπιτηδειόταται αἱ πηλοπατίδες καλεόμεναι· τοῦτο γὰρ ὑποδημάτων ἥκιστα κρατεῖται
ὑπὸ τοῦ ποδός, ἀλλὰ κρατεῖ μᾶλλον· ἐπιτήδειος δὲ καὶ ὁ Κρητικὸς τρόπος τῶν ὑποδημάτων.

56 jane draycott
a surgical or mechanical intervention which would definitely cause the
patient both a high degree of pain during and immediately after the
procedure and an albeit (hopefully) decreasing degree of discomfort over
the extended period of time during which they were recovering and
rehabilitating from it. This minimally invasive approach is not surprising,
considering the author’s criticism of those who attempt to cure incurable
conditions such as hunchback with unnecessary surgical and mechanical
interventions, in part out of a desire to perform in front of an audience of
rubberneckers, elsewhere in the text.37
According to Erotian, a Chian boot was a woman’s boot, and according
to Galen, a Cretan boot reached halfway up the leg, so in both cases the
foot would benefit from a considerable amount of support from the ankle
and lower leg.38 There are several ancient literary references to individuals
with impaired feet utilising corrective footwear. The first is a music master
named Damonidas, who lost a pair of boots that were specially made for his
impaired feet and, depending upon how you interpret his response, either
rather generously or rather snidely prayed that they might fit the feet of the
thief.39 The second is a musician named Dorion, who lost his specially
made shoe at a party, which led him to curse the thief that the shoe might
come to fit him or, potentially, her.40 It is probable that corrective footwear
was made by a shoemaker but one working according under the supervi-
sion of, or at least according to instructions given by, a physician. Several
examples of corrective footwear have been found amongst the 4,000 shoes
dating to between the first and fourth centuries ce that have been exca-
vated from the site of the Roman fort at Vindolanda.41 To correct an over-
supinated or over-pronated gait, it would appear that the shoemaker
attached a metal bar to the sole of the shoe on the side of the foot that
required extra support at the point at which the shoe was originally made.42
To provide additional support to a particular part of the foot, clusters of
hobnails were somewhat randomly inserted into the corresponding part of
the sole, but this time not necessarily at the point at which the shoe was
originally made and, considering the lack of finesse with which this was
done, probably not by the shoemaker.43
One step beyond employing corrective footwear as a means of treating
an impaired foot is employing an extremity prosthesis as a means of
37
Hippocratic Corpus, Joints 42.
38
Withington 1928, 351, n. 1, n. 2 referring to Erotian, Hippocratic Lexicon and Galen, Commentary on
Hippocrates’ On Joints.
39 40 41
Plutarch, Moralia 18d. Athenaeus, Dinner Sophists 8.338a. Greene 2019.
42 43
See for example Vindolanda L-1992–3745. See for example Vindolanda L-1988–2118.

replacing a lost foot, and doing so would potentially inspire a considerable
amount of awe. There is ancient literary, archaeological, and bioarchaeo-
logical evidence for the use of extremity prostheses in the form of toes, feet,
and legs in antiquity.44 The earliest mention of a prosthesis in classical
literature that can be classed as historical and attributed to a genuine
historical figure, rather than a mythological one, is the wooden foot of
Hegesistratus of Elis, which dates to the early fifth century bce.45
Hegesistratus, the most distinguished soothsayer from amongst the
Telliads, a Greek clan renowned for their knowledge of prophecy, was
captured and imprisoned by the Spartans. However, in an attempt to avoid
torture and execution at Spartan hands, he managed to amputate enough
of his foot to enable him to remove his shackles, then broke through the
wall of his cell and escaped. Once his wound healed, he acquired a wooden
foot and continued to work against the Spartans, actions which led to his
subsequent recapture and execution after the battle of Plataea.46 There are
two prosthetic great toes from Egypt that have been dated to the first half of
the first millennium bce that support the essentials of this story; both
prostheses exhibit wear on the base, indicating that they were worn during
life and utilised as assistive technology; experimental archaeological recon-
struction and experimentation has proved that both are quite comfortable
to wear either with or without sandals.47
Much more elaborate and impressive than the Egyptian prosthetic toes
is the famous ‘Capua Limb’, a prosthetic right leg recovered from a tomb in
Capua that can be dated to approximately 300 bce. It is one of the oldest-
known functional prosthetic limbs in the world. It was originally made
from wood and bronze, consisting of a wooden core covered in bronze
sheeting worn in conjunction with a leather and bronze belt to hold it in
place and, assuming that the prosthesis could be securely fastened at the
thigh and the waist, to facilitate a limited amount of movement in
conjunction with a crutch. Other finds recovered from the tomb were
a bronze urn and some locally produced red-figure pottery, and judging by
these and the materials used in the limb’s construction, it was likely worn
by a high-status individual, or at the very least, a wealthy one, perhaps
a veteran of the Second Samnite War (327–304 bce) or even a retired
gladiator. Capua is noted in ancient literature as a city of considerable
44
See Bliquez 1996; Draycott 2018a.
45
Herodotus, Histories 9.37; some details of this episode are also included in Plutarch, On Brotherly
Love 3.1.
46
Herodotus, Histories 9.38. For discussion of Hegesistratus, see Dillery 2005. 47 Finch 2018.

58 jane draycott
wealth and luxury and is particularly feted for its bronze, so it would have
been the perfect setting for an individual to wear such an item.
It is not until Late Antiquity and the early Middle Ages that we see more
archaeological and bioarchaeological evidence for the use of extremity pros-
theses from sites in northern Europe. A skeleton of a man missing a foot and
bearing in its place a prosthesis comprising a leather pouch with a wooden sole
attached to it by iron nails has been excavated at Bonaduz in Switzerland and
dated to between the fifth and seventh centuries ce. The pouch was filled with
hay and moss, presumably intended to cushion the stump but possibly also to
soak up pus from the wound, since there is bioarchaeological evidence of
minimal healing having taken place and, in any case, the individual lived for
a maximum of two years after the amputation. It is debatable whether this
prosthesis was functional in the sense of allowing the wearer to walk around,
since there seems to have been no means of attaching it to the ankle. A skeleton
of a man aged thirty-five to fifty years old missing his lower left leg and bearing
a wood and metal prosthesis in its place has recently been excavated from
a Frankish settlement at Hemmaberg in southern Austria and dated to the
sixth century ce. Bioarchaeological evidence of osteoarthritis in the knees and
shoulders indicates that he used the prosthesis in conjunction with a crutch.
A skeleton of a man aged fifty-seven to sixty-three missing his lower left leg
below the knee and bearing the remains of a wood and bronze prosthesis,
probably a wooden peg-leg tipped with bronze, in its place has been excavated
from a Frankish cemetery at Griesheim near Darmstadt in Germany from
a site dating to the seventh or eighth century ce. The skeleton’s left femur was
atrophied, indicating that the man had survived for a considerable time after
the amputation but had only restricted movement.
Is it fair to say that the use of assistive technology such as a staff or
specially made shoe or an extremity prosthesis renders an individual
a cyborg? The myth of Oedipus and the riddle of the Sphinx shows that
it was not unusual for a staff to be considered and described as a third foot
or leg, while an individual using corrective footwear or extremity pros-
theses sought to supplement or substitute their own feet or legs.48
Hybrids: Conveyances; Equids; Bearers

Hephaestus built a number of chariots, including one with wings for
himself that he is often depicted riding in on vase paintings.49 The extent
48
Pseudo-Apollodorus, Library 3.52–5; Diodorus Siculus, Historical Library 4.64.4.
49
See for example an Attic red-figure kylix, Berlin inv. F2273, Beazley 201595.

to which physically impaired individuals utilised assistive technology akin
to a wheelchair is unknown; there are a few depictions in ancient literature
and art of young children utilising wheeled walking frames similar to
modern Zimmer frames (ὁ δίφρος ὑπότροχος in Greek, sustentaculum in
Latin), but none of adults doing something equivalent.50 There are rather
more depictions of immobilised individuals being carried around in chairs
and litters, at least for journeys of short duration; one example of this is
Artemon, an immobilised siege-engine designer who was carried around in
a litter during the siege of Athens. Longer journeys were likely carried out
with the assistance of equids such as donkeys or mules.51 Here we should
bear in mind the mythological episode of the return of Hephaestus to
Mount Olympus, a popular motif on Greek black- and red-figure vases in
the sixth and fifth centuries bce, where Hephaestus is frequently depicted
riding a donkey, the explanation given for which being his impairment.52
This choice has raised questions: while it makes sense for Hephaestus to be
depicted upon a donkey when he is depicted in the company of other
Olympian deities who are on horses or in horse-drawn chariots, an indica-
tion of his lower status, why is he depicted on a donkey when away from
Mount Olympus and the other Olympians?53
It is debatable how much agency an immobilised individual had in classical
antiquity; someone who used a chair, or a litter, was reliant upon the
cooperation of their bearers, while someone who used an equid was restricted
to using that equid where equids were permitted or where access allowed.
Automata
Finally, the assistive technology utilised by Hephaestus that has received
the most attention from scholars to date are his automata. Hephaestus is
described as fashioning a range of automata, humanoid, animal, tripod in
form, all of which served and assisted either gods or mortals in some way.54
50
Soranus, Gynaecology 1.114.
51
For the example of Artemon, see Plutarch, Pericles 27.3–4. See Griffith 2006, 324, for discussion of
the ‘peculiarly close relationship’ between an equid and its rider.
52
Hedreen 2004; Brennan 2016. However, see also MacDonald 2015, 185–8, on the possibility of
viewing Hephaestus as a lame sinner.
53
Hedreen 2004 has suggested that this is due to the artists drawing on the real Dionysiac processions
that took place in classical Athens for inspiration, as it would make more sense for Hephaestus to
travel on or in one of his own creations, such as the winged chariot or winged shoes mentioned
above. See Griffith 2006, 348–51, on this.
54
See Faraone 1987 for the suggestion that the inspiration for these stories originated in the ancient
Near East.

60 jane draycott
The ones that concern us here are the golden maidens that he fashioned to
serve and assist him in his workshop.55 Whether such things ever existed is
beside the point; what is more intriguing is the fact that they were
imagined, and that alongside the automata with loftier purposes,
Hephaestus is imagined as creating for himself assistive technology.56 It
has been suggested that there is as specific reason why this assistive
technology was envisaged in human form; according to Paipetis, ‘the
greater diversity of jobs is required from a robot, the more its form tends
to become human, for a simple reason: The human body, through its age-
long adaptive evolution within terrestrial environments, is the perfect
“universal tool”, therefore, if robots are to substitute for humans in their
activities, they must assume their form’.57 Additionally, according to
Kalligeropoulos and Vasileiadou, ‘two mythical robots, two self-moving
manlike machines, having sense, speech and strength. Innovative techno-
logical visions: The strength, i.e. the feature that transforms low-power
commands into powerful mechanical movements, the speech, i.e. the
construction of machines producing sounds to communicate, and the
sense, i.e. the particular inner structure that results in skilful, learning
machines’.58 All three of these have a resonance in relation to the require-
ments of assistive technology for an impaired individual: strength is
necessary for physical assistance, speech for interaction and companion-
ship, sense for anticipation and empathy. Whether or not the poet and the
audience believed that Hephaestus had created automata/robots/androids
to assist him is not the point; the point is that it seemed to be a reasonable
supposition, that someone with his physical impairments would benefit
from such assistive technology. Aristotle uses this very episode as a starting
point to consider and debate the positive and negative aspects of artificial
intelligence as it pertains to slavery.59 In reality, it is likely that physically
impaired individuals sought the assistance of their family, friends, and
household staff (whether freeborn, formerly enslaved, or enslaved).60 In
some cases, this link between assistive technology and human assistance is
made explicit, as the individuals rendering this assistance are described in
those very terms, even as assistive technology personified.61 According to
55
Homer, Iliad 18.418–21; Philostratus, Apollonios of Tyana 6.11.18–19.
56
Kalligeropoulos and Vasileiadou 2008, 77; Paipetis 2010, 111. See also Berryman 2009, 24–7.
57
Paipetis 2010, 108. 58 Kalligeropoulos and Vasileiadou 2008, 79.
59
Aristotle, Politics 1253b35; for discussion, see Devecka 2013. See also LaGrandeur 2011.
60
For discussion of the crucial role that human caregivers played in the lives of the physically impaired
and immobile, see van Schaik 2018.
61
See for example Sophocles, Oedipus at Colonus 867; Euripides, Hecuba 65, 261. See also Greek
Anthology 9.13b.

Macrobius, the reason that the Cornelii Scipiones bore that particular
cognomen was that one of their family members ‘used to guide his blind
homonymous father, in place of a cane’.62
The Close Association of Impairment and Technology

There seems to have been a strong connection between impairment and
technology, and technology and impairment in the minds of Greeks and
Romans. Attempts to explain Hephaestus’ impairment have included
seeing his lameness as compensation for his technological talents (and
vice versa), seeing it as a visible symbol of his wisdom and intelligence,
seeing it as an indicator of fire needing to be crippled in order to be
controlled.63 Hephaestus is, after all, not the only example of a god of
craft in world mythologies who is physically impaired.64 It is also possible
that Hephaestus was depicted as impaired because impaired individuals
actually did have a tendency to undertake trades.65 According to the author
of the Hippocratic treatise Joints, the Amazons deliberately dislocated the
joints of their male offspring and set them to work as artisans.66
However, more commonly stated was the belief that those who under-
took trades would become impaired, as a combination of the sedentary
nature of the occupation and the repetitive physical activity it required
would deform the body.67 Certainly, repetitive exertion would lead to
disproportionate muscle development and build-up of callus. Additionally,
due to the dangerous conditions found in ancient workshops, it is likely
that individuals who worked in them were frequently scarred. It has also
been suggested that impairments could have arisen from arsenical neuritis,
poisoning due to the high concentrations of arsenic in metal being smelted
and worked.68
62
Macrobius, Saturnalia 1.26 (trans. R. A. Kaster): qui cognominem patrem luminibus carentem pro
baculo regebat.
63
Delcourt 1957, 121–8, views Hephaestus as a magician; Detienne and Vernant 1974, 272; Faraone
1992, 134. However, Bremmer 2010, 200, sees Hephaestus’ impairment as symbolic and not
a reflection of ancient reality.
64
For a survey, see Brandon 2004. 65 Garland 1995, 62, 32–5.
66
Hippocratic Corpus, Joints 53 ; see Mayor 2014, 156 on this.
67
See for example Xenophon, Estate-Management 6.2; Aristotle, Household-Management 1.2.3;
Aristotle, Politics 1.4.3–4, 8.2.1; Lucian, The Dream 6–13. It is worth noting that the funerary
monuments of artisans do not tend to depict them as suffering from impairments; see for example
the stele of the shoemaker Gaius Julius Helius, whose funerary monument is dominated by
a carefully rendered portrait of his nude upper body rather than an image of him participating in
his trade, see George 2006, 27–8, for analysis and discussion.
68
Rosner 1955, 362–3; see more recently Craddock 1976, 1977; Nriagu 1983; Harper 1987.

62 jane draycott
Whether impaired individuals were likely to undertake trades or those
undertaking trades were likely to become impaired, we do need to consider
the possibility that physically impaired artisans utilised their experiences of
being physically impaired to inspire and inform their work. After all, who
better to understand the requirements of an impaired patron, especially if
they were attempting to commission something out of the ordinary such as
a prosthetic limb, than their impaired client?
Just as assistive technology augmented, supplemented, and even trans-
formed the human body, so too did it augment, supplement, and on
occasion even transform the practice of medicine. It offered physicians
the opportunity to expand their horizons and initiate dialogues and col-
laborations with the practitioners of the τέχναι of leatherworking, wood-
working, and metalworking, amongst others.
Assistive Technology, Technê, and the History of Technology

Ancient philosophers such as Plato, Aristotle, and the Stoics all depicted
nature as operating in a manner characteristic of technê.69 Erasistratus
seems to have frequently likened nature to artisans, and the works of nature
to the works of artisans.70 Assistive technology, although underrepresented
in ancient technological treatises, accords with the nature/art dichotomy,
with the impaired individual utilising art to overcome their nature and
bring about what nature cannot.71 In the particular examples discussed and
analysed in this chapter, what nature cannot bring about is physical
mobility. Additionally, if one of the points of ancient technology such as
automata was to inspire ‘wonder’ in the viewer, might it have been equally
wonderful, in the eyes of an ancient viewer, to see an impaired individual
apparently restored to health?72
Conclusion
As I stated at the outset of this chapter, relatively little attention has been
paid to the different types of assistive technology that could or would have
been utilised by individuals in classical antiquity with impairments that
69
von Staden 2007, 38. 70 On Erasistratus, see von Staden 2007, 38.
71
Pseudo-Aristotle, Mechanical Problems 847a11–12: art imitates nature and brings to fruition what
nature cannot? Aristotle, Nature 199 b 28–9: art goes beyond nature and brings about results that
nature cannot? For discussion of this dichotomy/polarity, see Schiefsky 2007. See however Micheli
1995, 64; van Leeuwen 2016, 12–18.
72
On the miraculous healing of those with mobility impairments, see van Schaik 2018, 145–7.

affected their physical mobility. Assumptions have been made regarding
the capabilities (or rather lack of capability) of ancient technology in
relation to impairment and disability.73 While ancient assistive technology
was clearly not as advanced in actuality as it was in the poetic imagination,
all incarnations served the same purpose: art was used to improve upon
nature and ‘fix’ something was ‘broken’ or ‘replace’ something that had
been ‘lost’. The differences between actual assistive technology – wood,
leather, bronze, iron, litters, chairs, equids, people – and imagined assistive
technology – winged shoes, winged chariots, automata – are less significant
than they first appear.
Studying assistive technology in classical antiquity is also worthwhile
from the perspective of facilitating efforts to apply the theoretical frame-
works of the discipline of Disability Studies to the disciplines of Classics,
Ancient History, and Classical Archaeology. In the twenty years since
Robert Garland published the first edition of his The Eye of the Beholder:
Deformity and Disability in the Graeco-Roman World, impairment and
disability has become a popular topic in the disciplines of ancient history
and classical archaeology, and the theoretical approaches of Disability
Studies have been brought to bear on this material, with varying degrees
of success, as neither the Medical Model nor the Social Model are entirely
appropriate for dealing with impairment and disability in classical
antiquity. Martha Lynn Edwards (later Martha Lynn Rose) has proposed
that what she designates a ‘Community Model’ be utilised instead, arguing
that the impaired and disabled were thoroughly integrated into ancient
society, and the contribution that they made to their particular community
was what was important.74 Sarah Newman has elaborated upon this with
what she designates a ‘Civic Model’, in which the impaired and disabled
were viewed in the context of citizenship and belonging and thereby
rendered impaired and disabled if they were unable to function physically,
mentally, or morally within their community.75
However, it is with the Medical Model that I would like to conclude.
The Medical Model sees disability as a personal limitation arising from the
impairment that is part of the individual’s constitution, with the necessary
response being medical intervention, treatment, and cure by medical or
technological means. Resources are targeted at the individual in order to
‘fix’ them and render them able to participate more fully in the world
73
See for example Rose’s opinion of the technological capabilities of ancient extremity prostheses: ‘It is
difficult to believe that any prosthetic device would have been practical as well as cosmetic’, at Rose
2003 (2013) 26. Obviously, this has been proved incorrect by Finch (2018) discussed above, p. 000.
74
Edwards 1997. 75 Newman 2013, 12, 16.

64 jane draycott
around them. The onus is on the individual to do what is necessary to make
the effort to adjust and fit in and be ‘normal’. While the Medical Model
accords with the prevailing view of impairment and disability in some
historical periods, particularly more recent historical periods, it is not
considered to accord with impairment and disability in classical antiquity.
This is because medical intervention, treatment, and cure by medical
means was not the general rule in ancient cases, because ancient physicians
did not tend to treat what they perceived to be incurable conditions.76
Individuals who attempted to treat such conditions, such as hunchback or
dwarfism, were dismissed as quacks and charlatans.77 Individuals suffering
from conditions that physicians were unable or unwilling to treat are
believed to have resorted to other types of healing practice, such as the
religious or magical.78
Yet if we focus on medical intervention, treatment, and cure by techno-
logical means, as we have seen, it is clear from attestations in the earliest
surviving classical literature that when individuals were experiencing phys-
ical impairments such as lameness or the loss of a limb, they are depicted
utilising a variety of different types of technology to assist them in their
everyday activities, to ‘fix’ themselves, as it were. Relatively little attention
has been paid to the different types of assistive technology that could or
would have been utilised by individuals with impairments that affected
their physical mobility in antiquity, yet, as I have argued, paying closer
attention to Hephaestus can be informative and provide a starting point
from which to further examine not just physical impairment but also
assistive technology in classical antiquity. I have examined the use of
technology as a means of supplementing, augmenting, and even trans-
forming the impaired body, surveyed the mythological examples and
assessed the insights that they can offer into the ways in which the impaired
body was scrutinised and stigmatised, and investigated the extent to which
these insights can be applied to real examples of impaired individuals. It is
clear that impaired individuals in ancient Greece and Rome could utilise
technology as an aid or even a cure for their conditions, and it is clear that
under some circumstances and in some cases they did. In light of new
research into ancient technology, might it be time to reconsider the
76
See for example Hippocratic Corpus, Prognostic 1; Hippocratic Corpus, On the Art 8; Hippocratic
Corpus, Prorrhetic 2.8; Hippocratic Corpus, On the Sacred Disease 2. See von Staden 1990.
77
See for example Hippocratic Corpus, Joints 42, on attempting to straighten a spinal curvature with
a ladder.
78
See for example Wickkiser 2008 with specific reference to the growing cult of the Greek god
Asklepios during the classical period.

perceived inappropriateness of applying the Medical Model to impairment
and disability in classical antiquity?
RE F E RE N CE S
Abrams, J. Z. 1998. Judaism and Disability: Portrayals in Ancient Texts from the
Tanach through the Bavli (Washington, DC).
Acton, P. 2014. Poiesis: Manufacturing in Classical Athens (Oxford).
Avalos, H., S. Melcher, and J. Schipper, eds. 2007. This Abled Body: Rethinking
Disabilities in Biblical Studies (Atlanta).
Brandon, R. 2004. ‘Myth and Metallurgy: Some Cross-Cultural Reflections on
the Social Identity of Smiths’, in A. Andren, K. Jennbery, and C. Raudvere,
eds., Old Norse Religion in Long-Term Perspectives: Origins, Changes, and
Interactions (Lund), 99–103.
Berryman, S. 2009. The Mechanical Hypothesis in Ancient Natural Greek Philosophy
(Cambridge).
Bliquez, L. J. 1996. ‘Prosthetics in Classical Antiquity: Greek, Etruscan, and
Roman Prosthetics’, Aufstieg und Niedergang der römischen Welt, ii 37.3:
2640–76.
Breitwieser, R., ed. 2012. Behinderungen und Beeinträchtigungen/Disability and
Impairment in Antiquity (Oxford).
Bremmer, J. N. 2010. ‘Hephaestus Sweats or How to Construct an Ambivalent
God’, in J. N. Bremmer and A. Erskine, eds., The Gods of Ancient Greece:
Identities and Transformations (Edinburgh), 193–208.
Brennan, M. 2016. ‘Lame Hephaestus’, Annual of the British School at Athens, 111:
163–81.
Brommer, F. 1978. Hephaestus: Der Schmiedegott in der antiken Kunst (Mainz).
Brule, P. 2006. ‘Bâtons et bâton du mâle, adulte, citoyen’, in L. Bodiou, D. Frère,
and V. Mehl, eds., L’expression des corps: gestes, attitudes, regards dans l’icono-
graphie antique (Rennes), 75–83.
Capdeville, G. 1995. Volcanus: recherches comparistes sur les origins du culte de
Vulcain (Rome).
Clynes, M. E. and N. S. Kline. 1960. ‘Cyborgs and Space’, Astronautics,
September, 26–7, 74–6.
Cokayne, K. 2003. Experiencing Old Age in Ancient Rome (London).
Couvret, S. 1994. ‘L’homme au bâton: statique et statut dans la céramique attique’,
Metis, 9–10: 257–81.
Craddock, P. T. 1976. ‘The Composition of the Copper Alloys Used by the Greek,
Etruscan and Roman Civilizations 1: The Greeks before the Archaic Period’,
Journal of Archaeological Science, 3.2: 93–113.
Craddock, P. T. 1977. ‘The Composition of the Copper Alloys Used by the Greek,
Etruscan and Roman Civilizations 2. The Archaic, Classical and Hellenistic
Greeks’, Journal of Archaeological Science, 4.2: 103–23.
Dasen, V. 1993. Dwarfs in Ancient Egypt and Greece (Oxford).

66 jane draycott
De Ciantis, C. 2005. ‘The Return of Hephaestus: Reconstructing the
Fragmented Mythos of the Maker’. Unpublished PhD thesis, Pacifica
Graduate Institute. Available online: www.academia.edu/749603/The_Re
turn_of_Hephaestus_Reconstructing_the_Fragmented_Mythos_of_the_M
aker (accessed October 2018).
Delcourt, M. 1957. Héphaistos ou la legend du magician (Paris).
Detienne, M. and J.-P. Vernant. 1974. Les ruses de l’intelligence: la mètis des Grecs
(Paris).
Devecka, M. 2013. ‘Did the Ancient Greeks Believe in Their Robots?’, Cambridge
Classical Journal, 59: 52–69.
Dillery, J. 2005. ‘Chresmologues and Manteis: Independent Diviners and the
Problem of Authority’, in S. L. Jonston and P. T. Struck, eds., Mantikê:
Studies in Ancient Divination (Leiden), 167–232.
Draycott, J. 2015. ‘The Lived Experience of Disability in Antiquity: A Case Study
from Roman Egypt’, Greece & Rome, 62.2: 189–205.
Draycott, J. 2018a. Prostheses in Antiquity (London).
Draycott, J. 2018b. ‘Introduction’, in J. Draycott, ed., Prostheses in Antiquity
(London), 1–28.
Draycott, J. (in preparation). ‘Staff or Stick? Cane or Crutch? Mobility Aids in
Ancient Greece and Rome’, in I. Bonati, ed., Words of Medicine: Technical
Terminology in Material and Textual Evidence from the Greco-Roman World
(Berlin).
Edwards, M. L. 1997. ‘Constructions of Physical Disability in the Ancient
Greek World: The Community Concept’ in D. Mitchell and S. Snyder,
eds., The Body and Physical Difference: Discourses of Disability (Ann Arbor,
MI), 35–50.
Emery, P. B. 1999. ‘Old-Age Iconography in Archaic Greek Art’, Mediterranean
Archaeology, 12: 17–28.
Faraone, C. A. 1987. ‘Hephaestus the Magician and Near Eastern Parallels for
Alcinous’ Watchdogs’, Greek, Roman and Byzantine Studies, 28.3: 257–80.
Faraone, C. A. 1992. Talismans and Trojan Horses: Guardian Statues in Ancient
Greek Myth and Ritual (Oxford).
Finch, J. 2018. ‘The Complex Aspects of Experimental Archaeology: The Design
of Working Models of Two Ancient Egyptian Great Toe Prostheses’ in
J. Draycott, ed., Prostheses in Antiquity (London), 29–48.
Fishbane, S. 2007. Deviancy in Early Rabbinic Literature: A Collection of Socio-
Anthropological Essays (Leiden).
Garland, R. 1994. ‘The Mockery of the Deformed and the Disabled in Graeco-
Roman Culture’ in S. Jäkel and A. Timonen, eds., Laughter Down the
Centuries (Turku), 71–84.
Garland, R. 1995, reissued 2010. The Eye of the Beholder: Deformity and Disability in
the Graeco-Roman World (London).
George, M. 2006. ‘Social Identity and the Dignity of Work in Freedmen’s Reliefs’,
in E. D’Ambra and G. P. R. Metraux, eds., The Art of Citizens, Soldiers and
Freedmen in the Roman World (Oxford), 19–29.

Greene, E. M. 2019. ‘Metal Fittings on the Vindolanda Shoes: Footwear and
Evidence for Podiatric Knowledge in the Roman World’ in S. Pickup and
S. Waite, eds., Shoes, Slippers, and Sandals: Feet and Footwear in Classical
Antiquity (London), 328–42.
Griffith, M. 2006. ‘Horsepower and Donkeywork: Equids and the Ancient Greek
Imagination. Part Two’, Classical Philology, 101.4: 307–58.
Grmek, M. D. and D. Gourevitch. 1998. Les maladies dans l’art antique (Paris).
Halliwell, S. 2008. Greek Laughter: A Study of Cultural Psychology from Homer to
Early Christianity (Cambridge).
Harper, M. 1987. ‘Possible Toxic Metal Exposure of Prehistoric Bronze Workers’,
Occupational and Environmental Medicine, 44.10: 652–6.
Harris, W. V., ed. 2013. Mental Disorders in the Classical World (Leiden).
Hedreen, G. 2004. ‘The Return of Hephaestus, Dionysiac Ritual and the Creation
of a Visual Narrative’, Journal of Hellenic Studies, 124: 38–64.
Hermary, A. and A. Jacquemin. 1988. ‘Hephaestos’, Lexicon Iconographicum
Mythologiae Classicae iv.1: 627–54.
Hughes, J. 2010. ‘Dissecting the Classics Hybrid’, in K. Rebay-Salisbury,
M. L. S. Sorensen, and J. Hughes, eds., Body Parts and Bodies Whole:
Changing Relations and Meanings (Oxford), 101–10.
Kalligeropoulos, D. and S. Vasileiadou. 2008. ‘The Homeric Automata and Their
Implementation’, in S. Paipetis, ed., Science and Technology in the Homeric
Epics (Dordrecht), 77–84.
Kelley, N. 2007. ‘Deformity and Disability in Greece and Rome’, in H. Avalos,
S. Melcher, and J. Schipper, eds., This Abled Body: Rethinking Disabilities in
Biblical Studies (Atlanta), 31–45.
Krötzl, C., M. Mustakallio, and J. Kuuliala. 2015. Infirmity in Antiquity and the
Middle Ages: Social and Cultural Approaches to Health, Weakness and Care
(Farnham).
Laes, C. 2014. Beperkt? Gehandicapten in het Romeinse rijk (Leuven).
Laes, C., ed. 2017. Disability in Antiquity (London).
Laes, C. 2018. Disabilities and the Disabled in the Roman World: A Social and
Cultural History (Cambridge).
Laes, C., C. Goodey, and M. Rose. 2013. Disabilities in Roman Antiquity: Disparate
Bodies a Capite ad Calcem (Leiden).
LaGrandeur, K. 2011. ‘The Persistent Peril of the Artificial Slave’, Science Fiction
Studies, 38.2: 232–52.
Loebl, W. Y. and J. F. Nunn. 1997. ‘Staffs as Walking Aids in Ancient Egypt and
Palestine’, Journal of the Royal Society of Medicine, 90.8: 450–4.
MacDonald, D. R. 2015. The Gospels and Homer: Imitations of Greek Epic in Mark
and Luke-Acts (Lanham, MD).
Malten, L. 1912. ‘Hephaestus’, Jahrbuch des deutschen archäologischen Instituts, 27:
232–64.
Malten, L. 1913. ‘Hephaestus’, Realencyclopädie der classischen Altertumswissenschaft,
8: 311–66.

68 jane draycott
Mayor, A. 2014. The Amazons: Lives and Legends of Warrior Women across the
Ancient World (Princeton, NJ).
Micheli, G. 1995. Le origini del concetto di macchina (Florence).
Michler, M. 1963. ‘Die Krüppelleiden in “De morbo sacro” und “De articulis”’,
Sudhoffs Archiv, 45: 303–28.
Mitchell, A. 2013. ‘Disparate Bodies in Ancient Artefacts: The Function of
Caricature and Pathological Grotesques among Roman Terracotta
Figurines’, in C. Laes, C. F. Goodey, and M. L. Rose, eds., Disabilities in
Roman Antiquity: Disparate Bodies a Capite ad Calcem (Leiden), 275–97.
Mitchell, A. 2017. ‘The Hellenistic Turn in Bodily Representations: Venting
Anxiety in the Terracotta Figurines’, in C. Laes, ed., Disability in Antiquity
(London), 182–96.
Newman, S. 2013. Writing Disability: A Critical History (Boulder, CO).
Nriagu, J. O. 1983. ‘Occupational Exposure to Lead in Ancient Times’, Science of
the Total Environment, 31.2: 105–16.
Ohry, A. and E. Dolev. 1982. ‘Disabilities and Handicapped People in the Bible’,
Koroth, 8.5–6: 63–7.
Olyan, S. M. 2008. Disability in the Hebrew Bible: Interpreting Mental and Physical
Differences (Cambridge).
Paipetis, S. A. 2010. The Unknown Technology of Homer (Dordrecht).
Roberts, C. and K. Manchester. 2013. The Archaeology of Disease (Stroud).
Roberts, C., C. J. Knusel, and L. Race. 2004. ‘A Foot Deformity from a
Romano-British Cemetery at Gloucester, England, and the Current
Evidence for Talipes in Palaeopathology’, International Journal of
Osteoarchaeology, 14: 389–403.
Rose, M. 2003, reissued 2013. The Staff of Oedipus: Transforming Disability in
Ancient Greece (Ann Arbor, MI).
Rosner, E. 1955. ‘Die Lahmheit des Hephaestus’, Forschungen und Fortschritte, 29:
362–3.
Schiefsky, M. J. 2007. ‘Art and Nature in Ancient Mechanics’, in B. Bensaude-
Vincent and W. R. Newman, eds., The Artificial and the Natural: An Evolving
Polarity (Cambridge, MA), 67–108.
Trentin, L. 2015. The Hunchback in Hellenistic and Roman Art (London).
Trentin, L. 2017. ‘The “Other Romans”: Deformed Bodies in the Visual Arts of
Rome’, in C. Laes, ed., Disabilities in Antiquity (London), 233–47.
Van Leeuwen, J. 2016. The Aristotelian Mechanics: Text and Diagrams (Berlin).
Van Schaik, K. 2018. ‘Living Prostheses’, in J. Draycott, ed., Prostheses in Antiquity
(London), 140–58.
Von Staden, H. 1990. ‘Incurability and Hopelessness: The Hippocratic Corpus’,
in P. Potter, G. Maloney, and J. Désautels, eds., La maladie et les malades dans
la Collection hippocratique: actes du vie Colloque international hippocratique
(Québec, du 28 septembre au 3 octobre 1987) (Quebec), 75–112.
Von Staden, H. 2007. ‘Physis and Technē in Greek Medicine’, in B. Bensaude-

Wickkiser, B. 2008. Asklepios, Medicine, and the Politics of Healing in Fifth-Century
Greece (Baltimore, MD).
Withington, E. T. 1928. Hippocrates: On Wounds in the Head. In the Surgery. On
Fractures. On Joints. Mochlicon (Cambridge, MA).
Ziskowski, A. 2012. ‘Clubfeet and Kypselids: Contextualising Corinthian Padded
Dancers in the Archaic Period’, Annual of the British School at Athens, 111:
211–32.

chapter 3
Not Yet the Android

The Limits of Wonder in Ancient Automata
Isabel A. Ruffell
Whereas many of the papers in this book are investigating how far the
human body was regarded in mechanistic terms in antiquity, and are
accordingly focused on medical evidence, this chapter comes at the issue
from the opposite direction. I am asking not how far the human body was
conceived of as a machine, but rather to what extent the ancient Greeks
and Romans were minded to (or did) build machines in the image of
humans.
One of the reasons for investigating this is as a corrective to some of the
influence of modern popular culture on the history of technology. Ever
since I began working on Hero of Alexandria’s automata, everyone who had
not read his Automata1 (i.e. most people, even within the discipline)
assumed that Hero’s automata were some kind of robot, which is far
from the case.2 This misapprehension, however, raises the question of
what the role of the human body in ancient automata was and why there
was not more interest in representing self-powered, programmable versions
of the human body.
So I am going to approach this question in both historical and thematic
terms. I start by looking at the earliest appearances of automata in ancient
texts, where we are dealing above all with the idea of automation, including
(but not restricted to or even not especially) human forms. I then go on to
look at references to the earliest historical automata and their split between
the anthropomorphic (of sorts) and non-anthropomorphic. Then I will go
on to look at, primarily, the Hellenistic tradition of automata where we can
see them in more detail and have some hope of approaching how automata
1
The title of the treatise is first attested in Pappus, Coll. 8.2 (p. 1024 Hultsch). The manuscripts are
divided between On Automaton-Making (peri automatopoiêtikês) and On Elements of Automaton-
Making (peri automatopoiêtikôn). For full discussion of the titles, see Grillo 2018. All abbreviations in
this chapter are taken from the Oxford Classical Dictionary, 4th edition (https://oxfordre.com/clas
sics/page/3993).
2
This is not to deny that this is a useful hook for the topic, as notably in Mayor 2018.
70

Not Yet the Android 71
were built for actual performance. As I shall argue, the role of the human
body, or of animal bodies, is strictly limited, and those bodies that do
appear are themselves strictly limited in what they can do. Some of these
limitations are technical and follow from the restricted nature of avail-
able power sources and mechanisms, but others relate to the tradition of
wonder-working (thaumatopoiia) out of which the tradition of autom-
ata-making emerges and against which it is reacting. The wonder that is
sought has less to do with the ability to recreate or mimic the human
body and more to do with mechanical virtuosity, which is not the same
thing.
Finally, I will turn to water-clocks and the various displays that they
drove (in order to tell the time). These so-called parerga are closely related
to the tradition of automata, as recorded both in Hero’s Automata treatise
and in the Pneumatica treatises of both Hero and his (probably late-third-
century bce) predecessor, Philo.3 Vitruvius himself (9.8.5) traces them
back to Ctesibius and the earliest Hellenistic water-clocks, but our best
evidence for such parerga in their more complex forms is late or filtered
through the Arabic tradition. Such accounts do, however, provide some
way into understanding audience response to such parerga. I shall argue
that, in addition to technical reasons, the nature of ancient mimesis and
narrative was such as to render the mimicking of the human body in
a naturalistic way superfluous.
Fantasies of the Automated Human

It seems clear that the idea of creating a working version of the human (or
at least humanoid) body goes back to the very beginnings of Greek
literature, and who knows how much further than that.4 Appropriately
enough, it is the craftsman-god Hephaestus who is represented as having
a variety of automatic creatures and objects, some more obviously built
than others. Specifically, the human (and animal) versions are presented in
rather vaguer terms than those entities that are not representations of living
creatures and which are more obviously the products of Hephaestus’ craft.
3
A precise date for Philo is contested, but he may plausibly be located within a generation or two of
Ctesibius, as he discusses hearing reports of those who had witnessed a Ctesibian experimental
catapult design in Alexandria (Bel. p. 72Th). Vitruvius includes him in his list of Hellenistic forebears
at 7. praef. 14.
4
For a full treatment, see Mayor 2018 and also Draycott, this volume, Chapter 2. My use of this
material here is much more limited and, indeed, mechanistic.

72 isabel a. ruffell
The earliest humanoid forms which have some claim to being precursors
of automata are Hephaestus’ female servants, who are mentioned briefly in
book 18 of the Iliad.
ὑπὸ δ᾿ ἀμφίπολοι ῥώοντο ἄνακτι
χρύσειαι, ζωῇσι νεήνισιν εἰοικυῖαι.
τῇς ἐν μὲν νόος ἐστὶ μετὰ φρεσίν, ἐν δὲ καὶ αὐδὴ
καὶ σθένος, ἀθανάτων δὲ θεῶν ἄπο ἔργα ἴσασιν.
Servants supported their lord,
golden, like living young women.
In them there is intelligence in their minds, and a voice too
and strength, and they know their work thanks to the immortal gods.5
Homer, Iliad 18.417–20
Nothing is said about their origin, or how they were made, and it is not
clear, for example, whether they are mechanisms or whether they are, in
effect, statues which have been brought to life by some kind of divine
power.
Much the same can be said of Pandora in Hesiod, most fully in
the Works and Days (70–82), where Hephaestus’ role is in giving the
overall form to the woman, but her attributes come from all the
other gods. Both these constructions are said to ‘resemble’ human-
ity, using cognate forms: ζωῇσι νεήνισιν εἰοικυῖαι (Iliad 18.418);
παρθένῳ αἰδοίῃ ἴκελον Κρονίδεω διὰ βουλάς (‘like a modest girl,
by the will of Zeus’ Op. 71). We might compare later human-origins
stories where Prometheus is responsible for creating humanity in
a similar fashion in the fabular and later literary tradition (first
attested in Ovid), but we are much more in the territory of magic
than mechanics here.6
Much the same can be said for other humans and creatures in the
mythological tradition, for which the better evidence comes later. Thus
the best-known account of the bronze man Talos, guardian of Crete and
destroyed by Medea, comes in book 4 of Apollonius Rhodius’ Argonautica,
although it can be traced back much earlier.7
5
All translations are my own unless otherwise indicated.
6
For Prometheus as creator, see Ovid, Met. 1.82–8. It is a feature of the fable tradition: fab. 228 H&H.
It may underlie the version in Plato’s Protagoras 320d–323a where Prometheus and Epimetheus are
assigned the task of doling out attributes to living creatures, but Epimetheus runs out before he
reaches mankind, necessitating the theft of fire.
7
For iconography reaching back to the fifth century, see Robertson 1977 with bibliography; Mayor
2018, 7–32.

τοὺς δὲ Τάλως χάλκειος, ἀπὸ στιβαροῦ σκοπέλοιο
ῥηγνύμενος πέτρας, εἶργε χθονὶ πείσματ᾿ ἀνάψαι
Δικταίην ὅρμοιο κατερχομένους ἐπιωγήν.
τὸν μὲν χαλκείης μελιηγενέων ἀνθρώπων
ῥίζης λοιπὸν ἐόντα μετ᾿ ἀνδράσιν ἡμιθέοισιν
Εὐρώπῃ Κρονίδης νήσου πόρεν ἔμμεναι οὖρον,
τρὶς περὶ χαλκείοις Κρήτην ποσὶ δινεύοντα.
ἀλλ᾿ ἤτοι τὸ μὲν ἄλλο δέμας καὶ γυῖα τέτυκτο
χάλκεος ἠδ᾿ ἄρρηκτος, ὑπαὶ δέ οἱ ἔσκε τένοντος
σύριγξ αἱματόεσσα κατὰ σφυρόν· αὐτὰρ ὁ τήν γε
λεπτὸς ὑμὴν ζωῆς ἔχε πείρατα καὶ θανάτοιο.
But Talos, the man of bronze, kept breaking rocks
off the towering height and preventing them veering cables to land
as they were trying to reach the Diktaean harbour’s anchorage.
The last of the ash-born men of bronze
race still living among the heroes,
the son of Kronos gave him to Europa to watch over the island,
going around the island three times on his bronze feet.
But although in the rest of his body and limbs he had been made
of bronze and unbreakable, below the tendon he had
a pipe that carried blood through his ankle. A
thin covering secured it, the bounds of life and death.
Ap. Rhod., Argonautica 4.1638–48
Talos does not seem to be mechanical so much as a living creature,

last of a primeval race.8 Indeed, the point of the pipe in his ankle is to
let out the fluid that gives him life.9 Medea befuddles Talos with the
result that he scrapes his ankle on a rock, causes the fluid to be let out
and thereby is destroyed (4.1665–88). There may be a hypothesis here
about blood being critical for life, although that may be based on no
more than the observation that exsanguination causes death in
humans, but the dominant paradigm here is that of the divine:
notwithstanding the use of ‘bloody pipe’ (syrinx haimatoessa) as an
epic periphrasis for a vein, the fluid is later described specifically as
8
The meaning of μελιηγενής is opaque: ΣbT Hom., Il. 22.126–7a suggest that the term denotes the stones
that Deucalion threw and turned into men. A less well-attested version (Apollod., Bibl. 1.140) has Talos
being given as a gift to Minos by Hephaestus, which makes him much closer to the devices of Hephaestus
in Homer.
9
This is not unknown in science fiction: fans of the British series Blake’s 7 (Nation 1978–81)
will be familiar with cyborgs (‘mutoids’) that work on similar principles; and indeed the
android (synthetic human) in Aliens (Cameron 1986) seems to have a similar substance.

ichôr (4.159), the liquid that flows in the veins of the gods.10 Talos is
thus closer to magic than mechanics.11
The term syrinx is used in various medical contexts, but not in the sense
used here as a means to carry blood, and its generic qualities are being
exploited here, even if that may also include mechanical contexts.12 More
obviously biological is the membrane (hymên) that covers the vein, although
a less common meaning to refer to thin sheets of metal may be influencing its
use here.13 Neither mechanical nor biological models are perhaps the main
inspiration for Talos. The fact that he is bronze and anthropomorphic
inevitably suggests that the model for this sort of idea is not anything
mechanical but rather statuary. A similar metalworking context is suggested
by the comparison of Talos’ leaking fluid to molten lead (τηκομένῳ ἴκελος
μολίβῳ, 4.160). Indeed, arguably Pandora herself may have been suggested by
terracotta figurines. The analogy with (or origins in) statues is even clearer in
the story of Pygmalion, where, as told by Ovid, it is explicitly an ivory statue or
model that is brought to life (Met. 10.280–6). As the object becomes animate,
the yielding of her flesh is further compared to the moulding of wax. The
divine source of this animation is very clear: it is a gift from Venus (10.277–9).
As a rather creepy example of male entitlement, it has few peers, but the story
does not obviously reflect any mechanical interest in the human body.
In addition to the human creation stories, we may add stories of similarly
statuesque animals, such as, in one version, the Trojan horse, or Hephaestus’
dogs in the Odyssey.14 Neither are the best candidates for precursors of or
suggestions of real-world automata. For these we need to look at some others
of Hephaestus’ creations, namely his bronze tripods on wheels.
῾Ηφαίστου δ᾿ ἵκανε δόμον Θέτις ἀργυρόπεζα
ἄφθιτον ἀστερόεντα, μεταπρεπέ᾿ ἀθανάτοισι,
χάλκεον, ὅν ῥ᾿ αὐτὸς ποιήσατο κυλλοποδίων.
10
See, for example, Iliad 5.334–43; 416–17.
11
Part of the reluctance to see Mary Shelley’s Frankenstein, or the Modern Prometheus as a precursor of
science fiction has been the way in which it draws on these classical models of animation through
mysterious, divine or magical means, or, as in the case of Talos, a fluid. It has, however, been shown
that Frankenstein’s methods were a satire on contemporary speculative physics and biology: see
Butler 1994, 236.
12
In the Pneumatica, it is mainly used of pipes used to produce sound, but note 1.16 (a pipe fitted
around an axle), 2.3 (conveying air) and 2.18 (the pipe at the end of the syringe). For the central
(rectangular) tube in Aut.: see esp. 2.8–9, 19 (double syrinx) and 23.3. Less relevant, perhaps, is its use
in catapults (Hero, Bel. p. 86W).
13
For medical or biological uses, see LSJ q.v. A1; for ὑμήν as referring to thin hammered (or poured)
metal see Philo (of Alexandria), Quis rerum divinarum heres sit 217 (gold), Ath. 6.230d (silver).
14
Trojan horse: Arctinus fr. 2; Hephaestus’ dogs: Odyssey 7.91–4. On such anthropomorphised
animals, see Faraone (1987); on Hephaestus’ devices, see also Mayor 2018, 129–55.

τὸν δ᾿ εὗρ᾿ ἱδρώοντα ἑλισσόμενον περὶ φύσας
σπεύδοντα· τρίποδας γὰρ ἐείκοσι πάντας ἔτευχεν
ἑστάμεναι περὶ τοῖχον ἐυσταθέος μεγάροιο,
χρύσεα δέ σφ᾿ ὑπὸ κύκλα ἑκάστῳ πυθμένι θῆκεν,
ὄφρα οἱ αὐτόματοι θεῖον δυσαίατ᾿ ἀγῶνα
ἠδ᾿ αὖτις πρὸς δῶμα νεοίατο, θαῦμα ἰδέσθαι.
οἱ δ᾿ ἦ τοι τόσσον μὲν ἔχον τέλος οὔατα δ᾿ οὔ πω
δαιδάλεα προσέκειτο· τά ῥ᾿ ἤρτυε, κόπτε δὲ δεσμούς.
ὄφρ᾿ ὅ γε ταῦτα πονεῖτο ἰδυίῃσι πραπίδεσσι,
τόφρα οἱ ἐγγύθεν ἦλθε θεὰ Θέτις ἀργυρόπεζα.
Thetis, silver-shod, came to the house of Hephaestus
deathless amid the stars, conspicuous for the immortals,
made of bronze, which the club-footed one himself made.
And she found him sweating and moving around the bellows
intently; for he was making twenty tripods altogether
to stand around the wall of the well-built hall,
and for each one he was setting golden wheels under the base,
so that they might enter the gathering of the gods by themselves
and come back to his house, a wonder to behold.
That was their intended function, but their ears
finely wrought, were not yet set on them; he was making those,
and he forged the chains.
While he laboured at this with his ingenious wits.
the goddess Thetis, silver-shod, came near him.
Homer, Iliad 18.369–81
In this instance, there is a much more obvious sense of these being made
things, and the fact of their spontaneous action and will is a source of
wonder in the narrative. As self-powered mechanisms, to which can also be
added his bellows (18.468–73), these are much closer to automata as later
realised, even if there is no obvious place to locate their power source.
Clearly, the dream of automation continued, alongside the more prac-
tical reality, and it should be noted that not only did this dream apply to
individual creatures but to entire landscapes. That idea similarly goes back
to the earliest epic poetry, in this case Hesiod (in the Works and Days), but
is exploited throughout antiquity, not least in Old Comedy.15 The distinc-
tion between these fantasies and mechanical marvels was, however, clearly
understood.16
15
As has been much discussed: see, for example, Baldry 1953; Ruffell 2000; Pellegrino 2000.
16
See, notably, Procopius of Gaza, Horologium §§1–2 (§§2–6 = 22–57 Diels). ‘So these were a story for
me, and I thought it just talk, and Homer was showing away with his craft, talking excessively about
what never existed or came to be; but after seeing the artful devices of this modern Hephaestus, I am
stunned by these things, and I concede that the latter allow the truth of the former.’ ταῦτα μὲν οὖν

Figuring Early Automata and Pseudo-Automata
Not until the fourth century is there any hint that automata of any sort were
constructed. Perhaps the best known and most discussed are the references to
automata made by Aristotle.17 It is clear that many of the references that are
sometimes supposed to be to automata in the fourth-century philosophers
allude to puppets rather than automata, but that terminological confusion is
itself based on an ambiguity in our sources, where Aristotle uses the language
of puppets (neurospasta and cognates) to do duty, albeit with further clarifi-
cation, for a kind of anthropomorphic automaton.18 This device seems to
have limited articulation, at least in terms of the connection of limbs to the
body, but it is far from clear – from Aristotle’s description, from
the corruption of the text19 and from Aristotle’s use of it as an analogy for
the human body – how exactly it worked.
οὕτως μὲν οὖν ἐπὶ τὸ κινεῖσθαι καὶ πράττειν τὰ ζῷα ὁρμῶσι, τῆς μὲν
ἐσχάτης αἰτίας τοῦ κινεῖσθαι ὀρέξεως οὔσης, ταύτης δὲ γινομένης ἢ δι’
αἰσθήσεως ἢ διὰ φαντασίας καὶ νοήσεως. τῶν δ’ ὀρεγομένων πράττειν τὰ
μὲν δι’ ἐπιθυμίαν ἢ θυμὸν τὰ δὲ δι’ ὄρεξιν ἢ βούλησιν τὰ μὲν ποιοῦσι, τὰ δὲ
πράττουσιν. ὥσπερ δὲ τὰ αὐτόματα κινεῖται μικρᾶς κινήσεως γινομένης,
λυομένων τῶν στρεβλῶν καὶ †κρουόντων ἀλλήλας τὰς στρέβλας†, καὶ τὸ
ἁμάξιον, ὅπερ ὀχούμενον αὑτὸ κινεῖ εἰς εὐθύ, καὶ πάλιν κύκλῳ κινεῖται τῷ
ἀνίσους ἔχειν τοὺς τροχούς (ὁ γὰρ ἐλάττων ὥσπερ κέντρον γίνεται,
καθάπερ ἐν τοῖς κυλίνδροις), οὕτω καὶ τὰ ζῷα κινεῖται. ἔχει γὰρ ὄργανα
τοιαῦτα τήν τε τῶν νεύρων φύσιν καὶ τὴν τῶν ὀστῶν, τὰ μὲν ὡς ἐκεῖ τὰ
ξύλα καὶ ὁ σίδηρος, τὰ δὲ νεῦρα ὡς αἱ στρέβλαι· ὧν λυομένων καὶ ἀνιεμένων
κινοῦνται. ἐν μὲν οὖν τοῖς αὐτομάτοις καὶ τοῖς ἁμαξίοις οὐκ ἔστιν
ἀλλοίωσις, ἐπεὶ εἰ ἐγίνοντο ἐλάττους οἱ ἐντὸς τροχοὶ καὶ πάλιν μείζους,
κἂν κύκλῳ τὸ αὐτὸ ἐκινεῖτο· ἐν δὲ τῷ ζῴῳ δύναται τὸ αὐτὸ καὶ μεῖζον καὶ
ἔλαττον γίνεσθαι καὶ τὰ σχήματα μεταβάλλειν, αὐξανομένων τῶν μορίων
διὰ θερμότητα καὶ πάλιν συστελλομένων διὰ ψύξιν καὶ ἀλλοιουμένων.
μῦθός τε ἦν ἐμοὶ καὶ λόγος ἐδόκει, καὶ ῞Ομηρος ἐτρύφα τῇ τέχνῃ λέγων ἐπ’ ἀδείας ἃ μήτε ἦν μήτ’
ἐγένετο πώποτε· νυνὶ δὲ τοῦ παρόντος ῾Ηφαίστου ἔργα καὶ τέχνην ἰδὼν ὑπεράγαμαι ταῦτα
κἀκείνοις ἀληθέσιν εἶναι συγχωρεῖν ἐπιτρέπω (§2 (§6 = 48–57 Diels)).
17
See especially Cambiano 1994.
18
For neurospasta as (humanoid) puppets, see Plato, Laws 644d–e; Aristotle, de mundo 398b. See already,
metaphorically, Xenophon, Symp. 4.55, where performers are compared to puppets. The thaumata in
Plato’s Republic 514b are also clearly puppets rather than automata. Aristotle does refer to automata
within the (very broad) contemporary class of thaumata (GA 734b, 741b, Metaph. 983a), but, equally
clearly, not all thaumata are automata. Aristotle may well be responding to Plato, but their perspectives
are different: broadly epistemological (Plato on audience perception) or metaphysical (emotional or
psychological manipulation in Laws) rather than physical or mechanical (Aristotle’s interest in how
power and control is transferred). The choice of type of thaumata is clearly relevant. For automata and
mechanistic explanation in philosophical texts, not least Aristotle, see Berryman 2003, 2007, 2009.
19
The fullest discussion is Nussbaum 1976, 146–52; see also Nussbaum 1978.

So animals are prompted to move and act in this way, the furthest cause of
movement being an impulse (orexis), and this happening through percep-
tion or through imagination and thought. Of things that are impelled to act,
some through desire or anger, others through impulse act or achieve goals.
And just as automata move when a small initial movement happens, as cords
(streblai) are released and strike one another [?], and the model cart, which
goes in a straight line under its own power, and again moves in a circle by
means of having wheels of unequal size (for the smaller one acts, as it were, as
a pivot, just as with cylinders), so too do animals move. For such elements
have the nature of sinews and bones, the latter like the wood and iron there,
and the sinews like the cords: when they are released and let go, they move.
So in the case of automata and model carts (hamaxia), there is no alteration,
since if the internal wheels become smaller or bigger, the same thing still
moves in a circle; but in the case of animals, the same thing can become
bigger or smaller and change its patterns of movement (schêmata), by its
parts growing because of heat or contracting because of cold and by being
altered. Aristotle, de motu animalium 701a33–b16
For the history of automata, it is particularly regrettable that the key phrase
for the movement of the human puppet is corrupt. It does seem evident
that the limbs move by cords being released, and the passage also suggests
quite a violent movement (krouontôn, ‘striking’). The comparison of the
cords with human or animal sinews suggests a spring-like motion, but it
would be ill advised to push that comparison too far. Fortunately, however,
it is possible to make a little more progress by considering Hero’s general
remarks on the power sources of automata. Although Hero is writing
around four centuries after Aristotle, he was undoubtedly drawing on
earlier sources, including the third-century bce mechanical writers
Ctesibius and Philo.20 He claims that there are two power sources for
automata: a falling weight and a spring (hysplênx). The former powers all
the devices in the Automata. The latter is slightly mysterious, but clearly
involved the elastic properties of sinew. The comparison that he makes
with the torsion engine for catapults clearly suggests a mechanism that
functions like a spring, and one may imagine a catch-and-release system
too (again comparing the catapult).
20
Hero is silent on the specific sources of book one of his Automata, while frequently positioning his
text against his predecessors, but in the second book, he explicitly draws on Philo (Aut. 20).
Ctesibius, the founder of the Hellenistic mechanical tradition, is also said by Vitruvius (9.8.4) to
have written on automata and other entertainments (deliciae) and is the only other writer on the
topic in that tradition that is currently known.

νευρίνῳ δὲ οὐδενὶ δεῖ χρῆσθαι, ἐπειδὴ παρεκτείνεται καὶ συστέλλεται κατὰ
τὴν τοῦ ἀέρος περίστασιν, εἰ μὴ ἄρα ὅταν δέῃ ὕσπληγγι χρήσασθαι. ὁ δὲ
ὕσπληγξ ἔστω καθάπερ ἐν τοῖς καταπέλταις ὁ ἄξων κατατεταγμένος ἐν τῷ
ἡμιτονίῳ, ὡς ἑξῆς ἔσται δῆλον. πάντα δὲ ταῦτα τὰ ὑπάγοντα τὴν ἀρχὴν
λαμβάνει τῆς κινήσεως διὰ ὕσπληγγος ἢ λείας μολιβῆς.
You must not use any sinew (as it expands and contracts according to the
atmospheric conditions), except when you need to use a spring (hysplênx):
the spring should be just like the axle arranged in the torsion engine in
catapults, as will be clear in what follows. All these mobile automata receive
their impulse by means of a spring or a lead weight. Hero, Aut. 2.6
Hero does not use this hysplênx anywhere in his treatise as it currently stands,
which makes this comment a curious one, unless it refers to historical
automata, or derives from his sources, or both. He uses a diminutive form
of the term, hysplêngion, to refer to a different piece of apparatus, a star-
shaped wheel that repeatedly flicks up and down a counterweighted and
pivoted arm, in order to produce a rocking motion.21 The similarity with the
effect of a spring being released is undoubtedly the reason for the application
of the term to the rocking arm (although the latter is of course more regular).
A spring release would also offer an explanation for the violent motion of
Aristotle’s humanoid automaton, and of course in this kind of power source
the cords are indeed being released rather than pulled.22 The striking motion
is however the first major corruption in the passage. It seems very difficult to
understand the sinews clashing, as in the paradosis,23 and it would make
much more sense to have the limbs clashing.24
What has been overlooked in this passage is that Aristotle is drawing on
two different types of automaton here, and both are used as analogies for
the movement of the human or animal body. As well as the (broadly)
humanoid puppets, there is also the small cart (hamaxion). Although usually
21
See p. 84 below.
22
The precise sense of streblê is unclear. It could mean twisted cords (e.g. hair or sinew, as in the torsion
engine), but that is not a directly attested use. It is apparently used of a winch in Aeschylus
(Suppl. 441 with Hesychius s.v. σ 1976), but it is most commonly associated with a rack-like torture
device (already in Ar., Frogs 860), also implying a winch (certainly stretching).
23
Although Michael of Ephesus suggests this interpretation (117.20–4): it may not be anything other
than an inference from the passage itself. More circumstantial is the idea, probably taken from
Philoponus (In GA 77.17), that automata were associated with marriages.
24
Either interpretation requires emendation. The obvious first step is to delete τὰς στρέβλας as an
intrusive gloss, and further emendation of κρουόντων accordingly would give grammar, if not sense.
Better is to suppose that a word for limbs, or similar, has been displaced: thus, for example, τῶν
σκελῶν or Nussbaum’s τῶν ξύλων is needed, with ἀλλήλας modified accordingly. Philoponus talks
of pieces of wood (xyla) transferring motion to each other ‘through some device’ (διὰ τινός
μηχανῆς), which admits of many different reconstructions.

understood as a toy cart which is pushed (or ridden), it is difficult to derive that
sense from the passage (which again, in any case, requires emendation).25
A minor alteration gives the correct sense of a self-powered cart.26 In fact, self-
powered toy carts appear to have been a part of the fourth-century landscape.
There is some evidence (albeit late and difficult) to associate them with Sicily
and with the court of Dionysius I (and II) of Syracuse, who is probably not
coincidentally also reputed to be responsible for deploying the first catapults.27
In what purports to be a letter from Speusippos to Dion, the Syracusans are
praised for giving up various innovations of the elder Dionysius, including the
practice of sending clever devices (sopha) to Delphi. Specifically mentioned is
Apollo’s supposed reaction to ‘seeing the cart running around in the hippo-
drome under its own power’ (τὸ ἁμάξιον ἰδὼν τὸ ἐν τῷ ἱπποδρόμῳ
περιτρέχον αὐτόματον).28 The distinction between straight-line and circular
motion also seems highly suggestive of Hero’s first two types of motion in his
account of the mobile automaton.29 Such self-powered toy carts would share
with the marionettes the problem of a power source. In the case of toy carts, it
is very difficult to imagine how a vertically descending weight could be used to
power them – where would it go? A coiled sinew spring makes much more
sense, as it would for the clashing of the arms, legs or cables in the case of the
marionettes. What this also means is that the ability to programme these early
automata was highly limited or non-existent: wind them up and let them go, as
it were.
Both the humanoid and vehicular types of automaton to which Aristotle
refers represent very early stages of the tradition of automata, and indeed
there seems to be no consistent terminology at this time. More broadly, they
fall under the very wide category of thaumatopoiia or wonder-working, from
which Hero, in particular, was keen to unpick the mechanical achievement
of himself and his predecessors. In connection to this, it is worth mentioning
25
Most manuscripts have ὅπερ ὀχούμενον αὐτο κινεῖ, Nussbaum emends ὅπερ ὀχουμένον to ὁ γὰρ
ὀχούμενος ‘the one who pushes’, but the verb cannot mean that.
26
Emending αὐτό to αὑτό, adopted in the quoted text above. Such use of κινέω with a reflexive is seen
in Aristotle’s discussion of how something can be autokinêtos in Physics viii, where he explains that if
the whole moves itself, there is part that moves and part that is moved (Physics 258a1–5). Galen’s
remarks on Hephaestus’ devices as autokinêta (de Usu Partium 3.268.11–16) seems to be looking back
to and modifying this passage of the Physics. An alternative strategy is to take ὀχουμένον as referring
to a model rider, moved by the cart, but this makes less sense in the context of the analogy between
the machine and the body.
27
Diod.Sic. 14.42–3
28
Epistuli Socratici 33 (35 Orelli; Hercher). This often baffling letter and its counterpart (34, from
Dionysius II to Speusippos) are fictional but draw on a variety of anecdotes about Dionysius I and
II: see Sykutris 1933, 92–101, and Bolzan 2009, 465 n. 49.
29
Hero, Aut. 5–8.

another simple device used by what we would call magicians, who belonged
to the wide category of thaumatourgoi or thaumatopoiioi, to suggest
a modification of the body: the use of bladders under the clothing to project
liquids, apparently from the mouth.
Διοπείθης δὲ ὁ Λοκρός, ὥς φησι Φανόδημος, παραγενόμενος εἰς Θήβας καὶ
ὑποζωννύμενος οἴνου κύστεις μεστὰς καὶ γάλακτος καὶ ταύτας ἀποθλίβων
ἀνιμᾶν ἔλεγεν ἐκ τοῦ στόματος. τοιαῦτα ποιῶν ηὐδοκίμει καὶ Νοήμων ὁ
ἠθολόγος. ἔνδοξοι δ’ ἦσαν καὶ παρ’ ᾿Αλεξάνδρῳ θαυματοποιοὶ Σκύμνος ὁ
Ταραντῖνος, Φιλιστίδης ὁ Συρακούσιος, ῾Ηράκλειτος ὁ Μιτυληναῖος.
γεγόνασι δὲ καὶ πλάνοι ἔνδοξοι, ὧν Κηφισόδωρος καὶ Πανταλέων.
Diopeithes of Locri, as Phanodemus says, came to Thebes and strapped on
bladders full of wine and milk, squeezed them and claimed that he was
drawing them out of his mouth. Noemon the êthologos was also famous for
doing such things. And there were also famous wonder-makers in the time of
Alexander: Scymnus from Taras, Philistides from Syracuse and Heraclitus
from Mytilene. There were also famous wandering showmen (planoi),
amongst whom were Cephisodorus and Pantaleon.
Athenaeus 1.35 p. 20a–b
This simple pneumatic device is used to give the impression of bodily
modification. In terms of the place of automata and their effects on their
audience, it is striking that such projection of liquids (and indeed the same
liquids: milk and wine) is, like motion, a feature of later, more developed
automata. Unlike these earlier thaumatopoioi, the projection was associated
less with the human body than with other implements: a concession,
perhaps, to realism.
Realism and Articulation

For more controlled and perhaps even realistic representations of human or
indeed animal bodies, we need to look beyond Aristotle. Unfortunately,
none of the reports of such devices are straightforward, and to my mind
some of them are deeply suspect.
Proceeding in roughly chronological order, the first example is an outlier in
many ways. Attributed to the philosopher and statesman Archytas of
Tarentum, the report (in Gellius via Favorinus) claims that the wooden
dove rose up in the air and descended (apparently in a controlled fashion)
before coming to rest.30 The combination of fluid and counterweight (libra-
menta) is known from Hero’s Pneumatica (e.g. Pneum. 1.16) and is not in itself
30
Gell. 10.12.8–9 = Archytas test. A10a Huffman.

surprising, but it is here air rather than liquid, the container is unspecified and
the precise function of the release of air is unclear: it could be to release weight
and take it out of equilibrium,31 but it might also simply have been a bladder
that is released to provide impulse. Either way, we do not seem to have any
articulation of limbs or wings, so it is a non-naturalistic movement that is
represented, however it was realised.32
Better known than Archytas’ dove is Demetrius of Phaleron’s giant
snail, which, according to Demochares, was part of a procession through
the Theatre of Dionysus, probably at the Dionysia of 308/7: ‘a snail,
moving by itself, led the procession, spitting out slime’ (κοχλίας
αὐτομάτως βαδίζων προηγεῖτο τῆς πομπῆς αὐτῷ, σίαλον ἀναπτύων).33
Here, questions abound: is the snail really spitting fluid or simply extrud-
ing it? How was it powered? And why a snail? The answer to the latter
question is the easiest: it is to avoid any kind of need for articulation –
which would have been really difficult to achieve in any kind of controlled
or sustained fashion. A snail is also handy because it hides whatever is
powering it, down to the floor level. I am increasingly sceptical that this
was a genuine automaton, however, even if it appeared that way to
spectators: it could just as easily – more easily in fact – hide people inside
pushing as it could house a plausibly sized counterweight or even, as in one
highly implausible reconstruction, an internal treadmill.34 The theatrical
context of this display is surely significant, but the element of realism in the
display is distinctly limited by the size, if by nothing else.
Slightly later in the Hellenistic period, there is the Nysa automaton that
was carried on a cart in Ptolemy II’s great procession at his self-titled
festival.35 In a segment of the parade devoted to Dionysiac elements, the
nymph (associated with a mountain range in Egypt or India) who nursed
31
So Schmidt 1904, the most influential, if over-precise, reconstruction; see also D–K, i.425, and
Huffman 2005, 570–9, with references.
32
It may well have been part of a larger tableau, as Schmidt 1904 suggests; for such bird tableaux,
compare Pneum. 1.16 and 2.4. For other flying birds, see Schmidt 1904, 347–51. It is unlikely to have
been free-flying, as Huffman 2005, 573, 577–8, notes.
33
Demochares, no friend of Demetrius, is quoted by Polybius 12.13.9–12 = Demochares fr. 4 FGrH / fr.
7 BNJ. For the date, see Athenaeus 12.60 p. 352e; compare Rehm 1937, 317.
34
Rehm 1937. Problems with this reconstruction include the lack of evidence for a belt drive or for
internal treadmills elsewhere. Rehm’s reconstruction is based on a highly dubious reconstruction/
interpretation by Schramm of a siege engine described by Biton, who is elsewhere explicit that the
helepolis is dragged into position (p. 56 Wescher, acknowledged at Rehm 1937, 321). For criticism of
the Rehm–Schramm hypothesis, see Marsden 1971, 89 and fig. 3, Lendle 1983, 38–58 at 49–53, and
Campbell 2003, 12–13.
35
For the procession and the festival, see especially Rice 1983 and also Keyser 2014 for commentary on
Callixenus’ description.

the infant Dionysus36 allegedly stood up, poured milk (compare Hero) and
sat down again, without anyone intervening.
μετὰ δὲ ταύτας ἤγετο τετράκυκλος πηχῶν ὀκτὼ πλάτος ὑπὸ ἀνδρῶν
ἑξήκοντα, ἐφ᾿ ἧς ἄγαλμα Νύσης ὀκτάπηχυ καθήμενον, ἐνδεδυκὸς μὲν
θάψινον χιτῶνα χρυσοποίκιλον, ἱμάτιον δὲ ἠμφίεστο Λακωνικόν.
ἀνίστατο δὲ τοῦτο μηχανικῶς οὐδενὸς τὰς χεῖρας προσάγοντος καὶ
σπεῖσαν ἐκ χρυσῆς φιάλης γάλα πάλιν ἐκάθητο. εἶχε δὲ ἐν τῇ ἀριστερᾷ
θύρσον ἐστεμμένον μίτραις. αὕτη δ᾿ ἐστεφάνωτο κισσίνῳ χρυσῷ καὶ
βότρυσι διαλίθοις πολυτελέσιν. εἶχε δὲ σκιάδα καὶ ἐπὶ τῶν γωνιῶν τῆς
τετρακύκλου κατεπεπήγεσαν λαμπάδες διάχρυσοι τέτταρες.
After these came a four-wheeled cart, eight cubits wide, drawn by sixty
men, on which there was a seated statue of Nysa, wearing a gold-
embroidered yellow-coloured tunic (chiton), and a Spartan cloak was
drawn around her. This stood up by means of a mechanism (mêchanikôs)
without anyone touching it with their hands and after pouring a libation
of milk from a bowl (phialê) it sat down again. It had in its left hand
a thyrsos wreathed with bands. She was crowned with ivy, gold and
grapes, cunningly worked from crystals. She had a parasol and on the
corners of the cart four golden lamps had been fixed.
Callixenus fr. 2 = Athenaeus 5.198e–f [5.25 Kaibel]
Among the many interesting features of this device is that, if it did what is
claimed, it must have been articulated in at least the knees and perhaps hips
and even shoulders. It is possible that the mechanism was automatic, that
is, self-powered, although that is not said explicitly. Michael Lewis confi-
dently claims this as an example of an early large-scale mechanical trip,
centuries ahead of its first archaeological attestation, where a cam on one of
the cart’s axles that worked on a lever raised the statue, before dropping it
again. Such repeated action could be achieved easily enough, however,
using ropes and axles in the manner generally favoured by Hero – and, it
would seem, Philo.37 I am not sure myself, however, that this was not
powered by a man hidden in the cart, hauling on a rope periodically. Either
way, this is very much a divine body, even if humanoid in appearance, and
the divine associations – and their reflections on Ptolemy – are clearly
central to the wonder being evoked in this performance.
36
Homer, Iliad 6.130–7. Egypt or environs: H.Hymn Dionysus 9–14; Hdt. 2.146.2 and 3.97.2; India:
Cleitarchus 137 F 17 FGrH, Strabo 15.1.7–9 and Arrian, Anabasis 5.1. Diodorus reports both (1.15.6–
7, 1.19.7; 3.64.5–6, 3.66.3). See also Rice 1983, 62–8, and Keyser 2014 ad loc., with bibliography.
37
Mechanical trip: Lewis 1997, and see further below, p. 84, for the smaller-scale device. For repeated
actions using axles, see, for example, the doors of Hero’s mechanical theatre, following Philo
(Aut. 23).

Nysa is clearly an extension of the idea of the statue, albeit slightly
articulated, rather than any attempt to mimic the human body in great
detail. A very similar case is attested in a Roman context, at what is described
as an excessively showy and un-Roman celebration of Q. Metellus Pius,
proconsul of Further Spain, apparently in 74 bce.38 A statue of victory is
alleged to have laid a wreath on his head, to the accompaniment of sound
effects: ‘as he was sitting down a statue of Victory was lowered by means of
a spring, accompanied with the sound of thunder created by a machine, and
it placed a crown on his head’ (sedenti transenna demissum Victoriae simulac-
rum cum machinato strepitu tonitruum coronam capiti imponebat). As
a sequence of events perhaps (but not necessarily) caused by the same device,
this might count as an automaton, even though it is not explicitly called that.
As with the other devices, a human agent may well have been more directly
involved in causing these effects. Again, this is very much occurring in
a divine context, and even more clearly than Nysa is a modification of
a basic statue: it is the arm as a machine rather than the body.
Similarly limited articulation occurs when we move into the territory of
machines that are clearly automata and described as being built (or as if
they were intended to be built),39 and where the mechanisms are more or
less plain. Thus Hero, apparently quoting Philo, describes the earliest
automata in the form of mini-theatres, the pinax type, opening to display
a face with animated eyes before the doors close and reopen and show
a depiction of a story.
Οἱ μὲν οὖν ἀρχαῖοι κέχρηνται ἁπλῇ τινι διαθέσει· ἀνοιχθέντος γὰρ τοῦ
πίνακος, ἐφαίνετο ἐν αὐτῷ πρόσωπον γεγραμμένον. τοῦτο δὲ τοὺς
ὀφθαλμοὺς ἐκίνει καμμύον τε καὶ ἀναβλέπον πολλάκις. ὅταν δὲ πάλιν
κλεισθεὶς ἀνοιχθῇ ὁ πίναξ, τὸ μὲν πρόσωπον οὐκέτι ἑωρᾶτο, ζῴδια δὲ
γεγραμμένα εἴς τινα μῦθον διεσκευασμένα.
So the ancients used a simple arrangement. For when the display opened,
a painted face appeared in it. This moved its eyes, closing and opening them
frequently. But when the box was closed and opened, the face was no longer
seen, but painted figures were arranged to form some story.
Hero, Aut. 22.1
The repeated opening and closing of eyes could easily have been
achieved by means of a horizontal rotating axle behind the cut-out
backdrop.
38
Sallust, fr. 2.70 (Macr., Sat. 3.13.7–9).
39
The question of whether such devices could have been built or were intended to be built is a complex
one and outside the scope of this chapter.

In Philo’s five-act drama, by contrast, we have figures not only painted on
the backdrop, or in fact a series of backdrops, but animated in various ways:
a horizontally scrolling backdrop (act 3), partial animation of a body part of
a figure otherwise painted on the backdrop (act 1), revealing and hiding
figures (act 5) and complete animation (acts 3 and 5). The Greeks repairing
their ships for departure are otherwise painted but partially animated by
means of having a moving right arm (Aut. 24). Made of horn, the arm is
fixed by a peg to a counterweighted rod behind the backdrop. A star-shaped
wheel turns and presses down one end of the rod and then releases it. As the
rod rotates around the pivot/shoulder joint, the right arm of each figure
appears to move up and down. This device for causing rapid chopping
motion is the hysplêngion (‘little spring’) (Aut. 24.4).40
There are two sets of figures that are fully independent of the backdrop.
One consists of the dolphins that appear in the foreground (Aut. 27) as the
Greek ships go sailing by on a scrolling backdrop (Aut. 26). These dolphins
are rendered by an appropriate shape attached to a rod itself connected at
right-angles to an axle beneath the display. As the axle turns the dolphins
emerge from the floor of the tableau (the sea) and then dive back down
again. Different, if rather opaque, rotating mechanisms are in play for the
appearance, movement and disappearance of Athena in the final scene
(Aut. 29). She is a counterweighted figure that is initially lying down,
pulled upright and laid back down again by separate cords. While upright,
she then rotates once by means of a further cord. The precise mechanisms
require interpretation. The salient information is that:
1. The device is criticised by Hero (Aut. 20.2), who suggests that a simple
hinge (ginglumos) could have been used instead and only two cords
used, one to raise the figure and one to lay it down;
2. The figure’s base has ‘pegs in the appropriate places’ (ἔχουσα ἐν τοῖς
προσήκουσι τόποις τύλους);
3. The two separate cords that pull the figure up and then down are both
said to come ‘from behind the (hip) joint’ (ἐκ τοῦ ὄπισθεν μέρους τοῦ
ἰσχαρίου);
4. The first cord is further qualified by the phrase ‘at’ or ‘in accordance
with’ (kata) her counterweight’ (κατὰ τὸ σήκωμα αὐτῆς);
5. The cord that turns around (περιάγει) Athena is ‘lying around the
cover’ (περικειμένη περὶ τὸ θωράκιον);
40
As Lewis 1997, 86, points out, this is a small-scale instance of a cam, but whether it was realised on
a much bigger scale at this point in history is much more debatable.

6. The cord turns the figure around ‘until it comes to the same place’ (ἕως
ἂν ἔλθῃ ἐπὶ τὸν αὐτὸν τόπον).
All of these observations are problematic. Hero’s view is that Philo offers
a complicated device, mêchanê, rather than a simple hinge. His counter-
proposal in fact offers less movement than Philo’s, as it lacks horizontal
rotation, but there may be additional differences between Hero’s idea of
a ginglumos and Philo’s mêchanê. The function of the pegs is not specified.
The location of cords behind the joint (ischarion) is unparalleled in three
ways. First, the term itself is not otherwise attested but is a diminutive of
ischion or hip joint. Second, no other figures in either stationary or mobile
automata have their cords or axles specified with such apparent anatomical
precision. Third, if this location is correct, the cords would be visible to the
viewer. Where visible cords are used elsewhere, in Hero’s addition of
a lightning strike on Ajax, they are explicitly camouflaged.41
Furthermore, if the figure is rotated, it is difficult to see how the second
cord can avoid becoming tangled up. As for the counterweight (sêkôma), if
kata is to do with place, there would appear to be a contradiction to the
specification about the cords in relation to the hip joint. If it relates to
means, then the implication is that there is a counterweight below the
figure to maintain its upright posture and stop it tipping over.42 As for the
horizontal rotation of Athena, it remains unclear why the cord is specified
as lying around the thôrakion, or what ‘turns round so as to come to the
same place’ means.
Proceeding in reverse order, it is most straightforward to understand the
horizontal rotation as simply that: rotation on the spot through 360 degrees
to come to the same position.43 The specification of the cord as lying
around the thôrakion relates to the routing of the cord to achieve the
rotation: if the figure is balanced upright, then it is essential that the rope
pulling it round is horizontal.44 The thôrakion is the vertical cover that
41
Camouflage: Aut. 30.2; Philo’s omission of the lightning-strike mechanism: Aut. 20.3.
42
Schmidt’s translation is rather loose (‘in order that it stays in equilibrium’) and fudges this point. For
drawing in accordance with forces, the best example is Philo, Bel. p.73 (Th.). 12–16. For κατά +
acc. of place, see esp. Hero, Aut. 2.11, in relation to the bundles of slack.
43
Querfurth’s reconstruction, quoted by Schmidt (Schmidt, Nix, Schöne and Heiberg 1899–1914, i.
lxiii–lxviii), has Athena circulating around the front of the stage from side to side. This relies on
a very complex mechanism and also seems to rely on a very narrow interpretation of περιάγω as
‘lead around’ rather than ‘rotate’ or ‘turn round’, and τόπος as ‘place’ rather than ‘position’. But
neither explanation is compelling.
44
As usual, instructions for routing cords are incomplete (where they exist); perhaps the most
complete routing is given for the rotating figures of the mobile automaton: chain from
a secondary counterweight via a catapult-style trigger, and cords to connect and synchronise

shields the mechanisms below the stage of the stationary automaton, and
by extension the compartment itself. The cord and thus the point of
rotation is thus below the surface of the stage and not visible.45 In order
for Athena to pivot up and down, the statue base must be pivoted. This
must, in fact, be the reason for the rather cryptic pegs (tyloi), which
elsewhere are mostly used as the points around which ends of ropes are
looped on an axle (or similar rotating device).46 The base must thus be on
or connected to an axle.47 This also explains why there are two separate
cords coming from the same direction: they may be looped around the axle
and peg in different directions to achieve the movement up and down.48
They would also not need to be visible but can pull from below. The
counterweight would need to be below the horizontal axle, but perhaps not
very far below, and it might still be possible to understand kata loosely in
a locative sense, but that is unnecessary and probably undesirable.49 This
reconstruction implies that ischarion cannot be a point of attachment for
the cords. One solution is to observe that the expression relates to
a direction or area of the display rather than a point of attachment as
such, although the specificity of the body part remains odd. The solution,
I suggest, is that the term ischarion does not literally denote the hip joint at
all. The clue may be in the diminutive, which is typical of parts in Hero’s
mechanical and pneumatic displays. The mechanical joint that is produced
by a base–axle combination that rotates both in the vertical plane and in
the horizontal plane is both distinctive in terms of the mechanical treatises
and very similar to the movement of the anatomical hip joint, and it seems
likely, then, that the term was coined for the mechanical part on the basis
Dionysus in the naiskos with Nike on the apex, over pulleys and up one of the naiskos’s columns
(Aut. 13.7–9).
45
For use as a vertical cover for mechanisms elsewhere in the Automata, see 15.2–3, 16.3; 23.2–3. This
use descends from its more general use to describe a parapet or breastwork (LSJ, ii). Without further
specification, the term must be used to denote this cover/compartment, and neither a separate
parapet around Athena’s statue base nor its very rare anatomical sense (LSJ, i, citing Lucian,
Par. 49).
46
Thus Aut. 2.9, 5.5, 12.3, 23.6, 27.4, 28.6–7. The only other use in the Automata or Pneumatica is on
a rod (kanonion) for connection to a screw: Aut. 10.2; Pneum. 1.5 (compare the use of tylion to link
rods, kanonia in Pneum. 1.6).
47
Schmidt (n. 42 above) has a hinged arrangement for flipping up and down but clearly does not
understand the point about the pegs. Prou 1884, 207–8, does not specify the arrangement but reports
Hero’s criticism. It is noticeable that all apparent hinges in Hero, both Pneumatica and Automata,
are operated through rotating axles, and none of them are referred to as ginglumoi.
48
For the principle, see the use of pegs on an axle to open and close the doors of the display (23.6).
Thus Schmidt’s emendation of the second ὄπισθεν (‘from behind’) to ἔμπροσθεν is entirely
unnecessary. The use of pegs in Querfurth’s reconstruction, to pull the statue base along
a groove, is wholly unparalleled.
49
Thus Prou’s rejection of (1884, 243n) of Baldi’s translation as incomprehensible is slightly excessive.

of that similarity. In other words, what can be seen here, as in other places,
is the machine as body, rather than the body as machine.50
In fact, in the developed automata of Philo and Hero, representation of
complete free-standing human forms is relatively scarce, and they do not
usually do anything except rotate, either directly or on a turntable. In
addition to those in the Philonian stationary automaton and the Heronian
mobile automaton, to which I shall return, there are a certain number of
(and better known) devices collected in the Pneumatica. A handful of these
are indicated as having elements that happen automatôs or similar. Before
turning to these, it is necessary to consider the evidence of Philo’s
Pneumatica, which is only available through the Arabic tradition and is
highly problematic as evidence.51 In a few cases, Philo’s devices are clearly
precursors of Hero’s, but the works diverge quite significantly, and there is
dispute over how many of those in the Arabic Philo really go back to the
Greek tradition and how many are interpolated or adapted in more than
superficial terms.52 Even so, there are interesting and significant diver-
gences between the contents of Philo’s Pneumatica and Hero’s Pneumatica.
One distinctive subset of Philo’s devices is marked by the use of water levels
to animate creatures, particularly birds and snakes. I give an example here
in Carra de Vaux’s French translation (followed, for the sake of conveni-
ence, by Prager’s English adaptation53 of the same passage):
Quand on verse l’eau dans le récipient, elle s’élève jusqu’à ce qu’elle atteigne
le flotteur du vautour; celui-ci est soulevé et son tuyau aussi, et le vautour,
puisqu’il est fixé au tuyau. Tandis que le vautour monte, les articulations des
ailes et celles du cou sont tendues par en haut, la tête se lève et se dresse et les
50
For other components that draw on body/machine analogies, note the thôrakion itself (as above, pp.
85–86), the ‘leg’ (skelos) of the siphon, which is ubiquitous in the Pneumatica, and the ‘hand’ (cheir)
of the catapult trigger (Aut. 13.9, cf. Philo, Bel. p. 68, 73–6, 78 Th; Hero, Bel. pp. 78–9, 100, 111W
[§§6, 23, 30 D–S]); for catapult devices turned back into human form, compare the Heracles archer
device (Hero, Pneum. 1.41). Other parts are named after entire creatures or their distinctive body
parts, such as the screw, kochlias (literally ‘snail’ or ‘shell’), and chelônion or chelônê (‘tortoise’), used
to refer to blocks used in various contexts, including catapults and medical racks (LSJ s.v. χελώνη
iii, χελώνιον iii.9); cf. Schiefsky (2005).
51
The opening chapters, and the earliest witness to this tradition, are in a Latin translation (so, rightly,
Drachmann 1948, 42; Prager’s claim that this was a direct translation from the Greek is clearly
wrong). The Latin text was edited by Rose 1870 and reprinted without apparatus in Schmidt, Nix,
Schöne and Heiberg 1899–1914, i.458–89). The chapters that concern us are all from the considerably
later Arabic manuscripts. For texts, see Carra de Vaux 1902 with French translation; Prager 1974,
with an English version.
52
Drachmann argues that many come either from the post-Philonian Greco-Roman tradition or from
the Arabic tradition (summary at 1948, 67–8). By contrast, Lewis 1997 and Schomberg 2008 offer
maximalist views which seem to me optimistic.
53
Note that Prager did not translate the Arabic directly.

ailes se déploient. Ensuite quand l’eau parvient aux flotteurs des passereaux,
les flotteurs s’élèvent et les verges aussi; les articulations des ailes se relàchent,
les ailes tombent et se ferment. Cela dure ainsi jusqu’à ce que l’eau soit
évacuée du recipient. Quand elle est sortie, chaque chose retourne à sa place
comme elle était d’abord. Philo, Pneum. 42
When water is poured into the reservoir this water rises and reaches the
vulture float; the float is lifted, with its pipe and with the vulture fixed to its
pipe; while the vulture rises, his wing and neck articulations are pulled up;
his head rises; the wings unfold. When water then reaches the sparrow floats,
these floats rise, and also their rods; the wing articulations are relaxed; the
wings fall and fold. This condition prevails until all the water is evacuated
from the reservoir. Prager 1974
The birds rise in response to water level, with their necks and wings
both being articulated; other pneumatic devices in Philo’s work are
similar.54 Because of the difficulties of the tradition, it is not clear
whether Philo described this as an automaton, although the example
of Hero suggests that it is unlikely: all of Hero’s pneumatic devices that
are so qualified have an element that includes a falling weight.
Nonetheless, the complexity of the device, the use of multiple creatures
and their having articulated joints are certainly relevant to the theme of
this chapter. This device, in particular, shows a concern to present
a more naturalistically moving creature, even if it remains limited both
in the range and number of movements, depending on water being
introduced into the device and then being evacuated via a siphon. With
a continuous water supply, repeated effects might be achieved, but that
is often not specified in Philo’s or Hero’s devices.55
By contrast, Hero’s Pneumatica has fewer displays of that sort, but one of
his more complex displays is both described as an automaton and combines
weights and pneumatics to provide both repeated half-rotation and sound
effects.
κατασκευάζεται οὖν ἤτοι ἐν κρήνῃ ἢ ἐν ἄντρῳ ἢ καθόλου ὅπου ἐπίρρυτον
ὕδωρ ἐστίν, ὄρνεα πλείονα διακείμενα καὶ τούτοις παρακειμένη γλαύξ, ἥτις
ἐπιστρέφεται αὐτομάτως παρὰ τὰ ὄρνεα καὶ πάλιν ἀποστρέφεται· καὶ
ἀποστραφείσης μὲν φθέγγονται τὰ ὄρνεα, ἐπιστραφείσης δὲ πρὸς αὐτὰ
οὐκέτι φθέγγονται. καὶ τοῦτο πλεονάκις γίνεται.
54
Compare Philo, Pneum. 40 and 41.
55
Contrast Hero, Pneum. 1.16, which is explicitly flagged as automatos and is powered by a ‘constantly
flowing spout’ (κρουνισμάτιον ἀεὶ ῥἐον). The many constant-level devices require a flow of water:
see below, p. 92, on clocks.

In a spring or cave or generally anywhere where water can flow, birds are
constructed and arranged, and an owl to their side, rotating towards the
birds and rotating away again; when it is turned away, the birds sing, and
once it has turned towards them they no longer sing. And this happens
several times. Hero, Pneum. 1.16
In this example, the filling and emptying of a reservoir serves both to
create the sound effects and the movement, by the forcing of air through
connected pipes before it is evacuated via siphon into a counterweighted
container, which revolves an axle that in turn causes the owl to rotate.56
As for Hero’s mobile automaton, the human figure is not as dominant as
some of Hero’s observations suggest. In particular, some significant pre-
liminary remarks discuss the configurations of rotating figures in general,
either supported on pin-bearings (knôdakes in iron sockets) or having
bronze collars that are set into bronze axle-boxes or collars.
δεῖ δὲ καὶ ὅσ᾿ ἂν ἐγκυκλίους στροφὰς ἢ κινήσεις ποιῆται, ταῦτα ἔντορνά τε
ἀκριβῶς καὶ περὶ ἃ κινεῖται λεῖα καὶ μὴ τραχέα ὑπάρχειν, οἷον οἱ μὲν τροχοὶ
περὶ κνώδακας σιδηροῦς ἐμβεβηκότας εἰς ἐμπυελίδας σιδηρᾶς, τὰ δὲ ζῴδια
περὶ ἄξονας χαλκοῦς ἐμβεβηκότας εἰς χοινικίδας χαλκᾶς συνεσμηρισμένας
αὐτοῖς.
It is necessary for everything that make circular rotations or movements to
be turned precisely and the elements that things turn on to be smooth and
not rough, such as the wheels that move on pin-bearings that go into iron
sockets, and the figures that move on bronze axles that go into bronze collars
that are tightly mated with them. Hero, Aut. 2.3
Leaving aside the question of which is technically preferable,57 there are in fact
no examples of the choinikis (bronze-collar) type in the treatise in relation to
a human or animal figure; the knôdax is generally preferred.58 Indeed, there are
very few rotating figures of any kind at all. The display has Dionysus (with
panther) rotating 180 degrees and having milk or water come from his thyrsos
and wine from his cup, in a manner very similar to the figure in Ptolemy’s
procession. Slaved to Dionysus is the Nike on the naiskos roof.59 There is,
56
The same principle that is observed in opening gates: Pneum. 1.38 and 1.39; for the axle configur-
ation, compare the stationary automaton (Aut. 23).
57
For discussion of bearings in the treatise, see Keenan-Jones, Ruffell and McGookin 2016.
58
The only appearance is in relation to a rejected or superseded configuration for snake-like motion
(Aut. 11), a passage which makes plain (11.8) the general preference for knôdakes. This is indicative of
the layers of authorship to be observed in book 1, but it is beyond the scope of this chapter to tease
apart those layers.
59
Aut. 13; the latter, certainly, pivots on a knôdax. Note also that the Nysa device follows a cart which
has a statue of Dionysus with very similar iconography to that of Hero’s naiskos.

however, no articulation. The dancing maenads do even less, standing fixed
on a circular strip of wood that is made to rotate around the shrine (naiskos)
that is the centrepiece of the upper display. It is possible, even likely, that at
one stage the pre-Heronian treatise or treatises that dealt with automata
featured many more of these sorts of devices. Even so, the sense of limited
or even no articulation of the human figure appears to be a feature of ancient
technical writing about automata.
There are two fundamental limitations with ancient technology that
mean that developing a realistic facsimile of the human body could never
be a proposition that could be achieved in a realistic or naturalistic way. One
is the need for a power source: the fact that Nysa is carried on a vehicle on
a parade clearly solves that problem while the cart is moving, but in other
circumstances, the performance of any automaton was strictly circum-
scribed. The other limitation is the mechanisms that could be powered.
Almost all of Hero’s devices in his automata depend on rotational energy
from axles and drums or, in the case of fluids, rely on floats, weight or
pressure, including their interaction with counterweights and drawn cords.
Only occasionally, as in the case of the screw in the (highly impractical and
undoubtedly theoretical) mechanism for rectangular motion is there any
sense of how one might convert circular motion to straight-line motion (in
that case by the use of a leading screw to drive an engaged rod).60 This is
even leaving aside issues of fine motor control and the fact that Hero largely
avoids toothed drums or gears.61 It is not, of course, that such things were
unknown, either in theory or in practice.62 One reason for their avoidance
may, perhaps, have been a form of conservatism, for all that Hero (and
probably his sources) are concerned to distinguish their work from that of
predecessors. Even here, though, there may have been technical reasons
underlying this choice: in his description of an odometer, Hero discusses the
issues of wear, slipping and unreliability in relation to (bronze) toothed
60
Aut. 10. Conversion of straight-line to rotational force is a problem in water clocks, initially solved
by a rack and engaged wheel but later solved by axles and counterweights: see below, p. 92.
61
For an explicit statement of this preference, see Hero, Mech. 2.21 (Drachmann, 1963, 83); compare
Vitruvius 10.2.5–7.
62
For toothed wheels or drums, see the dioptra device (engaged with a screw, Dioptra §3) and other
devices in the same treatise, with various types of gearing (land and naval odometers, §34 and §37;
weight-lifting device, §36), and the possibly theoretical baroulkos (Drachmann, 1963, 22–32). The
Antikythera mechanism is the most obvious example of very precise gearing (Freeth, Bitsakis,
Moussas 2006, Freeth, Jones, Steele and Bitsakis 2008), albeit under minimal load. In Hero’s
Automata and Pneumatica, the only toothed wheel is a lantern pinion in Pneum. 2.32; the same is
found in the basic design of the ‘Archimedes clock’, on which see below, p. 93. For a judicious
summary of the place of toothed wheels and gears in ancient mechanics, see Drachmann 1963,
200–3.

drums,63 and it may well be that the problems relating to available toothed
gears under load should be added to the wider set of technical limitations
that meant that anything approaching a mechanical humanoid was well out
of any aspirational reach.
In addition to the practical limitations, there may have been other reasons
why the later automata were not concerned to push forward a naturalistic
representation of the human body in action. Hero’s own explanations for
why one should read about automaton-making at all at the beginning of the
Automata treatise offer some clues. He says that automaton-making is useful
because it contains examples of every kind of mechanics. In other words, it is
a technology demonstrator. This seems broadly true, in that the mobile
automaton, in particular, contains such things as pipework, taps and an
entirely gratuitous catapult trigger (13.9), but given the many gaps in the
instructions it might be better described as a technician demonstrator. Either
way, naturalistic representation is not the highest priority. Furthermore,
Hero is also at pains to distinguish the practice, as a mechanical discipline,
from the previous understanding of it as a form of ‘wonder-working’. The
way he does this is to argue that the thauma and ekplêxis come not from the
experience of watching something inexplicable (a snail or a statue moving by
themselves) but from an understanding of the mechanical expertise required
to build them and achieve these effects. There is undoubtedly special
pleading going on here, in the attempt to separate out a genre of mechanical
writing from and to define it against the more heterogeneous and downright
unrespectable performance form known as thaumatopoiia.64 All the same,
the reasons for not developing a more realistic rendition of the human body
may owe rather more to the performative roots of automata than to their
mechanical realisation, and for this I turn to one final set of evidence, which
describes some ancient water clocks.
Narrative and Performance: The Wonder of Water Clocks

Water clocks are closely related to both pneumatics and automata. Indeed,
Hero’s Pneumatica is presented as a pendant to a much larger, four-volume
work on clocks, now unfortunately lost.65 The relation is partly
63
Dioptra §34.
64
Tybjerg 2003 has suggested, in an influential argument, that Hero’s deployment of thaumata was to
pursue an epistemological point about science writing, as it is deployed in the Aristotelian
Mechanica (cf. also Cuomo 2007), but that is not the rhetoric of the Automata preface.
65
Hero, Pneum. 1. pref. Schmidt (Schmidt, Nix, Schöne and Heiberg 1899–1914, i.456) prints
a surviving fragment.

theoretical – the fundamental problem of clock development (described by
Vitruvius) was how to maintain a constant flow rate, because of decreasing
water pressure as the principal vessel (klepsydra) empties.66 The connection
is also more strictly mechanical, in that the displays that actually tell the
time for the viewer take advantage of the mechanical techniques used in the
Automata and both Pneumatica treatises. A float is used, as in Philo’s
Pneumatica, either to drive an axle to operate symbolic displays or to
raise a pointer to an adjacent scale.67 In early versions, the float raises
a rack that engages a toothed wheel, but more sophisticated versions use
a counterweight to take up slack and turn a drum as the float rises. The
drum can then turn further axles to create more sophisticated effects.
Unlike most of the automata I have been describing, the developed water
clocks have a long-lasting, constant power source. The klepsydra is kept
full, either by maintaining an overflow, as in earlier designs discussed by
Vitruvius, or by means of a float-valve at the top of the klepsydra. In the
latter type, a float is placed in the upper container, which (also) empties
into the klepsydra at a steady rate.68 Although these devices driven by the
float are rather dismissively labelled as parerga by Vitruvius (9.8.5), who
characteristically only describes simple effects such as a pebble dropping
every hour, they are essential to the utility of these devices; they also,
clearly, became the focus of considerable ingenuity and elaboration. These
displays provide a third connection with automata, a performative one,
and it is this that allows us to approach much more clearly what automatic
devices meant to an audience.
A monumental clock is described by Procopius of Gaza, a writer of the
late fifth and early sixth century ce, in an ecphrastic but unfortunately
incomplete work.69 Procopius’ interest is not in the internal mechanisms
but in the figures that adorn the clock, some of which are clearly automata.
66
See Vitruvius 9.8.5; compare Drachmann 1948, 17–19. The other key problem is not a pneumatic
one, but how to address the issue of lengthening and shortening days: Vitruvius 9.8.6–15;
compare Drachmann 1963, 20–36.
67
Vitruvius 9.8.5–6, the latter the so-called ‘parastatic’ clock, going back to Ctesibius. See Drachmann
1948, 20f.
68
For a natural power source for the overflow, see the ‘Tower of the Winds’ at Athens (Noble and de Solla
Price 1968). Vitruvius does not mention the latter (upper container plus valve) type, which may thus
post-date him, although perhaps not by much: Drachmann 1948, 41, argues that the crucial valve is later
than Hero’s version of a float valve (Pneum. 1.20) and finds similarities in the use of a floating siphon
(cf. Hero, Pneum. 1.5), but his suggestion that the use of a descending float to drive parerga is
reminiscent of Hero’s ‘automatic theatre’ misses the point that this goes back to Philo (and it is clearly
indicated in Vitruvius). On water clocks generally, see also Schürmann 1991, 252–73.
69
For text, French translation and commentary, see Amato 2014. I also give references for the edition
of Diels 1917, which may be more readily accessible and also contains a foundational discussion; see
also Bäbler and Schomberg 2010.

It is possible to infer some of the movements from his description and from
comparison with devices known from the Greek mechanical tradition.
Further insight into the functioning of the clock also comes from the so-
called ‘Archimedes clock’ described in Arabic manuscripts, which is
another monumental clock, explicitly an in-flow device, and which clearly
draws on Greek sources, including the Gaza clock or something very
similar.70 For all the problems of this treatise, the great advantage is that
this clock is described mechanically rather than ecphrastically, albeit that
the Arabic work has reinterpreted or misinterpreted a number of the
elements: a gorgon head becomes generic flashing eyes, Helios and the
labours of Heracles has become an executioner knocking the heads off
twelve captives, one per hour, and eagles descending have become riders
dropping onto horses. Other elements in this device resemble those from
the pneumatic tradition: in particular, the emerging snakes and shrieking
birds both recall devices in Philo’s Pneumatica.71 With care, then, it is
possible to say rather more about the devices that Procopius describes.
After comparing various stories (logoi) of marvellous technê, including
mythological automata, and contrasting them unfavourably with this
spectacle (§§1–3 (§§1–9 = 1–75 Diels)), Procopius describes the location
of the clock in the centre of Gaza, the opportunities for viewing it (§3
(end)–4 (§§10–13 = 76–95 Diels)), and then the displays and associated
sound effects, which represent a series of mythological figures and narra-
tives. These comprise a Gorgon-head at the top; twelve doors, from which
Heracles emerges with one of his labours, each door surmounted by an
eagle which crowns Heracles; below that, a larger representation of
Heracles, banging a gong; Pan and satyrs above his head, Diomedes with
trumpet to his right, and a shepherd to the left, themselves flanked by a pair
of servants; and finally a further larger rendition of Heracles and the apples
of the Hesperides, where the narrative breaks off.
70
English translation: Hill 1976, which I use here, cited as Kitāb Arshimadas; German translation:
Wiedemann and Hauser 1918. Hill argues that the treatise has accumulated Hellenistic, Byzantine
and Arabic (and Persian) elements, arguing that the opening sections and basic functions go back to
Archimedes and Philo (a proposition accepted by Simms 1995, 54–5, without comment; Lewis 1997,
37–41, associates it with Philo, in an optimistically unfiltered way). For discussion, see also
Drachmann 1948, 36–41, who explored the relationship with the Gaza clock and was (to my
mind rightly) sceptical about the attributions (so too Carra de Vaux 1891, 296, and Wiedemann
and Hauser 1918, 6). Bäbler and Schomberg 2010 also use the treatise in their reconstruction of
Procopius’ clock.
71
For snake and shrieking birds, §8; compare Philo, Pneum. 40. For sound effects, see n. 82 below.

Although Procopius’ description is thin on technical details, it is still
possible to see the relationship to automata in a number of these elements,
beginning with the Gorgon’s eyes which change colour to indicate the
hours.
᾿Αλλὰ καὶ Γοργὼ ἐφ’ὕψους βλοσυρὸν ἀπειλεῖ τοῖς ὅσοι γνώμῃ προσελθεῖν
αὐθαδεστέρᾳ τολμῶσιν, ἐναλλάττουσα τοὺς ὀφθαλμοὺς ὁπόσα τῆς ἡμέρας
τέμνεται·
The gorgon, on high, grimly threatens all those who dare to regard it too
rashly, alternating her eyes for all the divisions of the day.
Procopius, Horologium §4 (= §14 = 96–101 Diels)
This is clearly a large-scale version of the face with opening and closing eyes
that according to Hero – apparently following Philo – was revealed as the
first tableau in earlier incarnations of the stationary automaton, which was
easily achieved by means of a rotating axle.72
As for other movements, the principal effect is the emergence of
Heracles to mark every hour.
καραδοκῶν οὖν ἐφέστηκεν ἀετός, ὁπότε τῶν ἐπικειμένων ὁ τοῦ Διὸς
῾Ηρακλῆς ἐξίοι θυρῶν ὥραν ἀγγέλλων, ὁ μὲν πρῶτος τὴν πρώτην, πρὸς
δὲ τὸν ἀριθμὸν οἱ λοιποί. δώδεκα μὲν γὰρ ὧραι, πᾶσαι δέ εἰσιν ῾Ηρακλῆς,
οὐκ ἀργὸς μὰ Δία καὶ πράττων οὐδέν· ἀργεῖν γὰρ ὅλως οὐ φίλον ἦν
᾿Ηρακλεῖ.
So an eagle stands on the clock waiting for when, out of the doors that hide
him, comes the son of Zeus, Heracles, announcing the hour, the first
announcing the first hour, the rest according to the number. For there are
twelve hours, and they are all Heracles – no good-for-nothing slouch by
Zeus. For slacking was wholly anathema to Heracles. Procopius, Horologum
§7 (§§27–8 = 156–66 Diels)
The eagle is made of bronze (§§6 and 8 (§§22 and 43 = 133 and 211–12
Diels)) and has articulated wings, rather like birds in some of Philo’s
pneumatic devices.73 Heracles emerges, pushing open the doors (also
bronze), and the eagle places the wreath on his head.
Τὰς οὖν χαλκᾶς ἀπωσαμένῳ θύρας καὶ σὺν ἄθλῳ φανέντι ἐφέπεται μὲν
ἄνωθεν ἀετός, πτέρυγάς τε ἁπλώσας καὶ ποσὶν ἀμφοτέροις ἴσον ἴσῃ κεφαλῇ
τὸν στέφανον ἐπιφέρων· βραχὺ δὲ διατρίβει καθάπερ ἀπολαύων ἡρωϊκῆς
κεφαλῆς· εἶτα τοῦτον ἀφεὶς ῾Ηρακλεῖ καὶ διαστήσας τὼ πόδε μετεωρίζεταί
72
Aut. 22.1. See discussion above, p. 83. This is confirmed by the ‘Archimedes clock’ §3 whose flashing
eyes are created by an axle studded with gems of various colours.
73
Discussed above, p. 87–88.

τε καὶ ὃν εἶχε τόπον λαγχάνει, ταῖς πλευραῖς ἐπιθεὶς τὰ πτερὰ καὶ
συναγαγὼν αὖθις ἐφ’ ἑαυτόν, θήραν διδοὺς ἀγαθήν, ἀλλ’ οὐ συνῆλθε
λαχών.
So when he has thrust aside the bronze doors and appeared with his prize, the
eagle comes upon him from above, opening its wings and carrying with its
talons a wreath which exactly fits his head. It stays a short while, as if it enjoys his
heroic head; then after dropping the wreath for Heracles and moving its talons
away it flies up in the air and takes up the position it previously occupied, after
bringing his wings to his sides and furling them again against himself, providing
a good hunt – but after taking his position he did not meet him again.
Procopius, Horologium §9 (§§44–5 = 215–27 Diels)
The descent and ascent of the eagle can be easily managed if it is on the end
of a counterweighted arm that is pivoted about an axle. The axle is rotated
forward, the eagle descends, releases the crown, and then moves back up
again. If the wings are attached by cords to the facade, they will open and
close in the desired way. The axle might be operated by the use of cords,
but the mechanisms here look similar to those employed in a series of
horses and riders on the ‘Archimedes clock’: there, the rider is on
a counterweighted arm which is brought down by releasing a hinge or
catch.74 Cords for the doors are attached to the rider, and they slacken as he
descends. The doors, weighted and angled to fall outwards, thus open.75 As
for the wreath that the eagle carries, if the talons are angled carefully, the
wreath will drop off. If the wreath is sufficiently heavy, its absence will alter
the balance of the arm, allowing the eagle to ascend again.76 Alternative
mechanisms might be devised, but only at the expense of proliferating
cords and additional complexity.77
As for Heracles himself, he moves forward through the doors together
with a token of the appropriate labour before receiving the crown. He then
goes back into the clock.
πρός γε μὴν τοὺς στεφάνους προκύπτει μὲν ῾Ηρακλῆς, ὡς ἂν ὁρῷτο πᾶσιν
ὡς ἐν μέσῳ σταδίῳ, κἆτα πρὸς ἰδίαν ἐτράπετο χώραν, οὗπερ ἐπόθει
τυχών.
74
For catches in the mobile automaton, see the release of garlands (Aut. 15.4), as well as the catapult
trigger (Aut. 13.9). The instructions for the ’Archimedes clock’ allow for a series of catches to be
released in order by means of one cord.
75
Kitāb Arshimadas §6. Compared with the eagle, the cords would function in an opposite fashion
(taut to slack).
76
Again, there are similarities with the balls that mark the hour in the ‘Archimedes clock’. After being
released from a drum, they drop into a crow’s beak, whose lower jaw is hinged and counterweighted:
the jaw drops, the ball falls out and the jaw closes again (Kitāb Arshimadas §2).
77
For example, by means of having an additional catch for the wreath.

Heracles emerges for the crowns, in order to be seen by everyone as in the

middle of a stadium, and then he turned back to his proper place after
achieving what he desired.
Procopius, Horologium §9 cont. (§46 = 227–32 Diels)
The mechanism is perhaps slightly less obvious here. The verb prokuptô is
appropriate for Heracles emerging from an opening and need not imply
any ducking of the head, although it might imply that he comes out
leading with his head.78 This would be straightforward if he were descend-
ing a slightly sloping surface. Drawing him forward and backwards with
cords that lead to an axle that changes direction may be the most straight-
forward and most controllable method of moving him. The Greek need
not imply that he literally turns round, although that could be contrived,
similar to the case of Athena in the stationary automaton.
Reinforcing the use of Heracles to mark the hours, there is a figure of
Helios, a personification of the sun, who has his left arm raised to the sky,
the other pointing to the correct hour, as he moves along in front of the
doors.79 This figure moves, rather exceptionally, in a straight line. The
mechanism must be that described in ‘Archimedes clock’ for the execu-
tioner who proceeds knocking off the heads of twelve captives in a row. He
is simply pulled slowly along a hidden channel, with the rope leading to the
main rotating drum.80 Unlike the executioner, Helios does not knock off
anyone’s head.
Underneath the displays that indicate the changing of the hours come
the series of figures that according to Procopius relate to and centre around
a young figure of Heracles. He is depicted as beardless and naked apart
from the lionskin on his shoulders. He holds a club (korynê) in his right
hand and a cymbal or gong (êcheion) dangling from his right:
καθώπλισται δὲ κορύνῃ τὴν δεξιὰν, ἀντὶ καὶ τῆς ἐκείνου βοῆς ἀνατείνει τε
ταύτην καὶ δίδωσι πληγὴν τῷ χαλκῷ· ὁ δὲ μετέωρός τε ὢν καὶ τοσαύτῃ
ῥώῃ πληγεὶς βοᾷ τε μέγα καὶ παρατείνει τὸν ἦχον. ἐφ’ ἑνὶ μὲν οὖν ἄθλῳ μία
78
Given what is said earlier about emerging with a token of his task, he must emerge bodily, although
the verb could be compatible with just poking his head out of the door (LSJ, q.v. A.i.2). The
‘Archimedes clock’ (§5) has an executioner knocking the heads off a series of captives, whose heads
are hinged to fall forwards, but this does not seem to offer a useful parallel for Heracles.
79
Horologium §6 (§§24–6 = 144–155 Diels). Procopius compares him to the starter at horse races,
although that does not necessarily imply that the doors themselves functioned in the same way as the
starting gates (balbides).
80
Kitāb Arshimadas §5. The executioner and catch-release system for the horse riders work similarly in
this respect.

πληγή, ἐφ’ ὥρᾳ δὲ δευτέρᾳ διττή. συλλέγει γὰρ ἐφ’ ἑκάστῃ καὶ τῶν
φθασάντων ἄθλων τὸν ἀριθμόν
He is equipped with a club in his right hand, and to make the specific
sound of that instrument he stretches it out and strikes the bronze; as
it is in the air and struck with such force it makes a loud noise and
resounds loudly. So for one task there is one strike, and in the case of
the second hour there is a double strike. For at each hour he also
collects the number of previous tasks. Procopius, Horologium §10 (§§52–5
= 254–67 Diels)
It is not impossible that the figure of Heracles actually strikes the
gong, and if so it would be a very similar arrangement to the figures
in the stationary automaton that are chopping, hammering and saw-
ing, except that there are indications that in this instance the figures
would be in high relief rather than painted figures.81 Alternatively, and
I think preferably, the sound effects might have been generated from
behind the facade, in which case there are other options, in particular
the well-attested use of metal balls to strike a cymbal.82 It is interest-
ing that in this case the effects do not go up to twelve but restart at
seven, giving two groups of up to six (§10 cont. (§55 = 247–75 Diels)).
This may have been for clarity, control or some other technical
limitations. A different mechanism will have produced the sound
effect associated with Diomedes. He flanks Heracles and blows his
trumpet to mark the end of the labours and the twelve-hour period.
This is clearly a large-scale version of a number of pneumatic devices
attested by Hero and Philo which produce various sounds by filling
a container with water and causing the air to be expelled through
a thin tube.83 In this case, the pipe will have passed through the figure
to the instrument. In the ‘Archimedes clock’, a trumpet effect is
produced in exactly this way, with the water drawn out of the
receiving vessel every six hours by means of an enclosed siphon.84
81
The text describes Heracles as μεγάλῳ τε . . . καὶ προβεβλημένῳ τῶν ἄλλων, which can be translated
either as ‘large and set forward of everything else’ or as ‘large . . . and projecting beyond the rest’, §10
(§50 = 245–7 Diels)). For Philo’s painted figures, see above, p. 84.
82
Note the release of balls to mark the hour in the ‘Archimedes clock’, in probably the earliest layer of
the treatise, that strike a cymbal (§2). Also compare the release of lead balls to hit a cymbal in Hero’s
mobile automaton (Aut. 14).
83
Horologium §10 (§63–5 = 305–12 Diels). Compare Hero, Pneum. 1.15, 1.16, 1.17, 2.5, 2.10, 2.15, 2.21,
2.26, 2.32, 2.35.
84
Kitāb Arshimadas §9; the repeatedly twittering birds described in the preceding section function on
similar principles. The manuscripts that preserve this treatise also include a short work on
a pneumatic flute player attributed to one ‘Apollonius the carpenter, the geometrician’, which
Hill 1976, 5, takes to be a Byzantine writer (for a different view, see Lewis 1997, 49–57). Similar

There seems no doubt, however, that these sound effects were actually
generated rather than simply imaginary.85
It seems unlikely that many more, if any, of the other items on the clock
were animated, although there is some uncertainty because of the narrative
perspective offered by Procopius which makes mechanical interpretation
far from straightforward. Unlike Hero or indeed the writer(s) on the
‘Archimedes clock’, Procopius is not writing from the perspective of
a maker (or mechanic or engineer). Rather, he is writing from the perspec-
tive of the spectator, and specifically a theatrical perspective, which consists
of two intersecting elements: on the one hand, sight, spectacle and wonder,
and on the other, narrative and storytelling. This approach is applied both
to the clearly mechanical elements, the clearly non-mechanical figures and
those that are somewhat ambiguous.
There are two main narratives. The twelve labours are described as
provoking responses from eagles and from Diomedes, and also from the
servants. The central figure of the youthful Heracles and his gong set up
a further narrative involving the nymph Echo and the reactions of Pan who
desires her.
ποθῶν δὲ τὴν ᾿Ηχὼ αἰσθάνεται τοῦ χαλκοῦ καὶ [± 20] σαι καὶ περιφέρει τὸ
πρόσωπον, εἴ πως ἴδοι τὴν κόρην στροφὰς στρεφόμενος, ὁποίας ἔρως
ἀτυχῶν ἐμποιεῖ. εἴποις δ’ ἂν αὐτὸν καὶ τὸν ῾Ηρακλέα θαυμάζειν, ὁπόσος
καὶ οἷος. ῎Εδει δὲ Πανὸς παρόντος μηδὲ Σατύρους ἀπεῖναι. ἐπιγελῶσι δὲ καὶ
κωμικῶς μέσον εἰληφότες τὸν Πᾶνα, πρόσωπον ὁρῶντες ἐρωτικόν τε καὶ
ἄγριον καὶ ἦθος κεκραμένον, ἥμερόν τε καὶ ἀπηνές.
Out of desire for Echo, he hears the gong and . . . turns his face round, in
case he might somehow see the young woman twisting in the turns that
unlucky love causes. You would also say that he was amazed at the size and
nature of Heracles. And as Pan was present, it was necessary for satyrs too
not to be absent. They have trapped Pan in their midst and laugh at him,
seeing his love-struck and wild features and his mixed nature, tame and
wild. Procopius, Horologium §11 (§§59–61 = 285–301 Diels)
Motivation and movement are here attributed to these characters, even
where such movement seems absent or only tenuously suggestive of
a particular narrative. If Pan’s love-struck strophai are not purely
sound effects were said to have been produced in the throne room of Byzantium: see Brett 1954. For
pipes running through free-standing figures, compare both the Dionysus of the mobile automaton
(Aut. 13.2–3) and the Nysa automaton discussed above, p. 82–83.
85
This seems to be the clear implications of the concluding remarks of §13 (§71 = 328–30 Diels): ᾿Αλλὰ
ταύτῃ μὲν ἀμφὶ τοῖς ἄθλοις ἦχος καὶ θαῦμα καὶ σάλπιγξ (‘In this way the labours are accompanied
by clanging, wonder and a trumpet’).

psychological, this would indicate another instance of a rotating figure, like
so many automata discussed in this chapter. The satyrs, however, do not
seem to have any particular movement or mechanism associated with
them, as is the case for a number of other figures, both generic (servants
and shepherd) and specific (the lower figure of Heracles as an archer).86
Nonetheless, similar motivations and back stories are constructed for all
the elements. The combination of spectacle and narrativity is quintessen-
tially theatrical, but that quality is further reinforced by the presence of
these satyrs and their comic (kômikôs) mockery of Pan, by explicitly
comparing the look of the gorgon to tragic imagery87 and also by aspects
of Heracles himself. The servants who flank the youthful figure are held by
Procopius to be rushing to prepare domestic comforts for the hero after
returning from his labours. The mythical figure in a domestic scene is
suggestive of a comic context, and the contrasting entrances from agora and
home for the servants’ entrance is reminiscent of plays from New
Comedy.88 Indeed, the functions of these servants themselves suggest the
representation of Heracles in drama, particularly the servant bringing back
opson from the market, which alludes to the idea of Heracles as a glutton,
a staple of Greek comedy.89
The response that the spectator has to all this spectacle and narrative,
according to Procopius, and a response on which he insists, is one of
wonder. It is produced through interpretation of the various sign-
systems, both mechanical and figurative, and from the resulting narratives
that are constructed. As should be obvious, this wonder is not that which
Hero is articulating in the Automata, but it is undoubtedly the one against
which he is reacting.90
86
It is possible that the description of the lower Heracles may have been amplified, as the text breaks
off, but since he is described as ‘being about to let the arrow go’ (μέλλων ἀφεῖναι τὸ βέλος, §14 (§75 =
345 Diels)) there does not seem to be any dramatic movement. Although Hero presents a water-
powered archer, this is by its nature a single-shot device, not a true automaton, as is inevitable with
such a weapon.
87
§4 (§15 = 105–7 Diels), adapting Euripides’ description of the Furies in Orestes 255–6: Amato 2014,
153–4.
88
See, for example, Plautus’ Casina (based on Diphilus’ Klêroumenoi, or Men Casting Lots), which
contrasts off-stage agora and country estate. That opposition may have been encoded in use of the
side entrances, even though Pollux 4.126–7 insists the distinction is between intra- and extra-polis
entrances. It is, however, unclear when or indeed if this became a rigid convention. See Taplin 1977,
450–1.
89
Perversely, the best example is Euripides’ Alcestis, but Aristophanes’ claims about this character
(Wasps 60; Peace 741–3) also seem borne out by his own use of Heracles in Birds and Frogs, as well as
by the comic fragments.
90
For the descriptions of the Automata as ecphrastic and redirecting the wonder of thaumatopoiia, see
Roby 2016. Roby does not discuss Procopius.

This essentially theatrical quality of this wonder91 is precisely the reason
why there was no pressure to develop more realistic representation of
articulated human and animal bodies. Ancient spectators did not expect
verisimilitude in theatrical productions and were well used to supplement-
ing and interpreting visual representations in order to construct narratives
and imaginary worlds. Of course, such audience or reader cooperation is
always required with fictional narratives and static visual representations,
but more realistic forms of theatre can obscure the point, and that is far
from the case with ancient theatre. Tragedy, based almost exclusively on
the mythical repertoire, always required audiences to supplement the on-
stage narrative.92 Comedy, it is true, had long since moved away from the
grotesque towards domestic plots and more realistic forms of visual repre-
sentation, but this should not obscure the extent to which the genre was
stylised and also depended on audience supplementation through such
means as character types and masks.93 Both forms of supplementation are
used in Procopius’ interpretation of the scenes on the clock, taking the
visual cues and elaborating the story. Both automata and static figures are
thus being created in a visual economy that is as much symbolic as it is
literal. Other automata are more explicitly theatrical, particularly the
stationary automaton of Philo and Hero,94 but Procopius’ treatise demon-
strates that for spectators there was a clear continuum. As such there was no
great performative drive towards more literal rendering of the human form
in action.
Conclusion
The explanations I have offered for the limited articulation of the body in
ancient automata are partly functional, partly technical and partly cultural.
It remains to be seen whether there are distinctive differences in what
bodies are represented. It is striking that many automata were gods
91
Clearly, this is an ecphrastic work, but the emphasis on spectators and narrative foregrounds the
theatrical quality. This is closer to, for example, Theocritus, Id. 15 than, on the one hand, much
literary ecphrasis, and on the other the descriptions of statues by Philostratus or Callistratus.
92
The subject of a famous complaint in Antiphanes’ comedy, Poetry (Poiêsis (fr. 189); it also, clearly,
did not preclude tragic innovation, including at the story level: see Ruffell 2016 with bibliography.
93
See especially Wiles 1991. The rigidity and systematic nature of the roster of character types can be
overstated. See Ruffell 2014 for the debate and for further bibliography. Mime, the most popular
form of comedy in the imperial period, seems to have employed a more limited repertoire but still
required similar audience cooperation.
94
Marshall 2003 explores connections with classical tragedy. Beacham 2013 discusses the automaton in
the context of contemporary theatrical practice.

(particularly Dionysus and his entourage, Nike, Athena) or quasi-divine
(Heracles, in his classical or post-classical guise). Comparatively few are
human (Greeks in Hero/Philo’s stationary automaton; Diomedes the
trumpeter on the Gaza clock), although animals are frequent enough
(snail, dolphins, snakes, birds). Given the undoubted association of some
of these automata with festivals and symposia, there might be some
suggestion that there is an inhibition against representing an articulated
human form where it was acceptable to represent gods on the one hand and
animals on the other, perhaps with an eye on the role of the gods as creators
and prime movers. The evidence, however, suggests that representations of
the mobile human form were a consistent feature of the tradition
(Aristotle’s automata; Philo and Hero; Gaza clock), if never its most
dominant form, and that creation of mobile fish, reptiles and birds
would seem to obviate any such reservation of such powers to the divine
sphere.
To conclude, then, it seems clear that the attempts by ancient mechan-
ical practitioners to render the human form mechanically were strictly
circumscribed. Some of the explanation is purely practical, in terms of the
limitations of the technology available. Some can be afforded by the
changing priorities as the discipline of mechanics evolved. And perhaps
most significant of all, I would suggest, is the capability of ancient audi-
ences to construct their own narratives around representations that were
only in part realistic or literal. Or, as we have seen with the most grandiose
and spectacular of the possible automata I have discussed, literally only
a part was mechanical and realistic. The cooperative imagination of the
audience did the rest.
RE F E RE N CE S
Amato, E., ed. 2014. Procope de Gaza: Discoiurs et Fragments (Paris).
Bäbler, B. and A. Schomberg. 2010. ‘Prokop: Die Kunstuhr in Gaza’, in E. Amato,
ed., Rose di Gaza: gli scritti retorico-sofistici e le Epistole di Procopio di Gaza
(Alessandria), 528–59.
Baldry, H. C. 1953. ‘The Idler’s Paradise in Attic Comedy’, G&R, 23: 49–60.
Beacham, R. 2013. ‘Heron of Alexandria’s “Toy Theatre” Automaton: Reality,
Allusion and Illusion’, in K. Reilly, ed., Theatre, Performance and Analogue
Technology: Historical Interfaces and Intermedialities (Basingstoke), 15–39.
Berryman, S. 2003. ‘Ancient Automata and Mechanical Explanation’, Phronesis,
48.4: 344–69.
2007. ‘Teleology without Tears: Aristotle and the Role of Mechanistic
Conceptions of Organisms’, Canadian Journal of Philosophy, 37.3: 351–69.

2009. The Mechanical Hypothesis in Ancient Greek Natural Philosophy
(Cambridge).
Bolzan, J. 2009. ‘Socratis et Socraticorum Epistolae: studi preliminari, traduzione,
commento’. Unpublished PhD thesis, University of Padua.
Brett, G. 1954. ‘The Automata in the Byzantine “Throne of Solomon”’, Speculum,
29.3: 477–87.
Butler, M. 1994. Mary Shelley: Frankenstein, or the Modern Prometheus. The 1818
Text (Oxford).
Cambiano, G. 1994. ‘Automaton’, Studi Storici, 35: 613–33.
Cameron, J. 1986. Aliens (20th Century Fox).
Campbell. D. B. 2003. Greek and Roman Siege Engines, 399 bc to ad 363 (Oxford).
Carra de Vaux, B. 1891. ‘Notice sur deux manuscrits arabes’, Journal Asiatique, 17:
287–322.
Carra de Vaux, B. B., ed. 1902. Le Livre des appareils pneumatiques et des machines
hydrauliques de Philon de Byzance édité d’apres les versions arabes d’Oxford et de
Constantinople et traduit en français (Paris).
Cuomo, S. 2007. Technology and Culture in Greek and Roman Antiquity
(Cambridge).
Diels, H. 1917. ‘Über die von Prokop beschriebene Kunstuhr von Gaza’,
Abhandlungen der königlich preussischen Akademie der Wissenschaften,
Philosophisch-historische Klasse, 26.7: 1–39.
Drachmann, A. G. 1948. Ktesibios, Philon and Heron: A Study in Pneumatics
(Copenhagen).
1963. The Mechanical Technology of Greek and Roman Antiquity: A Study of the
Literary Sources (London).
Faraone, C. A. 1987. ‘Hephaestus the Magician and Near Eastern Parallels for
Alcinous’ Watchdogs’, GRBS, 28.3: 257–80.
Freeth, T., Y. Bitsakis, X. Moussas et al. 2006. ‘Decoding the Ancient Greek
Astronomical Calculator Known as the Antikythera Mechanism’, Nature,
444: 587–91.
Freeth, T., A. Jones, J. M. Steele and Y. Bitsakis. 2008. ‘Calendars with Olympiad
Display and Eclipse Prediction on the Antikythera Mechanism’, Nature, 454:
614–17.
Grillo, F. 2018. ‘An Edition, Translation and Commentary on Hero of
Alexandria’s Automata’. Unpublished PhD thesis, University of Glasgow.
Hercher, R., ed. 1873. Epistolographoi hellenikoi: Epistolographi graeci, recensuit,
recognovit, adnotatione critica et indicibus instruxit Rudolphus Hercher; acce-
dunt Francisci Boissonadii ad Synesium notae ineditae (Paris).
Hill, D. R., ed. 1976. On the Construction of Water-Clocks: Kitāb Arshimı̄ das fı̄
‘amal al-binkamāt, Turner and Devereux Occasional Papers 4 (London).
Huffman, C. 2005. Archytas of Tarentum: Pythagorean, Philosopher and
Mathematician King (Cambridge).
Keenan-Jones, D., Isabel A. Ruffell and E. McGookin. 2016. ‘Taking a Bearing on
Hero’s Anti-crane and Its Un-windlass: The Relationship between Hero of
Alexandria’s Mobile Automaton and Greco-Roman Construction

Machinery’, in J. DeLaine, S. Camporeale and A. Pizzo, eds., Arqueología de
la construcción V: Man-Made Materials, Engineering and Infrastructure,
Anejos de Archivo Español de Arqueología 77 (Mérida), 167–84.
Keyser, P. T. 2014. ‘Kallixeinos of Rhodes’, BNJ, 627.
Lendle, O. 1983. Texte und Untersuchungen zum technischen Bereich der antiken
Poliorketik (Wiesbaden).
Lewis, M. J. T. 1997. Mill-Stone and Hammer: The Origins of Water Power (Hull).
Marsden, E. W. 1971. Greek and Roman Artillery: Technical Treatises (Oxford).
Marshall, C. W. 2003. ‘Sophocles’ Nauplius and Heron of Alexandria’s
Mechanical Theatre’, in A. H. Sommerstein, ed., Shards from Kolonos:
Studies in Sophoclean Fragments (Bari), 261–79.
Mayor, A. 2018. Gods and Robots (Princeton).
Nation, T. 1978–81. Blake’s 7 (BBC).
Noble, J. V. and D. J. de Solla Price. 1968. ‘The Water Clock in the Tower of the
Winds’, AJA, 72.4: 345–55.
Nussbaum, M. C. 1976. ‘The Text of Aristotle’s De Motu Animalium’, HSPh, 80:
111–59.
1978. Aristotle’s De motu animalium: Text with Translation, Commentary, and
Interpretive Essays (Princeton).
Orelli, J. K. 1815. Socratis et Socraticorum, Pythagorae et Pythagoreorum quae
feruntur epistolae: graece / . . . recensuit notis . . . et suis illustravit, versionem
Latinam, . . . dissertationes . . . adjecit Io. Conradus Orellius (Leipzig).
Pellegrino, M. 2000. Utopie e immagini gastronomiche nei frammenti dell’Archaia
(Bologna).
Prager, F. D., ed.1974. Philo of Byzantium: Pneumatica, The First Treatise on
Experimental Physics: Western Version and Eastern Version (Wiesbaden).
Prou, V. 1884. ‘Les théâtres d’automates en grèce au iie siècle avant l’ère chrétienne
d’après les ΑϒΤΟΜΑΤΟΠΟΙΙΚΑ d’Héron d’Alexandrie’, Mémoires présentés
par divers savants a l’Académie des Inscriptions et Belles-Lettres de l’Institut de
France, 9.2: 117–274.
Rehm, A. 1937. ‘Antike Automobile’, Philologus, 92: 317–30.
Rice, E. E. 1983. The Grand Procession of Ptolemy Philadelphus (Oxford).
Robertson, M. 1977. ‘The Death of Talos’, JHS, 97: 158–60.
Roby, C. A. 2016. Technical Ekphrasis in Greek and Roman Science and Literature:
The Written Machine between Alexandria and Rome (Cambridge).
Rose, V., ed. 1870. Anecdota Graeca et Graecolatina: Mitteilungen aus Handschriften
zur Geschichte des griechischen Wissenschaft, vol. 2 (Berlin).
Ruffell, I. A. 2000. ‘The World Turned Upside Down: Utopia and Utopianism in
the Fragments of Old Comedy’, in J. Wilkins and D. Harvey, eds., The Rivals
of Aristophanes: Studies in Athenian Old Comedy (London), 473–506.
2014. ‘Character Types’, in M. Revermann, ed., The Cambridge Companion to
Greek Comedy (Cambridge), 147–67.
2016. ‘Tragedy and Fictionality’, in Isabel A. Ruffell and L. I. Hau, eds., Truth
and History in the Ancient World: Pluralising the Past (London), 32–54.

Schiefsky, M. J. 2005. ‘Technical Terminology in Greco-Roman Treatises on
Artillery Construction’, in T. Fögen, ed., Antike Fachtexte. Ancient Technical
Texts (Berlin), 253–70.
Schmidt, W. 1904. ‘Aus der antiken mechanik’, Neue Jahrbücher für das Klassische
Altertum, Geschichte und Deutsche Literatur und für Pädagogik, 13: 329–51.
Schmidt, W., L. Nix, H. Schöne and J. L. Heiberg, eds. 1899–1914. Heronis
Alexandrini opera omnia quae supersunt (Leipzig).
Schomberg, A. 2008. ‘Ancient Water Technology between Hellenistic Innovation
and Arabic Tradition’, Syria, 85: 119–28.
Schürmann, A. 1991. Griechische Mechanik und antike Gesellschaft: Studien zur
staatlichen Förderung einer technischen Wissenschaft (Stuttgart).
Simms, D. L. 1995. ‘Archimedes the Engineer’, HTechn, 17: 45–111.
Sykutris, J. 1933. Die Briefe des Sokrates und der Sokratiker (Paderborn).
Taplin, O. 1977. The Stagecraft of Aeschylus (Oxford).
Tybjerg, K. 2003. ‘Wonder-Making and Philosophical Wonder in Hero of
Alexandria’, SHPS, 34.3: 443–66.
Wiedemann, E. and F. Hauser. 1918. ‘Uhr des Archimedes und zwei andere
Vorrichtungen’, Nova acta academiae caesareae leopoldino-carolinae
Germanicae naturae curiosorum, 103.2: 159–202.
Wiles, D. 1991. Masks of Menander: Sign and Meaning in Greek and Roman
Performance (Cambridge).

p a r t ii
The Technological Body

chapter 4
Technical Physicians and Medical Machines

in the Hippocratic Corpus
Maria Gerolemou
Introduction
While there are some studies on medical tools in classical antiquity, only
a few seem to deal with what is called in scholarship medical machines such
as the Hippocratic ladder, the Hippocratic board or the Hippocratic
bench.1 The function of medical machines as instruments of traction,
succussion or focal pressure is discussed in On Fractures and On Joints,
which are two of the oldest works in the Hippocratic corpus,2 and also in
Mochlikon.3 These works provide an assessment of the technological possi-
bilities and consequences that the use of medical technology entails.
In classical antiquity, the ‘technical’ could be explained in two ways: The
first refers to technology as manual art, or, in other words, to lowly or
undignified work. The second one denotes a skill or craft and relates to the
mathematical science of mechanics, established as a theoretical and math-
ematical understanding of machines during the Hellenistic period. The
way technological means in general are presented in the Hippocratic
corpus does not, however, portray medicine as science, that is, it does
not signal a ‘scientific turn’, as De Groot clearly points out in this volume
1
See Schneider 1989, 222–7, on Hippocratic medical machines. On Hellenistic medical machines, see
Drachmann 1963, 171–84; Wilson 2008, 345–6; Berrey 2017a, especially on Andreas’ medical
machine, and Berrey 2017b. On surgical tools, see Mile 1907; Tabanelli 1958; Künzl 1996; Jackson
2005; Gazzaniga and Marinozzi 2015, 27–31. Bliquez 2014; Le Blay 2016.
2
Withington 1928, 84–5. Witt 2018, 227–8, 230
3
The On Joints, On Fractures and Mochlikon describe three reduction methods, however, not system-
atically and not always in direct relation to medical devices. The first method, called katatasis, traction,
is applied to or across the fracture or the dislocated bone, in order to manipulate the bone fragments to
return to their natural position. This method is usually applied to the arms and legs, the neck, the
backbone, or the pelvis (although we also have axial traction which corrects spinal deformities).
The second, succussion (kataseisis), is the vigorous shaking of the skeleton, with the purpose of healing
dislocations and deformities, mostly related to spine and hip injuries. The application of focal pressure
(katanagkasis) for the reduction of bone deformities is the third, less complicated, healing method (for
instance over the kyphotic area).
107

108 maria gerolemou
(Chapter 5). This assumption is based on the fact that both the invention
and use of machinery by physicians are anchored to needs of a practical
nature; hence, neither are Hippocratic medical machines linked to scien-
tific principles nor did they try to promote new scientific knowledge.4 By
examining the degree to which physicians transfer knowledge and skills
from the technical to the medical domain, we rather come to the conclu-
sion that Hippocratic physicians, while they apply methods of technical
knowledge to medicine, that is, by integrating the two domains when
designing medical technical solutions, do not act like engineers.
Moreover, medical technology is perceived as rendering doctors poor
healers, due to the fact that the main focus lies on the ‘sick’ body parts,
which they treat with devices, instead of caring for the patient’s body as
a whole (see e.g. Art 13).5 At the same time, practical and ethical issues are
raised in substituting or enhancing the physician’s manual skills with
devices. Physicians who choose to include medical machines in their treat-
ment are described as mere charlatans, since they are considered to care more
about increasing their reputation through impressive but unnecessary and
dangerous techniques than curing the patient. Such dismissal of medical
technology is premised on the Hippocratic idea that the body’s treatment is
better ensured with rational, practical means that build a relationship of trust
between the physician and the patient.6 Finally, by depending merely on the
physician’s physical strength, the use of medical devices machinery could
impede the development of the physician’s cognitive abilities and skills,
which are essential for a medical treatment. To this end, the aforementioned
Hippocratic texts call for ‘patient-friendly’ medical care and warn that
medical technology could strip medicine of its humanistic qualities. In
most cases, they also recommend the use of machines only if the application
of strength by the physician’s hand or other body parts which subject the
body to tension or traction proves to be unsatisfactory.
This chapter offers first a brief overview of the Hippocratic notion
of mêchanê by presenting three different types of machines (a. natural
machines; b. simple machines; c. complicated machines), asking in par-
ticular what the saying ‘patient-friendly or harmful machine’ could mean
in this context. In the second part of the chapter, I turn my attention to
three examples of complicated Hippocratic machines (the Hippocratic
board, ladder and bench) and suggest ways of approaching them. Crucial
4
See Lonie 1981, on the limited use of mechanics in the Hippocratic corpus; Berryman 2009, 67. See
also Schneider 1989, 226–7.
5
See recently Bartoš 2020 on Hippocratic holism.
6
See On the Physician 2 and Bliquez 2014, 19.

Technical Physicians and Medical Machines 109
to this discussion is, on the one hand, the possibility of augmenting the
physician’s strength and skills through the use of medical machinery, and
on the other hand, the possible unwanted consequences of such
technologies.
Machines in the Hippocratic Corpus

Before elaborating on medical machines and drawing on the issues raised from
their use, it is germane to our subject to consider how the notion of mêchanê is
captured in the Hippocratic corpus. In classical antiquity, the term μηχανή
could borrow from a diverse panoply of meanings: it could refer to the ‘means’
of accomplishing a task (see e.g. Aristophanes Thesmophoriazusae 765), to
a trick (Hesiod Th. 146) or to a real device (see e.g. Herodotus 2.125, mêchanê
as a lifting device), which in the Hellenistic period could even designate the
capacity of device to perform work, independently of constant human inter-
vention.7 The latter – the case of the real device – occurs frequently in the
Hippocratic collection, specifically in the surgical treatises here under discus-
sion, and refers not only to actual medical apparatuses, but also to their
making and use by physicians and their assistants (see the terms μηχάνησις,
μηχανοῦμαι and μηχανοποιέω in e.g. On Fractures 30; On Joints 42, 72). In
terms of real machines, Hippocratic treatises refer to both simple machinery,
which could be accessible made out of natural material (referring to the
physician’s, his assistant’s or the patient’s body parts or other natural mater-
ials) – or of simple construction (e.g. simple tools, such as a lever), and more
complicated devices (such as ladders, constructed beds). Simple and compli-
cated machines are designed to work by the application of power (ischus) set in
motion by a single person or several (the physician and his assistants),
depending on the construction and, consequently, on the machine’s size
and weight. Hence, the machine’s performance, which could be proved to
be beneficial or harmful to the patient, is contingent both on the physician’s
skills and on the machine’s properties.
Generally, medical machines in the Hippocratic corpus are appraised
according to their usability: they must be easy to find and easy to use
(euchrêstia). On the Surgery suggests that medical technology (here called
organon) must be small or handy, not to impede the physician’s activity and,
above all, to be managed easily by an operator’s hand (5, μηδὲ ἐμποδὼν τῇ
ἀναιρέσει, παρὰ τὸ ἐργαζόμενον δὲ τοῦ σώματος; see also On the Surgery 2). In
a similar vein, the author of Physician 2.19–20 argues that it is crucial for
7
On the meanings of mêchanê, see Schneider 1989, 217–22; Schiefsky 2007, 77–8.

110 maria gerolemou
a device – referring to swabs, sponges and bandages – to be light, fine and easy
to handle (τὰ δ’ ὄργανα πάντα εὐήρη πρὸς τὴν χρείαν ὑπάρχειν δεῖ τῷ
μεγέθει καὶ βάρει καὶ λεπτότητι). At On Fractures 30, a good mechanical effect
largely depends on well-arranged and easy-to-use machinery (εὔχρηστον τὸ
μηχάνημα). In his commentary on the Hippocratic On Joints 3,8 the empiricist
Apollonius of Citium emphatically expresses his approval of practical experi-
ence with regard the use of medical devices (ἡ περὶ τὰ ὄργανα ἀπειρία, on the
other hand, inexperience in the use of medical devices derives from ἀχειρία,
clumsiness);9 specifically, in describing various methods of bone reductions
which include machines, such as with small hard balls or a small ordinary
ladder, he also uses the term euchrêstos, practical and ready to be used anywhere,
such as in the palaestra or in remote locations, since devices of this kind do not
need particular expertise in their construction as well as use.
Now, as must have become clear by now, although the use of technical
means in treatments is generally labelled mechanopoiia (see e.g. On Fractures 15,
30), there seems to be a difference between reduction methods using light and
simple technical means and methods using heavier and more complicated
ones. In the Hippocratic treatise Decorum, one of the latest works in the
Hippocratic corpus, probably dated around the first century ce,10 the author
argues that a physician must be well prepared before starting to work with
a patient. Specifically, the physician’s case should include instruments, appli-
ances and small pocketknives (8, προκατηρτισμένα ὄργανά τε καὶ μηχαναὶ καὶ
σίδηρος).11 A physician should also have, according to the same text, a second,
‘simpler bag’ called parexodos (παρέξοδος ἡ λιτοτέρη – ἡ διὰ χειρέων), which
he can carry with him during a journey (ἀποδημία). The text does not refer to
the contents of this second bag; however, assuming that this must be lighter so
as to be carried everywhere by the physician, one may naturally speculate that
heavier tools have no place there.12
As noted earlier in this chapter, three categories of medical machines
appear in the Hippocratic corpus. The first consists of natural means:
physicians could be using simple instruments made out of natural mater-
ials, like wood, wool, linen, rod and even the hands, feasts, heels and other
8
Kollesch and Kudlien 1965.
9
On medical empiricism in classical antiquity, see, among others, Deichgräber 1965; Frede 1987,
1988; and Waldow 2010.
10
Craik 2015, 59.
11
See Bliquez 2014, 17–18, 24, on medical leather pouches, wooden and cylindrical cases made of
copper.
12
Ἄρμενα could mean generally implements (Fract. 31 and Galen Hipp. Off. Med. 717); however, in
Erotian fr. 37 (sv) ἄρμενα are tools, ἐργαλεῖα, such as σμιλία, scalpel, φλεβοτόμα, tool for opening
the veins, καὶ τἆλλα τὰ τούτοις ὅμοια, ‘and other similar things’.

parts of their bodies or of their assistant’s or the patient’s bodies, which are
practically turned into tools.13 In On the Surgery 4, the physician’s hands
and fingers are actually considered among other tools. This group of
devices, apart from being ready to be used, seems to work in total accord-
ance, for instance, with the bone’s natural inclination, since these devices
do not have the power to move beyond this, and, therefore, their use is
uncomplicated and regarded as safe. For instance, at the beginning of On
Joints, the methods of healing dislocated shoulders are being discussed.
According to the text, such cases could be sufficiently treated by the natural
power of both the physician and the patient. For example, the physician,
by placing his
fingers under the armpit inside the head of the dislocated bone . . . should
force it away from the ribs, thrusting his head against the top of the shoulder
to get a point of resistance . . . and with his knees thrusting against the arm at
the elbow, should make counter-pressure towards the ribs; it is well for the
operator to have strong hands or, while he uses his hands and head in this
way, an assistant might draw the elbow to the chest. (2)14
What the physician’s body parts simulate on this occasion is the activities of
a real device; this is reminiscent of Aristotle’s statement on the hand and its
shape:
Hands are an instrument . . . a hand is not all of one piece, but it branches
into several pieces [. . .] it is possible to use them [sc. the fingers] a singly, or
two at a time, or in various ways. Again, the joints of the fingers are well
constructed for taking hold of things and for exerting pressure.15 (On the
Parts of Animals, 687a–b)
To return to Hippocratic machineries: In this first category of machines,
body parts and tools are regarded as remarkably similar.16
13
On this see also De Groot in this volume, Chapter 5.
14
τὸν αὐτὸν δὲ τρόπον τοῦτον καὶ ὁ ἰητρὸς ἂν ἐμβάλλοι, εἰ αὐτὸς μὲν ὑπὸ τὴν μασχάλην ἐσωτέρω
τοῦ ἄρθρου τοῦ ἐκπεπτωκότος ὑποτείνας τοὺς δακτύλους ἀπαναγκάζοι ἀπὸ τῶν πλευρέων
ἐμβάλλων τὴν ἑωυτοῦ κεφαλὴν ἐς τὸ ἀκρώμιον ἀντερείσιος ἕνεκα, τοῖσι δὲ γούνασι παρὰ τὸν
ἀγκῶνα ἐς τὸν βραχίονα ἐμβάλλων, ἀντωθέοι πρὸς τὰς πλευράς—συμφέρει δὲ καρτερὰς τὰς χεῖρας
ἔχειν τὸν ἐμβάλλοντα—ἢ εἰ αὐτὸς μὲν τῇσι χερσὶ καὶ τῇ κεφαλῇ οὕτω ποιοίη, ἄλλος δέ τις τὸν
ἀγκῶνα παράγοι παρὰ τὸ στῆθος.
All translations of On Joints and On Fractures are taken from Withington 1928. The Greek text
used here is also taken from the same source.
15
Translated in Peck and Forster 1937. See also Galen, Usefulness of the Parts of the Body 3.5, αἱ χεῖρες δ’
ὄργανον
16
Demont 2014 claims that the term organon ‘could be used for different kinds of tools or instruments
(often musical instruments) and metaphorically concerning “(bodily) organs”’ (p. 13). At the On the
Art 8, the term organon is probably used in a general sense (for both tools and machines). See

112 maria gerolemou
The second category of medical machines includes more advanced tools,
such as pestles, windlasses, levers, axles, cupping vessels, wineskins and so
on.17 At On Fractures 31, the author states that the axle, the lever and the
wedge (περιαγωγὴ, καὶ μόχλευσις, καὶ σφήνωσις) are the most powerful
apparatuses contrived by men; specifically, ‘without someone, indeed, or all
these, men accomplish no work requiring great force’ (Ἄνευ δὲ τούτων, ἢ
ἑνὸς δή τινος, ἢ πάντων, οὐδὲν τῶν ἔργων τῶν ἰσχυροτάτων οἱ ἄνθρωποι
ἐπιτελέουσιν).18 The operation of such equipment largely depends on the
energy produced by the physical force applied by the physician, his assistants
or even the patient; that is, although those implements are, as Hero later puts
it (Mechanics 2.5; see also Pappus 8 page 1122), autotelês, self-reliant, as far as
their construction is concerned, their function as part of the treatment does
not run autonomously. In other words, they – and other devices of similar
construction – cannot release enough power on their own to restore, for
instance, a bone to its natural position. This is what we learn from On Joints
77, where a treatment with an askos, wineskin, cannot bring back the joint
into its normal anatomical position, due to its limited power (Εἴ τε οὖν τις
μικρὸν ἐνθήσει τὸν ἀσκὸν, μικρὴ ἡ ἰσχὺς ἐοῦσα ἀδύνατος ἔσται ἀναγκάζειν
τὸ ἄρθρον).19 In other instances, a small force could be beneficial for the
skeleton, hence instruments of this type are considered safer. The author of
On Joints refers to an instance of a dislocated shoulder bone and describes
its treatment with an ambê, a projecting rim.20 He argues that this must
be placed under the armpit to push the bone back into the socket. As he
states, ‘this is by far the most powerful method for reducing the shoulder,
for it makes the most correct leverage, if only the instrument is well on
the inner side of the head of the humerus. The counterpoise is also most
correct and without risk to the bone of the arm’ (7).21 What is being
Webster in this volume, Chapter 6, and hsi forthcoming book for more on the meaning of tools and
implements in the Hippocratic corpus.
17
On levers, mochloi, see On Fractures 13, 31, 35; On Joints 7, 72–3; on windlasses, onoi, oniskoi, see On
Fractures 42, 72; On Joints 43; on plêmnai, naves of a wheel, see On Fractures 13; trochiliae, pulleys, in
On Fractures 43; sideria, iron rods, On Fractures 31.
18
See Bliquez 2014, 40. For this reason, a physician can possess several of this type of devices. In the
treatise On Fractures at 31, the surgeon is urged to have levering devices prepared (or to prepare for
himself, χρὴ ποιέεσθαι). These elevators should be prepared in several sizes, each being broader at
one end and narrower on the other (τὸ μέν τι πλατύτερον, τὸ δέ τι στενότερον), and they should be
strong, so that they do not bend. On simple and complex machines, see also Schiefsky 2007 and
2008 and Wilson 2008.
19
‘If one inserts a small bag, its power being small, it will be unable to reduce the joint.’
20
On the various forms of ambê, see Eratian s.v.
21
δικαιότατα μὲν γὰρ μοχλεύει, ἢν καὶ μοῦνον ἐσωτέρω ᾖ τὸ ξύλον τῆς κεφαλῆς τοῦ βραχίονος·
δικαιόταται δὲ αἱ ἀντιῤῥοπαί, ἀσφαλέες δὲ τῷ ὀστέῳ τοῦ βραχίονος. As sufficient treatment similar
to this is the one described with a Thessalian chair or a double chair in On Joints 7.

assessed here is not merely medical technology per se but a mechanical
system that consists of certain machineries, the physician and the patient,
where everything – that is, the device, the physician and the patient – is
seen as mechanisms that produce force. What may prevent the effective
function of this system, besides possible technical malfunction of the
devices, is the threat that these devices may substitute the physician’s
strength and skills at the expense of the patient.
This threat becomes more obvious when medical treatments include
heavier, complicated devices: this refers to devices which consist of simple
parts (simple tools and they are constructed on the spot, easily and
immediately for serving specific needs. As Oribasius later in the fourth
century ce notes, medical devices like the plinthion of the Hellenistic
physician Neileus or the Hippocratic bench belong to this category
(49.4.23); their axles are described as autokinêtos, exposed and put into
motion by a single person, in contrast to other machines, like the trispaston
of Apellis or Archimedes or the glossokomos of Nymphodorus, whose axles
are covered and their motion generated internally through other mechan-
isms (Medical Collections 49.4.24, οἱ μὲν οὖν ἔκθετοι ἄξονες αὐτόθεν διὰ
τῶν χειρῶν στρέφονται, οἱ δὲ κρυπτοὶ διά τινων ἑτέρων μηχανημάτων,
and 49.5.15).22 In the latter case, the role of the operator/physician and
patient during the treatment seems to be even more limited. The hidden
mechanism often followed by a feeling of admiration for the well-hidden
cause and mechanical ingenuity behind technological operation is a topos in
Greek and Roman mechanical treatises (see e.g. ps.-Aristotle Mech. 848a
34–7, Hero’s Aut. 17.1 20.2–3, 28.2–3).23 In the case of medical
machines, however, the hidden mechanisms could point towards an
inexplicable cause of operation and a passive user/physician. In these
cases, even if the physician who uses them means well and is well
trained, medical devices could unexpectedly produce force beyond his
control and what is needed for the treatment of the body. Hence, such
machines could cause further deformations and fractures, forcing, for
instance, a bone to go para phusin (against its δικαία φύσις, see On
Fractures 1.6–7). On the other hand, simple medical devices entail fewer
22
Compare Vitruvius, at On Architecture 10.1.3, on machines acting mechanically (they contain many
subordinate parts, and, thus, are propelled by a greater power) and on those acting organically (they
are less complicated and are set in motion by a single person).
23
Tybjerg 2003, 451, and Bosak-Schroeder 2016, 128, argue specifically regarding Hero that hiding
mechanisms and techniques increase the feeling of wonderment (thauma). In Gerolemou 2022,
I argue that in Hero wonderment lies also in the explanation of the mechanical spectacle (chapter 3).

114 maria gerolemou
dangers of this kind, as they are designed to merely give a mechanical
advantage to natural aids such as the physician’s hands.
Oribasius, who summarizes bone reduction methods and surgical
machines from Hippocrates’ era onwards, refers in book 49.1 to
Heliodorus’ categories of medical devices (first century ce), which are
reminiscent of the aforementioned three Hippocratic types, tropoi, of bone
healing. The first, palaistrikos, depends on the physician’s hands (διὰ
ψιλῶν χειρῶν τελειούμενοι) and treats recent bone injuries.24 The force
that the physician could apply on the bone with his hands depends on the
patient’s age and gender (49.1.1). The second category is called methodikoi
katartismoi; this category makes use of simple tools, similar to the ones we
have noted above (διὰ τῶν κοινῶν τοῦ βίου ἐργαλείων συντελούμενοι),
which release greater power than the physician’s bare hands, and they can
be applied to everyone, independent of gender or age, as well as to older
injuries. A third category is called organikoi katartismoi. According to
Heliodorus, these are more effective (praktikoteroi) than the methodikoi;
they can handle every injury, even the ones that seem incurable (49.1). The
last category, the organikoi reduction methods, works with medical
machines: specifically, it includes tonia, extension machines, stasima,
upright machines, μηχανήματα, machines bigger than the tonia but smal-
ler than other devices, and ephedrana, sitting apparatuses (49.2), which all
practise extension and levering; they are constructed by various simple
tools, such as rods, ropes, wheels, screws, axles, chains, pivots and so on
(49.4; see also Mochlikon 38).
Medical machines that belong to the last two categories are used when
greater force is required. The same is true, as I have already shown, for the
Hippocratic medical machinery. So, the author of On Fractures, discussing leg
injuries, argues that ‘If one intends to do the work well and skillfully, it is
worthwhile to have recourse to mechanism, that the fractured part may have
proper but not violent extension. It is especially convenient to use mechanical
treatment for the leg’ (30).25 According to the text, the use of a mechanical
device could have proper and not violent results, δικαίη καὶ μὴ βιαίη. As von
Staden remarks, on the other hand, technê’s violence suggests that art could
violate phusis and transgress its boundaries.26 For this reason, the mechanical
24
See specifically On Joints 4, where something similar is described; here, reduction methods, useful
for the palaestra, are presented that do not require further apparatuses.
25
εἰ δέ τις μέλλοι καλῶς καὶ εὐχερῶς ἐργάζεσθαι, ἄξιον καὶ μηχανοποιήσασθαι, ὅκως κατάτασιν
δικαίην καὶ μὴ βιαίην σχήσῃ τὸ κατεηγὸς τοῦ σώματος· μᾶλλον δὲ ἐν κνήμῃ ἐνδέχεται
μηχανοποιεῖν.
26
Von Staden 2007, 28.

solution is described on several occasions in the Hippocratic collection as
anankê, both force, and need (see e.g. On Fractures 15, τῶν ἀναγκῶν τινὰ
προσφέρειν, offer a mechanical aid; see also 16 on the σωλήν: οὐ γὰρ
ἀναγκάζουσι οἱ σωλῆνες ἀτρεμεῖν, ‘the hollow splints do not compel
immobility’).27 This term emphasizes not only what marks the need for the
device’s invention and how physicians have devised to meet that need, but also
how machine power could force, ἀναγκάζειν, the body to act according to the
laws of art and not of nature.28
A further conclusion can be drawn from the language of force. Similar to
ἀνάγκη, ἰσχύς in the Hippocratic corpus is defined as a natural force
inherent in the body,29 which could, however, be enhanced, προστιθέναι
διὰ τέχνης, by art (Regimen 1.2).30 In the Hippocratic surgery treatises,
ἰσχύς specifically serves bodily schêmata, postures, which could be natural
or the result of various technae, arts, specifically of their armena, the
apparatuses used, and the erga, the tasks they can accomplish (see On
Fractures 2; On the Surgery 15).31 At On Fractures paragraphs 1 and 2, the
author refers to cases of physicians whom he calls σοφιζόμενοι, the theor-
izers. They wrongly attempt to treat fractured arms by forcing their
patients to hold a specific posture which they speculate that is the correct
one. The author argues that they could have let the patient hold their arm
in the way they felt was right instead (or according to Galen, κατὰ τὸ
ἀνωδυνώτατον σχῆμα διδασκόμενος ὑπὸ τῆς φύσεως, at In Hippocratis
librum de fracturis commentarii iii, p. 337).
27
On Joints 7, often on bandages 14, αἵ τε γὰρ ἐπιδέσιες οὐδέν τι μᾶλλον προσαναγκάζουσιν ἢ
ἀπαναγκάζουσιν, 16, see further at 47, the Hippocratic board is called κατασκευή τοῦ
διαναγκασμοῦ; 48, 72, 73, ἀνάγκαι ἰσχυραὶ, 74 on levering, 75, τῇ σανίδι καταναγκάζειν, 77,
reduction through a bag; see also Apollonius On Joints 21.
28
On anagkê in the Hippocratic treatises, see von Staden 2007, remarking that the plural form of the
term indicates torture; see, for example, Herodotus 1.116; Antiphon 6.25; Thucydides 1.99; Polybius
15.28.2; Herodas 5.5.
29
See, for example, On Fractures 31, ἄρμενα . . . ἰσχυρότατά; On Joints 47, ξύλον ἰσχυρὸν, strong
broad plank. On that, Schneider 1989, 224–5.
30
See Fractures 2, 19; Joints 77; compare On Art 8, on gaining control through tools provided by nature
or via the tools of the technai, ὧν γὰρ ἔστιν ἡμῖν τοῖσί τε τῶν φυσίων τοῖσι τε τῶν τεχνέων ὀργάνοις
ἐπικρατεῖν.
31
On schêmata and dunameis, see a definition at On Ancient Medicine 22: ‘What I mean is roughly that
a “power” is an intensity and strength of the humours, while “structures” are the conformations to be
found in the human body, some of which are hollow, tapering from wide to narrow; some are
expanded, some hard and round, some broad and suspended, some stretched, some long, some close
in texture, some loose in texture and fleshy, some spongy and porous.’ λέγω δέ τι τοιοῦτον, δύναμιν
μὲν εἶναι τῶν χυμῶν τὰς ἀκρότητάς τε καὶ ἰσχύν, σχήματα δὲ λέγω ὅσα ἔνεστιν ἐν τῷ ἀνθρώπῳ, τὰ
μὲν κοῖλά τε καὶ ἐξ εὐρέος ἐς στενὸν συνηγμένα, τὰ δὲ καὶ ἐκπεπταμένα, τὰ δὲ στερεά τε καὶ
στρογγύλα, τὰ δὲ πλατέα τε καὶ ἐπικρεμάμενα, τὰ δὲ διατεταμένα, τὰ δὲ μακρά, 10τὰ δὲ πυκνά, τὰ
δὲ μανά τε καὶ τεθηλότα, τὰ δὲ σπογγοειδέα τε καὶ ἀραιά.

116 maria gerolemou
By choosing to follow treatments that are outlandish, non-customary
(ξενοπρεπείς) and bizarre (ἀλλόκοτοι) rather than effective, this category of
physicians exhibits, on the one hand, wisdom in terms of theory; however,
at the same time, they prove to be inexperienced in terms of practice (see
also On Fractures 3), hence they cause additional pain to the patient. For
instance, the archer position that the theorizing physicians seem to favour
in bandaging a fractured arm ignores a simple principle: that nature, like
the arts and all those things which are executed by natural strength or
artifice, presupposes a variety of positions or schêmata developed according
to the various ways in which the bodies of living beings act (On Fractures 2).
In particular, the author of On the Joints 8 speaks, referring to joint sockets
and to the attachment of ligaments, of φύσιες φυσίων, of natural diversities.32
Explicit knowledge of these kinds of natures is crucial when treating a patient
and, moreover, when designing a treatment which includes machineries.
Art in general, and medical technology in particular, allow the physician
to change or repare incorrect bodily schêmata. According to the author of
On Joints 71, the majority of physicians, who attempt to do that, do not
yield to ordinary devices (τυχούσα παρασκευή); therefore, the author
remarks, at least they ‘should know the most powerful methods which
the whole art provides for each case, and use them severally where they
seem appropriate’ (διὰ τοῦτο ἐπίστασθαι μὲν χρὴ τὰ κράτιστα περὶ
ἑκάστου ἐν πάσῃ τῇ τέχνῃ· χρῆσθαι δὲ οἷσιν ἂν δόξῃ ἑκάστοτε). The
balance among the members involved in such treatment (see e.g. On Joints
73), that is, the device, the patient and the physician (and his assistants; see
On the Surgery 2, ὁ ἀσθενέων· ὁ δρῶν· οἱ ὑπηρέται· τὰ ὄργανα, etc.), can
change according to the amount of energy that the device releases and the
patient is able to receive. The ability of the physician to manipulate or
control the machine also plays an important role.
What we learn from the above cases of instruments and machines is that
they should be tailored to patient-specific needs. Any thoughtless applica-
tion of medical technology can have serious adverse consequences for the
patient (see e.g. On Joints 73). For this reason, medical texts advise that one
should avoid using devices when human labor and power could be used
effectively instead. In paragraph 48 of On Joints, the author discusses
possible methods of treating a bone injury from a fall or from lifting
a heavy weight and how the physician must be in a certain position to
correctly use the respective medical equipment. Specifically, the author
warns that the use of cupping instruments for removing the depressed
32
See von Staden 2007, 22, 25.

vertebrae (repontes sponduloi) constitutes a great error of judgement, for
these instruments are used for pushing in rather than drawing out; mistakes
in the use of medical technology show that the physician often lacks the
appropriate expertise in applying it safely.33
Thus, although the texts under discussion do not offer a comprehensive
and systematic knowledge about the materials, construction and function
of machines (as in the work of the Hellenistic engineers), they do discuss
their role in the healing process, their ideal operators, their effects on
patients and their impact generally on treatment. In other words, the use
of medical machines in Hippocratic healing practice forms an ethical issue
rather than an epistemological one. This is the topic I explore in the next
section. This will lead us to the examination of other, closely related
implications for the relationship between physicians, patients and medical
art, and, more explicitly, for the norms that guide medical intervention in
the body qua instruments.
Examples of Hippocratic Medical Machines

For the correction of spinal deformities with a board, in addition to the
plank (xulon) which is fixed on the ground or a board (stulos) placed parallel
to a wall, parts of the physician’s body and various medical tools are also
used. In most cases, a physician or a trained assistant uses his hands, feet or
even his whole body to exercise pressure on the bone deformity, while
traction is applied with the help of bands, axles, straps and wheels (On
Joints 47; see also Mochlikon 38). In some cases, if everything is well
arranged, no great harm can be done, unless one deliberately wants to do
harm (On Joints, 47, εἰ χρηστῶς σκευασθείη, εἰ μὴ ἄρα ἐξεπίτηδές τις
βούλοιτο σίνεσθα);34 in others, the success of the treatment depends on the
strength and knowledge of the physician and his assistant (ἰσχυρὸς καὶ μὴ
ἀμαθής). With the help of wheels and axles, in other cases, the extent of the
force produced through the board can be easily regulated (On Joints 47, καὶ
33
Similarly, in paragraph 9 of the On the Physician, the author differentiates between devices that one
should be trained to use, like cupping instruments, and devices, like forceps for drawing teeth, that
are completely uncomplicated and easy to use by anyone.
34
To cite another case, at On Fractures 30, the author argues against the use of circlets (sphaerae) for
healing bone fractures and states that mechanical fallacies sometimes occur because of the arrange-
ment of the device. In particular, he says ‘if then the circlets are supple, of good quality, soft and
newly sewn, and the extension by the bent rods suitably regulated as just described, the mechanism is
of good use, but if any of these things are not well arranged it will harm rather than help. Other
mechanisms also should either be well arranged or not used, for it is shameful and contrary to the art
to make a machine and get no mechanical effect.’

118 maria gerolemou
κατὰ φύσιν γε ἀναγκάζουσι). However, this latter combination, which the
author acknowledges as one of the best and most correct reduction
methods (καλλίους and δικαιοτέρας), ‘has such power that, if one wanted
to use such forcible manoeuvres for harm and not for healing, it is able to
act strongly in this way also.’ That is, the physician must be in a position to
control the machine (On Joints 47).35 For instance, an advanced treatment
method involving the Hippocratic board combined with the use of
a wineskin filled with air, positioned under the patient’s spine, proves to
be unsuccessful exactly because of the incompetence of the practitioner
who used it; he did not blow up the wineskin well (On Joints 47; see also
112–28). The author, however, refers to this case because of the useful
lessons, mathêmata, that this experience can provide (see also On Joints 10,
καὶ οὐκ ἀρκεῖ μοῦνον λόγῳ εἰδέναι τὴν τέχνην ταύτην, ἀλλὰ καὶ ὁμιλίῃ
ὁμιλεῖν, ‘it is not enough to only know the art in theory but also through
practice’).36 The effectiveness of the device is measured here through
errors: specifically through how many errors could users make when
using this machine and how severe these errors are.37
The Hippocratic ladder reduces spinal curvatures, such as scoliosis,
kyphosis, lordosis and so on.38 To achieve reduction, according to the
author of On Joints, the patient must undergo a succussion while being tied
to a ladder; the body has to be in a state of equilibrium; therefore, it should
be fastened, entirely or partially, on the ladder or a surface acting as a ladder
(6–7). The physician has to lift the ladder against some high tower or house
gable. If the spinal deformity is near the neck, the patient is placed on the
ladder in an erect position; if the lower spine has a deformity, then they are
placed on the ladder upside down (On Joints 43; see also for the shoulder
On Joints 7). For this, a solid ground and a well-trained assistant, capable of
lifting the body of the patient, are required. Then, the physician lets the
body come down smoothly and at the same time, with the assistance of
a lowering tackle made by a pulley, a wheel or an axle (see On Joints 76)
prevents the ladder from coming to the ground unevenly, or himself and
the assistants from being pulled forward. The trunk’s weight and the limbs
are the pulling force that straightens the spine.
35
Αὗται αἱ ἀνάγκαι εὐταμίευτοί εἰσι καὶ ἐς τὸ ἰσχυρότερον καὶ ἐς τὸ ἧσσον, καὶ ἰσχὺν ἔχουσι
τοιαύτην, ὥστε, καὶ εἴ τις ἐπὶ λύμῃ βούλοιτο, ἀλλὰ μὴ ἐπὶ ἰητρείῃ, ἐς τοιαύτας ἀνάγκας ἀγαγεῖν,
κἂν τούτῳ ἰσχυρῶς δύνασθαι.
36
The term mathêma here could not refer to mathematics such as in Archytas 1 D-K; see Snell 1923, 77.
37
On the epistemological principle of error in the Hippocratic corpus, Lloyd 1987, 125–6 and 140–42.
38
Galen in Hipp. Fract. p. 338 calls it mêchanikon klimakion and classifies it among the stasima organa.

The risk entailed in each machine is naturally related to the severity of
the injury. For example, the ladder can prove fatal as it deals with spinal
deformities. The author of On Joints warns that this device should only be
used on special occasions. Furthermore, as he puts it, it is disgraceful for
any art in general and especially the art of medicine to make use of such
a troublesome, complicated and high-risk procedure (42; cf. Apollonius
Hp. On Joints 28). This is because even the patient involved in this kind of
treatment must be experienced, so as, with the help of gravity, to success-
fully manipulate the suspension weight into putting pressure somewhere
between the perineum and over the sacrum (On Joints 76). Because of the
danger of this method, this is practised mainly by ignorant physicians
(skaioi):
[Those] who use this method are chiefly those who want to make the vulgar
herd gape, for to such it seems marvellous to see a man suspended or shaken
or treated in such ways; and they always applaud these performances, never
troubling themselves about the result of the operation, whether bad or good.
As to the practitioners who devote themselves to this kind of thing, those at
least whom I have known are incompetent. (42)39
The author of On Joints admits though that the machine’s benefits for
such injuries outweigh the possible complications and shortcomings. He
therefore praises the inventor of this device as well as other mechanical
contrivance that supports a natural way of healing (42, μηχάνημα κατὰ
φύσιν ἐπενοήθη). Apollonius, in his commentary on the Hippocratic
treatise On Joints, mentions that there seems to be two reports on what
the Hippocratic ladder might be: the one supports that the ladder is
a square device which resembles a ladder with a windlass at the bottom;
according to the other, it is a small regular ladder, such as, for instance, the
one builders use at their work. Apollonius agrees with the latter; otherwise,
as he states, if Hippocrates was referring to the first case, that is, to a more
complicated device, he would have referred to its construction process and
never called this kind of method the most drastic (6, κρατίστη). He
concludes by saying that, although he recognizes that succussion with
pulleys and windlasses could be more effective, it is shameful to even
39
χρέονται δὲ οἱ ἰητροὶ μάλιστα αὐτῇ οἱ ἐπιθυμέοντες ἐκχαυνοῦν τὸν πολὺν ὄχλον· τοῖσι γὰρ
τοιούτοισι ταῦτα θαυμάσιά ἐστιν, ἢν ἢ κρεμάμενον ἴδωσιν ἢ ῥιπτεόμενον, ἢ ὅσα τοῖσι
τοιούτοισιν ἔοικε, καὶ ταῦτα κληΐζουσιν αἰεί, καὶ οὐκέτι αὐτοῖσι μέλει ὁποῖόν τι ἀπέβη ἀπὸ τοῦ
χειρίσματος, εἴτε κακὸν εἴτε ἀγαθόν. οἱ μέντοι ἰητροὶ οἱ τὰ τοιαῦτα ἐπιτηδεύοντες σκαιοί εἰσιν, οὕς
γε ἐγὼ ἔγνων. However, as Witt 2018 argues (p. 229) ‘the desire to entertain an audience is not
altogether rejected [sc. in the Hippocratic corpus]’; see, for instance, On Joints 70, where it is stated
that suspension καὶ δή τι καὶ ἀγωνιστικὸν ἔχουσα, ‘has something striking about it’.

120 maria gerolemou
speak at length (ἀηδὲς μὲν καὶ μακρολογεῖν) about these methods (περὶ
αὐτῶν, sc. mechanical solutions, 15). Apollonius’ thinking about the
possible form of the Hippocratic ladder illuminates certain aspects of
scientific change, from an empirical and experimental knowledge of some-
thing being described as a machine to a knowledge of a specific machine.
He, however, advocates some sort of practically oriented conception of
medical devices: if a more elaborate construction could only have
a negative connotation (in terms of a technical aid that intervenes in the
natural order of the body), therefore the Hippocratic ladder cannot be one.
A third example of Hippocratic medical machines is the scamnum or
bench. As the Hippocratic author of On Joints puts it, no joint is incapable
of reduction when the mechanical aid of the scamnum or bench is involved
(72–5; cf. also On Joints 78).40 The Hippocratic bench is an adjustable table
with a mechanism that enables it to extend. This is the only medical
machine in the Hippocratic collection that is described in some detail; in
particular, the author refers to its construction material, length, breadth,
thickness and function. The Hippocratic bench works together with levers
and axles. The author of On Joints 72, by recognizing the great power that
this machine can release, highlights once again the fact that the physicians
who use this must be well trained. He concludes, however, by clarifying
that most cases of bone dislocation are simple, hence they do not need great
pulling forces or the use of such complicated machines (73, ἀσθενεστέρη
κατάτασις καὶ φαυλότερη κατασκευή, weaker extensions and more ordin-
ary apparatus).41
As we have seen, ethical issues related to the machine’s use are raised in
medical texts, mainly in regards to the machine’s operator, who could be an
inexperienced physician or, even worse, a charlatan who merely wishes to
attract the audience’s attention and uses machines on injuries that do not need
them.42 For this reason, the Hippocratic authors often warn the patients
against incompetent practitioners who not only incorporate machines into
their treatments to impress their audience, but also to harm instead of heal;
generally, the unnecessary use of complicated machines is chastised for being
absurd (On Fractures 15, καὶ γὰρ σολοικότερον μηχανοποιεῖν μηδὲν δέον). As
40
Celsus refers to it as a device of highest efficacy, without specifying its structure and functioning.
7.10.7; 25. Oribasius also asserts that the Hippocratic bench is the most efficient machine for cases of
bone reductions in big cities (49.6.1ff.) On its popularity and reception in later times, see Galen
Hipp. Art. pp. 746–9.
41
Αὗται πᾶσαι αἱ εἰρημέναι ἀνάγκαι ἰσχυραὶ, καὶ πᾶσαι κρέσσους τῆς ξυμφορῆς, ἤν τις ὀρθῶς καὶ
καλῶς σκευάζῃ. Ὥσπερ δὲ καὶ πρόσθεν ἤδη εἴρηται, πουλύ τι ἀπὸ ἀσθενεστέρων κατατασίων καὶ
φαυλοτέρης κατασκευῆς τοῖσι πλείοσιν ἐμπίπτει (Joints 73).
42
Berrey 2017a, 182, 190; Gleason 2009, 100–2.

the author of On Joints 78 notes, one should choose the least troublesome way
of healing (ἀοχλότατος); ‘this is more honourable and more in accord with
the art [καὶ γὰρ ἀνδραγαθικώτερον τοῦτο καὶ τεχνικώτερον] for anyone who
is not covetous of the false coin of popular advertisement [ὅστις μὴ ἐπιθυμεῖ
δημοειδέος κιβδηλίης]’. Apollonius emphasizes further the Hippocratic scep-
ticism on the use of machines in medical praxis (18, τῶν ἄρθρων χωρὶς
ὀργανικῆς μηχανοποιΐας καταρτίζειν). Although he endorses Hippocratic
machines, refers to their potential for success, praises their inventor’s intelli-
gence and depicts their materials and construction methods, at the same time
he demonizes their use, insisting on the fact that physicians should avoid
popular mechanical tricks and concentrate on the healing itself (28).43 He
emphasizes the dangers embedded in such machines and warns that one
should be cautious when one uses them, while he also asserts that it is
disgusting, ἀηδὲς, to even discuss such kinds of technical treatments (15).
In general, treatments for damaged bones that do not take into consider-
ation the bones’ natural inclination fail to meet the patient’s needs and,
eventually, are accused of atechnia. The term occurs several times in the
Hippocratic corpus and describes medical ignorance (On Arts 1; On Fractures
36) or the physician’s untested knowledge (Decorum 4; see also On Art 5); or
it is equivalent to boldness, thrasutês, in healing practice (Law 4). Thus, in
a manner corresponding to the Hippocratic doctrine described at the outset,
the expected, ideal relationship among machines, physicians and devices
should be the following: the proper use of instruments should be linked to
an informed, experienced operator’s/physician’s will and knowledge and
should not to be used for profit or any other unseemliness;44 such
a treatment is often called technikos (Decorum 2, though not in reference
to the use of devices), technikôteros (On Joints 78, On the Physician 14, more
according to the art) or technikôtatos (On Joints 67, most according to the
art) – in those cases, medical art is fulfilled with ease and non-complicated
treatments (ῥήϊστον, ἀοχλότατον) – or δίκαιος, just and proper (see e.g. On
Joints 7, δικαιότατα μὲν γὰρ μοχλεύει; On Fractures 7, 8, δικαιοτάτη
βραχίονος κατάτασις).45
The skilful joining of devices and body parts (of physicians, assistants and
patients) reflects the collaboration between technê and nature. However, at
the same time, the use of machines, by enhancing, in a way, the physician’s
43
καὶ γὰρ ἀνδρα<γ>αθικώτατον τοῦτο καὶ τεχνικώτατον ἄν τις ὑπολάβοι, ὅστις μὴ ἐπιθυμεῖ δημ{ι}
οειδέος κιβδηλίης.
44
See further Lucian Ignorant Bookseller 29.
45
On the term dikaios in Hippocratic surgical treatises, see Craik 2010 231–2, Craik 2017 and Eratian
s.v.

122 maria gerolemou
hands and strength, shakes confidence in the physician’s skills and questions
his expertise. In other words, it poses the question of whether physicians are
indeed more competent than a layman who simply uses effective machines
(cf. On Ancient Medicine 4; On the Art 5–6). Let us briefly, and selectively,
sketch some of the implications entailed in this threat. The use of instruments
could easily turn medical technê into banausikê technê which, in this case, while
not being accused of causing bodily degradation, as attested in Aristotle’s
Politics about crafts (1258b37; see also Xen. Oec. 6.5), contains the risk of
associating medical art with profit, deceit and charlatanism, αἰσχροκέρδεια
καὶ ἀσχημοσύνη (see e.g. Decorum 2, 5; The Sacred Disease 18).46 On the other
hand, it underscores the ignorance of the physicians who fail to incorporate
new technical knowledge into their healing praxis or properly understand
both the functioning of the human body and technology. In On Fractures 16,
the author, referring to hollow splints (sôlênes) which are put under fractured
legs, mentions that vulgar patients (dêmotai) often demand that their doctors
employ these devices, attributing to them important powers – ‘the practi-
tioner will be freer from blame (ἀναμαρτηρότερος) if a hollow splint is
applied’ – although this is a bad practice (ἀτεχνέστερος).47
*
To conclude, let me return to the general question of this chapter: what do
the Hippocratic authors acknowledge as a medical machine and how do
they describe it? I have suggested that Hippocratic authors determine the
type of medical devices according to accessibility, transportability, complex-
ity, the force they release and their dependence on the physician’s physical
skills. If the device is identified as a complicated machine, the texts deal with
the possibility of its prevalence over the natural bodily schêmata of the
patient. This type of machine can go beyond the control and power of the
physician to attain a goal that could not be otherwise achieved; hence, it
should be operated with caution. In other words, the technical intervention
in processes of healing sometimes seems to be counter-effective,48 precisely
because of the machine’s possible release of uncontrolled power and the
inability of the physician to restrain it. Most of the time, however, it is not
the machine per se that causes anxiety but the possibility that this may
converge into something that solely serves the unethical interests of its user.
46
See also On Ancient Medicine 4, where a physician is identified as an ἐπιστήμων in contrast to
a τεχνίτης, craftsman. On banausia, see, among others, Nightingale 2004, 117–23.
47
See Edelstein 1967, 92.
48
On the debate about whether mechanics acts against nature, see Krafft 1970 and Schiefsky 2007.

Thus, the issues raised by the use of medical devices is actually a problem of
instrumentality and control.49
RE F E RE N CE S
Bartoš, H. 2020. ‘Hippocratic Holisms’ in C. Thumiger, ed., Holism in Ancient
Medicine and Its Reception (Leiden), 113–32.
Berrey, M. 2017a. Hellenistic Science at Court: Science, Technology and Medicine in
Ancient Cultures (Berlin).
Berrey, M. 2017b. ‘Technology, Performance, Loss: Reconstructing Andreas of
Carystus’ Surgical Machine’, in T. Derda, J. Kilder and J. Kwapisz, eds.,
Fragments, Holes, and Wholes: Reconstructing the Ancient World in Theory and
Practice. Journal of Juristic Papyrology Supplement XXX: Polish Academy of
Sciences (Warsaw), 273–89.
(Cambridge).
Bliquez, L. J. 2014. The Tools of Asclepius: Surgical Instruments in Greek and Roman
Times (Leiden).
Bosak-Schroeder, C. 2016. ‘The Religious Life of Greek Automata’, Archiv für
Religionsgeschichte, 17: 123–36.
Craik, E. 2010. ‘The Teaching of Surgery’, in H. F. J. Horstmanshoff, ed.,
Hippocrates and Medical Education (Leiden), 223–34.
Craik, E. 2015. The ‘Hippocratic’ Corpus: Content and Context (London).
Craik, E. 2017. ‘Teleology in Hippocratic Texts: Clues to the Future?’, in J. Rocca,
ed., Teleology in the Ancient World: Philosophical and Medical Approaches
Deichgräber, K. 1965. Die griechische Empirikerschule: Sammlung der Fragmente
und Darstellung der Lehre, 2nd ed. (Berlin).
Demont, P. 2014. ‘The Tongue and the Reed: Instruments and Organs in the
Philosophical Part of Hippocratic Regimen’, JHS, 134: 12–22.
Drachmann, A. G. 1963. The Mechanical Technology of Greek and Roman
Antiquity: A Study of the Literary Sources (Copenhagen).
Edelstein, L. 1967. ‘The Hippocratic Physician’, in O. Temkin and C. L. Temkin,
eds., Ancient Medicine: Selected Papers of Ludwig Edelstein (Baltimore),
87–110.
Frede, M. 1987. ‘The Ancient Empiricists’, in M. Frede, Essays in Ancient
Philosophy (Minneapolis), 243–60.
Frede, M. 1988. ‘The Empiricist Attitude towards Reason and Theory’, in
R. J. Hankinson, ed., Method, Medicine and Metaphysics. Apeiron 21
(Edmonton), 79–97.
49
I have argued elsewhere that this analysis could be applicable to other contexts as well (2019a, 2019b,
2020 and 2022), even to ‘cognitive machines’.

124 maria gerolemou
Gazzaniga, V. and S. Marinozzi. 2015. ‘Orthopaedics, An Ancient “Specialty”?’, in
A. Piccioli, V. Gazzaniga and P. Catalano, eds., Bones: Orthopaedic
Pathologies in Roman Imperial Age (Dordrecht), 16–30.
Gerolemou, M. 2019a. ‘Staging Artificial Intelligence: The Case of Greek Drama’,
in P. Meineck, W. Short and J. Devereaux, eds., Routledge Companion to
Classics and Cognitive Theory (London), 345–55.
Gerolemou, M. 2019b. ‘Why Can’t I Have Wings? Aristophanes’ Birds’, in
G. M. Chesi and F. Spiegel, eds., Classical Literature and Posthumanism
(London), 175–81, 373–5.
Gerolemou, M. 2020. ‘Laughing against the Machine’, in P. Swallow and E. Hall,
eds., Aristophanic Humour: Theory and Practice (London).
Gerolemou, M. 2022. Technical Automation in Classical Antiquity (London).
Gleason, M. W. 2009. ‘Shock and Awe: The Performance Dimension of Galen’s
Anatomy Demonstrations’, in C. Gill, T. Whitmarsh, and J. Wilkins, eds.,
Galen and the World of Knowledge (Cambridge), 85–114.
Jackson, R. 2005. ‘Holding on to Health? Bone Surgery and Instrumentation in
the Roman Empire’, in H. King, ed., Health in Antiquity (London), 97–119.
Kollesch, J. and F. Kudlien. 1965. Apollonios von Kition: Kommentar zu Hippokrates
über das Einrenken der Gelenke. Corpus medicorum Graecorum 11.1.1 (Berlin)
Krafft, F. 1970. Dynamische und statische Betrachtungsweise in der antiken Mechanik
(Wiesbaden).
Künzl, E. 1996. ‘Forchungsbericht zu den antiken medizinischen Instrumenten’,
ANRW, 2.37.3: 2434–2639.
Le Blay, F. 2016. ‘Surgery’, in G. L. Irby-Massie, ed., A Companion to Science,
Technology, and Medicine in Ancient Greece and Rome (Chichester), 371–86.
Lloyd, G. E. R. 1987. The Revolutions of Wisdom: Studies in the Claims and Practice
of Ancient Greek Science (Berkeley).
Lonie, I. M. 1981. ‘Hippocrates the Iatromechanist’, Medical History, 25: 113–50.
Mile, J. S. 1907. Surgical Instruments in Greek and Roman Times (Oxford).
Nightingale, A. W. 2004. Spectacles of Truth in Classical Greek Philosophy: Theoria
in Its Cultural Context (Cambridge).
Peck, A. L. and E. S. Forster. 1937. Aristotle: Parts of Animals. Movement of
Animals. Progression of Animals. Loeb Classical Library 323
(Cambridge, MA).
Schiefsky, M. J. 2008. ‘Theory and Practice in Heron’s Mechanics’, in W. R. Laird
and S. Roux, eds., Mechanics and Natural Philosophy before the Scientific
Revolution (Dordrecht), 15–49.
Schneider, H. 1989. Das griechische Technikverständnis: Von den Epen Homers bis
zu den Anfängen der technologischen Fachliteratur (Darmstadt).
Snell, B. 1923. Die Ausdrücke für den Begriff des Wissens in der vorplatonischen
Philosophie (Göttingen).

Tabanelli, M. 1958. Lo strumento chirurgico e la sua storia: dalle epoche greca
e romana al secolo decimosesto (Forli).
Alexandria’, SHPS, 34: 443–66.
Von Staden, H. 2007. ‘Phusis and Techne in Greek Medicine’, in B. Bensaude-
Waldow, A. 2010. ‘Empiricism and Its Roots in the Ancient Medical Tradition’, in
C. T. Wolfe and O. Gal, eds., The Body as Object and Instrument of Knowledge:
Embodied Empiricism in Early Modern Science (Dordrecht), 287–308.
Webster, C. forthcoming, Tools and the Organism: Technologies and the Body in
Greek and Roman Medicine (Chicago).
Wilson, A. 2008. ‘Machines in Greek and Roman Technology’, in J. P. Oleson,
ed., The Oxford Handbook of Engineering and Technology in the Classical
World (Oxford), 337–68.
Withington, E. T. 1928. Hippocrates: On Wounds in the Head. In the Surgery. On
Fractures. On Joints. Mochlicon. Loeb Classical Library 149
(Cambridge, MA).
Witt, M. 2018. ‘Surgery’, in P. E. Pormann, ed., Cambridge Companion to
Hippocrates (Cambridge), 217–45.

chapter 5
The Empirical, Art and Science in Hippocrates’

On Joints
Jean De Groot
As much as we prize evidence of prehistoric or archaic uses of mechanical

tools or tropes, it can be difficult to assess what sort of understanding
accompanies their use.1 This is the case especially for devices that were in
use for millennia, like the stationary pulley, and proliferated widely, like
the potter’s wheel. The lever is one such device. This question of varieties
of understanding figures in Hippocrates’ justly famous accounts of how to
reduce shoulder dislocations in his On Joints (Peri Arthrôn) 1–11.2 My aim
in this chapter is to evaluate the several modes of reduction of the shoulder
described by Hippocrates in Joints 2–7 in terms of the mode of cognition or
understanding of leverage and associated physical phenomena present in
this treatise of art (technê), not science (mathêma or epistêmê).3 My further
1
Depending on context, understanding in this paper means sometimes the grasp of technique at the
level of art (technê) or the self-conscious discursive knowledge that goes by the name epistêmê . Since
epistêmê means knowing in a scientific way, I avoid referring to art as knowledge and refer to it as
a type of understanding instead. In the twentieth century, the word understanding was used by some
scholars of ancient philosophy to connote a high level of intellectual grasp of something.
Understanding was the translation for nous or noein in Aristotle. I use the word more broadly and
for cognitive phenomena at the level of experience.
2
I refer to the treatise hereafter as simply Joints. Along with On Fractures, Joints is considered by
scholars to be the writing of a highly skilled physician. Both treatises were probably written at Cos in
the middle to late fifth century bc. Indeed, they may originally have comprised a single text,
although no manuscripts attest to this original unity (Witt 2018, 227). Given their date and place,
I will call the author of the treatise Hippocrates to avoid repeated circumlocution when referring to
whomever in Hippocrates’ school was the author. On the history of Hippocrates and the Hippocratic
corpus in antiquity, see Craik 2015, xvii–xxiv. The Greek text used herein is the Loeb version,
Hippocrates, vol. 3 (Loeb Classical Library 149, Hippocrates 1928), which is based on Petrequin (1878).
For editions, see Jouanna 1999, 403. The most recent edition is by Kühlewein in the Teubner library
(1902). Perseus (www.perseus.tufts.edu) has made the edition by Adams (1868) more accessible. The
complete Littré edition (Hippocrates 1840) of the Hippocratic corpus is available online at www
.archive.org/details/oeuvrescompltes04hippgoog, with Peri Arthôn starting at p. 106. Withington,
the Loeb translator, and Witt point out that two manuscripts have the fuller title Peri Arthrôn
Embolês, ‘On the Reduction of Joints’ (Hippocrates 1928, 200; Witt 2018, 220). For more on the
history of the text of Joints, see Witt 2018, 225–33. Textual references to Joints in this paper will use the
abbreviation of the transliterated Greek, Arth.
3
Perilli 2018 gives a general treatment of the epistemology of ancient arts and of medicine.
126

The Empirical, Art and Science in Hippocrates’ On Joints 127
aim is to arrive at some points of distinction among empirical intuition of
physical forces, art that arises from such intuition and the physical science
of leverage that emerged a century or more later.
At the beginning of Joints, Hippocrates gives instructions for various

methods of reducing a shoulder dislocation. Perhaps because of its associ-
ation with the later treatise of the corpus named Mochlikon,4 the treat-
ments recommended in Joints are often bundled as all involving leverage as
understood in terms of the operation of a lever (Figure 5.1). Hippocrates,
however, uses the noun lever (mochlos) only four times5 in Joints. We
should bear in mind that the lever is a tool, and not every instance of
invoking forces to achieve reduction of a joint uses a tool. In what follows,
I will discuss in its own terms Hippocrates’ account of forcing (anagkazein)
and extending (katateinein) to achieve reduction.
Hippocrates does not use the word mochlos, lever, in presenting the
modes of reduction in Joints 2–7, and the methods he gives depart in
significant ways from scientific accounts of leverage developed in the
fourth and third centuries bc.6 This is what one might expect, given that
medicine is an art and that Hippocrates’ methods antedate the ancient
mathematical principles of leverage that have survived. Nevertheless, to
understand mechanics within the art of medicine, we need to explore these
differences with more precision.
My general conclusion, which is documented below (pp. 137–147), is that
Hippocrates had his own understanding of combining forced movements,
which is not anchored in equilibrium as later scientific accounts of leverage
are. Hippocrates’ technical understanding involves oppositions of movements
and oppositions of movement and stability. These oppositions are meant to
provoke some other movement remotely. The remote movement intended –
4
On the treatise Mochlikon, see Jouanna 1999, 398. Withington argues on grammatical grounds in
favor of a common early authorship of Fractures and Joints, and against a common authorship of
Joints and Mochlikon, the latter being a later work (Hippocrates 1928, xxiii–xxvii). He reasons that an
original Fractures/Joints treatise perhaps including an On the Nature of Bones, now largely lost to us,
was broken up into parts soon after its composition (87–9).
5
Hippocrates 1928: Arth. 72.30, 34, 39, 30; 74.4.
6
See the pseudo-Aristotelian Mechanics, as well as Aristotle’s biomechanical references in On the
Movement of Animals 1 and 7 and On the Progression of Animals. Given that the attribution of
Mechanics to Aristotle is still doubtful but that the treatise seems to be from his school, I will refer to
the Mechanics by the shorter adjectival, Aristotelian, not pseudo-Aristotelian. Archimedes’ Of the
Equilibrium of Planes is from a mathematical standpoint a more pleasing account of the lever.

128 jean de groot
Figure 5.1 Simple lever, FCIT (Florida Center for Instructional Technology),
J. L. Comstock, A System of Natural Philosophy: Principles of Mechanics (Pratt,
Woodford and Company, 1850), 69
to slide a bone back into its joint – is fairly highly calibrated in its effect. In
some cases, what Hippocrates describes can be construed as leverage, but in
other cases, it cannot. What Hippocrates offers are methods to be imitated for
applying force.7
There is a distinction from linguistic philosophy that can serve as a guide
to thinking about the differences among the kinds of cognitive awareness
ancient practitioners had about their productive techniques, namely, the
distinction between ostension and ostensive definition. Ostension is learning
a meaning or use of a word by someone else’s pointing or otherwise
gesturing to something. Given the human capacity for language, infants
and toddlers actively assimilate word usage by noticing the actions and body
language of adults around them. They do not explicitly assign words to
things and do not take note of having added a new word to their vocabulary.
In ostensive definition, on the other hand, both the person with mastery of
the language and the language learner are aware of the linguistic nature of the
word, so that the pointing is more like an account or definition that fits the
word into the syntax of language and the organization of the world.8
7
For another approach to the medical appropriation of natural patterns and forces before those
phenomena had been codified by principles and laws, see Webster in this volume, Chapter 6.
8
Wittgenstein addressed the difference between ostension and ostensive definition in Philosophical
Investigations 1–38. In these paragraphs, he deflates the idea that the proposition is the sole vehicle of
meaning, saying that language users, without being fully aware of it, engage in shared ‘language
games’. The smooth functioning of even simple language games can be compared to the sensory-
motor intuition that is eventually channelled into art. A difference between Wittgenstein’s ostension
and my treatment of the art of ancient surgery is that the latter clearly involves responses of the one
who understands at the interface between themselves and the physical world. Interface with the
natural world can only be inferred as part of Wittgenstein’s philosophy of language. For a recent
account of ostension and ways to interpret it, see Engelland 2014.

A similar distinction can be made with respect to mechanical techniques
as they gained currency in life around the Mediterranean from prehistoric
to Hellenistic times. There is a difference between learning to work
a device or perform a technical activity, on the one hand, and having an
account of how or why it works, on the other. A potter’s wheel or a pulley
as part of a ship’s tackle could be passed down within a geographical region
and used for a very long time without innovation being made upon the
basic arrangement or technique. Similarly, a novice learns to use a pulley by
watching others use it and noticing that the pulley eases the work of
moving a weight. From having watched, with sufficient industry they
could put together a pulley of their own by imitation. This is different,
however, from being able to say why the pulley works and to what other
devices it might be related.9
Enhancement of a technique can arise from watching and imitating,
however. A novice can think that, if one pulley lightens the load, adding
another along the same track may lighten the work even more. There is
a limit to the novice’s innovation at the level of imitation, however. He
likely would not come up with the idea of a genuine compound pulley,
which sets two pulleys operating apparently in opposition to one another.
The opposing of devices this way requires a level of reflection that need not
be present in imitation.10
Corresponding to ostensive definition, on the other hand, is the articu-
lation of a principle – for the motion of bodies, usually a mathematical
one – that organizes different phenomena under a single statement that
gives the principle underlying the phenomena. The Aristotelian Mechanics
proceeds under just such a plan for the principle of leverage. Its principle
concerns properties of circular motion. Mechanics applies a single theoret-
ical explanation to a range of craft phenomena and some natural
phenomena.
Understanding along scientific lines opens a wider avenue for enhance-
ment of the power than is available to the novice, something we have
learned from the long history of technology. For one thing, understanding
links a device to natural powers, so that the differences and the connections
between device and natural action come to light as a subject matter for
epistêmê, knowledge of the natural world that reflects principles at work in
9
In the Aristotelian Mechanics, the ease of movement enabled by the pulley (trochilea) is related to the
properties of circular movement that explain the lever (Mech. 853a33–b2 [Aristotle 2000, III, ch. 17 =
ch. 18, Teubner ed. and Loeb]).
10
Imitation, mimêsis, is a multivalent concept in antiquity. For present purposes, I choose a narrow
connotation of the word to match the human capacity for ostensive behaviour.

130 jean de groot
nature. Since so many devices originated before written records and even
before most archaeological evidence, it is important to distinguish between
the existence of a device known because of archaeological evidence or by its
being named in a context appropriate to its use, and an account of its
working, which brings it into the realm of scientific or theoretical interest.
Clearly, though, there is an intermediate level of cognitive awareness
between imitation and understanding, namely, know-how or skill. As I use
the term here, know-how has some aspects of both empeiria and technê as
Aristotle distinguished those abilities.11 The skilled person – the one with
know-how – notices about a pulley that the force exerted and the load lifted
move in contrary directions, that the pulley works better the longer the
length of rope or cable, and that the heavier the load, the longer the cable
and, following on from that, the greater the time required to raise or lower
the weight. The skilled person is able to use these details to modify or
compound devices. This is more than the imitator (comparable in my
analogy to the child learning words) can do. The one with know-how may
or may not be oriented towards causes or principles of nature to which the
working of the pulley could be traced back,12 although they are always very
attentive to relevant details. Theirs is a motor and memory-based skill, and
they are a clever person.
There are at least three ways of being acquainted with and using device,
then – repetition (comparable to ostension), know-how (which can be
taught but is the fruit of untutored cleverness) and knowledge (comparable
to ostensive definition). Of this threefold division, I focus on the first two
in analysing Hippocrates’ forcing, aligning and extending in his reducing
of joints. To focus on these two, however, it is useful first to have a point of
reference in the ancient science of mechanics. A brief excursion into the
mechanics of the lever thus follows.
II
There are two ancient formulations of the principle of the lever, that of the
Aristotelian Mechanics and the other of Archimedes’ Of Equilibria of Planes
11
Know-how in my sense is sometimes still within the realm of what Aristotle called experience
(empeiria). I affirm for Hippocratic medicine in this instance both sides of what Aristotle says in
comparing experience and art in Metaphysics A.1. Art recognizes a class (kata eidos hen) of ailment
where experience can only offer a treatment based on similarities. Nevertheless, in practice those
with experience may hit upon a solution better than those possessed of an explanation (logos) unaided
by practice (981a2–15, 28–30). Experience is knowledge (gnôsis) of particulars – in medicine, the very
subjects to be treated – while craft knows the universal (a16–17).
12
For this formulation of the relation of phenomena to principles, see Aristotle, Physics 1.1.

or Centres of Gravity (barôn) of Planes. In the fourth century, even before its
appearance in the Aristotelian Mechanics, the mathematical principle of the
lever as understood by the author of that treatise was cited by Plato and
Aristotle.13 This principle was expressed in two ways, in terms of a rotating
radius or in terms of the movement together of concentric circles.14 The
moving-radius formulation of the principle of the lever is as follows:
(1) Of a line rotating about one fixed end, different points trace out
longer and shorter arcs in the same time. Speeds of the moving points
are proportional to their distance from the centre of the circle.15
(2) Of concentric circles revolving around their centre, the smaller circles
cover in the same time a shorter distance than the larger circles. Points
on the smaller circles move more slowly than points on the larger
circles. Respective speeds of the circles are proportional to their
distances from their common centre.
If one imagines the moving radius visually, the end of the rotating line would
be moving faster than a point on the line close to the fixed end. The result of
the concentric-circles version is that, as Plato puts it, longer and shorter
distances are covered in the same time, ‘an effect one might have expected
to be impossible’.16 The two formulations of the principle are equivalent
(Figure 5.2), something of which Aristotle was aware.17 The principle in either
version falls ill on modern ears, which are not accustomed to measuring circles
without π. In Greek thought, Archimedes introduced the quantity π in the
third century bc. So, the original account of why and how a lever works
focused on circular motion but without benefit of π. The principle yields
a mathematical proportion of arc-speeds to distances from the centre.
This pre-Archimedean principle of balance has much in common with
what is called Archimedes’ Law of the Lever in his Of the Equilibrium of
Planes. Archimedes’ treatment, however, is better described as a law of
13
Plato cites the principle in relation to astronomy in Laws 10, 893c–d. In this same context, see
Aristotle, On the Heavens ii.8, 289b30–290a7. In relation to the amplification of power typical of
leverage, see Aristotle’s Movement of Animals 1, 698a18–24, and 7, 701b2–28.
14
Both formulations appear in the Aristotelian Mechanics (848a8–b18; 848b5). When points are taken
opposite one another along a diameter, then the moving radius becomes a mathematical version of
the lever (849b20–32). Krafft, the first contemporary scholar to draw attention to the principle,
conducts his analysis of mechanics before Archimedes using the concentric-circles formulation
(Krafft 1970). De Groot 2014 makes use also of the moving-radius formulation.
15
The Aristotelian author’s formulation in Mechanics is as follows: ‘Further, because none of the points
on a line drawn from the centre of the same circle are moving at the same speed but the one further
from the stationary limit always moves faster, many marvels in the movements of circles come about,
and these will be evident in the following problems’ (Mechanics 848a12–19).
16 17
Plato, Laws 10, 893d45. Aristotle, De Caelo 2.8, 290a2–5.

132 jean de groot
MOVING RADIUS CONCENTRIC CIRCLES
Figure 5.2 Moving radius, concentric circles
balance (isorrhopia). Archimedes begins with a beam with equal weights in

balance at equal distances from a fulcrum (Figure 5.3a). These weights will
not balance when placed at unequal distances from the fulcrum
(Figure 5.3b). Taking again the equal weights at equal distances from the
fulcrum, balance will be disturbed by the addition of weight at one end
(Figure 5.3c). When this additional weight is introduced, a rebalancing of the
weights is achieved by placing the now-dissimilar weights at different dis-
tances from the fulcrum rather than at the same distance (Figure 5.3d). The
heavier is on the shorter end of the beam, and the lighter on the longer end.
Archimedes shows that distances from the fulcrum and weights at ends of
the balance beam will be in inverse proportion to one another when they
balance. If W2 is the double weight in Figure 5.3d and D1 is the longer
distance supporting the lesser weight, then
D1 W2
∝
D2 W1
and the different weights will balance at these distances.
The earlier moving-radius principle of the lever present in Mechanics
stands in contrast to Archimedes’ Law insofar as it begins and ends with
movements taking place at opposite ends of the lever. This approach, though
not as elegant mathematically as Archimedes’, lies close to the purpose of the
lever, which is to move a greater weight by means of a lesser one (Mech.
847a21–4). In this earlier account of the lever, speeds of different points along
the rotating radius (or speeds of the revolving circles) are proportional to the
distance of the moving point from the centre of the circle. Speed is under-
stood as distance covered in a given time.18 A greater speed is one by which
the moving body covers a longer distance in the same time as some other
body covers a shorter distance. The speeds, though different, are all linked to
18
In Mechanics, this proportion is itself given an explanatory account in terms of a proportion of
rectilinear elements constructible within the circle (Mech. 849a22–b18).

(a) (b)
(c) (d)
Figure 5.3a A beam with equal weights at equal distances from a fulcrum.
5.3b Equal weights at unequal distances from a fulcrum.
5.3c Disequilibrium induced by the addition of weight to one end of a balance of equal
beams at equal distances.
5.3d Unequal weights in equilibrium at unequal distances from the fulcrum.
one another by sharing a radius or a common disc-like movement around

one centre. To this extent, proportion as an organizing principle of equilib-
rium is built in to the situation of movement itself. For any heavy body being
moved, the effective force is less than that body’s weight according to a ratio
inverse to the ratio of the lengths of the arms of the lever from the fulcrum.
The principle of circular movement of the Aristotelian Mechanics is never
treated as a model for the calculation of exactly how long the beam of a lever
should be or where the fulcrum should be placed. It is a theoretical principle
that the author shows to be at work in many kinds of devices, like simple oars,
ship’s tackle and levers for moving weights. The moving-radius principle
does, however, provide the aspect of mathematical ordering and correlation
that explains the lever’s ability to move a larger weight by means of a lesser
one, and its account could be given a quantitative expression.
My contrast between the Aristotelian and Archimedean principles of
leverage has been brief. Naturally, the contrast, along with the commonalities
pointed out between the two, raises many questions proper to just these two
accounts. These questions cannot be pursued here. For present purposes, it
suffices that the two accounts provide specific examples of what a scientific
principle governing the lever looked like in antiquity. Both the author of
Mechanics and Archimedes provide a demonstration or proof of the principle.
In this early science of the lever, certain aspects can be identified as
essential to the working of a lever. First, leverage as such requires a straight
beam by which the effects of weights are linked. Second, the weights, or
a weight and a force exerted, are matched against one another so that one
overpowers or exceeds the other. The effect of leverage, usually movement,
is registered at the location of the weight that is overmatched. Third, a lever
must pivot around an unmoved point, the fulcrum (hupomochlion).

134 jean de groot
The fulcrum can be at one end of the beam, as in the case of
a mortar and pestle. In this case, a greater weight does not move
a lesser, but a force exerted produces a grinding effect that would not
be possible without the pestle. So, there is the same resistance to the
movement. In every case, then, there must be an unmoved place of
pivot somewhere along the moving beam. When the fulcrum is located
along the beam, there is pivoting around that point. This pivoting is
obviously movement. The word pivot, as a noun, is appropriate for the
fulcrum, which does not itself receive any of the force (ischus) of the
contending weights.
The last and perhaps most important point about identifying why
a lever moves a greater weight with a lesser one is the presence of
equilibrium, which underwrites the possibility of forming proportions
such as Archimedes’ above. Archimedes’ principle is very directly
focused on equilibrium, insofar as his discourse gives the formula for
adjustments in weight and distance that will restore balance. In the
Aristotelian Mechanics, equilibrium is less obviously at issue, because
the equilibrium is not stasis, remaining in place, but exists mathemat-
ically and physically in the correlation of longer and shorter distances
covered in the same time. These movement are not independent of one
another.19
With this outline of the emerging science of the lever in the fourth to
third centuries bc, I will now analyse Hippocrates’ way of dealing with
phenomena that bear some resemblance to leverage and balance. I analyse
Joints 1–4 together and then chapters 5–7.
III
I begin with several observations about Hippocrates’ language. First is his

use of verbs instead of nouns in describing his techniques for relocating the
shoulder. In Section IV, I will develop the significance of this difference
further. For now, I simply assert that, in Hippocrates’ descriptions, osten-
sion in language, the very activity itself, overlaps with its parallel in
technique, namely, unreflected-upon imitation. Hippocrates shows with
19
On this last point of movements constrained by one another and thereby issuing in an orderly
motion, compare the pseudo-Aristotelian Problemata 16.3, 9 and 12. These chapters show that the
mutual constraint of weights or forces in motion was a topic of interest in antiquity. These chapters
of Problemata 16 are analysed by De Groot 2014, 173–80 and 195–216.

action verbs what his hearers or medical readers are to do.20 Accordingly,
nouns are scientific words, while verbs serve art, which focuses on action
with a product or outcome.
Hippocrates uses the verb mochleuô and the process noun mochleusis.
Among the nouns he uses, more common than mochlos is embolê, the
general word for reduction of a dislocation. Even in this case, the verb form
emballô is more frequent and is used in specific descriptions for applying
pressure, pushing or placing something in a position to act. The words
Hippocrates uses set up actions. As in the case of ostension, his is not the self-
conscious vocabulary of the balance beam or equilibrium. Hippocrates’
vocabulary is not only pre-scientific but is to some extent pre-technical.
He is establishing technique. Hippocrates’ language in Joints and Fractures is
notable too for its very many verbs compounded with prefix prepositions
(see below, pp. 137–140; 142–144). These compound verbs when linked to an
object or state of affairs take a simple es, whether the meaning is ‘to’, ‘at’,
‘into’ or some other position or direction.21
There is another difference in the use of words that is relevant to
Hippocrates’ understanding of physical forces acting in concert.
Descriptions of the lever from Aristotle through Archimedes to Heron of
Alexandria use the word zugon for the beam of the lever.22 Zugon referred
20
This topic, which I have characterized in philosophical terms, is related to the larger question of how
either understanding or instruction is conveyed to others by ancient technical texts. Relevant to
Hippocrates’ detailed instructions about relocating the shoulder are questions about the relation of
the text to oral instruction or clinical demonstration and also the question of how much knowledge
is presumed on the part of the hearer or reader. On these questions, see Formisano and van der Eijk
2017, 1–7, and Cañizares 2017, 92–109. Authorial presence in the text is also relevant to Joints. Both
personal authority (Arth. 1.1–35; 11.7–20) and a sense of wonder (7.45–7) are expressed in the author’s
voice. On the importance of asserting personal prestige in antiquity rather than taking an imper-
sonal stance in medical writing, see Holmes 2013, 432–5.
21
The history of Greek prepositions is beyond the scope of this chapter and perhaps not even relevant
given established usage in fifth century technical writing. An older semantic theory about Greek
prepositions held that they emerged from adverbs. More recently, emphasis has been placed on
prepositions as parallel to noun cases. For references, see Hessinger 1978 and Bortone 2010. The
important point for Hippocrates’ language is that the preposition/verb compound heightens the
sense of action as opposed to notion or concept and makes the action more specific.
22
Zugon is used in the Mechanics as a word for the balance rather than the lever (mochlos) (848a11–14).
Archimedes wrote a treatise, Peri zugôn, which is lost (Heath 2002, xxxvii, citing Pappus, Collectio
8.1068, Thesaurus Linguae Graecae). In his Of Equilibria of Planes or Centres of Gravity (barôn) of
Planes, Archimedes uses the word ‘length’ (Doric makos = mêkos), connoting length of a beam from
a centre. Heron of Alexandria refers to the balance (zugon) by analogy in his surviving Greek works
(Pneum.1.1, 18; Dioptra 37, 54) and directly in his Mechanics (Mech. Frag. 1.1, 57). Pappus uses xulon
more often than zugon, the balance beam, in Synagôgê. In Pappus’ account of the lever in book 8, he
mentions the beam (xulou makrou) as needing to be wedged under the weight to be moved by
digging beneath it (8, 54.1118). In book 8, where he treats Heron’s five powers, Pappus speaks of the
mutual effect of weights being acted upon by one another with reference to the balance beam (kata
ton antipeponthota tôn barôn en tois zugois logon) (1042.15–21).

136 jean de groot
originally to the yoke pairing oxen at the plough. So, the word originated
in agriculture and carries a connotation of pairing to induce an orderly
exertion of effort in a single direction. Hippocrates does not use zugon for
the beam but rather uses the words board (sanis), plank (xulon) or pestle
(huperos). In Joints, a piece of wood sometimes operates to exert pressure at
one of its ends – hence the word pestle – or to bring about extension as an
agent of traction. When the traction depends on suspending the body in
some way (Arth. 4–5), the pressure operates, as far as the physician is
concerned, more at one end than the other.
Hippocrates begins his treatment of dislocation of the shoulder by
insisting that, despite how it may appear, there is no forward and certainly
no backward dislocation of the shoulder but only dislocation downwards
out of the joint socket. Figure 5.4a shows a healthy shoulder joint and
Figure 5.4b a dislocation such as Hippocrates describes. The author notes
that, in a normal attitude of the body facing forward with arms at the
person’s sides, the shoulder blade protrudes slightly forward already. This
appearance is exaggerated when the humerus is displaced downwards, and
this is the source of the mistake in diagnosing a forward displacement of
the humerus (Arth. 1). Someone who dislocates a shoulder frequently, as
might an athlete in contact sports, is able to reduce the dislocation
themselves: ‘They place the knuckles of the other hand in the armpit and
push the ball of the joint upward, drawing the elbow of the dislocated arm
(a) (b)
Figure 5.4a A healthy shoulder joint

Figure 5.4b A dislocation such as Hippocrates describes

toward the chest’ (Arth. 2. 1–6).23 Experience shows that, with the contri-
bution of local tendons, ligaments and musculature, the bone tends to slide
back into the socket.
Hippocrates describes how the iêtros would perform the same reduction
surgically (Arth. 2). He would (1) position his fingers more inside the
armpit (towards the ribs) than the dislocated ball of the humerus and
force (apanagkazoi) the ball away from the ribs, while (2) bracing the
physician’s own head against (emballôn es) the injured person’s sound
acromion (the narrow bone of the shoulder above the joint) for resistance
(antereisios heneka); and (3) pushing the elbow of the dislocated arm
towards (emballôn es) the chest with his knees, the physician would push
the arm in the opposite direction (antôtheoi) (Arth. 2. 6–13). The ‘opposite’
direction in this case is opposite to the physician’s pressure with his fist in
the armpit. This procedure is clearly complicated enough to be the work of
a skilled physician able to perform all three actions in a short burst (‘It is
well for the one performing a reduction (emballôn) to have strong hands’,
Arth. 2.13–14).24 For help in this combination of pushing and bracing
against,25 the physician might have another person push the elbow towards
the chest (Arth. 2.14–17).
In this description, there is a need for an unmoved place (the patient’s
acromion) sturdy enough to be a prop against which to brace in order
for the primary action to proceed free of interfering ancillary movements of
the body. The primary action is pushing the ball of the humerus away from
the ribs (2.9–11). There is also a requirement of pressure applied at the
elbow in the direction opposite to the action of moving the ball of the
humerus (Arth. 2.13). This last manoeuvre, also undertaken by the athlete
who reduces his own dislocation, cannot be dispensed with but is necessary
to the procedure. Drawing the elbow to the ribs stabilizes the shoulder/arm
23
A similar technique can be seen as a rugby player relocates his own shoulder on the pitch at www
.youtube.com/watch?v=TvMj6WKP7r0 (accessed 26 April 2021).
24
It strains credulity that this account in Joints 2 was meant in its first composition as simply
a textbook entry. It is reasonable to think that the author, with a note-taker present, was providing
a visual example or clinical demonstration for medical students or assistants in training. This would
be the expectation also in the case of the procedure described in Joints 3. There is a growing
secondary literature on the topic of the rhetoric of medical texts, on the one hand, and the relation of
a given text to clinical training, on the other. See for example Craik 2010, 232; Dean-Jones 2003, 98–
9; Cross 2018, ch. 3.
25
Hippocrates uses the verb emballô, meaning to throw in or strike, for both the action of exerting
pressure to attain immobility and of positioning the fist in the armpit. There is a connotation of
forcible activity in either case, but I have translated the word as bracing, moving or exerting pressure.
The noun embolê is translated as reduction, for the entire accomplished repositioning of the bone in
the joint (Arth. 4.1).

138 jean de groot
configuration so that it does not swing with the pressure of the fist in the
armpit, and it opens momentarily the axilla cavity for the ball of the
humerus to resume its proper location. Although this method involves
counter-movements and an unmoved place for steadying the physician
himself, there is no beam (pestle or plank), unless we think of the practi-
tioner’s body itself as the beam. Nevertheless, as will become clear next, this
method seems more like genuine leverage than the second method that
Hippocrates presents in Joints 3.
In Joints 3, the leg and heel of the physician work together like a plank or
pestle. With the patient lying on the ground and the physician seated on
the same level, he places his heel into the armpit of the dislocated shoulder,
right heel for a right-shoulder dislocation and left heel for a left-shoulder
dislocation (Arth. 3.7–9). He pulls the affected arm out and with his heel in
the armpit exerts counter-pressure (antôthein) on the humerus with his heel
(3.6–8). To ensure he makes contact with the ball of the humerus in this
method, Hippocrates says the physician should place a hard ball, like the
balls stitched together from pieces of leather, at the base of his heel. This is
to prevent tendons from moving into the axilla when the arm is extended
(3.9–18). The ball occupies the place where the tendons would go, thereby
keeping the axilla open for the ball of the humerus to move. With the
leather ball and heel in place, the heel pushing while the patient’s arm is
pulled in traction makes the humerus return to its socket. During this
procedure, an assistant should be holding the other side of the patient’s
body immovable on the ground (Arth. 3.18–22). Another assistant, posi-
tioned beyond the head of the patient, exerts traction in the opposite
direction (antikatakeinein) by pulling on the leather ball towards the
patient’s head by means of a strap wrapped securely around it (Arth.
3.22–31). The procedure as outlined would involve at least three people.
Now it is questionable, despite Hippocrates’ language of pushing in
return (antôthein) and counter-tension (antikatakeinein), whether this
method of extending the arm and pushing the humerus into place with
the heel is a levered movement. It is simply pushing the end of the humerus
bone into place by means of the end of something else straight and rigid.
There is pushing at some angle, but is that leverage? A number of points are
worth considering in answer to this question.
The primary action of reducing the dislocation is pushing the ball of the
humerus back into the joint. This action is undertaken in a situation where
physical tension is induced locally by extension of the arm towards the feet
and, in opposition, extension of the armpit towards the head, the latter by
means of the strap and leather ball. Extension in both directions is crucial,

because it opens a space for the bulbous tip of the humerus to shift. Since
the leg-and-heel pestle pushes the ball of the humerus sideways and
upwards, what the pushing of the humerus accomplishes is that the ball
of the bone does not remain below the joint when traction on the arm
opens the space around the joint. This push into a cavity has no real
counterpart in mechanical leverage. Furthermore, stretched tendons and
adjacent musculature contribute confining forces to the physical tension
around the axilla, another feature without any counterpart in the lever.
Although tension produced by movement against resistance is present in
the traction, the repositioning of the humerus is not an example of
leverage, even though the leg of the physician, placed into the armpit,
resembles a beam, which is central to the action of a lever.
In the method of Joints 2, where the physician pushes the humerus back
into place, there is a fulcrum-type element, an unmoved place, the patient’s
acromion used by the physician as a prop. Around himself as the ‘beam’, the
physician produces counter-movements of pushing and pulling using his
hand and knees. The point of immobility is a key element of leverage, if we
regard the physician’s body itself as the beam. Still, the unmoved point lies
outside the primary action of relocation in the axilla.26
In the method of Joints 3, in which the physician pushes with his heel,
another place of immobility is the other side of the patient’s body, which is
being held down (3.18–22). Immobilizing the opposite side of the body
isolates the system of articulation and extension which is the focus of the
action. The immobility of the uninjured side seems less relevant to the
action of reduction than the physician’s immobilizing of his own body
with his head in the method of Joints 2. This immobilizing of the other side
of the body in the heel-and-leg method is like securing a catapult to its
launch site by using stakes or bolts to hold it in stone. This securing
prevents dissipation of force or misdirection due to the catapult moving
around. This securing is quite secondary to the action of the catapult.
Someone might argue that the unmoved point in the method of Joints 3
is instead where the leg–heel–leather-ball combination makes contact
through the skin with the ball or upper part of the humerus. The move-
ment brought about at this location, however, lacks the element of con-
trary inclinations to move (rhopê and antirrhopê) and the overmatching of
one inclination by the other that is central to leverage. The contrariety of
26
On this and other chapters of On Joints, see also Gerolemou in this volume, Chapter 4. Gerolemou
distinguishes three varieties of devices used, ranging from natural materials that are simple, or even
parts of the physician’s own body, to typical devices like a pestle or wedges, and on to heavier ladders
and traction machines.

140 jean de groot
actions is present in the traction, which is directed at opening the axilla for
the shift of the humerus. Indeed, the main reason for saying that the
primary action of reduction is not leverage in Joints 3 is that the overpower-
ing of one movement by another is lacking at the place of the primary
action.
The difference between the two methods of reduction in Joints 2 and 3
has to do with the absence of traction, or extension, in the method of Joints
2. A combined forcing of the elbow towards the ribs and of the humerus
away from the ribs could be considered a leveraged action, since it forms
a common movement combining opposite directional components in
a confined space. One could visualize in this case a circular movement
originating at two points: the fist at the ball of the humerus and, diagonally
from this point, the tip of the elbow being pushed towards the ribs. It is the
method of Joints 3 that I have questioned as involving anything more than
traction and a gentle push. Interestingly, traction is most often regarded as
leverage in ancient medical accounts, because tools for putting someone in
traction use levers and winches.
The contrast of ostension and ostensive definition applied to ancient
techniques helps to disentangle the strands of the physician’s experiential
grasp of mechanical phenomena in these methods. The Hippocrates of
Joints is not drawing upon any mechanical tradition that would make the
lever as such central to discussions of amplification of power, which is the
concern of the Aristotelian Mechanics. Simple shifting of a part of a body is
his main concern, not making a lesser power greater. Even considering his
account in its own terms, it cannot be said that he has codified his
multipartite ways of nudging or forcing bones into place in any terms so
simple as the mechanical accounts of the Aristotelian Mechanics and of
Archimedes. He does not possess a principle. He has instead a number of
methods of reduction.
His grasp of how to move the humerus back into its joint does reflect the
assimilation of detailed observation and the ability to modify techniques in
light of his experience. He shares these details, which lend themselves to
flexibility in the application others might make of the techniques. The gist
of the two chapters examined, however, is a very clear presentation of
techniques to be imitated by other practitioners. To this extent,
Hippocrates’ On Joints 2 and 3 are instances both of the natural and
unremarked-upon activity comparable to ostension – doing something
by intuitive and sensory cognitive habit but not reflecting upon it – and
of know-how, the close attention to detail, including of oppositions and
similarities, that fosters art (technê). He presents his techniques for novices

who wish to imitate this sensory cognitive habit and develop it in
themselves.
On the basis of the analysis above, Hippocrates’ accounts in Joints 2 and
3 can be seen as an intuitive grasp of physical forces that is in its own way
broader and deeper than the later principled accounts of the lever. This is
because his accounts are embedded in sensory-motor uptake of undergoing
forces in competition. This uptake is experience and involves perceptual
judgement. The role of know-how is evident in the account of Joints 2,
where an unmoved pivot (the practitioner’s head against the patient’s
acromion) lies outside the place of the primary action. In order to create
the conditions for relocation, consideration of a wider field of action than
the place of dislocation is required. This ‘insight’ is sensory motor in origin
but is developed by the physician by modification and coordination. The
outside pivot point is what sustains the counter-movements within the
triangle of (1) the lower arm (positioned horizontally), (2) the humerus
slightly askew downwards, and (3) the fist in the armpit. Hippocrates’
account thus incorporates pressures or positionings meant to counter one
another at a distance rather than the ordering of effects in a straight line.
His know-how presupposes what we would call a system of forces.
The notion of a system of movements in early modern mechanics
involved an underlying reference to an equilibrium of influences played
out as constrained motion. The equilibrium lies in the forces ordering
themselves, that is, constraining one another, by mutual influence. There is
something akin to this notion of system in Hippocrates’ gathering of
counterpoints of position, stationariness and movement to effect
a relocation of a shoulder. In Hippocrates’ case, the recognition of system
would be part of his practice, since the human skeletal framework is already
articulated and works in concert with tendons, cartilage and muscle. The
parts are in contact and respond to one another, action inducing reaction.
Although one could argue that he implicitly utilizes an equilibrium of
counterpoints of motion and stationariness, which are key features of
leverage, there is no reason why Hippocrates would think of what he
does in terms of equilibrium. His practice has an aim, the return of the
humerus to its functional position, and this is his cognitional point of
reference.
From his use of verbal forms having to do with counter-movements,
counter-pressure and propping against something for stability and resist-
ance, it is reasonable to conclude that Hippocrates had his own practice-
based grasp of a variety of phenomena that would have contributed, in
more clear-cut circumstances, to delineation of a science (mathêma) of the

142 jean de groot
lever. Restoring health in the human body does not lend itself to the
simplification sought in the mathematical sciences of antiquity.
Iatromechanics in this instance, then, is considerably different from what
is recognized as mechanics even 100 years after Hippocrates.
I remarked above (p. 140) that the association of leverage as convention-
ally understood with Hippocrates’ reductions of joints has much to do with
the use of simple machines to sustain traction. At the same time, traction
itself is not leverage. Traction does involve a contending of contrary
movements, which in a different context is one facet of leverage. The
methods of reduction given in Joints 2–4 do not involve machine aids, or
even tools – artificial ‘pieces’ like a beam. Moving on to Joints 4,
Hippocrates says quite directly that the method he describes there is
valuable because it does not require surgical apparatus (ta armena) and
can be done on site at the gymnasium (Arth. 4.16–19).
Joints 4 describes a person taller than the afflicted person placing his own
shoulder firmly into the other’s armpit. He draws the dislocated arm
towards his chest as best he can. Then:
In this configuration, let him shake up and about when he raises the man off
the ground, so the rest of the man’s body exerts weight in opposition
(antirrhepoi), set against (antios) the arm held fast. If the person is light,
let a child of light weight be suspended on him from behind. (4.11–16)
This method is a variation on the method of Joints 2, which involves
movement or immobility of parts of the physician’s own body. The lifter’s
shoulder may jostle the ball of the humerus like the physician’s fist in Joints
2. On the other hand, it may be more important that this method in Joints
4 makes use of the weight of the injured person’s body, which Hippocrates
describes in the quotation as exerting force over against the dislocated limb.
The additional weight of the child hanging on behind adds needed
counter-force. He uses a word, antirrhepô, which means to initiate, or
bring about, a movement in opposition.27 The weight of the body, all on
the side opposite the dislocation, opens up the axilla, with the shaking at
the point where the patient is lifted off the ground. Still, the opposition of
forces (rhopai) is made more explicit and takes place around an unmoved
place in the armpit, namely, where the operator’s shoulder is securely
placed. The formulation quoted above is the clearest indication of
Hippocrates’ understanding of weights acting against one another to effect
27
The noun cognate, rhopê, means an inclination to move that is effective enough to get movement
started or underway. For comparison, see Aristotle, De Caelo 4.1, 307b31–3; Mechanics (Aristotle
2000 iii.7 = ch. 8 Teubner and Loeb), 851b25–35.

movement. This method pitches the weight of the injured person’s body
against the stability of the lower arm held fast. Again, though, the result
seems to be traction and the consequent extension of the space of the joint,
which expedites the movement of the bone back into place. Like Joints 3,
the account in chapter 4 uses a rigid extender placed in the armpit.
Hippocrates almost always combines katatasis and mochleusis in his
description of reduction techniques but regards them as two different
aspects of the procedure. Extension is traction but mochleusis is the forced
moving of the bone.
Some of the reduction methods Hippocrates has presented are roughly
in accordance with nature (kata phusin), he says. The heel-and-leg pressure
method is ‘nearly in accordance with nature’ (eggus ti tou kata phusin)
(Arth. 3.2). According to Fractures 1.1–3, a natural method (hê dikaiotatê
phusis) in both fractures and dislocations involves stretching along
a straight line. Apollonius, a commentator on Hippocrates, says that
a method is not in accordance with nature unless it involves traction,
which is moving bones into alignment by stretching. Kata phusin on this
interpretation seems to mean allowing the joint to correct itself by provid-
ing the space for it to do so.28
Another method Hippocrates says is nearly natural to the same extent is
one in common use by those forcing movement around pestles (hoi peri ta
hustera anagkazontes) (Arth. 5). The end of the pestle is wrapped in a soft
cloth to keep it from slipping and is placed, like the knuckles of the hand in
Arth. 2, between the ribs and the ball of the humerus in the armpit. This
reduction can be done while the patient is seated, but it is best if the pestle
is longer than his arm so as to accomplish a slight suspension of the joint
above its normal position. The arm lies along the length of the pestle. This
method uses a tool but accomplishes the same thing as the human lifter of
Joints 4.
The main action inducing relocation in the method of Joints 5 takes
place on the other side of the body, where someone putting his arms
around the patient’s clavicle forces the body downwards on that side.
The dislocated shoulder would tend to follow the rotating clavicle around
with the push downwards on the other side of the body. Although similar
to the description of circular motion in the Aristotelian Mechanics,
Hippocrates does not call this a circular motion. To relocate the bone
depends on the humerus not following the other side of the body but
instead sliding into its socket. In the method of Joints 2, the fist in the
28
On Apollonius, see Hippocrates 1928, 205.

144 jean de groot
armpit pushes the ball of the humerus away from the ribs. In the method of
Joints 3, the leg and heel of the physician slide the humerus into place.
Similarly, in the method of Joints 5, the pestle obstructs the inclination of
the dislocated shoulder to follow the rotation of the body on the other side,
and this transfers the downward movement on the healthy side of the body
into a movement of the pestle’s pushing the humerus into its socket.
Although involving a tool, the pestle, Hippocrates says the method of
Joints 5 is ‘nearly natural’ (5.2).
The method of Joints 6 is similar but better, Hippocrates says, ‘because
the body is more safely compensated in one direction, then the other (to
men têi to de têi antisêkôtheiê), when raised’ (Arth. 6.2–4).29 This method
places the armpit on the step of a ladder, adding something rounded
between the armpit and the ladder, ‘which helps to force (prosdianagkazei)
the head of the humerus back into its natural place (es tên phusin)’ (Arth.
6.8–10). This method still involves raising the body up off the ground. We
can note in the succession of methods presented from Joints 4–6 the
addition in each case of something more in the way of artificial aids. In
Joints 7, Hippocrates presents what he regards as the most effective
(krastistê) method, which is, like the method of Joints 4, modelled on
utilizing the weight of the person’s body. The method of Joints 7 also
involves apparatus to an even greater extent than the methods of Joints 4–6.
A key part of the effectiveness of the method is a plank (xulon) that has a lip
or projection on one side of the piece of wood, which has been tapered
(Arth. 7.5–9) (Figure 5.5). It is important for the end with its lip to be set
well into the armpit (Arth. 7.7–12). With the arm extended and tied to the
outer reaches of the tapered board, the lip or rim should be above the
dislocated ball of the humerus in the armpit and the rest of its wood piece
between the humerus and the ribs (7.14–23). Then the arm with its wooden
support is placed on a crossbar situated between posts so that the arm and
pestle are both on one side of the crossbar and the person’s body is on the
other side. The bar should be high enough for the body to be slightly
suspended off the ground.
In this arrangement, the body and the arm are both forced downwards
(katanagkazein) on their respective sides of the crossbar. With the crossbar
keeping the body raised, this is the ‘strongest’ (kratistos) method of
29
Withington translates: ‘the body is more safely kept in equilibrium on either side’. The word
isorrhopia, for equilibrium or balance, is not in the text, and since the use of particular words is
important for my argument, I use the more basic meaning of antisêkoô as ‘compensate for + dative’,
noting the verb’s combination with ‘in one direction, in the other direction’ (datives) in a men . . . de
construction.

Figure 5.5 A bone-reduction method with a plank (xulon) that has a lip or projection
on one side of the wooden piece
reduction of the shoulder (7.26–34). Hippocrates makes two statements

relevant to his grasp of the physical forces at work in this operation. He says
dikaiotata men gar mochleuei (7.34) . . . dikaiotatai de hai antirrhopai (7.36).
The passage containing these phrases reads:
For this method (tropos) is by far the strongest for reduction of the shoulder.
For it moves the bone most justly (34), if only the wood is well on the inner side
of the head of the humerus. The counter-inclinations are most correct (36) and
without risk for the bone of the arm.

146 jean de groot
Dikaiotata in this context (used first adverbially and then as a plural
predicate noun) has connotations of naturalness, effectiveness and
accuracy.30 Hippocrates praises the natural perfection of the method. It
does the least harm to the arm and bone (7.37). In shoulders recently
dislocated, the reduction is accomplished remarkably quickly, and the
method is even successful with older dislocations (7.39–41). The second
statement reveals something of the meaning of mochleuei for Hippocrates.
In reduction (embolê), inclinations to movement (rhopai) are aroused by
forceable actions combined with appropriate positioning of those actions.
These inclinations to movement, positioned so as to oppose one another
(antirrhopai), cause a slight but highly targeted movement elsewhere in the
system, in the present case at a point between the two opposite points of
exertion of effort. The ball of the humerus slips into its socket.
A condition of the effectiveness of this method, Hippocrates cautions, is
that the lip of the tapered wooden pestle be positioned well inside the
armpit and above the head of the humerus (7.35–6). Concerning the
success of the method for older dislocations (7. 39–41), he says that, even
in the face of invading tissues and other impediments, it can succeed: ‘For
what would not correct mochleusis move?’ (7.46–7). He immediately notes
that an old dislocation, even if reduced by this method, might easily slip
back out (7.47–8). This fact only reinforces that mochleusis overcomes
obstacles, even when a habituated but impaired nature tends to undo the
effect.
The accuracy, easy healing and low failure rate that Hippocrates ascribes
to this method seems to take the author himself by surprise. Hippocrates
does not say this method is natural, perhaps because it has more than one
artificial component. He nevertheless acknowledges it having the highest
rate of success. Thus, his statement, ‘For what would not correct mochleusis
move?’ is a declaration of wonder at the automaticity of the action, and
perhaps also wonder at its acting not at the site of forced motion (on either
side of the body) but just where a precisely crafted action is needed
somewhere in between.
30
Compare hautê gar hê dikaiotata phusis of Fractures 1.3. In this passage, Hippocrates says that
stretching (katatasis) in a straight line is the most natural for resetting fractures and dislocations.
I understand dikaios here to mean just in the sense of rightness with respect to the physical situation
of the body – the restoring of the healthy disposition of the joint. Craik notes the original meaning
of the word as ‘path’, that is, right path. She sees in Fractures and Joints an additional connotation
indicative of the ‘moral superiority’ of Hippocrates’ method to that of others (Craik 2010, 231–32).
The implication is that competition with the sophistic strain in medical authority is a factor in the
meaning of dikaios here.

In presenting the series of methods of reduction of dislocation of the
shoulder found in Hippocrates’ On Joints 2–7, I have noted the similarities
and dissimilarities to leverage as that mechanical phenomenon has been
defined from ancient times. Hippocrates does not describe his methods as
leverage or his tools as levers. This is because he has his own understanding
of the opposition of physical forces within the human body and impinging
upon it from outside the body. In particular, he does not think in terms of
equilibrium. He does not use the word fulcrum. Some of his accounts do
not involve anything resembling a fulcrum, while others do. His accounts
do involve an opposition of movements or tendencies to move, but these
are generally pitted against one another so as to produce stretching or
extension of a joint. What coaxes the ball of the humerus into place is (1)
the situation of tension combined with (2) the opening of the space for the
bone to move. Mochleusis is pushing the bone into place.
IV
The distinction between ostension and ostensive definition applied to

Hippocrates’ language in Joints 2–7 sometimes plays very directly into
the making of judgements about the meaning of a technique in use,
because it is the very use of nouns, like lever or pulley, versus verb forms
of mochleuô that indicates whether principled knowledge of a phenomenon
is present or not. To use a noun ostensively, even without reflection, is to
acquire some mastery of the thing named, especially since a thing can be
named in its absence. The noun isolates a phenomenon and promotes
detachment in thought and speech from what is named. Verbal speech, on
the other hand, shows a process emerging.31 The process thereby becomes
replicable, but more importantly it becomes understandable in the ver-
nacular of art. This feature of verbal speech is amply displayed in a close
reading of Joints 2–7. Speech rich with verbal forms is closer to phenomena,
and its imperatives, optatives and participles commit the lecturer or writer
to iterating the action – to rehearsing and recreating it for another.
The predominance of verb forms in setting up the key elements in
reducing a joint testifies to the physician’s art as engagement in process.
Rather than nominalizing a situation, the key aspects of which are for the
most part simultaneous actions, Hippocrates adds prepositional prefixes to
verbs. I have noted these in the course of analysing the methods of
reduction in Joints 2–7. To lengthen and specify verbs with prepositions
31
On the features of nominalized versus verbal discourse, see Sokolowski 1978, chs. 1–4.

148 jean de groot
of direction and iterative action reinforces the verbal as opposed to the
nominalizing character of his discourse. These verb uses all show that the
physician is not reporting what he knows but relating an accomplishment.
His habitual success in this accomplishment resides in the ‘depth percep-
tion’ of forcing and counter-forcing, which may take place by applying
a fist, a palm of the hand, a heel or a stiff board. This difference between
reporting and accomplishing is one of the defining differences between
science, which reports authoritatively, and an art like medicine, which aims
at restoring some natural condition.
I have relied on Hippocrates’ language to make a claim that the modern
notion of leverage is applied somewhat too broadly to the surgical art of
reducing dislocations. An objection that also draws on Hippocrates’ lan-
guage can be raised to my claim.32 What appears to be a strong reason for
seeing Joints as being about the lever and whatever principles underlie it are
parallel passages in Fractures, a treatise closely related to Joints and the later
Mochlikon. In Fractures,33 the context is reducing fractures in which the
bone projects outside its normal alignment (Peri agmôn 31.45–46). The
bone is not just broken. It also projects outwards. Such a fracture can break
the skin and be part of a more extensive wound. In such a circumstance,
Hippocrates recommends the use of several levers (mochloi) such as stone-
masons use, made out of iron and narrower at one end while wider at the
other (to me ti platuteron, to de ti stenoteron) (Peri agmôn 31.48–52). Given
the confined spaces involved and his reference to several of the tools, he
may have in mind the tool combining the lever and wedge used to separate
deeply scored stone into blocks.34 The stonecutter places these in a series
along the score. For moving the fractured bone, Hippocrates says, the
levers should be pressed downwards on a stable bone situated below the
fracture while forcibly lifting the accompanying fractured bone into line.
Perhaps because the author treats this as a manoeuvre that may involve
32
In Fractures, Joints and Mochlikon, forms of the noun mochlos appear twelve times and the verbal
forms of mochleuô nineteen times. This seems to make the usage of the noun approach fairly closely
the use of the verb. Hippocrates uses the noun, however, to refer to particular tools in existence, like
‘the levers that the stonecutters use’ (Peri agmôn [On Fractures] 31.49) or in reiterative phrases with
the verb (Arth. 72.30, 74.4). The nominative mochlos appears most often in cases where a reduction
requires more strength because the bone moved requires a large piece of wood to move it, as in
relocating a hip (Arth. 74.7). Hippocrates does not understand his own uses of mochleusis to involve
a fulcrum, but he sometimes does involve propping or bracing against (ereisas) something, as in Arth.
74.10.
33
Peri agmôn 31.47–74.
34
The author of the Aristotelian Mechanics describes the wedge as two levers (853a21–22) (Aristotle
2000, III, ch. 16).

considerable strength, he praises leverage and goes on to praise other tools
the surgeon might need:
This is a great help, when the irons are suited to the task and someone exerts
leverage as fitting (mochleuêtai ôs chrê). For of all the apparatus that have
been devised by man these three are the strongest of all: the turning round of
the millstone (onou periagôgê), leverage (mochleusis) and use of the wedge
(sphênôsis).35 (Peri agmôn 31.58–63)
No feat requiring great strength will be accomplished without these.
Levering movement (mochleusis) is not to be despised, then, for the bone
will be reduced by it, or not at all (31.63–7).
Hippocrates shows in this passage his knowledge of the lever as
a simple machine used in a variety of practical pursuits, while also
praising the remarkable feats of the device. His noun mention of the
lever, however, is limited to citing the tools of the stonemasons to
indicate what the shape of the irons should be. His references to
leverage in the context of bone setting use the verbal noun (mochleusis)
or forms of the verb. The middle voice of mochleuô (Peri agmôn 31.60)
is used here for someone prising or prying in the proper way so as to
restore the bone to its place.
In a similar passage, the author of Mochlikon uses words for the windlass
(onos), wedge (sphêniskos) and press (ipos) (Moch. 38.1–3).36 He says that the
windlass is for extension (anagein), while the mochlos is for bringing
a dislocated bone along to its socket (paragein). The very separation of
these simple machines in speech from what they are to accomplish is
noteworthy. The author of Mochlikon gives a reason why a lever would
be needed in the kind of cases treated in Fractures 31:
In the case of finger or toe, foot, hand, wrist, spinal curvature, double
extension (dianagkasai) and forcing down the projection (katanagkasai)
are required; in the other cases, separation by hand-power is enough, but
one must force projecting parts (ta huperechonta) into position with the heel
or palm over something (epi tinos) taking care that a suitable soft pad is
placed under the projection. (Moch. 38.12–17)
35
I have revised Withington’s translation (Hippocrates 1928, 173) slightly.
36
The parallel passages citing the three kinds of device in both Fractures and Mochlikon raise the
question of interpolation of the passage in Fractures (31.61–4). The enthusiastic digression on their
indispensability for achieving feats of great strength (64–5) is out of keeping with Hippocrates’
otherwise sober discourse hewing close to the subject at hand. The text could proceed smoothly
from 60 to 68. On the other hand, the passage could be included as a response to a criticism about
using mochloi, since the author says that in these cases, the fracture will be corrected with the tools or
not at all (66–7).

150 jean de groot
A projecting fracture is sufficiently recalcitrant to forcing (anagkazô) that
use must be made of an adjacent bone actually in contact with the reducing
tool and underneath it. This bone underneath is a prop stable enough to
make the procedure work.
These two passages (with special emphasis on the one from Fractures 31,
which is probably earlier) show that Hippocrates knew the tools called
levers well from other craft contexts, and he recognizes the assets of the
device – providing additional strength on the condition of having a sound
platform for propping – that would serve especially difficult circumstances
in restoring articulations. In Joints 72, Hippocrates uses mochlos (30) for
a large plank (xulon) placed in an apparatus involving a windlass (oniskos)
on each side and a crossbar, which is used to reduce dislocations around the
hip (Arth. 72.7–9). Here the lever is combined in a complex device with
other simple machines. Do these explicit references to the lever weaken my
argument for how to understand iatromechanics in the treatise Joints from
the late fifth century bc?
I do not think so. Living in a preindustrial society, it would be difficult
for any craftsman, including a physician, to miss the enduring advantages
in many sorts of action made possible by simple machines, which were
ubiquitous in the ancient world. Just because someone knows a noun well
enough to refer to its object in its absence does not mean that they are
invoking that meaning in every other circumstance where we may suppose
its principles at work. G. E. M. Anscombe, in explicating Wittgenstein on
ostension, imagined a culture that lacked a word for horse but used ‘some
verb form signifying something like horse presence’. We could not say of
such a people that ‘they “fail to realize something that we realize”’.
Similarly, Hippocrates uses verbal forms for forcing actions, some of
which come close to what we call leverage. He is not unaware of the forces
later codified as leverage. Presenting the alternative case, that such a people
have the word ‘horse’, Anscombe says we cannot suppose they do not
understand what they are doing when, on newly discovering cows, they
apply the count-noun ‘horse’ to them.37 Similarly, Hippocrates uses
mochleusis for a wider variety of phenomena than simply what levers do.
That does not mean that he is unaware that levers, so named, do
a particular kind of work in particular circumstances. Hippocrates’ art
shows that there is not only one way to understand mechanical forces.
My argument has been that Hippocrates has his own sort of understand-
ing of the physical forces that sometimes fall into our category of leverage.
37
Anscombe 1981, 114–15.

It is notable that he separates traction and how to establish traction from
the movement called mochleusis. These are separate components of the
reduction of a joint.38 It is more often extension that is accomplished by
a lever-like tool. Mochleusis means in particular moving a bone to re-secure it
in its natural position.
It is possible that the fact that any reduction takes place within the physical
constraints of injury to a part of the body heightens the general awareness of
the closedness of a physical state of affairs and the awareness of counter-action
taking place only within the state of affairs. Closure under this description
seems to attend the working of a lever, especially in its scientific abstraction.
Perhaps mochleusis in medicine has a connotation of movement within
a similar state of affairs of confinement and counter-tension. Certain kinds
of movement to relocate the bone simply cannot be undertaken, given the
injury. Tissues are damaged, or stretched and out of place. The route to
pushing a bone back into place is blocked, because the tendons have closed
into the joint. In general, there is resistance from the surroundings to reducing
a joint, so that any reduction is overcoming obstacles. As I have shown through
examination of Joints 2–7, this does not mean that either traction or the actual
moving of the bone is leverage as understood in the later scientific tradition.
This interpretation of Hippocrates’ language in terms primarily of action
is reinforced by the claims of the author, possibly Hippocrates, of On the Art.
The author is responding to sophists or philosophers who would capitalize
on Parmenidean reasoning to deny the existence of technai. Jouanna links
this defence in On the Art to the ‘birth of epistemology’, and along with it
the birth of science, in the late fifth century bc.39 The arts must defend
themselves against the claims of a supposedly superior science:
Whereas the things-that-are always are in every case seen and known, the
things-that-are-not are neither seen nor known. Accordingly, the arts are
known only once they have been taught, and there is no art that is not seen
as an outgrowth of some form (ek tou eideos ouk horatai). In my opinion,
they acquire their names, too, because of their forms. For it’s absurd – not to
mention impossible – to think that forms grow (blastanein) out of names:
names for nature are conventions imposed by and upon nature (phusios
nomothetêmata), whereas forms are not conventions but outgrowths
(blastêmata). (Peri technês 2.2–3)40
38
Gerolemou in this volume (Chapter 4) also points out that the Hippocratics distrusted the elaborate
device, since it was often used more to impress and to cover incompetence than genuinely to heal.
39
Jouanna 1999, 243–258.
40
Mann 2012, On the Art of Medicine, 69. See Mann’s commentary on this passage, especially in
relation to the form of an art (97–100). The treatise is considered to be among the late fifth-century
contributions to the corpus (Jouanna 1999, 378; Mann 2012, 39). At 2.3, ta men gar onomata phusios

152 jean de groot
The author is defensive, but he is not just saying that of course the
arts must exist because we see them practised. An art is manifest in its
ability to be taught.41 Teachability reflects the reality and repeatability
of medicine’s formations of regular practice. These cannot exist without
some connection to nature. These (natural) formations of medicine are
actually seen and known. This is a subtle point meant to highlight the
grasp that medicine has upon the natural processes of health and its
restoration at which the art aims.42 He continues, saying that it is
absurd that the names of the arts, specifically medicine, could have
come first and a detailing of the articulations of the body, its parts like
bile and phlegm and the aim of health then be developed later to fit
the name. This defence of the medical craft emphasizes that names are
secondary to activities, and it makes a claim that these activities are in
cooperation with nature.
My purpose has been to loosen the terms within which Hippocrates’
treatises on fractures and joints are considered so as to understand better
the cognitive grasp of physical forces suited to medical art. The reason for
doing this is to sharpen the investigator’s view of the differences among
empirical intuition, art and science in the ancient period. I have used
characteristics of language itself to drive home certain points about the
difference among these pursuits. The focus has been on empirical intuition
and art in examination of a foundational issue in iatromechanics, the very
meaning of leverage. With this focus, my aim has been to lend substance to
the difference between the cognitive basis for the art of medicine in
experience and the art itself, which remains close to empirical intuition
while transforming it into something that is both an organized practice and
a way of understanding. This way of understanding is distinctive (in
a sense, cannot be otherwise), because characteristic to it is the mutual
involvement of medicine and nature. Medicine aims at the rectification of
health in the way that least impedes the natural state of the human body.
The human body in its natural condition shows what health is. The
implied circularity is interesting. A science calls for deduction, and circu-
larity is a flaw. In the medical art, however, the circularity frames the
expertise that is to be conveyed to another.
nomotethêmata could read ‘the names of nature (meaning “of things of the natural world”) are
conventions’. On phusis before the fourth century bc, see Zhmud 2018, 57.
41
That craft can be taught is a criterion noted by Aristotle, Metaphysics A.1, 981b8.
42
As Mann says, ‘there is no evidence that he countenances the possibility that medicine could be and
yet not be an art’ (Mann 2012, 1).

RE F E RE N CE S
Adams, F. 1849. The Genuine Works of Hippocrates (London).
Anscombe, G. E. M. 1981. ‘The Question of Linguistic Idealism’, in From
Parmenides to Wittgenstein: Collected Philosophical Papers i (Minneapolis),
112–33.
Archimedes. 1972 [1910–16] De planorum aequilibriis sine de centris gravitates
planorum. Opera Omnia ii, ed. J. L. Heiberg, 2nd ed. (Leipzig).
Aristotle. 1978. De Motu Animalium: Text with Translation, Commentary and
Interpretative Essays, ed. M. Nussbaum (Princeton).
Aristotle. 1888. De Plantis, de Mirabilibus Auscultationibus, Mechanica, de Lineis
Insecabilibus, Ventorum Situs et Nomina, de Melisso, Xenophane, Gorgia, ed.
O. Apelt (Leipzig).
Aristotle. 1980. Mechanical Problems, in Minor Works, trans. W. S. Hett
Aristotle. 2000. Problemi meccanici: Introduzione testo greco, traduzione italiana,
note, ed. M. E. Bottecchia-Dehò (Padua).
Aristotle. 1968 [1937]. Progression of Animals, Bekker text with revisions by
E. S. Forster, in Parts of Animals, Movement of Animals, Progression of
Animals, Loeb Classical Library 323 (Cambridge, MA).
Bortone, P. 2010. Greek Prepositions from Antiquity to the Present. Oxford
Linguistics (New York).
Cañizares, P. P. 2017. ‘On the Applicability of Hippocratic Therapy’, in
M. Formisano and P. J. van der Eijk, eds., Knowledge, Text and Practice in
Ancient Technical Writing (Cambridge), 93–111.
Craik, E. 2010. ‘The Teaching of Surgery’, in M. Horstmanshoff, ed., Hippocrates
and Medical Education: Selected Papers Presented at the XIIth International
Hippocrates Colloquium. Studies in Ancient Medicine 35 (Leiden), 223–34.
Craik, E. 2015. The ‘Hippocratic Corpus’: Content and Context (Abingdon).
Cross, J. R. 2018. Hippocratic Oratory: The Poetics of Early Greek Medical Prose
(Abingdon).
Dean-Jones, L. 2003. ‘Literacy and the Charlatan in Ancient Greek Medicine’, in
H. Yunis, ed., Written Texts and the Rise of Literate Culture in Ancient Greece
De Groot, J. 2014. Aristotle’s Empiricism: Experience and Mechanics in the Fourth
Century bc (Las Vegas).
Engelland, C. 2014. Ostension: Word Learning and the Embodied Mind
(Cambridge, MA).
Formisano, M. and van der Eijk, P. J. 2017. ‘Introduction’, in M. Formisano and
P. J. van der Eijk, eds., Knowledge, Text and Practice in Ancient Technical
Writing (Cambridge), 1–11.
Heath, T. L. 2002 [1897]. The Works of Archimedes (New York).
Heron of Alexandria. 2016. The Baroulkos and the Mechanics of Heron, eds.
G. Ferriello, M. Gatto and R. Gatto (Florence).
Hessinger, J. J. 1978. ‘The Syntactic and Semantic Status of Prepositions in Greek’,
Classical Philology, 73 (3): 211–23.

154 jean de groot
Hippocrates. 1840. Oeuvres Complètes d’Hippocrate, Emile Littré (Paris).
Hippocrates. 1928. Hippocrates iii, trans. E. T. Withington, Loeb Classical Library
149 (Cambridge, MA).
Holmes, B. 2013. ‘In Strange Lands: Disembodied Authority and the Role of the
Physician in the Hippocratic Corpus and Beyond’, in M. Asper, ed., Writing
Science: Medical and Mathematical Authorship in Ancient Greece (Berlin),
431–72.
Jouanna, J. 1999. Hippocrates, trans. M. B. De Bevoise (Baltimore).
Krafft, F. 1970. Dynamische und Statische Betrachtungsweise in der Antiken
Mechanik (Wiesbaden).
Mann, J. E. 2012. Hippocrates: On the Art of Medicine. Studies in Ancient Medicine
39 (Leiden).
Pappus of Alexandria. Synagôgê. Thesaurus Linguae Graecae, http://stephanus
.tlg.uci.edu.
Perilli, L. 2018. ‘Epistemologies’, in P. E. Pormann, ed., The Cambridge
Companion to Hippocrates (Cambridge), 119–51.
Sokolowski, R. 1978. Presence and Absence: A Philosophical Investigation of
Language and Being (Washington, DC).
Witt, M. 2018. ‘Surgery’, in P. E. Pormann, ed., The Cambridge Companion to
Hippocrates (Cambridge), 217–45.
Zhmud, L. 2018. ‘Phusis in the Pythagorean Tradition’, Philologia Classica, 13 (1):
50–68.

chapter 6
Hippocrates’ Diseases 4 and the Technological Body

Colin Webster
Πολλάκις γὰρ αὐτοῦ τοῦ δένδρου μέρος τι [αὐτοῦ] συνελήφθη ὑπὸ

θατέρου συμφυοῦς γενομένου. Καὶ ἐάν τις ἐκγλύψας θῇ λίθον εἰς τὸ
δένδρον ἢ καὶ ἄλλο τι τοιοῦτο, κατακρύπτεται περιληφθὲν ὑπὸ τῆς
περιφύσεως. Ὅπερ καὶ περὶ τὸν κότινον συνέβη τὸν ἐν Μεγάροις τὸν
ἐν τῇ ἀγορᾷ· οὗ καὶ ἐκκοπέντος λόγιον ἦν ἁλῶναι καὶ διαρπασθῆναι
τὴν πόλιν, ὥσπερ ἐγένετο < . . . > Δημήτριος· ἐν τούτῳ γὰρ
διασχιζομένῳ κνημῖδες εὑρέθησαν καὶ ἄλλ’ ἄττα τῆς ἀττικῆς
<σκευῆς> κρεμαστὰ [ὅ ἐστιν ἐν κοτίνῳ] οὗ ἀνετέθη τὸ πρῶτον
ἐγκοιλανθέντος.
(Theophrastus, HP 5.2.4; Amigues)1
For often some part of a tree itself is absorbed by another portion

growing around it. For instance, if someone carves out a hole and puts
a stone or some other such thing into a tree, the object is covered over,
having been enveloped by the surrounding growth. This very thing
occurred to the wild olive in the agora at Megara. An oracle stated that
if this were felled, the city would be conquered and plundered. And this
very thing happened [when] Demetrius [took it]. For, when the tree was
cut apart, greaves and other things of Attic workmanship were discovered
hanging inside the wild olive where the first object had been dedicated
and the hole made.
The wild olive in the Megarian agora swallows the votives affixed to it. The
process takes time, but eventually this tree pulls all the attached specimens
of Attic workmanship into itself as it grows around them, literally incorp-
orating these objects into its own mass. It does not fully assimilate these
tools by breaking them down. It absorbs them whole. For Theophrastus, as
for us, this tree displays no unique or monstrous power but simply repre-
sents a vivid instance of the common vegetal capacity to enfold adjacent
objects. This chapter will explore an analogous absorptive power displayed
1
All translations are my own unless otherwise noted.
155

156 colin webster
by the human body as a conceptual unit, seen most clearly in the
Hippocratic treatise Diseases 4, as well as two closely related texts,
Generation and On the Nature of the Child. These works envision a body
whose inner matter is punctuated with parts that behave like tools and
implements, many of which are drawn from medical practices. This
chapter will examine the question of iatromechanics in the fifth century
bce by exploring the physical and conceptual interface of medical tools
and the bodies that they touch. It will illustrate how the tools used by this
Hippocratic author to intervene in the body often merge with explanations
of those same processes, as the author conceptualizes the inner behaviours
of the body using the same tools with which he treats the patient physically.
Or, seen from another perspective, we might say that the implements,
objects, and substances applied to the patient are absorbed into its concep-
tual interior. Like the wild Megarian olive, the Hippocratic body can
enfold adherent technologies.
Debates about plants, tools, and mechanism have long surrounded this
cluster of treatises in particular. In his 1981 article ‘Hippocrates the
Iatromechanist’, Iain Lonie interrogated Friedrich Hoffman’s insistence
that Hippocrates was a mechanical thinker by outlining three different
ways in which the term ‘mechanical’ can be understood to describe
explanations: those that (1) involve mathematical laws; (2) rely on recog-
nizable physical forces alone; or (3) are modelled on machines or automata.
Lonie argues, contra Hoffman, that the corpus does not show a proclivity
towards mathematical laws, nor do Hippocratic authors cast the body as
a type of machine or automaton – indeed, how could they, since mechanics
as a discipline was not developed until the Hellenistic period?2 What Lonie
does assert is that Diseases 4, Generation, and On the Nature of the Child all
employ multiple analogies with tools and implements to explain how fluids
are moved around the body. For instance, Diseases 4 compares the head to
a cupping vessel, likens the interior organs of the body to interconnected
bronze bowls and invokes oil flasks, sediment in a swirled kylix, iron process-
ing, and an inverted amphora. Lonie sees these comparisons as part of this
particular Hippocratic author’s broader tendency towards mechanism of
the second type, that is, an attempt ‘to reduce all explanations to the contact
and pressure between solid and fluid bodies’ (Lonie 1981a: 130–1). Lonie
admits, however, that this same author will also mix this type of explanation
2
Compare Berryman 2009, who likewise argues that any assessment of ancient mechanistic thought
must be rooted in the historical development of mechanics as a formalized discipline in the
Hellenistic era.

Hippocrates’ Diseases 4 and the Technological Body 157
with other ‘animist’ or ‘vitalist’ arguments that seem to attribute agency to the
parts and substances of the body. The mechanism in these texts is thus, for
Lonie, inconsistent, unconscious, and avant la lettre, while the moments of
animism seem embarrassing and in need of defence.3
Other commentators place greater value on the abundant plant analo-
gies in these treatises, seeing them not as moments of regressive ‘animist’ or
‘vitalist’ tendencies, but sites of ontological interest.4 For instance, Brooke
Holmes has recently explored how Generation and On the Nature of the
Child repeatedly model the embryo on plant and arboreal growth, and she
suggests that these extended accounts actually reframe the accompanying
mechanical analogies.5 For example, when the Hippocratic author argues
that the embryo grows and articulates itself under the influence of pneuma
and the resultant collection of ‘like with like’, he illustrates this principle
with a comparison to a bladder filled with earth, sand, water, and lead
filings, which all separate into discrete layers when subjected to sustained
inflation (Nat. Puer 6, 7.498 L).6 This configuration forms an image of the
embryo itself, insofar as the inserted pipe entering the bladder mimics the
umbilical cord, which likewise supplies breath to the growing embryo.7 Yet
the author then likens the articulation of the embryo’s inner and outer
parts to a ramifying tree, the motive force provided by an external source of
inflation now replaced by the inner breath of the plant itself. This com-
parison expands into a long discourse on tree growth. For Holmes, this
later comparison reframes homogeneous attraction as self-directing and
remakes ‘the self-sustaining attraction of like to like as characteristically
vegetal’ (Holmes 2017: 365).8 Rather than occasionally resorting to animist
arguments when good mechanistic arguments are not available, Holmes
see the plants as the more dominant analogical partner in understanding
the embryo and its development.
3
Compare Lonie 1981b: esp. 72–86, which outlines the general scientific methodology of the treatises
and adopts the same defensive stance; and pp. 150–2, which poses the question of mechanism and
vitalism in the embryological treatises.
4
For the prevalence of plant analogies, see Lonie 1969; Lloyd 1978: 45–9; Jouanna 2002: 266–7;
Holmes 2017.
5
Holmes 2017.
6
This bladder with a pipe attached is likely a douche; see Lonie 1981b: 184. For the convenience of
English speaking readers, all references to the Genit./Nat. Puer/Morb. 4 treatises come from the
edition of Potter 2012, including the Greek text, which does not differ meaningfully from Joly's 2003
edition in the passages quoted.
7
See Lonie 1981b: 183–6 for a discussion of this passage and earlier interpretations.
8
Holmes 2017: 365. Because the umbilical cord never stops providing pneuma to the embryo, I am
hesitant to assert that the vegetal analogies make the ramification of the embryo completely self-
directing, but I would agree that the extended tree discourse facilitates certain types of physical claims
not as easily accessed through the filings-filled bladder.

158 colin webster
In general, then, we might say that tools and plants have been under-
stood to collect two apparently opposing types of physical claims – and two
modes in which the body is thought to work. On the one side, the tool
comparisons seem to construe the body’s interior fluids as flowing by a type
of automatic, lifeless mechanism, with some moistures moving downwards
by gravity, some liquids expanding through heat, and some precipitates
collecting through sifting or centrifugal force. On the other side, plant
comparisons have been seen to reframe the body and its parts as ‘animist’ or
‘vitalist’, whereby parts can actively attract what they need by some innate,
self-directing, and explanatorily basic power possessed by living matter.
This chapter does not seek to evaluate either Lonie’s or Holmes’s claims
directly9 but instead asks whether mechanism versus vitalism as a heuristic
lens occludes a deep congruity running between both types of comparisons
that can help make sense of some apparent tensions. That is, almost all of
these analogies – especially those in Diseases 4 – come from implements,
procedures, and techniques that a physician would have either administered
to the body, or encountered in the more general practice of medicine.
Cupping glasses, bronze bowls, amphora, wine-filled kylixes, and boiling
mixtures are all part of the physician’s toolkit or part of his medical practices.
And yet so too are plants in the form of drugs, food, and occasionally physical
tools. Plants and plant matter form a huge part of the basic medical technolo-
gies that the Hippocratics employ in their therapies, whether as simple dietary
prescriptions or more complex medicaments. Moreover, plants are often
transformed into pharmaka in the same vessels and with the same processes
used in the so-called mechanical analogies. Although this does not entirely
collapse the distinction between a growing seedling and a bronze vessel, it does
illustrate that both belong to the same practical sphere. Therefore, their
collective use to imagine the interior of the body potentially represents
a more consistent conceptual tendency than previously acknowledged.
With such a frame, mixing plant and tool comparisons neither represents
a failure of mechanistic rigour nor a subordination of one type of physical
claim to the other. Rather, it illustrates how, like the Megarian wild olive, the
Hippocratic body absorbs the medical technologies that it touches, pulling
tools and objects into its interior composition, whether they are made of
metal, wood, resin, fruit, pulp, or leaves. And like the wild olive, the tools in
the body’s interior can remain unintegrated and seemingly incompatible if
9
It should be noted that Holmes’s arguments concern Generation and On the Nature of the Child,
which both focus on the formation and growth of an embryo and use tree and plant comparisons to
a far greater extent than Diseases 4, which primarily concerns pathology and forms the core of my
focus.

the tree is cut open and its contents compared. The body articulated in these
Hippocratic treatises contains a discontinuous collection of inert and biotic
technologies, where the essential feature is not the compatibility of these
swallowed tools, but the dynamic system flowing around them.10 Indeed,
Diseases 4 in particular remains largely dedicated to explicating the humoral
production of pathologies, rather than imagining a body as an active, inte-
grated whole.11 Thus, we might ask whether in the fifth century bce, prior to
the formalization of pneumatics and mechanics, we can more productively
characterize the relationship of medicalized bodies and tools as iatrotechnics,
where the material technologies that provided the most powerful templates for
understanding inner corporeal processes come from within medical thera-
peutics themselves.12 Bodies have not yet become organized into machines,
but this does not mean that tools and technologies do not participate in the
conceptual construction of the body within these Hippocratic texts.
Diseases 4 and Its Pathogenic Framework

The group of Hippocratic treatises that includes Generation, On the Nature of
the Child, and Diseases 4 is thought to be either a single work, or, more likely,
the products of a single author. The content of Generation flows naturally into
On the Nature of the Child – both of which discuss the formation and
development of human life – and all three treatises propose a version of the
four-humour theory of the body not seen elsewhere the Hippocratic corpus.
They also contain multiple cross references to the gynaecological treatises,
which shows the broader context in which they have been composed.13
Diseases 4 uses this four-humour theory to establish a basic aetiology of
disease, as it roots illness in the imbalance of phlegm [phlegma], blood
[haima], bile [cholê], and water [hudôr, hudrôps]. The author associates
each of these moistures with its own ‘spring’ [pêgê], which serves as the
primary site of this humour in the body. The head collects phlegm, the
10
See Holmes 2014 for the Hippocratic emphasis on dynamic fluid exchange, rather than functional
parts.
11
The same might be said of all Hippocratic nosological texts, more or less by definition.
12
Von Staden 2007 has emphasized the deeply reciprocal relationship between physis and techne within
the medical tradition. He has also illustrated numerous ways in which mechanics and medicine
interacted during the Hellenistic era, most notably the interrelation of Ctesibius’ force pump and
the heart (1995, esp. 199–203; 1996) and the shared vocabulary of Andreas of Carystus and Philon of
Byzantium (1998). See also Berryman 2009; compare n. 1 above.
13
Genit. 4, 7.476 L; Nat. Puer 4, 7.496 L; and Morb. 4.26, 7.612 L all reference Diseases of Women,
which the author identifies as his own treatise. In all three cases, the passages he refers to appear in
Mul. 1, while Mul. 1.1, 8.10 L and Mul. 1.44, 8.102 L refer back to the Genit./Nat. Puer/Morb. 4
cluster. Doctrinal overlaps abound (see Lonie 1981b: 51–4).

160 colin webster
heart blood, the gallbladder bile, and the spleen water. Despite what the
connotations of ‘spring’ might indicate, these locations do not function as
the source of these humours but serve rather as something more akin to
reservoirs. When we eat food and drink that contain these humours, they
enter what the author calls the ‘cavity’ [koiliê], an interior hollow similar to,
but not identical with, the digestive track from stomach to rectum.14 From
the cavity, each of the four individuated springs draws off its own respect-
ive moisture through interconnected vessels or ‘veins’ [phlebes] according
to the principle ‘like moisture to like’ (that is phlegm goes to phlegm, blood
to blood, etc.). These springs draw their own individual humours from the
cavity and pass them along to the rest of the body, but the springs can also
draw moisture back from the body, if they are empty (Morb. 4.2, 7.544 L; cf.
Morb. 4.8, 7.556 L). The springs thus serve as a type of buffer to mitigate
excessive or insufficient humours in the body’s flesh.
Although moistures can and do flow from any one of the three locations
into any other – if emptiness or excess propel such a transfer – in a non-
pathological cycle, the consumption of moisture-containing foods introduces
new humours from the cavity to the springs, which pushes older moistures
from the springs into the flesh, and from the flesh back into the cavity, where
they are expelled through the nose, mouth, or, more generally through the
urethra and anus in the form of urine and stool. Health is maintained as long
as this process proceeds according to a three-day cycle for fluids and a two-day
cycle for solids (see esp. Morb. 4.11, 7.562–4 L). Illness results from the springs
attracting too much of their humour, humours lingering too long in the body
or getting caught in the veins, or one of the humours dominating the others
through excessive consumption (see Figure 6.1).15
Springs as Plants/Pharmaka
This Hippocratic author, like multiple others, envisions the body as a site
of fluid progression and exchange, whereby moistures move through an
interwoven network of vessels in a determined sequence. He spends much
14
Compare Nat. Puer 6, 7.498 L which provides a description of the formation of the cavity and the
intestines. Since they are both mentioned, it is not clear whether the intestines and cavity represent
two distinct parts or whether, as seems likely to my mind, the latter is included in the former. See
Holmes 2010 for the conceptual role of the cavity in Hippocratic thought, although she treats it
more broadly as the opaque interior of the entire body, not as restricted to a hollow therein.
15
For the latter, see Morb. 4.14, 7.570 L. The possible dynamics of this system are more complex than
outlined above, since moistures can move from any of the three stations into any other (as indicated
by the grey arrows). Both pathological and non-pathological circumstances produce such
movements.

Food/Drink
Cavity
Spring Spring
Body Body
Cavity
Figure 6.1 Flow chart of humours in Morb. 4, with normal flow lines marked in red,
pathological movements marked in grey
of his time articulating the dynamics of this system and expends consider-
able energy on articulating how each spring attracts its own unique
moisture. To explicate this process, he points to plants, which can (accord-
ing to his argument) draw their own moistures into themselves from the
ground:
Ἔχει γὰρ ὧδε ἡ γῆ ἐν ἑωυτῇ δυνάμιας παντοίας καὶ ἀναρίθμους. ὁκόσα
γὰρ ἐν αὐτῇ φύεται, πᾶσιν ἰκμάδα παρέχει ὁμοίην ἑκάστῳ, οἷην καὶ
αὐτὸ τὸ φυόμενον ἑωυτῷ ὁμοίην κατὰ συγγενὲς ἔχει, καὶ ἕλκει ἕκαστον
ἀπὸ τῆς γῆς τροφήν, οἷόν περ καὶ αὐτό ἐστι· τό τε γὰρ ῥόδον ἕλκει ἀπὸ
τῆς γῆς ἥτις ἐν τῇ γῇ ἰκμὰς τοιαύτη ἔνεστί, καὶ τὸ σκόροδον ἕλκει ἀπὸ
τῆς γῆς ἰκμάδα τοιαύτην, οἷόν περ καὶ αὐτὸ δυνάμει ἐστί, καὶ τἆλλα
πάντα τὰ φυόμενα ἕλκει ἐκ τῆς γῆς καθ᾽ ἑωυτὸ ἕκαστον. (Morb. 4.3,
7.544–6 L)
For in this way the earth contains countless manifold potencies. And for
everything that grows in it, the earth provides a moisture similar to what
each plant already contains in itself, similar according to its kind. And
each plant draws nourishment like itself from the earth. For the rose
draws the moisture from the earth that is of one sort, and garlic draws
moisture from the earth of a different sort, such as it already is itself in
potency. And all other plants likewise draw each thing from the earth
according to its kind.
As further proof, he points to the notoriously un-transplantable silphium
plant, which grows wild in Libya but resists all human attempts at cultiva-
tion in Ionia and the Peloponnese. Since these latter locations have both
sun and moderate weather, the author insists that the soil in Greece must

162 colin webster
not possess the specific moistures that accord with silphium’s nature. His
argument is that each of the four springs must attract its own particular
moisture from the cavity, just as a plant attracts its own proper moisture
[κατὰ συγγενὲς] according to the principle, ‘like moisture attracts like’
[ἕλκει ἡ ὁμοίη ἰκμὰς τὴν ὁμοίην] (Morb. 4.2, 7.544 L).
If using mechanism versus vitalism as a heuristic lens, the plant-model
represents a step away from a mechanical explanation and towards some-
thing which Holmes has called ‘proto-sympathetic attraction’. That is, the
plant comparison attributes each spring’s attractive capacity to some
intrinsic (and to our minds almost mystical) living power rather than
some physical composition and mechanical arrangement.16 Letting this
debate set the terms, however, downplays another crucial feature of this
comparison: rather than seeing a strict polarity between plants and tools,
we should recognize that plants themselves form a powerful medical tool –
perhaps the medical tool par excellence – in the form of food and drugs.
The Hippocratic author in fact highlights this aspect of the comparison
in several ways. First, his analogy describes how plants draw particular
potencies [dunamias] from the ground. It is precisely this concept that the
author of Regimen 1 uses to understand the effects that food and drink have
on the body more broadly. In other words, the doctor harnesses, mitigates,
and manipulates the potencies of plants as one of the core practices of his
technê.17 Moreover, Diseases 4 does not point to generic plant potencies but
identifies the phlegm-like and blood-like moistures that enter the body in
the form of food and drink:
ἕλκει δὲ ἕκαστον τῶν φυομένων βρωτῶν τε καὶ ποτῶν ἐς ἑωυτὸ πολλὰς
δυνάμιας ἀπὸ τῆς γῆς· ἐν παντὶ δέ ἐστί τι φλεγματώδεος καὶ αἱματώδεος·
ἀνάγκην οὖν τῷδε προσηγαγόμην, ὅτι ἀπὸ τῶν βρωμάτων καὶ τῶν
ποτῶν ἐς τὴν κοιλίην χωρεόντων ἕλκει τὸ σῶμα κατὰ τὰς πηγὰς ἃς
ὠνόμασα, ἡ ὁμοίη ἰκμὰς τὴν ὁμοίην διὰ φλεβῶν. (Morb. 4.3, 7.548 L)
But each of the plants consumed in food and drink draws many potencies to
itself from the earth; but in every one there is something phlegm-like and
blood-like. And so, I have added a compelling example to this, because from
foods and drinks entering into the cavity, the body draws [moistures] to the
16
Holmes 2014. Her analysis, while focusing on Places in the Human, rightly, I think, demonstrates
that Hippocratic physiologies privilege the attractive capacities inherent in the dynamic systems (that
is, in this case, the humours) and not specifically assigned to organs or organ tissue. Even as the
author of Diseases 4 describes the unique attractive capacity of each reservoir, these reservoirs still
function only within broader fluid exchanges driven by the action of the humours.
17
Dioscorides and Galen will also both turn δύναμις into the key conceptual rubric for understanding
the medical effects that drugs have on the body.

springs that I named, with like-moisture drawing like-moisture through the
vessels.
He then goes on to describe some food and drinks as mutual pharmaka
for each other insofar as they possess counteracting strengths. This once
again emphasizes their medical use:
ἕτερα γὰρ τῶν ἑτέρων τὰ ἐσθιόμενα καὶ πινόμενα φάρμακά ἐστιν· οὕτω δὲ
καὶ τὰ ἄλλα ὁκόσα σινεόμενα φάρμακα ἐστιν, † ὑφ᾽ ἑτέρου ἕτερον ἐσπεσὸν
ἐς τὴν κοιλίην, τῇ ἑωυτοῦ δυνάμει αἰτίην ἔχον κρατηθὲν ἔξω, καὶ σινέεται
τούτον. (Morb. 4.5, 7.552 L)
For some things eaten and drunk are drugs for other things; and so too with
the rest that are harmful drugs, one entering into the cavity because of
another, possessing a cause by its own potency, having been overpowered
outside, it harms this one [inside the body?] too.
This part of the text is regrettably garbled, so the meaning of this passage is
obscure at best. Nevertheless, the emphasis on pharmaka both entering the
body and being ruled outside suggests a conceptual continuum between
the external plant substances and the internal fluid systems of the body.
What is more, the three plants that he mentions, rose, garlic, and
silphium, are all used as drugs, especially in the gynaecological treatises
to which this cluster of texts is closely related. The Hippocratic corpus
mentions roses eleven times, always as part of gynaecological remedies.
Roses are used in potions consumed to clean out the lochia after birth,18 as
part of a vaginal suppository to treat post-birth uterine ulceration or ‘red
flux’,19 and as part of uterine fumigations,20 as well as an ingredient in
fomentations,21 sitz baths,22 and topical remedies.23 This is not to mention
the nearly four dozen times ‘rose unguent’ [rhodinon muron] and approxi-
mately two dozen times ‘rose oil’ [rhodinon elaion] are used in similar
gynaecological remedies. Hippocratic authors mention garlic in remedies
and regimens approximately seventy times, while silphium, which some-
times appears in the same passages as garlic,24 appears forty-six times.
Hippocratic authors prescribe silphium for multiple medical purposes
(and it can even be used as a simple condiment), although it is notably
used to induce abortion or expel a stillbirth, which once again suggests
a pharmacological connection to the gynaecological practices discussed by
18 19
Nat. Mul. 109.23, 7.428 L. Nat. Mul. 84, 7.406 L; Mul. 2.87.2, 8.380 L.
20
Nat. Mul. 34.11, 7.372 L; Mul. 1.51, 8.110 L; Mul. 2.86.7, 8.378 L; Mul. 2.97.2, 8.398 L.
21 22
Mul. 2.57, 8.344 L. Mul. 2.58, 8.346 L; Mul. 2.101, 8.406 L. 23 Mul. 1.49, 8.108 L.
24
For silphium and garlic together, see Mul. 2.24, 8.298 L.

164 colin webster
this author.25 Plants thus supply a conceptual model for understanding the
inner workings of the body, but only insofar as they also supply the literal
substances that get applied to the body, inserted into orifices, or introduced
into physical systems through consumption. In short, when conceptualiz-
ing the attractive capacity of the springs inside the body, the author does
not invoke plants in some generic way, but in their specific role as medical
tools.
In the process of becoming absorbed into the conceptual fabric of the
body and making human parts more vegetal, plants likewise become more
human – or at least the boundary between vegetal and human patients
dissolves. Not only do plants possess ‘blood-like’ and ‘phlegm-like’
humours, but they also get sick for the same basic reasons as people. The
Hippocratic author explains that the excess and defect of appropriate
moistures in the earth will lead to illness in the plant:
ὅτῳ δὲ τῶν φυομένων ἐν τῇ γῇ ἰκμὰς κατὰ συγγένειαν τοῦ δέοντος πολλῷ
πλέων ἐστί, νοσέει ἐκεῖνο τὸ φυτόν· ὅτῳ δὲ ἐλάσσων τοῦ καιροῦ, ἐκεῖνο
αὐαίνεται. (Morb. 4.3, 7.546 L)
If there is much more moisture than is necessary for any plant according to
its kind, this plant falls ill; and if there is less moisture than is appropriate for
any plant, this plant dries up.
The author explains the attractive powers of the humoral reservoirs in the
body by comparing them to plants, which themselves model the patho-
logical behaviour of the human body in possession of humoral reservoirs.
The physical and conceptual interfaces of the body merge, as they mutually
articulate each other. Neither the author’s understanding of plants nor the
human body can stand alone. This mutually determining interface resem-
bles a Hippocratic strategy in Airs Waters Places, which Fréderic Le Bley has
called ‘dual direction analogies’. In this latter treatise, the author explains
why rainwater has certain effects on the body, but only by explaining
rainwater’s nature with an analogy to the sun’s effects on the sweating
human.26 In other words, the Hippocratic author of Airs Waters Places
explains the effects of meteorology on the body by weaving the body back
into the same meteorological system both conceptually and physically.
Something similar happens in Diseases 4, except instead of a continuum
between the body and its environment, this author first establishes both
25
For a discussion of silphium, its medicinal use, and scholarship on the issues surrounding it, see
Totelin 2009: 158–60.
26
Le Bley 2005; compare Aër. 8.1–25, 2.32–6 L.

a physical and an epistemological continuum between a body and the
plants that enter it.27 We might ask, then, whether such a methodology can
truly be called ‘analogy’, since the target (tenor) and analogue (vehicle)
both operate within the same dynamic humoral system, which is mutually
determining.28 The author is not modelling the interior springs on plants,
because the plants themselves are modelled on humour-having patients.
We might more productively say that the body absorbs medicalized plants,
both literally and conceptually.
Cupping-Vessel Model
If there were any doubt that the Hippocratic author does not separate
mechanical from vitalistic explanations, immediately upon providing the
plant model of attraction for the springs’ innate capacity, the author
supplies another proof [sêmeion] that attributes the head’s ability to draw
up phlegm to its basic shape, which he claims takes the form of a cupping
vessel. Cupping vessels are among the most ubiquitous icons for the
practising physicians in Classical antiquity. They are mentioned in mul-
tiple places throughout the Hippocratic corpus, and nine pre-Roman
examples are extant, as well as a pictorial representation on the Peytel
aryballos of the so-called Clinician Painter, located at the Louvre
(Figure 6.2). Most often, a physician would place a small piece of burning
lint inside the vessel and then stick it against the skin. The lint would burn
out and create suction in the process, fastening the cup to the body.29 The
physician might also heat up the cup itself and then let it cool on the skin to
the same effect. Sometimes a small incision would be made first, over
which the cup would be placed, and this would draw out blood (see Morb.
2.55, 7.86 L, Internal Affections 51, 7.296 L). Affections 4, 6.212 L recom-
mends attaching two cups to the head to cure a headache. Aphorisms 5.50,
27
The author of Diseases 4 avoids complete assimilation between plant and human bodies, insofar as
plants only provide ‘blood-like’ [αἱματώδης] and ‘phlegm-like’ [φλεγματώδης] moisture, rather
than simply containing blood and phlegm. In further explicating the humoral system, however, he
makes no real distinction between these moistures as they enter the body and circulate within it.
28
A tremendous amount has been written about analogies and metaphors in ancient medicine and
science – so much so that productively synthesizing here is not possible. The locus classicus for such
scholarship remains Lloyd 1966, which explored the importance of analogical reasoning in early
Greek science (cf. Lloyd 2015, which revisited the subject with a focus on ‘semantic stretch’). More
recently, Roby 2017 has provided a succinct overview of the major interpretations of metaphor as
relevant for ancient science, ranging from models based on ‘similarity’ and ‘substitution’ to ‘class-
inclusion assertions’.
29
VM 22, 1.628–630 L and Medic. 7, 9.212 L both describe different shaped cups, and these were used
for distinct therapeutic tasks.

166 colin webster
Figure 6.2 Peytel aryballos, 480–470 bce (Louvre CA 2183), photo via Wikimedia
Commons (https://tinyurl.com/yc7vfbj9)
4.550 L and Epidemics 2.6.16, 5.136 L mention applying large cupping

vessels to the breasts to control menstruation, while Diseases of Women
2.110, 8.236 L also describes applying cupping glasses to the breasts.30
These cupping vessels both extract fluids from the body and move
moisture around its interior by applying topical suction. They operate
physically at a point of contact between the physician’s art and the patient’s
humoral systems, and in so doing, act as a type of embodied guarantor for
humoral theories by making visible the type of dynamic exchange the
Hippocratic physician describes going on unseen inside the body. Just as
the plant/pharmaka simultaneously articulate, embody and ‘prove’ the
pathological power of moistures, the cupping vessel ‘proves’ that humoral
flows produce and remove disease. This physical and conceptual interface
makes the cupping vessel a useful conceptual tool, which the author of
30
Many other instances of cupping vessels appear in the Hippocratic corpus; compare Bliquez 2014.

Diseases 4 uses to explicate the attractive power of the head as a reservoir for
phlegm:
φημὶ δὲ ὁκόσον ἐν τῷ βρώματι ἢ πόματι φλεγματῶδες ἔνι, κείνου ἐς τὴν
κοιλίην ἐλθόντος, τὸ μὲν τὸ σῶμα ἕλκει ἐς ἑωυτό, τὸ δὲ ἡ κεφαλὴ κοίλη
ἐοῦσα καὶ ὥσπερ σικύη ἐπικειμένη ἕλκει· τὸ <δὲ> φλέγμα, ἅτε γλίσχρον ἐόν,
ἕπεται [δὲ] τοῦτο τὸ ἕτερον διὰ τοῦ ἑτέρου ἐς τὴν κεφαλήν· τὸ μὲν νέον
φλέγμα τὸ ἐγγεννώμενον ἀπὸ τοῦ βρώματος μένει ἐν τῇ κεφαλῇ, τὸ δὲ
παλαιόν, ὁκόσῳ πλεῖον τὸ νέον, ὑπ᾽ ἐκείνου βιώμενον ἐξέρχεται, καὶ διὰ
τοῦτο ἐπήν τις πίῃ ἢ φάγῃ ὅ τι φλεγματῶδες, ἀποχρέμπτεται ὁ ἄνθρωπος
φλέγμα. (Morb. 4.4, 7.548 L)
I say that when whatever is phlegm-like in food or drink enters into the
cavity, the body attracts some to itself, and the head attracts some, insofar as
the head is hollow and set on the body, just like a cupping vessel. And the
phlegm, insofar as it is sticky, follows one part after another into the head.
And as new phlegm, arising from the food, occupies the head, an equal
amount of old phlegm is pushed out by it, and because of this, whenever
someone drinks or consumes anything phlegm-like, he will cough out
phlegm.
This is not the only Hippocratic author to employ the attractive capacity
of the cupping vessel to articulate the behaviours of various body parts.31
Ancient Medicine likewise imagines the head functioning like cupping
vessel, as well as the bladder and womb, insofar as their hollowness
produces attractive power:
τοῦτο δέ, αἱ σικύαι προσβαλλόμεναι ἐξ εὐρέος ἐς στενότερον συνηγμέναι
πρὸς τοῦτο τετεχνέαται, πρὸς τὸ ἕλκειν ἐκ τῆς σαρκὸς καὶ ἐπισπᾶσθαι,
ἄλλα τε πολλὰ τοιουτότροπα. τῶν δὲ ἔσω τοῦ ἀνθρώπου φύσις καὶ σχῆμα
τοιοῦτον. κύστις τε καὶ κεφαλή, καὶ ὑστέρη γυναιξίν. (VM 22 Schiefsky,
1.626–628 L)32
And with respect to this, cupping vessels, proceeding from a wide area and
tapering into a narrower, are constructed for this purpose: to draw and
attract [moisture] from the flesh, and there are many other things of such
a kind. Examples inside a human that have such a nature and shape are the
bladder and the head, as well as the womb for women.
31
Similarly, Morb. 4.24, 7.600 L states that the vessels leading to the bladder can attract phlegm-like
and earth-like components of dirty milk that enters the infant’s cavity, and he seemingly ascribes this
attractive capacity to these vessels’ wideness.
32
Compare Arist. Somn. Vig. 3, 457a21–5, where Aristotle suggests that excessively large heads, such as
those associated with dwarfism, draw upwards-tending vapours from digestion in even greater
amounts than normal. This claim seems to participate in this same logic.

168 colin webster
Not only does the author of Ancient Medicine make the basic physical
shape of the cupping vessel responsible for its attractive power, he seem-
ingly makes no fundamental distinction between the physician’s tool and
the other ‘things of such a kind’ that reside in the body. The cupping vessel,
bladder, head, and womb all manifest the same physical capacity based on
their ‘shape’ [schêma].
On the one hand, the appearance of the cupping vessel in Diseases 4
might seem to advance a mechanical account of the head’s capacity to
attract moisture upwards, insofar as its basic hollowness and physical
structure create the conditions for the upward suction, not any intrinsic
(and explanatorily basic) qualitative capacity. On the other hand, the head
does not primarily attract all moistures, only phlegm. This may be because
phlegm’s unique stickiness, viscosity, and thickness allow it to be drawn
upwards, while the other, thinner moistures do not have enough cohesion
to be pulled against gravity in this way.33 Nevertheless, the fluid dynamics
described in this passage cannot describe any essentially ‘mechanical’
action, since a cupping vessel only draws flesh and fluid into it until it
reaches equilibrium. This author, by contrast, describes the head drawing
and expelling equal amounts of phlegm at the same time, as well as
attracting too much phlegm, which is not possible within the fixed capacity
of the material tool.34 The cupping vessel thus enacts principles previously
guaranteed by the plant/pharmaka analogy, and the two comparisons work
in tandem, not in opposition, to explain how the head attracts phlegm
according to the principle of ‘like with like’. Both pharmaka and cupping
vessels supply a ready model for the body’s interior functions, but largely
because they are already components in the dynamic exchange between the
body and the medical intervention.
Three Bronze Vessels

Over the next several chapters, the Hippocratic author describes each of the
three remaining springs and explicates how they attract their respective
humours. His explanations mix arguments relying on the explanatorily
33
On the Nature of the Child supports this reading insofar as it argues that bones get filled with marrow
in utero because they are hollowed out by pneuma, and ‘since they are hollow, they draw into
themselves the fattiest component of the mass of blood from the flesh’ [κοῖλα δὲ ἐόντα ἕλκει ἐς ἑωυτὰ
ἀπὸ τῶν σαρκῶν τοῦ αἱμάλωπος τὸ πιότατον] (Nat. Puer. 8, 7.506 L).
34
Moreover, the author mentions ‘vessels’ [φλέβες] in the head in which the phlegm can get stuck
(Morb. 4.4, 7.550 L). This might indicate that the head’s functional ‘hollow’ shape operates alongside
different internal articulations, depending on the proximate needs of the author’s account.

basic capacity to attract and those invoking structure and shape. The
gallbladder ‘draws’ [ἕλκει] bilious moisture to itself from ingested foods,
but the author further suggests that bile alone ‘is separated from food and
drinks into the gallbladder’ [ἐς μὲν τὸ χωρίον τὸ ἐπὶ τῷ ἥπατι ἀπὸ τῶν
βρωτῶν καὶ τῶν ποτῶν ἀποκρίνεται ἡ χολὴ μοῦνον] because the thinness
of the gallbladder’s vessels prevents it from attracting the other, thicker and
heavier moistures (Morb. 4.9, 7.560 L). The spleen, which he also considers
to be hollow, attracts water, but much of the account seems to be based on
simple gravity flow, except that the cavity and the bladder draw water back
up from the legs insofar as there are many vessels that stretch downward
and draw water from the lower parts when they become dry (Morb. 4.6,
7.554 L). Lastly, the heart draws off some blood-like moisture that enters
the cavity (as does the body itself), but its attractive capacity does not
require any unique explanation aside from stating that the heart simply
attracts what is similar to itself. And yet, it can also draw too much blood,
although this is quickly pushed into the jugular veins, which operate as
a type of overflow. How the heart manages to do this, or avoid attracting
too much blood in normal circumstances, is left underdetermined. The
account thus blends the homogeneous attractive capacity and other types
of mechanisms without any apparent anxiety. These various explanations
are, for this author, complementary and do not represent different types of
physical behaviours.
Having supplied a separate account for each of the four springs, the
Hippocratic authors generalizes his discoveries, articulating that the springs
supply the body when they are each full but draw from it when they are
empty, just as the cavity does too. He illuminates the relationship with
another technological analogy involving three interconnected bronze dishes:
ἔχει γὰρ οὕτως ὥσπερ εἴ τις ἐς χαλκεῖα τρία ἢ πλείονα ὕδωρ ἐγχέας καὶ
συνθεὶς ὡς ἐπὶ ὁμαλωτάτου χωρίου καὶ συναρμόσας ὡς κάλλιστα
ἐπιδιαθείη, αὐλοὺς ἐναρμόσας ἐς τὰ τρυπήματα, καὶ ἐγχέοι ἡσυχῇ ἐς ἓν
τῶν χαλκείων ὕδωρ μέχρις οὗ ἐμπλησθῇ ἀπὸ τοῦ ὕδατος πάντα· ἀπὸ γὰρ
τοῦ ἑνὸς ῥεύσεται ἐς τὰ ἕτερα χαλκεῖα μέχρις ὅτου καὶ τὰ ἄλλα πλησθῇ·
ἐπὴν δὲ πλήρεα γένηται τὰ χαλκεῖα, ἤν τις ἀπὸ τοῦ ἑνὸς ἀπαρύσῃ τοῦ
ὕδατος, ἀνταποδώσειεν ὀπίσω ῥέον τὸ ὕδωρ ἐς τὸ ἓν χαλκεῖον, καὶ κενεὰ
ἔσται τὰ χαλκεῖα πάλιν ὥσπερ καὶ ἐδέξατο. οὕτω δὲ καὶ ἐν τῷ σώματι ἔχει·
ἐπὴν γὰρ ἐς τὴν κοιλίην πέσῃ τὰ βρώματα καὶ τὰ ποτὰ, ἐπαυρίσκεται τὸ
σῶμα ἀπὸ τῆς κοιλίης καὶ πληροῦται σὺν τῇσι πηγῇσιν· ἐπὴν δὲ ἡ κοιλίη
κενῶται, ἀποδίδοται αὖτις ὀπίσω ἡ ἰκμάς, ὥσπερ καὶ ἐπεδέξατο [ὡς] ἐκ
τῶν ἄλλων τὸ ἓν χαλκεῖον. (Morb. 4.8, 7.556–558 L)

170 colin webster
For it is just as if someone should pour water into three or more bronze
vessels, having positioned and fit them together on level ground, and should
arrange them as nicely as possible, and then, having fitted pipes into
openings in them, should gently pour water into one of the bronze vessels
until they were all full from the water. For water will flow from one into the
other bronze vessels until the rest are also filled. And whenever the bronze
vessels are full, if someone draws off water from a single one, the water
would flow backwards and give back to that one vessel. And the bronze
vessels will be emptied again in the same way as they were also filled. This is
also the case in the body. For whenever foods and drinks enter the cavity, the
body draws off [moistures] from the cavity and is filled along with the
springs. But when the cavity is emptied, moisture is given back again, just as
one bronze vessel received [water] from the others.
This example has often been taken as representing the relationship between the
four springs,35 but the three bronze vessels are supposed to represent the cavity,
spring, and body. First, the author explicitly introduces the bronze vessels as
way to conceptualize how these three elements mutually draw and disperse
fluids to each other, and he mentions these three components in explaining his
comparison. Second, the four springs do not directly communicate, nor do
they all have the same moistures running between them, whereas the cavity–
body–spring continuums do.36 The latter also has the advantage of making
better sense of the number of vessels that the Hippocratic author mentions,
which has otherwise vexed some commentators.
The author’s vivid description of ‘fitting pipes into openings’ and the
instructions to ‘pour gently’ and ‘arrange as evenly as possible’ invites the
reader to imagine that the author actually completed such a task. In fact,
this has encouraged some scholars to interpret this comparison as a quasi-
experiment. Although this is certainly possible, it is hard to imagine
drilling holes in a valuable bronze vessel for such a purpose, rather than
using something cheaper and easier to puncture, such as terracotta pots,
wooden buckets, or simple bowls. Yet even if the author did choose to
conduct this demonstration, why, we might ask, does he insist on bronze
vessels to recreate the corporeal fluid dynamics outside of the body? Why
do these particular implements form the basis of his vivid image? The
Peytel aryballos (Figure 6.2) mentioned above provides a possible answer,
since it depicts a bronze vessel as a prominent, perhaps even iconic medical
implement alongside the cupping vessel previously discussed.
35
Lonie 1981a: 130; Lloyd 1978: 45–9.
36
Morb. 4.9, 7.560 L describes how the gallbladder contains only bile, while the head, spleen, and heart
contain all four humours, although they each contain a majority of their own natural moisture.

In the image the bronze vessel rests at the physician’s feet, positioned
underneath the patient, who appears to be leaning forward slightly so as to
arrange himself directly over the top of it. The physician makes an incision
into the patient’s forearm, presumably to draw blood out, which the
bronze vessel will catch.37 A similar item from the late fifth or early fourth
century bce, labelled a ‘footbath with stand’ (Metropolitan Museum of
Art 38.11.5a, b, see Figure 6.3) gives a sense of the ornament and potential
worth of such an object, which makes destroying three of them hard to
fathom. Rather, I would suggest, bronze vessels serve as a model to
understand the internal reservoirs that catch excessive fluid, since they
can already function as such reservoirs as part of the medical intervention.
Like both the pharmaka and cupping vessel, they are physically integrated
into the very systems that the Hippocratic author conceptualizes by using
them as an imaginative tool. In other words, the bronze vessels are pulled
inside the body as part of the same process by which they help draw fluids
Figure 6.3 Footbath with stand, late fifth or early fourth century bce (Metropolitan
Museum of Art, New York, 38.11.5a, b)
37
For an alternative interpretation, see Krug 2012.

172 colin webster
out. Tools that interface with the body become amalgamated into its inner
landscape.
There are several other examples of medical tools and processes serving
as conceptual models for inner corporeal behaviours. For example, to
explain how bile can nourish fever, the author mentions boiling water
with unguent oil in a bronze vessel for a long time, so that the water boils
off, but only some of the unguent oil evaporates (Morb. 4.18, 7.580 L). This
comparison mirrors descriptions of how to make certain medicaments in
bronze vessels by subjecting them to low, sustained heat.38 The author also
describes how fluids can collect after a blow and cause swelling insofar as
they block their own way out, just as a leather oil flask [λήκυθος σκυτίνη]
can become blocked by the unguent inside – which may point to the
hypothetical medicament he has prepared in his earlier analogy (Morb.
4.20, 7.588 L).39 The author also explains why a hot environment can stir
up the body and cause diseases by using the analogy of Scythians agitating
mare’s milk to create butter and horse cheese [ἱππάκη] (Morb. 4.20,
7.584 L).40 He then explains how cold environments can cause disease by
describing how cold fig juice acts as rennet to coagulate milk into cheese
(Morb. 4.21, 7.590 L).41 Shortly thereafter he suggests that bladder and
gallstones are caused in infants when nursemaids eat phlegmatic foods –
which include cheese (Morb. 4.4, 7.548 L). He also states that stones
coallesce in the bladder like precipitate in a swirling kylix or bronze vessel
38
Ulc. 12, 6.412 L describes making a topical salve of grape juice, wine, honey, myrrh, and flower of
copper in a bronze vessel, while Haem. 3, 6.438 L also uses a bronze vessel to make a salve. Both
involve leaving the contents of the vessel out in the sun. Mul. 1.98, 8.224 L uses a bronze/copper
vessel to make a plaster for gout, while Mul. 102, 8.224 L and 104, 8.226 L both describe an eye
treatment that involves drying the contents of such a vessel in the sun. Mul. 105. 3, 8.228 L slowly
boils an eye medication in a bronze/copper vessel.
39
Lekythoi may also have funerary connotations, since they often contained oil used to anoint the dead
and many have been found in tombs (although not the leather variety). They could also hold oil used
to anoint brides before marriage, which would fit with the general connection this group of texts
bears with the gynaecological treatises of the Hippocratic corpus (cf. Kanowski 1984).
40
This mention of Scythian cheese might reflect borrowing from Hecataeus (see Lonie 1981b: 70), but
since it also appears at Aër. 18, 2.68–70 L, it may be considered part of Hippocratic medical
tradition, if only as a topos of exotic foodstuff.
41
Il. 5.902–5 compares the speed of fig juice coagulating milk to the speed at which Paiēon healed Ares,
which suggests that this particular comparison had long-standing medical overtones. Moreover,
Emp. DK31 B33, line 4 compares the effects of fig juice on milk to love’s capacity to bind and fix.
Similarly, Aristotle, GA 1, 729a12–14 compares the effects of semen on menstrual blood in the
formation of the embryo to this interaction of fig juice and milk; compare GA 2.737a14–16; 3,
771b19–27. Perhaps this supplies an even closer connection to embryology, which fits neatly with the
Hippocratic author’s use of it in describing diseases forming in utero. For the relationship of
Aristotle’s Generation of Animals to the Gen./Nat. Puer/Morb. 4 cluster, see Byl 1980; Oser-Grote
2004. Figs were also used in fertility medicaments, probably because fig was a slang term for vagina;
cf. Totelin 2009: 206.

(Morb. 4.24, 7.600 L), which presumably references wine sediment, espe-
cially since other Hippocratic physicians prescribe wine to alleviate bladder
stones in children (AWP 9, 2.36–8 L). He then describes how the bladder
stones harden, claiming that the phlegm and unclean sediment mix to form
a type of glue. This substance hardens into stone when the phlegm melts
out, like the slag burned out of iron ore in the process of refinement (Morb.
4.24, 7.602 L). Although iron production may not at first resemble a spe-
cifically medical technology, we should remember that physicians were
likely deeply involved in producing their specialized (often metal)
implements.42 Moreover, bladder stones are one of the early medical
conditions for which surgery was considered, and therefore a condition
for which metal knives, potentially designed and produced specifically for
this task, would have been used.43
This does not exhaust all of the analogic statements in the treatise, but it
provides an overview of how many medical tools have been pulled into the
body’s interior, especially to explain pathologies for which these tools
already provide treatment. Even the very formulation of ‘springs’ could
be seen to participate in this schema, insofar as Hippocratic physicians,
such as the author of Airs Waters Places, place great emphasis on how the
water source of a city affects its population’s health, and the direction its
springs face can play a crucial role in this. Moreover, the author’s treatment
of springs as reservoirs, rather than sources, more closely resembles
a human-made catch basin that serves as a collection point, rather than
a stream trickling from a hill (Figure 6.4). Springs themselves thus form yet
another potential image of a technological, medicalized site.
The capstone for the integration of medical technologies into the body
occurs when Diseases 4 discusses the purpose of the four reservoirs and the
humours they contain, which for the majority of the text seem primarily
employed to explain disease and characterize the patient as perpetually on
the cusp of illness. What role the moistures play in the health of the body at
42
Bliquez 2014: 14–16 emphasizes the large role physicians played in the design and production of their
own implements in antiquity. This accords not only with descriptions in the Hippocratic treatises
(for examples at Mul. 2.24, 8.280–302 L; Haem. 2, 6.436 L; Frac. 30–1, 3.518–524 L), but with later
comments by Galen (see Peri alupias 4–5 Boudon-Millot and Jouanna), who laments how a fire
destroyed not only tools that he had designed, but wax models he used to replicate these instru-
ments. See also Celsus 8.20.4; Galen, In Hipp. Artic. 18A.338–339K.
43
Bladder stones are described in multiple places in the Hippocratic corpus, and while Jusj. 22–3, 4.17
L explicitly prohibits lithotomy, this, to my mind, suggests that it was practised by at least some
physicians (otherwise, there would be little need to forswear it). This surgical intervention later became
a more common practice (see Celsus 7.26.2), and physicians such as Meges of Sidon (first century bce)
produced specialized medical tools for this surgery; see Bliquez 2014: 15, 98–102, 181–183; compare
Jackson 2010.

174 colin webster
Figure 6.4 Theagenes spring/fountainhead at Megara, ca. 600 bce (Photo by

O. Mustafin, via Wikimedia Commons, https://tinyurl.com/3zet5an3)
first seem primarily negative, aside from occasional vague claims about the
moistures supplying ‘nourishment’. Yet in section 8, the author links the
springs to the production of desire for foods. He claims that the springs
‘always interpret for the rest of the body the power of each thing eaten and
drunk, according to their own potency, before perception occurs, that is,
whatever is bile-like, phlegm-like, blood-like, and water-like’ [ἀεὶ καὶ πρὸ
τοῦ ἐσᾴσαι ἕκαστον κατὰ τὴν ἑωυτοῦ δύναμιν ἑρμηνεύει τῷ ἄλλῳ σώματι
τῶν ἐσθιομένων καὶ πινομένων ὅ τι χολῶδές ἐστι καὶ ὅ τι φλεγματῶδες καὶ
ὅ τι αἱματῶδες καὶ ὅ τι ὑδρωποειδές] (Morb. 4.8, 7.558 L).44 If a person has
excessive or defective amounts of a given moisture, the springs cause him to
desire foods that will equalize it with the other humours [ἰσώσει τῇσιν
ἄλλῃσι] (Morb. 4.8, 7.558–560 L). The springs thus create pleasure and
pain, desire and disgust, in order to maintain a type of equilibrium (ideally)
before imbalances manifest as full-blown diseases. They do this by
44
Holmes 2010: 111–12 discusses the ambiguity of the verb αἰσθάνομαι in medical contexts, where it
can mean the body perceiving without the subject noticing.

recognizing the potencies inherent in food – the same feature that facili-
tated the plant/pharmaka analogy in the beginning of the treatise.
Moreover, they hold in reserve limited amounts of these same moistures
to supply when necessary and a limited storage capacity to receive excess
moistures if needed. In this regard, we might say that the springs in the
body act like little doctors, both interpreting the foods beneficial for the
patient to consume in order to maintain health, and also dispensing these
same moistures to the body, should they be lacking one of them or another.
Of course, we do not always eat the correct foods, so sometimes medical
intervention is necessary. The physician’s technê works as an extension of
the natural activities of the body, but only insofar as the natural activities of
the body have been modelled on the practices of the physician. If the tools
of the physician have been pushed into the inner landscape of the body, the
author has finally made the conceptual leap of placing a functional surro-
gate of the doctor into the interior cavity of his patient.
Conclusion
This chapter has envisioned the body as a both a physical and a conceptual
interface, absorbing the technologies that are introduced into it in the form
of comestibles or attached prosthetically to it through treatments. Showing
the prevalence of this tendency helps clarify that what the author of Diseases
4 presents is not quite a series of vivid analogies, but a more complex
network of corporeal entanglements, where notions about the fabric of the
body are inseparable from the tools used to manipulate it and restore it to
health. In examining the question of iatromechanics in the Hippocratic
corpus, it would thus be, to my mind, a mistake to separate instances where
this interface involves plants, pharmaka, and biotic technologies, from
instances where the body meets with bronze bowls, cupping vessels, and
metal tools. In both case, external objects from the physician’s toolkit
merge with the corporeal interior of the patient, even as the interior fluids
interact with and get drawn out by the physician’s tools. In so doing, the
technical interventions employed by Hippocratic physicians do not simply
reveal the body; they articulate it. The technical interventions themselves
are only understood based on the effects that they have on the body that
they model. Like the votive-swallowing Megarian olive, the Hippocratic
body seen in Diseases 4 enfolds adherent tools. As is not the case with the
olive, the discontinuous inner landscape of the body stretches back out into
its surroundings, transforming some of the votives themselves into exten-
sions of its own inner behaviours.

176 colin webster
RE F E RE N CE S
Amigues, S. ed. and trans. 2006. Théophraste. Recherches sur les plantes, vol. 5 (Paris).
(Cambridge).
Bliquez, L. J. 2014. The Tools of Asclepius: Surgical Instruments in Greek and Roman
Times (Leiden).
Byl, S. 1980. Recherches sur les grands traités biologiques d’Aristote: sources écrites et
préjugés (Brussels).
in Ancient Greece (Princeton).
2014. ‘Proto-Sympathy in the Hippocratic Corpus’, in J. Jouanna and M. Zink,
eds., Hippocrate et les hippocratismes: médecine, religion, société: XIVe Colloque
International Hippocratique (Paris), 123–38.
2017. ‘Pure Life: The Limits of the Vegetable Analogy in the Hippocratics and
Galen’, in J. Z. Wee, ed., The Comparable Body: Analogy and Metaphor in
Ancient Mesopotamian, Egyptian and Greco-Roman Medicine (Leiden),
358–86.
Jackson, R. 2010. ‘Cutting for Stone: Roman Lithotomy Instruments in the
Museo Nazionale Romano’, Medicina nei Secoli, 22/1–3: 393–418.
Joly, R., ed. and trans. 2003. Hippocrate, De la génération, De la nature de l’enfant,
Des maladies IV, Du fœtus de huit mois. (Paris).
Jouanna, J. 2002. Hippocratis De natura hominis, rev. ed. Corpus Medicorum
Graecorum i 1, 3 (Berlin).
Kanowski, M. G. 1984. Containers of Classical Greece: A Handbook of Shapes (St
Lucia).
Krug, A. 2012. ‘Doktorspiele? – Der Aryballos Peytel’, Boreas, Münstersche Beiträge
zur Archäologie, 35: 11–23.
Le Bley, F. 2005. ‘Microcosm and Macrocosm: The dual direction of analogy in
Hippocratic thought and the meteorological tradition’, in P. J. van der Eijk,
ed., Hippocrates in Context (Leiden), 251–69.
Lloyd, G. E. R. 1966. Polarity and Analogy: Two types of argumentation in early
1978. Hippocratic Writings (New York).
2015. Analogical Investigations: Historical and Cross-cultural Perspectives on
Human Reasoning (Cambridge).
Lonie, I. M. 1969. ‘On the Botanical Excursus in de Natura Pueri 22–27’, Hermes,
97: 391–411.
1981a. ‘Hippocrates the Iatromechanist’, Medical History, 25: 113–50.
1981b. The Hippocratic Treatises “On Generation”, On the Nature of the Child,
“Diseases IV”: A Commentary, eds. G. Baader , F. Kudlien , C. Lichtenthaeler
and K. D. Fischer. (Berlin).
Oser-Grote, C. M. 2004. Aristoteles und das Corpus Hippocraticum: Die Anatomie
und Physiologie des Menschen (Stuttgart).
Potter, P. ed. and trans. 2012. Hippocrates. Generation. Nature of the Child. Diseases
4. Nature of Women and Barrenness. (Cambridge, MA).

Roby, C. 2017. ‘Animal, Vegetable, Metaphor: Plotinus’s Liver and the Roots of
Biological Identity’, in J. Z. Wee, ed., The Comparable Body: Analogy and
Metaphor in Ancient Mesopotamian, Egyptian, and Greco-Roman Medicine
(Leiden) 387–414.
Schiefsky, M.J. 2005. Hippocrates On Ancient Medicine: Translated with intro-
duction and commentary (Leiden/Boston).
Totelin, L. 2009. Hippocratic Recipes: Oral and Written Transmission of
Pharmacological Knowledge in Fifth- and Fourth-Century Greece (Leiden).
Von Staden, H. 1995. ‘Teleology and Mechanism: Aristotelian Biology and Early
Hellenistic Medicine’, in W. Kullmann and S. Föllinger, eds., Aristotelische
Biologie: Intentionem, Methoden, Ergebnisse (Stuttgart), 183–208.
1996. ‘Body and Machine: Interactions between Medicine, Mechanics, and
Philosophy in Early Alexandria’, in K. Hamma, ed., Alexandria and
Alexandrianism (Malibu), 85–106.
2007. ‘Physis and Technê in Greek Medicine’, in B. Bensaude-Vincent and
W. Newman, eds., The Artificial and the Natural: An Evolving Polarity

p a r t iii
Towards the Mechanization of the Human
Body

chapter 7
Aristotle on the Lung and the Bellows–Lungs Analogy

Giouli Korobili
Introduction
Even in the fifth century bce, when many of the treatises comprising the
Hippocratic corpus were composed, the inside of the living body remained
largely unknown and mysterious, an object of study difficult to explore but at
the same time fascinating for doctors and natural philosophers alike. Although
major organs such as the heart, brain, liver and lung seem to have been observed
in some detail, their specific functions remained largely opaque, and as such
were at times subjected to far-fetched and all but imaginary interpretations.1
Given these limitations, the search for the ‘analogous’ – positing similarities
between the ‘unseen’ processes within the living body and the external phe-
nomena of the visible world – was one of the most common types of explana-
tory model current in medical and philosophical circles of the time.2
Among the ‘invisible’ processes that take place inside the living body,
a prominent place is occupied by breathing. Before Aristotle, breathing had
been widely understood as referring simply to the passage of the outside air
into the body and its implications. Aristotle was the first to establish that
the lung was the instrument of breathing. In his treatise De Juventute et
Senectute, Vita et Morte, de Respiratione (henceforth De Juventute),3 the
I would like to thank Maria Gerolemou, George Kazantzidis, Julius Rocca, Konstantinos Stefou and
Rob Bostock for their helpful comments during the process of development of this paper. Immense
thanks go also to Robert Hahn for our correspondence regarding the current topic of investigation and
to John Stefou for the illustrations. The paper benefited greatly from the discussion in the
Montagskolloquium (Winter 2018) of the Institute of Classical Philology at Humboldt University in
Berlin. I am especially indebted to Philip van der Eijk, Sean Coughlin and Roberto Lo Presti. The
research for this paper was part of the project ‘Mapping the Vegetative Soul. Nutrition and Nutritive
Soul in Aristotle and Aristotelianism’, supported by the Berlin Cluster of Excellence ‘Topoi’.
1
Perhaps it would be more accurate to speak of processes or bodily works (ἔργα, ἐργασίαι) rather than of
functions, at least in reference to the Hippocratic, pre-Aristotelian writings. It is in Aristotle that the term
λειτουργία is first attested with reference to the mouth and its parts (PA ii 3.650a9; iii 14. 674b10, 19).
2
See Holmes 2010, 108–16.
3
The last part of Aristotle’s short treatises on natural science and philosophy (Parva Naturalia). I treat
this part of the collection as a single, unified treatise.
181

182 giouli korobili
oldest surviving systematic treatment of respiration, Aristotle castigates his
predecessors for either neglecting the subject completely or offering only
incomplete accounts. Only a few of them dealt with breathing, some
attributing it to incorrect causes, others ignoring its causes altogether and
all of them failing to correlate it with the lung.4 In the same work, Aristotle
also launches, for the first time in the extant corpus of Greek literature, into
a comparative analysis of the structural and functional similarities between
the lung and a pair of forge bellows.5 This analogy will form the subject of
the present chapter.
According to Aristotle, Empedocles attempted to describe the bodily
process of respiration as a kind of ‘pneumatic’ function, in which breathing
in and out can be explained as due to the successive mutual displacement of
blood and air in certain vessels, and went on to illustrate it by analogy with
the operation of the clepsydra, a device for transferring and perhaps
measuring small quantities of liquid.6 This analogy is dismissed as mislead-
ing by Aristotle, who instead introduces as a counterproposal the compari-
son of the lung to a pair of bellows. This he does at a crucial juncture in the
exposition of De Juventute, in which he has set himself the task of identify-
ing the causes of respiration. The text is strongly teleological in nature, but
in addition to explanations that focus upon the purpose for which respir-
ation takes place, Aristotle also offers a mechanistic account of the main
organ of respiration: the function or purpose of the lung is similar to that of
a bellows, in that it consists in keeping the (vital) fire alive and blazing.7
4
Juv. 7 (1).470b6–27. Cf. Thivel 2005, 248–9. Surveying Greek literature from Homer to the late fourth
century bce, Thivel distinguishes three main theories of respiration, which correspond to three distinct
periods: ‘In the first period, air is a material, visible substance, and, if it penetrates into the body, it causes
pain and disease; in the second period, air and blood are the source of life and circulate alternately in the
whole body through the vessels and the pores of the skin; in the third period the lungs are known as the
organs of respiration, which always breathe in and out, but air is no longer useful to breath, it is just
a cold flow whose function is to cool the heat of the heart’ (2005, 240).
5
Anaximander had already used the bellows–breathing analogy in his cosmological account of the sun
and moon and their ‘exhalation’ of fire (DK 12 A 21–2; Hahn 2010). The bellows continued to be used
by physicians such as Erasistratus and Galen as a model illustrating the functions of certain bodily
parts, most notably the heart: see, for example, De symptomatum causis libri ii 7.161.1–5 K.; De placitis
Hippocratis et Platonis p. 528, 24–8 de Lacy; von Staden 1997, 201, n. 80; Longrigg 1993, 206–7. In De
corde 8, the auricles of the heart are said to be intended by nature to function as a bellows. For
Erasistratus’ description of the heart, see Valleriani in this volume, Chapter 10. For Galen’s
mechanical explanations of the digestive process, see Lewis in this volume, Chapter 11.
6
See p. 189, n. 45, below. In all probability pneumatics had not been demarcated as a distinct field of
inquiry by that time (Berryman 2009, 78).
7
Causal explanations of natural phenomena can involve both teleological and mechanistic thinking:
see von Staden 1997 and Johnson 2017. Berryman 2009, 10, rather than speaking in terms of an
opposition between teleological and mechanistic views in ancient natural philosophy, prefers to
describe this dichotomy as an opposition between teleological and materialist approaches, reserving

Aristotle on the Lung and the Bellows–Lungs Analogy 183
In what follows, before focusing on the bellows–lungs analogy, I will
first present Aristotle’s account of the structure, texture and movement of
the lung as well as its connection to the heart. This will allow me to
adumbrate in some detail the theoretical background needed in order to
gain an initial insight into what appears to guarantee the validity of the
proposed analogy. I will then turn to the comparison itself, which I will try
to elucidate through a detailed description of the bellows as a forge instru-
ment, drawing upon material from archaeological sources depicting the
variety of bellows used up until Aristotle’s time. This in turn will allow me
to propose what seems to be the most likely candidate for the position of
the bellows Aristotle has in mind when claiming that both lungs and gills
resemble a brazier’s bellows. I will conclude by offering an explanation of
the rationale underlying Aristotle’s incorporation of this analogy into De
Juventute’s account of the lung.
1 Aristotle on the Lung
1.1 The Structure of the Lung

The organ to which Aristotle refers as pneumôn (in the singular) is double,
consisting of two halves, the right and the left (called ‘pulmonary lobes’
today), which, however, jointly perform the lung’s central function.8 The
distance that separates the two parts of the lung varies from species to
species. For example, in the case of egg-laying animals, as Aristotle carefully
observes, the two parts are separated by a greater distance from each other
than in the case of live-bearing animals, and especially in humans, where
the two parts are closer together.9 The lung is connected to both the
trachea and the heart. According to Aristotle, each of the lung’s parts is
joined to the heart in a twofold way: first, the right part of the lung is
connected to the largest cavity of the heart (right ventricle) by the great
vessel (megalê phleps, which refers to both the vena cava and the right
atrium),10 and the left part of the lung is connected to the middle cavity by
the term ‘mechanistic’ for a later, third approach inspired by Hellenistic mechanics; see also
Berryman 2009, 15–17, 209–10.
8
HA i 16.495a34; 495b5–7.
9
PA iii 7.669b23-25; HA i 16.495a34–b1, b3–6. Pace Oser-Grote 2004, 202. Here I use the terms ‘egg-
laying animals’ and ‘live-bearing animals’, following Lennox’s 2001 suggestion, for which see pp.
xii–xiii. Indeed, this variation in distance is confirmed if we compare the distance separating the two
halves, say, of a frog’s or a tortoise’s lung (animals whose lung Aristotle seems to have inspected, as is
clear from his descriptions in De Juventute) to that of the human lung.
10
See Dean-Jones 2017, 130, as well as her depiction of Aristotle’s conception of the heart at p. 131.

184 giouli korobili
the aorta;11 and second, the heart is connected to the lung by means of
passages (poroi) which are of considerable importance for cooling the heart
(more on this below). The lung is connected to the trachea (which Aristotle
calls artêria) at a particular point: the bifurcated trachea terminates in the
right and left parts of the lung via orifices (what we today call ‘bronchi’)
which are responsible for supplying air to the lung.12 These orifices, which
are thickest at the point where the trachea starts to branch and gradually
become thinner, are referred to by Aristotle as diaphuseis (germinations) or
suringes (pipes), whose thin ends (trêmata; ‘bronchioles’) spread over the
entire length and width of the lung.13
Considerable uncertainty remains in relation to the passages that con-
nect the heart to the lung so that the heart can be cooled. It is not at all clear
from the text whether air does or does not flow into the heart through these
passages. This uncertainty makes it difficult to give a complete picture of
how the lung performs its cooling function, and correspondingly difficult
to understand how such a function may resemble the blowing of air into
a furnace by a bellows. Aristotle himself concedes that the task of observing
such passages presents many difficulties, principally due to their minuscule
size.14 In a rather complicated passage in Historia Animalium, these poroi
are distinguished by their ability to take in breath and pass it on to the heart
‘in virtue of their contact’:
T1 Passages (πόροι) also lead into the lung from the heart, and branch off
just as the windpipe does, running all over the lung parallel with the passages
coming from the windpipe (παρακολουθοῦντες τοῖς ἀπὸ τῆς ἀρτηρίας).
Those from the heart are uppermost; and there is no common passage
(πόρος δ’ οὐδείς ἐστι κοινός), but in virtue of their contact (διὰ τὴν
σύναψιν) they receive the breath and pass it on (διαπέμπουσιν) to the
heart; for one of the passages conveys it to the right cavity, and the other
to the left. (HA i 17.496a27–34)15
11
HA i 16.495b7–8, 17.496a24–7. By ‘middle cavity’ Aristotle possibly refers to the left ventricle, which
is larger than the left atrium, the latter representing in all probability the left and smallest cardiac
cavity in Aristotle’s view; see Dean-Jones 2017, 130 and 132.
12
HA i 16.495b8–10.
13
HA i 16.495b5–12; iii 3.513b18; Juv. 27 (21).480b6–9. Within the Aristotelian corpus only in HA
i 16.495b9 do we find Aristotle referring to some hollow parts which probably exist at the terminat-
ing points of τρήματα, and to which the air passes during inhalation (διαδίδωσιν εἰς τὰ κοῖλα μέρη
τοῦ πνεύμονος τὸ πνεῦμα). For some additional Aristotelian references in which the term κοιλία is
used to denote the reservoir or end point of a fluid material, see Wee 2017, 147.
14
HA iii 3.513a35–7; compare 513b22–3 and 4.514a25–7. Galen makes a similar comment regarding the
beginning of the optic nerves (De placitis Hippocratis et Platonis p. 450, 2–3 de Lacy): they have
a passage (τρῆμα) which is ‘difficult to see’ (δυσθεώρητον).
15
All translations are taken from Barnes’s 1984 edition, with slight modifications.

While the phrase διὰ τὴν σύναψιν suggests that at some point there must be
contact between branches of the trachea and those stemming from the heart,
it is left unclear under what circumstances this contact takes place. Aristotle
is careful to underline that the pipes or canals originating from the heart (i.e.
the blood vessels) not only run all over the lung, but also closely follow
(παρακολουθοῦντες) the branched breath-tubes of the trachea, being con-
sistently uppermost.16 One might be tempted to take the word
διαπέμπουσιν literally, in the strict spatial sense that a certain amount of
pneuma is transferred to the heart via a common passage between the
branches of the trachea and of the heart; Aristotle, however, categorically
denies the existence of such a passage.17 Harris, after rejecting Huxley’s
proposition that some kind of transudation or diffusion must occur through
which the inhaled air passes into the blood vessels, follows a rather Galenic
approach: he claims that we should instead speak of some sort of breath
transfer, from the thinnest germinations of the trachea to the blood vessels,
which takes place at the point of their convergence.18 But Aristotle nowhere
says this explicitly;19 instead, he insists on underscoring the proximity and
‘contactual’ connection between the pipes. Similarly in De Juventute:
T2 But just as when [the heat] was increasing this part (i.e. the chest) rose, so too
when [the heat] decreases it (i.e. the chest) must contract, and when it collapses
the air which entered must pass out again. When it enters the air is cold, but on
issuing it is warm owing to its contact with (διὰ τὴν ἁφὴν) the heat residing in this
organ, and this is especially the case in those animals that possess a full-blooded
lung. For the air passes (ἐμπίπτειν) into numerous canal-like ducts in the lung,
along each of which blood-vessels lie (παρ’ ἑκάστην παρατέτανται), so that the
whole lung is thought to be full of blood. (27 (21).480b1–9)
The inhaled air manages to cool the heat of the blood vessels occupying the
lung (and thus itself to become warm) due to the contact of the bronchial
ramifications and the blood pipes.20 The expression διὰ τὴν ἁφὴν, as well
as the overall emphasis on the notion of proximity (παρ’ ἑκάστην
παρατέτανται), seems to discourage us from thinking of some kind of
intake of air that takes place in the blood vessels. So, if we had to decide
16
Similarly in HA iii 3.513b17–25.
17
Compare the meaning of διαπέμπειν in PA iv 5.681a30. Juv. 21 (15).478a23–6, mentioned by Althoff
1992, 161, presumably refers, more broadly, to the region of the lung where the air is still in motion.
18
Harris 1973, 156–7; compare 124 and 307–9.
19
Even in HA i 16.495b16 (εἰς αὐτήν) Aristotle refers not to the heart, but to the trachea, thus
suggesting that there is no direct connection between them.
20
See also Juv. 22 (16).478b17–18: θερμαίνει γὰρ ἡ τοῦ αἵματος θίξις ἑκάτερον. θερμὸν δ’ ὂν τὸ αἷμα
κωλύει τὴν κατάψυξιν.

186 giouli korobili
between the interpretations of Harris and Huxley, seeing that neither of
them enjoys direct textual support, we would tend to align ourselves with
Huxley.
However, is it in fact necessary that air should flow to the heart if the blood
is to be cooled? Let us take the example of the brain, which is, as Aristotle often
states, the coldest of all the bodily parts.21 On the principle that nature has
contrived to counterbalance the excess of heat and cold in every bodily part,
the brain accordingly serves as a counterpoise to the heat contained in the
heart. The heart, on the other hand, must be seen as an organ contributing,
along with the spinal marrow, to moderating the coldness of the brain;22 and it
does so by a network of numerous small blood vessels which are full of thin
and clear blood and surround the brain (περιέχουσιν αὐτόν).23 In other words,
the blood coming from the heart does not need to be mixed with anything else
(e.g. with air) in order to counterbalance the excess of heat. Why is it then that
in the case of the lung the very point of contact between the blood vessels and
the branched endings of the artery is not responsible for effecting cooling?
Something similar happens in fish:
T3 It is in the same way that the motion of the gills in fishes takes place. For
when the heat in the blood rises through the members, the gills rise too, and
let the water pass through, but when it retreats through its channels to the
heart and is chilled, the gills contract and eject the water. As the heat in the
heart continually rises, the heart continually receives it, and expels it again
when it is chilled. (Juv. 27 (21).480b12–17)24
The text clearly shows that in fish the heat of the blood and the (cold) water
can neither be mixed with one another nor can they coexist in the same
vessel. Their contact, which takes place at some point after the water has
entered a fish’s mouth and before it is expelled by the gills, allows the blood
to cool.25 A few chapters earlier, Aristotle, while touching upon this same
subject, had provided the following additional information on the
21
See, for example, PA ii 7.652a27–8. 22 PA ii 7.652a27-33; 652b17–653a10.
23
PA ii 7.652b31–3.
24
Aristotle frequently compares animals with lungs and animals with gills, given that the two organs
share the common function of cooling, the one by means of air, the other by water; see, for example,
Somn. 456a6–11. The case of animals, mostly insects, effecting cooling through their innate pneuma
(ἐμφύτῳ πνεύματι, see Juv. 15 (9).475a8) is often adduced as a parallel by scholars claiming that air
does flow to the heart of lunged animals. Nevertheless, I do not think we have enough evidence to
give a definite answer on this point, mainly because Aristotle does not tell us whether in the above-
mentioned animals innate pneuma and their ‘analogue of blood’ (since they are bloodless) coexist in
the same place/vessel, nor does he go into any detail about the way in which these two cooperate in
order to effect cooling.
25
Compare HA ii 13.504b29–30.

anatomy of the gills: as the water flows continuously through the gills
(diaulônizontos), ducts running from either side of the heart to each gill’s
extremity effect the cooling of the heart.26 One must therefore tentatively
conclude that we should not speak of air being transported to the heart, but
rather of a process in which the blood contained within the vessels (inside
the lung) becomes cooled, and this cooling effect is diffused in the heart.27
1.2 The Texture of the Lung

With regard to its texture, the human lung is neither smooth nor is it split
into many parts (poluschidês), but rather constitutes an intermediate con-
dition, exhibiting, as Aristotle puts it, a certain unevenness.28 One of its
most characteristic features is its sponginess, endowing it with elasticity
and the capacity to expand to allow an inflow of air (poiôn euruchôrian).29
In live-bearing animals, the lung is rich in blood due to the blood con-
tained in its veins.30 Broadly speaking, the richer in blood the lung is, the
more cooling the animal needs to receive.31 To be sure, the lung’s richness
in blood was a factor that students of animal anatomy tended to neglect or
ignore. This is clear from the censorious reference to contemporary dis-
sectors at HA i 17.496b4–6, just before the end of his description of the
lung. Here, Aristotle criticises those who claim that the lung is empty (i.e.
bloodless) based on erroneous observations after cutting into the lung. This
implies that either Aristotle himself or one of his consultants was in
a position to observe the texture and constitution of an animal’s lung
before it was removed from its location in the body. Whether or not that
body was still alive remains, of course, unclear.32
26
Juv. 22 (16).478b13–14.
27
Similarly King 2001, 120 (cf. 189, n. 287), who adds two more arguments: (a) Aristotle stresses that
the lung is full of blood, in this way somehow anticipating the importance of ‘proximity’; and (b)
Aristotle denies that breath is a form of nutrition for bodily heat, which means that breath is not
consumed when it is breathed in.
28
ἔχει ἀνωμαλίαν, HA i 16.495b3. Indeed, while in humans the left half is divided into two lobes, and
the right into three (see e.g. Oser-Grote 2004, 203, n. 58), the right and left lungs of a dog are divided
respectively into four or five, and three lobes (Fleming 18732, 471). In tortoises, to take another
example (although they are not live-bearing animals), each lung has five or six lobes (Gans and
Hughes 1967, 4).
29
HA i 17.496b4; PA iii 6.669a14–15. It should be noted that the lung was believed to be spongy in
texture due to its ability to let liquids pass through (Thivel 2005, 242).
30
HA i 17.496a35–b9. 31 Juv. 7 (1).470b25-26; PA iii 6.669a25–6.
32
Compare ἔκ τε τῶν ἀνατεμνομένων in Juv. 22 (16).478a26; cf. also Juv. 22 (16).478a35–b1 on
dissection of fish.

188 giouli korobili
1.3 The Movement of the Lung
When Aristotle speaks of the movement specific to the lung (oikeia kinêsis),
which takes place for the sake of cooling, he emphasises that this kind of
movement is of crucial importance for the maintenance of life, since
a breathing animal’s inability to move the lung, due to affection (pathos)
or old age, causes death.33 Over time, the lung gradually loses its elasticity
and becomes drier, resulting in a state of impeded motion, as it can no
longer easily expand or contract.34 Similarly, in the case of diseases which
are manifested in tumours, secretions or excess heat, the lung’s inability to
expand and contract, due to its hardening, results in an acceleration of the
breathing rate.35
Aristotle maintains that the lung’s movement is derived from the
heart.36 The efficient cause of this movement is the increase of the innate
heat and the central principle at work seems to be ‘the power of the void’
inside the lung.37 As De Juventute 27 (21) makes clear, the increase of the
heat in the region of the heart causes the lung to expand and so to draw in
air through the windpipe. This process can to some extent be perceived
externally, by observing the movement of the chest. As the heat increases
and causes the lung to expand, the bodily part surrounding it, the chest, is
forced to expand simultaneously.38 Now, in order for the external air to be
drawn into the lung we must assume that a certain force needs to be exerted
by the interior of the chest, and that the force is actuated by the increase of
the heat in the heart.39 Aristotle believes that, for animals that cool
themselves by means of air, a certain kind of attraction of the external air
takes place, which can be correlated directly with the movement of the
lung; yet this is not the case in fish.40 In De Somno et Vigilia, while
discussing movement, breathing and cooling in different kinds of animals,
Aristotle admits that it is not possible for movement to occur without
strength:
T4 And since to move anything (κινεῖν μέν τι), or do anything, is impossible
without strength (ἄνευ ἰσχύος ἀδύνατον), and holding the breath produces
strength – in animals that inhale, the holding of that breath which comes
from without, but in animals that do not respire, the holding of that which
is connatural. (2.456a15–18)
33
Juv. 15 (9).475a24; 7 (1).470b21–3; 22 (16).478b19–22. 34 Juv. 23 (17).479a11–15.
35
Juv. 23 (17).479a24–9; cf. HA vii 23.604a17–21. 36 PA iii 6.669a14.
37
See Berryman 2009, 75. 38 Juv. 27 (21).480a25–9.
39 40
Compare Juv. 10 (4).472a23–5; 27 (21).480a16–20. Juv. 9 (3).471a27–9.

Air has an inherent ability to slip through any place in the lung, and to do
so rapidly, due to its rarefied nature.41 As soon as the air reaches the
outermost branches of the tracheobronchial tree – the trêmata – its
dynamic movement enables it to produce a cooling effect on the branches
of the blood vessels stemming from the heart; and it does so due to the
proximity and contact between these separate branches. When the cooling
of the blood’s heat has reached a certain point, the lung begins to contract,
expelling the air that is now warmer after contact with the heat of the blood
vessels. In what follows, I will discuss the noteworthy device that Aristotle
employs to illustrate the movement of the lung: the analogy with a forge
bellows.
2 The Bellows–Lungs Analogy

In the course of discussing his predecessors’ views on respiration in De
Juventute, one of the main targets of Aristotle’s criticism is Empedocles. In
chapter 13 (7) specifically, Empedocles comes in for severe criticism on two
counts: (a) he does not discuss the purpose of respiration, and (b) he does
not specify whether or not all animals perform that function.42 On this
basis his views are, according to Aristotle, nothing but unsubstantiated
claims devoid of ‘science’.43 In the same vein, a little later, Aristotle
introduces a lengthy passage from Empedocles’ poem (Β100 DK) with
the dismissive comment: ‘[Empedocles] likens this process44 to what
occurs in a clepsydra’ (παρεικάζων τὸ συμβαῖνον ταῖς κλεψύδραις,
473b8).45 For Aristotle, the clepsydra analogy is quite inappropriate, since
it is highly doubtful that it tallies with the actual facts of respiration. That
is, it remains unclear how the structure and operation of the clepsydra can
41
Juv. 21 (15).478a18–20. 42 473a15–17. 43 Compare οἴεται, 473a18.
44
That is, the process of the air flowing in or out when the blood is carried downwards or upwards,
respectively.
45
A clepsydra was ‘a household implement in the shape of a bulb with a stem or an inverted funnel.
The extended tube was open at the end while the bell was enclosed with a floor which was perforated
with holes like a colander’ (Worthen 1970, 520). For a detailed description of the clepsydra’s
workings, see Last 1924 and esp. p. 170 for illustrations; compare Guthrie 1939, 228. Perhaps more
than any other Presocratic philosopher, Empedocles makes extensive use of analogical reasoning in
his physical and biological inquiries; very often, however, his views come in for criticism by
Aristotle, sometimes because he thinks them obscure (and so inadequate for understanding the
nature of things), sometimes because he finds that the dissimilarities outweigh the similarities
between the two things being compared by Empedocles (Lloyd 1966, 325–36 and 403). There has
been much scholarly debate about Empedocles’ clepsydra simile, but more extensive analysis is
beyond the scope of this chapter. Detailed discussions include Powell 1923; Last 1924; Timpanaro
Cardini 1957; Booth 1960; Lloyd 1966, 328–33; O’Brien 1970; Worthen 1970; Bremer 1980; Rashed
2008.

190 giouli korobili
adequately represent a model for the process of breathing. This analogy
also constitutes, in and of itself, a crucial methodological error due to its
reliance on a second-degree conjecture (cf. παρ-εικάζων) – on speculative
reasoning founded on an inconclusive supposition. Not only did
Empedocles merely surmise the existence of certain blood vessels in the
body which pull in and push out air during inhalation and exhalation (a
hypothesis Aristotle clearly rejects by referring to animals that evidently
breathe, T5), in addition he tried to buttress this erroneous assumption
with a second conjecture, the clepsydra analogy.46
But what is that Empedocles should have done in order to render his
theory consistent with, and fully confirmable by, the facts?
T5 This, then, is what he says about respiration. But, as we have noted, all
animals that evidently respire do so by means of the windpipe, using the
mouth and the nostrils at the same time. Hence, if it is of this kind of
respiration that he is talking, we must ask how the explanation given will
tally with the facts (πῶς ἐφαρμόσει ὁ εἰρημένος λόγος τῆς αἰτίας). (Juv. 13
(7).474a6–11)
Of great importance here is the absolute use of the verb ἐφαρμόζω, given
the emphasis it puts on asking how, if at all, Empedocles’ account of the
cause of respiration fits with the phenomenon of respiration. Cognate with
ἀραρίσκω, ἐφαρμόζω conveys the notion of something fitting together into
a perfectly coherent whole. The ‘joints’ that need to fit together closely in
order to form a complete theory of respiration refer to nothing other than
what in Aristotle’s view is the ideal conjoining of what we observe happen-
ing inside the body (the ‘seen’) and what we reasonably infer as happening
inside that same body based on external observation (the ‘unseen’). This, as
will become clear shortly, is Aristotle’s main criterion for assessing the
accuracy of a given analogy for breathing.
But if Empedocles’ explanation is so problematic, why does Aristotle
quote it at length? Why does he not simply refer to it in passing, just as he
did with the accounts of the other natural philosophers? A comprehensive
answer to this question requires a multifaceted response. I restrict myself
here to what I consider one of the prime reasons for dwelling upon
Empedocles’ account: the use of analogical thinking in physiological
46
Juv. 13 (7).473b2–9. A similar use is already found in Plato’s Politicus (260e3), where παρεικάζειν is
employed to refer to a conjecture that is to conform to an inference drawn from analogy. Other
instances of this signification of παρεικάζειν in Aristotle include Meteor. 369a30 and De Insomn.
461b20. In the latter, the residual movements that occur in sleep change form so quickly that they
can hardly be clearly likened (παρεικάζουσιν) to human beings or centaurs.

explanation. In presenting a detailed analysis of Empedocles’ analogy,
Aristotle seizes the opportunity to counterpropose a model that is capable,
both structurally and functionally, of illustrating with much greater clarity
the phenomenon of breathing.47 His text continues as follows:
T6 For what we see happening is the opposite (φαίνεται γὰρ τοὐναντίον
συμβαῖνον). By raising (ἄραντες) the region (i.e. of the chest), just like the
bellows in a forge (καθάπερ τὰς φύσας ἐν τοῖς χαλκείοις), they breathe in – it
is reasonable (εὔλογον) that heat should raise it, and that the blood should
occupy the hot region; but they breathe out by causing it to collapse
(συνιζάνοντες) and sink down (καταπλήττοντες), like the bellows in the
other case. (Juv. 13 (7).474a11–15)
Aristotle accordingly has no inherent objection to the use of analogies in
‘scientific’ discourse, provided that they do in fact elucidate the phenomena
under consideration. What he is objecting to is the cogency of the clepsydra
analogy and its use to illustrate the phenomenon of respiration. This is evident
from his immediately following suggestion that a more apt analogy for how
respiration occurs in animals is the bellows in a forge.48 Aristotle returns to the
analogy with the forge bellows in chapter 27 (21), where both heart and lung
are said to conform closely to that shape; but it should not escape us that in
both places he puts considerable emphasis on what an observer can see and
interpret, in this case the rise and fall of the chest. Significantly, the analogy
with the bellows appears twice in T6: its purpose is to stress emphatically that
the region of the chest bears a striking resemblance to forge bellows, both in
how it rises, when the breath is taken in, and how it sinks down until it
collapses when the breath is let out – a situation which is perhaps more visible
in the case of a bellows’ skin bag than in a lung.
This text invites consideration of the ‘agent’ or efficient cause of the
movement observed. In the case of the bellows, a human agent is clearly
responsible for the inflation and deflation of the bags.49 In the case of
47
It turns out therefore that, for Aristotle, whether a metaphorical model of viewing a natural object is
convincing or not is dependent on (at least) two considerations: how well the theory proposed in the
model accords with empirical data, and whether the model itself attempts to provide answers to
‘Why’-questions (i.e. questions relating to causes). These two considerations must be addressed for
the model to be regarded as playing an integral role in the conceptualisation of a given natural
object, in a manner acceptable to the science of nature. I thank G. Kazantzidis for pressing me on
this point.
48
The most common term for ‘bellows’ is φῦσα (mostly in the plural, φῦσαι). Less common terms
include πρηστήρ, φυσητήρ, φυσητήριον, ἀσκοί, ζωπύρια, ἀκροφύσια or, rarer still, ἀκροστόμια;
see Smith 18592, 543 and Healy 1978, 193.
49
See the pseudo-Aristotelian De Audibilibus 800a34–b3: ‘For, because [the lung] is hard and thick
and constricted, it does not admit of dilatation to any great extent, nor again can it force out the

192 giouli korobili
animals that breathe, on the other hand, the increase of the hot substance
in the body’s mid-region, specifically in the heart, appears to function as
the efficient cause of the lung’s movement. However, in T6, the repeated
used of participles (ἄραντες; συνιζάνοντες; καταπλήττοντες) seems to
point to an active agent other than the heat in the heart, one that could
be identified mutatis mutandis with the living, or, more precisely, the
human being. On this view, humans themselves are included among the
causes effecting breathing, inasmuch as they manage to inhale or exhale by
raising or collapsing the region of their chest; and the success of the analogy
lies in its ability to make humans apprehend their role as active agents in
a phenomenon that takes place within their own bodies.
An indication that Aristotle is not merely deploying a loose, ad hoc
analogy may be seen in the qualification that follows the text quoted above:
T7 Except that in that case (i.e. in a bellows) they do not admit and expel the
air by the same channel (οὐ κατὰ ταὐτὸν), but those who breathe do. (Juv.
13 (7).474a16–17)
This addition is significant since, as we shall see in Section 3, it helps us
limit the number of likely candidates among the different kinds of bellows
that were in use in Aristotle’s time: bellows were divided into two types,
depending on whether they had one or two openings for the admission
and/or expulsion of air.
A few chapters later, while setting forth his own account of respiration,
Aristotle once again has recourse to the bellows–lungs analogy:
T8 Respiration takes place owing to the increase of the hot substance (τοῦ
θερμοῦ) in which the nutritive principle is seated (ἐν ᾧ ἡ ἀρχὴ ἡ θρεπτική) . . .
Now as the heat increases it necessarily causes the organ to rise. This organ we
must take to be constructed (δεῖ δ’ ὑπολαβεῖν τὴν σύστασιν τοῦ ὀργάνου) in
a manner closely resembling (παραπλησίαν) the bellows in a smithy, for both
heart and lungs closely conform (οὐ πόρρω . . . τοῦ προσδέξασθαι) to this
shape. Such an organ must be double, for the nutritive principle must be
situated in the centre of the cooling50 force. (Juv. 27 (21).480a16–24)
Aristotle develops this analogy in a very deliberate manner in order to preclude
any possibility that explanans and explanandum may somehow be considered
as having no relation to each other. In a bid to avoid the same mistake
previously attributed to Empedocles (cf. παρεικάζων, p. 189), Aristotle is
breath by contracting after wide distension; just as we ourselves cannot produce any effect with
bellows, when they have become hard and cannot easily be dilated and closed.’
50
I follow Ross 1955, 340, in reading ψυκτικῆς instead of φυσικῆς. Siwek 1963, 367, n. 23, adopts
φυσικῆς. For an alternative way of interpreting the passage, see Althoff 1992, 160, n. 21.

careful to stress that the structure of the lung closely resembles (παρα-πλησίαν)
that of a forge bellows, in that the shapes of both the heart and the lung closely
conform (lit. are not too far from conforming) to a bellows’ shape. The same tone
is also continued by the syntactic construction of the passage, an impersonal
construction with δεῖ denoting a necessity followed by a series of verbal forms,
mostly infinitives. Now, the point of this analogy is not merely to draw
momentary attention to a close resemblance in terms of shape; were that the
case, Aristotle’s analogy would turn out to be as problematic as Empedocles’, as
it would remain just as unclear how it corresponds with (cf. ἐφαρμόσει, T5) the
true causes of respiration. Instead, in addition to the resemblance in shape,
Aristotle emphasises that this very double structure (a feature also characteristic
of the bellows, as we shall see in Section 3) functions as an ideal cooling
environment, at the centre of which lies the nutritive part of the soul. Such
a cooling environment necessarily has an inherently teleologically oriented
structure, one that must serve the purpose of cooling. Astonishing though
Aristotle’s reference to the similar structure of a bodily organ and of an artefact
may at first seem, it should not surprise us, since σύστασις (and its cognates) is
a term frequently used by Aristotle in descriptions of the purpose for which the
various organs of the body are formed.51 On this basis, then, the shape (of the
lung) and the pneumatic action this shape involves must be innate in the lung
already at the moment of its formation; and that the lung must therefore be
able to accommodate the purpose for which it has been constructed and has
been assigned its own proper function: namely, the cooling of the most vital
organ, the heart. More information on cooling, and how this is effected, is
provided by the following passage (parts of which we have already discussed):
T9 Thus as the heat increases it rises (αἴρεται), and as it rises the part
surrounding it must also rise (αἴρεσθαι). This is what those who breathe
seem (φαίνονται) to do; for they raise (αἴρουσι) their chest because the
principle of the organ described resident within (τὴν ἀρχὴν τὴν ἐνοῦσαν)
the chest does so. And as the chest rises (αἰρομένου) the air from outside,
being cold, must rush in as into a bellows, and by its chilling influence
reduce by extinction (σβεννύναι) the excess of the fire. But just as when [the
51
See, for example, the σύστασις of the testes, GA i 4.717a15; of female menstrual discharges, GA
i 19.727b32 and 20.729a22; or of nails or hoofs, GA ii 6.744b26. See also De Juventute 19 (13), where
teleology and necessity are combined in order to account for the reason why each kind of animal has
been assigned its own particular constitution. References to the lung’s structure can also be found in
Juv. 23 (17).478b27 and 25 (19).479b16. Sometimes Aristotle uses the term σύστασις figuratively to
refer, for example, to a state’s constitution (Polit. iv 11.1295b28, viii 13.1332a30) or to the structure of
a plot (Poet. 10.1452a18–19), of the incidents of a story (Poet. 6.1450a15, 14.1453b2) or even of tragedy
in general (Poet. 13.1453a23). I have not, however, been able to find any other reference to an artefact’s
structure, unless we call poetry an artificial object (see e.g. Poet. 13.1453a22–3).

194 giouli korobili
heat] was increasing, this part (i.e. the chest) rose, so too when [the heat]
decreases it (i.e. the chest) must contract, and when it collapses the air which
entered must pass out again. When it enters the air is cold, but on issuing it
is warm owing to its contact with (διὰ τὴν ἁφὴν) the heat resident in this
organ, and this is especially the case in those animals that possess a full-
blooded lung. For the air passes (ἐμπίπτειν) into numerous canal-like ducts
in the lung, and blood vessels extend along each of them (παρ’ ἑκάστην
παρατέτανται), so that the whole lung is thought to be full of blood. The
entry of the air is called inspiration, and its exit expiration, and this goes on
continuously just so long as the animal lives and keeps this part in continu-
ous motion; it is for this reason that life depends upon inspiration and
expiration. (Juv. 27 (21).480a25–b12)
Before leaving Aristotle’s use of the bellows–lungs analogy, let me dwell for
a moment on the peculiarity of the lung’s movement. As is clear from T9,
Aristotle understands the movement of the lung as a kind of upward-and-
downward movement following the way heat rises and falls in the heart.52
The verb consistently used to describe that movement is αἴρεσθαι, which
indicates a movement away from the place where the lung is situated, in
a direction parallel to the direction which the heart takes when the heat
inside it increases. However, it should be assumed that another type of
motion occurs at the same time. This motion or ‘spatial change’ is one that
any receptacle undergoes while adjusting its volume in accordance with the
material it is filled with – the lung is a sort of receptacle which takes in air
while being lifted up following the heat’s rising. Thus, besides the lung’s
rising and falling, we must assume that it changes its position also in terms
of inflation and deflation (which means that the organ moves spatially in its
whole perimeter). While this latter kind of motion is not mentioned by
Aristotle, we may assume that it is presupposed due to the particular nature
of adjustable receptacles, such as the lung. And indeed his analogy with the
bellows can support this conjecture due to the nature of the materials with
which a bellows is formed, as we will soon see.
3 The Archaeological Evidence

I will now turn to a brief examination of the surviving archaeological
evidence, invaluable for Aristotle’s claim about the resemblance between
the lung and the bellows. Considerable material evidence for bellows
survives for the period between the sixth century bce and the first century
52
Provided that we take the heart or the ‘vital fire’ it possesses as the point of reference when the body is
lying down rather than standing up. This is perhaps indicative of a body posture during dissection.

ce, including vase paintings, pieces of sculpture, votive plaques and even
remains of actual bellows (nozzles).53 Given that most bellows in antiquity
were made of perishable materials (such as leather or wood), the depictions
on vases in particular are an extremely valuable source of information
about a bellows’ structure and its everyday use. In what follows, I will
briefly review the various types of bellows used in ancient Greece and
Rome.54
An old (though not the oldest) type of bellows, the dish bellows,
consisting of two pots covered with animal skins has been excavated in
a metallurgical installation called Chrysokamino (‘Golden Furnace’) in
Crete, dating from late Neolithic/early Minoan times. It is thought to
have been active from the end of the fifth millennium bce.55 Each pot
is connected to a tube, and both tubes are inserted into a blast-pipe, so
that when air is drawn into the bellows it can be directed against, and
forced into, the fire. As one of these pots draws in air, the other is
pressed down and forces the air out by the nozzle. In this way, the fire
is constantly supplied with air as long as the bellows operates and the
pots work together in rapid alternation. However, the typical form of
bellows in classical Greece and Rome was the ‘skin bellows’.56
Depictions of such bags can be found on a wide range of Greek
vases.57 This type of bellows consists of one bag (single bellows) or
more usually two bags (double bellows), made from animal (mainly
goat) skins. Each bag’s blowpipe is attached at one end to the bag
(most often by being inserted into the ‘tube’-shaped skin from the
animal’s leg), and at the other end to a clay nozzle.
Bellows are broadly categorised into two kinds depending on the way in
which they receive external air. Some bellows have no other opening for taking
in air than the nozzle of their pipe. In this case the nozzle of the blast-pipe
should lie close to but not inside the furnace, to avoid drawing air back out of
the furnace and into the bellows.58 In most depictions, however, we find
53
See Ulrich 2008, 49, and Mattusch 2008, 432–4. Figures illustrating preserved nozzles can be found
in Hahn 2010, 105–6 and 108.
54
Detailed discussions of the illustrations presented here can be found in studies such as those by
R. Hahn 2010, Archaeology and the Origins of Philosophy, chapter 4 and K. G. Tsaimou 1997,
Αρχαιογνωσία των μετάλλων, 93–101. On the different kinds of metallurgical furnaces in antiquity,
see Forbes 1964, 72–83. For the means of supplying blast air to the furnace, see Forbes 1964, 83–8.
55
Hahn 2010, 108. Suggested reconstruction of the pot bellows in use can be found in Hahn 2010, 109.
See also Tsaimou 1997, 97. For further details and up-to-date bibliography on the smelting process
after the third millennium bce, see Craddock 2008, 102–4.
56
Forbes 1964, 84 and 87–8.
57
See Forbes 1964, 103, n. 346 for bibliography on the depiction of bellows on vases.
58
Hahn 2010, 98.

196 giouli korobili
bellows with two separate passages, one to take in air and another to expel it.
Through one passage on the top of each bag the bellows received air from the
outside. Through the other passage – the end of the blast-pipe – air was
discharged into the furnace.59
In Figure 7.1, from a red-figure cup dating to c. 475 bce, one can see
a single bellows bag, made from a skin with the shaggy coat still on, with
a characteristic opening on the top. The skin bag has slits with two wooden
rims as the openings for admitting air into it. These slits are adjusted by
hand – they are left open when air needs to be drawn into the bag, and are
closed when the bag had to be pressed down to force the air into the fire.60
This specific type of bellows is equipped with two pipes instead of one to
supply fire with air.
Figures 7.2 and 7.3 depict double bellows with openings on top. In
Figure 7.2, the right bag of the bellows is pressed down (by Hephaestus or
Heracles),61 while the left bag is left free. The left bag is made from the
integument of a headless animal: its legs can be seen at the top and bottom
Figure 7.1 Detail in red-figure cup by Douris (c. 475 BCE). Single bellows with two
tubes, after Stefou. See Hahn 2010, 101.
59
Recall Aristotle’s remark on the subject matter at hand (T7).
60
Forbes 1964, 84; Hahn 2010, 98.
61
Hephaestus: Hahn 2010, 99–100; Heracles: Tsaimou 1997, 96.

Figure 7.2 A double bellows. Detail from the Siphnian Treasury (Delphi, c. 525
bce), after Stefou. See Hahn 2010, 100.
Figure 7.3 The right bag of a bellows as it is stretched out, after Stefou. Black-figure
kantharos (c. 550 bce). See Hahn 2010, 99.
of the bag, while at its bottom left-hand side a short curly tail is visible. In
a black-figure kantharos from the mid-sixth century, the shape of the
double bellows is comparatively well preserved: creases can be seen in the
right bag, as it is being pressed, while the left bag is stretched out, full of

198 giouli korobili
air.62 In Figure 7.3, a depiction of the right bag of the bellows, we can
observe on the top Hephaestus’ palm as he holds the bag-sticks used to
press down the bag (more clearly seen in Figure 7.1). Yet what is really
remarkable here is the stippling depicted on the surface. The animal skin is
untanned and its porous nature is uniquely represented. This way of
portraying the outside of the bellows bears a striking resemblance to
Aristotle’s description of spongy lungs. In addition, the creases formed in
the bag as it deflates remind us of Aristotle’s observation that in some
animals the lung is split into many parts (πολυσχιδής) and in humans it
exhibits a certain unevenness.
4 Back inside the Body

With these considerations in mind, Aristotle’s emphasis on the lung’s
resemblance to the bellows is by no means without justification. If we take
a closer look at the construction of the heart and the lung in different
animals, we can easily identify the main points of resemblance which
Aristotle uses to construct his analogy in De Juventute. On Aristotle’s
account, the opening at the top of the bellows (where it takes in air) can
be likened to the lung’s opening(s) at that part of the windpipe where the
right and left main bronchi start to enter the right and left halves of the lung,
thus ensuring an adequate supply of air. On the other hand, the vessels
which convey blood to the lung from the heart might well be equated with
the bellows pipe. Naturally, when one examines the integral functioning of
both lung and bellows, certain differences will inevitably become apparent.
A case in point is the stream of air that is normally expected to be transmit-
ted in both cases: with bellows air is blown into the furnace, whereas in the
case of the lung we should rather speak of a cooling effect being diffused over
the region of the heart. But that does not mean that the bellows–lungs
analogy should be rejected as wholly invalid: its primary purpose, as Aristotle
repeatedly stresses, is not to be stretched to the point of identity, but rather,
from a methodological perspective, to draw attention to close visual resem-
blances based on correct empirical observation (in contrast with
Empedocles’ merely notional analogy with the clepsydra). A similar resem-
blance in shape can be traced in the case of gills as well. Their creasy contour,
specifically, is probably what impressed Aristotle and led him to believe not
only that the cooling organ of fish bears a striking likeness to bellows too, but
62
A similar (though only partial) depiction, preserved on a black-figure Attic vase from the sixth
century bce, can be found in Forbes 1964, 81, and in Tsaimou 1997, 85.

also that he has established, through the bellows analogy, a certain anatom-
ical link among species, especially with respect to the constitution of their
main cooling organ.
Two questions inevitably arise from the above descriptions of the lung
and the bellows, one relating to their structure and the other to the final
cause of their operation. To begin with the first: the main difficulty
confronting the bellows–lungs analogy is the discrepancy between how
in the case of a double bellows each bag must be inflated and deflated in
turn, whereas in the lung both parts expand and contract at the same time.
The solution is found in Aristotle’s further qualification that the lung,
whose structure must be double, should still be considered a single organ
performing a single function. If therefore we regard the bellows as a single,
albeit ‘double-structured’, artificial organ with a single function, then we
can reasonably say that the bellows’ dilation causes the air to be drawn into
it, while its contraction results in the expulsion of air and hence the stoking
of fire. This alternation occurs continuously as long as the supply of air
ensured by each of the two nozzles inserted into the blast-pipe is
maintained.
Now the second, more controversial question of the purpose behind
a lung’s or a bellows’ function (and how these functions can be equated
with each other) is mainly stimulated by the fact that, unlike lungs, bellows
discharge air in order to make a fire (or coals) hotter, rather than cooling
it.63 It seems to me, however, that the problem of this apparent divergence
in purpose, even though it might appear to undermine the validity of the
bellows–lungs analogy, is created by the way in which we attempt to
decode, interpret and evaluate Aristotle’s use of the analogy rather than
by his actual intent behind that use. On this reading, what Aristotle has in
mind when he compares the functioning of the lung and that of a bellows
needs to be considered within the broader context of life and death. Fire is
kept alive as long as the bellows continues to blow out a blast of air. Natural
heat, on the other hand, is kept alive as long as the lung cools down the
region of the heart. The regulation of the quantity of air passing into both
the furnace and the heart is in fact what prevents the fire from dying out.
And in the opening chapters of De Juventute, Aristotle has already
explained in detail that heat, when left to accumulate, dies out by consum-
ing itself if refrigeration does not manage to cool it down. In view of all the
63
In this respect it is worth noting Theophrastus’ comment (De Igne 37) on the high temperatures
obtained for metallurgy by means of the forge bellows, as opposed to the lower temperatures
required for the medical procedure of cauterisation. For another analogy between the field of ‘iron
working’ and the human body in Theophrastus, see the Mayhew in this volume, Chapter 8.

200 giouli korobili
above, I believe that Aristotle’s analogy between bellows and lungs still
preserves its validity.
Conclusion: Why Is This Analogy Introduced Here?

The use of figurative language to describe the inside of the body and the
functions of its various organs appears to have been a common practice at
the time when Aristotle was writing. For the lung in particular, Plato refers
to it as ‘the dispenser (tamias) of air to the body’,64 while Empedocles, as
we have seen, regards the whole process of respiration as similar to the
action of a clepsydra. It is interesting that this particular analogy between
the lung and the bellows is not attested in any surviving Greek text before
Aristotle.65 Whether it was invented by Aristotle or not is a question that
cannot be answered with certainty, although the cautious way in which it is
put forward in De Juventute may be indicative of a conscious attempt to
introduce a novel perspective on the issue under discussion. On the other
hand, if Aristotle invented the analogy, why is its use confined to De
Juventute? Why do we not encounter it in the main discussions of the
lung in De Partibus Animalium and Historia Animalium? The reason,
I believe, lies in the diversity of the purposes served by each individual
reference to, or account of, the lung as they occur throughout the
Aristotelian corpus as a whole. In De Juventute, specifically, Aristotle
presents a succinct account whereby he intends to enumerate and highlight
the main factors ensuring survival in living things.66 Respiration is among
these factors,67 but all previous treatments of the subject were, in Aristotle’s
view, inaccurate. As a result, there was as yet no sure footing upon which to
develop new ideas. In a bid therefore to make use of a familiar example in
order to illustrate plainly (but still informatively) how respiration takes
place, and also what the organ of breathing looks like, Aristotle introduces
the bellows–lungs analogy. It is highly likely that the analogy emerged in
direct response to Empedocles’ mechanistic analogy. First, the model of
the clepsydra is rejected on the basis that it cannot be reconciled with the
facts; then it is displaced by an alternative model, which pays special
attention to the shape of the bellows and its more evident similarities to
the lung. At the same time, the bellows–lungs analogy seems perfectly
64
Timaeus 84d2. 65 For Anaximander, compare n. 5 above.
66
I offer a more detailed account of this issue in Korobili 2022. It should be noted here that in his
digression on respiration (Juv. 7 (1)–14 (8)) Aristotle seems to assume some previous knowledge of
the anatomy of the chest.
67
Juv. 27 (21).480b11–13.

aligned with Aristotle’s own teaching purposes, functioning as a heuristic
tool for grasping the internal anatomy of the chest solely by recalling and
dwelling on the structure of a bellows.
RE F E RE N CE S
Althoff, J. 1992. Warm, kalt, flüssig und fest bei Aristoteles: Die elementarqualitäten
in den zoologischen Schriften (Stuttgart).
Barnes, J., ed. 1984. The Complete Works of Aristotle: The Revised Oxford
Translation, 1-vol. digital edition, Bollingen Series lxxi.2 (Princeton).
(Cambridge).
Booth, N. B. 1960. ‘Empedocles’ Account of Breathing’, JHS, 80: 10–15.
Bremer, D. 1980. ‘Aristoteles, Empedokles und die Erkenntnisleistung der
Metapher’, Poetica, 12: 350–76.
Craddock, P. T. 2008. ‘Mining and Metallurgy’, in J. P. Oleson, ed., The Oxford
Handbook of Engineering and Technology in the Classical World (Oxford),
93–120.
De Lacy, P. 2005. Galeni De placitis Hippocratis et Platonis, vol. 2, Corpus
Medicorum Graecorum v 4,1,2 (Berlin).
Dean-Jones, L. 2017. ‘Aristotle’s Heart and the Heartless Man’, in J. Z. Wee, ed.,
The Comparable Body: Analogy and Metaphor in Ancient Mesopotamian,
Egyptian, and Greco-Roman Medicine (Leiden), 122–41.
Fleming, G. 18732. The Comparative Anatomy of the Domesticated Animals by
A. Chauveau, translated and edited by George Fleming (London).
Forbes, R. J. 1964. Studies in Ancient Technology, vol. 4 (Leiden).
Gans, C. and G. M. Hughes. 1967. ‘The Mechanism of Lung Ventilation in the
Tortoise Testudo Graeca Linne’, J. Exp. Biol., 47: 1–20.
Guthrie, W. K. C. 1939. Aristotle: On the Heavens (Cambridge, MA).
Hahn, R. 2010, Archaeology and the Origins of Philosophy (Albany, N. York).
Harris, C. R. S. 1973. The Heart and the Vascular System in Ancient Greek Medicine
(Oxford).
Healy, J. F. 1978. Mining and Metallurgy in the Greek and Roman World (London).
in Ancient Greece (Princeton).
Johnson, M. R. 2017. ‘Aristotelian Mechanistic Explanation’, in J. Rocca, ed.,
Teleology in the Ancient World (Cambridge), 125–50.
Korobili, G. 2022. Aristotle. On Youth and Old Age, Life and Death, and Respiration
1–6. With Translation, Introduction and Interpretation (Cham).
Last, H. 1924. ‘Empedokles and His Klepsydra Again’, CQ, 18: 169–73.
Lennox, J. G. 2001. Aristotle: On the Parts of Animals i–iv (Oxford).
Longrigg, J. 1993. Greek Rational Medicine (London).

202 giouli korobili
Mattusch, C. 2008. ‘Metalworking and Tools’, in J. P. Oleson, ed., The Oxford
Handbook of Engineering and Technology in the Classical World (Oxford),
418–38.
O’Brien, D. 1970. ‘The Effect of a Simile: Empedocles’ Theories of Seeing and
Breathing’, JHS, 90: 140–79.
Oser-Grote, C. M. 2004. Aristoteles und das Corpus Hippocraticum (Stuttgart).
Powell, J. U. 1923. ‘The Simile of the Clepsydra in Empedocles’, CQ, 17: 172–4.
Rashed, M. 2008. ‘De qui la clepsydre est-elle le nom? Une interpretation du
fragment 100 d’Empédocle’, REG, 121.2: 443–68.
Ross, W. D. 1955. Aristotle: Parva Naturalia. A Revised Text with Introduction and
Commentary (Oxford).
Siwek, P. 1963. Aristoteles: Parva Naturalia graece et latine (Rome).
Smith, W. 18592. Dictionary of Greek and Roman Antiquities (Boston).
Biologie: Intentionen, Methoden, Ergebnisse (Stuttgart), 183–208.
Thivel, A. 2005, ‘Air, Pneuma and Breathing from Homer to Hippocrates’, in
P. J. van der Eijk, ed., Hippocrates in Context: Papers read at the xith
International Hippocrates Colloquium University of Newcastle upon Tyne 27–
31 August 2002 (Leiden), 239–49.
Timpanaro Cardini, M. 1957. ‘Respirazione e Clessidra’, PP, 12: 250–70.
Tsaimou, K. G. 1997. Αρχαιογνωσία των μετάλλων: αρχαία μεταλλευτική και
μεταλλουργική τεχνική (Athens).
Ulrich, R. 2008. ‘Representations of Technical Processes’, in J. P. Oleson, ed., The
Oxford Handbook of Engineering and Technology in the Classical World
(Oxford), 35–61.
Wee, J. Z. 2017. ‘Earthquake and Epilepsy: The Body Geologic in the Hippocratic
Treatise On the Sacred Disease’, in J. Z. Wee, ed., The Comparable Body:
Analogy and Metaphor in Ancient Mesopotamian, Egyptian, and Greco-Roman
Medicine (Leiden), 142–67.
Worthen, T. D. 1970. ‘Pneumatic Action in the Klepsydra and Empedocles’
Account of Breathing’, Isis, 61.4: 520–30.

chapter 8
The Ill Effect of South Winds on the Joints

in the Human Body
Theophrastus, De ventis 56 and Pseudo-Aristotle,
Problemata 1.24
Robert Mayhew
The focus of this chapter is Theophrastus’ description in De ventis 561 of

a negative effect of Notos (the south wind) on the physical condition of
humans and a problêma clearly related to it in the Aristotelian Problemata
physica, namely Pr. 1.24.2 I think it much more likely that Pr. 1.24 is raising
a question about Vent. 56 than that Theophrastus wrote Vent. 56 in
response to this problêma. I cannot argue for that chronology here, but
read alongside one another, Vent. 56 seems (to me at least) likely to be the
original or earlier text (I present them together in the Appendix). In any
case, I believe I am justified in treating Vent. 56 and Pr. 1.24 in this order
chronologically, at least as a hypothesis.3
In brief, what I aim to show in this chapter (and why chronology is
relevant) is that Theophrastus in Vent. 56 employs material principles to
explain certain features of both living beings and inanimate things, and
that the author of Pr. 1.24 uses these same principles but in a different way,
1
Standard chapter divisions in De ventis are the creation of Schneider 1818. For ease of reference,
however, I refer to these chapters and not to, for example, ‘what we now call Vent. 56’.
2
The text of Vent. 56 is from Mayhew 2018, 62, with some modifications. The text of Problemata 1.24 is
from Marenghi 1999, 56. See the Appendix for details. (A second, briefer problêma, Pr. 26.42, also
related to Vent. 56, is of little interest. See the following note.) All translations are my own unless
otherwise indicated.
3
I also think it likely that the author of Pr. 1 (on medicine) first raised a question about Vent. 56, with
the author or compiler of Pr. 26 (on winds) later taking Pr. 1.24, emending and abridging it, and
making it part of the set of problêmata on winds (i.e. Pr. 26.42 – see note 36 for text and translation).
But certainty is hard to establish here, and other relationships among these texts are possible. For
instance, there is a lost Problemata by Aristotle, which seems to have included problems on wind (see
Flashar 1962, 303–16, Mayhew 2011, 1: xvi–xxi, and Bodnár 2015, 1–4). It is at least possible that this
was the ultimate source (directly or indirectly) of Vent. 56, Pr. 1.24, and Pr. 26.42. Such a possibility,
however, is no reason to alter my view of the priority of Vent. 56.
203

204 robert mayhew
treating the parts of certain artifacts as models of, or analogous to, parts of
the human body.
Theophrastus begins De ventis with the claim that “The nature of the
winds [sc. generally] . . . has been considered earlier; but it is [now]
necessary to attempt to explain why for each [of the winds] the
capacities and in general the accompanying attributes accompany it
according to reason (αἱ δυνάμεις καὶ ὅλως τὰ παρακολουθοῦντα κατὰ
λόγον ἀκολουθεῖ).” One attribute used to distinguish the different
winds is their effect on plants and animals.4 This is why
Theophrastus ends chapter 1 (his introduction to the treatise):
“Indeed, speaking generally, the inquiries turn out to be into these
issues [e.g. temperature, strength, frequency], and concern those in
which animals and plants are included as well.”
Theophrastus discusses plants in Vent. 13, 14, 38, 43, 45, and 58. Wind
and plants are also sometimes discussed in his botanical works: for
instance, CP 5.12.4–11 is a lengthy account of the negative effect of cold
wind on trees. As for animals, the only kind appearing in De ventis are
humans.5 Vent. 13–14 discuss the effect of winds on crops and so deal
directly with plants and indirectly with human life. Vent. 53 mentions
waves, waterspouts, and (as a consequence) the destruction of ships. There
are other references to nautical proverbs and/or ships as well (see Vent. 5, 6,
28, 33, 46, 49, 50, 51). It is in Vent. 56–58,6 however, that Theophrastus
deals explicitly with the effects of winds on the human condition.
The opening line of Vent. 56 marks a transition between two sections:
Vent. 46–55, in which Theophrastus discusses assorted common opinions
concerning phenomena occurring in “the air and the sky as a whole” (περὶ
τὸν ἀέρα καὶ τὸν ὅλον οὐρανόν); and Vent. 56–58, in which (as just noted)
the focus is on the effect of winds (especially Notos)7 on humans (on the
ground, so to speak). First, he discusses the effects of Notos on the
functioning of the joints in the human body, and the consequences thereof
(56); next, he explains why Notos is (under certain conditions) fever-
producing (57); and finally, he mentions the effects of winds on crops,
4
On capacities and other accompanying attributes, and on the effects on plants and animals being
accompanying attributes that are not capacities, see Mayhew 2018, 83–85.
5
De ventis likely is unfinished or has survived incomplete. Theophrastus may have discussed other
animals (or planned to), especially domesticated ones (e.g. livestock and poultry).
6
As I point out from time to time in my commentary (e.g. Mayhew 2018, 128, on Vent. 5), sometimes
Schneider’s chapter divisions (see above, note 1) make little sense. Here the division is quite sensible,
however, even though Vent. 56–58 clearly form a unit.
7
Boreas and Notos are Theophrastus’ paradigmatic winds, and so it is in this section. In Vent. 56 and
58, he contrasts Notos and Boreas (though the emphasis is on Notos); in 57, he discusses Notos alone.

The Ill Effect of South Winds 205
on certain artifacts, and on iron working (58).8 My interest here is in Vent.
56. But to understand this chapter, it is necessary to look first at some key
summarizing statements in Vent. 58. And before doing that, I need to say
a bit more about how Theophrastus conceives of Notos.9
Notos is the south wind. It is directly south on the wind rose and
opposite Boreas. Boreas comes from the Arctic (Vent. 49), but
Theophrastus does not say where precisely in the south Notos origin-
ates. As I argue elsewhere, however (Mayhew 2018, 106–7), he likely
held (as Aristotle did) that Notos blows “from the summer tropic”
(ἀπὸ τῆς θερινῆς τροπῆς, Mete. 2.5.362a31) – what we call the Tropic of
Cancer – and specifically from the mountains there (Aethiopian moun-
tains in Libya, Silver mountains in Egypt).10 What follows are its
general attributes, though Theophrastus makes it clear that these are
all variable, depending on location, season, and other factors: Notos is
a powerful wind that blows frequently (Vent. 2) – specifically, it blows
throughout winter and at the beginning of spring and the end of
autumn (10). Although it is clear in the south (6), it brings rain to
the north (4 and 54), which includes Greece. And as it blows from the
south, it is in general a hot wind (3).
To understand how Notos affects the human body, we must understand
how according to Theophrastus winds affect the world generally. And in
this respect, Vent. 58 is quite informative, where he summarizes Vent. 56–57
before turning to crops and artifacts and other human undertakings.11 All
such phenomena as he discusses in Vent. 56–57, he says – that is, the effects
of various winds on humans and other living things – “will be reduced to
(εἰς . . . ἀναχθήσεται) [i.e. explained in terms of] moisture and dissolution,
and density and solidification, and however many other pairs there are.”
The main pairs of opposite characteristics, and their corollary processes,
are: hot and cold, moist and dry, dense and “rare”; and, heating and
cooling, moistening and drying, solidifying and dissolving. “And it is
similar,” he says, “in the case of inanimate things, for instance the breaking
of gut strings and the noises of things glued together and whatever else
happens when objects are moistened and loosened.” And presumably other
8
For my commentary on Vent. 56–58, see Mayhew 2018, 330–40. I now regret not saying more there
about Vent. 56 and am glad to have the opportunity to do so here.
9
I do not have the space (but neither is it necessary) to describe Theophrastus’ account of the natural
processes that give rise to Notos winds. See, however, Vent. 2 and Mayhew 2018, 103–9.
10
See Arist. Mete. 1.13.350b10–14, and compare Vent. 5.
11
The text of Vent. 58 is corrupt and lacunose and has been much emended by editors. I cannot discuss
textual issues here. For a defense of my reading of the text, see Mayhew 2018, 336–40.

206 robert mayhew
effects will be explained in terms of opposite characteristics and processes,
that is, when things are dried and solidified. (He focuses on moistening and
dissolution here, no doubt because Notos is the primary wind of interest in
these chapters, with Boreas mentioned from time to time by way of
contrast.) So in the case of iron working (περὶ τὴν τοῦ σιδήρου
κατεργασίαν), he writes:
For they say they beat out [iron] better with Notos winds than with Boreas.
And the reason is that Boreas winds dry and harden, whereas Notos winds
moisten and relax; and everything is easier to work when made spreadable
than when hardened. But at the same time too, [smiths] are stronger and
[work] more intensely in Boreas winds.
As will become clear shortly, it is important in moving forward that we
know (to the extent possible) what Theophrastus means by “the noises of
things glued together” (οἱ ψόφοι τῶν κεκολλημένων).12 In Vent. 58, he is
referring to inanimate things; but which ones? Although κολλάω comes
from κόλλα (“glue, paste”), and so its original meaning was “to glue,” it
eventually came to refer to any sort of fastening.13 So τῶν κεκολλημένων
could refer to any objects consisting of parts fastened together or securely
put together. Nevertheless, judging by Theophrastus’ comments in
Historia plantarum, I think he is likely referring to things held together
(at least in part) by glue; and specifically, to the wooden parts of a house
(frames, doors, and windows) as well as certain (probably large) pieces of
furniture.14 (More specifically, I assume he has in mind the use of glue in
forming a mortise and tenon joint.) He writes (5.6.2):
With a view to the needs of carpenters, pine holds glue15 best (πρὸς δὲ τὰς
τῶν τεκτόνων χρείας ἐχέκολλον μὲν μάλιστα ἡ πεύκη) because of its loose
texture and straight grain; for they say that it never breaks apart when it is
glued (ἐὰν κολληθῇ).
Theophrastus claims that in housebuilding, drier wood is better at least
when gluing is employed (πρός γε τὴν κόλλησιν ἡ ξηροτέρα συμφέρει)
(5.7.4); and in the case of gluing (κατὰ τὴν κόλλησιν), oak does not attach
12
The anonymous De signis, at one time attributed to Theophrastus, states (30.207–9): “if when there
are Notos winds there is a noise from things glued together (ψοφῇ τῶν κεκολλημένων),” and so on.
The author or compiler of De signis may have taken this from Vent. 58. On the origin and authorship
of De signis, see Sider and Brunschön 2007, 4–5 and 40–43.
13
See Liddell, Scott, and Jones 1996 and Danker 2000 s.v. κολλάω.
14
See Sider and Brunschön 2007, 161–62.
15
The adjective I translate as “holds glue” is ἐχέκολλος (ἔχω + κόλλα), which according to LSJ and
BDAG means “glutinous, sticky.” But LSJ adds: “ἐχέκολλον μάλιστα ἡ πεύκη takes glue best, Thphr.
HP 5.6.2.”

well to pine or fir (5.7.2). Among the wooden artifacts he mentions are
beds, chairs, tables, chests, doors, and looms (3.10.1, 5.7.6).
So, I conclude that the noises of things glued together are the creaking of
large pieces of furniture (beds, chests, looms), as well as the wooden parts of
the house as it “settles.” And Theophrastus takes the cause of these noises to
be Notos winds, as the humidity they create moistens and loosens the
joints where the pieces of wood are attached by glue, and so loosened they
creak (which does not occur when the joints are secure).16 Now, irritating
noises are hardly much of a negative effect on the human condition. But as
De ventis is an unpolished and possibly incomplete work, and perhaps
originally a set of lecture notes (which would not have contained all of
Theophrastus’ elaborations),17 I think that “noises of things glued
together” was likely meant to point to a cluster of related, and often
more irritating, outcomes: for instance, stuck doors and windows, warped
wooden objects, loose table and chair legs, and so on.
Theophrastus is not here treating the human body as a machine or artifact,
nor do I think he is engaging in analogical reasoning. Rather, he maintains
that the same fundamental material principles explain phenomena in cases
both of living beings and of inanimate things. This invites someone with the
same basic outlook to treat an artifact as a model or analogy of the human
body. And I believe this is what the author of Pr. 1.24 does, however tersely.
Armed with this information from Vent. 58, we can turn to Vent. 56 and
better understand what Theophrastus says there about the ill effects of
Notos winds on the human body.
I divide Vent. 56 into four parts (of one sentence each, as I punctuate the
text). The first is the opening (transitional) sentence, mentioned above
(p. 204). The second gives an example of a wind (i.e. Notos) directly affecting
the human condition, and one reason for it; the third provides a further reason
for this effect; the fourth describes the effects of Boreas, by way of contrast.
Here is the second part:
οἷον ⟨βαρύ⟩τερον ἐν τοῖς νοτίοις ἔχουσιν ⟨οἱ⟩ ἄνθρωποι καὶ ἀδυνατώτερον·
αἴτιον δ’ ὅτι ἐξ ⟨ὀλίγου⟩ πολὺ ὑγρὸν γίνεται καὶ [ἡ] ὑγρότης βαρεῖα ἀντὶ
κούφου πνεύματος.18
16
That the cause of such creaking was an issue of interest among Peripatetics, note [Arist.] Pr. 11.28: “Why
do some things, such as chests, make a noise (ψοφεῖ) and move suddenly, when nothing perceptible
moves them?” The lengthy response, however, has nothing to do with Notos or humidity.
17
See Mayhew 2018, 18.
18
Pr. 1.24.862a27–28 (cf. 26.42.945a14–15) is useful in emending the text in two places: changing the
manuscripts’ ἕτερον into the much better βαρύτερον, and filling the lacuna in the manuscript
following ὅτι ἐξ with ὀλίγου.

208 robert mayhew
For instance, people feel heavier and more powerless in Notos winds; and
the reason is that a great deal of moisture comes out of ⟨a small amount⟩, and
heavy moisture comes instead of light wind.19
This states what Notos does – makes “people feel heavier and more
powerless” – and one reason for it: the additional moisture it brings
(which is more than one would expect from what seems to be a light
wind). Instead of a balanced state of moist and dry, there is excessive
moisture. (There is no mention of excessive heat, but that may be implied.)
This is a clear reference to a high level of humidity, which is experienced as
oppressive and quite capable of making people feel sluggish.
So Theophrastus seems to be saying that a great deal of humidity comes
out of a small amount of moisture. Given that he says Notos is clear in the
south but brings moisture to those in the north (see Vent. 4, 6, and 54),
I assume he is referring to the amount of moisture gathered by Notos as it
travels north, which – however much it is – one would not expect to
produce so much humidity. In the same way, it seems, heavy moisture (i.e.
oppressive humidity) comes out of what is, or is felt to be, a slight wind – as
opposed, I take it, to a powerful, stormy, cloud-bearing wind, which one
expects to bring and produce a great deal of moisture.
The first half of Pr. 1.24 (862a27–30) presents this same material as
a problem to be solved.
Διὰ τί ἐν τοῖς νοτίοις βαρύτερον ἔχουσι καὶ ἀδυνατώτερον οἱ ἄνθρωποι; ἢ
ὅτι ἐξ ὀλίγου πολὺ ὑγρὸν γίνεται διατηκόμενον διὰ τὴν ἀλέαν, καὶ ἐκ
πνεύματος κούφου ὑγρὸν βαρύ;
Why do people feel heavier and more powerless in the Notos winds? Is it
because a great deal of moisture, melted by the warmth, comes out of a small
amount, and heavy moisture comes out of light wind?
Not only is this passage useful in emending the second part of Vent. 56
(405–7), it also briefly refers to a process not (explicitly) mentioned there:
“being melted by the warmth” (διατηκόμενον διὰ τὴν ἀλέαν, 1.24.867a29;
cf. 26.42.945a42.16), which I assume refers to evaporation. This either was
in Vent. 56 as well but dropped out (like ὀλίγου and part of βαρύτερον), or
it was added by the author of Pr. 1.24. Again, I assume (though this is not
entirely clear) that the author is referring to warm Notos winds causing
evaporation and accumulating moisture as they travel north.
19
On heaviness and fatigue, see Theophrastus Lass. 3.21–23, quoted below (note 30). And see Lass. 4–5
on the connection between fatigue and “an abundance of moisture” (πλῆθος ὑγρότητος, 4.31).

In the present context – examining ancient authors on the mechanics of
the human body – the third part of Vent. 56 (407–9), along with the second
part of Pr. 1.24, is much more interesting than the rest of the chapter. But it
is also even more corrupt textually. Here is the line that concerns me (the
entire third part), exactly as found in Vaticanus gr. 1302:20
ἔτι δ’ ἡ μὲν ἰσχὺς καὶ δύναμις ἐν τοῖς ἀθρόοις τι· ἔπειτα μέντι κινεῖ λίαν δ’
ὑγρὰ κειμένη συντίθεσθαι·
Editors have relied on the second half of Pr. 1.24 (862a30-33) to repair this
text:
ἔτι δ’ ἡ δύναμις ἡμῶν ἐν τοῖς ἄρθροις ἐστί, ταῦτα δ’ ἀνίεται ὑπὸ τῶν
νοτίων. δηλοῦσι δ’ οἱ ψόφοι τῶν κεκολλημένων. τὸ γὰρ γλίσχρον ἐν τοῖς
ἄρθροις πεπηγὸς μὲν κινεῖσθαι κωλύει ἡμᾶς, ὑγρὸν δὲ λίαν ὂν συντείνεσθαι.
Further, our power is in the joints, and these are made slack by Notos winds.
Now the noises of things glued together make this clear. For the viscous
material21 in the joints, having been solidified, prevents us from moving, but
being too moist [it prevents us from] exerting ourselves.
I offer possible interpretations of this passage in due course. But first I must
attempt to make sense of the third part of Vent. 56. Furlanus 1605,
Schneider 1818, Wimmer 1862 and 1866, Coutant and Eichenlaub 1975,
and Gigon 193722 all made ample use of Pr. 1.24 in editing it.23 What
follows is the text so edited.24
ἔτι δ’ ἡ μὲν ἰσχὺς καὶ δύναμις ἐν τοῖς ἄρθροις ⟨ἐσ⟩τί· [ἔπειτα μέντι] ⟨ταῦτα δὲ
ἀνίεται ὑπὸ τῶν νοτίων. δηλοῦσι δ’ οἱ ψόφοι τῶν κεκολλημένων· τὸ γὰρ
γλίσχρον ἐν τοῖς ἄρθροις πεπηγὸς μὲν κωλύει⟩ κινεῖ⟨σθαι ἡμᾶς⟩, ὑγρὸν δὲ
λίαν [κειμένη] ⟨ὂν⟩ συντείνεσθαι·25
Further, strength and power are in the joints, and these are made slack by
Notos winds. Now the noises of things glued together makes this clear; for the
20
This is the manuscript on which all the others depend (see Burnikel 1974).
21
On τὸ γλίσχρον, see below p. 213, with note 32.
22
On Gigon 1937, long thought to be lost, see Mayhew 2018, 9–10.
23
Many of these emendations, though not the inserted material, were first suggested by the sixteenth-
century French classicist Turnebus, in his marginalia in a copy of Vascosanus 1551 now in the Leiden
University Library (call no. 757 D 32:2). See Mayhew 2018, 8–9.
24
Pointed brackets indicate material inserted from Pr. 1.24, italics words emended based on Pr. 1.24,
square brackets words in Vaticanus gr. 1302 that were omitted. Such indications are often absent,
unclear, or inaccurate in the editions of the aforementioned scholars.
25
There are two variations among these editions: Wimmer printed οὐ in place of ὂν, and only Gigon
included δηλοῦσι δ’ οἱ ψόφοι τῶν κεκολλημένων in the long insertion (which I have italicized in the
translation that follows).

210 robert mayhew
viscous material in the joints, having been solidified, prevents us from moving,
but being too moist [it prevents us from] exerting ourselves.
In editing Vent. 56, I found this approach to our line too intrusive (though
not entirely implausible). After all, Theophrastus could here have made
some general claim about power being in our joints – the text now mangled
and/or lacunose – and moisture negatively affecting them, with the author of
Pr. 1.24 later making use of, and expanding on, Vent. 56, offering a more
specific account of Notos’ negative effect on our joints. So although I too
have relied on Pr. 1.24 in editing Vent. 56,26 I stick much closer to the
manuscript reading, choosing to emend the words ἔπειτα μέντι rather than
obelizing or replacing them.27 There is in fact a comparatively easy fix – by
rearticulating the letters and inserting one word – though the result is rather
inelegant Greek (not that that is a rarity in De ventis). I have changed ἔπειτα
μέντι to ἐπεί τὰ μέν τι,28 and inserted ὄντα where Furlanus et al. (following
Pr. 1.24) print ὂν. So here is my best attempt at understanding the line:
ἔτι δ’ ἡ μὲν ἰσχὺς καὶ δύναμις ἐν τοῖς ἄρθροις ⟨ἐσ⟩τί, ἐπεί τὰ (sc. ἄρθρα) μέν
τι κινεῖ· λίαν δ’ ὑγρὰ ⟨ὄντα⟩ κωλύεται συντείνεσθαι.
Further, strength and power are in the joints, since the [joints] set something
in motion; but ⟨being⟩ too moist they prevent us from exerting ourselves.29
This has the advantage of being closer to the manuscript tradition (though any
attempt to make sense of the line remains speculative). In addition, it fits my
above speculations about the relationship between Vent. 56 and Pr. 1.24. That
is, it leaves open the possibility (to my mind highly likely) that the author of
Pr. 1.24 was expanding on Vent. 56. Theophrastus claims that Notos makes us
feel less powerful by making our joints too moist, without explaining how
overly moist joints prevent us from exerting ourselves – other than what is
implied from Vent. 58, namely, that this loosens our joints.30 The author of Pr.
26
Not unlike Furlanus et al., I made these two textual emendations (first proposed by Turnebus) based
on Pr. 1.24: ἄρθροις ἐστί for ἀθρόοις τι, and κωλύεται συντείνεσθαι for κειμένη συντίθεσθαι.
27
In Mayhew 2018, 62–63, I obelized ἔπειτα μέντι κινεῖ and marked a lacuna in my translation. I have
since come up with a better (however imperfect) solution. I would like to thank David Sider for help
with this text (though he should not be held responsible for any lingering problems).
28
Turnebus (see note 23) first suggested ἐπεὶ τὰ for ἔπειτα, followed by μέντοι. μέντοι is the reading of
Mediol. Ambrosianus P 80 sup. and other manuscripts, and of the Aldine and most early editions –
including Vascosanus 1551, which Turnebus followed in the case of μέντοι. That is, he left it
untouched in making his handwritten comments on that edition.
29
This is a more accurate rendering of λίαν ὑγρά than the one in Mayhew 2018, 63.
30
This is consistent with what Theophrastus says in his De lassitudine about fatigue and our joints. See,
for example, Lass. 3.21–23: “To speak simply, fatigue occurs in the bent parts (ἐν τοῖς καμπτομένοις
μέρεσι) especially and the most sinewy ones, whenever some liquefaction comes to the sinew and the

1.24 expands on this somewhat, offering a fuller (and perhaps quite different)
explanation. I return to Pr. 1.24 shortly; but I first must look at the last part of
Vent. 56.
If Notos winds slacken our joints, which (somehow) precludes exertion,
under what conditions (and specifically with what wind) does the human
body function best? Theophrastus tells us in the fourth part of Vent. 56:
Boreas winds, however, produce a certain proportion (τινὰ συμμετρίαν),
such that we are strong and exert ourselves (συντείνεσθαι) more.
Boreas is (normally) a cold and dry wind (Vent. 57), which dries and
hardens things (58). I assume Theophrastus is claiming that under normal
conditions, the cold and dry Boreas winds, balanced against the warm
climate of Greece (that is the “certain proportion”), produce a moderate
temperature with no extraneous moisture. Our joints are thus strong
enough for us to move well and so exert ourselves, but not so stiff as to
undercut mobility.
I want now to look more closely at Pr. 1.24.862a30–33 and how it differs
from Vent. 56. A few of the differences are unimportant: Vent. 56 refers to
strength and power, Pr. 1.24 to power alone; Vent. 56 says the joints set
something in motion, Pr. 1.24 merely implies this; Pr. 1.24 says the joints
are made slack by Notos winds, Vent. 56 merely implies this. I am inter-
ested in the significant differences.
First, whereas Vent. 56 considers two possibilities – Notos negatively
affecting our joints, Boreas allowing them to function properly – Pr. 1.24
describes two negative possibilities (omitting or leaving to implication
a positive one): In one negative possibility (the one that is the focus of
the problêma), Notos winds cause the viscous material around the joints to
become too moist, in which case they provide insufficient support, such
that we cannot exert ourselves. In the other, opposite negative possibility,
the viscous material around the joints is solidified, causing immobility.
(No wind is mentioned, but the cause is likely any extremely dry and cold
climate.) The author does not tell us what wind has a positive effect on our
joints, but it must be one that contributes to conditions that fall some-
where between these extremes. (Boreas is a good candidate, for the reasons
given in Vent. 56.)
bends (τὰς καμπάς, sc. of the limbs). And heaviness (ἡ βαρύτης) results from this.” (Text and
translation from Sollenberger 2003, translation slightly modified.) Compare καμπή, “bend” (in Lass.
passim) and ἄρθρον, “joint” (in Vent. 56 and Pr. 1.24), though they may well be synonyms. See note
19 above. There is a connection between joints or bends (καμπαί) and fatigue in [Arist.] Pr. 5.5, 35,
and 40 as well.

212 robert mayhew
Two other significant differences are that Vent. 56 says simply that our
joints are affected, whereas Pr. 1.24 refers specifically to the viscous material
(τὸ γλίσχρον) in our joints; and Pr. 1.24 refers to the noises of things glued
together (οἱ ψόφοι τῶν κεκολλημένων) – likely taken from Vent. 58 –
whereas Vent. 56 does not. A third significant difference is that the author
of Pr. 1.24 employs analogical reasoning. But let us first consider the noises
of things glued together.
I can think of a few ways of interpreting οἱ ψόφοι τῶν κεκολλημένων
here. First, one could hold that, unlike Vent. 58 (where this is a clear
reference to artifacts), the author of Pr. 1.24 is instead referring to the
noises of the connected parts of the human body, that is, the joints. Some
scholars have held this view.31 Although this is not impossible, it is hard to
make sense of it. For how would noises in our connected parts (e.g. cracked
knuckles) make clear (δηλοῦσι) that the power is in our joints and that
Notos makes them slack?
Alternatively, one could conclude that in Pr. 1.24 the reference is the
same as in Vent. 58: wooden artifacts, like furniture (only in Pr. 1.24 this is
an analogy for the joints in our body). On this interpretation, there are two
extremes that are not good: when the viscous material joining parts
together (presumably glue) has solidified and so prevents movement; and
when this viscous material has become so moist that the connection loses
its strength and support. But then this is a bad analogy, as a fixed piece of
furniture, like a bed, does not well illustrate what the author of Pr. 1.24
seems to be claiming: for the optimal condition in the case of a bed or table
would be precisely that the glue be solidified and hold the parts together
firmly (such that there is no creaking), not that it be neither too solid nor
too loose.
In light of this, one may want to argue that the line about things glued
together is a limited, parenthetical remark, intending to show only that
a high level of humidity can loosen joints, but that it was not meant as an
analogy with noises in the joints in the human body. The γάρ that follows
makes this unlikely, however. One might also argue that δηλοῦσι οἱ ψόφοι
τῶν κεκολλημένων was originally a marginal gloss, which a scribe or scholar
copied from Vent. 58, and so there was in the original passage no reference
to furniture or any other artifact (and consequently no analogy). That
would certainly solve our interpretive problems. In fact, I assume this
31
The clearest case is Barthélémy-Saint Hilaire 1891, 23, who translates the line On peut s’en convaincre
par les craquements des pièces du corps soudées ensemble, and in a footnote refers to the crepitations in
our knees and knuckles. Compare the translations of Louis 1991, 19 and Marenghi 1999, 57.

apparent lack of fit explains why Furlanus, Schneider, Wimmer, and
Coutant (but not Gigon) left out δηλοῦσι δ’ οἱ ψόφοι τῶν κεκολλημένων
when they inserted material from Pr. 1.24 into Vent. 56. But this strikes me
as a rather artificial move.
I can think of one way to salvage the analogy, though it is highly
speculative (and not without problems of its own). It requires narrowing
the analogous furniture, and reconsidering to what precisely τὸ γλίσχρον
refers (which might reasonably be taken to refer to glue, given its close
proximity to “the noises of things glued together” at Pr. 1.24.862a31–32).
This parts-of-furniture–joints-in-the-body analogy would work better if the
author of Pr. 1.24 had in mind furniture and sections of a house with moving
parts: for instance, looms and doors. In this case, the optimal condition is
somewhere between parts being solidified and thus becoming stuck, and
parts being so loosely or tenuously connected that the loom or door or such
like does not function properly. This analogy fails, however, and for the
reasons given earlier, if τὸ γλίσχρον refers to glue. But the analogy might
succeed if τὸ γλίσχρον refers instead to some lubricating material (oil or
grease) used to unstick a door or make a loom function better. The basic
meaning of γλίσχρος (LSJ and BDAG s.v.) is “sticky,” “gluey,” “viscous” –
and this last is the closest to what Aristotle describes in Meteorology 4.32 In
Aristotle’s biology, it generally means “viscous,” “glutinous,” “slimy”: for
instance, in hairy animals, the male contribution to generation (σπέρμα) is
described as γλίσχρος (HA 3.22.523a15–17); and parts of the cuttlefish are
“slimy to the touch” (τὴν ἁφὴν γλίσχραν) (HA 4.2.527a24–28). Further,
Aristotle actually refers to something that sounds like an excellent analogue
to this material in the human body: At HA 3.5.515b10–18, he describes the
liquid around the sinew as white and gluey (κολλώδης). So in both the
moving parts of wooden artifacts and the joints in human bodies, if this
material solidifies, the parts or joints cannot function. And if the viscosity of
this material becomes too low – as would happen if Notos winds blow and
raise the level of humidity – then the parts or joints cannot function
properly. (I think this is the weakest part of the analogy.) But under optimal
conditions, the human body is a well-oiled machine.33
32
See Mete. 4.7.383b19–384a2, 4.8.385a10–18, 4.8.385b1–5, 4.9.387a11–15, and Tassios 2018, 152–53
and 158.
33
Compare Lass. 18, which takes the form of an Aristotelian problêma. It begins (139–43): “Why are
moist bodies more subject to fatigue (τὰ ὑγρὰ σώματα κοπιαρώτερα) than those which are not
moist? For they are weaker (ἀσθενέστερα). The hard body is stronger and less apt to be affected, not
in relation to every absence of injury . . . but <in relation to the> more viscous things (⟨πρὸς τὰ⟩
γλισχρότερα), some of which are soft and moist in nature.” (Sollenberger’s text and translation,
2003, 276–77, translation slightly revised.)

214 robert mayhew
I confess to not having a strong opinion about which interpretation of
Pr. 1.24 is correct.
To conclude: For Theophrastus, the same basic principles of natural
philosophy are at work in explaining the functioning of both inanimate
objects and the human body. This is explicit in Vent. 58 but left to implica-
tion in Vent. 56. The author of Pr. 1.24, however, in attempting to explain
why people feel heavier and less powerful under the influence of Notos
winds, makes this more explicit (on most interpretations), relying on what
Theophrastus said in Vent. 58: Wooden artifacts with attached parts and the
human body and its joints are analogous, so certain facts about the attached
parts of wooden artifacts apply (in some way) to the joints in our bodies.
Colin Webster, in the conclusion to his study of the use of technological
artifacts as analogies of natural processes in ancient thought, writes (2014, 273):
In many ways, I have been describing a theory of cognition that takes material
tools to be cognitive tools. Perhaps we have a box full of such tools, but they
work in different, sometime opposed ways, and depending on what we are
attempting to explain, we reach for one, then another, whenever we see fit.
I think it likely that the author of Pr. 1.24 took a cognitive tool he found in
Theophrastus and used it in a different way, employing analogical reason-
ing where Theophrastus did not. It should go without saying, however,
that there is no suggestion that the author of Pr. 1.24 thought that this
implied a mechanistic conception of the human body in some sense at odds
with an orthodox Aristotelian teleological one.34
Appendix: The Texts of Vent. 56 and Pr. 1.24
Vent. 56
[ἢ] ταῦτα μὲν οὖν καὶ ὅσα ἄλλα τοιαῦτα περὶ τὸν ἀέρα καὶ τὸν ὅλον
οὐρανὸν συμβαίνει, τάδε ⟨δέ τινα⟩ εἰς τὰς ἡμετέρας διαθέσεις. οἷον ⟨βαρύ⟩τερον 405
ἐν τοῖς νοτίοις ἔχουσιν ἄνθρωποι καὶ ἀδυνατώτερον· αἴτιον δ’ ὅτι ἐξ ⟨ὀλίγου⟩
πολὺ ὑγρὸν γίνεται καὶ [ἡ] ὑγρότης βαρεῖα ἀντὶ κούφου πνεύματος. ἔτι δ’ ἡ
34
This should be clear from the fact that Aristotle was quite capable of employing analogical
reasoning. See Korobili in the present volume, Chapter 7. To give another example, in his De
memoria Aristotle writes: “for the change occurring makes an imprint, like some stamp, of the thing
perceived, just like those sealing things with signet rings” (ἡ γὰρ γιγνομένη κίνησις ἐνσημαίνεται
οἷον τύπον τινὰ τοῦ αἰσθήματος, καθάπερ οἱ σφραγιζόμενοι τοῖς δακτυλίοις) (450a30–32). See also
Lloyd 1966, 374–5 and Webster 2014, 157–170. I want to thank Maria Gerolemou for gently pushing
me to say a bit more about the analogical reasoning in Pr. 1.24. (Any errors resulting from my
expanding on this topic are solely my own.)

μὲν ἰσχὺς καὶ δύναμις ἐν τοῖς ἄρθροις ⟨ἐσ⟩τί, ἐπεί τὰ μέν τι κινεῖ· λίαν δ’ ὑγρὰ ⟨ὄντα⟩
κωλύεται συντείνεσθαι. τὰ δὲ βόρεια ποιεῖ τινὰ συμμετρίαν, ὥστε καὶ ἰσχύειν
καὶ συντείνεσθαι μᾶλλον. 35 410
404 ἢ A, secl. Turn. 405 ⟨δέ τινα⟩ scripsi : lac. 6 litt. A || βαρύτερον Turn. :
ἕτερον A 406 ἄνθρωποι Turn. : ἀνών A (nom. sacr. = ἀνθρώπων) || αἴτιον
Turn. : ἄρτιον A || ὀλίγου Turn. : lac. 9 litt. Α 407 post γίνεται add.
διατηκόμενον διὰ τὴν ἀλέαν Bon. ex Pr. 1.24 || [ἡ] ὑγρότης βαρεῖα ἀντὶ
κούφου πνεύματος Turn. : ἡ πνεύματος ὑγρότης βαρέα ἀντὶ κούφου A 408
ἄρθροις ἐστί Turn. : ἀθρόοις τι Α : post ἐστί add. ταῦτα δ’ ἀνίεται ὑπὸ τῶν
νοτίων· τὸ γὰρ γλίσχρον ἐν τοῖς ἄρθροις πεπηγὸς μὲν κωλύει κινεῖσθαι
ἡμᾶς ὑγρὸν δὲ λίαν ὂν συντείνεσθαι Furl. ex Pr. 1.24 (post νοτίων add.
δηλοῦσι δ’ οἱ ψόφοι τῶν κεκολλημένων Gigon) || ἐπεί τὰ μέν τι scripsi (cf.
Turn.) : ἔπειτα μέντι A : om. Furl. || ὄντα addidi 409 κωλύεται
συντείνεσθαι Turn. : κειμένη συντίθεσθαι Α || βόρεια Turn. : βόρια A ||
ἰσχύειν Turn. : ἰσχὺν A
Pr. 1.24
Διὰ τί ἐν τοῖς νοτίοις βαρύτερον ἔχουσι καὶ ἀδυνατώ-

τερον οἱ ἄνθρωποι; ἢ ὅτι ἐξ ὀλίγου πολὺ ὑγρὸν γίνεται,
διατηκόμενον διὰ τὴν ἀλέαν, καὶ ἐκ πνεύματος κούφου
ὑγρὸν βαρύ;36 ἔτι δ’ ἡ δύναμις ἡμῶν ἐν τοῖς ἄρθροις ἐστί, 862a30
ταῦτα δ’ ἀνίεται ὑπὸ τῶν νοτίων. δηλοῦσι δ’ οἱ ψόφοι τῶν
κεκολλημένων. τὸ γὰρ γλίσχρον ἐν τοῖς ἄρθροις πεπηγὸς
μὲν κινεῖσθαι κωλύει ἡμᾶς, ὑγρὸν δὲ λίαν ὂν συντείνεσθαι. 37
862a30–32 ἐστί . . . ἄρθροις om. L (homoiotel.) 33 κινεῖσθαι κωλύει : κωλύει
κινεῖσθαι γ
35
Mayhew 2018, 62; line numbers are from this edition. The apparatus is abridged, and I have made
changes to the text in lines 406 and (especially) 408. “A” refers to ms. A (Vaticanus gr. 1302 – see
above note 20); “Turn.” refers to Turnebus (see above note 23); “Bon.” refers to Bonaventura 1593;
“Furl.” refers to Furlanus 1605.
36
Cf. Pr. 26.42 (which I present here in its entirety): Διὰ τί ἐν τοῖς νοτίοις βαρύτερον ἔχουσι καὶ
ἀδυνατώτερον οἱ ἄνθρωποι; ἢ διότι ἐξ ὀλίγου πολὺ ὑγρὸν γίνεται διατηκόμενον διὰ τὴν ἀλέαν, καὶ
ἐκ πνεύματος κούφου ὑγρὸν βαρύ; εἶτα ἡ δύναμις ἀτονεῖ. (Text: Mayhew 2011, 2: 196.) Nothing in
this problêma corresponds to the second part of Pr. 1.24.
37
Marenghi 1999, 56; Bekker line numbers. γ is one of four families of manuscripts; L is a minor
manuscript in family δ. (See Marenghi 1999, 21–22 for details.)

216 robert mayhew
RE F E RE N CE S
Barthélémy-Saint Hilaire, J. 1891. Les Problèmes d’Aristote, vol. 1 (Paris).
Bodnár, I. 2015. “The Problemata physica: An Ιntroduction,” in R. Mayhew, ed.,
The Aristotelian Problemata physica: Philosophical and Scientific Investigations
(Leiden), 1–9.
Bonaventura, F. 1593. Anemologiae pars prior, id est De affectionibus, signis,
causisque ventorum ex Aristotele, Theophrasto, ac Ptolomaeo tractatus
(Urbino).
Burnikel, W. 1974. Textgeschichtliche Untersuchungen Zu Neun Opuscula
Theophrasts (Wiesbaden)
Coutant, V. and V. L. Eichenlaub. 1975. Theophrastus: De Ventis (South
Bend, IN).
Danker, F. W., ed. 2000. A Greek–English Lexicon of the New Testament and other
Early Christian Literature, 3rd ed. (Chicago).
Flashar, H. 1962. Aristoteles: Problemata Physica (Berlin).
Furlanus, D. 1605. Theophrasti Eresii: Peripateticorum post Aristotelem Principis
(Hanover).
Gigon, O. 1937. “Theophrastos Über die Winde, Text, Kommentar und
Einleitung.” Diss. University of Basel.
Liddell, H. G., R. Scott, and H. S. Jones. 1996. Greek–English Lexicon. Revised
Supplement (Oxford).
Louis, P. 1991. Aristote: Problèmes, vol. 1: Sections i–x (Paris).
Marenghi, G. 1999. Aristotele: Problemi di Medicina: Testo Critico, Traduzione
e Commento, 2nd ed. (Milan).
Mayhew, R. 2011. Aristotle: Problems: vol. 1 (Books 1–19), vol. 2 (Books 20–38, with
D. Mirhady, [Arist.] Rhet. to Alex.) (Cambridge, MA).
2015. “Problemata 26 and Theophrastus’ De ventis: A Preliminary Comparison,”
in R. Mayhew, ed., The Aristotelian Problemata physica: Philosophical and
Scientific Investigations (Leiden), 294–310.
2018. Theophrastus of Eresus: On Winds (Leiden).
Schneider, I. G. 1818. Theophrasti Eresii quae supersunt opera et excerpta librorum,
vol. 1 (Leipzig).
Sider, D. and C. W. Brunschön. 2007. Theophrastus of Eresus: On Weather Signs
(Leiden).
Sollenberger, M. G. 2003. “Theophrastus of Eresus, On Fatigue,” in
W. W. Fortenbaugh, R. W. Sharples, and M. G. Sollenberger. eds.,
Theophrastus of Eresus, On Sweat, On Dizziness and On Fatigue (Leiden), 251–324.
Tassios, T. P. 2018. “Mechanical Properties of Solids in Aristotle’s Meteorologica,”
in D. Sfendoni-Mentzou, ed., Aristotle: Contemporary Perspectives on His
Thought (Berlin) 151–64.

Vascosanus, M. 1551, Teophrasti De Ventis (Paris).
Webster, C. 2014. “Technology and/as Theory: Material Thinking in Ancient
Science and Medicine.” Dissertation, Columbia University.
Wimmer, F. 1862. Theophrasti Eresii opera quae supersunt omnia, vol. 3 (Leipzig).
1866. Theophrasti Eresii opera quae supersunt Omnia (Paris).

chapter 9
The Beauty That Lies Within

Anatomy, Mechanics and Thauma in Hellenistic Medicine
George Kazantzidis
Introduction
The great anatomical discoveries of Herophilus and Erasistratus are counted
among the most distinctive features of Hellenistic medicine. These discov-
eries go hand in hand with an increasing assimilation, attested during this
period, between parts of the human body and mechanical devices. While this
mechanical model has been thoroughly discussed in scholarship, the
emphasis has been usually placed on the interaction between the fields of
medicine and mechanics in Ptolemaic Alexandria and the ways in which this
interaction helped doctors to understand the function and the properties of
the human body better. In this chapter, I will deal with a different, though
closely related, question: I will set out to examine the extent to which the
discovery of little ‘machines’ and ‘sub-machines’ operating within the body
is also significant on an aesthetic level. Aristotle, as we shall see, claims that
the interior of the human body looks messy and disgusting; still, as soon as
a bodily organ is found to serve a specific purpose (assigned to it by Nature), it
immediately claims a place in the realm of the beautiful. By focusing on the
case of Erasistratus, I will argue that Hellenistic medicine nourishes
a different aesthetic model. In this case, the expression of wonder for the
artful design of the human body is not so much a matter of teleology as it is
more tightly linked to, and becomes consolidated on the basis of, figural
analogies and similarities with products of human ingenuity and craft. But
this ingenuity involves also a considerable degree of deception – one that
becomes manifest in a mêchanê’s inherent capacity to instil feelings of
bafflement and confusion. Unlike Aristotle, who proposes that the body
should be fully comprehended before we proceed to marvel at it properly,
the machine–body analogy thus reinstates a more elusive kind of wonder in
which informed admiration and a simultaneous sense of bewilderment blend
inextricably with each other.
218

The Beauty That Lies Within 219
1 The Beauty of Having a Purpose
Imagine the most beautiful of human bodies: well-shaped and smooth-
skinned, with its individual parts harmoniously placed alongside each
other and forming an impeccable whole, where nothing stands in excess
and every single detail adds its own share of perfection – a body full of kallos
and charis. And then imagine that same body stripped of its life and placed
upon a dissection table, with its thorax and abdomen sliced open and its
innards exposed.1 What lies inside is a mess,2 a world of bloody and sticky
organs that make no apparent sense. As Aristotle admits, it would indeed
be extremely difficult for someone to come across such a sight and not feel
sick to the stomach (Part. an. 645a28–30):
οὐκ ἔστι γὰρ ἄνευ πολλῆς δυσχερείας ἰδεῖν ἐξ ὧν συνέστηκε τὸ τῶν
ἀνθρώπων γένος, οἷον αἷμα, σάρκες, ὀστᾶ, φλέβες καὶ τὰ τοιαῦτα μόρια.
For it is impossible to look at that from which mankind has been consti-
tuted – blood, flesh, bones, blood vessels, and other such parts – without
considerable disgust.3
The image emerging from this passage is that of a fragmented body which
has ceased to operate as a whole and whose parts, disclosed to the naked eye
through dissection,4 stand in isolation from each other. The observer’s
response is described as duschereia, a word which involves feelings of nausea
and repugnance but is also suggestive of the kind of annoyance that we
experience when something looks puzzling and confusing.5 And yet, as
Aristotle observes in the same context, there is a way to overcome this
annoying feeling; in fact, not only overcome it but also replace it with
sentiments of immense, ‘extraordinary pleasure’ (amêchanous hêdonas, Part.
1
The tension which I am laying out here becomes especially relevant when it comes to female
cadavers exposed for dissection in the (male-dominated) anatomy theatres of early modern
Europe and beyond. As the Edinburgh anatomist Dr Robert Knox points out in his Great
Artists and Great Anatomists (published in 1852), the external form of women presents the
‘perfection of Nature’s works’; he consequently warns artists to stay out of ‘a dissecting room’
for in such a place their minds ‘may become accustomed . . . to all that is detestable; familiar
with horrors, with the emblems of destruction and death’; for a discussion of Knox’s attitude, see
MacDonald 2005, 34–5 and Hegele 2022, 93–4. Compare Malland 2022, and Park 2006, 81 (on
the female body as ‘the paradigmatic object of dissection’).
2
For the idea that the (interior of the) human body is ‘worse than dung’, a sack loaded with excrement,
which is attributed to Heraclitus, see Glucklich 2001, 27.
3
Translation in Lennox 2001, 14.
4
What Aristotle is visualizing here is a dissected human body (Lloyd 1975, 139; Lennox 2018, 260), although,
as we shall see shortly below, p. 000, he never really practised dissection on humans, only on animals.
5
For duschereia as expressive of an (instinctive) feeling of disgust, see Fisher 2017, 107 and Kazantzidis
2017, 55; for the intrinsic link between duschereia and mental confusion, see Walker 2000, 83.

220 george kazantzidis
an. 645a9).6 Once we stop looking at each bodily part individually and see
them instead as pieces of an interconnected whole,7 and once we start grasping
that each organ is meant to serve a specific function,8 then we begin to marvel
at the delicate craftsmanship of the human body.9 To assign a thing with
a purpose, Aristotle tells us, instantly allows it to enter and occupy a place in
the realm of the beautiful;10 from this, it follows that there is nothing in nature
which is not thaumasion, ‘graceful’/‘marvellous’, in some sense or another11
(provided, of course, that we are able to understand why it exists and
functions the way it does); for nature does nothing in vain.12 Beauty, then,
for Aristotle, is mainly about us making sense of the world and placing its bits
6
The translation is by Lennox 2001, 13. The locution ἀμηχάνους ἡδονὰς is extremely rare in ancient
Greek, which makes it trickier to translate. Before Aristotle, we find it only in Plato, in a passage from
Philebus discussing a mixed state of pleasure and pain (46d–e): Λέγε δὴ τὰς μέν [i.e. μίξεις], ὅταν πλείους
λῦπαι τῶν ἡδονῶν γίγνωνται – τὰς τῆς ψώρας λεγομένας νῦν δὴ ταύτας εἶναι καὶ τὰς τῶν
γαργαλισμῶν – ὁπόταν ἐντὸς τὸ ζέον ᾖ καὶ τὸ φλεγμαῖνον, τῇ τρίψει δὲ καὶ τῇ κνήσει μὴ ἐφικνῆταί
τις, τὰ δ᾿ ἐπιπολῆς μόνον διαχέῃ, τοτὲ φέροντες εἰς πῦρ αὐτὰ καὶ εἰς τοὐναντίον, ἀπορίαις
μεταβάλλοντες ἐνίοτε ἀμηχάνους ἡδονάς, τοτὲ δὲ τοὐναντίον τοῖς ἐντὸς πρὸς τὰ τῶν ἔξω λύπας
ἡδοναῖς ξυγκερασθείσας, εἰς ὁπότερ᾿ ἂν ῥέψῃ. Plato’s passage leaves us enough space to assume that
an ἀμήχανος ἡδονή is not just about a pleasure that is a felt in a deep and rewarding way; crucially, it is
also about a feeling of satisfaction which seems awkward and extraordinary precisely because it is
experienced side by side with an attendant feeling of pain and discomfort; this affective script could
fairly well be said to apply also in the case of Aristotle’s passage. On pleasures mixed with pains in Plato,
see Erginel 2019.
7
Compare, especially, Part. an. 645a30–6: Ὁμοίως τε δεῖ νομίζειν τὸν περὶ οὑτινοσοῦν τῶν μορίων ἢ
τῶν σκευῶν διαλεγόμενον μὴ περὶ τῆς ὕλης ποιεῖσθαι τὴν μνήμην, μηδὲ ταύτης χάριν, ἀλλὰ τῆς
ὅλης μορφῆς, οἷον καὶ περὶ οἰκίας, ἀλλὰ μὴ πλίνθων καὶ πηλοῦ καὶ ξύλων· καὶ τὸν περὶ φύσεως περὶ
τῆς συνθέσεως καὶ τῆς ὅλης οὐσίας, ἀλλὰ μὴ περὶ τούτων ἃ μὴ συμβαίνει χωριζόμενά ποτε τῆς
οὐσίας αὐτῶν, ‘Just as one who discusses the parts or equipment of anything should not be thought
of as doing so in order to draw attention to the matter, nor for the sake of the matter, but rather in
order to draw attention to the overall shape (e.g. to a house rather than bricks, mortar, and timbers);
likewise one should consider the discussion of nature to be referring to the composite and the overall
substantial being rather than to those things which do not exist when separated from their
substantial being’ (translation in Lennox 2001, 14).
8
The idea goes back to Socrates. As Sedley 2017, 41 points out: ‘Beauty, whether in an artefact, in
a human body, or in an animal, is not an entirely independent value but an index of something’s
efficient adaptation to its function’. For teleology in Aristotle’s biology, see Gelber 2021.
9
This kind of ‘teleological wonder’, instilled by the recognition of the human body’s supreme
craftsmanship and the distinctive purpose assigned to each of its parts, becomes especially apparent
in Galen’s writings; see Tieleman 2013, with Flemming 2009. Compare Rocca 2003, 240, on
anatomical autopsy and wonder. The wonders of the body have to be seen to be believed, Galen
tells us, and in this respect they are not inferior to the thrills and excitement reserved for those who
are initiated in the Eleusinian or Samothracian mysteries; see De usu partium 7.15 (= 15.6 K.), with
Grant 2011, 13. This is a fitting reminder that, even when connected to solid, empirical knowledge,
thauma remains an emotionally challenging experience.
10
Part. an. 645a23–6: Τὸ γὰρ μὴ τυχόντως ἀλλ’ ἕνεκά τινος ἐν τοῖς τῆς φύσεως ἔργοις ἐστὶ καὶ
μάλιστα· οὗ δ’ ἕνεκα συνέστηκεν ἢ γέγονε τέλους, τὴν τοῦ καλοῦ χώραν εἴληφεν.
11
Part. an. 645a15–16: Ἐν πᾶσι γὰρ τοῖς φυσικοῖς ἔνεστί τι θαυμαστόν.
12
See, for example, Gen. an. 741b4–5: ἡ δὲ φύσις οὐδὲν ποιεῖ μάτην. For the same axiomatic statement,
compare Cael. 271a33; 291b13; IA 711a18; Part. an. 658a8; 695b18. There are always, of course,

and pieces in the bigger picture of an ordered reality to which they have been
originally designed to belong. However ignoble and appalling certain things
may appear at first sight – from small, ugly insects to the human spleen and
intestines – there is always a way to go beyond appearances and discover
behind them the fine artistry of nature.13
Aristotle says here something for which there is no express parallel in
Hippocratic medicine. Physicians in the Hippocratic corpus do not seem to
be interested in passing aesthetic judgements about the interior of the human
body. In part, this could be related to the fact that, despite their boldly
speculative and sometimes obstinately fixed claims about what is taking place
beneath the skin’s surface, that interior remains practically invisible due to the
absence of human dissection; consequently, there is only so much that can be
said about it, which is mainly along the lines of associating visible symptoms
with inner processes that take place beyond what Brooke Holmes calls the
‘threshold of perception’.14 Aristotle deplores precisely this lack of knowledge –
which, it is implied, the previous medical tradition did not manage to over-
come – when he observes, in Hist. an. 494b21–4, that the ‘inner parts of man’
are ‘largely unknown’ (ἄγνωστα γάρ ἐστι μάλιστα τὰ τῶν ἀνθρώπων) and
claims that one has therefore to rely on the dissection and examination of
animals whose nature is similar to that of a human being.15 But even if we allow
for a moment the hypothesis that the Hippocratics would have felt as comfort-
able (or as epistemologically justified) as Aristotle to speak in detail of specific
human ‘organs’,16 they would still have lacked the robust teleological frame-
work that would have allowed them to combine empirical observation with
aesthetic considerations. Galen admits at points that even the great
Hippocrates himself is not always at his best when it comes to explaining the
chreia of certain body parts;17 and there are passages which betray an unfailing
exceptions to the rule. In Part. an. 676b16–677b10, Aristotle concedes that bile, cholê, a residue
produced regularly in the course of an animal’s life, has no evident purpose – it is more of a necessary
by-product of the teleological operations of the liver; see Scharle 2015, 83.
13
See the detailed discussion in Leunissen 2010, 77–81.
14
Holmes 2010, 129; compare Holmes 2018, 67. For the absence of human dissection in Hippocratic
medicine (fifth to fourth centuries bc), and for the cultural and religious reasons which account for
such absence, see Lloyd 1975 and von Staden 1992; compare Nutton 2004, 131, who draws attention
to the exceptional practice of dissection in ancient Egypt, noting, in this context, Herodotus’
‘mixture of fascination and disgust’ when the historian reports on ‘this strange procedure’.
15
The full passage reads as follows: Ἄγνωστα γάρ ἐστι μάλιστα τὰ τῶν ἀνθρώπων, ὥστε δεῖ πρὸς τὰ τῶν
ἄλλων μόρια ζῴων ἀνάγοντας σκοπεῖν, οἷς ἔχει παραπλησίαν τὴν φύσιν. Compare Leunissen 2021,
76–8.
16
I am putting organs in inverted commas because in the case of the Hippocratics it is more precise to
speak of ‘body parts’ instead. See Holmes 2018, 73–4.
17
See Galen, UP 1.7 and 8, with Flemming 2009, 66–7. For Galen’s retrojection of his own teleological
ideas onto ‘Hippocrates’, see Holmes 2014, 136. As Gundert 1992, 465, writes, for the Hippocratics, ‘parts

anti-teleological stance, for instance in [Hipp.] Anat. (8.540 L.) where we
encounter the striking idea of the spleen as ‘a false liver’: that is, an organ
without a purpose.18 With these observations in mind, it is by no means
a coincidence that the only time that a doctor in the corpus is expressly
marvelling at the fine design of a certain organ is found in a treatise entitled
On the Heart,19 which is roughly dated at around 300 bc and is believed by
many to have been influenced by Aristotle in a number of significant ways.20 In
discussing the delicate nature of ‘hidden membranes’ (ὑμένες ἀφανέες) in the
area of the heart – ‘a piece of craftsmanship which deserves description above
all others’ (ἔργον ἀξιαπηγητότατον)21 – the author of Cord. 10 (9.86–8 L.)
notes how these ‘spread out like cobwebs through the cavities and surround the
orifices on all sides and implant filaments into the solid wall of the heart’,
serving as a foundation to the arteries. And he continues:
ἔστι δὲ αὐτῶν ζεῦγος καὶ θύραισι μεμηχάνηνται τρεῖς ὑμένες ἑκάστῃ,
περιφερέες ἐξ ἄκρου περ ὁκόσον ἡμίτομα κύκλου, οἵ τε ξυνιόντες
θαυμάσιον ὡς κλείουσι τὰ στόματα, τῶν ἀορτέων πέρας.
Now there is a pair of these arteries, and on the entrance of each three
membranes have been contrived, with their edges rounded to the approxi-
mate extent of a semicircle. When they come together it is wonderful to see
how precisely they close off the entrance to the arteries.22
The passage clearly rests on advanced anatomical knowledge which is not
paralleled by any other treatise in the corpus; there is in fact evidence in the text
that human dissection has been performed to yield all this detailed informa-
tion.23 Equally exceptional, by Hippocratic standards, is the author’s ‘marked
teleological slant’,24 as it can be seen, for instance, in his observation that, on
close inspection, the human heart looks like ‘the work of a skilled craftsman’.25
may perform particular roles because they have given structures; there is never any hint that they have
particular structures in order to fulfill given roles’. For a recent discussion arguing in favour of some form
of ‘incipient teleology’ in certain treatises of the Hippocratic corpus, see Craik 2017.
18
See Craik 2006, 148; 2017, 216.
19
For the generally negative associations of thauma in Hippocratic medicine, and for its exceptional
use with a positive meaning in the Hippocratic On the Heart, see Kazantzidis 2018, 42–3.
20
First and foremost, by virtue of the fact that its author is proposing that the highest psychic faculties
are located in the heart. For Aristotle’s cardiocentric model, see van der Eijk 2000, 68; 2005, 124–5;
for a detailed discussion of the On the Heart, see Lonie 1973, with Thumiger 2017, 37–8.
21
For the extremely rare ἀξιαπήγητος, an Ionic form of ἀξιαφήγητος, compare Herodotus, Hist. 2.100, 137.
22
Translation by I. M. Lonie in Lloyd 1978.
23
Its advanced anatomical knowledge is the reason why some scholars propose an even later date for
the treatise, placing it at around 250 bc and considering its author as approximately contemporary
with Herophilus and Erasistratus; see Lloyd 1975, 134.
24
Craik 2015, 55.
25
Cord. 8 (9.84–86 L.): δοκέω τὸ ποίημα χειρώνακτος ἀγαθοῦ. Compare also the significant use of
μεμηχάνηνται in the passage cited above.

Aristotle looms here in the background,26 and it would overall be fair to say that
without his intervention the aesthetic experience of looking at the marvellous
properties of the interior of the human body does not yet exist as a possibility.
2 The Body as Machine

In what follows, I wish to consider a different aesthetic model applied to the
human interior, which is first made possible in the context of the systematic
assimilation between bodily organs and mechanical devices in Hellenistic
medicine. This model, it should be noticed in advance, rests on conditions
of clearer visibility, as it were, when compared to Aristotle who, though
undeniably a more acute anatomist than the Hippocratics, still bases his
observations on comparative findings derived from dissected animals.27 With
Herophilus and Erasistratus, on the other hand, we move directly to human
dissection and vivisection and, along with them, to unprecedented and
remarkably precise anatomical discoveries which constitute – to use
Heinrich von Staden’s words – ‘a stunning moment in the history of
science’.28 At the same time, mechanics is consolidated as a discipline only
when we reach the Hellenistic period; consequently, it is only during the third
century bc, as Sylvia Berryman writes, that ‘the development of mechanical
theory and mechanical technology’ allowed certain ancient Greek thinkers –
doctors included – to make use of the concrete insights afforded by this new
discipline and to draw ‘analogies to mechanical devices as a guide to investi-
gating the natural world’.29 Although it is not always easy to identify whether
a doctor was influenced by a mechanician or the other way around, the
evidence speaks, at any rate, of a highly interactive environment which we
have to take into account as we are approaching the new-fangled and, in many
ways, extraordinary world of the Hellenistic anatomists.
In light of these remarks, let me move straight to a number of represen-
tative testimonies which directly connect the name of Erasistratus – whose
26
For the importance of the craft analogy in Aristotle’s natural teleology, see Witt 2015 and Johansen 2020.
27
For the ‘unsystematic, sporadic and tentative’ use of animal dissection before Aristotle – whose own
advanced anatomical research still remains ‘primitive and crude compared with that of some of his
successors’, most notably the Hellenistic anatomists, see Lloyd 1975, 138 and 143 respectively.
28
von Staden 1992, 224.
29
Berryman 2009, 7, who in the same context observes: ‘Evidence from the fourth century is mixed.
Some ideas important to mechanics can be found in Plato and Aristotle, used only in piecemeal
fashion. It seems to be in the Hellenistic period that mechanics consolidated as a discipline’. When it
comes to Hippocratic medicine and practice, as De Groot discusses in this volume, Chapter 5, we
are only entitled to speak of an ‘intuitive grasp’ of physical forces. For some recent helpful accounts
of the history of Greek mechanics, see Cuomo 2018, Rihl 2018 and Berryman 2020.

anatomical discoveries refine and extend Herophilus’ previous findings30 –
with the idea of marvelling at the artful design of the human body. All of
these testimonies derive from Galen, our main, though also generally
critical and hostile, source for Erasistratus:31
(a) προνοητικὴν τοῦ ζῴου καὶ τεχνικὴν αὐτὸς ὁ Ἐρασίστρατος ὑπέθετο
τὴν φύσιν. ἀλλὰ καὶ τὸ χολῶδες ὑγρὸν ἄχρηστον εἶναι παντάπασι τοῖς
ζῴοις ἔφασκεν. οὐ συμβαίνει δ’ ἀλλήλοις ἄμφω ταῦτα. πῶς γὰρ ἂν ἔτι
προνοεῖσθαι τοῦ ζῴου δόξειεν ἐπιτρέπουσα συναναφέρεσθαι τῷ αἵματι
μοχθηρὸν οὕτω χυμόν; (Nat. Fac. 2.2, 2.78 K.)
Erasistratus himself supposed that Nature is capable of forethought for the
living being and capable of expert craftsmanship; and at the same time he
maintained that the bilious fluid was useless in every way for the animals.
Now these two things are incompatible. For how could Nature be still
looked on as exercising forethought for the living being when she allowed
a noxious humour such as this to be carried off and distributed with the
blood?32
(b) Οὐδὲ γὰρ ζῆν οὐδὲ διαμένειν οὐδενὶ τῶν ζῴων οὐδ’ εἰς ἐλάχιστον
χρόνον ἔσται δυνατόν, εἰ τοσαῦτα κεκτημένον ἐν ἑαυτῷ μόρια καὶ οὕτω
διαφέροντα μήθ’ ἑλκτικῇ τῶν οἰκείων χρήσεται δυνάμει μήτ’ ἀποκριτικῇ
τῶν ἀλλοτρίων μήτ’ ἀλλοιωτικῇ τῶν θρεψόντων. καὶ μὴν εἰ ταύτας ἔχοιμεν,
οὐδὲν ἔτι πόρων μικρῶν ἢ μεγάλων ἐξ ὑποθέσεως ἀναποδείκτου
λαμβανομένων εἰς οὔρου καὶ χολῆς διάκρισιν δεόμεθα καί τινος ἐπικαίρου
θέσεως, ἐν ᾧ μόνῳ σωφρονεῖν ἔοικεν ὁ Ἐρασίστρατος ἅπαντα καλῶς
τεθῆναί τε καὶ διαπλασθῆναι τὰ μόρια τοῦ σώματος ὑπὸ τῆς φύσεως
οἰόμενος. ἀλλ’ εἰ παρακολουθήσειεν ἑαυτῷ φύσιν ὀνομάζοντι τεχνικήν,
εὐθὺς μὲν ἐξ ἀρχῆς ἅπαντα καλῶς διαπλάσασάν τε καὶ διαθεῖσαν τοῦ
ζῴου τὰ μόρια, μετὰ δὲ τὴν τοιαύτην ἐνέργειαν, ὡς οὐδὲν ἔλειπεν, ἔτι
προαγαγοῦσαν εἰς φῶς αὐτὸ σύν τισι δυνάμεσιν, ὧν ἄνευ ζῆν οὐκ
ἠδύνατο, καὶ μετὰ ταῦτα κατὰ βραχὺ προσαυξήσασαν ἄχρι τοῦ
πρέποντος μεγέθους, οὐκ οἶδα πῶς ὑπομένει πόρων σμικρότησιν ἢ
μεγέθεσιν ἤ τισιν ἄλλαις οὕτω ληρώδεσιν ὑποθέσεσι φυσικὰς ἐνεργείας
ἐπιτρέπειν. ἡ γὰρ διαπλάττουσα τὰ μόρια φύσις ἐκείνη καὶ κατὰ βραχὺ
προσαύξουσα πάντως δήπου δι’ ὅλων αὐτῶν ἐκτέταται. (Nat. Fac. 2.3,
2.81–2 K.)
For there is not a single animal which could live or endure for the shortest
time if, possessing within itself so many different parts, it did not employ
faculties which were attractive of what is appropriate, eliminative of what is
30
See von Staden 1996, 91; cf. von Staden 1975.
31
For the reasons of Galen’s hostile attitude towards Erasistratus, see Hankinson 1997, 328; 1998a, 31–2;
compare von Staden 1997; Vegetti 1999.
32
Translation in Brock 1916, 123; slightly modified.

foreign, and alternative of what is destined for nutrition. On the other hand,
if we have these faculties, we no longer need channels, little or big, resting on
an unproven hypothesis, for explaining the secretion of urine and bile, and
the conception of some favorable position–in which point alone Erasistratus
shows some common sense, since he does regard all the parts of the body as
having been well and truly placed and shaped by Nature. But let us suppose
he remained true to his own statement that Nature is ‘artistic’ – this Nature
which, at the beginning, well and truly shaped and disposed all the parts of
the animal, and, after carrying out this function (for she left nothing
undone), brought it forward to the light of day, endowed with certain
faculties for its very existence, and, thereafter, gradually increased it until
it reached its due size. If he argued consistently on this principle, I fail to see
how he can continue to refer natural functions to the smallness or largeness
of canals, or to any other similarly absurd hypothesis. For this Nature which
shapes and gradually adds to the parts is most certainly extended throughout
their whole substance.33
(c) ἀλλ’ οὐχ Ἱπποκράτης, ὦ Ἐρασίστρατε, μηδέν τι χείρων ἰατρός σου περὶ
φλεβοτομίας οὕτως ἐπεγίνωσκεν· ἀλλ’ ἃ σὺ θαυμάζεις λόγῳ, ταῦτ’ ἔργῳ
ποιῶν εὑρίσκεται. θαυμάζεις μὲν γὰρ τὴν φύσιν, ὡς τεχνικήν τε ἅμα καὶ
προνοητικὴν τοῦ ζώου, μιμῇ δ’ αὐτὴν οὐδαμοῦ· ἢ διὰ τί πολλάκις ἰδὼν
αἵματος κενώσει τὴν φύσιν ἰασαμένην πολλὰ νοσήματα, τοῦτο οὐδ’ ἐφ’ ἑνὸς
ἔπραξας οὐδὲ πώποτε; τί δὲ σιγᾷς τὰ τῆς φύσεως ἔργα, ἣν ἐπαινεῖς; (Ven.
Sect. Er. 4, 11.158 K.)
But Hippocrates did not take this view of phlebotomy, O Erasistratus, and
he was in no way a worse physician than you are. What you applaud in
theory is found in him put into practice. You marvel at nature, as something
craftsmanlike and at the same time providential towards the living creature;
but nowhere do you imitate her. Why is it, if this is not so, that although you
have often seen nature healing diseases by evacuation of blood, you have
never yet practiced it, even in a single one? And why have you nothing to say
about the works of the nature whom you praise?34
33
Translation in Brock 1916, 127–9. The issue here concerns the central place of channels/passages
(poroi) in Erasistratus’ mechanical explanation of the processes through which various substances
fundamental to the functioning of the body are conducted to its various organs. Simply put,
according to Erasistratus, matter is being attracted to that which is being emptied (pros to kenou-
menon akolouthia). In the case of hunger, for instance, our appetite derives from an empty stomach
which, once filled by the nutritive elements which find their way in it through ‘passages’ in the body,
rests content. To Galen’s mind, the whole process should rather be attributed to the stomach’s
inborn faculty to attract food, which is more in line with the teleological conception according to
which organs have been so designed as to perform specific functions. See Lloyd 1973, 80; Furley 1989,
158–9; compare Polito 2006, 302.
34
Translation in Brain 1986, 21. Galen’s critique at this point is focused on Erasistratus’ silence on the
subject of phlebotomy; see Brain 1986, 15.

(d) ἄχρι μὲν οὖν τοῦ τὴν φύσιν ὑμνεῖν ὡς τεχνικὴν κἀγὼ γνωρίζω τὰ τοῦ
περιπάτου δόγματα, τῶν δ’ ἄλλων οὐδὲν οὐδ’ ἐγγύς (Nat. Fac. 2.4,
2.88 K.)
Now, in so far as he sings a hymn to Nature as being an artist in construc-
tion, even I recognize the Peripatetic teachings, but in other respects he does
not come near them.35
There are two important things that I would like to keep from these
passages. On the one hand, we cannot fail but notice that Erasistratus is
consistently reported by Galen as marvelling (thaumazein) at the supreme
craftsmanship of Nature (phusis technikê),36 for which the artful design of
the (interior of the) human body stands as solid proof (see, especially, the
language used in passage (b): ἅπαντα καλῶς τεθῆναί τε καὶ διαπλασθῆναι
τὰ μόρια τοῦ σώματος ὑπὸ τῆς φύσεως).37 On the other hand,
Erasistratus’ aesthetic awareness is not, and cannot be, exactly the same
as that of Aristotle, with which we started our discussion; for unlike
Aristotle, who assigns purposiveness to almost every single part of the
human body38 and speaks of beauty in terms of final causes and clearly
assigned functions, Erasistratus is not presented by Galen as a committed
teleologist: though extremely sensitive to the ‘artful design’ of things, the
Hellenistic anatomist is said to admire a kind of body that reserves
considerable space for some really useless stuff as well. This applies not
only to the bilious fluid mentioned above in passage (a) but also – if we
trust Galen’s critical remarks in Nat. Fac. 2.4 – to the spleen, the omentum,
the renal arteries and yellow bile.39 Overall, the body according to
Erasistratus is not something where everything makes sense as serving
a purpose – this is what Aristotle would have liked to believe and would
have then proceeded to raise our aesthetic awareness along these lines.
Consequently, if Erasistratus has only a circumscribed sense of
teleology,40 we need to ask on what other grounds he could be speaking
35
Translation in Brock 1916, 139; slightly modified.
36
Notice, for instance, the enthusiastic tone conveyed by the phrase τὴν φύσιν ὑμνεῖν in passage (d)
above. I haven’t been able to spot any other doctor in the Galenic corpus who is presented as ‘singing
a hymn’ to Nature.
37
Erasistratus’ praise of nature is also attested in non-medical sources, for example, in Plutarch, De
amore prolis 495C: πανταχοῦ μὲν γὰρ ἡ φύσις ἀκριβὴς καὶ φιλότεχνος καὶ ἀνελλιπὴς καὶ ἀπέριττος,
“οὐδέν” ὡς ἔφησεν Ἐρασίστρατος “ἔχουσα ῥωπικόν”, ‘for, everywhere nature is exact, fond of
technē, without deficiency, and without superfluity, having, as Erasistratus says, nothing tawdry’
(translation in von Staden 1996, 95). Garofalo (1988, 89) believes that the entire sentence between
πανταχοῦ and ἀπέριττος reproduces Erasistratus’ exact phrasing.
38 39
For some exceptions, see n. 8 above. See von Staden 1996, 95; 1997, 195.
40
Von Staden (1997, 194–5) attributes this circumscribed sense to the influence of Theophrastus who,
in comparison to Aristotle, imposes additional limits on teleology; see Lennox 1985. For the close

of the beauty of the body in the intense terms that he is presented to be
doing so by Galen. This question, I submit, cannot be answered properly
unless we place Erasistratus’ groundbreaking anatomical work side by side
with the emerging field of mechanics during the third century bc and
associate it with the pioneering technological discoveries of his time, to
which the word technikos in Galen’s testimonies must be alluding, to some
extent at least.41 Erasistratus may have a looser sense of teleology when
compared to Aristotle, but what has been enhanced in the meantime, and
considerably so, is the visibility of the human interior. The upgraded access
gained through human dissection gives rise to an extremely detailed
anatomy in the context of which the body becomes increasingly assimilated
to the simultaneously expanding world of machines. This assimilation, as
I will proceed to argue, yields novel results and insights about the function
of individual organs, but on another level, and for what concerns us here, it
creates also the space for appreciating aesthetically the human body in new
ways.
During the early Hellenistic period, the close interaction between
mechanical devices and the human body has an obvious practical side,
but it is also aesthetically significant. We can think here, for instance, of
Herophilus’ elaborate pulse theory. What makes this theory fascinating is
its precision, allowed by Herophilus’ construction of a portable clepsydra
(in all probability borrowing from advancements in contemporary water-
clock technology) which could be perfectly adjusted to fit the age group of
each individual patient. A closer look at the evidence reveals an ‘artistic’
angle which should not pass unnoticed. Herophilus did not just differen-
tiate between various pulse rhythms according to age, but he proceeded to
do so by distinguishing between different types of actual metrical units:
thus, while we all start as infants with a naturally pyrrhic pulse rhythm (˘ ˘)
we move on to a trochaic pulse (¯ ˘) in adolescence to a spondaic prime of
life (¯ ¯) and, finally, to an iambic pulse rhythm (˘ ¯) in old age.42 It has been
relationship between Theophrastus and Erasistratus, see Diog. Laert. 5.57; Sextus Empiricus,
Adversus mathematicos 1.258 and Galen, Nat. fac. 2.5 (2.90 K.). See Hankinson 1998b, 302–3.
41
As von Staden (1996, 96) remarks with reference to Erasistratus’ use of the concept of technikê phusis:
‘his view seems to be that, just as technê proceeds in a methodical, goal-directed fashion to the
construction of mechanical devices, so nature methodically (re)produces purposively structured
natural machines, including the human body’. For the pair phusis/technê in Greek medicine, see von
Staden 2007.
42
According to Herophilus’ theory of the pulse, both the dilation and contraction of the artery in the
majority of newborn children consist of one primary time unit. In adolescence, each cycle of diastole
and systole consist of three primary units, the dilation lasting for two units, the contraction for one.
When we reach the prime of life, the pulse rhythm consists of four primary time units equally
divided between contraction and dilation. Finally, in old people the pulse consists of three primary

argued that Herophilus’ idea of a primary perceptible time unit as analo-
gous to a short unit in the feet of musical metres was partly inspired by the
metrical theory of his contemporary Aristoxenus of Tarentum, expounded
in a treatise entitled Elements of Rhythm.43 However this may have been,
what is certainly intriguing is the fact that the new findings which become
possible through the use of a technological device (a clepsydra) are not
conveyed by Herophilus in a coldly clinical discourse but create instead the
conditions for an aesthetically enhanced engagement with previously
unexplored details of the human body. Pulse rhythm – which can now
be explained via mathematical proportions of absolute regularity –
becomes embedded in a metaphor that invites us, more or less, to think
of blood’s upbeats and downbeats as ‘nature’s music in our arteries’.44
I would suggest that, likewise, the increasing assimilation between
specific human organs and mechanical devices (wherein a device is not
put literally into practice in order to help us understand better, for example,
a patient’s pulse rhythm but features instead in the context of an extended
figural analogy between man-made constructions and the mechanics of
certain bodily parts and functions) gives rise not only to a new understand-
ing but also to a new aesthetics of the human interior. Erasistratus – my
case study in this chapter – is generally known for developing a more
‘mechanistic’ version of the body in comparison to Herophilus: on the one
hand, he abandoned Herophilus’ concept of invisible innate faculties
(dunameis) which regulate material processes, proposing instead
a corpuscular model of the human body where every single function is
explained by the simple principle that matter will rush into any space that is
being emptied. On the other hand, and while acting on this material basis,
Erasistratus made systematic use of theories developed in the context of
Alexandrian pneumatics, hydraulics and hydrostatics in order to explain
the material properties of a mechanistically operating body. Thus, while
already with Herophilus mechanical means are being employed in medical
practice, Erasistratus appears to have gone a step further, by discovering, as
it were, a number of ‘machines’ and ‘sub-machines’ inside the body.45 The
kidneys, liver and bladder – according to later testimonies – act for
Erasistratus as filters; the stomach grinds and crushes ingested food like
a corn mill.46 Most notably, the heart bears remarkable similarities to
a water pump – of the kind which was invented by the mechanician
time units, the dilation lasting for one unit, the contraction for two. For a detailed discussion, see
von Staden 1989, 276–83 and Berrey 2017, 60–73.
43
See von Staden 1996, 89. 44 I borrow the phrase from von Staden 1996, 89.
45
See von Staden 1996, 92. 46 See Nutton 2004, 137.

Ctesibius during Erasistratus’ lifetime.47 Below, and for the sake of clarity
and convenience, I reproduce the exhaustive list of correspondences as
provided by von Staden (1996, 93–4):
(a) Like Erasistratus’ version of the heart, Ctesibius’ water pump has two
chambers.
(b) Both the cardiac pump and the water pump are equipped with valves
to ensure the irreversibility of the flow.
(c) As in Erasistratus’ model of the heart, so in Ctesibius’ pump there are
four sets of valves, two controlling intake and two regulating outflow
from the two chambers.
(d) Ctesibius uses valves to ensure the irreversibility of the flow of either
liquid or air (both here and in several of his other machines).
Erasistratus likewise describes the heart valves as ensuring the unidir-
ectional flow of either air (πνεῦμα, breath) or liquid (blood): two sets
of cardiac valves, he says, control the flow of breath (respectively into
the left chamber of the heart from the lungs and out of this left
chamber into the aorta), while two other valve sets ensure unidirec-
tional flow of blood into and from the right cardiac chamber.
(e) Ctesibius’ water pump has forked pipes (fistulae furcillae), and
Erasistratus’ vascular system is similarly dependent on forking vessels.
(f) Both Erasistratus’ version of the heart and Ctesibius’ pump depend
centrally on the principle of an intermediate valved chamber (medius
catinus). The Erasistratean heart serves as a double intermediate
chamber, on the one hand, for blood between the vena cava (coming
from the liver) and the pulmonary vessels that carry blood to the
lungs, and, on the other hand, for breath between the lungs and the
aorta.
(g) Just as Ctesibius’ water pump is constructed with twin cylinders
(modioli gemelli) sitting in a round space, so Erasistratus’ heart is
a two-chambered machine that sits in a larger roundish space, the
thorax.
(h) As compression and expansion alternate in each chamber of
Ctesibius’ water pump, so contraction continuously alternates with
dilatation in Erasistratus’ cardiac bellows-pump.
(i) More fundamentally, the mechanical principles are similar in the two
cases: propulsion of matter into a contiguous space by compression or
contraction and drawing in of contiguous matter by expansion or
47
For Ctesibius’ water pump, see Oleson 1984, 301–25; compare Wilson 2008, 353–5.

dilatation, based on the recognition that continuous – as opposed to
disseminate – void does not exist naturally. (It should not be over-
looked, however, that Erasistratus in his extant writings explicitly
applies the theory of interstitial void only to the movement of liquids
through the body, not to the compressibility of air.)
Such lists of correspondences have been usually explored in relation to
the question of ‘who influenced whom’. Some scholars – including Matteo
Valleriani in the present volume (Chapter 10) – support the view that it was
the doctors who influenced the mechanicians; others maintain that the
interaction worked the other way around, and that it was primarily the
field of mechanics which gave rise to a corpus of practical knowledge that
was subsequently applied to the human body. A third approach speaks of
a cross-disciplinary environment, choosing to avoid the model of one-
directional influence and speaking instead of an ongoing dialogue between
different epistemic fields which constantly feed each other with new
findings and insights.48
As things stand, the question cannot be answered conclusively. But let
us, for a moment, allow the sound possibility – as has been done by many
so far – that Erasistratus was the one who borrowed from technical details
and the conceptual imagery of mechanical devices,49 and that the more he
proceeded to use them in order to speak of his new anatomical discoveries,
the more intricate similarities he found himself spotting between brilliantly
conceived man-made constructions and bodily organs. Naturally, this
would seem to create a context in which the emerging correspondences
do not relate only to matters of precision and functionality but also have an
aesthetic aspect; Erasistratus’ anatomy is so deeply immersed in the world
of mechanical analogies that it would be hazardous to assume that his
48
For the hypothesis that Erasistratus follows Ctesibius, see Lonie 1973, 138–9, and, more recently,
Netz 2020, 407; compare Longrigg 1993, 207–9, and Berryman 2009, 200. Russo 2004, 147, argues
that the inspiration went the other way around. For the cross-fertilization between mechanics and
medicine, working in both directions, see von Staden 1996, 94–5; 1998, 163; compare Vegetti 1993;
1998; Nutton 2004, 135–9.
49
Such ‘borrowing’ does not necessarily require that Erasistratus should have kept among his
bookshelves textual or even diagrammatic codifications of the construction of machines (see the
reservations raised by Valleriani in this volume, Chapter 10); what I have in mind is rather a closely
interacting intellectual milieu where knowledge travels from one discipline to the other seamlessly
and freely, sometimes in written form and at others orally – as is the case with much of the culture
produced in Hellenistic Alexandria. Particularly as regards medicine, what has survived from
Herophilus and Erasistratus shows these versatile physicians to have kept an extremely sensitive
and open eye on ongoing developments in other fields and disciplines, from philosophy and music,
to technology, literary criticism and poetry; see Roby 2016, 69–71; Berrey 2017, passim; Schironi
2018, 749–52; compare Thomas 2021, 5–17.

‘marvelling’ at the artful design of the human body – so richly attested in
Galen’s testimonies – is not somehow related to the latter’s machine-like
nature. If that is the case, this would also seem to entail a new model of
beauty concerning the human interior; for it is one thing to marvel at the
beautiful design of the human body, and another thing to marvel at that
same body as a machine. Thauma in the first case is a question of how
Nature or God created us; in the second case, where what we ‘see’ involves
the presence of little ‘machines’ and ‘sub-machines’ merging conceptually
with flesh and bones, thauma is yet again about nature’s wisdom but,
crucially, it involves this time a significant process of association with man-
made constructions. A piece of art – a statue or a self-moving automaton,
for example – is often marvelled at for its lifelike qualities, because ‘it looks
so real’.50 In the body–machine analogy which I have been laying out above
an inverse process takes place: the beauty of animate matter becomes
increasingly appreciated because it resembles so closely something inani-
mate. For Aristotle our instinctive response of disgust when we come face
to face with the interior of the human body can be overcome once we
discover the gracefully purposive nature of things. Alexandrian medicine
nourishes a different aesthetic model: the human body, understood
through a system of close correspondences with mechanical devices, can
instil feelings of wonder and admiration because parts of it look almost
identical to products of human ingenuity and craft. What is more, unlike
Aristotle who attributes the artistry of Nature to an ultimate authority of
divine proportions, the body–machine analogy puts human inventiveness
centre stage: reading, understanding and marvelling at the body as
a machine requires that we keep as our main point of reference, both
intellectually and aesthetically, the human realm of manufactured devices.
It is important to be reminded at this point of the distinction between
the κατὰ φύσιν and the παρὰ φύσιν thauma, as we find it in the introduc-
tion of the ps.-Aristotelian Mechanics (847a11–24):
Θαυμάζεται τῶν μὲν κατὰ φύσιν συμβαινόντων, ὅσων ἀγνοεῖται τὸ αἴτιον,
τῶν δὲ παρὰ φύσιν, ὅσα γίνεται διὰ τέχνην πρὸς τὸ συμφέρον τοῖς
ἀνθρώποις. ἐν πολλοῖς γὰρ ἡ φύσις ὑπεναντίον πρὸς τὸ χρήσιμον ἡμῖν
ποιεῖ· ἡ μὲν γὰρ φύσις ἀεὶ τὸν αὐτὸν ἔχει τρόπον καὶ ἁπλῶς, τὸ δὲ χρήσιμον
μεταβάλλει πολλαχῶς. ὅταν οὖν δέῃ τι παρὰ φύσιν πρᾶξαι, διὰ τὸ χαλεπὸν
ἀπορίαν παρέχει καὶ δεῖται τέχνης. διὸ καὶ καλοῦμεν τῆς τέχνης τὸ πρὸς
τὰς τοιαύτας ἀπορίας βοηθοῦν μέρος μηχανήν. καθάπερ γὰρ ἐποίησεν
50
See Berryman 2009, 26–7; Neer 2010, 105–8. On verisimilitude in ancient Greek sculpture and art,
see also Steiner 2001, 27–32.

Ἀντιφῶν ὁ ποιητής, οὕτω καὶ ἔχει· τέχνῃ γὰρ κρατοῦμεν, ὧν φύσει
νικώμεθα. τοιαῦτα δέ ἐστιν ἐν οἷς τά τε ἐλάττονα κρατεῖ τῶν μειζόνων,
καὶ τὰ ῥοπὴν ἔχοντα μικρὰν κινεῖ βάρη μεγάλα, καὶ πάντα σχεδὸν ὅσα τῶν
προβλημάτων μηχανικὰ προσαγορεύομεν.
Marvellous things occur in accordance with nature, the cause of which is
unknown, and others occur para phusin, which are produced by skill for the
benefit of mankind. For in many cases nature produces effects against our
advantage; for nature always acts consistently and simply, but our advantage
changes in many ways. When, then, we have to produce an effect contrary to
nature, we are at a loss, because of the difficulty, and require skill. Therefore
we call that part of skill which assists such difficulties, a device. For as the
poet Antiphon wrote, this is true: ‘We by skill gain mastery over things in
which we are conquered by nature.’ Of this kind are those in which the less
master the greater, and things possessing little weight move heavy weights,
and practically all those problems that we call mechanical.51
The passage has been subjected to various interpretations. Some have
argued that we should understand παρὰ φύσιν as referring to things
which happen ‘against/contrary to nature’; on this view, mechanics
would seem to amount to a kind of magic – a cheating of sorts, which
breaks or suspends regular patterns of cause and effect and brings about
supernatural results.52 Others, more recently, have laid the emphasis on the
fact that παρὰ φύσιν ‘need not imply transgression against or opposition
to, rather than merely going beyond, a given category’.53 On this reading,
the phrase in question could be taken as a reference to forced motion,
indicating little more than the results which cannot occur without some-
one’s intervention. Be that as it may, it is important to stress that the first
extant theoretical treatise on the subject of mechanics in antiquity opens
with a significant distinction between natural and artificial thaumata. The
author of the text displays a keen awareness that the world of mechanical
devices he is about to unveil can claim its own distinctive place in the realm
of wonder and admiration. It is these two different types of wonder,
I suggest, which merge in Hellenistic anatomy: marvelling at the heart by
virtue of the fact that its precise structure and function resemble those of
a water pump quite simply means that you marvel at something which is
51
Translation in Hett 1936, 331; slightly modified.
52
See, especially, Krafft 1967; for the connection between mechanics and magic, see Krafft 1973, 7 and
Mayor 2018, 3: ‘the more advanced the technology, the more it seems like magic’. Mayor is
discussing at this point Arthur C. Clarke’s famous dictum: ‘Any sufficiently advanced technology
is indistinguishable from magic’ (which first appeared in Clarke’s 1973 (revised) essay ‘Hazards of
Prophecy: The Failure of Imagination’).
53
Berryman 2009, 47; compare Schiefsky 2007.

happening κατὰ φύσιν by referring it to a device whose wondrous effects
occur παρὰ φύσιν.
What I have been trying to raise so far is the possibility that, among
Hellenistic anatomists, the expression of wonder when it comes to the
artful design of the human body is filtered through and becomes effectively
enhanced by a process of assimilation between body and machine – the
latter being already conceived as a (παρὰ φύσιν) wonder in and of itself. If
this hypothesis holds true, it would appear to give an interesting twist to
the Aristotelian text with which I have started my discussion. In Part. An.
645a15–16, Aristotle mentions thauma in order to speak of the admiration
that is instantaneously produced when a thorough knowledge of the body’s
interior has been acquired, and the specific function and purpose of each
individual organ has been explained; this admiration has only little to do
with the wonder which in Metaph. 982b is said to lie ‘at the beginning of
philosophy’, when things still look confusing and perplexing,54 and
emerges instead as a deeply felt response to the marvels of life, when
puzzles have been solved and the beauty of nature has become unveiled
to its full extent. By contrast, the ‘wonderful’ world of mechanics – which,
as I have been arguing, informs to a considerable degree the way in which
Hellenistic medicine sees and appreciates the elegance of the human body –
always retains an element of elusiveness and even deceptiveness. As
Courtney Roby observes, ‘pneumatic-mechanical wonders play with the
tension between audience awareness of their complex and delicate machin-
ery and its obfuscation by their decorative and concealing elements’.55
54
It is because of wonder, Aristotle says, that human beings undertake philosophy. The person who is
in a state of loss and wonder proceeds to examine more closely whatever it is that is puzzling him
(Metaph. 982b12–19). Interestingly enough, Aristotle compares our – at first perplexing and
mystifying – encounter with the world with what happens when we attend a puppet show
performed by mechanical marionettes (tôn thamatôn tautomata, Metaph. 983a14), and we wonder
about who is controlling the figurines. See Nightingale 2004, 253–4, and Nussbaum 1986, 259–60.
55
Roby 2016, 46. See, especially, Hero, On Automaton Construction 1.1: ‘The study of automaton
construction has been considered an acceptable pursuit because of the complexity (to poikilon) of the
craftsmanship involved and because it produces a baffling (ekplêkton) spectacle. For, briefly put,
automaton construction encompasses every part of mechanics in its step by step construction’
(translation in Cuypers 2010, 333). Tybjerg 2003 argues that Hero’s preface does not mention
wonder for wonder’s sake; these mechanical devices are not mere toys and gadgets, as was once
thought, but they can lead to serious philosophical investigation about the properties of matter and
the causes behind things; compare Cuomo 2008, 24 who remarks that Hero wants to engage his
readers in ‘the philosophically validated act of curiosity’. That having been said, the element of
playfulness is never fully erased. According to Asper 2017, 43, the understanding of automata in an
education context should not be exaggerated; often, the main point of a mechanical device, as
described in Hero’s writings, ‘is not to do anything. Its only function, by hiding the purely
mechanical causes of the unexpected movements, is to make the observer marvel’ at what they are
seeing. Compare Vernant 1983, 283: ‘Wherever the machines described by the engineers have

At the level of performance, the evidence we have allows us to place
machines in Hellenistic Alexandria in a court culture of public display
and wonder:56 sources speak of such things as a flying wooden dove which
rose up in the air and then landed or of a gigantic self-moving snail which
led the procession through the Theatre of Dionysus, probably at the
Dionysia of 308/7, while ‘spitting out slime’ in order to produce
a realistic effect.57 Ctesibius, whose water pump has been discussed as
providing a ‘blueprint’ for Erasistratus’ model of the heart, is also known
for creating a number of fanciful mechanisms;58 in fact, later mechanical
authors go so far as to accuse him of creating complicated constructions for
‘wonder’s sake’ alone.59 At the same time, the image of the illusionist
mechanic emerges also from the theoretical treatises devoted to the subject.
Recent scholarship has demonstrated that even in cases of analytical texts,
such as the ps.-Aristotelian Mechanics where the laws of physics and
mathematics are constantly invoked, machines continue to pose
a problem, in the sense that they are embedded in a deliberately self-
involved, ultra-technical and effectively obscure idiom which often
makes it difficult for readers to understand what exactly is being discussed.
Some of the descriptions sound way too theoretical, and they do not seem
to be directly concerned with issues of practical application; in this case, the
author appears to create a kind of marvellous, impressive discourse which
suspends information precisely at the moment when he pretends to pro-
vide it amply within the text. As Markus Asper points out – drawing from
the polemical remarks of Athenaeus the Mechanic in the first century bc –
a utilitarian purpose, they are used and conceived of as instruments for multiplying human
strength . . . When they call upon other sources of energy and, instead of amplifying a force given
at the outset act as automata producing their own movement, they turn out to be constructions that
follow a whole tradition concerned with objects to be marveled at, and, as such, they are marginally
relevant to technology in the strict sense. They are thaumata made to astonish people. Their value
and interest lie not so much in their usefulness as in the admiration and pleasure they arouse in the
spectator.’ For the central notion of ekplēxis (attested in Hero’s preface) in Galen’s wonder-inducing,
public anatomical demonstrations, and its connection with thaumatopoiia, see Gleason 2009.
According to Gleason, Galen’s demonstrations were intended to be instructive, but they were also
staged as a form of entertainment which was meant to astonish and baffle the watching audience.
56
See Berrey 2017, 163–90. Compare Cuomo 2007, 54: even in cases of war technology, ‘the aesthetics
of the machine played an important role, almost as if appearance was an integral part of efficacy’.
57
For the wooden dove, attributed to Archytas of Tarentum, see Gell. 10.12.8–9 = Archytas test. A10a
Huffman. For the snail, see Polybius 12.13.9–12 = Demochares fr. 4 FGrH / fr. 7 BNJ. For
a discussion of these marvels, see Ruffell in this volume, Chapter 3.
58
Vitruvius (9.8.4) mentions Ctesibius as a writer on automata and other ‘entertainments’ (deliciae).
59
See, for example, Athenaeus, Mech. p.31 W. = p. 58 W.-B. (passing a rather critical judgement on
Ctesibius’ construction of a movable giant tube inside which men could walk and climb to a wall of
any height during a siege): Γενναίου δὲ τοῦτο ἄξιον οὐθενός, ἀλλ’ ἐκ θαυμάτων τὸ μηχάνημα
συγκείμενον καὶ μάλιστα τὸν τεχνίτην τὸ θαυμάσαι.

Hellenistic authors on the subject are criticized ‘for saying’ but ‘not doing’;
when Athenaeus notes that some of his predecessors ‘wasted too much time
in useless language’ (πολυγραφοῦντες εἰς οὐκ ἀναγκαιους λόγους
καταναλίσκουσι τὸν χρόνον), displaying their ‘polymathy’ at the expense
of utility,60 ‘the sarcastic point’ he seems to make is that, ‘unlike in
constructing a real machine, machines on paper bury the mechanical
knowledge involved in obscure writings, and thus the reputation following
from it is nothing, a waste of effort’.61
There is an interesting parallel which can be traced at this point with
Hellenistic medicine. Scholars have aptly observed that, for all their remark-
able accuracy and precision, Herophilus’ and Erasistratus’ anatomical dis-
coveries did not yield any tangible results in actual medical practice, nor did
they have a lasting impact, in the sense that, after the third century bc, no
ancient scientists ever seem to have resumed systematic human dissection –
when one would have in fact expected the opposite, considering the ground-
breaking advancements in the field in the period before.62 Von Staden
speaks of ‘a mysteriously abrupt disappearance’ of human dissection follow-
ing the death of Herophilus and Erasistratus.63 In addressing this question,
Rebecca Flemming argues that the paradox could be resolved once we move
away from the strictly scientific context in which anatomical survey was
being conducted in Hellenistic Alexandria and consider its wider ideological
implications. The Hellenistic anatomists’ intention, she notes, was in the
first place to colonize the human body with new knowledge, in the same way
that previously uncharted territories of an ever-expanding empire were
discovered and catalogued by geographers and other natural scientists. If
we follow this hypothesis, it then transpires that:
conquest is not something that needs to be repeated. It is, by definition, a one-
off occurrence, though often the platform for further developments. Once the
knowledge of the body has been won – the territory mapped and named – then
that is the end of it, though it may be the beginning of something else. In these
terms, human vivisection and dissection thus came to its natural conclusion as
Herophilos and Erasistratos completed their project.64
At this point, Flemming invites an intriguing connection between
systematic, medical survey and less ‘scientific’ branches of knowledge,
mentioning among the latter paradoxography – a genre which is equally
60
The passage is quoted and discussed by Whitmarsh 2011, 237.
61
Asper 2017, 28–9. See also pp. 35–6: on the ‘absurdly mathematized diagrams’ found in mechanical
treatises.
62
See the discussion in Flemming 2005. 63 Von Staden 1992, 223. 64 Flemming 2005, 455.

fuelled by a drive to map previously unheard ‘marvels’, thaumata, and to
unveil the ‘wonderful’ qualities of the natural world, which have so far
remained hidden. While the differences between Hellenistic medicine and
paradoxography persist, tracing a conceptual link between the two65 – on
the common basis of a broadening of horizons, which calls for methodical
inquiry and yields extended lists of new findings – can be sufficiently
rewarding: on the one hand, it allows us to move beyond simple binary
models of opposition and to see ‘the development of bodies of occult
knowledge about nature’ as ‘an extension of methods of organization, of
systematization, established in more “rational” areas of natural knowledge,
rather than their contradiction’.66 On the other hand, it provides us with
a solid basis to start thinking that, despite its rationalistic vein, Hellenistic
anatomy remains intent on disclosing and laying before us a brand new
world of ‘wonderful’ discoveries, explaining the function of the human
body in thorough detail while at the same time re-mystifying it as
a remarkably constructed entity with all sorts of astonishing assets and
properties.
The machine–body analogy, I submit, could be set in this context: while
the analytical language which runs through the assimilation between
bodily organs and mechanical devices is based on the premise that the
body can be seen and explained down to its last detail, its final effect can
nonetheless be profoundly bewildering, not only because we end up
looking into our bodies as ‘machines’ – an act of perception that entails
a certain degree of self-alienation given that we attempt to comprehend
animate matter, our very own selves, by invoking the function of inanimate
objects67 – but also because a mêchanê remains by its nature, as we have
seen, an elusive and perplexing thing; grounded as it may be on a strict set
of established physical and mathematical principles, it never ceases to instil
feelings of wonder, ekplêxis, by acquiring a life of its own in ways which are
not always fully comprehensible. Compared to the loosely conceived
65
For the complex relationship between medicine and paradoxography in antiquity, see the recent
collection of essays in Kazantzidis 2019.
66
Flemming 2005, 460.
67
We can cite here as a parallel the de-familiarizing effects of atomism in antiquity. According to
Kennedy 2007, 384: ‘Below the proliferation of colours, tastes, sounds, and smells which is the world
of our senses, Epicurean physics asks us to imagine atoms that have only size, shape, weight, and
movement. Moreover, these atoms are themselves lifeless. What we call “life” and invest with such
significance is merely a phenomenon of particular atoms moving in particular ways. The “reduc-
tionist” thrust of such explanations runs the risk of producing an alienating effect [emphasis added].
Readers may well be convinced by the theory, but be left with feelings of disorientation about things
they had previously felt to be of central value, all the things which make us “human”.’

interior of the human body in Hippocratic medicine, there is no doubt that
Herophilus’ and Erasistratus’ anatomical discoveries break substantially
new ground. However, the extra-analytical discourse that is being
employed in these contexts – of which the machine–body analogy forms
an integral part – does not necessarily lead to conclusive results, nor is it
always communicated in a straightforward way. According to von Staden,
‘like practically all science’, the science of the Hellenistic anatomists is
a combination ‘of insight and blindness, of uncovering and covering, of
unwittingly supplementing and suppressing, of augmentation and elision,
in the slippery process of turning observation, hypothesis, and experiment
into text’.68 As a result, what we witness is a slide from the visible back to
the invisible. Erasistratus’ anatomy, rather than settle, once and for all, the
question of why certain organs work the way they do on the basis of
observable analogies with mechanical devices, ends up constructing
a model of the body ‘as another surface concealing even smaller parts
visible only to reason’.69 When the Empiricist school of medicine comes
to the fore as early as the second half of the third century bc, insisting that
doctors should primarily appeal to experience and autopsia and should
avoid referring their explanations to invisible entities which can only be
grasped by the mind’s eye, it is largely as a reaction to Erasistratus, who is
believed by his opponents to invest excessively in ‘mental seeing’ and
inferential reasoning.70 Considered in these terms, the machine–body
analogy does not yield as concrete and conclusive a scanning of the body
as it appears at first sight; rather, it seems to be embedded in a metaphorical
discourse which revives, as it were, a perception of the body as an elusive
entity – an entity whose wonderful qualities may become visible at the level
of its artful design and structure but cannot be fully grasped unless through
our imagination.
Richard Dawkins has demonstrated that science is not just about
dispelling the mysteries of the world around us; it can also help us recapture
a lost sense of awe and admiration for things which have been long taken
for granted. By exposing us to a theoretical and often difficult language
which sets out to explain the real nature of things, and by invoking ‘mind
images and allusions that go beyond the needs of straightforward
understanding’,71 scientific discourse shakes off ‘the anaesthetic of
68
von Staden 1992, 224. 69 Holmes 2010, 17.
70
The idea of things ‘seen with the mind’ is formalized in Hellenistic medicine as Erasistratus’ τὰ
λόγῳ θεωρητά (frs. 76–7 Garofalo). For the Empiricist school of medicine, see Hankinson 1995;
compare Frede 1988. For Erasistratus as a target of the Empiricists, see von Staden 1992, 235.
71
Dawkins 2000, 180.

familiarity, a sedative of ordinariness which dulls the senses and hides the
wonder of existence’.72 The ‘poetry of science’, as Dawkins calls it,73 is at once
revealing and de-familiarizing; the wonder produced by it combines informed
admiration with bewilderment and confusion. I would suggest that the
discovery of the body’s interior machinery by the Hellenistic anatomists
could be seen as producing a similar effect: though grounded on an unprece-
dented degree of scientific precision and detail, its final outcome is to lay bare
before our eyes a new and impressive image of the body as a mêchanê, leaving
us to wonder what the difference really is between us qua living organisms and
a carefully constructed device which operates in the same fashion as we do.
This wonder, while originally deriving from an in-depth knowledge of the
body, cannot but involve also a certain degree of aporia – an aporia which
revolves around the little machines and sub-machines that make up our bodies
and can even be said to extend to the very meaning of ‘life’ itself.
Conclusion
I have tried to briefly sketch out two scripts of marvel relating to the human
body’s interior: the first derives from assigning a purpose to everything;
the second from discovering that bits and pieces of the body look like little
machines. Considerable attention has been placed recently on the re-
evaluation of the supposed polarity between teleology and mechanics in
ancient Greek philosophy and medicine. Rather than assume a mutually
exclusive relationship between the two, scholars argue that the two models
can be seen to converge and combine with each other in a number of
significant ways.74 An organ which looks like a machine – for instance,
Erasistratus’ heart – is still working with a specific purpose; in fact, its machine-
like design can be adduced to further confirm the idea that nature did
everything in wisdom.75 Differences, however, persist, and one of them relates
72
Dawkins 2000, 6. 73 Dawkins 2000, 26–7.
74
Johnson 2017, 132–3, notes, for instance, that while Aristotle discusses respiration in a teleological
context (it exists as a function for the sake of cooling the internal organs), this does not prevent him
from providing also a mechanistic explanation of the whole process. In Resp. 21, we read that ‘both
heart and lungs conform pretty well to the shape of the bellows in a smithy’; when people respire ‘they
raise their chest because the motive principle of the organ described resident within the chest causes an
identical expansion of this organ. When it dilates the outer air must rush into it as into a bellows, and,
being cold, by its chilling influence reduces by extinction the excess of fire’ (translation by Ross,
reproduced in Monte Johnson 2017, 132–3); for the Aristotelian passage, see also Korombili in this
volume, Chapter 7. For teleology and mechanics in Hellenistic medicine, see von Staden 1997.
75
See, however, Netz 2020, 407, who does not exclude the possibility that the presence of mechanical
devices in Erasistratus’ accounts of the human body ‘could . . . have been, among other things, non-
teleological accounts’.

to the important issue that teleology ascribes the purpose of things to an
invisible force, whereas a mêchanê has a human constructor. To argue that the
body consists of machines could thus be seen as opening a new avenue into
looking at its marvellous properties: kallos in this case, while still being linked to
a super-human designer,76 is in a sense more concretely understood and
appreciated in practice with direct reference to the human mind’s inventions.
As a consequence, I would suggest that the beauty which lies within the body
in Hellenistic medicine implicates human ingenuity more directly. But this
ingenuity involves also a certain degree of deception – one that projects onto
a mêchanê’s intrinsic capacity to baffle and confuse. Unlike Aristotle, who
proposes that the body should be fully comprehended before we are able to
marvel at it properly, the machine–body analogy thus reinstates a more elusive
kind of wonder in which informed admiration and a simultaneous sense of
bewilderment blend inextricably.
RE F E RE N CE S
Asper, M. 2017. ‘Machines on Paper: From Words to Acts in Ancient Mechanics’,
in M. Formisano and P. J. van der Eijk, eds., Knowledge, Text and Practice in
Ancient Technical Writing (Cambridge), 27–52.
Berrey, M. 2017. Hellenistic Science at Court (Berlin).
(Cambridge).
2020. ‘Ancient Greek Mechanics and the Mechanical Hypothesis’, in L. Taub,
ed., The Cambridge Companion to Ancient Greek and Roman Science
Brain, P. 1986. Galen on Bloodletting: A Study of the Origins, Development and
Validity of His Opinions, with a Translation of the Three Works (Cambridge).
Brock, A. J. 1916. Galen: On the Natural Faculties (Cambridge, MA).
Craik, E. 2006. Two Hippocratic Treatises, On Sight and On Anatomy (Leiden).
2015. The ‘Hippocratic’ Corpus: Content and Context (New York).
2017. ‘Teleology in Hippocratic Texts: Clues to the Future?’, in J. Rocca, ed.,
Teleology in the Ancient World: Philosophical and Medical Approaches
Cuomo, S. 2007. Technology and Culture in Greek and Roman Antiquity
(Cambridge).
2008. ‘Technology’, in J. P. Oleson, ed., The Oxford Handbook of Engineering
and Technology in the Classical World (Oxford), 15–34.
2018. ‘Greek Mechanics’, in A. Jones and L. Taub, eds., The Cambridge History
of Science: Volume i: Ancient Science (Cambridge), 449–67.
76
I am referring here to the notion of technikê phusis which Galen repeatedly attributes to
Erasistratus – though, as we have seen, in a context where teleology is present only in
a circumscribed sense.

Cuypers, M. 2010. ‘Historiography, Rhetoric, and Science: Rethinking a Few
Assumptions on Hellenistic Prose’ in J. J. Clauss and M. Cuypers, eds.,
A Companion to Hellenistic Literature (Chichester), 317–36.
Dawkins, R. 2000. Unweaving the Rainbow: Science, Delusion and the Appetite for
Wonder (Boston).
Erginel, M. M. 2019. ‘Plato on Pleasures Mixed with Pains: An Asymetrical
Account’, OSAP, 56: 73–122.
Fisher, N. 2017. ‘Demosthenes and the Use of Disgust’, in D. Lateiner and
D. Spatharas, eds., The Ancient Emotion of Disgust (Oxford), 103–24.
Flemming, R. 2005. ‘Empires of Knowledge: Medicine and Health in the
Hellenistic World’, in A. Erskine, ed., A Companion to the Hellenistic
2009. ‘Demiurge and Emperor in Galen’s World of Knowledge’, in C. Gill,
T. Whitmarsh and J. Wilkins, eds., Galen and the World of Knowledge
Frede, M. 1988. ‘The Empiricist Attitude Towards Reason and Theory’ in
R. J. Hankinson, ed., Method, Medicine and Metaphysics: Studies in the
Philosophy of Ancient Science. Apeiron, 21: 70–97.
Furley, D. 1989. Cosmic Problems: Essays on Greek and Roman Philosophy of Nature
(Cambridge).
Garofalo, I. 1988. Erasistrati Fragmenta (Pisa).
Gelber, J. 2021. ‘Teleological Perspectives in Aristotle’s Biology’, in S. M. Connell,
ed., The Cambridge Companion to Aristotle’s Biology (Cambridge), 97–113.
Gleason, M. 2009. ‘Shock and Awe: The Performance Dimension of Galen’s
Anatomy Demonstrations’, in C. Gill, T. Whitmarsh and J. Wilkins, eds.,
Galen and the World of Knowledge (Cambridge), 85–114.
Glucklich, A. 2001. Sacred Pain: Hurting the Body for the Sake of the Soul (Oxford).
Grant, R. M. 2011. Miracle and Natural Law in Graeco-Roman and Early Christian
Thought (Eugene, OR).
Gundert, B. 1992. ‘Parts and Their Roles in Hippocratic Medicine’, Isis, 83:
453–65.
Hankinson, R. J. 1995. ‘The Growth of Medical Empiricism’, in D. Bates, ed.,
Knowledge and the Scholarly Medical Traditions (Cambridge), 60–83.
1997. ‘Hellenistic Biological Sciences’, in D. Furley, ed., From Aristotle to
Augustine: Routledge History of Philosophy, vol. 2 (London), 320–55.
1998a. Galen: On Antecedent Causes, Edited with an Introduction, Translation
and Commentary (Cambridge).
1998b. Cause and Explanation in Ancient Greek Thought (Oxford).
Hett, W. S., ed. 1936. Aristotle: Minor Works (Cambridge, MA).
Hegele. A. 2022. Romantic Autopsy: Literary Form and Medical Reading (Oxford).
in Ancient Greece (Princeton, NJ).
2014. ‘Proto-Sympathy in the Hippocratic Corpus’, in J. Jouanna and M. Zink,
eds., Hippocrate et les hippocratismes: médécine, religion, société (Paris), 123–38.

2018. ‘Body’, in P. E. Pormann, ed., The Cambridge Companion to Hippocrates
Johansen, T. K. 2020. ‘From Craft to Nature: The Emergence of Natural
Teleology’, in L. Taub, ed., The Cambridge Companion to Ancient Greek
and Roman Science (Cambridge), 102–20.
Johnson, M. R. 2017. ‘Aristotelian Mechanistic Explanation’, in J. Rocca, ed.,
Teleology in the Ancient World: Philosophical and Medical Approaches
Kazantzidis, G. 2017. ‘Empathy and the Limits of Disgust in the Hippocratic
Corpus’, in D. Lateiner and D. Spatharas, eds., The Ancient Emotion of
Disgust (Oxford), 45–68.
2018. ‘Medicine and the Paradox in the Hippocratic Corpus and beyond’, in
M. Gerolemou, ed., Recognizing Miracles in Antiquity and Beyond (Berlin),
31–61.
Kazantzidis, G., ed. 2019. Medicine and Paradoxography in the Ancient World
(Berlin).
Kennedy, D. 2007. ‘Making a Text of the Universe: Perspectives on Discursive
Order in the De Rerum Natura of Lucretius’, in M. Gale, ed., Oxford Readings
in Lucretius (Oxford), 376–96.
Krafft, F. 1967. ‘Die Anfänge einer theoretischen Mechanik und die Wandlung
ihrer Stellung zur Wissenschaft von Natur’, in W. Baron, ed., Beiträge zur
Methodik der Wissenschaftsgeschichte (Wiesbaden), 12–33.
1973. ‘Kunst und Natur: Die Heronische Frage und die Technik in der klas-
sischen Antike’, Antike und Abendland, 19: 1–19.
Lennox, J. G. 1985. ‘Theophrastus on the Limits of Teleology’, in
W. W. Fortenbaugh, P. M. Huby and A. A. Long, eds., Theophrastus of
Eresus: On His Life and Work (New Brunswick), 146–63.
2001. Aristotle: On the Parts of Animals i–iv (Oxford).
2018. ‘Aristotle, Dissection, and Generation: Experience, Expertise, and the
Practices of Knowing’, in A. Falcon and D. Lefebvre, eds., Aristotle’s
Generation of Animals: A Critical Guide (Cambridge), 249–72.
Leunissen, M. 2010. Explanation and Teleology in Aristotle’s Science of Nature
(Cambridge).
2021. ‘Empiricism and Hearsay in Aristotle’s Zoological Collection of Facts’, in
S. M. Connell, ed., The Cambridge Companion to Aristotle’s Biology
Lloyd, G. E. R. 1973. Greek Science after Aristotle (London).
1975. ‘Alcmaeon and the Early History of Dissection’, Sudhoffs Archiv, 59:
113–47.
1978. Hippocratic Writings (Harmondsworth).
Longrigg, J. 1993. Greek Rational Medicine: Philosophy and Medicine from Alcmaeon
to the Alexandrians (London).
Lonie, I. M. 1973. ‘The Paradoxical Text ‘On the Heart”’, Medical History, 17: 1–15
and 136–53.
MacDonald, H. 2005, Human Remains: Dissection and Its Histories (New Haven).

Malland, L. R. 2022, ‘Her Body, His Evidence: Female Subjugation in the Early
Modern Anatomy Theaters’, in L. R. Malland, ed., The Spaces of Renaissance
Anatomy Theater (Wilmington, DE), 89–108.
Mayor, A. 2018. Gods and Robots: Myths, Machines, and Ancient Dreams of
Technology (Princeton, NJ).
Neer, R. 2010. The Emergence of the Classical Style in Greek Sculpture (Chicago).
Netz, R. 2020. Scale, Space and Canon in Ancient Literary Culture (Cambridge).
Nightingale, A. 2004. Spectacles of Truth in Classical Greek Philosophy: Theoria in
Its Cultural Context (Cambridge).
Nussbaum, M. 1986. The Fragility of Goodness: Luck and Ethics in Greek Tragedy
and Philosophy (Cambridge).
Nutton, V. 2004. Ancient Medicine (London).
Oleson, J. P. 1984. Greek and Roman Mechanical Water-Lifting Devices: The History
of a Technology (Toronto).
Park, K. 2006, Secrets of Women: Gender, Generation, and the Origins of Human
Dissection (New York).
Polito, R. 2006. ‘Matter, Medicine, and the Mind: Asclepiades vs. Epicurus’,
OSAP, 30: 285–335.
Rihl, T. E. 2018. ‘Mechanics and Pneumatics in the Classical World’, in
P. T. Keyser and J. Scarborough, eds., The Oxford Handbook of Science and
Medicine in the Classical World (Oxford), 337–59
Rocca, J. 2003. Galen on the Brain: Anatomical Knowledge and Physiological
Speculation in the Second Century ad (Leiden).
Roby, C. 2016. Technical Ekphrasis in Greek and Roman Science and Literature
(Cambridge).
Russo, L. 2004. The Forgotten Revolution: How Science Was Born in 300 bc and
Why It Had to Be Reborn (transl. S. Levy) (Berlin).
Scharle, M. 2015. ‘“And These Things Follow”: Teleology, Necessity, and
Explanation in Aristotle’s Meteorologica’, in D. Ebrey, ed., Theory and
Practice in Aristotle’s Natural Science (Cambridge), 79–99.
Schironi, F. 2018. The Best of the Grammarians: Aristarchus of Samothrace on the
Iliad (Ann Arbor, MI).
Sedley, D. 2017. ‘Socrates, Darwin, and Teleology’, in J. Rocca, ed., Teleology in
the Ancient World: Philosophical and Medical Approaches (Cambridge), 25–42.
Steiner, D. 2001. Images in Mind: Statues in Archaic and Classical Greek Literature
and Thought (Princeton, NJ).
Thomas, J. J. 2021. Art, Science, and the Natural World in the Ancient
Mediterranean, 300 bc to ad 100 (Oxford).
Thumiger, C. 2017. A History of the Mind and Mental Health in Classical Greek
Medical Thought (Cambridge).

Tieleman, T. 2013. ‘Miracle and Natural Cause in Galen’, in S. Alkier and
A. Weissenrieder, eds., Miracles Revisited: New Testament Miracles Stories
and their Concepts of Reality (Berlin), 102–14.
Alexandriaʼ, Studies in History and Philosophy of Science, 34: 443–66.
Van der Eijk, P. J. 2000. ‘Aristotle’s Psycho-physiological Account of the Body–
Soul Relationship’, in P. Potter and J. P. Wright, eds., Psyche and Soma:
Physicians and Metaphysicians on the Mind–Body Problem from Antiquity to
Enlightenment (Oxford), 57–78.
2005. Medicine and Philosophy in Classical Antiquity: Doctors and Philosophers on
Nature, Soul, Health and Disease (Cambridge).
Vegetti, M. 1993. ‘I nervi dell’ anima’, in J. Kollesch and D. Nickel, eds., Galen und
das Hellenistische Erbe: Verhandlungen des IV. Internationalen Galen-
Symposiums (Stuttgart), 63–77.
1998. ‘Between Knowledge and Practice: Hellenistic Medicine’, in
M. D. Grmek, ed., Western Medical Thought from Antiquity to the Middle
Ages (transl. A. Shugaar) (Cambridge, MA), 72–103.
1999. ‘Historiographical Strategies in Galen’s Physiology’, in P. J. van der Eijk,
ed., Ancient Histories of Medicine: Essays in Medical Doxography and
Historiography in Classical Antiquity (Leiden), 383–95.
Vernant, J.-P. 1983. ‘Some Remarks on the Forms and Limitations of
Technological Thought among the Greeks’, in J.-P. Vernant, Myth and
Thought among the Greeks (transl. J. Lloyd) (London), 279–301.
Von Staden, H. 1975. ‘Experiment and Experience in Hellenistic Medicine’, BICS,
22: 178–99.
1989. Herophilus: The Art of Medicine in Early Alexandria. Edition, Translation
and Essays (Cambridge).
1992. ‘The Discovery of the Body: Human Dissection and Its Cultural Contexts
in Ancient Greece’, The Yale Journal of Biology and Medicine, 65: 223–41.
1996. ‘Body and Machine: Interactions Between Medicine, Mechanics, and
Philosophy in Early Alexandria’, in J. Walsh and T. F. Reese, eds., Alexandria
and Alexandrianism (Malibu, CA), 85–106.
1997. ‘Teleology and Mechanism: Aristotelian Biology and Early Hellenistic
Anatomy’, in W. Kullmann and S. Föllinger, eds., Aristotelische Biologie:
Intentionen, Methoden, Ergebnisse (Stuttgart), 183–208.
2007. ‘Physis and Technē in Greek Medicine’, in B. Bensaude-Vincent and
Walker, J. 2000. ‘Pathos and Katharsis in “Aristotelian” Rhetoric: Some
Implications’, in A. G. Gross and A. E. Walzer, eds., Rereading Aristotle’s
Rhetoric (Carbondale, IL), 74–92.

Whitmarsh, T. 2011. Narrative and Identity in the Ancient Greek Novel: Returning
Romance (Cambridge).
Wilson, A. I. 2008. ‘Machines in Greek and Roman Technology’, in J. P. Oleson,
ed., The Oxford Handbook of Engineering and Technology in the Classical
Witt, C. 2015. ‘In Defense of the Craft Analogy: Artifacts and Natural Teleology’,
in M. Leunissen, ed., Aristotle’s Physics: A Critical Guide (Cambridge),
107–20.

chapter 10
The Mechanics of the Heart in Antiquity

Matteo Valleriani
Erasistratus (304–250 bce) is considered the first scientist to conceive of

the heart as a pump. This view is based in large part on Erasistratus’
recognition of the four valves of the heart and their functions (Garofalo
1988, 38).1 According to all interpreters of Erasistratus’ fragments, there is
no doubt that he considered the movement of the left ventricle responsible
for filling the arteries and the aorta with pneuma (having received it from
the lungs). Less clear from the fragments is the supposed function of the
movement of the right ventricle, which in principle, if we follow the
previous assertion, should be responsible for filling the veins with blood.2
Even more than these fragments, Galen’s (129–ca. 210 ce) references to
the work of his predecessor form our knowledge of Erasistratus’ doctrines.
Throughout his life, Galen discussed many aspects of Erasistratus’s medical
doctrines, and references can be found in many of his works. It is Galen who
helps us understand the novelty of Erasistratus’ conception. In his De placitis
Hippocratis and Platonis, for instance, he refers to Erasistratus when he
writes, “For he says that matter does not flow in of its own accord as into
some lifeless receptacle, but the heart itself expands like blacksmiths’ bellows
and draws in it” (de Lacy 1984, vi, 6, 9–10). Thus, Galen identifies
a parallelism between the heart and a technological device – the blacksmiths’
bellows – in the doctrine of Erasistratus, without, however, giving any
historiographical hint as to the relation between medical and anatomical
knowledge on one side and the developments in mechanics on the other.3
According to a historical reconstruction initiated in the 1970s by Heinrich
von Staden, Erasistratus’ idea is usually associated with developments in the
field of mechanics from the same period, and in particular with the textual and
1
For the analysis of the fragments of Erasistratus, the present work relies on Garofalo 1988.
2
For an introduction to the interpretative problems concerned with the function of the right ventricle
according to Erasistratus, see von Staden 1973, 213.
3
For Galen’s interpretation of Erasistratus’ conception of the heart and its relation to the Pneumatic
School, see Lonie 1973a, 8.
245

246 matteo valleriani
perhaps diagrammatic codification of the construction and functioning of the
hydraulic hand pump by Ctesibius (285–222 bce). The association between
mechanics and anatomy is conceived of, in line with Galen’s references, as
a parallelism between developments in the field of mechanics (and specifically
pneumatics) on one side and developments in the fields of anatomy and
physiology on the other.4
Von Staden laid out his interpretation of Erasistratus’s heart model exten-
sively in “The Alexandrians,” a chapter of a book entirely dedicated to the
history of conceptions of the heart and the vascular system in antiquity (von
Staden 1973).5 According to von Staden, Erasistratus understood the action of
the left ventricle of the heart as an air pump: The heart worked by sending
pneuma into the arteries and the aorta, which accordingly were not believed to
contain blood (von Staden 1973, 181). After having taken into consideration all
interpretative difficulties concerned with the function of the right ventricle,
von Staden concludes that Erasistratus understood the heart as a kind of twin-
cylinder pump (von Staden 1973, 197).
In von Staden’s analysis, the appearance of the physical principle of horror
vacui in Erasistratus’ doctrines serves as the evidential basis for associating his
work with the mechanical pump. This principle is not only invoked to explain
the functioning of the pump but also to explain why blood comes out from the
opened arteries of a living being in spite of the theory that arteries contain only
pneuma.6 What is more relevant concerning the principle of horror vacui is the
fact that, according to von Staden’s interpretation, Erasistratus’ conception of
void is equivalent to that of Strato of Lampsacus, which was in turn cited in the
Pneumatika by Hero of Alexandria (ca. 10–ca. 70) around three centuries later
(von Staden 1973, 203). Moreover, as Diels already pointed out in 1893, Strato’s
physical system was most probably equivalent to that of Ctesibius (Diels 1893).
For all these circumstantial reasons, von Staden concludes that the heart
model of Erasistratus must have been related to the model of the mechan-
ical pump as developed by Ctesibius. In 1990, for instance, he wrote:
Erasistratus . . . recognised the function of the heart as a pump . . . He
apparently depicted the heart as two-stroke suction- and force-bellows,
4
For an extensive study concerned with Hellenistic scientific and technological developments, see
Russo 2004. For a specific study concerned with hydrostatics and pneumatics, see Valleriani 2016.
5
In the same year, the same interpretation was expressed by I. M. Lonie too, though as a secondary
point in his argument, which instead aimed at dating the text “On the Heart.” For more information,
see Lonie 1973b, 138. More recently, the same argument, though in the context of a more general
debate on the relation between technology and theoretical developments in antiquity, has been
expressed in Webster 2014.
6
For a detailed discussion concerning strategies to avoid objections related to the idea that arteries do not
contain blood, especially by the followers of Erasistratus, see Garofalo 1988, 38–44; von Staden 1973.

The Mechanics of the Heart in Antiquity 247
whose double action alternately (a) pushes and thrusts pneuma through the
arterial system from the left ventricle and (b) somehow or other moves blood
through the veins from the right ventricle. . . . whenever the left ventricle of
the heart contracts, it expels pneuma into the aorta, causing the vessel to
dilate. Since the left ventricle is empty after contraction, pneuma from the
lungs moves into it via the pulmonary vein (in accordance with what later
became known as the horror vacui principle), this dilating the ventricle once
more. The cycle continues, the systole of the left ventricle of the heart
apparently always coinciding with the diastole of the arteries, and vice versa.
(von Staden 1990, 386)
In 1997, von Staden revisited the topic and wrote:
Like Erasistratus’ version of the heart, Ctesibius’ water pump is two-
chambered, and both the heart and the water pump are equipped with
valves to ensure the irreversibility of the flow. More specifically, in
Erasistratus’ model of the heart, as in Ctesibius’ pump, there are four
sets of valves, two controlling the intake, two regulating outflow from the
two chambers. Furthermore, Ctesibius uses valves to ensure the irreversi-
bility of the flow of either liquid or air . . . and Erasistratus likewise
describes the heart valves as enduring the uni-directional flow of either
air (pneuma, “breath”) or liquid (blood): two sets of cardiac valves, he
says, control the flow of pneuma (respectively into the left chamber of the
heart from the lungs and out of this left chamber into the aorta), while
two other sets of valves ensure uni-directional flow of blood into and
from the right cardiac chamber. In addition, Ctesibius’ water pump has
forked pipes, and Erasistratrus’ vascular system is similarly dependent on
forking vessels. Similarly, both Erasistratus’ version of the heart and
Ctesibius’ pump centrally depend on the principle of an intermediate
valved chamber . . . The Erasistratean heart serves as a double intermedi-
ate chamber, on the one hand for pneuma between the lungs and the
aorta and, on the other hand, for blood between the vena cava (coming
from the liver) and the pulmonary vessels that carry blood to the lungs.
Moreover, just as Ctesibius’ pump is constructed with twin cylinders
sitting in a round space, so Erasistratus’ heart is a two-chambered device
that sits in a larger roundish space, the thorax, and just as compression
and expansion alternate in each chamber of Ctesibius’ water pump, so in
Erasistratus’ cardiac “bellows” contraction continuously alternates with
dilatation. . . . it seems evident that Erasistratus’s version of the body has
much in common with early Alexandrian technology. (von Staden 1997,
202–203)
In the following sections, a series of ancient works and fragments from the
Hellenistic period are analyzed in order to establish the extent to which such
a parallelism between early Alexandrian technology and Erasistratus’

mechanical model of the heart as a pump is historically justified in light of
the level of scientific reflection at Erasistratus’ time. By way of an analysis of
archeological findings for periods as close as possible to the third century
bce, it is here shown that, contrary to what is usually assumed, such
a parallelism cannot be demonstrated, and that, contrary to what has been
assumed, it is medical knowledge that may in fact have influenced and
stimulated Hellenistic developments in the fields of mechanics, and not
vice versa.
Ctesibius, Vitruvius, or Hero of Alexandria?

The first problem of von Staden’s narrative is the fact that no single
document or historical source that can be ascribed to Ctesibius has reached
us. Von Staden’s argument is based on the earliest available description of
Ctesibius’ water pump, which was in fact furnished by Vitruvius (died 15
bce) (Figure 10.1) in his famous De architectura, written at least 250 years
later. Vitruvius describes the device in the following way (10.7.1–3):
Now it remains to demonstrate Ctesibius’ machine, which conducts water
to the height. It should be made of bronze. At its roots it has twin cylinders,
standing slightly apart, with pipes that connect together in the figure of
a fork, running together into a tank placed between them. In this tank disk
valves should be closely fitted over the upper outlets of the pipes, so that
when they are blocking these outlets they will prevent the escape of whatever
water has been pushed into the tank by pressure. Above the tank a hood
rather like an inverted funnel is fitted and secured to the tank by a clasp with
a wedge through it, so that the pressure of the incoming water will not raise
it. Above this, a pipe called “trumpet” should be set up, fitted in at the very
top of the machine. The cylinders also have disk valves installed above the
openings at the bottom of the pipes. Hence, from above, pistons, turned and
finished on the lathe and worked with oil, terminating in armatures and
levers, compress whatever air is present there [in the cylinders] along with
the water from above, with the valves obstructing the mouths of the pipes
the pressure of the pistons pushes the water on through the outlets of the
pipes and into the tank, where a little extra pressure is added, and finally
forces it out upward through the “trumpet.” By this means, from a reservoir
in a lower place, water may be supplied for a fountain jet.7
The stated function of the machine is to lift water (in this case, for running
a fountain), but Vitruvius never really specifies why or on which occasions
one should choose it over the other existing devices used at that time to lift
7
Translation in Rowland, Howe, and Dewar 1999, 125.

Figure 10.1 Visual display of the twin-cylinder force pump as described by Vitruvius.
The visual scheme is based on early modern reconstructions. From Vitruvius,
Frontinus, and Giocondo 1513, 183 v. Courtesy of the Library of the Max Planck
Institute for the History of Science
water. The fact that the device is made of bronze suggests a certain
robustness and the possibility of reaching a high pressure (and thus
powering a stronger water jet). The pump is made up of two cylinders
that meet in a middle space, while two pairs of valves, one at the bottom of
the cylinders and the other where the pipes converge on the middle
chamber, facilitate the flow of water. There is an inverted funnel at the
top of the pipe from which water exits, which seems to suggest that the exit
pressure was quite high. Finally, a rod connects the pistons and acts as
a lever handled by an externally applied force.
Vitruvius’ description of the construction and functioning of the water
pump is indeed extremely similar to von Staden’s description of
Erasistratus’ model, and Vitruvius clearly ascribed it to Ctesibius.
A further testimony concerning the twin-cylinder water pump, written
shortly after Vitruvius’ work, can be found in the Pneumatics of Hero of

Alexandria.8 This is all the more relevant, as von Staden’s argument relies
on the fact that Hero’s conception of void, as expressed in the Proemium of
the same work, is the same as that of Erasistratus.
Hero begins his description with the statement that the force pump is
used to extinguish fires (Figure 10.2). The construction details as well as
most of the details concerned with its functioning are more or less equiva-
lent to those of Vitruvius. The most interesting aspect here is related to the
fact that Hero’s work also contains a chapter on the valve itself, the use of
Figure 10.2 Visual display of the twin-cylinder force pump as described by Hero of
Alexandria. The visual scheme is based on early modern reconstructions. From Hero
of Alexandria and Woodcroft 1851, 44.
8
See Woodcroft 1851, 1. Although it is still uncertain when Hero of Alexandria was active, there is
widespread agreement that it was some time during the first century ce.

which is also treated with greater detail and precision in the general
explanation of the pump:
Now, if the cylinders . . . be plunged into a vessel containing water, I J U Z, and
the beam A0 A0 be made to work at its extremities A0 , A0 , which move alternately
about the pin D, the pistons, as they descend will drive out the water through
the tube E0 S0 and the revolving mouth M0 . For when the piston MN ascends it
opens the aperture T, as the cup WY rises, and shuts the valve R; but when it
descends it shuts T and opens R, through which the water is driven and forced
upwards. The actions of the other piston, K L, is the same.9
Hero’s water pump is mechanically more elegant: It integrates the mouth
of the ejecting pipe with a rod that works as a lever; it is equipped with sorts of
filters at the bottom of the cylinders; and it has only one valve in the collection
chamber in the middle (at the bottom). More explicitly than in Vitruvius’
text, Hero’s description particularly focuses on the fact that this device works
with high air pressure. Hero even suggests using pipes produced specifically
for this particular device. It can therefore be inferred that technical develop-
ments in machine building allowed for a more hermetic effect, and thus for
the creation of higher pressure and the construction of devices that might have
actually been useful for extinguishing fires.
The technological innovation of the valve and especially its textual and
diagrammatic codification are the fundamental aspects of the argument for
the equivalence between Erasistratus’ model and the actual mechanical pump.
However, in spite of the presumed equivalence between the cutting edge of
mechanics in the third century bce and that of the first century ce, the
possible role of such a temporal gap should first be further investigated, this
time on the basis of archeological findings.10
The Diffusion of Water Pumps

Using a census of all archeological findings (in the Database of the Ancient
Water Technologies)11 concerned with various kinds of water-lifting machines
in antiquity, it is possible to gain an idea of the real diffusion of such
9
Translation in Woodcraft 1851, 45.
10
Von Staden’s uncritical assumption that Vitruvius’ description of the Ctesibian pump really refers to
the device developed by Ctesibius and not to the force pump, which had also been in development
up until Vitruvius’ time, has been accepted by Lonie 1973b and Webster 2014, 78–79.
11
The Database of the Ancient Water Technologies, curated by Gül Sürmelihindi, is accessible at
https://drupal.mpiwg-berlin.mpg.de/watermachines (last accessed April 21, 2021). There is extensive
literature on Roman water technology. In this work, Oleson 1984 in particular has been used. For
further reading, see also Hodge 1983, 2005 and Wikander 2000.

technology as a material artifact. The database includes artifacts from the
Archaic period through the fifth century ce and contains 159 archeological
findings. The lion’s share of findings date to the period between the beginning
of the first century bce and the end of the second century, a range of about 300
years. For the period in question, the findings lie within an area that stretches
from present-day Portugal to the eastern Mediterranean region and from
present-day England to Egypt, with earlier findings situated in regions lying
further east.
If, however, only force pumps are considered, the results are much less
spectacular. A total of only twenty-one ancient pumps populates our
museums and sites. The span of time covered by their dates corresponds
exactly to the peak period of production of water-lifting devices in general,
and therefore to the aforementioned 300-year period. However, if the
materials constituting the pump are also taken into consideration, the
number decreases even more. Of the twenty-one pumps, fifteen are consti-
tuted of mixed materials, primarily lead and wood. Only six of them are
made of bronze. Obviously, these numbers do not correspond to the total
number of water pumps effectively built and used in antiquity, but consid-
ering the fact that wooden devices are more likely to disappear over time, it
can be inferred circumstantially that the diffusion of twin-cylinder force
pumps made of bronze, as described by Vitruvius and Hero of Alexandria,
was quite limited and by far not the first choice for the task of lifting water.
We know, for instance, from the Syrian-Greek engineer Apollodorus of
Damascus (50–130 ce) that the use other contrivances instead of force
pumps was often suggested, especially for extinguishing fires. In particular,
cow intestines and hollow reeds (“like those of the birdcatchers”) were
considered a useful alternative.12 According to John Peter Oleson, know-
ledge concerning and related to the force pump was spread over signifi-
cantly large geographic areas, but, at the same time, robust devices could
not be built in most of these areas due to a weakness in production
management and the distribution of technical devices (Oleson 1984, 29).
The peak of the diffusion of bronze pumps indicated by an analysis of the
archeological findings might ultimately have been due to the Roman
regulations of the period concerning fire security. Domitius Ulpianus
(died 228 ce), for instance, confirms in his Ad sabinum (book xx, par.
18) that pump-like devices (similar to siphons), which were used as
hydrants, were much more widespread in the second century ce than in
previous times. In particular, such devices belonged to the fire-safety
12
See Whitehead 2010, 174.

equipment mandatory for all great and applicable buildings (Scott 1932,
Vol. 7).
The geographic diffusion of the bronze force pumps that have reached
us is more limited and covers only the Iberian peninsula, northern France,
England, and central Europe. It roughly corresponds to the great mining
sites of the Roman empire, which suggests a possible different use of the
device aside from extinguishing fires. A particularly interesting finding is
the bronze force pump found at Sotiel Coronada (today Valverde del
Camino) in Portugal (Figure 10.3) and preserved at the Museo
Arqueológico Nacional of Madrid. The pump is dated to the first
or second century ce and was discovered in a Roman copper mine in
1889. Thanks to the detailed analysis of Oleson (Oleson 1984, 268–269), it
is now clear that the pump was operated by hand, using a handle posi-
tioned at its base. The cylinders are made of bronze and protected against
corrosion by a hydrous silicate of zinc. In total, the pump consists of twenty-
six pieces of bronze, the biggest ones being the cylinders with a height of
Figure 10.3 Bronze force pump of Sotiel coronada (Valverde del Camino, Portugal),
first century ce. Museo Arqueológico Nacional of Madrid. Photo by Elena Paulino
Montero, 2018.

27 cm. The cylinders are plugged at the bottom by means of spindle valves; the
valves would have moved up and down within a cage fixed inside the
cylinders. There is a collection chamber for water in the middle, which two
single valves would have opened into. The pistons within the cylinders are
connected by a rod, which must have been placed atop the tall pipe to jet the
water. According to modern calculations, the pump must have been able to
deliver about 1.5 liters of water per cycle (up and down once for each cylinder).
This pump was apparently conceived in order to manage the displace-
ment and the direction of the water jet very precisely, that is, by adjusting
both the angle and elevation of the jet. This aspect, together with the
information concerning the site in which the pump was found, led Oleson
to the conclusion that the pump might have been used for cooling and
therefore splitting rocks previously heated by fire within the mine – in
other words, that the pump was a mining tool.
While the analysis of the few remaining bronze force pumps certainly
testifies to a surprisingly perfect match between the textual and diagram-
matic sources from around the first century ce and the actual technological
realizations, this can be taken as supportive of von Staden’s argument only
in a limited manner, because the findings themselves do not allow one to
go back in time far enough to approach the period of Erasistratus himself.
Another way to explore von Staden’s argument is to investigate the use
of the force pump as a model to explain natural phenomena on a more
general basis, not only limited to physiology or anatomy. This approach,
however, cannot prove outright the supposed parallelism between devel-
opments in mechanics and in physiology and anatomy during the third
century bce, but it would increase the plausibility of such an argument.
The Force Pump as Explanans of Natural Phenomena

The force pump and pneumatic-mechanical devices in general were indeed
used at times to explain natural phenomena, though mostly (if not only) in
the context of what we would nowadays call geology. But once again, the
available textual testimonies for such uses all date to the first century ce.
In his Natural History, Pliny the Elder (died 79 ce) states in his descrip-
tion of the behavior of water streams on and within the mountains (2.166):
We must believe that the great artist, Nature, has so arranged it, that as the
arid and dry earth cannot subsist by itself and without moisture, nor, on the
other hand, can the water subsist unless it be supported by the earth, they are
connected by a mutual union. The earth opens her harbours, while the
water pervades the whole earth, within, without, and above; its veins

running in all directions, like connecting links, and bursting out on even the
highest ridges; where, forced up by the air, and pressed out by the weight of
the earth, it shoots forth as from a pipe, and is so far from being in danger of
falling, that it bounds up to the highest and most lofty places.13
Here, Pliny is investigating how it could be that waters shoot upward at
great heights, in spite of their natural tendency to move downward. In
general, he investigates the relationship between earth and water. To
explain such phenomena, he not only refers to an “artifex” but also to
pneumatic technology. In particular, he refers to the siphon, which,
according to the ancient theoretical systematization of pneumatics, works
according to the fundamental principle of suction by means of which all
devices, including the force pump, are explained.
Pliny the Younger (61–113 ce), moreover, also appears to have been quite
familiar with pneumatic devices, and the force pump specifically. He
mentions the pump as a device for irrigation and as a hydrant. For instance,
in a letter to Trajan he asks for permission to form a group of 150 trained
firemen in Nicomedia (Izmit).14
A further instance of the use of the force pump as a model to explain
natural phenomena can be found in the poem Aetna, dated to the first
century ce.15 The poem, which has as its subject the Sicilian volcano,
furnishes an explanation of volcanic activities (lines 323–8):
the (volcanic) wind feels the impact of the struggle which compresses it,
wraps its own strength within its heavy mass and impels its close-packed
particles through fiery passages. Wherever a path is found, it speeds on,
ignoring any wind that would stay its course, until, driven by the confluent
air-stream, as by so many forcing-pumps, it leaps forth and all over Aetna
discharges itself in blasts of angry fire.16
The volcanic activity of Etna is described here by focusing on the motion of
the internal, subterranean winds, which is compared to the action of the force
pump, presumably with water. While these sources clearly show that the
mechanical explanation of the force pump could and was indeed used as an
13
Translation in Rackham 1952.
14
It is worth noting that the emperor declined Pliny’s request explicitly because he did not want the
knowledge to be transferred to Nicomedia, as this region was considered to be animated by
rebellious tendencies. This reasoning suggests that technical knowledge of pneumatics was believed
to play a key role in maintaining supremacy. See Pliny the Younger, Letters 10.33 and 34.
15
The Aetna forms part of the Appendix Vergiliana (see Hine 1996). Scholars date it in the first century
ce, with a terminus ante quem of 79 ce (Porter 2016, 508). It is also believed that its author shows
himself to be familiar with Seneca’s (died 65 ce) Naturales quaestiones. See Setaioli 2015, 260.
16
Translation in Duff 1934.

explanatory model in natural philosophy, these nevertheless remain tempor-
ally confined to the first century, just like the textual sources directly dealing
with the device as well as the pertinent archeological findings. There is,
however, another resource that is closer to Erasistratus’s time: the treatises of
Philo of Byzantium (ca. 280–220 bce).
A More Realistic View of Erasistratus’ Mechanical Context

Philo lived after Ctesibius and worked mostly in Alexandria. He wrote
a treatise on pneumatics, which has reached us only by way of an Arabic
translation from the fourteenth century (A.S. 3713, Ayasofia Museum,
fourteenth century ce). This work was discovered by Carra de Vaux and
published in 1902 (Philo of Byzantium and Carra de Vaux 1902).17 The fact
that Philo’s work is available only in a later translation prompts legitimate
doubts as to how accurately this translation represents the original treatise
and might be the reason why von Staden completely ignored it as a possible
source.18 A closer analysis of the Arabic Philo, however, might still be
helpful for a better contextualization of Erasistratus’s model.
In chapter 64 of the original manuscript, it is possible to find the only
“pump proposal” of the text (Figure 10.4):
Let there be a strong and solid wooden box, reinforced with beams and
tightened with tar all around. Also, let a square-section wood pipe be made,
which starts in the middle of the box and which also is strongly built; its top
rises above the mouth of the well, so that its total height is one fathom
higher. Let there be provided a wheel with teeth, near the tube’s end. To lift
water, one has the wheel turned and the box raised to above the water level.
It can then be allowed to fall, for which purpose it has a lead weight. When
the box reaches the water, you see water spring from the pipe, with a strong
wind. This continues for some time, until all air has been driven out of the
box. The operation is then repeated.19
This device is designed to lift water out of deep wells without using buckets.
The description is probably corrupted, so it is not possible to assert how the
device functioned with full certainty or clarity.20 The pump is a one-cylinder
17
This section is based on David Prager’s edition published in 1974: Philo of Byzantium and Prager
1974.
18
For Prager’s argument in favor of the interpretation that the Arabic translation corresponds to the
original, see Philo of Byzantium and Prager 1974, 126. The oldest Latin manuscript extant, dated to
the thirteenth century ce, contains only a partial translation and does not mention the pump.
19
Translation in Prager 1974, 230.
20
The most detailed attempt to explain the functioning of this device is Drachmann 1948, 6–11.

Figure 10.4 Pump proposal. Philo of Byzantium, Pneumatics. A.S. 3713, f. 82a,
Ayasofia Museum, fourteenth century. From Philo of Byzantium and
Prager 1974, 229
device that is positioned within a well. Once a certain depth of immersion is

reached, the pipe is filled entirely with water, and by raising the piston up to
a certain point, air would probably enter with enough strength to push up the
water. If this is how the pump functions correctly, then the cylinder would
have to be very close to all sides of the well so that the well is hermetically
closed. The discharge of water would depend on the depth of the installation.
The fact that some doubts about the functioning of this device cannot be
resolved does not factor significantly into the present work. Two relevant
aspects can be stated with certainty: The pump is not a twin-cylinder device,
and its functioning does not involve valves, as in the case of the model under
discussion.21
21
There is another Arabic manuscript, entitled The Extracts of Irun, preserved as Eastern Manuscript
954, Bodleian Library, which was first considered by Carra de Vaux as a further testimony of Philo’s
mechanics and pneumatics. This manuscript does indeed contain the description of a twin-cylinder
force pump. However, more recent studies came to the conclusion that the MS 954 is, in fact, a later

In his treatise on Belopoeica, Philo comes back to the description of
a stone-throwing device powered by air compressed in cylinders by means
of pistons (Marsden 1971, 77.8–78.28, 151–155; Schiefsky 2015). Philo men-
tions here the works of Ctesibius as a basis for certain devices that he later
developed himself. For instance, Philo mentions the water organ.
However, the stone thrower does not make use of any water and, what is
more relevant, does not indicate the use of valves either. Philo mentions
a construction with double cylinders but without communication between
them (through a middle chamber, for instance).
In conclusion, if we consider the historical sources concerned with
mechanics and pneumatics that are older than the first century ce (i.e.,
much closer in time to Erasistratus), the framework that emerges does not
seem to exhibit the parallelism hypothesized by von Staden.
Conclusions
The available sources do not entirely support von Staden’s argument,
which favors a straightforward parallelism between anatomy and physi-
ology on one side and mechanics and pneumatics on the other during the
third century bce. Devices, such as bellows, and contrivances, such as
spindle valves, were certainly already known, as they are mentioned in
Homer’s Iliad. It is also certain, thanks to Philo’s and Vitruvius’ testi-
monies, that Ctesibius independently worked on the development of
devices such as water pumps that made use of air compressions. But the
twin-cylinder pump facilitated by an alternating valve system probably did
not yet exist at the time of Erasistratus.
Instead of continuing with arguments for parallelism, as suggested by
von Staden in his later works as well as in more recent works by other
authors (Webster 2014, 1–8), or of how technological developments
oriented theoretical research, the sources indicate that we need to take
a new look at the dichotomic developments between anatomy and physi-
ology, and mechanics and pneumatics, respectively. As Giouli Korobili
shows in this volume (Chapter 7), analogical arguments that built up
a relationship between mechanical contrivances and the human body
already existed in the fourth century bce and were particularly concerned
collection of achievements of ancient and Hellenistic pneumatics and mechanics in general,

therefore citing many authors and engineers. For this reason, this device is not considered Philo’s
own. For more information, see Prager 1974, 234–237. Carra de Vaux’s wrong ascription of the twin-
cylinder force pump to Philo was used to support the argument developed by von Staden in Lonie
1973b as well. The same is true for the recent dissertation of Colin Webster: Webster 2014, 83.

with the analogical relation between the lungs and the bellow. From an
epistemological point of view, it is worth considering that the two fields –
anatomy and mechanics – are characterized by both a theoretical and
a practical approach: anatomy and machine building on one side, and
physiology and mechanics on the other.22 The realization and, especially,
the codification of knowledge concerning the monocylindric pump in texts
and perhaps even in diagrams as early as Ctesibius’ time, might have given
the impulse to a more theoretical approach. As explained elsewhere, codi-
fied practical knowledge possesses a higher degree of circulation and has
a significantly higher likelihood of being integrated into arguments whose
focus differs from the original one to which the codified knowledge
originally referred. First of all, a written description of an air pump cannot
avoid describing the behavior of air. Even if the description aims to remain
at the practical level as much as possible, by describing the construction
and the usage of the pump for example, it nevertheless needs to provide
some information on the elemental composition of air, if its behavior is to
be understood. In this way, even the most pedestrian description of the
device has to touch the abstract subject of the constitution of matter. The
process of codification turns into a process of theorization, and such
a transformation of knowledge is inherently unavoidable any time practical
knowledge is codified. Once the knowledge has left the workshop in its
codified form, therefore, it can be subsumed to other concepts or inte-
grated into other scientific arguments under the conditions that those
arguments also have reached at least the same level of codification and
theorization. It is at such a theoretical level that arguments can then be
integrated into each other and, for instance, an analogy can be fruitfully
expressed and used in an explanatory narrative. The observation of a device
by a non-expert alone cannot be the historical basis of a relation between
anatomy and theoretical mechanics.
Finally, it seems more plausible that the diffusion of knowledge con-
cerning the compressibility of the air, as well as its material realizations in
the form of monocylindric pumps, had indeed influenced the arguments of
those analyzing the behavior of the heart such as Erasistratus. But
Erasistratus was dealing with an extraordinarily more complicated
“device,” contained within an even more difficult body of “devices,” and
especially within a highly elaborated and well-developed theoretical med-
ical framework. It might therefore be necessary to turn the argument in
22
For the epistemology of the relation between practical and theoretical knowledge in science, see
Valleriani 2017.

a different direction: Perhaps Erasistratus’ model of the heart influenced
the development of mechanics to such an extent that around the first
century bce a double-cylinder air pump could, in fact, not only be
conceived but also realized.
RE F E RE N CE S
Carra de Vaux, B. 1902. Le livre des appareils pneumatiques et des machines
hydrauliques par Philon de Byzance (Paris).
De Lacy, P. 1984. Galen on the Doctrines of Hippocrates and Plato (Galeni de placitis
Hippocratis et Platonis), translated by Phillip de Lacy. Vol. 2, books vi–ix
(Berlin).
Diels, H. A. 1893. “Über das physikalische System des Straton,” Sitzungsberichte
der Königlich Preussischen Akademie der Wissenschaften zu Berlin, 1: 101–127.
Drachmann, A. G. 1948. Ktesibios, Philon and Heron: A Study in Ancient
Pneumatics (Copenhagen).
Duff, A. M. 1934. Minor Latin Poets, Vol. 1. Loeb Classical Library (Cambridge, MA).
Garofalo, I. 1988. Erasistrati Fragmenta (Pisa).
Giocondo, G., 1513. Vitruvius: De architectura libri x iterum et Frontinus à Iocundi
revisi repurgatique quantum ex collatione licuit (Florence).
Hine, H. M. 1996. “Aetna,” in S. Hornblower and A. Spawforth, eds., The Oxford
Classical Dictionary (Oxford), 31.
Hodge, T. A. 1983. “Siphons in Roman Aqueducts,” Papers of the British School at
Rome, 51: 174–221.
2005. Roman Aqueducts & Water Supply (London).
Lonie, I. M. 1973a. “The Paradoxical Text ‘On the Heart.’ Part 1,” Medical History,
17 (1): 1–15.
1973b. “The Paradoxical Text ‘On the Heart.’ Part 2,” Medical History, 17 (2):
136–153.
Marsden, E. W. 1971. Greek and Roman Artillery: Technical Treatises (Oxford).
Oleson, J. P. 1984. Greek and Roman Mechanical Water-Lifting Devices: The History
of a Technology (Toronto).
Porter, J. I. 2016. The Sublime in Antiquity (Cambridge).
Prager, F. D. 1974. Philo of Byzantium: Pneumatica (Wiesbaden).
Rackham, H. 1952. Pliny: Natural History (reprint 1983) (Cambridge, MA).
Rowland, I. D., T. N. Howe, and M. J. Dewar, eds. 1999. Vitruvius: Ten Books on
Architecture, 10 vols. (Cambridge, MA).
Russo, L. 2004. The Forgotten Revolution: How Science Was Born in 300 bc and
Why It Had to Be Reborn (Berlin).
Schiefsky, M. 2015. “Technē and Method in Ancient Artillery Construction: The
Belopoeica of Philo of Byzantium,” in H. Brooke and K. D. Fischer, eds.,
The Frontiers of Ancient Science: Essays in Honor of Heinrich von Staden
(Berlin), 613–651.
Scott, S. P., ed. 1932. The Civil Law, including the Twelve Tables: The Institutes of
Gaius, the Rules of Ulpian, the Opinions of Paulus, the Enactments of Justinian,

and the Constitutions of Leo. Translated from the Original Latin, Edited, and
Compared with All Accessible Systems of Jurisprudence Ancient and Modern, 17
vols. (Cincinnati).
Setaioli, A. 2015. “Seneca and the Ancient World,” in S. Bartsch and A. Schiesaro,
eds., The Cambridge Companion to Seneca (Cambridge), 255–265.
Valleriani, M. 2016. “Hydrostatics and Pneumatics in Antiquity,” in G. L. Irby,
ed., A Companion to Science, Technology and Medicine in Ancient Greece and
Rome (Hoboken, NJ), 145–160.
2017. “The Epistemology of Practical Knowledge,” in M. Valleriani, ed.,
Structures of Practical Knowledge (Dordrecht), 1–19.
Von Staden, H. 1973. “The Alexandrians,” in C. R. S. Harris, ed., The Heart and
the Vascular System in Ancient Greek Medicine: From Alcmaeon to Galen
(Oxford), 177–233.
1990. “Cardiovascular Puzzles in Erasistratus and Herophilus,” in R. Bernabeo,
ed., Atti del xxxi Congresso Internazionale di Storia della Medicina (Bologna,
30 agosto – 4 settembre 1988) (Bologna), 385–391.
1997. “Teleology and Mechanism: Aristotelian Biology and Early Hellenistic
Medicine,” in W. Kullmann and S. Föllinger, eds., Aristotelische Biologie:
Intentionen, Methoden, Ergebnisse. Akten des Symposions über Aristoteles’
Biologie vom 24.–28. Juli 1995 in der Werner-Reimers-Stiftung in Bad
Homburg (Wiesbaden), 183–208.
Webster, C. 2014. “Technology and/as Theory: Material Thinking in Ancient
Science and Medicine.” Unpublished PhD thesis. University of Columbia.
Whitehead, D. 2010. Apollodorus Mechanicus: Siege-Matters (Poliorkētika)
(Stuttgart).
Wikander, Ö., ed. 2000. Handbook of Ancient Water Technology (Leiden).
Woodcroft, B. 1851. The Pneumatics of Hero of Alexandria (London).

chapter 11
The Mechanics of Galen’s Theory of Nutrition

Orly Lewis
A Note on the Illustrations

This chapter includes seven illustrations which are intended to assist
readers in following the anatomical details depicted in this chapter and
thus in understanding the discussion of Galen’s theory. The illustrations
are the product of the collaborative work between: the author, the illustra-
tor Yotam Giladi, and experts in anatomy Joshua Milgram and Esteban
Marroquín Arroyave of the Koret School of Veterinary Medicine at The
Hebrew University of Jerusalem (HUJI). All are members of the research
group ATLOMY (Anatomy in Ancient Greece and Rome: An Interactive
Visual and Textual Atlas). It was not possible to discuss here in detail the
considerations and sources informing each illustration, but I highlight
a few points:
1. The illustrations are based on close readings of key passages in Galen.
They are visualisations of the body parts as described by Galen in terms
of their general shape, location, connections and mutual proportions.
They are not intended as an exhaustive depiction of his anatomical
understanding of these body parts. They depict the features discussed
in the chapter and additional key properties required for understand-
ing the discussion.
This research has received funding from the European Research Council (ERC) under the European
Union’s Horizon 2020 research and innovation programme (GA 852550 – project ATLOMY). Earlier
stages of the research were funded by the Martin Buber Society of Fellows for the Humanities and
Social Sciences at The Hebrew University of Jerusalem. I am grateful to the editors for their patience
and comments as well as for their warm hospitality during the conference from which this volume
stems. Thank you also to Giouli Korobili, Julius Rocca, Peter Singer and Chiara Thumiger for their
helpful comments and to Dimitry Ezrohi, Premshay Hermon and Nir Propper for their assistance.
A special thanks also to Esteban Marroquín Arroyave, Joshua Milgram and Yotam Gildai for their
assistance in interpreting Galen’s anatomical descriptions and their work in producing the illustrations;
and to Assaf Marom at the Anatomy and Human Evolution Lab and the instructors at the Anatomy
Research and Education Centre, both at the Israel Institute of Technology, for their advice and
practical instruction on the anatomy of the digestive system in humans.
262

The Mechanics of Galen’s Theory of Nutrition 263
2. When Galen’s writings provide insufficient details for an illustration
(e.g. as regards the exact course of the duodenum), we reverted to
modern observations. These observations were either performed by
the author and those advising and producing the illustrations or were
available in printed or digital resources. Galen’s anatomical under-
standing is at times very similar to modern ideas but also differs on
some points. I have noted key differences in the captions accompany-
ing the illustrations and in relevant places throughout the chapter.
3. Galen performed most of his anatomical research on animals. He often
emphasises the differences between them as regards internal anatomy,
although he notes that the digestive organs are generally the same in all
animals on which dissection is practised.1 Despite the non-human
origin of most of Galen’s descriptions, he describes anatomy with
the physician in mind, that is, a clinician treating humans, who
must be familiar with human anatomy. The illustrations present,
therefore, the anatomy on the vertical axis – the upright human
posture – and situate it in a human figure.
Introduction
Nutrition, says Galen, is one of the three main faculties (dunameis), actions
(erga) or activities (energeiai) of nature.2 Growth and genesis are the other
two. These are all faculties which both animals and plants possess. By
contrast, sensation and voluntary motion are faculties of the soul and as
such are unique to ensouled beings and animals.3 Genesis (genesis) is the
alteration and shaping of the underlying material of the animal. Growth
(auxêsis, to auxanesthai) is an addition to the existing (generated) matter
1
Galen, AA 6.3 (Garofalo 2.354–6 = K. 2.547–8). Abbreviations of Galenic works follow those listed in
Hankinson 2008, 391–7, with references to editions both by Karl G. Kühn (K.) and modern editors,
when available.
2
Galen explicitly distinguishes between ἔργον (‘effect’), ἐνέργεια (‘activity’) and δύναμις (‘faculty’),
but as he admits, there is some overlap and his terminology remains flexible (Nat.Fac. 1.2 (K. 2.2, 2.6–
7); cf. PHP 6.1.5–9 (De Lacy 360,22–361,9 = K. 5.506–7)). There is still a need for a close consider-
ation of these terms in Galen (especially of the former two) and the relation between them; for some
discussion, see for example Hankinson 2014, Harari 2016, 212–19, and, more broadly, Corcilius 2015.
3
Gal., Nat.Fac. 1.1, 1.5, 1.9 (K. 2.1, 2.10–1, 2.19–20). In Symp.Diff. 4.1–2 (Gundert 228,1–17 = K. 7.62–
3), Galen lists the following natural ‘activities’ (ἐνέργειαι): appetite (ὄρεξις), digestion (πέψις),
nutrition (θρέψις), distribution (ἀνάδοσις), generation of blood (ἡ τοῦ αἱμάτου γένεσις), pulse
(σφυγμός), separation of residues (διάκρισις τῶν περιττωμάτων) (cf. Nat.Fac. 1.4 (K. 2.9–10)) that
plants do not share. He does not refer to growth and genesis (perhaps since they were less relevant in
a practical diagnostic context). Galen’s classification of the natural faculties/activities and the
hierarchical relation between them requires further study.

264 orly lewis
and entails dimensional extension or distension (diastasis), whereas nutri-
tion (threpsis, to trephesthai) is an addition to matter without such
extension.4 Galen defines nutrition as the ‘assimilation of that which
nourishes to that which is nourished’ (omoiôsis tou trephontos tôi
trephomenôi).5 It entails an alteration of foodstuff (e.g. vegetables, meat,
pulses) into blood and from blood into the different types of the body’s
basic parts such as bone, cartilage, arteries, nerves and so forth (the so-
called homoiomerous or unified parts).
Like other ancient authors, Galen does not think that ingested food
breaks down into carbohydrates, fats, proteins, vitamins and minerals
which are then absorbed and carried by the blood to the different parts
of the body. Blood does not circulate through the body ferrying around the
nutritive components from food (and the waste residues produced during
digestion) like we believe today. Rather, for Galen, digestion is what
produces blood in the first place: blood is produced directly from the
food we eat. Our food becomes the blood itself, and this blood is then
further processed in each of the body’s parts to turn into the particular
substance from which each part is made. Mastication in the mouth and
then churning and heating in the stomach turn the food into a thick fluid.
This fluid flows through the intestines to the liver, where it undergoes
further alteration (‘digestion’), which turns it into the red fluid known as
blood. From the liver, it flows to the other body parts, where it is assimi-
lated to their respective qualities and thereby consumed or used up.6 For this
reason, some Greek and Roman authors refer to blood as the ‘ultimate’ or
‘final’ nutriment since blood itself is the substance from which the parts of
the body are directly nourished. The digestive-nutritive process thus
entails, according to Galen, an alteration of food from solid or semi-solid
foods of varying colours to the red liquid blood and then to the various
colours and solidity of the different parts. The green lettuce, red beetroot
or yellowish cheese all turn into blood before becoming soft red flesh, hard
white bone and any other form of homoiomerous parts.
This process of change and assimilation is, unsurprisingly, long and
elaborate. It requires different body parts, which perform different stages of
the gradual alteration of food. Other parts (e.g. the oesophagus, intestines
and veins) are required in order to transmit the useful (chrêston) nutritive
matter from one digestive (or altering) part to the other and finally to the
4
Gal., Nat.Fac. 1.5 (K. 2.10–1). 5 Gal., Nat.Fac. 1.11 (K. 2.24)
6
For discussion of blood and blood flow in antiquity, see Siegel 1968, Harris 1973, Boylan 1982, 2015.
The main sources for Galen are PHP book 6; Nat.Fac. and UP books 4–5.

parts nourished by it. Other parts (e.g. kidneys, colon, gall bladder and
spleen) are required on account of the residues (perittômata) produced
during the digestive alteration (e.g. urine, faeces and various kinds of bile).
Different parts perform the separation, collection and carrying of these
residues.7
The little scholarship there is on Galen’s ideas concerning digestion
focuses on a particular aspect of the process: the faculties (dunameis) of
attraction (helktikê),8 retention (kathektikê), alteration (alliôtikê) and
expulsion (apokritikê), which are the ‘servants of nutrition’ (hupêretides
threpseos).9 According to Galen, each part possesses these four natural
faculties innately (sumphuton), and it is these faculties which allow each
part to attract to itself the blood that is beneficial to it, retain the matter for
the time required to alter it and expel the material which is alien or harmful
to the part.10 Scholars have emphasised the role of these faculties. For
example, Armelle Debru states that Galen
conceived of digestion as the exercise of natural faculties residing in each of
the different organs through which the food passes, and not as a mechanical
process. . . . the process of nutrition involves the complex action of faculties
residing in the numerous organs which take part in it. Their capacities of
attraction, retention, alteration and expulsion also serve to explain for Galen
the progress of the food through the oesophagus, and the action of the
stomach and the other organs involved.11
C. R. S. Harris stresses particularly the role of the faculty of attraction.
He claims that Galen considered it ‘responsible for most, if not every,
movements of the blood’ and as ‘the real force which distributes the blood
supply’.12 While these faculties are indeed crucial for the digestive-nutritive
process, focusing on them alone overlooks a key part of Galen’s
7
Gal., Nat.Fac. 1.10 (K. 2.20–4).
8
Throughout this chapter, I render ἕλκειν, ὁλκή and cognates interchangeably as pull/pulling,
attraction/traction or draw/drawing in.
9
Gal., Nat.Fac. 3.9 (K. 2.178), compare Nat.Fac. 1.9 (K. 2.20), and see pp. 288–92 below. An
exception in scholarship is Meyer-Steineg 1913, which is still very useful. Studies such as De Lacy
1988, Schiefsky 2012 and Trompeter 2018 on Galen’s conception of the soul, particularly the
desiderative soul, are important for the general framework of Galen’s theory of nutrition, but
they do not discuss his ideas of the actual process and its different stages.
10
Gal., Nat.Fac. 2.3, 2.6, 3.1, 3.9 (K. 2.80, 2.96, 2.143–5, 2.178,1–3); Loc.Aff. 5.8 (K. 8.358). On the
‘innateness’ of these faculties, see below, n. 84. The idea that the body parts function through
faculties is essential in Galen’s physiology and is present in earlier sources too: for example, Hipp.,
Morb.Sac. 16.1 (= 19 Jones) (L. 6.390,10–1 = Jouanna 29,4–5); compare 17.2 (= 20 Jones) (L. 6.392,6–7
= Jouanna 30,5–6); Praxagoras of Cos., frs. 9–10 Lewis (= Gal., Diff.Puls. 4.2 (K. 8.701); PHP 6.7.3
(De Lacy 406,4–8 = K. 5.561)), who refers also to an ‘innate’ faculty with respect to the pulse.
11
Debru 2008, 273–4. 12 Harris 1973, 327, 329.

266 orly lewis
explanation of the process, namely, the perceptible motions of the body
parts involved in digestion. Galen states that ‘anything that is moved
must be either pulled or propelled by something, or conducted (sc.
carried)’ (ἢ γὰρ ἕλκεσθαι χρὴ τὸ μεθιστάμενον ἢ πέμπεσθαι πρός τινος
ἢ παραπέμπεσθαι).13 This is true also for the nutritive matter in its
different stages (including blood). As this chapter will show, the physical
contractions and extensions of parts such as the stomach and intestines
are responsible for the motion of the nutritive matter; it is they that
actively and ‘mechanically’ pull and propel food into and through the
body. Without them, there would be no ‘stream’ of nutritive matter for
the different parts to draw from and feed off. Hence, understanding these
motions and their roles in the digestive-nutritive process is crucial for
understanding Galen’s explanation of this process and his conception of
the nutritive faculties.
This chapter brings to light this essential but overlooked mechanical
part of Galen’s theory. I term these motions mechanical in so far that they
can be paralleled to the working of inanimate, lifeless (dunamis-less)
devices.14 Galen himself refers to the stomach’s motions and their expul-
sive-propulsive effect as a mêchanê (‘device’, ‘contraption’, ‘mechanism’)
and demonstrates that the contraction motions involved in swallowing, for
instance, can occur also in a dead body (see below, n. 31).15 In the living
body, these motions are not entirely mechanical, however. They are
caused, argues Galen, by the four faculties of attraction, retention, assimi-
lation and expulsion, which Nature’s wise design embedded in each part.
Galen seeks to prove this in his work On Natural Faculties. His elaborate
argument in that treatise is directed at those who explain digestion and
nutrition without reference to faculties (in particular Erasistratus and
Asclepiades and their respective followers), and he thus emphasises the
role of the faculties.16 His fierce polemic and emphasis of the faculties has
13
Gal., UP 6.14 (Helmreich 1.350–1 = K. 3.480–1), trans. May, slightly modified.
14
For a similar conception of ‘mechanical’, see Berryman 2007, 365–7; and Gotthelf’s summary of
Berryman’s argument: ‘the term “mechanism” should . . . only (be used) in cases where regular
outcomes are (for the most part) guaranteed by an internal structure analogous in its operation to
a machine’ (Gotthelf 2012, 81). I do not distinguish in this chapter between ‘mechanical’ and
‘mechanistic’ (for the distinction, see for instance Johnson 2017, 127).
15
Gal., Nat.Fac. 2.1 (K. 2.76).
16
The length and tangled nature of Galen’s polemic in themselves emphasise the strongly mechanical
aspect and appearance of these motions: he has to work hard to fit them into a teleological
framework in which unseen faculties are the cause of such motions (and one wonders, at times,
whether his argument is convincing throughout). The difficulty is manifest also in his ‘flexible’ ideas
on the relation of these faculties to the desiderative soul and its anatomical source (ἀρχή), namely,
the liver, on which see below, n. 84.

led scholars to focus on the teleology of Galen’s theory of nutrition and
oppose it to mechanical explanations of this process.
Heinrich von Staden has shown that Galen’s professions of teleological
explanations and his argument against Erasistratus on this account have
‘tended to obscure’ Erasistratus’ teleology;17 however, they have, con-
versely, also obscured the mechanics involved in Galen’s explanation of
the body’s activities. The teleological explanation by means of faculties
endowed by Nature for particular aims is only one level, and one part, of
Galen’s explanation of nutrition (and other physiological processes). On
another and no less essential level, these processes are explicable in terms of
tangible motions of the anatomical parts, which can be perceived by the
senses and not only reason. And while these are a result of a teleological
design, they are also mechanical in practice. Saying that the function – the
telos – of the intestines is to deliver nutritive matter and that this is
performed by means of the expulsive faculty is only part of an explanation
for Galen. He would not be satisfied with anyone who stops there. One
must complete it by reference to the anatomical structure (which is
designed in order to fulfil the function) and just as importantly by refer-
ence to the manner in which this structure fulfils its function. It is on this
fundamental, material level that mechanical motions such as contractions
and expansions are at play in Galen’s theory.
I do not mean to argue in this chapter that Galen’s theory was not
teleological, but rather to shed light on the complexity of his teleology. As
this chapter will show, these motions and Galen’s physiological theory of
digestion-nutrition are an illustrative example of how teleology and mech-
anical explanations go hand in hand in explaining the complex whole
which is the living and active body.18 The meeting point, so to speak, of
teleology and mechanics is the body’s material structure – its anatomy.
Galen is well known for his systematic empirical research into the structure
of the body and its parts.19 The foundation of Galen’s teleological view is
his idea that the parts of the body are shaped and fitted together in
accordance with the activity or function they are assigned (by Nature) in
the body. In other words, in exploring the shape, size, location and texture
of the different parts, one must always ask why nature designed it in such
17
Von Staden 1997, 203.
18
For the close relation between mechanical and teleological explanations, see von Staden 1997
regarding Erasistratus, Berryman 2007, Gotthelf 2015, most neatly on 172, and Johnson 2017
regarding Aristotle.
19
Laid out most clearly in his fifteen-book work On Anatomical Procedures and his ‘minor anatomical
works’ on the dissection of nerves, vessels and muscles. See also Garofalo 1991, 1:5–58, Rocca 2003.

268 orly lewis
a way. He sets out this scheme in detail in the seventeen books of his On the
Usefulness of the Parts.20 From these material qualities, one can conclude
the function and the way in which a part performs it (e.g. which kind of
motion it is capable of).21 It is no surprise, therefore, that anatomy is the
foundation of Galen’s theory of the digestive-nutritive process.
In order to explain Galen’s theory, I follow the anatomical path the
nutritive matter takes from the mouth to the individual parts and explain
how it is moved along this path, both by the mechanical motions of the
parts and the natural faculties embedded in these parts. I discuss, too, the
final stage of digestion, which I call ‘local’ digestion and which takes place
in the individual parts.22 The chapter is concerned with the motion of what
Galen calls the ‘useful’ (chrêston) matter, that is, the substance which
actually serves to nourish the body, as opposed to the residues produced
in the process. I focus, therefore, on the parts through which this useful
nourishing matter passes, and I touch upon the production or motion of
residues and their respective body parts only when this is closely related to
the progression of the useful nutritive matter. In order to make it easier to
follow the discussion of the different stages, I begin with a brief overview of
the anatomical path the nutritive matter takes and its main ‘stations’,
according to Galen.
The Course of Food in the Body according to Galen: An Overview

In the mouth, the teeth break down the food to small pieces, and the
moisture of the saliva (phlegma) and the mouth’s flesh (chrôs) alter it
further.23 From the mouth, the nutritive matter passes through the follow-
ing body parts (I describe the parts as Galen does and with his terminology;
and see Figure 11.1 below for a visual overview):
1. Oesophagus (oisophagos/ stomachos) – this is the ‘pipe’ which extends
down from the larynx at the back of the mouth to the stomach. It ends
20
On Galen’s teleological framework, see, for example, Schiefsky 2007 and the recent overview in
Marechal 2020.
21
On a more fundamental anatomical level of the qualitative ‘chemical’ structure of the part (a
histological level, to use modern terminology), each part is composed from a particular mixture
(κρᾶσις) of the four qualities of hot, dry, cold and wet. This mixture is determined by Nature in
accordance with the function it assigned the part, and the natural healthy functioning of the part is
harmed if the mixture changes. For some discussion of this elaborate and multifaceted theory of
Galen and its practical implications, see Singer and van der Eijk 2019 and van der Eijk 2014, 2015.
22
This stage can be very roughly paralleled to what modern theory calls the metabolic stage of
digestion, which occurs on the cellular level of different parts.
23
Gal., Nat.Fac. 3.7 (K. 2.162–3).

Figure 11.1 An overview of the digestive-nutritive system and the course of food
according to Galen: anterior view showing the mouth (1), oesophagus (2), stomach
(3), intestines (4), intestinal-liver veins (5) culminating in the portal vein (6), liver
(7), the hollow vein (extending down and up from its starting point at the liver) (8),
the heart (9). Veins are depicted in purple and only select veins and branches are
depicted. Some of the parts responsible for receiving and expelling bile and urine are
included in pale green for the sake of orientation (spleen (10), gall bladder (12),
kidneys (13–14), ureters (15–16), bladder (17)). Image by Yotam Giladi and Project
ATLOMY

270 orly lewis
at the ‘mouth of the stomach’ (στόμα τῆς γαστρός / στόμα τῆς
κοιλίας), an orifice in the stomach wall, through which the food enters
the stomach’s cavity. (See ‘(i) From Mouth to Stomach’.)
2. Stomach (gastêr) – a large elongated cavity surrounded by thick walls; it
lies in the left upper part of the abdomen. The stomach begins to change
the food from solid to liquid form. At the opposite end from its ‘mouth’,
there is a second orifice, the pylorus (pulôros). Through it, the stomach
expels the semi-digested food into the intestines. The stomach is also the
instigator of eating, for appetite arises from it: ‘(nature) has granted to
the stomach alone and particularly to the parts near its mouth the
sensation of lack (aisthêsis endeias) which arouses the animal and stimu-
lates it to consume food’.24 (See ‘(i) From Mouth to Stomach’.)
3. Intestines (ta entera) – a long hollow conduit which Galen divides into
six different sections or parts (see below, pp. 280–281). The intestines
perform some further alteration of the nutritive matter, but are mainly
assigned with conducting and pushing the matter forward: the ‘useful’
part into the veins leading to the liver, the residues out of the body in
the form of excrement. (See ‘(ii) From Stomach to Intestines’.)
4. Intestinal-liver veins (phlebes) – veins are hollow pipes which have softer
and thinner walls than arteries. At this stage, the nutritive matter passes
through the veins connecting the intestines and liver: they begin as
numerous narrow veins (‘the veins of the mesentery’, αἱ κατὰ τὸ
μεσεντέριον φλέβες) which merge to become the portal vein (ἡ ἐπὶ
πύλαις φλέψ – lit., ‘the vein which is at the portals’), which enters the
liver at its concave (‘bottom’ or caudal) part. (See ‘(iii) From Intestines to
Liver’.)
5. Liver (hêpar) – a large porous and fleshy organ at the upper right part
of the abdomen. It performs the substantial and final stages of chan-
ging food into blood. (See ‘(iv) From the Liver to the Hollow Vein’.)
6. The hollow vein (koilê phleps) – a large vein which stems from the many
smaller veins in the ‘upper’ (convex/cranial) part of the liver (i.e. the
opposite side from the portal vein). At this part, Galen’s hollow vein
corresponds to the modern inferior vena cava. According to Galen,
from the liver the hollow vein splits: one branch leads up and the other
down (see below, p. 285 for its exact course according to Galen). When
the nutritive matter (by now, blood) exits the liver and enters the hollow
24
Gal., UP 4.7 (Helmreich 1.201,25–202,2 = K. 3.275), trans. May, slightly modified; cf. UP 5.9
(Helmreich 1.277,4–12 = Κ. 3.378). This appetite is transmitted to the brain via the pair of sensory
nerves extending to it (UP 4.7, 16.5, Helmreich 1.203, 2.394–5 = K. 3.278, 4.289; AA 14.10 (Simon 278
= Duckworth, Lyons and Towers 220).

vein, different residues are separated from it and are deposited and
processed further in their respective parts and systems: spleen, gall
bladder and urine system. (See ‘(v) Distribution through the Hollow
Vein and Its Branches’.)
7. Venous system – the nutritive matter proceeds from the hollow vein to
its branches and from them to their smaller branches until it reaches the
smallest veins distributed through each part of the body. Some of the
blood enters the heart, and a small portion of it reaches the left cavity of
the heart (the modern left ventricle), where it is further processed and
sent into the arterial system in the form of fine pneumatised blood. (See
‘(v) Distribution through the Hollow Vein and Its Branches’.)
8. The substance of each part in the body draws to itself nutritive matter
from the small branches of veins which spread through it. (See ‘(vi)
Nourishing the Parts: The Final Stages of Nutrition’.)
Throughout this journey some of the digestive matter is drawn into the
digestive parts themselves (e.g. the stomach, liver) for the sake of their own
nourishment. They draw it into their walls or flesh, which surround the
nutritive matter passing through them. (See ‘(vi) Nourishing the Parts:
The Final Stages of Nutrition’.)
How the Nutritive Matter Moves through the Body
(i) From Mouth to Stomach

The motion of ingested food from the mouth to the stomach is performed,
according to Galen, by means of the fibres (ines) in the walls of the stomach
and oesophagus. As we proceed, we shall see that, for Galen, such fibres in
different digestive parts are the key to the motion of the nutritive matter
throughout the process. The walls of the stomach and oesophagus are
composed, Galen explains, of two layers of fibres:25
καὶ μὴν δύο χιτῶνας ἡ γαστὴρ ἔχει πάντως ἕνεκά του γεγονότας καὶ
διήκουσιν οὗτοι μέχρι τοῦ στόματος. . . . ὅτι μὲν οὖν ἐναντίας ἀλλήλαις
τὰς ἐπιβολὰς τῶν ἰνῶν ἔχουσιν οἱ χιτῶνες οὗτοι, τὸ φαινόμενον αὐτὸ
μαρτυρεῖ. τίνος δ᾽ ἕνεκα τοιοῦτοι γεγόνασιν, Ἐρασίστρατος μὲν οὐδ᾽
ἐπεχείρησεν εἰπεῖν, ἡμεῖς δ᾽ ἐροῦμεν. ὁ μὲν ἔνδον εὐθείας ἔχει τὰς ἶνας,
ὁλκῆς γὰρ ἕνεκα γέγονεν· ὁ δ᾽ ἔξωθεν ἐγκαρσίας ὑπὲρ τοῦ κατὰ κύκλον
περιστέλλεσθαι. ἑκάστῳ γὰρ τῶν κινουμένων ὀργάνων ἐν τοῖς σώμασι
25
He later remarks that the stomach also has a third type, namely fibres which extend obliquely (see
below, pp. 275–277).

272 orly lewis
κατὰ τὰς τῶν ἰνῶν θέσεις αἱ κινήσεις εἰσίν. . . . καταπίνεται μὲν ἀμφοῖν τῆς
γαστρὸς τῶν χιτώνων ἐνεργούντων, τοῦ μὲν ἐντὸς ἕλκοντος, τοῦ δ᾽ ἐκτὸς
περιστελλομένου τε καὶ συνεπωθοῦντος.
The stomach has two tunics, which certainly exist for some purpose; they
extend as far as the mouth. . . . Now simple observation will testify that these
tunics have their fibres inserted in contrary directions. And, although
Erasistratus did not attempt to say for what reason they are like this, I am
going to do so. The inner (tunic) has its fibres straight, since it exists for the
purpose of traction. The outer (tunic) has its fibres transverse, for the purpose
of circular compression. In fact, the movements of each of the mobile organs
of the body depend on the setting of the fibres. . . . Swallowing takes place
when both tunics of the stomach are in action, the inner one exerting a pull
and the outer one by circular compression and propulsion. Gal., Nat.Fac. 3.8
(K. 2.168–9 . . . 2.172), trans. Brock, slightly modified
We see that according to Galen, the two layers of the stomach’s fibres
extend all the way to the mouth, that is, they encompass the oesophagus too.
In each layer, the fibres extend in a particular direction which defines their
respective function and allows the stomach to perform different actions. The
fibres of the inner layer extend along the length of the stomach and oesopha-
gus and thus enable these parts to pull matter into themselves. The fibres of the
outer layer extend around these parts, encircling them and enabling them to
compress and push matter and thus propel it forwards.26 Galen compares the
actions and effects of these fibres to those of muscle fibres:
ὥσπερ γὰρ ἐν τοῖς μυσὶν ἑκάστης τῶν ἰνῶν τεινομένης τε καὶ πρὸς τὴν
ἀρχὴν ἑλκομένης αἱ κινήσεις γίγνονται, κατὰ τὸν αὐτὸν λόγον κἀν τῇ
γαστρί· τῶν μὲν οὖν ἐγκαρσίων ἰνῶν τεινομένων ἔλαττον ἀνάγκη
γίγνεσθαι τὸ εὖρος τῆς περιεχομένης ὑπ᾽ αὐτῶν κοιλότητος, τῶν δ᾽
εὐθειῶν ἑλκομένων τε καὶ εἰς ἑαυτὰς συναγομένων οὐκ ἐνδέχεται μὴ οὐ
συναιρεῖσθαι τὸ μῆκος.
For just as the movements in the muscles take place when each of the fibres
becomes tightened and drawn towards its origin, such also is what happens
in the stomach: when the transverse fibres tighten, the breadth of the cavity
contained by them necessarily becomes smaller; and when the longitudinal
fibres contract and draw in upon themselves, the length must necessarily be
curtailed. Gal., Nat.Fac. 3.8 (K. 2.169–70), trans. Brock, slightly modified
This description demonstrates that the pulling and compressing which
Galen describes are not metaphorical, nor are they motions in the more
26
Galen stresses this point in order to argue against Erasistratus, who seems to have thought that the
stomach does not aid in the pulling in but passively receives the masticated food which is like
a weight that was dropped down.

general sense of qualitative alteration, which does not involve movement in
space.27 Rather, Galen depicts these motions of the fibres as actual motions of
the stomach and oesophagus (namely contraction and extension), similar to
the motion of muscles and their fibres. Elsewhere, Galen offers an even more
vivid picture, when he compares the straight fibres to hands (καθάπερ χερσί)
which stretch from the stomach to the mouth and pull the food in and
down.28 In On Natural Faculties 3.8, he describes in great detail the mechanical
effects of the pulling of these fibres: the contraction of the longitudinal fibres
in the stomach and oesophagus causes these parts to curl and contract upon
themselves – the oesophagus is actually seen to move down, whereas the
larynx, to which it is connected at the top, moves up. This is what happens
when the animal swallows.29 Galen stresses the sensory perceptibility of these
motions.30 Such motions will occur even in a dead person, if one pours water
into their throat (i.e. oesophagus), says Galen.31 This comparison with a dead
body demonstrates just how mechanical these motions are – the structure of
the parts (namely, their fibres) means that certain motions can take place
regardless of whether the body is alive or not. In other words, regardless of
whether their faculties exist, so to speak, or are activated.
Nevertheless, in the living animal the motions are not entirely mechanical,
for they depend on the natural faculties, which activate the fibres. Just as in the
case of muscular motion, where the fibres are activated by a motor faculty
reaching them from the brain through the nerves, so are the stomach fibres
activated by the faculties of attraction, retention, alteration and expulsion
inherent in the stomach.32 ‘All four exist (huparchousai) in the stomach’, says
Galen: ‘the attractive faculty in swallowing, the retentive in digestion (pettein),
the expulsive in vomiting and in the descent of digested food into the “thin
intestine”, and digestion itself is alteration (alloiôsis)’.33 Different stomach
fibres are activated by different faculties in accordance with requirement,
27
See Gal., Nat.Fac. 1.2 (K. 2.2–3) for Galen’s concept of ‘qualitative motion’ (κατὰ τὴν ποιότητα
κινήσις) as a change and alteration (ἀλλοίωσις) in a substance’s colour, temperature, flavour or
moistness.
28
Gal., UP 4.8 (Helmreich 1.206,24–207,1 = K. 2.282), and more generally UP 4.7–8 (Helmreich
1.205–7 = K. 2.281–2); compare Nat.Fac. 3.8 (K. 2.174) for Galen’s description of the oesophagus as
the ‘hand’ of the stomach (he makes the comparison with respect to particular animals, but his point
is nonetheless to demonstrate the general mechanism).
29
Gal., Nat.Fac. 3.8 (K. 2.168–73).
30
Gal., Nat.Fac. 3.8 (K. 169–74) and see Nat.Fac. 3.2–4 (K. 2.145–6, 2.149–50, 156–7) for the
perceptibility of the fibres’ motions during the exertion of the retentive, expulsive and assimilative
faculties.
31
Gal., Nat.Fac. 3.8 (K. 2.171).
32
See Gal., Mot.Musc. 1.1 (K. 4.371) on the faculty of voluntary motions reaching the muscles through
the nerves from the brain.
33
Gal., Nat.Fac. 3.8 (K. 2.177).

274 orly lewis
Figure 11.2 Schematic depiction of the two layers of fibres in the oesophagus and
three layers in the stomach: transverse (green; outer), longitudinal (purple; inner);
oblique (yellow; mixed with longitudinal). For the oblique fibres, see pp. 275–277.
(Image by Yotam Giladi and Project ATLOMY)
beginning with the initial stimulation by appetite.34 Galen does not explain
how exactly the faculties stimulate or activate the different fibres, but this
should not surprise us. As Galen himself says, he uses the term ‘faculty’
34
See above, p. 270, as well as Gal., Nat.Fac. 3.8 (K. 2.172–4).

(dunamis) when he is ‘ignorant of the substance of the activating cause’
(ἀγνοῶμεν τὴν οὐσίαν τῆς ἐνεργούσης).35 Galen’s motivation for referring
to faculties here, and not only to the mechanical motions of fibres, is his
underlying perception of the body as something that works through the
design of nature and cannot be reduced to explanations which rest solely on
the rules of mechanics. In this particular context, he is arguing against
Erasistratus’ purely mechanical explanation of swallowing.36
The food enters the stomach from the oesophagus through the ‘mouth
of the stomach’ (stoma tês gastros) – an elongated orifice which connects the
two parts and can open and close.37 Once the food has reached the
stomach, digestion occurs with the aid of the heat in the stomach.
The heat causes the digestive concoction (pepsis) which changes the bits
of the ingested food into a mixed fluid substance.38 Although Galen stresses
the ‘chemical’ process of qualitative change through heating, this process is
aided by the anatomical motions of the stomach, namely its contraction,
which retains and compresses the food inside the cavity:
ἡ γαστὴρ τῇ καθεκτικῇ δυνάμει χρωμένη συναγαγοῦσα πανταχόθεν
ἑαυτὴν καὶ συσπειράσασα καὶ περισταλεῖσα τοῖς ἐνυπάρχουσι περὶ τὴν
πέψιν αὐτῶν καταγίγν[ε]ται. τηνικαῦτα γὰρ ἑκάτερον τῶν στομάτων εἰς
ἐλάχιστον συνῆκται καὶ μέμυκεν.
the stomach exercises its retentive faculty and is actively engaged in the
digestion of its contents, gathering itself together, contracting and clasping
and compressing its contents. At that time both openings draw very tightly
together and close. Gal., UP 4.7 (Helmreich 1.205,25–206,5 = K. 3.281),
trans. May, slightly modified
This contraction of the stomach and its openings during the period of
digestion occurs by means of the oblique fibres (see Figure 11.3). These are
a third kind of fibre (see Figure 11.4). They are mixed in with the longitudinal
fibres and are substantially fewer in number.39 By means of these oblique fibres,
the stomach can ‘tightly contract and stretch over its contents at every point’
35
Gal., Nat.Fac. 1.4 (K. 2.9); compare Caus.Puls. 1.2 (K. 9.11) and see Hankinson 2014.
36
Erasistratus claimed that swallowing occurs simply on account of food entering from above (Gal.,
Nat.Fac. 1.16, 3.8 (K. 2.60–3, 176–7)).
37
In modern anatomy, this orifice is known as the cardiac or gastrooesophageal sphincter, a circular
muscle which acts as a kind of valve.
38
Gal., Nat.Fac. 2.4, 3.7 (K. 2.89, 2.166).
39
Gal., Nat.Fac. 3.11 (K. 2.181). In modern anatomy, the different fibres described by Galen are
recognised as muscle fibres, and the oblique (muscle) fibres, which Galen perceives as part of the
longitudinal layer, are distinguished as a third and innermost layer, which lies below the circular
muscle fibres. It enables the stomach to churn the food and break it up. It is noteworthy that this
inner layer is more prominent in particular parts of the stomach and that in animals (on whom

276 orly lewis
Figure 11.3 The stomach during digestion; it is contracted by means of the oblique
fibres (yellow); both openings (from oesophagus and at the pylorus) are tightly
closed. (Image by Yotam Giladi and Project ATLOMY)
Galen performed most of his anatomical observations) it forms a ‘very incomplete’ layer (Visible
Body 2017: ‘oblique (inner) muscle layer’; Dyce, Sack and Wensing 2010, 127).

Figure 11.4 The three layers of the stomach fibres: transverse (green; outer), longi-
tudinal (purple; inner); oblique (yellow; mixed with longitudinal). The transverse
fibres have been partially omitted to reveal the inner fibre layers. (Image by Yotam
Giladi and Project ATLOMY)
(πανταχόθεν ἐσφίγχθαι καὶ περιτετάσθαι τοῖς ἐνυπάρχουσι τὸ μόριον) until

the stomach’s role in digesting (i.e. altering) the food is completed.40
40
Gal., Nat.Fac. 3.11 (K. 2.181), trans. Brock; compare Nat.Fac. 3.4 (K. 2.152–13, 2.156). The same
occurs in the (pregnant) uterus (Nat.Fac. 3.3, 3.11 (K. 2.147–50, 2.181)).

278 orly lewis
(ii) From Stomach to Intestines
Once the alteration inside it ends, the stomach expels the digested matter
and propels it into the intestines through the second opening at the bottom
end of the stomach, the pylorus (pulôros). This is accomplished by the
contraction of the transverse fibres of the stomach (the outer layer) – which
afford circular pressure (τοῦ κατὰ κύκλον περιστέλλεσθαι) – which
‘squeezes’ the matter inside the cavity and expands the exit hole, thus
allowing the matter to flow into the intestines (Figure 11.5):
Figure 11.5 Expelling digested matter from the stomach into the intestines through
the widened pylorus (4), by means of the transverse fibres. (Image by Yotam Giladi
and Project ATLOMY)

αὖ πάλιν ὅταν τῇ προωστικῇ καλουμένῃ δυνάμει χρωμένη τὰ μὲν ἄλλα
πάντα στενώσῃ καὶ συναγάγῃ καὶ σφίγξῃ, τὸν πόρον δ’ ἀναπετάσῃ, δι’ οὗ
χρὴ κενωθῆναι τὰ προωθούμενα.
When the stomach exercises the faculty called expulsive all its other parts are
drawn in, contracted and held tightly together, while the passage through
which the material to be expelled must pass (sc. the pylorus), is spread wide.
Gal., UP 4.7 (H. 1.206,5–8 = K. 3.281), trans. May, slightly modified41
We see here too that the process begins when the faculty is activated.
This produces a mechanical action in the form of motions of the stomach
(on account of its fibres). Galen assumes that this effect of the propulsion
action of the stomach (which he calls a ‘mechanism’ – mêchanê),42 extends
far into the digestive tract, reaching past the liver all the way to the veins
branching from the hollow vein.43 We shall return to this later, when
discussing the motion through the veins (below, p. 288–000). The imme-
diate effect of the propulsion ‘mechanism’ is that the concocted food is
transferred from the stomach to the intestines.
(iii) From Intestines to Liver

In the intestines, there is now a flow of the fluid mass of partially digested
food. The intestines continue to digest – that is, alter – the food, but that is not
their main function. Rather, they act first and foremost as instruments of
‘distribution’ (anadosis): leading the nutriment into the liver via the veins.44
The intestines are long ‘tubes’ folded in multiple coils. The walls of these tubes
have, according to Galen, two layers of transverse fibres. In some parts of the
intestines and in some animals, he tells us, there are also longitudinal fibres,
but their role is not to pull, as of those in the stomach, but to protect the
transverse ones.45 The double layer of transverse fibres facilitates the strong
expulsive motion of the intestinal walls. They do not draw from the stomach
but rather receive from it the matter which the stomach’s expulsive force sends
into them. The intestines need then only propel this matter further forwards,
and the double layer allows them to perform this action with vigour
41
Compare Gal., Nat.Fac. 3.4 (K. 2.156–7). 42 Gal., Nat.Fac. 2.1 (K. 2.76).
43
Gal., Nat.Fac. (K. 2.76–7) and hence, there is no real need to explain the motion of the nutritive
matter through the veins with the aid of the mechanical principle of horror vacui, as Erasistratus does
(on this principle see below, pp. 287–288).
44
Gal., UP 4.17 (Helmreich 1.237–40 = K. 3.323–7, in particular Helmreich 1.237–8 = K. 3.323–4), 5.3
(Helmreich 1.255–6 = K. 3.349).
45
Gal., UP 4.17 (Helmreich 1.242–3 = K. 3.330–1), cf. UP 5.12 (Helmreich 1.283–4 = K. 3.387–8) and AA
6.7 (Garofalo 2.381 = K. 2.568). Modern anatomy identifies a longitudinal layer and transverse layer
for the intestines, continuing those of the stomach and oesophagus.

280 orly lewis
(sphodrotês) and hence more productively.46 Moreover, the double layer offers
insurance in cases in which the inner tunic is harmed to an extent that it stops
functioning (in dysentery, for example). In such cases, the outer transverse
layer of fibres continues to provide the expulsive motion of the intestines.47
Numerous veins connect the intestines to the liver (Figure 11.6). Their
mouths, explains Galen, open directly into the cavities of the intestines: ‘the
whole circumference of the intestines is pierced by innumerable openings that
extend to the inside and seize upon the useful part of the nutriment as it is
going by’ (καταπεπύκνωται γὰρ ἐν κύκλῳ πᾶν ἔντερον ἀναρίθμῳτινὶ πλήθει
στομάτων εἴσωκαθηκόντων, ὑφ’ ὧν ἀναρπάζεται τὸ χρηστὸν τῆς
διερξερχομένης τροφῆς).48 The stream of nutriment runs into these open
veins, aided by the contraction of the intestines’ transverse fibres which pushes
it along the many coils and into the veins. The continuous stream through the
long and winding path means that whatever is not ‘caught’ in the first row of
veins is caught by the next row and so forth.49 The multitude of veins from the
intestines lead to the portal vein, which enters the liver at ‘its most concave
part’.50
The majority of the nutritive matter is pushed into the veins from the so-
called fasting (nêstis) part of the intestines (known today as the ‘jejunum’
from the Latin translation of the Greek name). This is the first coiled and
long part of the intestines – it is connected at its beginning to the
‘outgrowth onto the intestine’ (ἡ εἰς τὸ ἔντερον ἔκφυσις) which extends
from the stomach, and at its end to the ‘thin intestine’ (τὸ λέπτον ἔντερον).
The latter is the ileum of modern anatomy, the former (the ‘outgrowth’) is
the duodenum.51 The ‘fasting’ intestine is so called since it is always found
empty when dissected, on account of the swift delivery of the nutriment
46
Gal., UP 4.17 (Helmreich 1.242 = K. 3.330); compare Nat.Fac. 3.8 (K. 2.168).
47
Gal., UP (Helmreich 1.243 = K. 3.330–1).
48
Gal., UP (Helmreich 1.240,9–12 = K. 3.327), trans. May, slightly modified, and see also UP 5.1–2
(Helmreich 1.249–53 = K. 3.340–5). Note that today we identify the veins as connected to the
intestines only from the mesentery side, that is, the dorsal (back) side of the intestines; Galen’s
formulation points explicitly to venous connections on all sides of the circumference of the intestinal
folds (ἐν κύκλῳ πᾶν ἔντερον).
49
Gal., UP 4.17 (Helmreich 1.240 = K. 3.326–7).
50
Gal., AA 6.10–11 (Garofalo 389 = K. 2.574–5), 13.1–2 (Simon 173–4, 178, 180–1 = Duckworth, Lyons
and Towers 139, 142, 144–5).
51
Galen tells us that Herophilus coined the term ‘duodenum’ (twelve-finger length –
δωδεκαδάκτυλον) – Ven.Art.Diss. 1 (Garofalo 77,13–15 = K. 2.780–1). Galen sometimes refers to
both coiled parts of the thin intestines (the jejunum and ileum) together as ‘thin intestine’, similar to
modern anatomical terminology; see the note by May (1968: 247, n. 5) on Galen’s terminology. After
the ‘thin intestine’ come the ‘blind’ (τὸ τυφλόν) intestine, the colon (τὸ κόλον) and the ‘straight’ (τὸ
ἀπευθυσμένον) intestine (i.e. the rectum). See Gal., UP 5.3 (Helmreich 1.253 = K. 3.345–6) for the
parts of the intestines.

Figure 11.6 The veins from the intestines (9) through which the digested matter
moves into the liver (12) – they merge into the portal vein (10). Most veins extend
from the ‘fasting intestine’ (jejunum (4)). We could depict here the veins only from
the anterior/ventral side, but Galen states that they are connected from all sides of
the small intestines, i.e. from the dorsal (back) and lateral sides as well (see n. 48).
According to the modern understanding, they are connected only to the mesentery
side (i.e. dorsal side). The colon (7) and rectum (8) are the passage through which
residues are expelled (as excrement). (Image by Yotam Giladi and Project
ATLOMY)

282 orly lewis
from this part to the liver.52 Galen lists several reasons for this swift process.
First is the large number of veins extending from it. Second is its proximity
to the liver, which means the nutriment has a shorter distance to cover.
Third, the expulsive force in this ‘fasting’ part of the intestine is strongest,
says Galen, on account of the purity of the black bile flowing into it from
the spleen through the duodenum. The black bile here is not yet mixed
with the residues of the digestive process (which it will accumulate in the
lower parts of the intestines), and it is thus ‘sharper’ and irritates the walls
of the intestine more strongly. In so doing, it encourages the ‘fasting’ part
to expel the matter inside it more swiftly.53 Finally, the distribution to the
liver is swiftest when the liver is empty and thus its appetite is strongest so
that it exercises its own attractive faculty most fiercely.54
(iv) From Liver to the Hollow Vein

The nutritive matter thus enters the liver through the portal vein and
spreads through the flesh of the liver.55 The liver performs the final
arrangement or structuring (teleos ho kosmos)56 and concoction (pepsis) of
the nutritive matter and turns it into ‘proper’ (akribes) and ‘useful’
(chrêston) blood.57 This process entails the heating, boiling (zesis) and
52
Gal., UP 5.3 (Helmreich 1.253,13–17 = K. 3.345).
53
According to Galen, residual black bile from the spleen enters the stomach and ‘encourages’ it to coil
and clasp onto the food and then moves out of the stomach with the partially digested food into the
duodenum and the rest of the intestines – Gal., UP 4.15, 5.4 (Helmreich 1.233, 1.265 = K. 3.317, 3.263–
4). Yellow bile also flows into the intestines through bile ducts running from the gall bladder to the
duodenum and bypassing the stomach, which it would otherwise harm on account of its corrosive
power (Gal., UP 5.1, 5.4 (Helmreich 1.250, 1.257–64 = K. 3.360)); in some animals, bile does enter the
stomach, but only near its lower part, close to the exit into the intestines (the pylorus sphincter) – see
Gal., Temp. 2.6 (Helmreich 77 = K. 1.631–2); compare May 1968, 244 n. 1, 252 n. 16; and Singer and
van der Eijk 2019, 143, n. 163. On the production of black and yellow biles, see below, n. 60.
54
Gal., UP 5.3 (Helmreich 1.255–6 = K. 3.348–50) and more generallyUP 5.4 (Helmreich 1.257–65 =
K. 3.351–62); compare Gal., Nat.Fac. 3.13 (K. 2.189–90) for the liver’s appetite affecting the speed of
the process.
55
On the branches of the portal vein inside the liver, see Gal., AA 13.2 (Simon 181 = Duckworth, Lyons
and Towers 145).
56
Gal., UP 4.3 (Helmreich 1.197,8–9, 197,16 = K. 3.269). May translates κόσμος as elaboration. While
this reflects the general idea, Galen seems to be distinguishing κόσμος from the more general
‘elaboration’: a few lines later, Galen uses the verb προκατεργάζομαι (‘pre-elaborated’, UP 4.3
(Helmreich 1.197,20 = K. 3.270)), a cognate of κατεργάζομαι, which is most commonly translated in
such contexts as ‘elaborate’. His point seems to be to stress the liver’s role in completing the
alteration of food into blood (i.e. in ‘ordering’ it), rather than just ‘processing’ or ‘elaborating’ the
nutritive matter.
57
Gal., UP 4.3 (Helmreich 1.197,8–9, 198,6 = K 3.269, 3.270); compare PHP 6.4.9–10, 6.4.17, 6.8.10–36
(De Lacy 384–6, 410–4 = K. 5.535, 5.537, 5.566–72). For the importance of the liver’s blood-making
role for the spirited and rational activities of the body see most recently Trompeter 2018, particularly
198–201.

purification (katharsis) of the nutritive matter by the innate heat in the
liver.58 The actual part which performs this alteration is the flesh of the
liver, which makes the matter thicker and redder by assimilating it to its
own properties.59 Galen compares the whole process to the production of
wine: the nutritive matter entering the liver is like the liquid derived
directly from pressing grapes and poured into casks; in the casks, it
ferments, and the heavy parts sink to the bottom whereas the airy part
rises and settles on top of the liquid. As in the case of wine, the digestion in
the liver produces two residues (perittômata). One is muddy and thick
(ἰλυῶδές τε καὶ παχύ), and its weight carries it down; it moves through
a vein to the spleen. The other is fine and light (λέπτον τε καὶ κοῦφον); it
rises and flows to the gall bladder through the bilious ducts which reach
deep into the liver.60 The different directions of motion of the two residues
(to the spleen and gall bladder respectively) are determined by the location
of the conduits into which they flow and by the particular power of
attraction of the bladder and spleen: each attract the substance most fitting
to it.61 The newly generated blood (the ‘wine’, as it were) rises and enters
the venous system:
ἀποθέμενος οὖν ὁ παρασκευαζόμενος ἐν ἥπατι χυμὸς εἰς τροφὴν τῷ ζῴῳ τὰ
προειρημένα δύο περιττώματα καὶ τὴν ἐκ τῆς ἐμφύτου θερμασίας πέψιν
ἀκριβῆ κτησάμενος ἐρυθρὸς ἤδη καὶ καθαρὸς ἐπὶ τὰ κυρτὰ μόρια τοῦ
ἥπατος ἀνέρχεται . . . ἐκδέχεται δ’ αὐτὸν ἐνταῦθα μία φλὲψ μεγίστη τῶν
κυρτῶν τοῦ ἥπατος διαπεφυκυῖα πρὸς ἄμφω τοῦ ζῴου τὰ μέρη φερομένη
τό τ’ ἄνω καὶ τὸ κάτω.
After the juice prepared in the liver for the nourishment of the animal has
deposited the two aforementioned residues (sc. to the gallbladder and
spleen), and after it has been perfectly concocted by the innate heat, it
becomes pure and red and rises to the convex part of the liver. . . . This juice
58
Gal., UP 4.3–4 (Helmreich 1.197–9 = K. 3.269–72).
59
Gal., UP 4.12 (Helmreich 1.217–20); compare PHP 6.7.8–31 (De Lacy 408–14 = K. 5.565–71).
60
Gal., UP 4.3–4 (Helmreich 1.197–9 = K. 3.269–72), AA 6.12 (K. 2.577–8) – this is in fact the process
in which the black, yellow and white humours are generated and cleansed (the gall bladder collects
the yellow bile and the spleen the black bile, whereas the phlegm has no specific organ and some of it
is assimilated in the body), see Gal., Nat.Fac. 2.9 (K. 2.134–40).
61
On this kind of attraction on account of qualitative affinity, see below, p. 290. Although the gall
bladder lies ‘below’ the concave part of the liver (the inferior or caudal side), the bilious ducts which
lead to the bladder reach to the ‘upper’, convex (cranial) part of the liver (cf. Gal., AA 13.2 (Simon
181 = Duckworth, Lyons and Towers 145)). According to Galen, the vein which runs to the spleen
branches from the portal vein which is inserted into the lower part of the liver (see Gal., UP 4.4
(Helmreich 1.198 = K. 3.271), AA 13.2 (Simon 175–7 = Duckworth, Lyons and Towers 140–1)). On
the flow of matter in both directions through the portal vein (nutritive matter from the intestines
into the liver, blood and yellow bile from the liver), see Gal., Nat.Fac. 3.13 (K. 2.196–8, 2.201).

284 orly lewis
is then received by one very large vein which grows from the convex portion
of the liver and leads to both the upper and lower parts of the animal. Gal.,
UP 4.4–5 (Helmreich 1.199,9–13 . . . 200,16–19 = K. 3.272 . . . 273–4), trans.
May, modified
This useful concocted material has become red and pure and rises
(ἀνέρχεται) to the convex (that is, cranial, which is ‘upper’ in bipeds)
portion of the liver, where it is received (ἐκδέχεται) by a large vein (the
‘hollow vein’, which we shall discuss shortly), from which it will flow to
most parts of the body. Some blood passes back through the portal vein on
the other (concave) side of the liver (by which the nutritive matter had
entered earlier) and goes to the intestines and stomach. Once the blood
enters the ‘hollow vein’ running from the other side of the liver, it ‘sheds’
the light, watery fluid which has been easing the passage of the nutritive
matter through the small vessels. This liquid makes the raw and thick
nutritive matter in its earlier stages more fluid and thus more easily
transmittable. It helps conduct (παραπέμπειν) the nutriment and without
it the matter could not be distributed up (ἀναδοθῆναι) from the stomach
to the liver. The source of this liquid is the water we drink and swallow into
our stomach.62 After the liver, the vessels are wide and the blood is hotter
(thanks to the liver, and later the heart) and consequently more fluid and
mobile; hence, the liquid is not required any more to ease the blood’s way.
The kidneys draw this thin liquid from the hollow vein by means of their
attractive faculty and then expel it as urine through the ureters, bladder and
urethra.63
(v) Distribution through the Hollow Vein and Its Branches

From the liver the blood is distributed to all parts of the body by means of
the venous system. According to Galen, the liver is the origin or source
(archê) of the system: all veins are ultimately connected to it through two
main veins which stem from it directly: the hollow vein and the portal vein.
Some fully prepared blood nourishes the stomach and intestines through
branches of the portal vein, which is connected to the concave (i.e.
‘bottom’) part of the liver.64 Most of the body is nourished, however,
through the branches of the hollow vein (ἡ κοιλὴ φλέψ), which has its
62
Gal., UP 4.5 (Helmreich 1.199–200 = K. 3.272–3); Galen cites there Hippocrates as referring to this
liquid as the ‘vehicle’ (ὄχημα) of nutrition; compare Hipp., De alim. 55 (Littré 9.120–1)).
63
Gal., Nat.Fac. 1.15–17 (K. 2.60–73), UP 5.5–13 (Helmreich 1.265–86 = K. 3.362–91), AA 6.13
(Garofalo 2.395 = K. 2.579–84).
64
Gal., PHP 6.5.16–8 (De Lacy 390 = K. 5.542), UP 5.2 (Helmreich 1.250–2 = K. 3.341–4).

many roots in the veins inside the liver, and which grows from the convex
(upper, cranial) part of the liver.65
This large vein extends, according to Galen, in two directions from the
liver – up and down. The upper ‘branch’ extends through the diaphragm and
passes by the right side of the heart to the throat, and its branches reach the
heart, arms, head and the muscles in the thorax and upper vertebrae.66 The
lower part extends down and sends branches to the kidneys, pelvic muscles and
organs, lower vertebrae and the muscles in that area, abdominal muscles, legs
and so forth.67 Galen compares the hollow vein to an aqueduct (agôgos) of
blood: ‘this vein, you would say, is a sort of aqueduct full of blood, having very
many conduits (ochetoi), both large and small, leading off from it and branch-
ing into every part of the body’.68 On Galen’s account, blood from the liver
enters this aqueduct and from there follows the course of its upper and lower
branches; some blood ‘spills’ into the lower branch from which it is delivered to
all the parts below the diaphragm.69 The rest of the blood rises towards the
thorax from which it continues to the head, arms, ribs and upper back.
The rising of the blood through the upper part of the hollow vein is
facilitated by the pulsating motion of the heart which ‘forcibly draws’
(ἐξαρπάζειν σφοδρῶς) blood up in its direction.70 Galen, like other ancients,
did not identify the atria (i.e. the upper cavities of the heart) as part of the
heart, but considered them part of the vessels which open into these atria (on
the right side, this is the hollow vein) (Figure 11.7).71 The pulsation of the
heart draws blood up from the direction of the liver through the hollow vein
on account of the principle of ‘the following into what is emptied’ (ἡ πρὸς
65
The argument for the liver being the origin of the veins (and the source of blood) is the topic of book
6 of Galen’s On the Doctrines of Hippocrates and Plato (De Lacy 360–426 = K. 5.505–85). To support
his argument, Galen offers empirical evidence from dissection and from the stages of development
of embryos and also from Plato and Hippocrates. See also below, n. 84 on the liver as a ‘source’.
66
Gal., AA 13.4–7, 13.10 (Simon 193–210, 222–5 = Duckworth, Lyons and Towers 154–67, 177–9), UP
16.14 (Helmreich 2.434–5 = K. 4.342–3); on the hollow vein, see also UP 6.4 (Helmreich 1.305–8 =
K. 3.418–22). Galen’s hollow vein corresponds to the modern inferior and superior vena cava: the
inferior extends from the common iliac veins in the pelvis to the right atrium in the heart, and
carries deoxygenated blood from the lower half of the body to the heart; the superior part extends
from below the neck at the junction of both brachiocephalic veins to the right atrium and carries
deoxygenated blood from the upper parts to the heart. Galen considers these as a single vessel (PHP
6.5.2–6 (De Lacy 388 = K. 5.538–40)) and does not distinguish it from the right atrium, which he
considers part of the vessel (see below, p. 285).
67
See in particular Gal., AA 13.3 (Simon 183–90 = Duckworth, Lyons and Towers 147–52).
68
Gal., UP 4.4 (Helmreich 1.199,19–22 = K. 3.272).
69
Compare Gal., PHP 6.4.11–2 (De Lacy 386 = K. 5.535–6).
70
Gal., Nat.Fac. 1.16, 2.1 (K. 2.64, 2.77), UP 4.6 (Helmreich 1.200–1 = K. 3.274); and see PHP 6.8.10–
12 (De Lacy 406–8 = K. 5.563) for the pulsation of the parts of the hollow vein closest to the heart.
71
Gal., AA 7.9, 7.11 (Garofalo 2.439–40, 2.449–50 = K. 2.616–17, 2.623–5), UP 6.3 (Helmreich
1.306,16–26 = K. 2.420).

Figure 11.7 The heart with its ventricles (3, 5) revealed. Note the perforations (4) in
the wall between the heart’s ventricles (septum). The hollows above the ventricles
(2, 6) are the connections between the hollow vein (vena cava) and the right ventricle
(2) and between the arteries entering from the lungs (the modern pulmonary veins)
and the left ventricle (6); these are known today as the atria of the heart, but are not
part of the heart according to Galen. (Image by Yotam Giladi and Project
ATLOMY)

τὸ κενούμενον ἀκολουθία), known as horror vacui (‘fear of void’). Like many
ancient authors, Galen believed that there is no complete void. Accordingly,
whenever changes in matter create an empty space, nature ensures that this
space is filled by new matter which is ‘automatically’ attracted into the
hollow space.72 The expansion of the heart upon dilation produces empty
spaces in its left and right cavities (the modern ventricles), and these are filled
by substances from the adjacent parts.73 The right cavity fills with blood
from the branch (apoblastêma) which extends, according to Galen, from the
hollow vein to the right cavity (see #2 in Fig. II.7).74
So of the blood which reaches the hollow vein, some proceeds to the right
cavity. Another portion flows into the veins nourishing the substance (i.e.
the walls) of the heart itself (coronary veins – ‘those which surround and
encircle the heart’).75 The rest of the blood continues up towards the neck
and nourishes the head, arms and upper-back area through the different
branches. From the portion of blood that enters the right cavity, most of the
blood proceeds to the lung through the artery-like vein (he artêriôdês phleps,
the pulmonary trunk in modern anatomy), which extends from a second
orifice in the right cavity and branches to both lungs.76 The rest of the blood
from the right cavity moves into the left cavity of the heart through the small
perforations which Galen observed in the wall separating the two cavities
(the septum).77 In the left cavity, the blood undergoes pneumatisation by
the vital pneuma (to zôtikon pneuma) present inside that cavity and becomes
finer and more yellow.78 This is the blood found in the arteries; for from the
72
Gal., Nat.Fac. 3.15 (K. 2.206–7). Galen explains this principle through an analogy with bellows; see
Korobili in the present volume, Chapter 7, as well as Furley and Wilkie (1984: 32–6).
73
Gal., Nat.Fac. 3.15 (K. 2.206–7), Nat.Fac. 3.14 (K. 2.204).
74
Gal., Nat.Fac. 1.14 (K. 2.64), PHP 6.5.2 (De Lacy 388,12–14 = K. 5.539). The blood entering the right
cavity passes through the tricuspid valve (the valve of ‘three barbs’ – τριγλώχις – in Erasistratus’
terminology) which covers the orifice connecting the cavity with the vein (in modern anatomy it
connects the right ventricle with the right atrium). On the cardiac valves, see PHP 6.6.4–11 (De Lacy
396 = K. 5.548–50) and UP 6.14–16 (Helmreich 1.347–58 = K. 476–92).
75
Gal., AA 13.5 (Simon 195–6 = Duckworth, Lyons and Towers 156), Nat.Fac. 3.15 (K.2.209); compare
PHP 1.7.5 (De Lacy 82,21–2 = K. 5.189–90).
76
Gal., UP 6.10–11 (Helmreich 1.324–37 = K. 3.444–62, especially Helmreich 1.324–30 = K. 3.444–53).
For Galen’s understanding of the vessels connecting the heart with the lungs, see Gal., UP 6.10–11
(Helmreich 1.324–34 = K. 3.444–58) and Harris 1973, 281–7 and Rocca 2020.
77
Gal., Nat.Fac. 3.15 (K. 2.207–8), where he also describes the shape of the perforations. Galen remarks
that the walls of the right cavity may draw to themselves some of the blood in the hollow of the
cavity, thereby adding nourishment to that which they receive from the (coronary) veins which
reach the walls directly. He is sceptical of this, however, and notes that if such additional nourish-
ment occurs, it is of a particularly small quantity (Nat.Fac. 3.15 (K. 2.207, 2.209)).
78
Gal., PHP 6.8.32–41 (De Lacy 414–16 = K. 5.571–3). The source of the vital pneuma is respiration:
when the heart dilates, the left cavity draws pneuma from the ‘vein-like artery’ (ἡ φλεβώδης
ἀρτηρία – the pulmonary arteries in modern anatomy) which connects it to the lungs. On this
see Gal., Nat.Fac. 3.15 (K. 2.207–8), UP 6.16 (Helmreich 1.354–5, 357–8 = K. 3.487, 3.490–2).

288 orly lewis
left cavity it enters the arterial system through ‘the large artery’ (hê megalê
artêria – the aorta) which extends up from the left cavity and which is the
‘trunk’ of the entire system. The motion of blood through the different
hollow parts of the heart and into the vessels attached to it occurs on account
of the dilation of the heart and these vessels. These parts draw the blood
through the connecting mouths and perforations by the mechanic principle
of horror vacui.79 The constant motions of these parts is the natural pulsation
(sphugmos) which occurs on account of another faculty of nature – the
‘pulsating faculty’ (he sphugmikê dunamis).80
Although the veins do not pulsate continuously like the heart and
arteries,81 Galen suggests that the walls of all veins exert a certain transverse
pressure which pushes the matter onwards. Moreover, he remarks that the
compression and expulsive force of the stomach remain effective, even if
weakened, throughout the digestive path and thus assist in propelling the
matter.82 Through the connected system of veins which branch from the
main ‘aqueduct’, blood reaches all the parts of the body. Veins lead into
the substance of each part and inside these parts branch into small veins.
It is from these veins that each part nourishes itself. Galen compares this
layout to an irrigation system in which many small conduits (ochetoi)
extend through the substance: in the garden, this substance is the earth
patches between the pipes; in the body it is the substance of the
individual parts in the body, namely, flesh, bone, cartilage and so forth.83
(vi) Nourishing the Parts: The Final Stage of Nutrition

Once blood arrives at the individual parts, the final, local, stage of nour-
ishment begins. This takes place by means of the four natural faculties
79
Gal., Nat.Fac. 3.14 (K. 2.204–5).
80
Gal., Caus.Puls. 1.2 (K. 9.4–5), PHP 6.7.1–7 (De Lacy 404–6 = K. 5.560–2).
81
Galen claims at times that the veins attached directly to the heart pulsate to a certain degree, though
he does not commit to the cause of this pulsation, that is, whether it is occurs on account of a faculty
inside the veins’ tunics as in the case of the arteries, or for a more ‘mechanical’ reason, namely their
physical attachment to the pulsating heart, which causes them to move with it – see Gal., PHP
6.8.10–12 (De Lacy 406–8 = K. 5.562–3); compare Arist., Resp. 20, 480a10–3 for this ‘mechanical’
pulsation of the vessels on account of their connection to the heart.
82
Gal., Nat.Fac. 2.1 (K. 2.76, 2.77). Galen claims this in the course of arguing against Erasistratus (who
claims that the motion through the veins depends on the horror vacui principle), and it is not entirely
clear whether Galen truly believes that the veins have compression and propelling force, or whether
his claims are merely rhetorical for the sake of proving the absurdity (as he sees it) of his rival’s view.
83
Gal., Nat.Fac. 3.15 (K. 2.210–12), UP 16.14 (Helmreich 2.432–3 = K. 4.339–40); compare Arist., Part.
An. 3.5, 668a27–b1 for the irrigation pipes analogy. Nature has designed the layout of the veins in the
parts in a manner fitting to the underlying substance and requirement of nourishment in each part
(Gal., UP 16.14 (Helmreich 2.432–3 = K. 4.339–40); cf. Nat.Fac. 3.15 (K. 2.211–12)).

present innately in each of the parts.84 Galen summarises this local process
in the following manner:
ἕκαστον τῶν ὀργάνων εἰς ἑαυτὸ τὴν πλησιάζουσαν ἐπισπᾶται τροφὴν
ἐκβοσκόμενον αὐτῆς ἅπασαν τὴν χρηστὴν νοτίδα, μέχρις ἂν ἱκανῶς
κορεσθῇ, καὶ ταύτην . . . ἐναποτίθεται ἑαυτῷ καὶ μετὰ ταῦτα προσφύει
τε καὶ ὁμοιοῖ, τουτέστι τρέφεται.
Each of the organs draws into itself the nutriment alongside it, and devours
all the useful fluid in it, until it is thoroughly satisfied; this nutriment . . . it
stores up in itself, afterwards making it adhere and then assimilating it – that
is, it becomes nourished by it. Gal. Nat.Fac. 3.13 (K. 2.198), trans. Brock
During this local nutritive activity, the faculties are not assisting in prepar-
ing and supplying nutritive matter for the entire body, but in ensuring that
a given part benefits from the nutritive matter it receives. Let us follow the
course step by step.
First comes ‘the actual ferrying (αὐτὸ τὸ παράγεσθαι) of nutriment
from the veins into each of the parts’; this occurs when the attractive
faculty – the ‘faculty to attract the appropriate quality’ (ἡ ἑλκτικὴ τῆς
οἰκείας ποιότητος δύναμις) – is active (ἐνεργούσης).85 In each body part,
the substance surrounding the small veins (the ‘irrigation conduits’)
attracts from the closest veins and by means of its own attractive faculty
84
When laying out his views of the tripartite soul, Galen depicts the liver as the source of natural
and nutritive faculties: Prop.Plac 10 (Boudon–Millot and Pietrobelli 181); PHP 7.3.2–3 (De Lacy
440,3–8 = K. 5.601). However, in Nat.Fac. there is no reference to a relation of the liver to these
faculties, nor indeed to the soul. Moreover, in his diagnostic works he explicitly distinguishes natural
faculties and the activities they produce as ‘innate’: Loc. Aff. 1.7.1 (Gärtner 302,17–22 = K. 8.66);
Symp.Diff. 3.1 (Gundert 216,16–18 = K. 7.55); compare also Mot.Musc. 1.1 (K. 4.371). This inconsist-
ency has often been noted, for example, by De Lacy 1981, 61–3, and von Staden 2002, 107–9, but
I would like to note some further points relevant to the topic of this chapter, which emphasise the
independence of these faculties from the liver in Galen’s method. First, Galen’s detailed discussions
of digestive-nutritive problems demonstrate that, by his account, such problems are caused by local
impairments in the parts (e.g. Symp.Diff. 4.6–20, Gundert 230–44 = K. 7.65–74), and no reference
is made to an effect caused by the liver, unless the liver itself is impaired (in PHP 6.3.5 (De Lacy
374,1–4); he only claims that it takes time for liver problems to disrupt nutrition). For example,
problems in the stomach’s digestive-assimilative function do not derive from a problem in the liver,
and problems in the liver do not undermine the stomach’s function; for example, Caus.Symp. 2.3, 3.1,
3.3 (K. 7.166–70, 7.207–8, 7.221–2); Symp.Diff. 5.7 (Gundert 248–50 = K. 7.77–8) is particularly
telling. It should be noted that in book 6 of PHP, which is concerned with proving the liver’s role as
a source of the desiderative soul and functions related to it, the focus is on the liver’s role as the source
of the veins and blood and the faculty of the production of blood in the liver and its distribution
thereof, not as a source of the faculties producing the different stages of the process. The whole
argument is essentially directed against the claim that the heart and not the liver is the source of the
veins, for example, PHP 6.4.1–9, 6.8.74 (De Lacy 384, 424,9–11 = K. 5.532–5, 5.582–3).
85
Gal., Nat.Fac. 1.13, 3.1 (K. 2.30,13–5, 2.144,7–10), trans. Brock, modified. See Nat.Fac. 1.14–17 (K.
2.45–73) as well as book 2 for Galen’s demonstration of the existence of an attractive faculty and his
arguments against Erasistratus and Asclepiades on this matter.

290 orly lewis
the blood it requires and which is most appropriate to its particular nature
and qualities.86 The kind of attraction at play here is one we may term
‘qualitative attraction’ or ‘specific attraction’. Galen opposes this kind to
the attraction occurring on account of the horror vacui principle. The latter
occurs ‘mechanically’ as soon as a new space is created and will attract the
lightest matter. The former is more ‘selective’: a given matter attracts to
itself what is ‘closer in nature’ (τῇ φύσει συγγενέστερον) on account of
‘appropriateness of quality’ (οἰκειότητι ποιότητος). Whereas Galen com-
pared the horror vacui attraction to the flow of air into bellows, he
compares the qualitative attraction to the effect of the lodestone and
cathartic drugs. The latter attract (ἐπισπᾶν) and draw up (ἀνέλκειν) and
out of the body different substances such as phlegm, poisons and arrow-
heads embedded deeply in the flesh, ones which even direct and forced
traction by fingers did not manage to extract.87 The lodestone and drugs
attract matter on account of a similarity between their respective qualities
and the quality of the matter they attract. Galen’s distinction between the
two types of attraction and his insistence on the involvement of a ‘faculty’
and an ‘informed’ process of attraction is part of his argument against other
explanations of digestion and nutrition, which rested on purely mechanical
principles or atomistic ideas.88
Once a body part has attracted to itself the nutritive matter it requires, it
must assimilate it to itself. ‘Actual nourishment’ (θρέψις αὐτή), explains
Galen, is the alteration and ‘complete assimilation’ (τελέως ὁμοιοῦται) of
the common nutritive matter to the particular needs (i.e. particular qual-
ities) of each part.89 This process is not immediate. In order for the part to
be able to process and alter (‘digest’) the nutritive matter from the veins
(i.e. blood), this matter must remain inside the part and not flow back into
the veins. Every part thus applies its retentive faculty in order to retain the
matter it has attracted: the parts ‘envelop their nutriment and in this
manner clasp it all round’ (περιπτύσσεται τῇ τροφῇ καὶ οὕτω σφίγγει
πανταχόθεν αὐτήν).90 The duration of this process differs according to the
86
Gal., Nat.Fac. 3.15 (K. 2.210–1). Each homoiomerous part has a unique qualitative mixture, and each
organ is a unique combination in quality and shape of these homoiomerous parts. Nature shaped
them so according to their respective functions in the body (Gal., Nat.Fac. 1.6 (K. 2.13–15)). The
teleological, functional blueprint of the anatomical design of the body and its parts is a core part of
Galen’s thought and the underlying theme of his On the Usefulness of Parts.
87
Gal., Nat.Fac 1.14, 3.1 (K. 2.53–4, 2.206–7, 2.210).
88
See the references above, nn. 36 and 85, as well as Nat.Fac. 1.16–17 (K. 2.66–71).
89
Gal., Nat.Fac. 1.11, 3.13, 3.15 (K. 2.24, 2.198, 2.211–12).
90
Gal., Nat.Fac. (K. 2.199), trans. Brock, slightly modified; and see Nat.Fac. 3.1 (K. 2.144–5) for his
proof for the retentive faculty.

type of homoiomerous matter from which the part is made. The red,
viscous blood requires only a short alteration to become flesh, but
a longer one to become white-yellow solid nerve or cartilage, and longer
still to turn into hard bone. Those parts which require significant time have
‘cavities or caverns or some analogue to the caverns’ (κοιλίαι ἤ σήραγγες ἤ
τι ταῖς σήραγξιν ἀνάλογον) of diverse dimensions in which they hold the
nutritive matter while it is being assimilated.91
The retention of the nutritive matter entails two processes: ‘addition’
(prosthesis) and then adhesion (prosphusis). Galen does not fully explain
how these take place, but he clearly envisions these as necessary stages of the
process.92 We must imagine here a flow of matter entering the substance of
the part (due to the attractive faculty which draws it from the veins); this
matter then ‘stands’ inside this part but separately from the part’s substance
(think of a mixture of oil and water), and the part then exerts an adhesive
force which prevents the matter from ‘slipping away’ and allows the
assimilative faculty to begin altering it. Galen compares the stage from
attraction through to presentation to the process of eating:
ὥσπερ οὖν τοῖς ζῴοις αὐτοῖς ὅρος ἐστὶ τῆς ἐδωδῆς τὸ πληρῶσαι τὴν
γαστέρα, κατὰ τὸν αὐτὸν τρόπον ἑκάστῳ τῶν μορίων ὅρος ἐστὶ τῆς
προσθέσεως ἡ πλήρωσις τῆς οἰκείας ὑγρότητος.
As in the case of the animals themselves the end of eating is that the stomach
should be filled, similarly in the case of each of the parts, the end of addition
is the filling of this part with its appropriate liquid. Gal., Nat.Fac. 3.13 (K.
2.198–9), trans. Brock
When we eat, we chew and swallow food, it reaches the stomach (thanks to
the attraction and propulsion of the fibres in the walls of the oesophagus
and stomach, as discussed earlier). Once a sufficient amount of food has
91
Gal., Nat.Fac. 3.15 (K. 2.211). The longer, more complex processes of assimilation entail the
production of intermediate (ἐν μέσῳ) substances: for bones we have marrow (μυελός); for cartilage
the mucus substance surrounding it (τὸ περικεχυμένον μυξῶδες); for ligaments, nerves and mem-
branes a ‘viscous liquid which is spread inside them’. All these ‘intermediate substances’ are
composed of many fibres (ἶνες, reading ἔκαστον in K. 2.212, line 18 as referring back to ὑγρός and
τὸ περικεχυμένον) among which the newly attracted nutritive matter is held and on which the
assimilative faculty acts – see Gal., Nat.Fac. 3.15 (K. 2.212–13). Galen’s theory here is worthy of closer
consideration, though this cannot be undertaken here. I mention two brief points to take into
account in such a study. First, the exact relation between these substances (e.g. marrow), their fibres
and the ‘caverns’: does Galen consider these substances as the ‘analogue to the caverns’ (Nat.Fac.
K. 2.212)? Second, the reference to ‘fibres’ in a liquid recalls Aristotle’s description of the blood of
some animals which helps it congeal (Arist., Hist. An. 3.6, 515b30–516a7).
92
See Gal., Nat.Fac. 1.11 (K. 2.24–5) for his demonstration of the difference between addition and
adhesion.

292 orly lewis
entered and filled the cavity of the stomach, we stop ingesting more food,
that is, we stop eating; then the process of ‘digestion’ begins. On the local
level of the individual parts, ‘eating’ ends when the part is full with the
nutritive matter it has attracted from the veins and this matter is added to it
and ready for use. Once this process is complete, the part can begin
digesting and processing the food – it first makes it stick and then begins
to alter it in order to assimilate it to its own qualities.
Because different parts lie at different ‘stops’ along the digestive tract and
venous system, the locative nutrition takes place in different parts at
different times. Unlike the attraction based on the horror vacui principle,
qualitative attraction works on matter lying very close (τὰ ἐγγύτατα), and
each part attracts its nutriment as it passes nearby.93 This is true also for the
main digestive parts which contribute to the overall nutrition of the body:
the stomach, intestines, liver and veins. These parts also require nourish-
ment. They attract into their substance (their walls or flesh) a portion of the
nutritive matter passing through them (at whatever stage of digestion it
may be) before passing it on for further digestion and distribution.94 After
passing on the matter, the part performs the local digestive process on the
portion of matter it attracted to itself. As Galen explains, while the stomach
is making adherent (prosphuesthai) the nutritive matter which its walls have
attracted and added to themselves, the matter which the stomach has
expelled is now moving through the intestines and liver. These parts, in
turn, attract from this flow the amount which they require for their
respective nutrition and add (prostethenta) it to themselves. In the next
stage, while the (walls of) the stomach are completing the assimilation of
their nutritive matter (which has already become adherent to them –
ὁμοιώσασαν ἑαυτῇ τελέως τὰ προσφύντα), the matter which was earlier
added (τὸ προστεθέν) to the intestines and liver now becomes adherent
(προσφύεσθαι) to them – at the same time, throughout the body there is
distribution (ἀνάδοσις) through the veins and addition (πρόσθεσις) to the
other parts. While the intestines and liver are assimilating their nutritive
matter, the body parts make their own nutritive matters adherent and later
assimilate them.95
93
Gal., Nat.Fac. 3.15 (K. 2.210), compare Nat.Fac. 3.12 (K. 2.183).
94
On Galen’s account, under normal conditions, the stomach and intestines do not need the liver for
their nourishment – they attract from the nutritive matter which passes through them; only if no
food is ingested for a long period do the stomach and intestines attract nutritive matter from the
liver through the branches of the portal vein – see Gal., Nat.Fac. 3.13 (K. 2.197–8, 2.201–2). This
implies that the stomach and intestines are not nourished by blood as such.
95
See Gal., Nat.Fac. (K. 2.200–1) for Galen’s description of these stages.

Conclusion
Galen’s theory of the digestive-nutritive process demonstrates his complex
view of the body as a living, teleologically designed and operated machine,
one in which faculties and anatomical structures, teleology and mechanics,
are close counterparts generating the workings of the body and maintaining
it in accordance to Nature’s plan. Digestion and nutrition are facilitated by
natural nutritive faculties of attraction, expulsion, retention and assimilation.
These enable each part to attract to itself exclusively the nutritive matter
most appropriate to it and to hold it for long enough in order to process it
until the matter is fully assimilated and can actually nourish the part. In the
parts assigned with nourishing the entire body – that is, preparing the
common nutritive matter from food and distributing it to the parts –
these faculties perform their actions on a ‘larger’ scale. They attract, retain,
assimilate and propel (expel) the whole supply of nutritive matter moving
through their cavities. Whereas Galen is vague on how the local digestive-
nutritive process occurs in practice, he describes in detail the actions which
the faculties cause in order to perform the digestive-nutritive process of the
body as a whole. They cause the contraction and relaxing of the different
kinds of fibres which extend through the walls of the digestive parts. These
cause the motion and change of shape in the parts in accordance to the
action required – pulling, pushing, squeezing and holding in place. These
motions and actions of the digestive parts produce and move the nutritive
matter for all the parts of the body. These motions do not depend, however,
only on the ‘mysterious’ faculties. As Galen stresses time and again in all his
physiological explanations, the activities depend also on the anatomical
structure of the parts; in the case of digestion and distribution of nutritive
matter, this includes the layout of the fibres of the individual organs involved
in the process. It is my hope that this close analysis of the different stages,
processes and parts involved in healthy digestion and nutrition according to
Galen will facilitate closer consideration and a better understanding of his
diagnostic, pathological and therapeutic conceptions concerning digestive-
nutritive diseases and impairments. This can assist, in turn, in gaining
a closer understanding of the considerations shaping and directing Galen’s
diagnostic and therapeutic practical methods.
RE F E RE N CE S
Balme, D.M. and Gotthelf, A., 2002. Aristotle: ‘Historia Animalium’ (Cambridge).

294 orly lewis
Berryman, S. 2007. ‘Teleology without Tears: Aristotle and the Role of
Mechanistic Conceptions of Organisms’, Canadian Journal of Philosophy,
37.3: 351–69.
Boylan, M. 1982. ‘The Digestive and “Circulatory” Systems in Aristotle’s Biology’,
Journal of the History of Biology, 15.1: 89–118.
2015. The Origins of Ancient Greek Science: Blood. A Philosophical Study
(New York).
Brock, A. J. ed. 1916. Galen: On the Natural Faculties (Cambridge, MA) (reprint,
2006).
Boudon-Millot, V., A. Pietrobelli. 2005. ‘Galien ressuscité : édition princeps du
texte grec du De propriis placitis’, Revue des Études Grecques, 118: 168–213.
Corcilius, K. 2015. ‘Faculties in Ancient Philosophy’, in D. Perler, ed., The
Faculties: A History (Oxford), 19–58.
De Lacy, P. 1978–1984. Galen: On the Doctrines of Hippocrates and Plato (Berlin).
1988. ‘The Third Part of the Soul’, in P. Manuli and M. Vegetti, eds., Le opere
psicologiche di Galeno: atti del terzo Colloquio galenico internazionale, Pavia,
10–12 settembre 1986 (Naples), 43–63.
Debru, A. 2008. ‘Physiology’, in R. J. Hankinson, ed., The Cambridge Companion
to Galen (Cambridge), 263–82.
Duckworth, W. H. L., M. C. Lyons and B. Towers. 1962. Galen on Anatomical
Procedures: The Later Books (Cambridge) (reprint, 2010).
Dyce, K., W. O. Sack and C. J. G. Wensing. 2010. Textbook of Veterinary Anatomy
(4th ed.). (St Louis).
Furley, D. J. and J. S. Wilkie. 1984. Galen on Respiration and the Arteries (Princeton).
Garofalo, I., ed. 1986–2000. Galeni anatomicarum administrationum libri qui
supersunt novem. 2 vols. (Napoli).
Gärtner, F., ed. 2015. Galeni De locis affectis i–ii. CMG v 6,1,1 (Berlin).
Gotthelf, A. 2012. Teleology, First Principles, and Scientific Method in Aristotle’s
Biology (Oxford).
2015. ‘Teleology, First Principles, and Scientific Method in Aristotle’s Biology’,
in B. Holmes and K.-D. Fischer, eds., The Frontiers of Ancient Science: Essays
in Honor of Heinrich von Staden (Berlin), 139–74.
Gundert, B., ed. 2009. Galeni De symptomatum differentiis. CMG v 5,1 (Berlin).
Hankinson, R. J., ed. 2008. The Cambridge Companion to Galen (Cambridge).
Hankinson, R. J. 2014. ‘Galen and the Ontology of Powers’, British Journal for the
History of Philosophy, 22.5: 951–73.
Harari, O. 2016. ‘Alexander against Galen on Motion: A Mere Logical Debate?’,
Oxford Studies in Ancient Philosophy, 50: 201–36.
Harris, C. R. S. 1973. The Heart and the Vascular System in Ancient Greek Medicine:
From Alcmaeon to Galen (Oxford).
Helmreich, G., ed. 1904. Galeni De temperamentis libri tres (Leipzig).
1907–9. Galeni De usu partium libri xvii. 2 vols. (Leipzig) (reprint, Amsterdam,
1968).
Johnson, M. R. 2017. ‘Aristotelian Mechanistic Explanation’, in J. Rocca, ed., Teleology
in the Ancient World: Philosophical and Medical Approaches (Cambridge).

Kühn, C. G., ed. 1821–33. Claudii Galeni opera omnia. 22 vos. (Leipzig) (reprint,
Cambridge, 2011).
Lewis, O. 2017. Praxagoras of Cos on Arteries, Pulse and Pneuma: Fragments and
Interpretation (Leiden).
Littré, É., ed. 1861. Oeuvres completes d’Hippocrate, vol. 9 (Paris)
Marechal, P. 2020. ‘Reflection 1: Teleology and Function in Galenic Anatomy’, in
J. K. McDonough, ed., Teleology: A History (Oxford), 64–70.
May, M. 1968. Galen: On the Usefulness of Parts of the Body. 2 vols. (Ithaca).
Meyer-Steineg, T. 1913. ‘Studien zur Physiologie des Galenos’, Archiv Für
Geschichte Der Medizin, 6.6: 417–48.
Rocca, J. 2003. Galen on the Brain: Anatomical Knowledge and Physiological
Speculation in the Second Century ad (Leiden).
2020. ‘Galen on Pneuma: Between Metaphysical Speculation and Anatomical
Theory’, in S. Coughlin, D. Leith and O. Lewis, eds., The Concept of Pneuma
after Aristotle (Berlin), 283–312.
Schiefsky, M. J. 2007. ‘Galen’s Teleology and Functional Explanation’, Oxford
Studies in Ancient Philosophy, 33: 369–400.
2012. ‘Galen and the Tripartite Soul’, in R. Barney, T. Brennan and C. Brittain,
eds., Plato and the Divided Self (Cambridge), 331–49.
Siegel, R. E. 1968. Galen’s System of Physiology and Medicine: An Analysis of His
Doctrines and Observations on Bloodflow, Respiration, Humors and Internal
Diseases (Basel).
Simon, M. 1906. Sieben Bucher Anatomie des Galen, 2 vols. (Leipzig).
Singer, P. N. and P. J. van der Eijk, eds. 2019. Galen: Works on Human Nature,
vol. 1: Mixtures (Cambridge).
Van der Eijk, P. J, 2014. ‘Galen on the Nature of Human Beings’, in P. Adamson,
R. Hansberger and J. Wilberding, eds., Philosophical Themes in Galen (Bulletin
of the Institute of Classical Studies Supplement 114) (London), 89–134.
2015. ‘Galen on the Assessment of Bodily Mixtures’, in B. Holmes and K.-
D. Fischer, eds., The Frontiers of Ancient Science: Essays in Honor of Heinrich
von Staden (Berlin), 675–98.
Biologie: Intentionen, Methoden, Ergebnisse: Akten des Symposions über
Aristoteles’ Biologie vom 24.–28. Juli 1995 in der Werner-Reimers-Stiftung in
Bad Homburg (Wiesbaden), 183–208.
2002. ‘Body, Soul, and Nerves: Epicurus, Herophilus, Erasistratus, the Stoics, and
Galen’, in J. P Wright and P. Potter, eds., Psyche and Soma: Physicians and
Metaphysicians on the Mind–Body Problem from Antiquity to Enlightenment
(Oxford).
Trompeter, J. 2018. ‘The Actions of Spirit and Appetite: Voluntary Motion in
Galen’, Phronesis, 63.2: 176–207.
Visible Body: Human Anatomy Atlas. 2017. Version 2017.2.08 [Computer
Software], Argosy Publishing.

chapter 12
Iatromechanism and Antiquarianism in Morgagni’s

Studies on Celsus, 1720–1761
Marquis Berrey
It is not my habit to write diffusely in the theoretical part of medical

consultations; since I always remember that most patients are usually
like the Empirics, convinced as Celsus [1.pr.38] says, that these con-
jectures about hidden matters are not relevant, since it is not the cause
of the disease but the cure that counts.
Giovanni Battista Morgagni, Consultation 23, August 14, 17251
The effect that readers can have on authors is profound. Our

Enlightenment heritage disposes us to believe that the relationship of
cause and effect runs from authors to readers, rather than the reverse.2
Yet the history of medicine includes numerous examples of readers affect-
ing authors. In the origin story of early modern iatromechanics, Friedrich
Hoffmann (1660–1742) came to see in the Hippocratic writings
a predecessor of his own mechanical understanding of pathological action
within the activities of the body.3 While iatromechanics was not a term of
Hoffmann’s own use, it captures his explanatory model of assigning the
proximate causes of disease to the dynamic processes of force, pressure, and
motion within bodily anatomy. Over the course of the eighteenth century,
1
Translated in Jarcho 1984, 77–78. All references to Celsus (as book, chapter, and section) follow the
edition of Marx 1915; all references to Morgagni (as title, volume, page, and column) follow the
edition of Morgagnus 1762–1765. All translations are my own. I quote liberally from Morgagni’s
works, since, with the exception of the De Sedibus et Causis Morborum and the Consulti Medici, there
are no English translations available of these texts. Materials enclosed within square brackets [ ] in
translated primary sources are my editorial insertions.
2
de Certeau 1988, 166–167.
3
Lonie 1981, who gives a listing and analysis of passages from Hippocrates Seed, Nature of the Child,
and Diseases 4 (130–131) “in which there is, perhaps, a deliberate attempt to reduce all explanations to
the contact and pressure between solid and fluid bodies” in a prelude to Hoffmann’s reading. See
Webster this volume, Chapter 6, for a rereading of these texts between mechanism and animism.
Celsus cites none of these Hippocratic texts; compare Marx 1915, 433–435. Müller 1991, 25–28, contains
a more expansive view of Hoffmann’s relation to ancient medicine.
296

Iatromechanism and Antiquarianism 297
other contemporary physicians reacted to Hoffmann’s medical mechanics,
itself a synthesis of activities in Newtonian and Cartesian physics and
William Harvey’s discovery of the circulation of the blood.4 The advance
of contemporary European medicine raised the question whether the
traditional ancient Greco-Roman sources of medicine were compatible
with the new medicine, a usual question since the sixteenth century now
placed within a new framework. An underappreciated example of readers
affecting authors comes from the Paduan professor of anatomy, Giovanni
Battista Morgagni (1682–1771), whose lifelong engagement with the med-
ical text of the Roman encyclopedist A. Cornelius Celsus (fl. 30 ce)
produced a reading of Celsus situated in iatromechanist debates: under-
appreciated, because Morgagni’s studies show the iatromechanist shift
from Hoffmann and because Morgagni’s reception of Celsus continues
to mark our editions of the ancient Latin text.
After all, Aulus Cornelius Celsus was once a major figure in the history of
medicine. For over 400 years, from the fifteenth century until well into the
nineteenth century, European physicians treated his text De Medicina with the
same esteem and value as the works of Hippocrates or Galen. In addition to
Celsus’ medical guidance, early modern physicians consulted Celsus about the
translation of Hippocratic works into Latin; they consulted him about med-
ical ethics and the relative value of surgeons in comparison with university-
trained physicians.5 Italian physicians in particular seem to have read Celsus
with great interest, perhaps because of his putative Italian origins.6
Giovanni Battista Morgagni was one of Celsus’ Italian readers.7
Morgagni was not only a leading physician of the eighteenth century but
also, as an accomplished Latinist and an antiquarian, an important corres-
pondent in the pan-European Republic of Letters. His eight open letters
on Celsus, written in Latin as Epistulae in Celsum, form the most important
study of Celsus in the eighteenth century.8 They were written in two
4
Harvey’s discovery of the circulation of the blood was central to the querelle between ancients and
moderns; Lo Presti 2014.
5
Celsus’ role in the history of early modern medicine is richer than previous studies’ benign neglect.
Conde Parrado 2003 provides the only full-scale study of the reception of Celsus but limits his
evidence to c.1500–1600. Conde Parrado 2008, Fischer 1994, and Ratcliffe 2013, 17–41, are useful
supplements.
6
Morgagni In Celsum 1 5.50 col.1, In Celsum 4 5.66 col.1 argues that Celsus 5.21, 7.26 show that he was
Roman; he is critically following Rhodius’ seventeenth-century Vita Celsi, widely reprinted over the
early modern period; see Rojouan 1997, 2.6–16.
7
Zampieri 2016 is the best recent study, contextualizing Morgagni’s evolving mechanist ideas and the
De Sedibus et Causis Morborum (1761); Jarcho 1984, Nicholson 1992, Cunningham 2010, Bresadola
2018 are useful supplements.
8
Marx 1915, lxvii–lxviii.

298 marquis berrey
groups: the first three were written in 1720 to 1721; the last five were written in
an extraordinary burst of creative energy in a month from mid-December
1749 to mid-January 1750.9 Morgagni addressed them all to the same individ-
ual, who was Morgagni’s assistant dissector in the anatomy theater at Padua;
the open letters were dedicated to Lorenz Heister (1683–1758), Morgagni’s
great surgical contemporary. Morgagni’s letters, in combination with his other
works, show his lifelong engagement with Celsus’ text. Previous scholarship
has seen Morgagni’s letters simply as a repository of variae lectiones to the text
of Celsus, rather than part of a longer dynamic strategy of understanding the
medical content of (in Morgagni’s view) a largely accurate ancient medical
author.10 I contend that Morgagni situates the evidence of Celsus in the
iatromechanic debates contemporary with Morgagni himself, without com-
mitting Celsus to the defense or attack of iatromechanism: Celsus’ testimony
is valued in pursuit of the larger medical problem.
Morgagni as a Philologist
Morgagni combined the activities of a philologist with the activities of
a physician. Consider first his philological activity. In Letter 1, Morgagni
estimates that he might suggest 200 emendations.11 In Letter 2, Morgagni
makes his famous textual discovery.12 Morgagni used the Venetian manuscript
Marcianus Latinus vii 8, a copy of its ancestor manuscript J kept in Florence.13
Like its ancestor, the Marcianus provides a chapter listing of topics. While the
9
Letter 1 dated December 14, 1720; Letter 2 dated August 1, 1721; Letter 3 dated December 4, 1721;
Letter 4 dated December 15, 1749; Letter 5 dated December 21, 1749; Letter 6 dated December 26,
1749; Letter 7 dated January 2, 1750; Letter 8 dated January 10, 1750.
10
Rojouan 1997 is the only study of all Morgagni’s letters on Celsus known to me; he includes
references to earlier, limited scholarship. He includes a French translation (with which it is possible
to disagree), a discursive commentary on each letter, a listing of Morgagni’s emendations and
subsequent assimilation by editors of Celsus, and a brief guide to the early modern editors and
scholars mentioned by Morgagni. The listing of Morgagni’s variae lectiones in Rojouan 1997, 2.40–
139 is a startling monument to Morgagni’s importance in determining Celsus’ text as printed in
Marx’s 1915 edition. Yet Rojouan too echoes Marx’s 1915, lxviii complaint: “In these letters, however,
where [Morgagni] speaks about the emendation and interpretation of individual places of Celsus, he
is rather copious in his language and more generous with his eloquence than his emendations.”
11
Morgagni In Celsum 1 5.52 col.1. Rojouan 1997, 2.136 counts 180 modifications of Celsus’ text.
12
Morgagni In Celsum 2 5.59 col.1: “Understand now what I took from that [Venetian] codex [=
Marcianus Lat. vii 8] . . . the codex has a large lacuna after the words subjicienda sunt [Celsus
4.27.1D], not only on the remains of the page but onto the beginning of the next page until coeuntia.
id faciunt etiam albae olivae &, and beside this lacuna it is written in the same hand ‘two folia are
lacking in the oldest exemplar’ . . . These chapters, therefore, which I judge as most complete
interrupted only by the smallest series of numbers, are as follows: cx. Vulva exulcerata est. cxi. De
Vesica. cxii. De calculis in vesica. cxiii. In omni dolore vesica.”
13
Marx 1915, 16 provides a stemma, updated by Maire 1994 in light of the discovery of the Toledo
supplement, described below, pp. 310–311.

chapter list was probably not made in antiquity, someone using a complete
manuscript of Celsus must have made it.14 Using this table of contents,
Morgagni was able to identify materials missing in the text of Celsus and the
size of the lacuna. The missing material concerned the bladder and calculi in
the bladder, a topic to which I will return. Morgagni used other manuscripts
beyond the Venetian one, such as a manuscript at his hometown of Forli that is
now lost.15 Morgagni was a keen enough Latinist to recognize in Celsus’ usage
new definitions of words that were not listed in the standard Latin dictionaries
of his time.16 In sum, Morgagni acted much like a philologist modern
philologists would recognize: he compared at least ten previous printed
editions of Celsus, he used manuscripts to establish an authoritative text, he
identified novel uses of language peculiar to the author, he suggested
emendations.
Previous scholarship has tried to assimilate Morgagni’s editorial work to
our contemporary standard. For instance, Friedrich Marx, editor of the
standard edition of Celsus (1915), acted as if Morgagni’s medical interests
were separate from his philological study.17 To be sure, some of the above
description of Morgagni’s editorial work conforms largely to our own
standards. In our own contemporary philological activity, founded on
the practices of stemmatics, emendatio ope codicum, and an author’s precise
lexical usage, authorial intent has little role to play prior to the manuscript
recension of the text. Yet in the wake of New Historicism’s comparative
focus on contexts, scholars have come to recognize that our contemporary
philological activity is a selective type of reading of evidence: it places more
weight upon the materialities of preservation than a reader’s insight.
Setting a historicizing gaze upon humanist inquiry has shown that our
contemporary disciplinary practice of stemmatic philology shares historical
roots with certain scientific activities that arose in the early modern period,
as scholars developed methods of records-keeping for observing phenom-
ena that stretched beyond a human lifetime. Both of these disciplinary
activities, dependent on archival practices and the comparison of different
reported authorities, effect a removal of subjective error that aims to
14
Gautherie 2017, 97–99, has suggested that the table of contents found in the J and T branch of the
manuscripts may derive authorially from Celsus. But Riggsby 2019, 22–29, has shown that such
information technologies were so rare in ancient Latin books that Roman authors who used them
were forced to explain their presence and use to unwitting readers. Celsus makes no explicit reference
to a table of contents in the extant text.
15 16
Morgagni In Celsum 3 5.60 col.1; Marx 1915, lviii. Morgagni In Celsum 4 5.71 col.1.
17
Marx 1915, lxviii: “Nevertheless Morgagni asked the learned to take comfort in the edition of
[Johannes Antonides van der] Linden, who, even if he was inferior in medical science to
Morgagni, was far superior as a grammarian and philologist.”

300 marquis berrey
propagate objectivity. Furthermore, the recent efforts of the movement
within contemporary philology called New Philology or Material
Philology to ground the text on material artifacts is a theoretical elabor-
ation of a more textually conservative approach to traditional editing. The
practice of either a conservative stemmatics or a materialist philology
denies readers’ emendatio ope ingenii, that is, the resource of their own
critical insight into authorial intent. The result is that critics rarely chal-
lenge the paradosis of the text.18
Should we conclude with Marx that Morgagni distinguished his discip-
linary study of Celsus as a philological activity from his disciplinary study
of anatomy as a medical activity? Nothing could be further from the truth.
Not only did Morgagni use a historicizing method to scientific inquiry
throughout his medical career, described below, p. 309, but it was
Morgagni’s belief that Celsus was a medical author, with medical inten-
tions, which stood at the center of his studies on Celsus. Morgagni was able
to bring his own anatomical and therapeutic knowledge to bear on
improving the text of Celsus. In this, Morgagni of course follows an
older line of thinking that placed the author’s intention at the center of
literary criticism.19
Morgagni as a Medical Reader of Celsus

Morgagni’s historical awareness imbues his study of Celsus. Already in an
early lecture in 1716 as the new professor of anatomy at Padua, Morgagni
had declared that he was awestruck by the responsibilities of his new
position in measuring up to his predecessors, like Vesalius, previous
professor of anatomy at Padua, as well as encompassing the vast new
amount of medical knowledge that has taken place even in the last 200
years.20 Morgagni realized acutely the vast gulf of specific knowledge that
separated him from his medical predecessors. Morgagni maintained that
18
Timparano 2005 characterizes and historicizes Lachmannian stemmatics; Daston and Most 2015
contextualize and historicize the archival practices of disciplinary philology; Trovato 2014, 21–23,
77–108, 188–190, critiques the current state of New Philology and its older roots; West 1973, 55–59,
argues for challenging the paradosis.
19
Burke 2007, 90–92, has argued that a strict reading of authorial intent did not become widespread
among European intellectuals until circa 1800, although his evidence comes from vernacular
languages, which were treated with greater liberty than works written in classical languages.
Morgagni’s consistent attribution of authorial intent, not only to classical medical authors such as
Celsus but also to his own contemporaries and prior scholar-physicians, reveals his elite academic
conception of reading. See Cunningham 2010, 210–211, on Morgagni’s elitism toward patients.
20
Morgagni Cum Primum ad Anatomen Tradendam in Theatrum Venisset 5.10 col.1.

Celsus’ knowledge was for the most part anatomically accurate, even if
historically limited.
He offers a variety of evidence to substantiate this position. For example,
Morgagni claims that Celsus in the following passage was the first writer of
which we have knowledge to enumerate all three sections of the anatomy of
the ear:21
The passages (foramina) inside the skull are greatest of the eyes, then the
nose, then what we have in the ears. From those which are of the eyes, they
go straight to the brain. The passages of the nose are divided in the middle
by a bone . . . [AA] In the ear the path is first straight and simple; [BB] as it
continues it becomes twisted. [CC] That which is next to the brain is
divided into many thin passages (multa et tenua foramina), through which
the faculty of hearing is.22
My editorial insertion of the doubled letters charts the three passages of the
ear’s anatomy: AA is the outer ear, visible to our eyes; BB is the middle ear
with tympanic cavity; CC is the inner ear, invisible, with the bony labyrinth.
Morgagni had first argued the position that Celsus was the first writer to
enumerate all sections of the ear’s anatomy in his Epistulae Anatomicae
(1728), twenty-one years before this particular letter on Celsus. Modern
commentaries on Celsus reaffirm Morgagni’s anatomical judgment.23
Morgagni maintains the attitude in the course of his emendations that
Celsus’ medical knowledge was accurate. For example, Celsus 8.1.2 cor-
rectly describes the third suture of the skull. Yet others have wanted to read
ad latera “toward the sides” rather than ad aures “toward the ears”:
This suture, called the lambdoid, is thus described by Celsus: tertia ad aures
per verticem tendens, occipitium a summo capite diducit “the third suture,
directed toward the ears along the crest, divides the occiput from the top of
the skull.” But [Pieter] Paaw [1616, A. Corneli Celsi de Re Medica liber
octavus] judges that not ad aures but ad latera ought to be read; following
[van der] Linden [1657, A. Cornelii Celsi de medicina libri octo], [Theodorus
Janssonius van] Almeloveen [1713, Aurelii Celsi de medicina libri octo]24
agrees. Indeed, whenever I find in my remaining ten editions, if you except
that of [Johannes] Caesarius [1528 Castigationes in Cornelium Celsum de re
21
Morgagni In Celsum 4 5.69 col.2. 22 Celsus 8.1.5–6.
23
Scheller and Frieboes 1906, 760. The German medical commentary is a product of Scheller,
a professor of medicine and surgery in Braunschweig, who first published his commentary in
1846. Whether he was influenced by Morgagni’s great authority, I do not know; he certainly had
independent anatomical prowess and thereby the capability to make an independent evaluation of
the passage.
24
Almeloveen published multiple editions of Celsus; Rojouan 1997, 2.142, argues that Morgagni used
his 1713 edition.

302 marquis berrey
medica] in which ad nares is read in most evident error, whenever I find the
same thing in the codex, I sustain the [medical] sense ad aures; which, even if
another codex should be put forward that had ad latera, no less would
I sustain.25
Morgagni would sustain the medical sense over against the evidence of
the manuscript, even if manuscript evidence should confirm these appar-
ently nonsensical readings. We see here a difference in how Marx would
have treated a similar problem. For Marx, stemmatic recension would be
the appropriate solution to arrive at Celsus’ authorial intent. But for
Morgagni, the correct medical sense must be Celsus’ authorial intent.26
And how did Celsus know such correct anatomy all those years ago,
when anatomical dissections were forbidden for so long? Morgagni more
or less attributes Celsus’ correct anatomical knowledge to the lost works of
Herophilus and Erasistratus, the two Hellenistic authorities who made the
first systematic explorations of the human body.27 Morgagni nevertheless
believes that Celsus wrote the De Medicina; it was not a translation from
some Greek text.28 Morgagni would place Celsus among the anatomici
auctores, the anatomical authorities, and rebukes Daniel LeClerc’s (1702,
Histoire de la médecine) claim that Celsus provided only generalities in his
anatomical descriptions.29 Only a man of sophisticated medical observa-
tion and experience could have written the De Medicina.30 For example,
25
Morgagni In Celsum 4 5.69 col.1–2.
26
Pace Nardo 1981, 32–33, who is too eager to claim Lachmann’s editorial principle of primum recensere
for Morgagni. Morgagni was not dull to the value of manuscript evidence, but his lament of
manuscript sources In Celsum 6 5.86 col.2 is motivated by the poor medical state of the extant text,
not its lack of critical character.
27
Morgagni In Celsum 4 5.70 col.2: “But on the other hand, when his character and the long
succeeding centuries, adverse to the incision of human cadavers, come into your mind, you will
not easily think that what Celsus transmits, if they do not disagree about the intimate structure of
a human, is from a source different than the inspection of the wounded and much more was taken
from reading the books of Herophilus and Erasistratus when they were extant.”
28
Morgagni In Celsum 4 5.67 col.1. Much ink was once spilled on Celsus’ authorship, a debate about
Quellenforschung now ignored when classicists recognize that Latin technical authors contribute as
much ingenuity and originality to their handling of technical material as Greek sources; compare
Gautherie 2017, 77–80. Our modern historiography cares whether Celsus was a physician because
the historical accuracy of the information he provides about drugs and surgeries is at stake. The
contemporary doubt as to whether Celsus was a practicing physician arises from his putative noble
status and from the fact that the De Medicina was the second part of an encyclopedia embracing also
agriculture and rhetoric. Morgagni holds the position that Celsus was a physician with experience
and intelligence, even with the knowledge that Celsus wrote an encyclopedia called Artes: Morgagni
took this knowledge, like we do, from the manuscripts of the J class that preserve the alternate book
numbers of the encyclopedia.
29
Morgagni In Celsum 4 5.70 col.1.
30
Morgagni In Celsum 4 5.66 col.2–67 col.1: “For a long time the most serious men, among them
[Jerome] Rossi, men who differ in their opinions as to whether he was a physician, have noted that

Celsus’ recognition of new diseases in 1.pr.49 wins Morgagni’s praise in
a retrospective diagnosis for his accurate observation of the time of death.31
Morgagni’s biographical reading of Celsus’ De Medicina yields an author of
medical skill and professional competence, who was concerned with med-
ical accuracy as it was known in his time.
Yet Morgagni is aware of the historical limitations of Celsus’ medical
knowledge and is not above criticizing him. In the De Sedibus et Causis
Morborum (1761), Morgagni writes:
For although Celsus says [8.4.7], “it rarely but sometimes happens that the
bone remains whole and, broken from a true blow, some blood vessel in the
membrane of the brain emits blood,” yet the frequent dissection of bodies
has since shown us that something of this kind, or even a more violent
injury, very frequently happens without any injury to the bone.32
Celsus’ pathology recognizes the correct injury to the meninges but
mistakes the frequency of its occurrence. As we saw before (pp. 301–302),
Morgagni draws on his expertise in recognizing Celsus’ anatomical knowledge
when describing the passages of the ear. Here, he again brings his own ana-
tomical experience and expertise to bear in criticizing Celsus. Importantly, his
experience is the tool he wields to judge Celsus’ medical intent.
Pathology and Iatromechanism

I turn now to a particular case study of Morgagni’s reading of Celsus.
Morgagni was fundamentally concerned with pathology: that is, how
diseases occur and localize within the body. He conducted dissections of
very many corpses throughout his life.33 His postmortem anatomical
studies aimed to determine correct structure; his postmortem pathological
studies aimed to discover how disease was localized in the body after death.
Zampieri has argued that, since Morgagni’s work treats postmortem
morbid anatomy and in vivo clinical symptoms as two distinct orders of
phenomena, it was necessary to compare them on some theoretical basis of
a conjectural mechanical medicine.34 Space prevents me from exploring
Morgagni’s empirical sense of mechanism in depth. At least Morgagni
employed mechanistic ideas throughout his career, although not all his
only a man could have written this text who excelled not only in doctrine but practice and
observation.” See Rojouan 1997, 2.157 on Rossi.
31
32
Morgagni De Sedibus et Causis Morborum §52.1 4.297 col.1.
33
Cunningham 2010, 206, relying on Long, A History of Pathology (1965), estimates at least 700.
34
Zampieri 2016, 325–327.

304 marquis berrey
understanding aligns with Friedrich Hoffmann’s iatromechanism.35
Morgagni brought his understanding of pathology and mechanism to his
work on Celsus.
Morgagni’s manuscript discovery was the lacuna at Celsus 4.27.1D,
where Celsus should have discussed the condition of the bladder and the
origin of bladder calculi. In 7.26, Celsus gives a famous and extant
description of surgery for bladder calculi. Celsus lays the groundwork for
the surgery by describing the anatomy of the urethra in both men and
women and describing where calculi are caught.36 By Morgagni’s day the
Celsian surgery was still one of four operative procedures for bladder
calculi, although it had nearly gone into abeyance until rescued by
Morgagni’s prominent surgical contemporary and dedicatee of the
Epistulae in Celsum, Lorenz Heister.37 Morgagni’s editorial activity gave
much attention to this surgery, since both Letter 1 and Letter 7 consider the
Celsian surgery in detail. I quote a lengthy excerpt from Morgagni’s
commentary in Letter 1:
In the different words that occur in [Celsus] 7.26.1 Nonnunquam etiam
prolapsus in ipsam fistulam (that is the urethra) calculus, qui subinde ea
extenuatur, non longe ab exitu inhaerescit. In these words [Robert]
Constantin does not show what we want. For he marks in the margin
urina, vel subaudi “urine, or it be understood,” too far from what seems to
be sought, since the word urina is not properly placed: nor moreover does
urine thin the calculus that sticks but enlarges it; or if it thins, why then
thinned does it stick? I hoped that for qui either quia or qua could be found
in some book, so that Celsus would say: because or where (quia vel qua) the
urethra is thinned; therefore, the calculus sticks in that part. . . . But that
according to nature the urethra is accustomed to be contracted on its very
end, or on its exterior orifice as I have observed even in women [Morgagni
Adversaria Anatomica 4.24], as you know. I mention the matter particularly
because where Celsus in the next section 4 of this chapter speaks about the
lesser calculi of women, when all books which I have seen, unless my
memory deceives me, have the following [Celsus 7.26.4]: ergo et per se
saepe excidit, et si in primo quod est angustius inhaeret, eodem tamen unco
sine ulla noxa educitur “therefore it often falls out on its own, and if it sticks
in the terminus where it is narrower it is nonetheless draw out by a hook
35
Zampieri 2016, 239–278 shows in detail the importance of synthesis between the new physics
brought into medicine by Malpighi and classical empiricism in Morgagni’s mechanical conception
of medicine; see, especially, pp. 276–277.
36
Celsus 7.26.1C, 7.26.2O, 7.26.4.
37
See Conde Parrado 2008, 238–240, on the disappearance of the Celsian operation. Ellis 1969, 4–25,
contains an excellent overview of the history of the Celsian, the Marian, the high, and the lateral
operations for bladder stone.

without any harm,” where [Theodorus Janssonius van] Almeloveen [1713]
and [Johannes Antonides van der] Linden [1657] have primo, et si in urinae
itinere, quod est angustius, inhaeret “firstly, even if it sticks in the urethra
where it is narrower.”38
Morgagni’s discursive passage blends philology, his own medical experi-
ence, the surgical perspective, and correct medical evaluation of pathology
to produce an improved text of Celsus. In the quoted passage, Celsus’
initial Latin must mean “sometimes the calculus slides into the urethra
itself, the calculus is thinned by it from that point and sticks not far from
the exit.” The editor Constantin wanted it to mean “the calculus is thinned
by the urine.”39 Morgagni, on the other hand, wanted the text to mean
“because the urethra is thinned from that point, the calculus sticks there.”
The dispute between the two versions hangs on how to construe the
relative pronoun qui and the pronoun ea. The Latin text that previous
editors print forces the reader to construe the masculine qui as “calculus”
and the feminine ea as “urine” in the ablative case. Morgagni objects to the
empirical accuracy of the reading, not the grammar: the urine is responsible
for enlarging the calculus, not thinning it. Morgagni invokes the criterion
of medical accuracy in Celsus that he defends in extenso elsewhere to note
the insufficiency of the printed Latin. Morgagni’s philological solution is
economical and elegant. The masculine relative pronoun qui should be
emended to the feminine relative pronoun qua “where” or the conjunction
quia “because”; in either case, the pronoun ea should now be read as the
nominative subject of the verb extenuatur. Morgagni invokes his own
experience to defend the text that now reads “because the urethra is
thinned from that point.” His own anatomical investigations have shown
in detail the narrowing of the urethra in women at the terminus.40 Further,
38
Morgagni In Celsum 1 5.57 col.2–58 col.1.
39
Constantin published Aurelii Corn. Celsi de re medica libri octo cum adnotationibus et correctionibus
in 1566, but the first time he is mentioned in Morgagni In Celsum 1 5.51. col.1 readers are referred to
“Adnotat. ad Cels.,” which might be this volume or abstracts preserved elsewhere. Rojouan 1997,
2.147 notes that Almeloveen reprinted Constantin’s notes and scholia in Almeloveen’s 1697 edition
of Celsus.
40
Morgagni dated the Adversaria Anatomica IV in February 1719; he dated the first letter on Celsus in
December 1720. Morgagni Adversaria Anatomica IV §24 1.124 col.1: “In the description of the female
urethra [Regnier de] Graaf omitted a few things which I have added. I shall pass over the fact, as I am
accustomed to observe, that it becomes narrower on the very end; it certainly does not lack
canaliculi, by which it is lined along the interior. You could recognize this from [Laurentius]
Terraneus describing them beyond those ‘which Graaf placed in the beginning of the urethra and
called lacunae, which the anatomists only commonly describe,’ as if I had not described them three
years before him: Lorenz Heister has recently noted this. Since moreover Terraneus calls them
‘small’ and ‘produced from a glandular body underneath,’ you may know that, even if rarely they
appear but not so rarely, often, in fact, as even now in the three or four women I recently

306 marquis berrey
Celsus’ own text in a later passage shows that he himself was aware of this
anatomical fact. It is possible to understand the final Latin phrase (in
Celsus 7.26.4) in primo adverbially, “at first, in the beginning, principally,”
equivalent to the adverbial primo “firstly” printed by Almeloveen and van
der Linden, but Morgagni is certain in primo must refer to the anatomical
terminus of the female urethra. Morgagni’s anatomical reading of Celsus
leads to a philological restoration of the text. In a succeeding passage,
Morgagni continues investigating Celsus’ description for bladder calculi.41
He emends a detail about where the surgery must happen in order to
remove calculi successfully. Modern editions continue to follow
Morgagni’s emendations.42 Morgagni’s medical assurance about the ana-
tomical location of calculi is the key to his reading of the passage. Where
are calculi caught? They are found in the narrowing of the urethra and in
the bladder neck. Celsus’ text ought to reflect these facts. We see how
Morgagni’s experience in pathology and anatomy drives his philological
reasoning.
So much for the surgery of calculi. But how do calculi arise so that they
are caught in the neck of the bladder or urethra terminus in women? In the
forty-second letter of his masterpiece, De Sedibus et Causis Morborum per
Anatomen Indagatis (1761), Morgagni dealt with the pathology of bladder
stones from the evidence of postmortem dissections. In relating the history
of a particular case, he upbraids a lithotomist whose rashness killed the
patient by quoting the warning of Celsus 7.26.2I: “a calculus, when
removed by force, makes its own path unless it has one ready.”43 He cites
Fredrich Hoffmann in defending the view that red wines are healthy and
do not provide calcareous material.44 Morgagni winds his way through
these and numerous other authorities; evidence from multiple case histor-
ies seems to divide between calculi that originate in the kidneys and those
that originate in the bladder. At last Morgagni arrives at a conclusion.,
From what Aristotle has said elsewhere [Historia Animalium 1.3.15], “not
only a humor, but even some dry sediments descend” into the bladder “from
which calculi may be formed,” we understand the most ancient origin of the
dissected . . . the mouth of the urethra was sometimes elliptical, sometimes semielliptical, sometimes
circular, many were found by me to be no less than those he drew in the moderate ductus of the male
urethra, some were equal to the larger ductus, and several even larger. Some of them, chiefly the
lesser, I clearly saw pertain to the glandular body of the urethra; most, however, even when they were
lesser, had insertions into canaliculi, most similar to those in the male urethra.”
41
42
Marx 1915 accepts Morgagni’s proposed emendations quia in 7.26.1 and patefacta in 7.26.20.
43
Morgagni De Sedibus et Causis Morborum §42.9 4.158 col.1–2.
44

opinion of those who maintain that all the beginning of calculi of the
bladder arise from the kidneys and that in these further a peculiar nucleus
is always found in the center. Although I do not deny that this is true in
many people, sometimes nonetheless I should more readily join with
Hippocrates [De Aere Aquis Locis 22–23], who taught that calculi are
excreted from urine which is long pressed and too thin in itself: “that
which is thickest and most turbid is heaped up together and made
sediment” . . . whether incipient calculus, or the material of calculus, des-
cends from the kidneys into the bladder, or is generated in the bladder;
certainly calculus takes its start from the same material; nor foreign from the
truth do those seem who maintain that particles of material join together
more firmly in proportion to a slower start; less surely in proportion to more
speed; they seem right who maintain that it is greater in summer than
winter, since in the summer the material is much more diluted by the water,
then the greater portion of the urine goes off through the skin: this seems to
me be another reason why, if we have the option, the excision of the calculus
should be put off from autumn until spring, rather than spring to autumn.45
In the long letter of this article, Morgagni describes a variety of evidence for
the conclusion that the substance of the urine itself is responsible for
bladder calculi. Central to his argument are the numerous cases of
women who have introduced a pin-needle into their urethra, which is
subsequently fallen partially or completely into the bladder; a calculus then
began to accumulate around the needle.46
It seems as if the debate contemporary to Morgagni as to whether
calculi originate in the kidneys or bladder has ancient roots in the dispute
between Aristotle and Hippocrates. Morgagni claims that the ancient
dispute reflects an early modern division between the explanations of
iatromechanists and humoralists, following Aristotle and Hippocrates
respectively. Friedrich Hoffmann, whose Medicina Rationalis Systematica
(1730) stands as the best source of iatromechanist thought, systematizing
his own lifelong work, himself attributed bladder calculi to their forma-
tion in the kidneys:
It is notable that calculi endowed with large, notable branching in the
substance of the kidneys, unnoticed without any unusual pain, can hide
for many years; but as soon as they move from their seat and descend into
the muscular canals narrowed with sheath called the ureters and effect their
exit through them into the bladder, the most severe symptoms arise. . . . And
it ought to draw notice that calculi reside more frequently in the left kidney
than the right; whence it happens that pains from renal calculi are observed
45
Morgagni De Sedibus et Causis Morborum §42.18 4.163 col.1–col. 2.
46
Morgagni De Sedibus et Causis Morborum §42.20–27 4.164. col.2–169 col.1.

308 marquis berrey
by far more frequently on the left than the right . . . nor is the reason of this
outcome in doubt: for in the right kidney, because the hepatic duct covers it
with great mass and warms it with greater heat than the other, the progres-
sion of blood comes much more promptly and quickly through the vessels of
this kidney, and likewise the separation of urinary fluid: hence not so easily
are there stagnations of blood and urine in this kidney, as in the left, which,
since it meets with the bend of the colon, is more pressed by frequent flatus
and restagnation. From which it happens that the more freely the course of
the blood is intercepted in compressed vessels, the more difficult it makes
the transportation of urinary solution by tubules, and the more promptly it
causes there the stagnation, separation, and concretion of tartarous
material.47
Hoffmann here argues that the calculi arise from the substance of the
kidney itself. He provides an anatomico-mechanical basis for his conclu-
sion. The right kidney is more heated, which allows urine to pass more
easily into the bladder. The left kidney, however, is more compressed by
the colon at the splenic flexure; the pressure hinders the passage of urinary
fluid. The tartarous material, an archaic chemical term for the calcium
oxalate salts that form the majority of calculi, accumulates, precipitates,
and solidifies.48 The concretion of the material of the urine under pressure
was the iatromechanistic explanation for calculi; it was primarily the
pressure exerted in the body itself upon the kidneys that was responsible
for their production. Hoffmann provides an explanatory mechanism for
calculi that assigns proximate cause to the dynamic structure and motion of
bodily anatomy: in a word, iatromechanism.
Morgagni, ever the master of the graceful sidestep, did not disavow in
the passage quoted above from the De Sedibus that it was possible for
calculi to begin in the kidneys, an indirect reference to Hoffmann’s
iatromechanistic explanation. But he argued that it was not pressure
that caused the formation of most of the calculi. He cited the testimony
of Hippocrates that it was the material in the urine itself that caused
calculi, principally in the bladder. Supporting evidence also came from
the cases of women who introduced a pin-needle into their urethra,
around which calculi accumulated. It is notable that, while Hoffmann,
who was not averse to citing predecessors, offered no ancient authorities
to support his claims, Morgagni charitably offered him the support of
47
Hoffmann 1734, §2.6.4, §2.6.6 in pp. 362–364. Compare Müller 1991, 127.
48
A standard technical overview of the modern understanding of the etiology and pathogenesis of
calculi is Pearle, Antonelli, and Lotan 2016, a reference I owe to E. J. Wright MD, Chief of Urology
at Johns Hopkins Bayview Medical Center. The comparison of modern and historical understand-
ing of calculi deserves a deeper account than I have space for here.

Aristotle: “we understand the most ancient origin of the opinion of those
who maintain that all the beginning of calculi of the bladder arise from
the kidneys.” Morgagni’s account of the difference separates – if we may
so call it – an humoralist explanation (the material of the urine) from an
iatromechanistic one (the pressure upon the kidneys). By assigning the
testimony of Hippocrates and Aristotle respectively to the ancient ter-
minus of each explanatory claim, Morgagni thereby constructed
a historical genealogy to his contemporary debate over the origin of
calculi.
Are we wrong to follow Morgagni in construing the question about the
origin of calculi as a debate stretching from antiquity to Morgagni’s own
contemporaries? We have seen that Morgagni treated the history of
medicine in which ancient and modern authorities were engaged in the
evaluation of the same evidence available to the senses. Many scholars
have called Morgagni’s medicine an empirical ordering of the body, no
doubt in reference to the postmortem anatomical pathology on display in
the De Sedibus.49 Yet an historical ordering of empirical testimony is also
evident in Morgagni’s discussion of the origin of calculi, one uniting past
and present causal disputes. The contemporary historian’s hesitation to
impose a consistent terminology of humoralism or iatromechanism on
such disparate sources – ancient vitalism and teleology in Hippocrates
and Aristotle, early modern mechanism and physicalism in Hoffmann
and Morgagni – arises from our postmodern awareness of how termino-
logical homogeneity across time can conceal the rupture in contextual
meaning.50 Still, whether Morgagni is imposing a unity of authorial
concern to these disparate sources is a modern historian’s question. He
himself gave no heed to the cautions of nominalism; his own authorial
intention was certainly to map the opinions of previous authors and
contemporaries onto the transhistorical pathology of the human body.
Once we have properly understood Morgagni’s historical genealogy, we
ought to read Morgagni’s emendations to Celsus in the same spirit. In his
emendation of Celsus 7.26.1, quoted above in this section (p. 304),
Morgagni objected to previous editors’ texts on the grounds that the
urine adds to the calculus. Celsus’ text ought to read calculus, quia subinde
ea extenuatur, non longe ab exitu inhaerescit “because the urethra is thinned
from that point, the calculus sticks not far from the exit.” We have seen in
49
For example, Foucault 1973, 125–129.
50
Several papers in this volume have emphasized the at best provisional and limited way in which texts
from ancient medicine or mechanics anticipate concepts central to later conceptions of their
disciplines: see Gerolemou (Chapter 4), De Groot (Chapter 5), Ruffell (Chapter 3).

310 marquis berrey
Morgagni’s De Sedibus the reason for the objection: Hippocrates, not
Aristotle, correctly described the mechanism for the calculus’
enlargement.51 Furthermore, as Morgagni argues in his emendation, calculi
are found in the narrowing of the urethra and in the bladder neck, so that
surgery can be performed there. Any surgical intervention proceeds on the
basis of this pathological and anatomical reasoning.52 Morgagni offers
evidence that Celsus is not a iatromechanistic author in the lineage of
Hoffmann about the pathology of calculi in virtue of the fact that Celsus
believed the origin of calculi to be the bladder, not the kidneys.
Morgagni’s account of Celsus’ mechanism for the formulation of calculi
infers directly from the Hippocratic model and only indirectly from
Celsus’ own text. Morgagni knew that the very passage 4.27.1D where
Celsus might have laid out the pathogenetic mechanism of calculi
Morgagni discovered was missing, thanks to the table of contents in the
Venetian manuscript. In the 1970s, this passage was rediscovered in
a previously unknown manuscript in Toledo. The recovery of the
Toledo supplement provides direct evidence of Celsus’ views on the
pathogenesis of bladder calculi:
Beyond these calculi also arise in the bladder, with difficulty of urination
and great pain equally resulting. Again, when urine descends without pain
beyond due measure, in an opposite manner wasting consumes a patient.
From this wasting sometimes the flow is more thin, sometimes fast and
thick. Sediments are generally accustomed to fall in this part. . . . By far the
most dangerous matter in calculi is when they arise from a bladder some-
times healthy, sometimes rough [from an injury]. Arising in a healthy
bladder they always are moved in it, which the Greeks call ΠΛΩΤΟΥΣ
[swimmers] from their swimming, or they arise in the very neck of the
bladder, where at first they remain stationary, then afterwards, growing in
size, they slip into the bladder because of their weight. Those calculi arising
from wounds first adhere to them as a crust, then because they are furrowed
in [the wound] are not retained; they fall out by their own weight. Calculi
become mobile from immobile, and so on.53
51
Morgagni’s conclusion of De Sedibus et Causis Morborum §42.18 4.163 col.1–col. 2 is already implicit
in the observation that urine enlarges the calculus from In Celsum 1 5.57 col.2–58 col.1. Morgagni’s
opinion evidently held firm over the forty years separating the publication of these two texts.
52
Morgagni De Sedibus et Causis Morborum §42.18 4.163 col.2 (quoted above, p. 307) suggests putting
off the surgical intervention for calculus until the spring, following Celsus 7.26.2A: “And not in
every season nor in every age nor for every complaint ought this operation [for bladder calculi] be
tried, but only in spring.”
53
Celsus, Toledo supplement to 4.27.1D (not in Marx 1915 but published in Capitani 1974, 170.38–
171.44, 171.70–78).

Celsus suggests that calculi either arise in the bladder itself, where sediment
accumulates, and slip into the neck of the bladder or arise in the neck of
bladder and slip back into the bladder. It is clear that Celsus echoes the
Hippocratic viewpoint quoted in Morgagni’s De Sedibus.54 Calculi arise in
the bladder, because of compaction of sediment.
To reiterate, Morgagni did not know the Toledo supplement. Morgagni
does not use the evidence of Celsus’ humoralism to attack Hoffmann or
argue against him, on grounds that he gives elsewhere in the same letter of
the De Sedibus:
You understand that, in this controversy difficult in other respects on
account of the observations disagreeing so much, I reject none of them by
following the equipoise (aequitatem) of Celsus and the judgment of the most
eminent men. “It is probable,” Celsus [7.14.1] says on a different topic, “that
what one does not know is passed over by each person; and that what one
has not seen is imagined by no one.”55
After all, Morgagni positively cites both Celsus and Hoffmann for different
ends in his discussion of calculi in the De Sedibus. Rather than attack
Hoffmann with Celsus’ ostensible humoralism, Morgagni acted as if it was
necessary to elucidate Celsus’ authorial intent in order to weigh up the
value of his testimony in empirical pursuit of the larger medical problem at
hand: where calculi originate and the mechanism of their pathogenesis.
Morgagni’s investigation of Celsus’ authorial intent is as much a medical
investigation into the plausibility and accuracy of Celsian pathogenetic
mechanism and surgical intervention as it is a philological investigation of
Celsian Latinity and manuscript evidence. In turn, the causal, mechanical
consistency of nature and the human body facilitates Morgagni’s empirical
ordering of Celsian testimony and the evidence of contemporary path-
ology. As Zampieri points out, “In this sense, starting from an empiricist
methodology which was based on observation, Morgagni, by means of
employing a conjectural line of thought, was able to give a historical
turning point to diagnostics and general pathology.”56 While Celsus’
empirical content and ambiguous attitude toward causal etiologies allowed
various medical readers of the early modern period to claim Celsus for their
own theoretical perspective, Morgagni does not infer that Celsus was
54
Celsus often quotes or summarizes Hippocratic writings, although he is a creative imitator and
measured in his Hippocratism; compare Gautherie 2017, 51–52.
55
56
Zampieri 2016, 325–326 (Italian in the original).

312 marquis berrey
a humoralist per se.57 In Morgagni’s treatment, Celsus’ evidence remains
empirical rather than theoretical in nature.
Antiquarianism
Morgagni found Celsus useful for his own medical science. Celsus offered
some facts, or a philosophy, necessary for medical practice and know-
ledge. Morgagni turned to Celsus as a medical authority because he saw
his writing as a part of shared medical enterprise. In open acknowledg-
ment of his sources, Morgagni cited Celsus and used him to authorize
a point: the historical method that lists previous opinions and authorities
was central to Morgagni’s methodological approach to medicine.58
Indeed, Morgagni continued his lifelong engagement with Celsus not
only in his philological letters but throughout his written corpus, as we
have seen.
Morgagni’s studies raise the question whether scholarship has anything
to gain from reading the past through the lens of the present in the case of
historical medical texts. Yes, I answer, subject to certain conditions. The
Latinist Charles Martindale’s influential analysis of the later reception of
ancient poetry, arguing that meaning is translated at the point of reading, is
opposed to the Enlightenment assumption that authorial meaning can be
communicated unproblematically to readers.59 Yet Morgagni’s style of
reading historical medical writings deserves an analysis deeper than the
expected conclusion that it is a product of his times mediated at the point
57
Celsus’ attitude toward the ancient medical sects is ambiguous beyond his explicit plea for a middle
way in 1.pr.45. In the most recent discussion of Celsus’ own authorial intent, Gautherie 2017, 236–
238, argues that the Celsian text shows a complementarity between Rationalist ratio and Empiricist
usus. Celsus’ own ambiguity notwithstanding, early modern readers saw in Celsus an exponent
variously of Empiricism or Rationalism, usually in accord with their own views. Conde Parrado
2003, 80–106, tracks how Celsian passages on surgery, deontology, and causation were received in
the sixteenth century by medical readers who believed that the ancient debate between Empiricists
and Rationalists formed the essential differences between themselves and their contemporary
medical opponents.
58
It is worth underscoring how wide the gap in knowledge had grown between historical medicine and
Morgagni’s contemporaries by 1700 (recall Zampieri 2016, 276 “medicina classica, che aveva esaurito
del tutto la propria vitalità e constituiva, perciò, un freno allo sviluppo ulteriore della teoria e della
practica medicina”). Morgagni’s continued intensive reading of ancient authorities for medical
content should surprise us. We might therefore borrow the title of Najman 2017, a piece whose
themes overlap considerably with the present essay, to call Morgagni’s transhistorical medical
synthesis authorized through ancient and contemporary professional witnesses to the human
body (including himself) an “ethical reading.” Najman applies her coinage to both historical and
contemporary readings of biblical literature, whose participants evaluate their own moral agency by
the ethical philosophy of the source text or textual traditions.
59
Martindale 1993.

of reception. There is a space in scholarship for crediting the professional
insight of historical readers, who judge the intention of other authors
working within the same scientific tradition. To that end, some academic
proponents of tempering historicism have argued that certain generic
modes of scientific writing presuppose a readership supplying an under-
stood frame of intention coincident with the dominant presuppositions of
that science.60 Such an argument does not imply the historiographical
naiveté that ancient and early modern categories of scientific reading or
indeed science at large are identical.61 Reading the past through the present
can be a useful and productive procedure when dealing with authors whose
authorial intention was to hit upon transhistorical truths about the natural
world. Reading the anatomy of antiquity as capturing (some) genuine
knowledge about human anatomical structures in fact aids our historio-
graphical insight into the intent of such authors. Medical history can
operate successfully with a framework that stresses similarity and identifi-
cation with the present. While recognizing that reading the past with
present knowledge presents pitfalls too, I offer Morgagni’s studies of
Celsus with insight into the probable conjectural reconstruction of the
text as proof that the procedure has merits.
I have argued that Morgagni’s reading is sustained by his dual commit-
ments to medicine and philology. In a field such as art or literature, the
attitude of adaptors toward the role that a particular Roman encyclopedia
played in the development of early modern European thought would be
called classicism. Classicism, in turn, is a species of antiquarianism.
Antiquarianism refers to the deep past its authors considered ancient and
has as its object not a dramatic narrative but a series of facts to be assembled
and described.62 Such a strong interest in the deep past contradicts our own
contemporary sensibilities and expectations of novelty in medicine, since
“antiquarianism” in contemporary English is often pejorative. Perhaps
interdisciplinarity would be the positive modern term. I have argued
elsewhere that historical interdisciplinarity is a discourse of belated aesthet-
ics, because it must appeal to two previously existing disciplinary sets of
explanation.63 Belatedness is the referencing of prior models and texts, so
interdisciplinarity – far from an aesthetic of novelty – is a backward-
looking aesthetic of explanation. Applying the lesson that interdisciplinar-
ity is a belated aesthetic of explanation, Morgagni’s researches on Celsus
60
Blåsjö and Hogendijk 2018; Stolberg 2016.
61
Netz 2021 emphasizes historical contingency in a defense of historical objectivity.
62
Bravo 2007, 517. 63 Berrey 2017, 227–239.

314 marquis berrey
are interdisciplinary in the best sense of antiquarian: they concern the
distant past, and they appeal to the belated explanations of two disciplines.
Morgagni knew that medical classicism was antiquarian. Immediately
preceding the title page to the first volume of Morgagni’s collected works,
authorized and published when Morgagni was eighty years old, is the
poetic fragment of a Greek play.
ἆ μάκαρ, ὅστις ἔην κεῖνον χρόνον ἴδρις ἀοιδῆς,
Μουσάων θεράπων, ὅτ᾽ ἀκήρατος ἦν ἔτι λείμων·
νῦν δ᾽ ὅτε πάντα δέδασται, ἔχουσι δὲ πείρατα τέχναι,
ὕστατοι ὥστε δρόμου καταλειπόμεθ᾽, οὐδέ τοι ἔσται
πάντῃ παπταίνοντα νεοζυγὲς ἅρμα πελάσαι.
O happy is he who, skilled at song,
was a servant of the Muses at that time when the meadow was uncut;
now all things are dispersed and the arts have their ends.
We are left as the last in the race; nor will it be possible,
although one looks about everywhere, to drive a newly yoked chariot.64
The Greek author Choerilus uses the traditional metaphors of poetic
composition, the uncut meadow and fresh driving team, to suggest that all
literary styles, all sciences, all arts already have defined genres. There is
nothing new under the sun for the creative and innovative person. It is
a lament of the poverty of coming afterward in time: we are forever
condemned to unoriginality, to antiquarian concerns only. Assuming that
the placement and deployment of Choerilus is authorial, is Morgagni setting
down his mark against those succeeding him in time with the claim that he
has found the ends of all the arts? Or is he commenting ironically on his
attempt at originality in an effort to synthesize and unify the disparate
threads of past knowledge? In view of Morgagni’s antiquarianism, open
acknowledgment of historical sources, and consistent honor toward his
teachers Malpighi and Valsalva, I favor the latter, more melancholy inter-
pretation. Morgagni’s own life work is deeply unoriginal, he announces
programmatically, but it is medical classicism nonetheless. Classicism in
medicine is the contemporary appreciation of a shared medical enterprise
with the past, just as he himself taught the subject.65 Morgagni’s love of
Celsus effected what other readers who practice the antiquarian appreciation
64
Supplementum Hellenisticum 317 is the contemporary critical edition, but I print the poem as found
in Morgagni Adversaria Anatomica 1.ii.
65
Morgagni De Via atque Ordine ab Se in Tradenda Publice Medicina et Anatome Servato 5.8 col.1:
“You are not ignorant which books were interpreted in the traditional manner by teachers of
theoretical medicine: the first Fen of Avicenna’s First Canon, Galen’s Ars Medica, Hippocrates’
Aphorisms. The first two I thought were given to me not so much for interpreting as taking their

of the past always do: he found value in an historical author and reinter-
preted his understanding of the past in light of his own experience. The
effect that readers can have on authors is profound.66
RE F E RE N CE S
Berrey, M. 2017. Hellenistic Science at Court (Berlin).
Blåsjö, V. and J. P. Hogendijk, 2018. “On Translating Mathematics,” Isis, 109.4:
774–781.
Bravo, B. 2007. “Antiquarianism and History,” in J. Marincola, ed., A Companion
to Greek and Roman Historiography, 2 vols. (Malden, MA), 2.515–527.
Bresadola, M. 2018. “The Problems of Anatomica Practica and How to Solve
Them: Pathological Dissection around 1700,” in S. De Renzi, M. Bresadola,
and M. Conforti, eds., Pathology in Practice: Diseases and Dissections in Early
Modern Europe (London), 56–75.
Burke, P. 2007. “Lost (and Found) in Translation: A Cultural History of
Translators and Translating in Early Modern Europe,” European Review,
15.1: 83–94.
Capitani, U. 1974. “Il recupero di un passo di Celso in un codice del De Medicina
conservato a Toledo,” Maia, 26: 161–212.
de Certeau, M. 1988. The Practice of Everyday Life, transl. S. Rendell (Berkeley).
Conde Parrado, P. P. 2003. Hipócrates Latino: El De Medicina de Cornelio Celso en
el Renacimiento (Valladolid).
Conde Parrado, P. P. 2008. “‘Por el orden de Celso’: aspectos de la influencia del
De medicina en la cirugía europea del Renacimiento,” Dynamis: Acta
Hispanica ad Medicinae Scientiarumque Historiam Illustrandam 28: 217–241.
Cunningham, A. 2010. The Anatomist Anatomis’d: An Experimental Discipline in
Enlightenment Europe (Farnham).
Daston, L. and G. Most. 2015. “History of Science and History of Philologies,”
Isis, 106.2: 378–390.
Ellis, H. 1969. A History of Bladder Stone (Oxford).
Fischer, K.-D. 1994. “Der Artz Johann Friedrich Clossius (1735–1787): Leben und
Werk eines Celsusliebhabers im Zeitalter der Aufklärung,” in P. Mudry and
G. Sabbah, eds., La Médecine de Celse: aspects historiques, scientifiques et
littéraires (Sainte-Étienne), 359–371.
Foucault, M. 1973. The Birth of the Clinic: An Archaeology of Medical Perception,
transl. A. M. Sheridan (London).
argument that could be treated and expounded by a man respectful for truth and for his listeners’
utility. Thus I set out clearly and briefly, as I could, the view of this or that author, and then I passed
on to inquire more fully what, about those same matters that Avicenna or Galen had once written,
the former or the latter should think regarding so many new inventions and experiments if [as
Horace Satires 1.10.68 says] he should have been transported by fate to our age.”
66
I owe thanks to audiences in Nicosia, Boston, and Iowa City for feedback, to David Riesbeck and
Luis Salas for their thoroughgoing comments, and to the Rare Books Room at the Bernard Becker
Memorial Library of Washington University in St. Louis for research support.

316 marquis berrey
Gautherie, A. 2017. Rhétorique et thérapeutique dans le De Medicina de Celse
(Turnhout).
Hoffmann, F. 1734. Medicinae Rationalis Systematicae Tomi Quarti: quo specialis
morborum pathologia et huic superstructa solida therapia cum medendi methodo
cautelis clinicis et morborum enarrationibus adiuncta epicrisi exhibentur: pars
secunda doctrinam haemorrhagiarum et dolorum perspicua et demonstrativa
methodo tradens (Halle).
Jarcho, S. 1984. The Clinical Consultations of Giambattista Morgagni (Boston).
Lo Presti, R. 2014. “La place d’Hippocrate dans les grandes querelles médicales du
xviie siècle: le cas du débat sur la circulation sanguine,” in J. Jouanna and
M. Zink, eds., Hippocrate et les hippocratismes: médecine, religion, société.
xive Colloque International Hippocratique (Paris), 461–482.
Lonie, I. 1981. “Hippocrates the Iatromechanist,” Medical History, 25: 113–150.
Maire, B. 1994. “Proposition d’un nouveau ‘stemma codicum’ de la tradition
manuscrite du ‘De medicina’ d’Aulus Cornelius Celsus,” in M. E. Vázquez
Buján, ed., Tradición e innovación de la medicina latina de la Antigüedad y de
la Alta Edad Media (Santiago de Compostela), 87–99.
Martindale, C. 1993. Redeeming the Text: Latin Poetry and the Hermeneutics of
Reception (Cambridge).
Marx, F. 1915. A. Corneli Celsi quae supersunt (Berlin).
Morgagnus J. B. 1762–1765. Opera Omnia in Quinque Tomos Divisa, 5 vols.
(Venice).
Müller, I. W. 1991. Iatromechanische Theorie und ärztliche Praxis (Stuttgart).
Najman, H. 2017. “Ethical Reading: The Transformation of the Text and the
Self,” Journal of Theological Studies, 68.2: 507–529.
Nardo, D. 1981. “Scienza e filologia nel primo Settecento padovano. Gli studi
classici di G.B. Morgagni, G. Poleni, G. Pontedera, L. Targa,” Quaderni per
Storia dell‘Università di Padova, 14: 1–40.
Netz, R. 2021. “The Problematic Greek Miracle,” Syllecta Classica, 31: 1–37.
Nicholson, M. 1992. “Giovanni Battista Morgagni and Eighteenth-Century
Physical Examination,” in C. Lawrence, ed., Medical Theory, Surgical
Practices: Studies in the History of Surgery (London), 101–134.
Pearle, M., J. Antonelli, and Y. Lotan. 2016. “Urinary Lithiasis: Etiology,
Epidemiology, and Pathogenesis,” in A. J. Wein, L. R. Kavoussi,
A. W. Partin, and C. Peters, eds., Campbell-Walsh Urology, 11th ed.
(Philadelphia), 1170–1199.
Ratcliffe, J. F. 2013. ‘The De Medicina, Cornelius Celsus and Surgery in the First
Century ce’, unpublished PhD thesis (University of Queensland).
Riggsby, A. 2019. Mosaics of Knowledge: Representing Information in the Roman
World (Oxford).
Rojouan, J. 1997. “Huit lettres ‘In Celsum’: Morgagni – médecin et philologue,” 2
vols. PhD thesis (University of Nantes).
Scheller, E. and W. Frieboes. 1906. Aulus Cornelius Celsus: Über die
Arzneiwissenschaft in acht Bücher 2nd ed. (Braunschweig).

Stolberg, M. 2016. “Epilogue: Approaches to the History of Patients: From the
Ancient World to Early Modern Europe,” in G. Petridou and C. Thumiger,
eds., Homo Patiens: Approaches to the Patient in the Ancient World (Leiden)
497–518.
Timpanaro, S. 2005. The Genesis of Lachmann’s Method, transl. G. Most
(Chicago).
Trovato, P. 2014. Everything You Always Wanted to Know about Lachmann’s
Method: A Non-Standard Handbook of Genealogical Textual Criticism in the
Age of Post-Structuralism, Cladistics, and Copy-Text, transl. F. Poole (Padua).
West, M. L. 1973. Textual Criticism and Editorial Technique Applicable to Greek
and Latin Texts (Stuttgart).
Zampieri, F. 2016. Il metodo anatomo-clinico fra meccanicismo ed empirismo:
Marcello Malpighi, Antonio Maria Valsalva et Giovanni Battista Morgagni
(Rome).

Conclusions or From Antiquity to the Early Modern
The idea of the body as a machine constitutes one of the central analogies
in early modern Western thought. From Descartes’ Treatise of Man (writ-
ten in the 1630s) to the Iatromechanist School of medicine, and from La
Mettrie’s Man a Machine (1747) and de Vaucanson’s automata to science
fiction’s fascination with cyborgs, robots and androids, mechanical models
have been employed to reproduce and mimic one aspect or another of life
itself.1 One of the aims of the present collection of essays has been to show
that the conceptual origins of this early modern body–machine concept
can be traced back to texts, scientific theories and ideas of classical
antiquity. The technological artefact – be it a simple device or a more
complex machine – in the texts and authors which we have been exploring
does not stand in isolation from the flesh, bones, fluids and organs that
make up the human body; on the contrary, they intersect with the latter in
a number of significant ways.
That said, tracing the history of the body–machine concept in Greek
and Roman sources reveals that connections with the early modern period
should not be stretched too far. What we have been identifying as
a mechanical understanding of the body in antiquity does not always
require the transference of scientifically concrete mechanical principles
into the domain of medicine; there might be exceptions (e.g. in the case
of Hellenistic medicine’s direct engagement with mechanical theories), but
we need overall to remain aware of the fact that, for the most part, the
assimilation between body and machine in Greek and Roman antiquity is
articulated in the context of imaginative, figural associations between the
characteristics of living beings and those of the products of human craft. In
this context, it is also important to bear in mind that, in contrast to the
1
On Descartes’ mechanical philosophy, see Snider 2000; Des Chene 2001; Vaccari 2008; Riskin 2016,
ch. 2. On de Vaucanson’s automata, see Riskin 2003; Landes 2011. See further on seventeenth-century
mechanical philosophy Garber 2013; Roux 2018.
318

Conclusions or From Antiquity to the Early Modern 319
early modern body–machine concept, the ancient evidence which speaks of
certain bodily parts operating/looking like technological artefacts and
machines does not exclude teleology from the picture.2 On the one
hand, the close resemblance between bodily organs and mechanical devices
is often evoked as a proof of Nature’s superior artistry. On the other hand,
mechanical devices are considered to supplement and improve Nature’s
perfect plan. Material models are first deployed as a means of representing
biological processes, for instance, human respiration (see e.g. Empedocles’
clepsydra, B100 D–K); these models, from the fourth century bce
onwards, are increasingly seen as sharing similar final causes to those that
pertain also to living organisms. Finally, unlike what we find in the early
modern period, the ancient body–machine imaginary does not seem to
have any bearing on questions regarding the relationship between body
and soul.3
Our volume first introduces the readers into the beginnings of
a conceptual blending between organic and fabricated elements and high-
lights the concept of hybridity as a notion that helps us think about the
intrinsic relationship between the human body and technological devices.
In this context, man-made artefacts are seen as expanding the limits and
enhancing the abilities of the body; at the same time, they also serve on
a figural level, as points of reference on the basis of which body parts, which
otherwise remain hidden under the skin, are analogically visualized and
understood. With the advent of systematic dissection, a mechanical under-
standing of the human interior becomes more firmly rooted in medical
theory and praxis. The assimilation between body and artefacts of various
sorts and kinds is already boldly attempted in Hippocratic medicine; but it
is only with Hellenistic medicine that it reaches a systematic and, in many
ways, more sophisticated form. By the time we reach Galen, the mechanics
of the human body are so deeply ingrained in medical imagination that
they are seen to be combined, unproblematically, with teleological exeget-
ical frameworks. The potential of human inventiveness thus blends with
nature’s craft, the one reflecting on the other; at the same time, the body
emerges as an instantiation of the grand design that pervades all things in
nature, but also as a living thing that replicates the products of human
wisdom.
2
On Descartes’ anti-teleological account of the human body, see Shapiro 2003. See, though,
Distelzweig 2015, who argues that some of the uses of organic parts in Descartes’ medical writings
are explained with reference to their final causes.
3
See Kakoudaki 2014, ch. 2.

The body–machine interface, as this volume claims, is already explored
in antiquity with the intention of casting light on the mysteries of the
human body and its function, on mechanical worldviews and on questions
of teleology and mechanization. In this respect, this interface represents
a novel way of thinking not just about the body itself, but also about the
ways in which humans as living organisms intersect with products of their
own design.
RE F E RE N CE S
Des Chene, D. 2001. Spirits and Clocks: Machine and Organism in Descartes
(Ithaca).
Distelzweig, P. 2015. ‘The Use of Usus and the Function of Functio: Teleology and
Its Limits in Descartes’ Physiology’, Journal of the History of Philosophy, 53:
377–99.
Garber, D. 2013. ‘Remarks on the Pre-history of the Mechanical Philosophy’, in
D. Garber and S. Roux, eds., The Mechanization of Natural Philosophy
(Dordrecht), 3–26.
Kakoudaki, D. 2014. Anatomy of a Robot: Literature, Cinema, and the Cultural
Work of Artificial People (New Brunswick).
Landes, B. J. 2011. ‘Vaucanson’s Automata as Devices of Enlightenment’,
Sjuttonhundratal, 8: 50–9.
Riskin, J. 2003. ‘The Defecating Duck, or, The Ambiguous Origins of Artificial
Life’, Critical Inquiry, 29: 599–633.
Riskin, J. 2016. The Restless Clock: A History of Centuries-Long Argument over What
Makes Living Things Tick (Chicago).
Roux, S. 2018. ‘From the Mechanical Philosophy to Early Modern Mechanisms’,
in S. Glennan and P. Illari, eds., The Routledge Handbook of Mechanisms and
Mechanical Philosophy (New York), 26–45.
Shapiro, L. 2003. ‘The Health of the Body-Machine? Or, Seventeenth Century
Mechanism and the Concept of Health’, Perspectives on Science, 11: 421–42.
Snider, A. 2000. ‘Cartesian Bodies’, Modern Philology, 98: 299–319.
Vaccari, A. 2008. ‘Legitimating the Machine: The Epistemological Foundation of
Technological Metaphors in the Natural Philosophy of René Descartes’, Oeuvres
Complètes d’Hippocrate, Emile Littré (Paris). in C. Zittel, G. Engel, R. Nanni,
and N. C. Karafyllis, eds., Philosophies of Technology: Francis Bacon and His
Contemporaries (Leiden), 287–336.

Index of Passages
(ps.-) Apollodorus Apollonius Rhodius

Bibliotheca Argonautica
1.140: 73 1.202-6: 51
3.52–5: 58 1.203: 51
(ps.-) Aristotle 4.159: 74
De audibilibus 4.160: 74
800a34–b3: 191 4.1638–48: 73
Mechanica 4.1665–88: 73
847a11–12: 62 Aristophanes
847a11–24: 231 Frogs
847a21–4: 132 860: 78
848a11–14: 135 Peace
848a12–19: 131 741–3: 99
848a34–7: 113 Thesmophoriazusae
849a22–b18: 132 7.65: 109
853a21–2: 148 985: 40
853a33–b2: 129 Wasps
Problemata physica 60: 99
862a27–8: 207 Aristotle
862a27–30: 208 De caelo
862a30–3: 209 271a33: 220
862a31–2: 213 289b30–290a7: 131
290a2–5: 131
Aeschylus 307b31–3: 142
Suppliants De generatione animalium
441: 78 729a12–14: 172
Aetna 734b: 76
324–9: 255 741b4–5: 220
Alcman De insomniis
Partheneion 461b20: 190
3.4–10: 24 De juventute et senectute
Anaximander 470b25–6: 187
DK 12 A 21–2: 182 470b6–27: 182
Apollonius of Citium 473b2–9: 190
In Hipp. De artic. comm. 473b8: 189
15: 120 474a11–15: 191
18: 121 474a16–17: 192
3: 110 474a6–11: 190
3.2: 143 475a8: 186
42: 119, 120 480a16–24: 192
6: 120 480a25–b12: 194
321

322 Index of Passages
Aristotle (cont.) 369a30: 190
480b11–13: 200 383b19–384a2: 213
480b12–17: 186 Physica
480b1–9: 185 258a1–5: 79
De memoria Politica
450a30–2: 214 1253b27–32: 19
De motu animalium 1253b35: 60
698a18–24: 131 1258b 37: 122
701a33–b16: 77 Rhetorica
701b2–28: 131 1411a: 1
De mundo Athenaeus
398b: 76 Deipnosophistae
De partibus animalium 198e–f = Callixenus fr. 2: 82
645a15–16: 220, 233 20a–b: 80
645a23–6: 220 230d: 74
645a28–30: 219 338a: 56
645a30–6: 220 Athenaeus Mechanicus
645a9: 220 p.31 W. = p. 58 W.-B: 234
650a9: 181 Aulus Gellius
669b23–25: 183 Noctes Atticae
674b10, a19: 181 10.12.8–9: 80, 234
681a30: 185
687a–b: 112 Celsus
676b16–677b10: 221 De medicina
De respiratione 4.21.1D: 304
473a15–474a34: 30 4.27: 13
De somno et vigilia 7.10.7: 120
456a15–18: 188 7.14.1: 311
456a6–11: 186 7.26, 13: 304
457a21–5: 167 7.26.1: 304
Historia animalium 7.26.2A: 310
494b21–4: 221 7.26.2I: 306
495a34: 183 7.26.4: 304
495b16: 185 8.1.2: 301
495b3: 187 8.1.5–6: 301
495b5-12: 184
495b7–8: 184 Diodorus Siculus
495b8–10: 184 Bibliotheca historica
495b9: 184 4.64.4: 58
496a24–7: 184 Diogenes Laertius
496a27-34: 184 Vitae philosophorum
496a35–b9: 187 5.57: 227
496b4: 187
496b4–6: 187 Empedocles
515b10–18: 213 DK B33: 172
523a15–17: 213 Euripides
527a24–8: 213 Hecuba
Metaphysica 65: 60
981a2–15: 28–30, 130
981b8: 152 Galen: 263
982b: 233 De anatomicis administrationibus
982b12–19: 233 6.10-1 = 2.574–5 K.: 280
983a: 76 6.3 = 2.547–8 K.: 263
Meteorologica De locis affectis
362a31: 205 1.7.1 = 8.66 K.: 289

De motu musculorum 13.7–9: 86
1.1 = 4.371 K.: 289 13.9: 87, 91
De naturalibus facultatibus 17.1: 114
1.10 = 2.20–4 K.: 265 2.11: 85
1.15-17 = 2.60–73 K.: 284 2.3: 89
1.2 = 2.2 K.: 263 2.6: 78
1.2 = 2.2–3 K.: 273 2.8–9: 19, 74
1.4 = 2.9–10 K.: 263 20: 77
1.5 = 2.10–11 K.: 264 20.2: 84
2.1 = 2.76 K.: 266, 288 22.1: 83, 94
2.2 = 2.78 K.: 224 23.3: 74
2.3 = 2.81–2 K.: 224 24: 84
2.4 = 2.88 K.: 226 24.4: 84
3.11 = 2.181 K.: 277 26: 84
3.13 = 2.198 K.: 289 27: 84
3.13 = 2.198–9 K.: 291 29: 84
3.14 = 2.204–5 K.: 288 30.2: 85
3.15 = 2.206–7 K.: 287 5–8: 79
3.15 = 2.207–8 K.: 287 Belopoeica
3.15 = 2.210–11 K.: 290 p.86 W.: 74
3.15 = 2.212–13 K.: 291 Dioptra
3.7 = 2.162–3 K.: 268 §3: 90
3.8 = 2.168–73 K.: 273 §34: 91
3.8 = 2.168–9 . . . 2.172: 272 Mechanica
3.8 = 2.169–70 K.: 272 2.5: 112
3.9 = 2.178 K.: 265 2.21: 90
De placitis Hippocratis et Platonis Pneumatica
6.5.2 = 5.539 K.: 287 1.16: 74, 80, 81, 88, 89
6.5.2-6 = 5.538–40 K.: 285 1.20: 92
6.7.3 = 5.561 K.: 265 1.41: 87
De symptomatum differentiis 1.5: 92
4.1-2 = 7.62-3 K.: 263 14: 97
4.6-20 = 7.65–74 K.: 289 2.18: 74
5.7 = 7.77–8 K.: 289 2.3: 74
De usu partium 2.32: 90
4.17 = 3.330–1 K.: 279 2.4: 81
4.3 = 3.269 K.: 282 Herodotus
4.3–4 = 3.269–72: 283 Histories
4.4-5 = 3.272 . . . 273–4 K.: 284 2.125: 109
4.5 = 3.272–3 K., 284 5.77: 43
4.7 = 3.275 K.: 270 9.37: 57
4.7 = 3.281 K.: 275, 279 9.38: 57
5.3 = 3.345–6 K.: 280 Hesiod
5.3 = 3.348–50 K.: 282 Opera et dies
6.14 = 3.480–1 K.: 266 70–82: 72
De venae sectione adversus Erasistratum 71: 72
4 = 11.158 K.: 225 Theogonia
De venarum arteriarumque dissectione 146: 109
1 = 2.780–1 K.: 280 Hippocrates
De aere, aaquis et locis
Hero of Alexandria 18 = 2.68–70 L.: 172
Automata 8 = 2.32–6 L.: 164
10: 90 9 = 2.36–8 L.: 173
11: 89 De affectionibus
13: 89 4 = 6.212 L.: 165

Hippocrates (cont.) 8 = 7.506 L.: 168
De affectionibus interioribus De ulceribus
51 = 7.296 L.: 165 12 = 6.412 L.: 172
De arte De vetere medicina
2 = 6.2–4 L.: 151 22 = 1.626–8 L.: 115, 167
De articulis Mochlicon
10 = 4.102–4 L.: 118 38 = 4.382–6 L.: 149
2 = 4.80–2 L.: 111 Homer
42 = 4.182–4 L.: 119 Iliad
47 = 4.200–12 L.: 118 1.584–600: 51
48 = 4.212–16 L.: 117 18.369–81: 75
52 = 4.226–32 L.: 54 18.372–80: 22
58 = 4.248–54 L.: 53 18.373–9: 19
62 = 4.262–8 L.: 55 18.380: 30
7 = 4.88–94 L.: 113 18.416: 51
71 = 4.292–6 L.: 116 18.417–20: 72
73 = 4.300–2 L.: 120 18.418: 20, 72
77 = 4.308–12 L.: 112 18.418–21: 60
De corde 18.468–73: 28, 75
8 = 9.84–6 L.: 222 18.590–605: 22
10 = 9.86–8 L.: 222 18.608: 30
De fracturis 2.490: 43
30 = 3.516–24 L.: 110, 115, 118 4.141–9: 33
31 = = 3.524–8 L.: 112 5.334–43: 74
De genitura 5.902–5: 172
4 = 7.476 L.: 159 6.130–7: 82
De morbis 8.420: 28
4.11 = 7.562–4 L.: 160 9.609–10: 28
4.14 = 7.570 L.: 160 Odyssey
4.18 = 7.580 L.: 172 18.196: 33
4.2 = 7.544 L.: 160, 162 19.564–5: 34
4.20 = 7.588 L.: 172 23.159–62: 34
4.21 = 7.590 L.: 172 5.456: 45
4.24 = 7.600 L.: 172 7.91–4: 74
4.24 = 7.602 L.: 173 7.94: 20
4.3 = 7.544–6 L.: 161 8.258–64: 24
4.3 = 7.546 L.: 164 8.265: 25
4.3 = 7.548 L.: 163 8.274: 23
4.4 = 7.548 L.: 167, 172 8.379–80: 24
4.4 = 7.550 L.: 168 Homeric Hymns
4.5 = 7.552 L.: 163 Hymn to Apollo
4.6 = 7.554 L.: 169 161–4: 44
4.8 = 7.556–8 L.: 170 203: 25
4.8 = 7.558 L.: 174 Hymn to Dionysus
4.8 = 7.558–60 L.: 174 9–14: 82
4.9 = 7.560 L.: 169 Hymn to Hermes
De morbo sacro 38–54, 475–88: 21
16.1 = 6.390 L.: 265 Hyginus
De mulierum affectibus Fabulae
2.57 = 8.344 L.: 163 166: 55
De natura mulierum
109 = 7.428 L.: 163 Macrobius
84 = 7.406 L.: 163 Saturnalia
De natura pueri 1.26: 61
6 = 7.498 L.: 157, 160 3.13.7–9: 83

Oribasius 9.117–19: 25
Collectiones medicae Plato
49.1.1: 114 Leges
49.4.23: 113 644d–e: 76
49.4.24: 113 893 c–d: 131
49.5.15: 113 893d45: 131
Ovid Philebus
Metamorphoses 46d-e: 220
1.82–8: 72 Politicus
10.277–9: 74 260e3: 190
10.280–6: 74 Protagoras
320d–23a: 72
Pausanias Respublica
Description of Greece 514b: 76
5.13.4–6: 33 Timaeus
Philo of Byzantium 84d2: 200
Automata Pliny the Elder
23: 82 Historia naturalis
Belopoeica 2.166: 254
p.72 Th.: 71 28.34: 33
p.73 Th.: 85 7.11.50: 51
Pneumatica Pliny the Younger
40: 93 Epistulae
41: 88 10.33: 255
42: 88 10.34: 255
Syntaxis Plutach
4.4: 40 De amore prolis
Philoponus 495 C: 226
In Aristotelis De generatione animalium Plutarch
77.17: 78 Pericles
Philostratus 27.3–4: 59
Vita Apollonii Polybius
3.27.2–3: 51 Historiae
6.11: 42 12.13.9–12: 81
6.11.18–19: 60 Procopius
Pindar Horologium
Hymn to Zeus §§1–2 (§§2–6 = 22–57 Diels): 75
fr.33d.5–9: 21 §§1–3 (§§1–9 = 1–75 Diels): 93
Nemean §§6 & 8 (§§22 & 43 = 133 & 211–12
11.15–18: 32 Diels): 94
4.82: 33 §10 (§§52–5 = 254–67 Diels): 97
Olympian §11 (§§59–61 = 285–301
1.105: 34 Diels): 98
1.25–7: 31 §13 (§71 = 328–30 Diels): 98
1.27: 33 §14 (§75 = 345 Diels): 99
1.28–32: 34 §3–4 (§§10–13 = 76–95 Diels): 93
1.47–51: 32 §4 (= §14 = 96–101 Diels): 94
6.91–2: 21 §6 (§§24–6 = 144–55 Diels):, 96
7.52: 21 §7 (§§27–8 = 156–66 Diels: 94
Paean 8 §9 (§§44–5 = 215–27 Diels): 95
fr. 52i.116 S.-M.: 43 §9 (§46 = 227–32 Diels): 96
fr.52b.100–1 S.-M.: 43
fr.52i.65–84 S.-M.: 39 Sextus Empiricus
Pythian Adversus mathematicos
12.8–10, 25: 21 1.258: 227

Sophocles Historia plantarum
Oedipus at Colonus 1.3.2: 53
867: 60 5.2.4: 155
Soranus 5.6.2: 206
Gynaecia 5.7.4: 206
1.114: 59
Strabo Ulpian
Geography Ad Sabinum
1.4.40: 40 20.18: 252
Theophrastus Vitruvius
De causis plantarum De architectura
5.12.4–11: 204 10.1.3: 113
De igne 10.2.5–7: 90
37: 199 10.7.1–3: 248
De lassitudine 7, apraef.: 14, 71
18: 213 9.8.4: 77, 234
3.21–3: 210 9.8.5: 71, 92,
De signis
30.207-209: 206 Xenophon
De ventis Oeconomicus
56: 211, 214 6.5: 122
56–7: 205 Symposium
58: 206, 214 4.55: 76

General Index
aesthetics (of the human body) artefacts, 1, 2, 4, 14, 19, 23, 37, 44, 45, 300, 319,
absence of aesthetic judgements in the artistry
Hippocratic Corpus, 221 human, 33
in Aristotle, 218 Nature’s, 221, 231, 319
in Hellenistic medicine (Erasistratus), 223, poetic, 21
226, 230, 231 audience, response, 21, 24, 40, 43, 52, 60, 71, 80,
agency, 3, 4, 59, 157 92, 100, 101, 121, 233
agôn, 24 automata, 21, 28, 48, 51, 52, 59, 60, 63, 70, 71, 74,
Alexandria, Ptolemaic, 218, 234, 235, 247, 256 75, 76, 77, 78, 79, 80, 83, 85, 87, 90, 91, 92,
analogy 94, 100, 101, 318
and metaphor, 1 autopsy, 2, 237
bellows-breathing analogy in axle, 78, 82, 83, 84, 86, 89, 92, 94, 95, 96, 112, 113,
Anaximander, 182 118, 119, 120
between craft and nature in Aristotle’s
teleology, 223 balance / equilibrium, 8, 85, 116, 118, 127, 131, 132,
between the bellows and the lungs, 191, 192 133, 134, 135, 141, 147, 168, 174
‘dual direction’, 164 bellows, 22, 28, 29, 75, 182, 183, 189, 191, 192, 193,
anatomy, 222 194, 195, 196, 198, 199, 200, 245, 246, 258
and admiration of the human body, 226, 236 bench, Hippocratic, 107, 113, 120
and bodily structure, 267 Berryman, Sylvia, 2, 223
and mechanics, 227, 230, 246, 259 bioarchaeology, 48, 57, 58
Hellenistic (Herophilus, Erasistratus), 218 breath, 28, 29, 30, 43, 45, 157, 181, 184, 185, 188,
human, 263 190, 191, 247
of animals, 187
parts of crafted objects resembling human, 5 catapults, 77, 79, 139
androids, 4, 60, 318 cauldron, 22, 23, 25, 26, 31, 33
animals, 20, 49, 74, 77, 101, 183, 185, 187, 188, 189, causes
190, 192, 194, 198, 204, 213, 223, 263, 291 efficient, 192
animation (of inanimate matter/objects), 5, 20, final, 226
21, 27, 34, 37, 38, 84, 98 of disease, 13
anthropomorphism, 5, 23, 42, 70, 74, 76 of respiration, 182, 193
antiquarianism, 313, 314 chariot, 51, 58, 59, 63, 314
Apollo charis/kallos (i.e. beauty), 219
Delphic temple of, 5, 37 circle, 132
Arabic tradition, 71, 87, 93, 256 concentric circles, 131
archaeological evidence, 13, 14, 26, 48, 57, 82, measuring circles, 131
183, 194 moving into a, 24, 77
Archimedes, 132 semicircle, 222
clock, 95, 96, 97, 98 ‘community model’, 63
Law, 131, 132, 134 contraction, 199, 247
architecture, 37, 40, 41, 42, 249 alternating with dilatation, 229, 247
327

328 General Index
contraction (cont.) fire, 30, 33, 50, 182, 193, 195, 196, 199, 250, 251, 252,
causes of, 293 254, 255
involved in swallowing, 266 Flemming, Rebecca, 235
of body parts, 275, 278 fluids, 156, 158, 160, 166, 170, 171, 172, 175
vs. extension of body parts, 266 force
Ctesibius, 71, 77, 229, 234, 246, 247, 248, 258 adhesive, 291
cyborgs, 4, 7, 48, 49, 51, 52, 58, 318 and counterforce, 142
and motion, 146, 232
dancing, 22, 24, 25, 28, 41, 90 centrifugal, 158
Dawkins, Richard, 237, 238 controlled dissipation of, 139
death, 20, 28, 73, 188, 199, 303 expulsive, 279, 288
deception, 34, 218, 233, 239 invisible, 239
deformity, 61, 114, 117, 119 machine-generated, 113, 150
Demetrius of Phaleron, 81 physical, 112, 114, 115, 127, 135, 141, 147, 150, 156
Descartes, 297, 318 pump, 249, 250, 252, 253, 254, 255
diet, 158 fulcrum, 132, 133, 134, 139, 147
Dionysius of Syracuse, 79
Dionysus, 81, 82, 86, 89, 98, 101 genesis, 33, 34, 263
theatre of, 234 pathogenesis, 303, 310, 311
disability, 48, 63, 64 gods, 20, 31, 32, 44, 45, 55, 59, 72, 74, 75, 100, 101
and stigma, 64 Gorgons, 21, 25, 26, 42
disease, 166, 172, 173, 188, 225, 296, 303 gynaecological treatises, 159, 163
aetiology of, 159
disgust, 121, 174, 218, 219, 231 Harvey, William, 297
dislocation, 53, 54, 120, 126, 127, 135, 136, 137, 138, Hephaestus, 5, 19, 20, 22, 23, 24, 26, 28, 30, 33, 49,
141, 142, 146, 147, 150 50, 51, 52, 54, 58, 59, 60, 61, 71, 72, 74, 198
dissection, 219, 221, 235, 263, 302, 303, 319 Heracles, 93, 94, 95, 96, 97, 98, 99, 101, 196
abrupt disappearance of human, 235 Hoffmann, Friedrich, 296, 297, 304, 306, 307,
and vivisection, 223 308, 311
human, 2, 222, 227 Holmes, Brooke, 2, 157, 162, 221
of animals, 221 horror vacui, 246, 247, 287, 288, 290, 292
postmortem, 306 humanoids, 59, 71, 72, 78, 79, 91
distributed cognition, 4 humoralism
dunameis (i.e. innate faculties), 228, 263, and production of pathologies, 159
265 in Celsus, 311
in Hippocratic medicine, 115 in Galen, 14
dwarfism, 64 and iatromechanist explanations, 309
humours
ecphrasis, 93, 100 bile, 159, 169, 170, 172, 221, 282
ekplêxis, 91, 236 blood, 73, 159, 169, 182, 185, 186, 187, 219, 229,
embolê, 135, 137, 146 247, 264, 265, 271, 283, 284, 285, 287, 288,
embryo, 157 303, 308
empeiria, 130 phlegm, 159, 165, 167, 168, 173, 283, 290
Empiricism hybridity, 19, 21, 27, 30, 49, 52, 319
school of medicine, 237 hydraulics, 12, 228
engineering, 6, 7, 12, 108, 117
entasis, 40 iatromechanics, 142, 150, 152, 156, 175, 296
entertainment iatrotechnics, 9, 159
and astonishment, 234 imagination, 77
and automata, 77 medical, 319
of crowds through medical practice, 119 poetic, 63
epistêmê, 126, 129 the audience in Hippocratic medicine, 121
ethics, medical, 108, 121, 297 the audience’s, 101
experiment, 40, 57, 170, 237 vivified objects in Greek, 20

General Index 329
imitation, 128, 129, 130, 134, 141 forced, 127, 143
impairment, 63, 64 impaired, 25, 57, 58
interdisciplinarity, 12, 63, 230, 313 mechanical, 60, 84
intuition, 8, 127, 141, 152 of liquids, 230
of the blood, 265
klepsydra, 92, of the heart’s ventricles, 245
of the lung, 183, 188, 192, 194
ladder, Hippocratic, 107, 118, 119, 120, 144 of the muscles, 272
lever (mochlos)/leverage, 109, 112, 113, 126, 127, oppositions of, 127
129, 131, 132, 133, 134, 139, 140, 142, 148, 149, prevention of, 212
150, 151, 249 up and down, 86
limitation violent, 77
of mechanical devices in reproducing human vs. immobility, 142
forms and life, 76 music, 21, 28, 29, 41, 56, 228
of the human body, as a consequence of mythology, 49, 57, 59, 61, 64, 72, 93
disability, 63
Lloyd, Geoffrey, 2 nature
and skill, 34, 232
magic, 44, 64, 72, 74, 80, 232 contrary to (para physin), 232
marionettes, 79 in accordance with (kata physin), 143, 232
mathêma, 118, 126, 141 neurospasta, 76
mathematics, 107, 127, 131, 133, 142, 156, 234 New Materialism, 3, 4
mêchanê, 85, 108, 109, 218, 236, 238, 239, 266, 279 nutrition, 262, 264, 265, 266, 267, 288, 290, 292, 293
mechanics, 40, 91, 101, 107, 127, 130, 141, 142, 156,
159, 218, 223, 227, 228, 230, 232, 233, 238, 245, oesophagus, 265, 268, 271, 272, 273, 275
248, 251, 258, 259, 260, 267, 293, 319 organic (vs. material/artificial), 20, 21, 32, 33, 319
mechanopoiia, 110 organs/body parts, 148, 170, 275
metals, 75 armpit, 54, 111, 113, 136, 137, 138, 139, 142,
bronze, 21, 23, 26, 36, 39, 43, 57, 58, 72, 73, 74, 143, 144
75, 89, 94, 97, 156, 158, 169, 170, 172, 175, bladder, 167, 172, 228, 265, 284, 299, 306, 307,
249, 252, 253 308, 310
gold, 20, 22, 25, 33, 42, 53, 72, 82, 195 bones, 121, 140, 143, 303
iron, 58, 63, 77, 89, 148, 149, 156, 173, 205, 206 brain, 186, 273, 301
metaphor, 1 ‘cavity’, 160, 162, 167, 170, 275, 292
and analogy, 1, 165 chest, 111, 137, 142, 185, 188, 191, 193
based on analogies between the artificial and clavicle, 143
the natural, 1 ears, 23, 75
in Aristotle, 191 flesh, 34, 74, 160, 167, 219, 264, 268, 271,
meteorology, 164, 213 283, 291
metrical hands, 82, 111, 114, 117, 137, 273
theory, 228 head, 9, 25, 35, 37, 83, 93, 94, 96, 111, 138, 141,
units, 227, 228 156, 165, 167, 168, 287
mimesis, 4, 5, 6, 7, 71 heart, 43, 169, 184, 185, 186, 187, 188, 192, 193,
moisture, 158, 160, 161, 162, 164, 166, 168, 169, 170, 194, 198, 222, 228, 229, 238, 245, 246, 247,
175, 205, 208, 210, 254 248, 271, 285, 287, 288
Morgagni, Giovanni, 296, 297, 298, 299, 300, intestines, 221, 264, 266, 267, 270, 278, 279,
301, 302, 303, 304, 305, 306, 307, 308, 310, 311 280, 282, 292
movement, 40, 41, 89, 94, 98, 133, 134, 137, 139, joints, 61, 88, 112, 130, 142, 152, 190, 204, 207,
146, 151 209, 210, 211, 212, 213
accompanied by music, 41 kidneys, 228, 284, 285, 307, 308, 309, 310
as a result of leverage, 133, 138, 149 legs, 57, 122, 196
causes of, 77, 143 liver, 222, 228, 229, 264, 270, 280, 282, 284,
circular, 133, 140 285, 292
counter-, 138, 139, 141 lungs, 29, 30, 181, 189, 192, 198, 199, 200,
downward, 144 229, 247

330 General Index
organs/body parts (cont.) robots, 4, 60, 318
mouth, 80, 160, 190, 251, 264, 268, 270, rotation, 84, 85, 86, 89, 90, 99, 131, 144
271, 272
pulmonary lobes, 183 satyrs, 29, 93, 98, 99
skin, 3, 5, 33, 139, 165, 221, 307, 319 sculpture, 50, 195
spine, 118, 119 shoulder, 29, 31, 32, 33, 34, 84, 111, 113, 119, 126, 134,
spleen, 169, 221, 222, 226, 265, 271, 282, 283 136, 138, 141, 142, 144, 145, 147
stomach, 160, 228, 265, 268, 271, 272, 273, 275, Sirens, 41, 43, 44
278, 279, 284, 291, 292 slaves, 19, 60, 89
sutures, 301 sound effects, 83, 88, 93, 97, 98
trachea, 184, 185 speed, 54, 131, 132, 307
veins, 160, 169, 247, 270, 271, 279, 280, 284, statues, 19, 21, 36, 37, 72, 74, 82, 83
287, 288, 289, 291 storytelling, 98
windpipe, 184, 188, 190 strength, 60, 188, 209, 211, 255, 257
womb, 167, 168 and skill, 113
ostensive definition, 128, 129, 140, 147 applied by the physician, 108, 109, 118
enhanced via tools, 149
pain, 56, 116, 174, 307, 310 succusion, 107, 120
Pan, 93, 98 surgery, 115, 173, 304, 306
Pandora, 20, 72, 74 sympathy
Pelops, 5, 30, 31, 32, 33, 34 between body parts and organs, 9
pepsis (i.e. digestion), 275, 282 proto-sympathy in the Hippocratic
performance Corpus, 162
of dance, 22, 24, 25, 41
performative aspects of machines, 90, 92 Talos, 72, 73, 74
pestle, 134, 136, 138, 139, 143, 144, 146 technê, 34, 62, 93, 122, 126, 130, 140, 162, 175
pharmaka, 9, 158, 160, 163, 166, 168, 175 technology
philosophy assistive, 48, 49, 50, 51, 52, 58, 60, 61, 62, 63, 64
and medicine, 181 history of, 70, 129
and wonder, 233 limitations of ancient, 90, 101
natural philosophy and the human body, mechanical, 223, 227
214 medical, 107, 108, 110, 113, 116, 117, 173
phusis, 114, 115, 143, 144, 151, 226, 232 pneumatic, 255
plants, 53, 157, 158, 161, 162, 164, 165, 204 teleology
pneumatics, 88, 91, 159, 228, 246, 255, 256, 258 and beauty, 239
posthumanism, 3 and wonder, 220
proportion, 211, 307 combined with mechanistic explanations,
inverse, 132 238, 267
mathematical, 131, 228 in Aristotle, 182, 221
prosthesis/prosthetic limbs, 32, 52, 56, 57, 58, 175 in Galen, 226
pulley, 119, 120, 126, 129, 130 in Hellenistic medicine (Erasistratus),
226
radius, 131, 132, 133 incipient teleology in the Hippocratic
realism, 80, 81 Corpus, 221
regularity, 228 vs. mechanistic explanations, 182
religion, 64 texture of bodily organs, 183, 187, 206, 267
prohibition of dissection, 221 thauma, 91, 220, 231, 232, 236
religious festivals, 81, 101 thaumatopoiia, 71, 79, 80, 91
‘reservoirs’ (of humours), 160, 164, 167, 171, 173 theatre, 81, 83, 98, 99, 100
residues, 264, 265, 268, 271, 282, 283 anatomy theatre, 298
respiration, 2, 10, 182, 189, 190, 191, 192, 193, 200 theory (vs. practice), 90, 116, 118, 190, 225, 319
rhopê/antirrhopê, 139 tools, 112, 114
rhythm (rhuthmos), 40, 41, 227, 228 craftsman’s, 27, 29, 52

General Index 331
material vs. cognitive, 214 void, 188, 230, 246, 250, 287
medical, 111, 142, 147, 149, 156, 172 volcanic activity
traction, 107, 108, 136, 138, 139, 140, 142, 143, of Aetna, 255
151, 272 von Staden, Heinrich, 2, 223, 229, 235, 237, 245,
tragedy, 100 246, 247, 249, 254, 267
tripods, 19, 20, 22, 23, 24, 26, 27, 28, 74
tropos (i.e. method), 145 water
clocks, 71, 91, 92, 93, 94, 95, 100, 227
vegetal pumps, 2, 4, 12, 228, 229, 232, 234, 247, 249,
imagery/analogies between body and plants, 251, 252, 254, 255, 257, 258
157, 164 weight, 77, 78, 79, 81, 85, 88, 90, 109, 117, 119, 129,
verisimilitude 130, 132, 133, 134, 142, 144, 232, 255, 256, 283,
in ancient Greek sculpture, 231 299, 310
in theatre, 100 wheels, 22, 24, 25, 74, 75, 77, 89, 118
vessels, cupping, 9, 112, 156, 165, 166, 167, 168 winds, 204, 205, 206, 207, 208, 209, 211, 213,
vitalism, 9, 158, 309 214, 255
voice, 39, 43 wings, 42, 58, 81, 88, 94, 95
an automaton’s, 28, 72 wonder, 24, 62, 71, 75, 80, 82, 91, 98, 99, 146, 218,
and motion, 43, 45 222, 231, 232, 233, 234, 236, 238, 239
and singing, 24, 41, 42 wooden dove, 80


Body and Machine in Classical Antiquity

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Body and Machine in Classical Antiquity

Uploaded by

Copyright:

Available Formats

BODY AND MACHINE IN CLASS ICAL AN TIQUI TY

This innovative and wide-ranging volume is the ﬁrst systematic

maria gerolemou is currently a research fellow at the Center for

Published online by Cambridge University Press

Published online by Cambridge University Press

Cambridge University Press is part of Cambridge University Press & Assessment,

Published online by Cambridge University Press

List of Figures page vii

An Introduction to Body–Machine Intersections 1

part i blended bodies 17

part ii the technological body 105

Published online by Cambridge University Press

Index of Passages 321

Published online by Cambridge University Press

1.1a–b Eleusis amphora by the Polyphemus Painter, dated page 26

Published online by Cambridge University Press

Published online by Cambridge University Press

Published online by Cambridge University Press

marquis berrey is currently a second-year medical student at the

Published online by Cambridge University Press

Published online by Cambridge University Press

Published online by Cambridge University Press

Published online by Cambridge University Press

Explanations of the human body through technological artefacts and machines

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.001 Published online by Cambridge University Press

Published online by Cambridge University Press

More Than a Thing

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

In Hephaestus’ Forge: The Twenty Tripods and Bellows

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

Homeric poet before him, the artist has effected a combination of

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

Figure 1.3 A red-ﬁgure krater by the Kleophrades Painter of ca. 500–490.

visual repertoire where, on a red-ﬁgure krater by the Kleophrades Painter

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

The Pindaric Pelops

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

https://doi.org/10.1017/9781009085786.002 Published online by Cambridge University Press

a quarter-century before the composition of Olympian 1, shows in its tondo