Professional Documents
Culture Documents
Political Phenomenology
Experience, Ontology, Episteme
Edited by Thomas Bedorf and Steffen Herrmann
Philosophy’s Nature
Husserl’s Phenomenology, Natural Science, and Metaphysics
Emiliano Trizio
Marek Pokropski
First published 2022
by Routledge
605 Third Avenue, New York, NY 10158
and by Routledge
2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN
Routledge is an imprint of the Taylor & Francis Group, an
informa business
© 2022 Marek Pokropski
The right of Marek Pokropski to be identified as author[/s] of
this work has been asserted by him/her/them in accordance with
sections 77 and 78 of the Copyright, Designs and Patents Act
1988.
All rights reserved. No part of this book may be reprinted
or reproduced or utilised in any form or by any electronic,
mechanical, or other means, now known or hereafter invented,
including photocopying and recording, or in any information
storage or retrieval system, without permission in writing from
the publishers.
Trademark notice: Product or corporate names may be
trademarks or registered trademarks, and are used only for
identification and explanation without intent to infringe.
Publication of this book was supported by Grant No. 2017/27/B/
HS1/00735 financed by the National Science Centre, Poland.
Library of Congress Cataloging-in-Publication Data
Names: Pokropski, Marek, author.
Title: Mechanisms and consciousness : integrating phenomenology with
cognitive science / Marek Pokropski.
Description: New York, NY : Routledge, [2022] | Series: Routledge research
in phenomenology | Includes bibliographical references and index.
Subjects: LCSH: Phenomenology. | Cognitive science | Mechanism (Philosophy)
Classification: LCC B829.5 .P6155 2022 (print) | LCC B829.5 (ebook) | DDC
142/.7--dc23
LC record available at https://lccn.loc.gov/2021032724
LC ebook record available at https://lccn.loc.gov/2021032725
ISBN: 978-0-367-46525-4 (hbk)
ISBN: 978-1-032-13792-6 (pbk)
ISBN: 978-1-003-03536-7 (ebk)
DOI: 10.4324/9781003035367
Typeset in Sabon
by SPi Technologies India Pvt Ltd (Straive)
To Paula and Malina
Contents
List of Figures x
Acknowledgments xi
Epigraphs xii
Introduction 1
PART I
Integrating Phenomenology with Cognitive Science 11
PART II
Phenomenology and Mechanism: In Search of Constraints 137
Index 201
Figures
DOI: 10.4324/9781003035367-1
2 Introduction
Figure I.1 T
he fable of the blind men and an elephant. From World Stories for
Children (p. 14), by S. Woods, 1916, Ainsworth & Co. In the public
domain.
References
Bechtel, W. (2008). Mental mechanisms: Philosophical perspectives on cognitive
neuroscience. Routledge.
Craver, C. F. (2007). Explaining the brain: Mechanisms and the mosaic unity of
neuroscience. Clarendon Press.
Cummins, R. (1975). Functional analysis. The Journal of Philosophy, 72(20),
741–765.
Dreyfus, H. L., & Hall, H. (1982). Introduction. In H. L. Dreyfus and H. Hall
(Eds.), Husserl, intentionality and cognitive science (pp. 1–27). MIT Press.
Gallagher, S., & Brøsted Sørensen, J. (2006). Experimenting with phenomenol-
ogy. Consciousness and Cognition, 15(1), 119–134.
Gallagher, S., & Schmicking, D. (Eds.). (2010). Handbook of phenomenology and
cognitive science. Springer.
Glennan, S. (2017). The new mechanical philosophy. Oxford University Press.
Husserl, E. (1969). Formal and transcendental logic (D. Cairn, Trans.). Springer.
Husserl, E. (1977). Phenomenological psychology (J. Scanlon, Trans.). Martinus
Nijhoff.
Lutz, A. (2002). Toward a neurophenomenology as an account of generative pas-
sages: A first empirical case study. Phenomenology and the Cognitive Sciences,
1(2), 133–167.
Madary, M. (2017). Visual phenomenology. MIT Press.
McIntyre, R. (1986). Husserl and the representational theory of mind. Topoi,
5(2), 101–113.
Moran, D. (2013). ‘Let’s look at it objectively’: Why phenomenology cannot be
naturalized. Royal Institute of Philosophy Supplements, 72, 89–115.
Neisser, J. (2015). The science of subjectivity. Palgrave Macmillan.
Petitmengin, C., Navarro, V., & le van Quyen, M. (2007). Anticipating seizure:
Pre-reflective experience at the center of neuro-phenomenology. Consciousness
and Cognition, 16(3), 746–764.
Introduction 9
Petitmengin, C., Remillieux, A., & Valenzuela-Moguillansky, C. (2019).
Discovering the structures of lived experience: Towards a micro-phenomeno-
logical analysis method. Phenomenology and the Cognitive Sciences, 18(4),
691–730.
Petitot, J., Varela, F. J., Pachoud, B. & Roy, J.-M. (Eds.). (1999). Naturalizing phe-
nomenology: Issues in contemporary phenomenology and cognitive science.
Stanford University Press.
Reynolds, J. (2017). Phenomenology, naturalism and science: A hybrid and hereti-
cal proposal. Routledge.
Roy, J.-M., Petitot, J., Pachoud, B., & Varela, F. J. (1999). Beyond the gap: An
introduction to naturalizing phenomenology. In J. Petitot, F. J. Varela, B.
Pachoud, & J.-M. Roy (Eds.), Naturalizing phenomenology: Issues in contem-
porary phenomenology and cognitive science (pp. 1–80). Stanford University
Press.
Sellars, W. (2007). Philosophy and the scientific image of man. In K. Scharp & R.
B. Brandom (Eds.), In the space of reasons: Selected essays of Wilfrid Sellars.
Harvard University Press.
Thompson, E. (2007). Mind in life: Biology, phenomenology, and the sciences of
mind. Harvard University Press.
Varela, F. J. (1996). Neurophenomenology: A methodological remedy for the
hard problem. Journal of Consciousness Studies, 3(4), 330–349.
Woods, S. (1916). World stories for children. Ainsworth & Co.
Part I
Integrating Phenomenology
with Cognitive Science
1 The Concept of Phenomenology
1.1 Introduction
In this chapter, I elaborate a moderate concept of phenomenology that
is amenable to integration with the explanatory framework of cognitive
science. I focus on Husserlian phenomenology, which in my view is the
most advanced and methodologically aware version of phenomenology.
In the broadest sense, phenomenology is the philosophical investigation
of first-person experience and subjectivity. Edmund Husserl is considered
to be the founder of the phenomenological movement, which includes
such important figures as Heidegger, Merleau-Ponty, and Sartre, and
developed an advanced philosophical approach driven by the reductive
method (phenomenological reduction). For Husserl, phenomenology was
a fundamental discipline that adopts the transcendental attitude; that is,
it investigates transcendental subjectivity and the conditions of possibil-
ity for experience as such.1 The concept of phenomenology I propose in
this chapter is weaker and adjusted to the aims of this book; it relies on
Husserl’s idea of phenomenological psychology and takes phenomenol-
ogy primarily as a theory of consciousness and cognitive acts, playing
down its transcendental aspect. Because our topic concerns the natural-
ization of phenomenology and its integration with cognitive science, I
mainly discuss methodological issues. As I argue, the proposed concept of
phenomenology is compatible with the naturalistic framework of cogni-
tive science.
The chapter is structured as follows: First, I introduce selected aspects
of Husserlian phenomenology, such as reduction and intentionality, and
elaborate the distinction between the static and the genetic approaches. I
also introduce the idea of phenomenological psychology, which occupies
the space between transcendental phenomenology and empirical psychol-
ogy and thus seems open to integration with naturalistic frameworks.
Second, I refute some common myths about phenomenology. I argue
against interpretations of phenomenology that see it as a form of intro-
spectionism that stress its alleged anti-naturalism and suggest that phe-
nomenological analysis concerns the qualitative character of experience.
DOI: 10.4324/9781003035367-3
14 Integrating Phenomenology
Finally, I explain why we need phenomenology in cognitive science, why
we need it in the field of consciousness studies in particular, and why it is
important to consider integration.
1.2 What Is Phenomenology
In this book, when discussing phenomenology, I mainly refer to the philo-
sophical tradition started by Edmund Husserl in the beginning of the
20th century. Of course, there were other important figures in the phe-
nomenological movement, but I find Husserlian phenomenology to be
the most advanced version of the position, the most driven by method-
ological considerations; hence, it fits well into the discussion surrounding
the methodological grounds for naturalization—granted, that is not to
say that it is the easiest version of phenomenology to naturalize. By natu-
ralization of phenomenology, I understand its methodological integration
with an explanatory framework for cognitive science (e.g., Roy et al.,
1999). I will return to the issue of naturalization in Chapter 2.
Phenomenology, especially Husserlian phenomenology, is a highly
advanced and sophisticated philosophy, and I cannot reasonably intro-
duce it here in its entire complexity, nor can I address all its important
contributions to various philosophical debates. That would certainly be
beyond scope of this book (for an overview of Husserlian phenomenol-
ogy, see, e.g., Landgrebe 1981, Moran, 2005; Zahavi, 2003). I focus on
methodological issues and those issues in phenomenology that can find
application in cognitive science. Thus, I hardly address the transcendental
level of phenomenological investigation. That level is also played down
in my considerations because, as I will argue, successful integration with
cognitive science requires it. That said, I do not deny that transcendental
phenomenology is an important philosophical endeavor, though separate
from the project of naturalization. I take phenomenology primarily as a
theory of consciousness and mental processes, as “the eidetic theory of
lived-processes” (Husserl, 1980, p. 38), which investigates the structure
of experience. This understanding of phenomenology is incorporated in
the idea of phenomenological psychology (Husserl, 1977), which I elab-
orate later. With respect to the phenomenological theory of conscious-
ness, I discuss mainly methodological topics relevant to naturalization
and integration with cognitive science, as well as selected issues such as
intentionality and consciousness of time, which are among what makes
phenomenology a research field of interest to cognitive science.2
it is necessary to distinguish the two: The epoché is the term for our
abrupt suspension of a naïve metaphysical attitude, and it can con-
sequently be likened to a philosophical gate of entry (Hua 6/260). In
contrast, the reduction is the term for our thematization of the cor-
relation between subjectivity and world.
(p. 46)
1.2.4 Phenomenological Psychology
Husserl often contrasted empirical psychology with transcendental
phenomenology and called for a revision of the former. As he writes in
“Philosophy as Rigorous Science,”
In later works, Husserl came up with the idea of an eidetic and pure psy-
chology, also called phenomenological psychology, which was thought
The Concept of Phenomenology 21
to be the science mediating between transcendental phenomenology and
empirical psychology (Husserl, 1971, 1977; see also Reynolds, 2017;
Thinès, 1977). Accordingly, phenomenological psychology was thought to
‘‘supply the essential insights needed to give meaning and direction to the
research presented under the title ‘empirical psychology’’ (Kockelmans,
1987, p. 6). These insights concern the eidos of consciousness’ acts
expressed in the “scientific concepts of internality” (1977, p. 166), which
cannot be obtained in an inductive and naturalistic approach but are
apprehended in intuition.
Husserl (1977) characterizes phenomenological psychology as a priori,
eidetic, grounded in intuition or pure description, and focusing on inten-
tionality. Phenomenological psychology is an a priori science because it
is primarily interested in “all those essential universalities and necessities,
without which psychological being and living are simply inconceivable”
(Husserl, 1977, p. 33). To put it differently, phenomenological psychology
investigates generic structures of experience, such as the perspectival char-
acter of visual perception. It is eidetic because “only subsequently does it
proceed to the explanation of psychological facts, to theory, precisely their
eidetic explanation, which is naturally for us the first interest” (Husserl,
1977, p. 33) or, in other words, its objective is to explain psychological
facts not in terms of the particular mental states of an individual but in
terms of discovered generic structures of consciousness. Next, intuition
(eidetic seeing) designates the source of the a priori. Finally, intentionality
states that consciousness is always a consciousness of something. Thus,
investigations under the heading of phenomenological psychology con-
cern a twofold aspect: acts of consciousness and the related objects of
experience or, to put it in phenomenological terminology, the noetic and
the noematic. I will return to the issue of the explanatory force of phe-
nomenological “eidetic explanation” in Chapter 3. For now, it is impor-
tant to remember that phenomenological psychology is a nonempirical,
eidetic investigation of how the human mind functions and that it pro-
vides experimental psychology with key concepts, such as intentionality.
It is also important to remark that phenomenological psychology applies
a reduction that is different from the transcendental reduction of tran-
scendental phenomenology. Husserl calls it “eidetic reduction” (1971) or
“phenomenological reduction” (1977). Eidetic reduction relies on exclud-
ing (suspending) all reference to the physical basis of the mental and all
scientific prejudices concerning the physical and the mental. In this way,
the reduction opens the sphere of the purely psychological to intentional
analyses and imaginative variation. The results of these methods are then
described, conceptualized, and applied in empirical psychology. According
to Joseph Kockelmans (1987), phenomenological psychology, although it
employs the reduction, remains in the natural attitude (p. 20). He argues
that “the purpose of the phenomenological-psychological reduction is
not to bring the transcendental subjectivity to light. Phenomenological
22 Integrating Phenomenology
psychology hopes to expose only the foundations of empirical psychology”
(Kockelmans, 1987, p. 21). Eidetic reduction makes it possible to see and
describe the structures of experience, such as noesis and noema, which in
the normal attitude are hidden. It can be applied to psychological research,
and it does not have to lead to transcendentalism.3 Phenomenological
analysis is therefore theoretical work preceding experimental research.
Husserl (1980, pp. 44–45) also suggests that psychological experiments
can be useful for phenomenology; thus, the relation can be considered
reciprocal. When we acknowledge this phenomenological middle ground,
phenomenology appears to be complementary to psychology rather than
opposed to it. As we will see, the difference between transcendental and
eidetic reduction is also of key importance for considering the integration
of phenomenology with contemporary cognitive science.
In the third book of Ideas, Husserl (1980) gives an example of how
empirical psychology and phenomenology are related to each other in
the study of perception:
the subject comports itself toward the Object, and the Object stimu-
lates or motivates the subject. The subject is subject of an undergoing
or of a being-active, is passive or active in relation to the Objects
present to it noematically, and correlatively we have ‘effects’ on the
subject emanating from the Objects. The Object ‘intrudes on the sub-
ject’ and exercises stimulation on it (theoretical, aesthetic, practical
stimulation).
(Husserl, 1989a, p. 231, emphasis in the original)
Let us consider the state of affairs more closely. This is all the more
necessary since at the present time the naturalism predominating so
greatly among psychologists, as among all natural scientists, has as
its consequence an almost universal misunderstanding of the sense
of phenomenology and of its possible achievements for the psycho-
logical science of experience. With this is connected the basically per-
verted view that with phenomenology it is a matter of a restitution
of the method of inner observation or of direct inner experience in
general. Only in this way also are explained those superficial (indeed
not even superficial, because not understanding at all the sense of the
matters) literary rejections of the claim that phenomenology makes,
and must make through its own specific character, of paving the way
for a reform of psychology (as also, on the other hand, of philoso-
phy) that in the literal sense is fundamental and novel.
(p. 33)
And in the third book of Ideas, Husserl (1980) argues that “experimen-
tal psychology should not be abandoned, but rather made incompara-
bly more fruitful through the phenomenological founding” (p. 42). This
higher scientific level of psychology and the fruitfulness of its research
can be achieved by integrating experimental psychology with what
Husserl (1971, 1977) calls phenomenological or eidetic psychology. As
shown earlier, phenomenological psychology would be an a priori science
that provides experimental psychology a systematic theoretical frame-
work, including a typology of intentional processes and their composi-
tion (Husserl, 1971, p. 79).
To sum up, the Husserlian view of naturalism is twofold. On the one
hand, it acknowledges the importance and achievements of the natural
The Concept of Phenomenology 31
sciences, including empirical psychology. On the other hand, it recog-
nizes the naturalistic attitude as naïve, because it assumes the existence
of nature as simply “ready-made” and encourages the application of its
method in studying every phenomenon, including consciousness. These
naturalistic fallacies result in attempts at reductive explanations of men-
tal phenomena, which Husserl is opposed to. Thus, he calls for rethinking
the naturalistic attitude and revising empirical psychology in the light of
phenomenology. In particular, he develops the project of phenomenologi-
cal psychology, which he considers a means to integrating phenomenol-
ogy with empirical psychology.
1.5 Conclusion
In this chapter, I introduced a basic concept of phenomenology that is
amenable to integration with a naturalistic framework. My concept of
phenomenology is based on Husserlian phenomenological philosophy, in
particular on his idea of phenomenological psychology, but it downplays
the transcendental character of his philosophy (in the next chapter, I elab-
orate more on why downplaying transcendentalism is necessary for the
project of naturalization). I take phenomenology primarily as a method
of investigation and theory of consciousness. However, as I argue, a phe-
nomenological theory of consciousness is not about its qualitative charac-
ter, and phenomenological method is not introspection. Phenomenology
investigates the structure of consciousness, for example, various forms of
intentionality, and the genesis of conscious experience, including passive
and pre-reflective processes of constitution. The concept of phenomenol-
ogy I endorse is not anti-naturalistic, although it is in opposition to reduc-
tive naturalistic explanations of consciousness. As I have argued, Husserl’s
view of naturalism is twofold. On the one hand, he explicitly opposes the
naturalistic reductive approach to consciousness. On the other hand, he
acknowledges the importance of empirical psychology but calls for its revi-
sion and integration with phenomenology. I discussed Husserl’s project of
phenomenological psychology, which was thought to provide empirical
psychology with phenomenological concepts of consciousness, including
the concepts of intentionality and intentional functions. Finally, I argued
that integrating phenomenology with cognitive science, also known as
naturalizing phenomenology, is important for studies of consciousness.
I concluded that phenomenology can deliver informative descriptions of
first-person experience and analyses of the structure of experience, and in
doing so, it may provide constraints on other fields. This possibility opens
the way to integrating phenomenology with cognitive science.
integrating
phenomenology with
cognitive science =
naturalizing
phenomenology?
The Concept of Phenomenology 37
Notes
1 Kantian examples of transcendental conditions of experience are time and
space as forms of all sensory intuitions.
2 Another important aspect of the phenomenological theory of consciousness,
and one which I am not addressing in detail in this book, is embodiment
(e.g., Husserl, 1989; for an overview of Husserl’s theory of embodiment, see,
e.g., Taipale, 2014). Husserl’s and Merleau-Ponty’s conceptions of embodi-
ment are key contributions to cognitive science, in particular to the field of
embodied cognition (see, e.g., Gallagher, 2005). Although I agree with the
embodied approach to cognition, I am not going to discuss it in detail for
two reasons. First, it is simply too vast a topic to include in a book mainly
concerned with methodology and the philosophy of science. Second, I take
it for granted, because most of the approaches to cognition that I discuss in
this book acknowledge, more or less, that cognition is embodied and situated
in the environment. Naturally, there are more issues in Husserlian phenom-
enology of potential interest to cognitive science, to mention only a few—
intersubjectivity (Husserl, 1960; for an overview of Husserl’s conception of
intersubjectivity, see, e.g., Zahavi, 2001), memory and image consciousness
(Husserl, 2005), and affectivity (for an overview, see, e.g., Lotz, 2007).
3 According to Płotka (2021), the clear distinction between transcendental phe-
nomenology and phenomenological psychology is problematic because, as he
argues, Husserl operates with clear-cut methods.
4 Note that Merleau-Ponty emphasizes the difference between introspection
and phenomenology in a similar manner. In Phenomenology of Perception he
writes, “this phenomenal field is not an ‘inner world,’ the ‘phenomenon’ is not
a ‘state of consciousness,’ or a ‘mental fact,’ and the experience of phenomena
is not an act of introspection or an intuition in Bergson’s sense” (Merleau-
Ponty, 2002, p. 66). For him, the idea of internal perception expresses a return
to the conception of the “immediate data of consciousness” and therefore is
“a hopeless enterprise.”
5 For a critique of IIT’s axiomatic method and explanation of consciousness,
see Pokropski (2018, 2019).
6 Notice that Husserl was even more critical of philosophy than natural science.
As he writes, “I am not saying that philosophy is an imperfect science; I am say-
ing quite simply that it is still not a science, that it has yet to begin as science”
(Husserl, 2002, p. 250). Thus, the motivation behind Husserl’s project of phe-
nomenological philosophy is to create a truly scientific and rigorous philosophy.
7 It is also important to notice when and to whom Husserl is addressing his
critique, namely, to such figures as Wundt and others, who applied a com-
bination of psychophysical and physiological methods in the early days of
experimental psychology (Kockelmans, 1987).
References
Balduzzi, D., & Tononi, G. (2009). Qualia: The geometry of integrated infor-
mation. PLOS Computational Biology, 5(8), 1–24. doi:10.1371/journal.
pcbi.1000462
Campbell, J. (2002). Rationality, meaning and the analysis of delusion. Philosophy,
Psychiatry and Psychology, 8, 89–100.
Chalmers, D. J. (1996). The conscious mind: In search of a fundamental theory.
Oxford University Press.
38 Integrating Phenomenology
Chalmers, D. J. (2010). Facing up to the problem of consciousness. In The char-
acter of consciousness (pp. 3–27). Oxford University Press.
Colombo, M., & Heinz, A. (2019). Explanatory integration, computational phe-
notypes, and dimensional psychiatry: The case of alcohol use disorder. Theory
& Psychology, 29(5), 697–718.
Cummins, R. (1975). Functional analysis. The Journal of Philosophy, 72(20),
741–765. doi:10.2307/2024640
Craver, C. F. (2007). Explaining the brain: Mechanisms and the mosaic unity of
neuroscience. Clarendon Press.
Dennett, D. C. (1991). Consciousness explained. Little, Brown.
de Warren, N. (2009). Husserl and the promise of time: Subjectivity in transcen-
dental phenomenology. Cambridge University Press.
Engel, G. (1977). The need for a new medical model: A challenge for biomedicine.
Science, 196(4286), 129–136.
Gallagher, S. (2005). How the body shapes the mind. Oxford University Press.
Gallagher, S., & Zahavi, D. (2012). The phenomenological mind (2nd ed.).
Routledge.
Høffding, S., & Martiny, K. (2016). Framing a phenomenological interview: What,
why and how. Phenomenology and the Cognitive Sciences, 15(4), 539–564.
Husserl, E. (1960). Cartesian meditations: An introduction to phenomenology (D.
Cairns, Trans.). Kluwer.
Husserl, E. (1971). “Phenomenology” Edmund Husserl’s article for the
Encyclopaedia Britannica (1927): New complete translation by Richard E.
Palmer. Journal of the British Society for Phenomenology, 2(2), 77–90.
Husserl, E. (1973). Experience and judgment: Investigations in a genealogy of
logic (J. S. Churchill & K. Ameriks, Trans.). Routledge & Kegan Paul.
Husserl, E. (1977). Phenomenological psychology (J. Scanlon, Trans.). Martinus
Nijhoff.
Husserl, E. (1980). Phenomenology and the foundations of the sciences: Third
book of Ideas pertaining to a pure phenomenology and to a phenomenological
philosophy (T. E. Klein & W. E. Pohl, Trans.). Martinus Nijhoff.
Husserl, E. (1982). Ideas pertaining to a pure phenomenology and to a phenom-
enological philosophy; First book: General introduction to a pure phenomenol-
ogy (F. Kersten, Trans.). Springer.
Husserl, E. (1989a). Ideas pertaining to a pure phenomenology and to a phenom-
enological philosophy; Second book: Studies in the phenomenology of consti-
tution (R. Rojcewicz & A. Schuwer, Trans.). Springer.
Husserl, E. (1989b). The crisis of European sciences and transcendental phenom-
enology: An introduction to phenomenological philosophy (D. Carr, Trans.).
Northwestern University Press.
Husserl, E. (1991). On the phenomenology of the consciousness of internal time
(1893–1917) (J. B. Brough, Trans.). Springer.
Husserl, E. (1997). Thing and space: Lectures of 1907 (R. Rojcewicz, Trans.). Kluwer.
Husserl, E. (1999). The idea of phenomenology (L. Hardy, Trans.). Martinus
Nijhoff.
Husserl, E. (2001a). Analyses concerning passive and active synthesis: Lectures
on transcendental logic (A. J. Steinbock, Trans.). Springer.
Husserl, E. (2001b). Logical investigations (Vol. 2; J. N. Findlay, Trans.).
Routledge.
The Concept of Phenomenology 39
Husserl, E. (2002). Philosophy as rigorous science (M. Brainard, Trans.). New
Yearbook for Phenomenology and Phenomenological Philosophy, 2, 249–295.
Husserl, E. (2005). Phantasy, image consciousness, and memory (1898–1925)
(Vol. 11; J. B. Brough, Trans.). Springer Science & Business Media.
Irvine, E. (2013). Consciousness as a scientific concept: A philosophy of science
perspective. Springer.
Kałwak, W., Reuter, M., Łukowska, M., Majchrowicz, B., & Wierzchoń,
M. (2018). Guidelines for quantitative and qualitative studies of sen-
sory substitution experience. Adaptive Behavior, 26(3), 111–127. doi.
org/10.1177/1059712318771690
Kendler, K. S., Zachar P., & Craver, C. (2011). What kinds of things are psychiat-
ric disorders? Psychological Medicine, 41, 1143–1150.
Kockelmans, J. J. (1987). Husserl’s original view on phenomenological psychol-
ogy. In J. J. Kockelmans (Ed.), Phenomenological psychology: The Dutch
school (pp. 3–29). Martinus Nijhoff Publishers.
Kortooms, T. (2002). Phenomenology of time. Springer.
Krueger, J., & Overgaard, S. (2012). Seeing subjectivity: Defending a perceptual
account of other minds. Consciousness and Subjectivity, 47, 297–319.
Landgrebe, L. (1981). The phenomenology of Edmund Husserl. Cornell
University Press.
Levine, J. (1983). Materialism and qualia: The explanatory gap. Pacific
Philosophical Quarterly, 64(4), 354–361.
Levine, J. (2001). Purple haze: The puzzle of consciousness. Oxford University
Press.
Lotz, C. (2007). From affectivity to subjectivity: Husserl’s phenomenology revis-
ited. Palgrave Macmillan.
Merleau-Ponty, M. (2005). Phenomenology of perception (C. Smith, Trans.).
Routledge.
Moran, D. (2005). Edmund Husserl: Founder of phenomenology. Polity Press.
Moran, D. (2013). ‘Let’s look at it objectively’: Why phenomenology cannot be
naturalized. Royal Institute of Philosophy Supplement, 72, 89–115.
Moskalewicz, M. (2016). Lived time disturbances of drug addiction therapy
newcomers: A qualitative, field phenomenology case study at Monar-Markot
Center in Poland. International Journal of Mental Health and Addiction, 14(6),
1023–1038.
Moskalewicz, M., Schwartz, M. A., & Gozé, T. (2018). Phenomenology of intuitive
judgment: Praecox-feeling in the diagnosis of schizophrenia. Avant, IX(2), 1–12.
Nagel, T. (1974). What is it like to be a bat? The Philosophical Review, 83(4),
435–450.
Neisser, J. (2015). The science of subjectivity. Palgrave Macmillan.
Oizumi, M., Albantakis, L., & Tononi, G. (2014). From the phenomenology to
the mechanisms of consciousness: Integrated information theory 3.0. PLOS
Computational Biology, 10(5), 1–25. doi:10.1371/journal.pcbi.1003588
Parnas, J., Møller, P., Kircher, T., Thalbitzer, J., Jansson, L., Handest, P., & Zahavi,
D. (2005). EASE: Examination of anomalous self-experience. Psychopathology,
38(5), 236–258.
Petitmengin, C., Navarro, V., & le van Quyen, M. (2007). Anticipating seizure:
Pre-reflective experience at the center of neuro-phenomenology. Consciousness
and Cognition, 16(3), 746–764.
40 Integrating Phenomenology
Petitmengin, C., Remillieux, A., & Valenzuela-Moguillansky, C. (2019).
Discovering the structures of lived experience: Towards a micro-phenomeno-
logical analysis method. Phenomenology and the Cognitive Sciences, 18(4),
691–730.
Piccinini, G. (2009). First-Person data, publicity & self-measurement.
Philosophers’ Imprint, 9(9), 1–16.
Płotka, W. (2021). A commentary on Pokropski’s functionalist reading of Husserlian
phenomenology. Frontiers in Psychology, 12. doi:10.3389/fpsyg.2021.614842
Pokropski, M. (2015). Affectivity and time: Towards a phenomenology of
embodied time-consciousness. Studies in Logic, Grammar and Rhetoric, 41(1),
161–172.
Pokropski, M. (2018). Commentary: From the phenomenology to the mechanisms
of consciousness: Integrated information theory 3.0. Frontiers in Psychology,
9. doi:10.3389/fpsyg.2018.00101
Pokropski, M. (2019). Phenomenology and mechanisms of conscious-
ness: Considering the theoretical integration of phenomenology with
the mechanistic framework. Theory & Psychology, 29(5), 601–619. doi.
org/10.1177/0959354319868769.
Ratcliffe, M. (2014). Experiences of depression: A study in phenomenology.
Oxford University Press.
Reynolds, J. (2017). Phenomenology, naturalism and science: A hybrid and hereti-
cal proposal. Routledge.
Roy, J-M., Petitot, J., Pachoud, B., & Varela, F. J. (1999). Beyond the gap: An intro-
duction to naturalizing phenomenology. In J. Petitot, F. J. Varela, B. Pachoud,
& J.-M. Roy (Eds.), Naturalizing phenomenology: Issues in contemporary phe-
nomenology and cognitive science (pp. 1–80). Stanford University Press.
Rupert, R. D. (2009). Cognitive systems and the extended mind. Oxford
University Press.
Sass, L. A. (2014). Explanation and description in phenomenological psychopa-
thology. Journal of Psychopathology, 20, 366–376.
Schacter, D. L. (1996). Searching for memory: The brain, the mind, and the past.
Basic Books.
Schwitzgebel, E. (2019, Winter). Introspection. In E. N. Zalta (Ed.), The Stanford
encyclopedia of philosophy. Metaphysics Research Lab. https://plato.stanford.
edu/entries/introspection/
Smith, J. (2010). Seeing other people. Philosophy and Phenomenological
Research, 81(3), 731–748.
Taipale, J. (2014). Phenomenology and embodiment: Husserl and the constitu-
tion of subjectivity. Northwestern University Press.
Thinès, G. (1977). Phenomenology and the science of behaviour. George Allen
& Unwin.
Tononi, G. (2004). An information integration theory of consciousness. BMC
Neuroscience, 5, Article 42. doi:10.1186/1471-2202-5-42
Welton, D. (1982). Husserl’s genetic phenomenology of perception. Research in
Phenomenology, 12(1), 59–83.
Varela, F. J. (1996). Neurophenomenology: A methodological remedy for the
hard problem. Journal of Consciousness Studies, 3(4), 330–349.
Vincini, S. (2020). The epistemological contribution of the transcendental reduc-
tion. Husserl Studies, 37, 39–66. doi:10.1007/s10743-020-09265-8
The Concept of Phenomenology 41
Yoshimi, J. (2007). Mathematizing phenomenology. Phenomenology and the
Cognitive Sciences, 6(3), 271–291.
Zahavi, D. (2001). Husserl and transcendental intersubjectivity. Ohio University
Press.
Zahavi, D. (2003). Husserl’s phenomenology. Stanford University Press.
Zahavi, D. (2013). Naturalized phenomenology: A desideratum or a category
mistake?. Royal Institute of Philosophy Supplements, 72, 23–42.
Zahavi, D., & Kriegel, U. (2015). For-me-ness: What it is and what it is not. In D.
Dahlstrom, A. Elpidorou, & W. Hopp (Eds.), Philosophy of mind and phenom-
enology (pp. 36–53). Routledge.
2 Naturalizing Phenomenology
Reconsidered
2.1 Introduction
In the broadest sense, naturalism acknowledges that we ought to explain
phenomena exclusively in terms of the natural sciences. But philosophy
also investigates phenomena that are within the scope of these sciences,
such as the mind. Thus, a question arises about the relation between
philosophical investigations and those conducted by the natural sciences.
Roughly speaking, we can approach this issue from two perspectives (see
Godfrey-Smith, 2003, pp. 149–153). The first belongs to the radical posi-
that’s
tion of reductive naturalism (e.g., Quine, 1969), which holds that philo-
stupid
sophical issues, specifically epistemological ones, ought to be given over why
to empirical psychology. Accordingly, epistemology ought to be replaced
with psychology, which will ultimately deliver a fully naturalistic account
of the origin of knowledge and belief formation. In this version of natu-
ralism, there is no such thing as a genuine philosophical question. Every
question is to be answered by the natural sciences. A more recent ver-
sion of naturalistic reductionism with respect to the mind and cognition
is motivated by rapid developments in neuroscience (e.g., Bickle, 2003;
Churchland, 1986). Bickle’s take on the position, which he calls “ruthless
reductionism,” is essentially fundamentalistic, as he aims to explain men-
tal phenomena purely in neurobiological terms and considers neurobiol-
ogy a fundamental level of explanation.
The second belongs to a weaker version of naturalism, according to
which there are specifically philosophical issues, theories, and methods,
and thus holds that philosophy cannot be replaced by the natural sci-
ences—and even that the products of the natural sciences can be used as
a resource in philosophical considerations. As such, there is a one-way
relation between philosophy and the natural sciences: scientific research
can help us answer some of our philosophical questions, for example,
questions concerning the mind, but that relation does not work in the
opposite direction—scientific endeavors do not benefit from philosophy,
with the possible exception of the philosophy of science. That being said,
naturalism may be understood as a claim about a continuity between
DOI: 10.4324/9781003035367-4
Naturalizing Phenomenology Reconsidered 43
philosophy and science. In which case, the relation is mutual and nonre-
ductive; philosophy and science coexist, interact, and can cooperate with
each other. There is no consensus, however, as to how this continuity
ought to be understood, that is, what the logical, conceptual, and meth-
odological grounds are for this promising cooperation. Even so, there are
proposals for naturalizing phenomenology that occupy this position (e.g.,
Varela, 1996).
My approach to naturalism and the naturalization of phenomenology1
is close to the second nonreductive approach but differs on two important
points. First, I am skeptical of there being specifically philosophical issues
reserved exclusively for philosophers, with the possible exception of meta-
philosophical issues. I think that there are many phenomena, including
cognitive and conscious phenomena, which may be approached from vari-
ous perspectives and studied with various methods. But I do believe that
next to the perspectives of the special sciences, there is a specific philo-
sophical point of view or, to put it differently, philosophical understanding
of the phenomenon in question. Although addressed in detail in Chapter
3, this picture of philosophical understanding can be briefly character-
ized here in two points: (1) it investigates the constitutive conditions of
a studied phenomenon, rather than its enabling conditions (see Wheeler,
2005, 2013), and (2) it offers a meta-perspective on scientific disciplines
or, as Wilfrid Sellars puts it, an “eye on the whole,” that is, “knowing
one’s way around with respect to the subject-matters of all the special
disciplines” (Sellars, 2007, p. 370). The second point communicates that
the most interesting aspect of the relation between scientific endeavors
and philosophy is not how the natural sciences can help in philosophical
investigations, although I am sure they can, but how philosophical investi-
gations can contribute to scientific, naturalistic explanations. With respect
to the naturalization of phenomenology, that would be recognizing a way
in which a phenomenological understanding of mental phenomena could
contribute to naturalistic explanations of them.
Before developing my position further in this chapter, I review natu-
ralistic approaches to phenomenology and evaluate the idea of continu-
ity between phenomenology and natural science expressed in terms of
constraints. First, I introduce the very idea of the naturalization of phe-
nomenology and, drawing on Jean-Michel Roy, Jean Petitot, Bernard
Pachoud, and Francisco J. Varela, the editors of the seminal volume titled
Naturalizing Phenomenology (Petitot et al., 1999), mathematization as a
method of naturalization. Next, I consider several proposals for natural-
izing phenomenology, including front-loaded phenomenology, the formal-
ization of phenomenological descriptions, semantic bridging using applied
ontology, and neuro- and micro-phenomenology. The main focus of my
approach to the debate is the notion of constraints, which characterizes
the relation between phenomenology and cognitive sciences. However,
I argue against the types of constraints that have been proposed in the
44 Integrating Phenomenology
debate; in particular, I argue that these positions offer either conceptual
constraints, which are too weak, or stronger isomorphic or homeomor-
phic constraints, which are, however, implausible. This critical evaluation
serves as the ground for developing, in subsequent chapters, an alternative
approach to naturalization based on the idea of mechanistic integration
of cognitive science and proposing a different type of constraint that phe-
nomenology could deliver to research fields in cognitive science.
2.3 Mathematization
The naturalization of phenomenology is closely related to the idea of
mathematizing phenomenological descriptions, which was one of the
main issues in Naturalizing Phenomenology. As the editors write,
dynamical models fit much better with the nature of our experience
than computational models do. Other things being equal, this gives
us good reason to prefer dynamical models over computational mod-
els; and if we accept this, then we have accepted that phenomenology
can substantially constrain cognitive science.
(van Gelder 1999, p. 258)
But what kind of constraints between the phenomenal and the dynamic
are we talking about? Gallagher (1997) reads van Gelder’s proposal as
one of isomorphic constraints, that is, that there is a one-to-one mapping
between the phenomenal and neural levels. As Gallagher (1997) remarks,
“the reasonableness of this strategy will depend, I think, on whether one
52 Integrating Phenomenology
takes phenomenology as a strong constraint or a weak one,” and accord-
ing to him, van Gelder appears “to take phenomenology as a strong con-
straint, in the following manner: for van Gelder and Dreyfus the causal
mechanism to be identified on the sub-personal level needs to be isomor-
phic with details explicated in the phenomenological account” (p. 202).
The problem with such a reading is, however, that van Gelder does not
explicitly use the concept of isomorphic constraints, nor does he apply a
strong conception of similarity between the phenomenological and neural
levels. Van Gelder does not clearly articulate what kinds of constraining
relations he has in mind. His claim is very general—that a phenomenology
of time-consciousness can constrain models of underlying mechanisms, for
example, that they should exhibit the dynamics of a network rather than
the sequential order of computational models. I do not feel convinced by
van Gelder, whose proposal of constraints seems very weak and unclear.
It is important, therefore, to consider in detail what types of constraints
phenomenology can place on explanations in cognitive science.
2.4.1 Conceptual Constraints
2.4.1.1 Front-Loaded Phenomenology
Front-loaded phenomenology is one way of incorporating phenomeno-
logical theory into the research practices of naturalistic cognitive science
Naturalizing Phenomenology Reconsidered 53
in a nonreductive manner. It requires neither training research subjects nor
acknowledging a complete phenomenological theory; it simply proposes the
use of phenomenological concepts and distinctions from phenomenologi-
cal literature in the early stages of research, namely, in experimental design.
“Phenomenology comes into the picture by contributing to the experimental
design, by providing clear phenomenological distinctions, which also inform
part of the analytic framework for interpreting the results” (Gallagher &
Brøsted Sørensen, 2006, p. 126). As an example of such a phenomenologi-
cally informed experiment, proponents of this approach refer to research on
the sense of agency (Chaminade & Decety, 2002) and the sense of ownership
(the so-called alien-hand experiment first done by Nielsen, 1963). Roughly
speaking, front-loaded phenomenology favors experimental practice over
methodological considerations. If phenomenological concepts and distinc-
tions taken from the works of Edmund Husserl, Maurice Merleau-Ponty,
and other phenomenologists seem to contribute to our understanding of
experience, then let us use them in experimental designs and see what the
results are. However, it is not entirely clear what constraining role phenom-
enological concepts play in the whole scientific process.2
One possibility is that phenomenological terminology, next to scientific or
folk psychological concepts, contributes to the body of knowledge underly-
ing an experimental design. This is certainly plausible, but such a formulation
of conceptual constraints does not give phenomenology a strong position. It
seems unclear why we should choose the phenomenological tradition instead
of any other approach to first-person phenomena or any other method of
describing and analyzing subjective experience, including folk psychology.
Another possibility is that phenomenological insights could support
the formulation of general heuristics for how to approach the issue in
question and specific testable hypotheses. The former, again, does not
give priority to phenomenology, because proposing an explanatory heu-
ristic, such as decomposition, can be done on the grounds of any scientific
or folk theory, and it is not clear why the phenomenological approach
should be preferred. A solution to this which I develop in Chapter 4
relies on showing that phenomenological decomposition of experience
provides specific functional constraints on functional and mechanistic
models. Testable hypotheses are more plausible. Gallagher and Brøsted
Sørensen (2006) claim that experiments can “test and verify the phenom-
enological description” (p. 131) and that
the mutual constraints not only share logical and epistemic account-
ability, but are further required to be operationally generative, that is
to say, to be in a position to generate in a principled manner eidetic
descriptions that can directly link to explicit processes of biological
emergence. For this to happen at least both sides of the wavy line
must be joined at a level of description sufficiently abstract that it
rightly belongs to both sides at the same time. In other words, the
neurobiological and the phenomenological accounts have to achieve
62 Integrating Phenomenology
a level of formal and mathematical precision to make this passage
possible.
(Roy et al., 1999, p. 68)
2.5 Conclusion
Although the naturalization of phenomenology has been the subject of
debate for over three decades, it is still a controversial topic. On the one
hand, some phenomenologists (e.g., Moran, 2013) reject the idea of natu-
ralizing phenomenology as implausible and at odds with Husserl’s project
of phenomenological philosophy, and others (e.g., Thompson, 2007) call
for a phenomenological redefinition of nature. On the other hand, more
pragmatically oriented cognitive scientists and philosophers who merge
different traditions (e.g., Gallagher & Brøsted Sørensen, 2006; Varela,
1996) argue that naturalization is not only possible but also vital both for
empirical studies of consciousness and phenomenology. The latter mid-
dle road to naturalization, endorsed by positions that I discussed in this
chapter, is a nonreductive integration of phenomenology with cognitive
science. The key issue in this project is finding a methodological bridge to
make such naturalization possible, in particular, considering the mutual
constraints between phenomenology and cognitive science.
As I argued, although conceptual constraints introduce phenomenol-
ogy to research practice, they only constrain explanations in a weak
sense. Constraints understood as “generative passages” or isomorphic
and homeomorphic relations between the phenomenal and the neurobio-
logical appear implausible because they assume some sort of similarity or
continuity between these levels (Roy et al., 1999). There is no satisfac-
tory empirical research supporting there being any such relations, and the
assumption backing it appears unjustified on its own.
I do not see constraining relations as necessarily involved in some
sort of similarity or continuity. There are two reasons for this. First, the
idea of continuity between phenomenology and empirical science seems
amenable to reductionism; that is, if the phenomenological level can be
derived from the neurobiological level, then it may be reducible in prin-
ciple. In my view, naturalizing phenomenology is not a reductive project
but an integrative one; that is, it seeks to answer how scientific and philo-
sophical views on mental phenomena and their representations “hang
together.” Second, I generally agree with the view that science is charac-
terized by disunity (Dupré, 1995; Fodor, 1974) and explanatory plural-
ism (Mitchell, 2002) rather than unity (Oppenheim & Putnam, 1958).
According to the former view, science consists of heterogenic structures
and disciplines that are incongruent and thus irreducible to one another.
Naturalizing Phenomenology Reconsidered 67
That does not mean, however, that heterogenic scientific endeavors
cannot inform or constraint each other. As I argue in the next chapter,
drawing on Lindley Darden and Nancy Maull’s (1977) conception of
interfield theories and from Craver’s proposal of mechanistic integration
of cognitive science (2007), different fields of research can constrain one
another and together contribute to mechanistic multilevel explanations.
These issues shed new light on novel ways of naturalizing phenomenol-
ogy, including alternative types of constraints phenomenology can pro-
vide and thus contribute to explanatory mechanistic models. Thus, it is
of great importance to understand what naturalistic explanation is, what
models of naturalistic explanation are applied in cognitive science, how
different fields of research may constrain each other, and, last but not
least, whether phenomenology can provide such constraints. This is a
clear weakness of the naturalization debate—it hardly addresses the vari-
ety of available explanatory models or the conditions of their integration,
nor does it consider whether phenomenology offers a sui generis model
of explanation and how it might relate to other models. I address these
issues in the following chapter.
Notes
1 For an interesting discussion of the relation between phenomenology and
naturalism, see Reynolds (2017).
2 An earlier version of the discussion of front-loaded phenomenology appeared
in “Phenomenology and Mechanisms of Consciousness: Considering the
Theoretical Integration of Phenomenology with the Mechanistic Framework”
(Pokropski, 2019).
3 The application of formal ontologies in science, for example, computer science and
the life sciences, is a growing field of research (see, e.g., Smith & Ceusters, 2010).
An important contribution to this field was Ingarden’s phenomenological ontology
(e.g., Ingarden, 1965a, 1965b; see also Chrudzimski, 2004; Płotka, 2020).
4 For an earlier version of the discussion neurophenomenological method, see
Pokropski (2019).
5 See Chapter 3 for a more detailed discussion of dynamic explanation and Chapter
5 for a critique of the application of DST in neuro- and micro-phenomenology.
References
Ashburner, M., Ball, C. A., Blake, J. A., Botstein, D., Butler, H., Cherry, J. M., &
Sherlock, G. (2000). Gene ontology: Tool for the unification of biology. Nature
Genetics, 25(1), 25–29. doi:10.1038/75556
Bechtel, W. (2008). Mental mechanisms: Philosophical perspectives on cognitive
neuroscience. Routledge.
Bechtel, W., & Abrahamsen, A. (2010). Dynamic mechanistic explanation:
Computational modeling of circadian rhythms as an exemplar for cognitive
science. Studies in History and Philosophy of Science Part A, 41(3), 321–333.
Beer, R. D. (2000). Dynamical approaches to cognitive science. Trends in Cognitive
Sciences, 4, 91–99.
68 Integrating Phenomenology
Bickle, J. (2003). Philosophy of neuroscience: A ruthlessly reductive approach.
Kluwer Academic Publishers.
Chalmers, D. J. (1996). The conscious mind: In search of a fundamental theory.
Oxford University Press.
Chaminade, T., & Decety, J. (2002). Leader or follower? Involvement of the infe-
rior parietal lobule in agency. Neuroreport, 13(1528), 1975–1978.
Chemero, A. (2009). Radical embodied cognitive science. MIT Press.
Chrudzimski, A. (2004). Roman Ingarden: Ontology from a phenomenological
point of view. Reports on Philosophy, 22, 121–142.
Churchland, P. S. (1986). Neurophilosophy: Toward a unified science of the
mind-brain. MIT Press.
Cornsweet, T. N. (1970). Visual perception. Academic Press.
Craver, C. F. (2007). Explaining the brain: Mechanisms and the mosaic unity of
neuroscience. Clarendon Press.
Cummins, R. (2000). “How does it work” versus “what are the laws?”: Two
conceptions of psychological explanation. In F. Keil & R. A. Wilson (Eds.),
Explanation and cognition (pp. 117–145). MIT Press.
Darden, L., & Maull, N. (1977). Interfield theories. Philosophy of Science, 44(1),
43–64.
Dupré, J. (1995). The disorder of things: Metaphysical foundations of the dis-
unity of science. Harvard University Press.
Fernandez, A. V. (2019). Phenomenology and dimensional approaches to psychi-
atric research and classification. Philosophy, Psychiatry, & Psychology, 26(1),
65–75.
Fodor, J. (1974). Special sciences (or: The disunity of science as a working hypoth-
esis). Synthese, 28, 97–115.
Fuchs, T. (2010). Phenomenology and psychopathology. In S. Gallagher & D.
Schmicking (Eds.), Handbook of phenomenology and cognitive science (pp.
546–573). Springer.
Gallagher, S. (1997). Mutual enlightenment: Recent phenomenology in cognitive
science. Journal of Consciousness Studies, 4(3), 195–214.
Gallagher, S. (2005). How the body shapes the mind. Clarendon Press.
Gallagher, S. (2018). Rethinking nature: Phenomenology and a non-reductionist
cognitive science. Australasian Philosophical Review, 2(2), 125–137. doi:10.1
080/24740500.2018.1552074
Gallagher, S., & Brøsted Sørensen, J. (2006). Experimenting with phenom-
enology. Consciousness and Cognition, 15(1), 119–134. doi.org/10.1016/j.
concog.2005.03.002
Gallagher, S., & Schmicking, D. (Eds.). (2010). Handbook of phenomenology and
cognitive science. Springer.
Godfrey-Smith, P. (2003). Theory and reality: An introduction to the philosophy
of science. University of Chicago Press.
Godfrey-Smith, P. (2014). Philosophy of biology. Princeton University Press.
Hastings, J., Ceusters, W., Jensen, M., Mulligan, K., & Smith, B. (2012).
Representing mental functioning: Ontologies for mental health and disease. In
Proceedings of the third international conference on biomedical ontology (pp.
1–5). University of Graz.
Hipólito, I. (2018). Perception is not always and everywhere inferential.
Australasian Philosophical Review, 2(2), 184–188.
Naturalizing Phenomenology Reconsidered 69
Hohwy, J. (2018). Phenomenology and cognitive science: Don’t fear the reduc-
tionist bogey-man. Australasian Philosophical Review, 2(2), 138–144.
Høffding, S., & Martiny, K. (2016). Framing a phenomenological interview: What,
why and how. Phenomenology and the Cognitive Sciences, 15(4), 539–564.
Husserl, E. (1977). Phenomenological psychology (J. Scanlon, Trans.). Martinus
Nijhoff.
Husserl, E. (1980). Phenomenology and the foundations of the sciences: Third
book of Ideas pertaining to a pure phenomenology and to a phenomenological
philosophy (T. E. Klein & W. E. Pohl, Trans.). Martinus Nijhoff.
Husserl, E. (1982). Ideas pertaining to a pure phenomenology and to a phenom-
enological philosophy; First book: General introduction to a pure phenomenol-
ogy (F. Kersten, Trans.). Springer.
Husserl, E. (1989). The crisis of European sciences and transcendental phenom-
enology: An introduction to phenomenological philosophy (D. Carr, Trans.).
Northwestern University Press.
Husserl, E. (1991). On the phenomenology of the consciousness of internal time
(1893–1917) (J. B. Brough, Trans.). Springer.
Husserl, E. (1997). Thing and space: Lectures of 1907 (R. Rojcewicz, Trans.).
Kluwer.
Husserl, E. (2001a). Analyses concerning passive and active synthesis: Lectures
on transcendental logic (A. J. Steinbock, Trans.). Springer.
Husserl, E. (2001b). Logical investigations (Vol. 2; J. N. Findlay, Trans.).
Routledge.
Husserl, E. (2005). Phantasy, image consciousness, and memory (1898–1925)
(Vol. 11; J. B. Brough, Trans.). Springer Science & Business Media.
Ingarden, R. (1965a). Der Streit um die Existenz der Welt, Vol. 1: Formalontologie,
Pt. 1: Form und Wesen. Niemeyer.
Ingarden, R. (1965b). Der Streit um die Existenz der Welt, Vol. 2: Formalontologie,
Pt. 2: Welt und Bewußtsein. Niemeyer.
Kaplan, D. M., & Bechtel, W. (2011). Dynamical models: An alternative or comple-
ment to mechanistic explanations? Topics in Cognitive Science, 3(2), 438–444.
Kelso, J. A. S. (1995). Dynamic patterns: The self-organization of brain and
behavior. MIT Press.
Kockelmans, J. J. (1987). HusserI’s original view on phenomenological psychol-
ogy. In J. J. Kockelmans (Ed.), Phenomenological psychology: The Dutch
school (pp. 3–29). Martinus Nijhoff Publishers.
Larsen, R. R., & Hastings, J. (2018). From affective science to psychiatric disor-
der: Ontology as a semantic bridge. Frontiers in Psychiatry, 9. doi.org/10.3389/
fpsyt.2018.00487
Larsen, R. R., & Hastings, J. (2020). Mapping the patient’s experience: An
applied ontological framework for phenomenological psychopathology.
Phenomenology and Mind, 18, 200–219.
le van Quyen, M. (2003). Disentangling the dynamic core: A research program
for a neurodynamics at the large-scale. Biological Research, 36(1), 67–88.
Lutz, A. (2002). Toward a neurophenomenology as an account of generative pas-
sages: A first empirical case study. Phenomenology and the Cognitive Sciences,
1(2), 133–167.
Lutz, A., Lachaux, J. P., Martinerie, J., & Varela, F. J. (2002). Guiding the study
of brain dynamics by using first-person data: Synchrony patterns correlate
70 Integrating Phenomenology
with ongoing conscious states during a simple visual task. Proceedings of the
National Academy of Sciences, 99(3), 1586–1591.
Lutz, A., & Thompson, E. (2003). Neurophenomenology integrating subjective
experience and brain dynamics in the neuroscience of consciousness. Journal of
Consciousness Studies, 10(9–10), 31–52.
Marbach, E. (2010). Towards a formalism for expressing structures of conscious-
ness. In S. Gallagher & D. Schmicking (Eds.), Handbook of phenomenology
and cognitive science (pp. 57–81). Springer.
Mitchell, S. D. (2002). Integrative pluralism. Biology and Philosophy, 17(1),
55–70.
Moran, D. (2013). ‘Let’s look at it objectively’: Why phenomenology cannot be
naturalized. Royal Institute of Philosophy Supplements, 72, 89–115.
Nielsen, T. I. (1963). Volition: A new experimental approach. Scandinavian
Journal of Psychology, 4, 225–230.
Olivares, F. A., Vargas, E., Fuentes, C., Martínez-Pernía, D., & Canales-Johnson,
A. (2015). Neurophenomenology revisited: Second-person methods for the
study of human consciousness. Frontiers in Psychiatry, 6. doi.org/10.3389/
fpsyg.2015.00673
Oppenheim, P., & Putnam, H. (1958). Unity of science as a working hypothesis.
Minnesota Studies in the Philosophy of Science, 2, 3–36.
Parnas, J., Møller, P., Kircher, T., Thalbitzer, J., Jansson, L., Handest, P., & Zahavi,
D. (2005). EASE: Examination of anomalous self-experience. Psychopathology,
38(5), 236–258.
Petitmengin, C., Baulac, M., & Navarro, V. (2006). Seizure anticipation: Are neu-
rophenomenological approaches able to detect preictal symptoms? Epilepsy &
Behavior, 9(2), 298–306.
Petitmengin, C., Navarro, V., & le van Quyen, M. (2007). Anticipating seizure:
Pre-reflective experience at the center of neuro-phenomenology. Consciousness
and Cognition, 16(3), 746–764.
Petitot, J., Varela, F. J., Pachoud, B., & Roy, J.-M. (Eds.). (1999). Naturalizing
phenomenology: Issues in contemporary phenomenology and cognitive sci-
ence. Stanford University Press.
Płotka, W. (2020). A controversy over the existence of fictional objects: Husserl
and Ingarden on imagination and fiction. Journal of the British Society for
Phenomenology, 51(1), 33–54. doi:10.1080/00071773.2019.1629553
Pokropski, M. (2019). Phenomenology and mechanisms of conscious-
ness: Considering the theoretical integration of phenomenology with
the mechanistic framework. Theory & Psychology, 29(5), 601–619. doi.
org/10.1177/0959354319868769
Reynolds, J. (2017). Phenomenology, naturalism and science: A hybrid and hereti-
cal proposal. Routledge.
Roy, J.-M., Petitot J., Pachoud, B., & Varela F. J. (1999). Beyond the gap: An intro-
duction to naturalizing phenomenology. In J. Petitot, F. J. Varela, B. Pachoud,
& J.-M. Roy (Eds.), Naturalizing phenomenology: Issues in contemporary phe-
nomenology and cognitive science (pp. 1–80). Stanford University Press.
Quine, W. V. (1969). Epistemology naturalized. In Ontological relativity and
other essays (pp. 69–90). Columbia University Press.
Salmon, W. C. (1989). Four decades of scientific explanation. University of
Pittsburgh Press.
Naturalizing Phenomenology Reconsidered 71
Sellars, W. (2007). Philosophy and the scientific image of man. In K. Scharp
& R. B. Brandom (Eds.), In the space of reasons: Selected essays of Wilfrid
Sellars. Harvard University Press.
Smith, B., & Ceusters, W. (2010). Ontological realism: A methodology for coor-
dinated evolution of scientific ontologies. Applied Ontology, 5(3), 139–188.
Stanghellini, G., Broome, M., Fernandez, A. V., & Raballo, A. (Eds.). (2019). The
Oxford handbook of phenomenological psychopathology. Oxford University
Press.
Thompson, E., Noë, A., & Pessoa, L. (1999). Perceptual completion: A case study
in phenomenology and cognitive science. In J. Petitot, F. J. Varela, B. Pachoud,
& J.-M. Roy (Eds.), Naturalizing phenomenology: Issues in contemporary phe-
nomenology and cognitive science (pp. 161–195). Stanford University Press.
Thompson, E. (2007). Mind in life: Biology, phenomenology, and the sciences of
mind. Harvard University Press.
Wheeler, M. (2005). Reconstructing the cognitEive world: The next step. MIT
Press.
Wheeler, M. (2013). Science friction: Phenomenology, naturalism and cognitive
science. Royal Institute of Philosophy Supplement, 72, 135–167.
van Gelder, T. (1999). Wooden iron? Husserlian phenomenology meets cognitive
science. In J. Petitot, F. J. Varela, B. Pachoud, & J.-M. Roy (Eds.), Naturalizing
phenomenology: Issues in contemporary phenomenology and cognitive science
(pp. 245–265). Stanford University Press.
van Gelder, T., & Port, R. F. (1995). It’s about time: An overview of the dynamical
approach to cognition. In R. F. Port & T. van Gelder (Eds.), Mind as motion:
Explorations in the dynamics of cognition (pp. 1–43). MIT Press.
Varela, F. J. (1996). Neurophenomenology: A methodological remedy for the
hard problem. Journal of Consciousness Studies, 3(4), 330–349.
Varela, F. J. (1999). The specious present: A neurophenomenology of time con-
sciousness. In J. Petitot, F. J. Varela, B. Pachoud, & J.-M. Roy (Eds.), Naturalizing
phenomenology: Issues in contemporary phenomenology and cognitive science
(pp. 266–314). Stanford University Press.
Vermersch, P. (1994). L’entretien d’explicitation. Editions ESF.
Yoshimi, J. (2007). Mathematizing phenomenology. Phenomenology and the
Cognitive Sciences, 6(3), 271–291.
Zahavi, D. (2010). Naturalized phenomenology. In S. Gallagher & D. Schmicking
(Eds.), Handbook of phenomenology and cognitive science (pp. 3–19). Springer.
Zahavi, D. (2013). Naturalized phenomenology: A desideratum or a category
mistake? Royal Institute of Philosophy Supplements, 72, 23–42.
3 Models of Explanation in Cognitive
Science
3.1 Introduction
Science is thought to describe, categorize, predict, and explain phenomena
that occur in the world. Between them, however, explanation is conceived
as the core of scientific endeavor (Nagel, 1961). Despite the distinct role
of explanation in science, just what explanation is and what it means to
explain something remains notoriously ambiguous. In fact, one of the
great debates in 20th-century philosophy of science concerned the nature
of scientific explanation. In this chapter, I introduce a few important con-
cepts, distinctions, and types of explanation that are relevant to the topic
of this book. I address neither all of the available conceptions of explana-
tion nor all of the related issues, as doing so would take us well beyond
the scope of this work (for an overview, see, e.g., Salmon, 1989).
The chapter is structured as follows: I introduce the explanatory mod-
els applied in cognitive (neuro)science and the project of mechanistic
integration of cognitive science and consider how phenomenology fits
into this picture. First, I present the theoretical background of the sci-
entific explanation debate and discuss the key models of explanation,
including the deductive-nomological model, personal explanations, func-
tional explanations, dynamical explanations, and, last but not least, the
mechanistic model of explanation. The mechanistic model of explanation
is especially important because it provides an integrative framework for
cognitive science, according to which different research fields contribute
to a multilevel mechanistic explanation by providing constraints on the
space of mechanisms. At the end, I consider what kind of explanations,
if any, are put forward by phenomenology and whether phenomenology
could be integrated with other explanatory models and feature as part of
a multilevel explanation. I argue that phenomenology does not provide
a genuine model of explanation but rather that it provides a constitutive
understanding of mental phenomena in question. Despite this, I hypoth-
esize that phenomenological constitutive understanding can provide con-
straints on mechanistic explanations, namely, functional and dynamical
constraints.
DOI: 10.4324/9781003035367-5
Models of Explanation in Cognitive Science 73
3.2 Scientific Explanation—Background
It is generally agreed upon that an explanation should deliver us a better
understanding of the explanandum phenomenon, that is, the phenom-
enon we want to explain. More specifically, an explanation consists of a
phenomenon to be explained, assumptions that amount to the explanans,
and a dependence relation that shows how the former and the latter are
related (Bangu, 2017, pp. 105–106). A classic example of this depen-
dence relation is the logical relation endorsed in the deductive-nomo-
logical (D-N) model, although causal explanations replace it with causal
dependence. This is the general idea of explanation. But phenomena may
be of a different nature, complexity, and regularity. Depending on these
aspects, explanations can vary in form and scope. Physical phenomena
that occur regularly, such as tides and eclipses, are typically explained
according to the D-N model (Hempel & Oppenheim, 1948), that is, by
subsuming their occurrences under a scientific law. The scientific law can
either take the form of a universal law of nature or a lawlike statistical
generalization, as it is in the deductive-statistical (D-S) and inductive-
statistical (I-S) models (Hempel, 1965). A given explanation can have dif-
ferent levels of certainty depending on the kind of law used and its form
of inference. D-N explanations, in which the explanans includes univer-
sal laws of nature, are certain. When the lawlike statement is a statistical
generalization, the explanation is not certain but probable.
We expect a different kind of explanation for singular phenomena,
such as the extinction of the dinosaurs. These explanations rather indi-
cate a cause, a combination of causes, or a causal mechanism than some
universal law of nature (e.g., Scriven, 1975). There are also phenomena
that are not strictly speaking natural and depend on the sociocultural con-
text (e.g., Watkins, 1957). For example, an explanation of the Bolshevik
Revolution would refer to various historical, economical, and social
causes and factors. Biological phenomena, such as respiration, require
yet another type of explanation, namely, mechanical explanation, which
relies on describing the causal mechanism responsible for the target phe-
nomenon. Finally, there are mental phenomena concerning which there is
still debate over what model of explanation is optimal. On the one hand,
there are psychological phenomena to which different models of explana-
tion apply, including causal mechanistic, functional, and statistical mod-
els. For example, the cognitive function of memory can be explained in
a mechanistic manner by establishing levels of organization with respect
to the underlying neural mechanisms (e.g., for the mechanistic model of
spatial memory in rats, see Craver, 2007). But explanations of memory
can also abstract away from the details of its physical realization and
focus on its functional organization (e.g., Baddeley, 2007). Another set
of examples of mental phenomena that are addressed from different
perspectives are mental maladies, such as schizophrenia, explanations
74 Integrating Phenomenology
of which have been proposed on the grounds of belief-desire psychol-
ogy (Campbell, 2002), phenomenology (Sass, 2014), and neurobiology
(Andreasen, 1999). On the other hand, there is a class of mental phenom-
ena that relates to the issue of consciousness—to first-person experience.
It is argued that these subjective phenomena are the “hard problem” of
cognitive science, and there is no consensus with respect to which model
of explanation is the best for tackling them (Chalmers, 1996). Different
proposals for explaining consciousness vary from purely theoretical and
top-down (e.g., Rosenthal, 2005) to bottom-up and empirical (e.g., Prinz,
2012).
According to Salmon (1989, pp. 117–121), we can roughly distinguish
three conceptions of scientific explanation: modal, ontic, and epistemic.
Modal explanations tell us what the modal status of an explanandum
phenomenon is in terms of nomological necessity and possibility; for
example, they communicate that in a given situation, an event had to hap-
pen because it instantiated a law of nature. Ontic explanations assume
that events occur in a world that is full of regularities. These regulari-
ties can be of different sorts; usually, they are causal, deterministic, and
expressed in universal laws of nature. Such explanations rely on fitting
the explanandum event into these causal patterns. Finally, the epistemic
conception conceives explanations as arguments that consist of different
sorts of scientific representations, such as lawlike sentences, graphs, mod-
els, and so on. According to this conception, explanations are the product
of scientific work.
Addressing the epistemic conception, notice that the type of explana-
tion called for also depends on the questions we ask about the target
phenomenon, and so our epistemic interests (e.g., Grobler & Wiśniewski,
2005). It is thought that the question of “why” is what scientific explana-
tion ought to answer. But there are different senses in which we ask “why,”
and explanatory answers may differ in form. Why does water boil? Why
did the Bolshevik Revolution take place? Why do humans have lungs?
Why do some people gamble? These questions, despite all asking “why,”
require different explanatory strategies to formulate an answer. We can
also ask “what.” The “what” question is related to causal explanation
(e.g., Salmon, 1984; Scriven, 1975). What caused the extinction of the
dinosaurs? What was the cause of that car accident? What is the cause of
Creutzfeldt–Jakob disease? Answering these questions requires indicat-
ing a cause or a collection of causes relevant to the explanandum phe-
nomenon. Establishing causal relevance is key to finding the explanation.
Not all causal explanations require knowing the actual causal mecha-
nism responsible for the target phenomenon. Thus, this type of explana-
tion is sometimes called a “bare causal explanation” (Glennan, 2017, pp.
224–228) and is distinguished from causal mechanical explanations. A
causal-mechanical explanation answers the question of “how,” that is,
how a specific phenomenon was or is being produced. Answering such
Models of Explanation in Cognitive Science 75
a question usually requires discovering the necessary conditions for the
phenomenon to occur and describing the underlying mechanism along
with its internal organization. For example, an explanation of how blood
circulates in the human organism requires describing a multilevel mecha-
nism, including components such as the heart, veins, arteries, blood, and
so on, which generate and sustain the circulation. The mechanistic model
of explanation is the most common explanatory strategy in the life sci-
ences (for an overview, see Craver & Darden, 2013) and neuroscience
(Craver, 2007).
By accepting the scientific standpoint, we agree that all natural phe-
nomena have natural causes, but it does not follow that all explanations
have to be causal or causal-mechanical. To put it differently, there are sci-
entific explanations that are not causal (see Glennan, 2017, pp. 230–236).
Examples of this sort of explanation are some dynamical explanations
(e.g., Chemero, 2000; van Gelder, 1998) or what Elliott Sober (1983)
calls “equilibrium explanations.” Consider a physical system the state
of which changes in time, say, an iron ball rolling on a curved surface.
Obviously, this is a causal process, but “what the equilibrium explanation
does that makes it non-causal is that it describes the features of the space
that constrain the causal process, rather than describes the causal process
itself” (Glennan, 2017, p. 231). In general, an equilibrium or dynamical
explanation shows how a target system moves through its state space
and reaches its equilibrium together with what the key parameters are
for describing the process. In the following sections, I address selected
explanatory models in more detail.
This neat and formal model of explanation fits physical science best, but
even then there are counterexamples that it cannot handle and that lead
to paradoxes. The most famous counterexamples come from Salmon
(1989, pp. 46–50). Three of them are especially interesting in the context
of this chapter. First, the barometer case—assuming that there is a gen-
eral law stating that each time the reading of a barometer falls a storm is
coming, we can infer from noticing a drop in the reading of a barometer
that a storm is coming. Such reasoning fits perfectly well to the D-N
model. However, a drop in the reading of a barometer is not causally
relevant to storms and thus cannot explain them, although it can serve as
a good predictor. Both of these events have a common cause—a drop in
atmospheric pressure—and a full explanation of a storm should include
generalizations about such conditions.
Second, the flagpole example: a vertical flagpole stands on flat ground
and casts a shadow. Knowing the length of the flagpole, the position of
the sun, and the law of rectilinear propagation of light (premises included
in the explanans), we can deduce, in accordance with the D-N model,
Models of Explanation in Cognitive Science 77
the shadow’s length (explanandum). But we can reverse this reasoning
and from knowledge of the length of the shadow, the position of the
sun, and the previously mentioned law, we can infer the length of the
flagpole. If that is the case, then the length of the flagpole is explained
by the shadow it casts. That conclusion seems counterintuitive, as we are
inclined to think that the flagpole causes the shadow and not vice versa.
Explanations should not run in both directions in that sense, they should
not be symmetrical, precisely because that leads to absurd consequences.
The third example concerns eclipses. We can explain the past occurrence
of a lunar eclipse by deducing it from the current positions of the earth,
sun, and moon and the laws of celestial mechanics. And we can produce an
analogous inference deducing future occurrences of eclipses. Here we touch
on a similar problem we did with symmetry earlier. The D-N model does
not recognize temporal restrictions and so does not distinguish between ret-
rospective and predictive explanations. To put it differently, if we knew the
relevant laws of nature before the explanandum phenomenon happened,
then we could predict its occurrence, and the logical form of such a predic-
tive inference would be the same as the retrospective deduction. But expla-
nations are different from predictions. For example, consider the symptoms
of a disease. Symptoms can be used as predictors for the disease that causes
them. And the disease can serve as an argument in the prediction of symp-
toms. But only the disease explains the symptoms, not vice versa.
The D-N model seems least applicable to biology, in which, as it is
argued (e.g., Bechtel, 2008; Godfrey-Smith, 2014), there are no strict
laws. This does not mean, however, that there are no patterns in biology.
For instance, Mendel’s first law was considered an example of a biologi-
cal (genetical) regularity. The “law” captures patterns in gene inheritance,
which are not incidental. But there are many exceptions to it, such as
Down syndrome in humans, so the “law” is not universal. Hence, biology
is able to capture regularities occurring in organisms and their behav-
ior, yet its laws are essentially different from the universal laws of phys-
ics. According to Peter Godfrey-Smith (2014), “rather than a two-way
distinction between laws and accidental regularities, biological patterns
show different amounts of what can be called resilience or stability” (p.
14). Such biological patterns are better explained according to the mech-
anistic model rather than the D-N model.
It is also hard to find universal laws in psychology. There are a few
candidates for lawlike statements, such as Weber’s psychophysical law
(ΔI/I = k), where I stands for the intensity of a stimulus, ΔI is the mini-
mum detectable increment over the stimulus intensity, and k is a constant
value (Bechtel & Wright, 2009). But such nomic regularities, according
to William Bechtel and Cory D. Wright, should themselves be subject to
explanation. Similarly, Robert Cummins (2000) argues that regularities
in psychology are called effects and that they are something that scien-
tists want to explain rather than use as explanans. Generally speaking,
78 Integrating Phenomenology
psychological explanations typically refer to psychological states, such
as beliefs and desires, and thus fall under personal explanations or use
concepts of functions, functional modules, and cognitive capacities as in
the functional type of explanation. That being said, under the influence
of rapidly developing neuroscience, explanations of mental phenomena
have been more recently formulated in terms of neural mechanisms.
3.3.2 Personal Explanations
The distinction between the personal and sub-personal level of explana-
tion was proposed by Daniel Dennett (1969) in his book Content and
Consciousness. According to Dennett, explaining human behavior can
take two forms: one personal which relies on concepts of mental states,
such as pain, and one sub-personal which refers to mental states’ func-
tional organization and physical realization. Dennett’s objective was to
show that sub-personal explanations can be understood as a type of psy-
chological explanation. The idea of personal explanations was adapted
by John McDowell and the so-called Pittsburgh school (Drayson, 2014).
McDowell (1986) extended the scope of personal explanations to all
sorts of mental states with propositional content such as beliefs, desires,
and intentions that are thought to be reasons for action. Accordingly, we
can explain human behavior by indicating those reasons which under-
lie the studied behavior. In his later works, Dennett (1981) developed
this personal type of explanation and connected it with the idea of an
explanatory strategy called the “intentional stance.” Adopting the inten-
tional stance allows to us treat studied subjects as rational agents and
to explain and predict their behavior in terms of different sorts of inten-
tional states, including beliefs, in an analogous manner to the “functional
stance,” in which the target system is described in functional terms. At
this stage, Dennett’s explanatory strategies were instrumental; that is,
the intentional strategy was a useful tool for explaining and predicting
the behavior of rational agents, but it did not tell us anything about the
true nature of mental states themselves. Later (e.g., Dennett, 1991b), he
shifted his position toward realism and characterized mental states in
dispositional terms. According to Zoe Drayson (2014), the Pittsburgian
version of personal-level explanation differs from the Dennettian version
in several important respects. The most important one concerns the nor-
mative character of personal explanation related to the assumption that
humans are rational agents, which was stressed by McDowell and his
colleagues. Thus, according to the Pittsburgian picture, personal explana-
tions are a different type of explanation than sub-personal ones, which
rely on causal constraints and natural laws.
Personal explanations are often considered to be at the top of the hier-
archy of explanatory levels, above the functional and physical levels. It
is often argued that the personal level of explanation, which is taken to
Models of Explanation in Cognitive Science 79
be distinct and autonomous from other levels, does not require refer-
ence to explanatory facts from other levels. Autonomy is also expressed
therein, given that personal explanations can only appeal to psychologi-
cal states, which means that they ultimately run out; that is, the moment
we employ a nonpsychological state, we leave the personal level and
move to the sub-personal (Dennett, 1969). According to others (e.g.,
Bermúdez, 2000) personal explanations, which are also sometimes called
horizontal explanations, are compatible with the vertical type of explana-
tions common in cognitive science. A vertical explanation shows how a
certain mental state or cognitive capacity is realized by a mechanism or
a set of mechanisms on a sub-personal level. Proponents of the explana-
tory autonomy of personal explanations hold that vertical sub-personal
explanation reveals only enabling conditions; that is, it only shows how
the personal level is physically possible.
However, there are good arguments suggesting that sub-personal
facts can play a more significant role, which in consequence breaches
the autonomy thesis (Bermúdez, 2000; Colombo, 2013). The arguments
usually refer to abnormal or irrational behaviors, such as those in psy-
chopathological cases, which contradict the rationality constraint crucial
for personal explanations. Cases such as blindsight, unilateral neglect,
or addiction show that there are mental phenomena for which personal
explanations are incomplete or just impossible. Thus, it is argued that the
sub-personal can also be constitutive for explanations of mental phenom-
ena. It is important to emphasize that the concept of constitutive rele-
vancy is understood here in accordance with what McDowell (1994) calls
constitutive understanding or constitutive explanation, which is different
from mechanistic constitutive explanation (Craver, 2007). According to
McDowell, constitutive explanation relies on conceptual distinctions and
logical relations and provides us with an understanding of how and why
people behave in a certain way, in particular what rational motives and
norms make a specific mental phenomenon what it is. So a constitutive
explanation is a personal level explanation with a rationality constraint. In
contrast, an enabling explanation deals with causal relations and empiri-
cal facts, which makes certain mental phenomena possible but does not
explain their nature. McDowell separates these types of understanding,
but Colombo (2013) argues that at least “some facts about subpersonal
states and events are constitutively relevant to some personal-level phe-
nomena, and therefore can, and sometimes should, inform personal-level
explanations” (pp. 567–568). For example, addiction can be considered
a phenomenon the explanation of which requires understanding a neural
mechanism—for example, the specific activity of dopamine in the reward
circuit. Rewards as conceptualized on the personal level, such as acquir-
ing money, are different from a reward for the brain, which is a flood
of dopamine in the case of addiction. Understanding these sub-personal
facts about the functioning of an addicted brain not only indicates the
80 Integrating Phenomenology
enabling conditions but, more important, gives us an understanding of
the whole phenomenon of addiction and thus helps us to understand and
predict the behavior of an addicted person. It also helps us distinguish
genuine addiction, such as a gambling addiction, from other behaviors
that, on the personal level, might look similar but do not share this neural
characteristic. If that is the case, then the conceptual apparatus employed
in personal level explanations can be revised in the light of sub-personal
facts provided by empirical cognitive science.
To conclude, roughly speaking, personal level explanations are a dis-
tinct type of explanation of human behavior in terms of beliefs, desires,
and intentions (BDI). Their autonomy is, however, doubtful, as there are
cases of first-person phenomena that require information from the sub-
personal level in order to explain. In those cases, sub-personal facts are
not merely enabling or causal but also constitutive for our conceptualiza-
tion and understanding of the target phenomenon.
3.3.2.1 Heterophenomenology Explained
Much of the debate about personal-level explanations concerns the reli-
ability of first-person evidence. Although many psychological studies
were built on introspective reports, such as the classic studies on prob-
lem-solving conducted by Allen Newell and Herbert A. Simon (1972),
introspection itself is still a controversial topic. On the one hand, there
is no consensus on whether the source of introspective knowledge is sin-
gular or plural, that is, whether it consists of multiple processes (e.g.,
Komorowska-Mach, 2019; Schwitzgebel, 2012). Furthermore, it is often
argued that introspection is an unreliable source of evidence because sub-
jects are often unaware of their subjective responses and give meaning
to their experiences with antecedently held theories (Nisbett & Wilson,
1977). It is also argued that introspective reports are unverifiable and
have no truth value (Dennett, 1991a). On the other hand, it is difficult
to imagine a study of consciousness without any kind of first-person
insights. Thus, some researchers (e.g., Dennett, 1991a, 2003; Hurlburt
& Schwitzgebel, 2007; Jack & Roepstorff, 2002; Piccinini, 2003, 2010)
argue for a more developed and rigorous methodology for validating
and analyzing first-person data generated on the basis introspection. The
results of such methods could serve as evidence supporting first-person
explanations and could contribute to a broader explanatory framework
that integrates different research fields, including those using the third-
person methods of neuroscience.
Talking about first-person data raises the issue of phenomenology. For
Dennett, first-person phenomenology is not explanatory because men-
tal states, such as pain, are unanalyzable qualities (e.g., 1969) and sole
introspection, which Dennett (1991a) erroneously equates with Husserl’s
phenomenological method, is an unreliable source of evidence. I share
Models of Explanation in Cognitive Science 81
Dennett’s skepticism about qualia, but his reading of Husserlian phenom-
enology is wrong. As I argued in Chapter 2, phenomenological method
differs significantly from introspection and is not interested in studying
the ineffable qualities of particular experiences but in general structures
of experience (for other arguments against Dennett’s reading of phenom-
enology, see, e.g., Gallagher & Zahavi, 2012; Zahavi, 2007). According
to Dennett, the only reliable phenomenology is heterophenomenology
(1991a), which is a third-person phenomenology or, to be more precise,
an analysis of the studied subject’s verbal reports in terms of BDI. In
brief, the method consists of two steps. First, the experimenter records
the verbal utterances of the studied subject. These recordings are the
raw data that are then interpreted in terms of beliefs. The experimenter
adopts an “intentional stance” (Dennett, 1981) and ascribes beliefs to the
subject about their experiences. As Dennett (2003) writes, “these are the
primary interpreted data, the pretheoretical data, the quod erat explica-
tum (as organized into heterophenomenological worlds), for a science of
consciousness” (p. 21). Next, the experimenter “composes a catalogue of
what the subject believes to be true about his or her conscious experience”
(Dennett, 2003, p. 20, emphasis in the original). Importantly, the experi-
menter does not assume that the subject is right but only acknowledges
that it is what the subject believes. For example, in simple perceptual
tasks, when a subject says that they saw a light, a heterophenomenologist
cannot acknowledge that the subject actually saw a light; the hetero-
phenomenologist is only allowed to infer that the subject believes that
they saw the light. So heterophenomenological method is interested in
the subject’s beliefs about first-person experience, not in the experience
itself. The explanatory question that a heterophenomenologist can try to
answer is why a subject in an experimental situation has a specific set of
beliefs about his or her mental life.
Dennett (2005) claims that heterophenomenology is a widely accepted
approach to first-person evidence, and “it has been practiced, with vary-
ing degrees of punctiliousness about its presuppositions and prohibitions,
for a hundred years or so, in the various branches of experimental psy-
chology, psychophysics, neurophysiology, and today’s cognitive neurosci-
ence” (p. 36). Even if that is the case, it is so not without controversy.
The very idea that first-person evidence is private and incorrigible is
already doubtful. On the one hand, one can argue from phenomeno-
logical positions that our mental lives are to some extent observable and
public thanks to bodily expressions (e.g., J. Smith, 2010). On the other
hand, one may concede that first-person evidence is private without con-
ceding that it should be removed from science. It just means that we
need to develop a methodology for how to process first-person data and
extract information from it. Heterophenomenology sets aside our mental
lives, our conscious experience, and focuses on beliefs about one’s mental
states. Why should we call it “phenomenology” then? Phenomenology, as
82 Integrating Phenomenology
well as any kind of first-person methodology, studies experience or, to put
it more abstractly, mental states, not just beliefs.
Dennett argues that heterophenomenology suspends judgments about
the truth value of subjects’ beliefs in a manner similar to Husserl’s method
of bracketing (see Chapter 1). The difference is, however, that Husserl’s
method is performed by a phenomenologist on their own beliefs and
judgments. In Dennett’s proposal, the method of bracketing is performed
by a researcher and concerns the studied subject’s beliefs. Thus, hetero-
phenomenology relies on a naïve attitude that tacitly assumes that the
observer’s beliefs about the subject are true. Recognition of the observ-
er’s biases touches upon the issue of “primary interpreted data,” which
according to Dennett (2003) is “pretheoretical” (p. 21). It seems that
Dennett uses this term because the interpreted data precedes a theory
of consciousness. However, if the data is interpreted in terms of BDI, it
cannot be completely pretheoretical, since belief is a folk-theory concept.
Belief talk presupposes that thinking has a propositional character, and
that is clearly a theoretical statement. As Piccinini (2010) points out, a
heterophenomenologist who applies the concept of belief is committed
to at least two problematic assumptions: first, that first-person reports
are always mediated by beliefs, that is, they are not directly caused by
perceptual or emotional experiential states, and, second, that first-person
reports are adequate representations of the contents of beliefs. The for-
mer assumption seems implausible since it is possible that some cases
of first-person reports, especially some first-person behaviors, may fol-
low directly from an experience; for instance, a painful experience can
directly cause a verbal expression. The latter assumption breaks down in
the light of an experiential subject’s confabulations or cognitive distor-
tions. Obviously, we can try to control the process to rule out confabula-
tions and other errors in reports, but why not, as Piccinini argues, cut
out the middle level corresponding to beliefs and think about reports as
direct expressions of mental states. After all mental states are what we
want to investigate and explain.
Taking these arguments into account, Piccinini proposes an improved
version of heterophenomenology. The improvements are based on his ear-
lier proposal of a third-person methodology for first-person data in which
subjects are considered self-measuring instruments generating first-person
behaviors (Piccinini, 2003, 2010). The aim of the method is to validate
first-person data and extract information about the subject’s mental states.
This version of heterophenomenology features several important revisions.
First, he extends the notion of first-person data to all first-person behav-
iors, not just utterances. That is completely fine with Dennett (2003), who
himself suggests such possible extensions. However, in contrast to Dennett,
Piccinini undermines the mediating role of beliefs in producing first-person
behavior, which is caused directly by occurrent mental states. Furthermore,
he argues against agnosticism concerning the subject’s report. For him,
Models of Explanation in Cognitive Science 83
we should embrace our everyday attitude toward others’ expressions and
beliefs and evaluate the credibility of first-person data. If the studied subject
is nothing more than a self-measuring instrument, then they can be treated
similarly to a telescope, the output of which is validated to check whether it
shows what it is supposed to show. Evaluation of the subject’s first-person
report is based on commonsense folk-psychology, and one should express
the platitudes one refers to. This is a different strategy for how to deal with
the observer’s tacit assumptions and theory ladenness than the one applied
in Husserlian phenomenology and Dennett’s heterophenomenology, which
was to bracket such beliefs as much as possible. Piccinini argues for making
them public in order to control these aspects in an intersubjective fashion.
Self-measurement methodology also drops the incorrigibility claim:
3.3.3 Functional Explanations
In philosophy of mind and cognitive science, functionalism can be
approached not only as a metaphysical conception of mental states, which
argues mental states are functional states individuated by causal relations
to other mental states and inputs and outputs (e.g., Block, 1980), but also
as an explanatory strategy called functional analysis. Functional explana-
tions were initially recognized as a way of explaining biological as well
as social phenomena (e.g., Nagel, 1961; Salmon, 1989) in teleological
terms such as goal, purpose, and function. It is important to notice that
we may operate with two different notions of function: a systemic and an
etiological one (e.g., Godfrey-Smith, 2014, pp. 59–65). The systemic con-
ception of function understands it as the causal role of some component,
which it plays in the studied system in such a way that it contributes to
the capacity we want to explain. For example, door keys contribute to
opening doors, but under certain conditions, they can serve as a bottle
opener. So, in the systemic conception of function, the ascription of a
function to a system’s part may vary depending on one’s explanatory
interests. The etiological conception of function, which is more common
in biology, focuses on the question of why the component is in the target
system (why is it there?) and seeks answers in the effect that this part
produces. This notion of function relates to the proper function of the
system’s component in question. So, with etiological notion of function,
we can explain, for example, why cats have claws, referring to the effect
claws have on catching prey and climbing trees.
84 Integrating Phenomenology
Carl Hempel (1965) was one of first to describe the pattern definitive
of this functional explanatory strategy. In short, we take an explanandum
that is a relatively persistent feature of an item, say, an organ in an organ-
ism, and we show that in certain conditions the item’s feature obtains
and contributes to the system’s proper functioning. Hempel adopts the
etiological understanding of function. For example, we can explain why
there is a heart in an organism by appealing to the heart’s disposition
to pump blood in its vascular system, which, in turn, contributes to the
organism’s well-being. But we cannot explain the heart’s disposition of
emitting sounds that way, because its emitting sounds does not affect the
organism’s functioning. Hempel’s structure of functional explanations
has one important problem, namely, the problem of equivalent realiza-
tions. We cannot explain the presence of a heart in an organism by the
fact that it pumps blood because there are other possible mechanisms
that could realize the same pumping function. If that is the case, then the
functional strategy has no explanatory power and can only have heuris-
tic value. Attempting to overcome this issue relies on showing that the
analyzed item was purposefully placed in the system in order to realize
this disposition. In the case of natural systems, one can argue (e.g., Nagel,
1961) that, when we consider a specific organism, the target disposition
of an organ cannot be realized by anything else because the organ is a
result of evolution. Thus, that particular organ is evolutionarily optimal
for realizing the target disposition.
However, in the seminal paper “Functional Analysis,” Cummins (1975)
argues that the problem comes from assumptions underlying Hempel’s
functional model and confusing a teleological approach with a functional
one. According to Cummins, we cannot explain why a specific item is
part of a system functionally—what we can explain is the behavior of the
containing system. The explanandum is a system’s disposition (Cummins
calls it a capacity), which occurs with lawlike regularity (so-called dis-
positional regularity). The functional analytic strategy does not subsume
the disposition’s occurrences under a law but decomposes the target dis-
position into a number of sub-dispositions, the manifestations of which
amount to the occurrence of the analyzed disposition. Cummins adopts
the systemic notion of function because the ascription of a function to
an item depends on the contribution of the function to the explanandum
disposition.
Cummins argues for distinguishing teleological thinking from func-
tional thinking. The reason behind identifying the functional approach
with teleological reasoning relies on explicit or implicit reference to the
Aristotelian idea of the final cause, that is, the idea that some future state
can causally affect the present. For example, one could argue that the
present behavior of an animal, for instance hunting for prey, is deter-
mined by a future state (eating the prey). But functional explanations of
such cases do not have to lead to thinking in terms of the future causing
Models of Explanation in Cognitive Science 85
present behavior. Apart from Cummins’s functional analysis, the theory
of “consequence-etiology” (Wright, 1976) shows that we can separate
the teleological approach from the functional one and explain the occur-
rence of an animal’s behavior not through it pursuing some goal in the
future but through that behavior having been causally effective in achiev-
ing goals in the past. So, for example, hunting for prey is goal-directed
behavior that is caused not by the future goal (eating prey) but by past
hunting behavior that was effective in keeping the hunting organism well
nourished. Accordingly, the real cause of the explanandum behavior is
its efficacy in the past, which is a feedback process spanning a long time
scale. Importantly, the past in such explanations can mean, in this case,
the evolutionary history of the organism’s species.
In contemporary philosophy of science and philosophy of mind, there
are several types of functional analysis. Piccinini and Craver (2011)
distinguish three types of such analysis: functional analysis by internal
states, so-called boxology, and task analysis. Computational function-
alism’s preferred form of explanation is functional analysis by internal
states, since it explains the target capacity using a set of internal states
and their relations to each other and to inputs and outputs. The key idea
behind this version of functionalism is that mental states can be under-
stood as Turing machine states (Putnam, 1960) and the mind as a sort of
computer program. Internal states can be described in representational
terms and their manipulation as computations. The form of analysis
called boxology, related to graphical representations of cognitive mod-
ules, can be found in 20th-century cognitive psychology and was utilized
in Jerry Fodor’s conception of the mind’s modularity (Fodor, 1983). The
conception encapsulates cognitive functions and subfunctions in special-
ized modules that have inputs and outputs; they are domain-specific (they
process only information from a specific domain), informationally encap-
sulated (external processes do not have access to the information being
processed in a module), and mandatory in their operations.
Finally, there is the already introduced functional analysis as task anal-
ysis proposed by Cummins (1975, 2000). According to Cummins (2000),
“functional analysis consists in analyzing a disposition into a number
of less problematic dispositions such that programmed manifestation of
these analyzing dispositions amounts to a manifestation of the analyzed
disposition” (p. 125). This method is quite universal and can be applied
to artificial as well as natural systems, including cognitive systems. An
example of an artificial system is a production line, which is composed of
a number of simple tasks. An example of a natural system is a situation in
which cats hunt small birds, where the disposition to hunt should include
sub-dispositions such as recognizing birdlike objects, stealing, running,
claw-catching, and so on. Each of these sub-dispositions can be, in turn,
divided into yet simpler dispositions. Research on working memory is an
example of such functional decomposition applied to cognitive functions.
86 Integrating Phenomenology
For instance, in the model proposed by Alan Baddeley (2007), working
memory is decomposed into four subfunctions: the central executive, the
visuospatial sketchpad, the phonological loop, and the episodic buffer.
This type of functional analysis can be roughly divided into three steps.
First, we identify the explanandum phenomenon, which in the case of
cognitive psychology is one of the target system’s cognitive capacities.
Then we define the capacity in dispositional terms; that is, the system’s
cognitive capacity φ is a complex dispositional property D, which means
that the system exhibits a lawlike regularity in its behavior. According
to Cummins, the best way to explain this kind of dispositional prop-
erty is decomposition of the disposition D into a set of sub-dispositions
d1, d2, …, dn. The final step is representing the system’s target capacity
decomposed in functional analysis in the form of a functional design. The
functional design represents the system’s functional architecture, which,
according to Cummins, explains the target phenomenon.
Cummins not only shows how the functional explanatory strategy
should proceed but also argues that the nature of psychological explana-
tion is functional. His line of thought is as follows: there are no general
laws of nature in psychology, thus those explanations cannot take the
form of subsumption under a law; regularities discovered by psychology
are not explanatory laws, but effects and capacities, which themselves
require explanation; an explanation of capacity consists in showing the
underlying functional structure. As Cummins concludes, “capacities
and their associated incidental effects are to be explained by appeal to
a combination of functional analysis and realization, and the currently
influential explanatory frameworks in psychology are all frameworks
for generating this sort of explanation” (Cummins, 2000). The frame-
works mentioned by Cummins include explanatory frameworks such as
BDI, computationalism, connectionism, the evolutionary approach, and
neuroscience.
Martin Roth and Cummins (2017) argue that functional analysis is a
distinct and autonomous explanatory strategy that delivers a functional
design of the target system’s capacity. Such a design does not entail struc-
tural decomposition; that is, it does not say anything about how specific
functions might map onto the components of a possible physical mecha-
nism. The autonomy of psychological explanations is defended in a simi-
lar vein (e.g., Barrett, 2014; Weiskopf, 2011). On the contrary, Piccinini
and Craver (2011) argue that functional explanations are neither autono-
mous nor distinct because they lack explanatory power. They can be per-
ceived as incomplete mechanistic explanations. I will return to the issue
of autonomy at the end of this chapter.
Cummins’s approach faces several important problems. One issue con-
cerns his conception of function relative to the arbitrariness of functional
analysis. In Cummins-style functional analysis, the way one decomposes
the target capacity, how many functions and subfunctions one includes in
Models of Explanation in Cognitive Science 87
the description, how many levels of functional organization are included—
they all depend on the interests and knowledge of the researcher. In other
words, there is no ultimate functional decomposition of a system but a
number of decompositions depending on what we want to explain and
what we already know about the target system and the explanandum phe-
nomenon. The problem results from Cummins’s (1975) systemic under-
standing of function: something has a function in the target system if it
contributes to the system’s performance (capacity) that we want to explain.
As Ruth Millikan (2002) notes, in Cummins’s functional explanations, we
do not ask why a system has a certain function but how it functions. This
approach uses a notion of function that is detached from the system’s (evo-
lutionary) history. Thus, Millikan proposes a different conception of func-
tion ascription to living systems called proper function, which is related to
the system’s evolutionary history. A proper function is a function of an ani-
mal trait, an organ, or animal behavior that is a result of natural selection.
The other problem with applying Cummins-style task analysis in
explanations of our mental lives is that dispositions occur under some
conditions. For example, salt has a disposition to dissolve in an envi-
ronment containing water, or a glass window has a disposition to
break when struck by something hard. To put it differently, a disposi-
tion always obtains when certain conditions are satisfied. This allows
one to see them as lawlike regularities dependent on causal conditions.
However, this is not always the case for mental dispositions, which often
are not exhibited with such regularity and for which conditions could be
difficult to define. Consider, for example, someone’s disposition to feel
touched when watching dramas. That behavior depends in large part on
the mood, place, companions, and, of course, the movie. But even if one
felt emotionally touched, their reaction can take various forms—one can
hide one’s emotions, someone else may cry, and so on. It is important,
therefore, to be aware that thinking in dispositional terms, and thus, in
accordance with a functional approach, has limitations in explaining our
mental lives. However, as I argue in Chapter 4, phenomenology shares
some important similarities with functional explanations.
3.3.4 Dynamical Explanations
Another major form of explanation corresponds to dynamical systems
theory (DST). It consists of a mathematical theory that enables modeling
and analyzing the behavior of dynamical systems. A dynamical system is,
generally speaking, any state-determined system the behavior of which
evolves in time. They may vary in nature and composition, from physical
systems through biological ones, up to the collective behavior of popula-
tions. Thus, DST finds application across the sciences, including physics,
biology, ecology, sociology, and recently cognitive science. DST calculus
is formed by differential equations or discrete equations in the case of
88 Integrating Phenomenology
discrete-time systems. These equations include variables that represent key
features of the target system’s change over time and parameters that stand
for other conditions relevant to the behavior. A set of these equations con-
stitute a formal dynamical model of the target system. The target system’s
behavior can be represented as a trajectory in n-dimensional state space,
where n depends on the number of variables describing the system’s states.
DST offers methods for both quantitative and qualitative analysis of
the behavior of a dynamical system. Quantitative methods require formal
models, that is, equations that allow for simulating the target system’s
behavior. Qualitative analysis is a more general approach that focuses on
the overall behavior of the system and reveals the features of its long-term
dynamics, such as its stable and unstable states or attractors (regions in
the state space towards which the system evolves). The long-run behavior
of the system may be visualized with the use of diagrams, such as phase
portraits, bifurcation diagrams, or parameter charts (see Beer, 2000).
Besides quantitative analysis (mathematical model) and qualitative anal-
ysis (analysis of behavior on the basis of phase portraits, etc.), van Gelder
and Port (1995) discuss a more general and conceptual approach called
“dynamic description,” which relies on the use of DST categories, such as
stable state, attractor, and bifurcation, to describe the system’s behavior.
It is important to remember that, in this case, dynamical categories are
used in a highly abstract fashion unless a proper dynamical analysis is
produced. A dynamic description may be the first step in building a com-
plete dynamical explanation.
Although the dynamic approach in science, especially in physics, has a
long history (see, e.g., Abraham & Shaw, 1992), in cognitive science it is
relatively new. It is argued that dynamical systems theory can be used to
explain cognitive systems (e.g., Chemero, 2000; Kelso, 1995; van Gelder
& Port, 1995). In his seminal paper, “What Might Cognition Be, If Not
Computation?” van Gelder (1995) provocatively proposes to think about
cognition in terms of dynamical systems rather than computational ones.
Such an approach assumes that cognition is an ongoing process that
evolves in time and spans the brain, body, and environment. Van Gelder
rejects the computer metaphor and offers a new one, namely, the Watt’s
governor. It is a simple but elegant device designed in the 18th century
by James Watt for steam engines in order to automatically control the
throttle and maintain a constant flywheel speed (see Figure 3.1) despite
changing steam pressure. It consists of a spindle with two arms attached
to the flywheel. Metal balls are attached at the ends of these arms. Faster
spindle rotation moves the arms upward and slower rotation downward.
Importantly, the arms are linked to the engine’s throttle valve; when the
speed of the flywheel increases, the arms go up, which closes the valve
reducing the flow of steam and speed; when the wheel’s speed decreases,
the arms go down and the valve opens enabling a greater flow of steam.
The simple yet brilliant design allows for controlling the speed of the
Models of Explanation in Cognitive Science 89
flywheel, which was crucial for using steam engines, for example, in the
cotton industry.
According to van Gelder, the behavior of the device can be described
with a differential equation:
d 2 g d
n cos sin sin r
2
dt 2 l dt
si si Ii x, y; i 1, , 7
7 12
i si si
j 1
w ji g s j w s i 8,,12
j 8
ji j j
12
y 3
Figure 3.2 M
odel of the perceptual categorization agent. (I) Simulated agent in
the environment. (II) The agent's neural network brain including seven
sensory neurons, five interneurons, and two motor neurons. Adapted
from “The Dynamics of Active Categorical Perception in an Evolved
Model Agent,” by R. D. Beer, 2003, Adaptive Behavior, 11(4), p. 213.
Copyright 2003 International Society for Adaptive Behavior.
first, the agent’s “neural network brain” and body (14 equations) and,
second, the environment (Equation 15 and 16). The coupling is math-
ematically expressed by putting variables S13 and S14 (responsible for two
motor neurons) into the second to last equation and variables x and y
(the object’s horizontal and vertical position relative to the agent) into
Equations 1 through 7.
Beer’s model shows two important things. First, dynamical systems can
be decomposed. As Beer (2003) writes,
Second, the dynamical approach has the potential to model complex cog-
nitive capacities but, as I argue, that is so because this model goes beyond
its own limitations and includes other perspectives to some extent, namely
the mechanistic one.
There is ongoing debate concerning the nature of dynamic descriptions
and their explanatory power. Some argue that dynamical explanations
are a subtype of D-N explanations (e.g., Raja et al., 2017; Walmsley,
2008); that is, they seek to formulate lawlike generalizations in the lan-
guage of differential calculus which explains the target system’s particu-
lar behavior. For example, the HKB model (Haken et al., 1985) consists
of a coordination law expressed in a single equation. The model allows
one to predict the system’s behavior and supports counterfactuals; that
is, it allows one to say what would happen with the system if certain
parameters or variables changed. Moreover, developments of the HKB
model show that it can be generalized to other oscillatory activities
(Kelso, 1995; Oullier et al., 2008; Port, 2003); thus, it can be conceived
of as a general law for coupled oscillatory motion. There are, however, at
least two serious problems with such a nomological reading of dynamical
models. First, as Joel Walmsley (2008) argues, if dynamical explanations
are D-N explanations, then they are deductive; that is, the explanandum
phenomenon or prediction is deduced from the lawlike formula and data
about the current state of the system. But if that is the case, then the
explanatory model becomes reductive—all the facts about the studied
phenomenon can be deduced. Second, the dynamical approach and D-N
explanations are committed to so-called predictivism; that is, they share
the argument that prediction equals explanation. The flagpole example,
discussed in the context of the D-N model of explanation, shows that
that is not necessarily the case—that we can have relatively good predic-
tive power without having a genuine explanation.
There is yet another issue with dynamical explanations in cognitive sci-
ence. The dynamical approach is not interested in the system’s structural
composition but in its overall behavior and its evolution in time. Thus,
for some researchers, dynamical analysis merely delivers a description
of the explanandum phenomenon, the explanatory power of which is
weak, even if it proves to be a reliable basis for accurate predictions.
To establish explanatory power, the dynamical model has to be supple-
mented with a model of the underlying mechanism, describing its parts,
internal organization, and causal interactions (e.g., Bechtel, 1998; Kaplan
& Craver, 2011). For example, according to Craver (2006), the Hodgkin
94 Integrating Phenomenology
and Huxley (HH) model of action potential (Hodgkin & Huxley, 1939)
is not explanatory but merely describes the behavior of neurons. The
HH model can, however, still be heuristically useful and help in find-
ing details about actual mechanisms responsible for neuron behavior, for
example, ion channels. Generally speaking, dynamical explanations can
be an important source of information guiding the study of the target
system and the building of a model of its parts and internal organization.
The constraints delivered by the dynamical approach concern the global
behavior of a target system, but, importantly, they can also character-
ize the behavior of mechanisms’ components (Zednik, 2011). A good
example of such a dynamical-mechanistic model is Beer’s model noted
earlier. In it, equations characterize the dynamics of the system’s compo-
nents, which are interconnected neurons responsible for different func-
tions organized in three layers. Separate equations describe the changing
position of the object. The overall dynamics generated by the whole set
of equations is therefore a derivative of component behavior. Dynamical
and mechanistic explanations can, therefore, be treated as complemen-
tary rather than exclusive explanatory strategies.
3.3.5 Mechanistic Explanations
Mechanistic explanations are in many ways as influential as those pro-
duced by the D-N model. Looking into the not-so-distant past, we may
note Rene Descartes or Julien Offray de La Mettrie as precursors of
mechanism (for a historical overview, see, e.g., Miłkowski, 2017; Roux,
2017). Descartes initiated a line of thinking about living organisms in
terms of mechanisms. A more recent predecessor of the contemporary
(neo)mechanistic model of explanation is the causal-mechanical model
of explanation developed by Salmon (1989), although his main area of
interest was physical science. According to Salmon, to explain a phe-
nomenon is not to show that it could be inferred from a law of nature or
lawlike generalizations, but to show how the phenomenon is produced
by causal processes. The (neo)mechanistic model of explanation1 (e.g.,
Bechtel, 2008; Craver, 2007; Glennan, 2017) that I discuss in this section
is a development of causal-mechanical explanation, which may also be
referred to as a constitutive or componential explanation. This type of
mechanistic explanation focuses on describing the mechanism, that is, the
components, activities, and organization that underlie or constitute the
phenomenon to be explained.
Whereas D-N explanations answer the question of “why,” mechanistic
ones are dedicated to answering the question of “how”; for example,
how is the target system’s behavior being produced, or how is a specific
function being realized in the target system? The mechanistic approach
emphasizes the particularity of the world rather than its generality, unlike
the D-N model. It focuses on real entities, particular mechanisms diverse
Models of Explanation in Cognitive Science 95
in form and organization, situated in an external context. Mechanistic
explanations are, contrary to D-N explanations, asymmetrical; that is,
the mechanism explains the target phenomenon and not vice versa. This
does not mean, however, that there is only one direction of determina-
tion, namely bottom-up determination from the parts of the mechanism
to the produced phenomenon. As I will show, there are cases of top-down
relations in which the mechanism’s behavior at a higher level can affect
its parts at a lower level.
In the following sections, I stress a few important features of mechanis-
tic explanations, namely that they are causal, contextual, multilevel, and
nonreductionistic in the strong sense. Mechanistic explanations are causal
because they focus on the causal mechanisms underlying an explanandum
phenomenon. They are contextual, as they situate a given mechanism in
the context of their environment, emphasizing that the context is impor-
tant for understanding the mechanism’s functioning. They are multilevel
because they address different levels of a mechanism’s organization and
their interlevel relations. Addressing different levels often requires inte-
gration of multiple research fields. Finally, mechanistic explanations are,
as I argue, nonreductionist in a strong sense because they acknowledge
that, in some cases, the behavior of the whole surpasses the behavior
of its components and that to describe a mechanism on multiple levels
we need to integrate different research fields without reducing one to
another.
3.3.5.1 Mechanisms
There is a common misunderstanding with respect to mechanisms that
needs to be addressed at the outset: a mechanism is not a simple aggre-
gate of parts. An aggregate is a collection of parts in which the order of
the parts is irrelevant to the whole. Think of a pile of stones as an exam-
ple. In an aggregate, we can interchange parts with one another without
any effect on the behavior of the whole. What matters for mechanisms is
just that: the organization of their parts and their activities. Consider the
following definition:
Figure 3.3 M
echanistic explanation. Adapted from Explaining the Brain:
Mechanisms and the Mosaic Unity of Neuroscience (p. 7), by C. F.
Craver, 2007, Clarendon Press. Copyright 2007 by Carl F. Craver.
3.3.5.2 Models of Mechanisms
Following Bechtel (2008), who states that “mechanisms do not explain
themselves” (p. 18), I adopt the epistemic view of mechanistic explana-
tion.3 Accordingly, a mechanistic explanation is a product of scientists’
work and consists of scientific representations of an actual mechanism
and its behavior. These representations may take the form of a simplified
sketch or a complete model, a detailed scheme describing the mecha-
nism’s parts, activities, and their organization. The explanatory process
goes from a sketch, which still includes black boxes and empty spaces,
toward a complete scheme, which covers all the details relevant to the
explanation.
When talking about mechanistic models, it is important to mention
the distinction between how-possibly and how-actually models (Craver,
2007, pp. 112–114; Glennan, 2017, pp. 68–73). How-possibly mod-
els are built on the basis of hypotheses about mechanisms that might
produce the explanandum phenomenon. Such a model does not refer
to an actual mechanism but expresses one of the possible realizations
of the phenomenon. On the contrary, a how-actually model describes
an actual mechanism responsible for the target phenomenon. Between
these two, we can situate a great many how-plausibly or how-roughly
Models of Explanation in Cognitive Science 103
models, which are more or less accurate representations of the underly-
ing mechanism. It is important to remember that there can be multiple
realizations of a function; thus, there can be several how-actually models
of mechanisms responsible for the same function but, for instance, in dif-
ferent organisms. Note also that on the epistemic account of mechanistic
models even a complete model does not include, and should not have to
include, all of the mechanism’s parts and their activities. It consists only
of those components that are relevant to the explanandum phenomenon.
To put it differently, models are always a matter of abstraction and ide-
alization. The former procedure omits some aspects or parts of the target
system that are irrelevant to the explanation. The latter adds features or
changes parameters in order to provide a better understanding of the
target mechanism.
An equally important dimension of mechanistic explanation is the
notion of a phenomenal model:4
A phenomenal model not only defines the explanandum but also indi-
cates its key properties, which, in turn, can serve as heuristics in search-
ing for mechanisms. Indeed, a complete mechanistic explanation consists
of two complementary parts: a detailed model of the mechanism and a
phenomenal model of the explanandum.
It is argued that, in contrast to phenomenal models, mechanistic mod-
els are explanatory because they support interventionist counterfactu-
als; that is, they allow one to answer “what-if–things-had-been-different”
questions (Woodward, 2017). In short, a mechanistic explanation tells us
what would change in the target phenomenon if we modified the behavior
of some part of the mechanism, changed the environmental conditions,
or level of stimulation. The explanatory power of mechanistic models
is also tied to our ability to control and manipulate the system (Craver,
2006). If direct manipulation is impossible, then simulation plays an
analogous role. Some mechanists (e.g., Craver) argue that phenomenal
models do not satisfy these conditions and thus that they are not explan-
atory. Accordingly, one can argue that the previously mentioned HKB
dynamical model of oscillatory movement (Haken et al., 1985) is a purely
descriptive phenomenal model. Although it can serve as a predictive tool,
it is not explanatory because it does not describe the system’s compo-
nents to the degree that would allow one to control its behavior (it isn’t
necessarily the case that it fails to support counterfactuals; as I argued,
104 Integrating Phenomenology
many dynamical models do support them). Others, however, argue that
the distinction between the descriptive and explanatory is rather concep-
tual and that these aspects of explanations should not be perceived as
opposites.
An example from neuroscience of the previously mentioned HH
model of action potential (Hodgkin & Huxley, 1939) shows how models
which are initially thought to be merely phenomenal are important in
the explanatory process. The original HH model is a dynamical model
and consists of equations describing the dynamics of neuronal behavior,
which, as it was hypothesized, was produced by specific components and
activities of neurons, such as the flow of ions across cell membranes.
According to Glennan (2017), the HH model cannot be categorized as
purely phenomenal or mechanistic but rather as a how-roughly model of
a mechanism that consists of two dimensions: a description of behavior
and actual or possible components producing the behavior. This early
model informed further research and the discovery of mechanistic com-
ponents such as ion channels. The discovery, in turn, supplemented the
model with further relevant details.
Allow me to also note that modeling neuronal behavior points to the
importance of thinking about neurons as active mechanisms rather than
something passive and simply responsive. Neurons do not remain silent
when not stimulated but exhibit basal activity and firing rate, which can
be modulated up or down. Thanks to positive feedback, the subsequent
behavior of a neuron depends on its previous behavior; for example, it
fires with a fixed rate until it receives inhibitory stimuli. Indeed, individ-
ual neurons as well as their assemblies can be modeled as coupled oscilla-
tors (Bechtel, 2008, pp. 229–233). For that purpose, again, the dynamical
approach is of key importance.
3.3.5.3 Levels of Mechanisms
One of the key features of mechanistic explanations is that they are mul-
tilevel. As I mentioned earlier, one of the results of applying heuristic
strategies such as decomposition to the target mechanism is division into
components and operations that can be grouped at different levels of
organization. For example, a mechanistic explanation of the phenom-
enon of spatial memory in mice (their learning and navigating in an envi-
ronment) consists of four levels: mouse behavior (navigating in a maze), a
neural structure encoding a spatial map in the hippocampus, a long-term
potentiation mechanism in neurons, and neurotransmitters along with
their receptors (activation of the NMDA receptor; Craver, 2007, p. 166).
The notion of level is notoriously ambiguous (for an overview of
different conceptualizations of it in science, see Craver, 2007, ch. 5).
First, we can think of levels of nature, that is, strata and distinctions
that actually exist in the world, for instance, as biological hierarchies
Models of Explanation in Cognitive Science 105
captured in taxonomies or as related to science as such. For example,
Paul Oppenheim and Hilary Putnam (1958) introduced six levels of
nature: elementary particles, atoms, molecules, cells, organisms, and
societies. Each level corresponds to a scientific discipline with its own
theory, laws, and explanations. These levels are, according to Oppenheim
and Putnam, not independent; that is, higher levels can be nomologically
reduced, via bridge laws, to lower levels. This picture, however, does not
fit mechanistic explanations. Consider the example of spatial memory
presented earlier. The levels of explanation do not constitute a nomo-
logical theory interrelated by bridge laws; rather, these levels represent
different research fields that are related because they refer to the same
explanandum phenomenon. Each of these fields delivers constraints and
thus contributes to the complete explanation. So how should we think
about levels in mechanisms?
The are several ways we can distinguish levels in mechanisms. First, we
can distinguish different spatial levels: mechanisms and parts of mecha-
nisms come in different sizes. Usually, bigger mechanisms are composed
of smaller parts, which, in turn, can have yet smaller parts. Certainly,
spatial levels play an important role in science. In neuroscience we can
talk about larger brain regions and areas, smaller cell assemblies, and
neural columns, ultimately reaching the level of individual neurons and
even smaller parts, such as neurotransmitters and ions. In mechanisms
entities of different sizes can causally interact. Furthermore, mechanisms
not only have parts of different sizes, but they also work in different
time scales. Temporal levels are critical for studying the brain, behavior,
and our subjective experience. Events in the brain usually happen on the
scale of milliseconds, whereas perceptual apprehensions and actions typi-
cally happen on the scale of seconds. Finally, there are causal levels. For
example, we can divide the processing of visual information into steps
and localize specific operations in cortical regions such as V1, V2, V3,
and so on. Accordingly, each region is dedicated to processing specific
information and participates in a causal chain of processes producing
visual experience. All these types of levels can contribute to the study
of the internal organization of the target mechanism and the relations
between its parts, but none of them gives us an answer to the more gen-
eral question concerning the nature of levels of mechanisms.
Craver (2007, pp. 188–195) proposes to think about levels of mecha-
nisms in terms of composition. As he argues, “levels of mechanisms are
levels of composition, but the composition relation is not, at base, spa-
tial or material. In levels of mechanisms, the relata are behaving mecha-
nisms at higher levels and their components at lower levels” (Carver,
2007, pp. 188–189). To put it differently, the higher level is the behavior
of the mechanism constituted or produced by the lower level, on which
we find organized parts constituting the mechanism (see Figure 3.4). In
an analogous manner, we can decompose this lower level, taking a part
106 Integrating Phenomenology
Figure 3.4 L
evels of mechanisms. Adapted from Explaining the Brain: Mechanisms
and the Mosaic Unity of Neuroscience (p. 194), by C. F. Craver, 2007,
Clarendon Press. Copyright 2007 by Carl F. Craver.
from that level and showing how its behavior is generated by parts and
operations at yet a lower level. A consequence of this understanding of
levels is that mechanistic explanations span levels and must have at least
two: a higher level corresponding to the phenomenon and a lower level
corresponding to the underlying mechanism.
Mechanists agree that relations between parts at a given level, intra-
level relations, are causal. How should we understand interlevel rela-
tions? First, although mechanisms and their parts have sizes, the levels
of mechanisms are not spatial. Thus, the relation of containment is ruled
out. The levels of mechanisms are levels of composition, which means
that the parts of lower levels are the components of higher levels. But can
we say that there is any causal interaction between these levels? Craver
(2007, p. 195) rejects the notion of interlevel causation, as it would lead
to the “strained” idea that the behavior of the whole mechanism can
interact with its parts. In a similar manner, Bechtel (2008) writes that
“certainly there are relations between levels within a mechanism. This is
the reason researchers investigating mechanisms need to bridge between
levels. But there are good reasons not to characterize the relation between
a mechanism and its parts causally” (p. 153). For instance, that would be
problematic in the case of a mechanism involving the temporal succes-
sion of causes and effects, as it is thought that causes precede their effects.
But that is not necessarily the case in mechanisms generating behavior;
that is, the overall behavior is collateral to the parts’ activities at the
Models of Explanation in Cognitive Science 107
lower level. Interlevel relations are constitutive; to be precise, they are a
hybrid of constitution and causation which Bechtel (2008) calls “mecha-
nistically mediated effects” (p. 153). One might ask whether “mechanisti-
cally mediated effect” is not just another name for emergence; in many
ways, it is. However, Bechtel is careful to avoid using “emergence,” as
he is aware that for some philosophers it conveys a mysterious relation.
But there is nothing mysterious about the emergent behavior of mecha-
nisms unless emergence is understood as referring to the behavior of the
whole that goes beyond the activity of its parts. In accordance with such a
notion of emergence, the higher level phenomena have a certain degree of
independence: they require a different mode of investigation than study-
ing entities and processes at lower levels (Bechtel, 2008). It does not fol-
low, however, that such emergent behavior cannot be explained in terms
of mechanistic composition. More recently, Bechtel (2019) has used the
notion of constraints to characterize interlevel relations. Accordingly, the
higher levels of a mechanism’s organization constrain the behavior of its
components on lower levels. To put it differently, higher level behavior
reduces the degrees of freedom of lower level components, consequently
producing the target phenomenon. Bechtel calls the higher level control
mechanism, which is mainly engaged in alteration of lower level compo-
nents’ activities, the production mechanism. This is yet another proposal
for how to address interlevel relations without causality.
Notice that the relations between fields have nothing in common with the
derivational relation between theories endorsed by proponents of theo-
retical reduction. An interfield theory emerges not from the deduction of
one theory from another but from complementing and constraining rela-
tions between research fields. So, for instance, one field may supplement
a description of an entity postulated in another field with structural or
physical details. Investigation of the causal nature of an entity in one field
may constrain research conducted in another field.
The integrative story of different research fields investigating the
same complex and multifaceted phenomena fits very well with cogni-
tive (neuro)science. Cognitive phenomena are complex and have mul-
tiple aspects requiring study from different perspectives using various
methods and theoretical vocabularies. Thus, any attempt to offer a com-
plete and reductive explanation, that is, an explanation in terms of one
fundamental scientific discipline such as physics, seems like a dead-end.
114 Integrating Phenomenology
Rather, the integrative strategy seems to be the best option at the current
stage of cognitive science’s development. It was argued recently that it
is the mechanistic model of multilevel explanation that can deliver the
framework required to integrate such a multidisciplinary endeavor (e.g.,
Craver, 2007; Miłkowski, 2016). As Miłkowski and Hohol (2020) argue,
“the mechanistic account of integration is compatible with nonextreme
versions of explanatory pluralism”. The mechanistic model of integration
requires, however, that various methodologies and theories are not fully
autonomous and that they can be mutually constrained.
3.4.3 Autonomy
The autonomy of integrated fields of research is a matter of ongoing
debate. A strong notion of autonomy assumes that scientific disciplines
are autonomous from each other if there are no direct constraints
between them. As I have shown above, this kind of autonomy is some-
times ascribed to first-person and psychological-functional explanations.
For instance, Fodor (1965) argues that psychological-functional explana-
tions of mental states tell us nothing about their neural realizations and,
vice versa, that neurological research will not shed light on functional
relations between mental states. Thus, psychological explanations can be
developed independently of neuroscience. Such a notion of autonomy
is, however, too strong if we want to integrate such different domains in
virtue of mutual constraints. As Piccinini and Craver (2011) argue, “func-
tional analysis and mechanistic explanation are not autonomous because
they constrain each other; in addition, we argue that they can’t possibly
be autonomous in this sense because functional analysis is just a kind of
mechanistic explanation” (p. 290). To put it differently, functional analy-
sis describes the functional properties of a possible mechanism responsi-
ble for the target phenomenon. Accordingly, functional analysis delivers a
Models of Explanation in Cognitive Science 117
functional sketch of a hypothetical mechanism responsible for the target
capacity. This sketch describes the system’s functional organization, but,
at the same time, the sketch ignores the structural details of the mecha-
nism, such as its physical components and their organization. When the
omitted details are added, the sketch becomes the scheme of the hypothet-
ical mechanism. Ultimately, the scheme becomes a complete explanation,
that is, a multilevel model of the actual mechanism responsible for the
explanandum phenomenon. In reply to this, Roth and Cummins (2017)
argue that proponents of mechanistic integration confuse explanatory
autonomy with confirmation autonomy. Functional-psychological expla-
nations are fully autonomous and thus cannot be integrated with mecha-
nism, but mechanistic neuroscience can verify which functional design is
correct. A mechanistic model can confirm a functional design, but it does
not follow that the model actually explains the target capacity. It is the
functional design that is the key to explaining the target phenomenon.
However, as Roth and Cummins (2017, p. 40) conclude, the autonomy
of functional explanation does not exclude that both functional analysis
and mechanism can constrain each other’s hypotheses.
That being said, we can consider integrated fields autonomous when
adopting a weaker notion of autonomy. According to Piccinini and
Craver (2011),
3.5.2 Phenomenological Understanding
The interpretations of phenomenological explanation discussed earlier
seem flawed. Understanding phenomenology as a personal explana-
tion seems to narrow down its scope and objective, that is, investigating
generic structures of experience. The second interpretation focuses on
the transcendental character of phenomenology and grounds phenom-
enological explanation in intuitive apprehension. Although it is true that
Husserlian phenomenological “seeing” is grounded in intuition, such a
reading does not elucidate what the form of this intuitive explanation
is. The third proposal also considers phenomenology a transcendental
philosophy and argues that the form of intuitive explanation is nomo-
logical; that is, phenomenological intuitions apprehend eidetic laws that
are expressed in the form of generic and universal judgments. However,
the interpretation of phenomenology as a sui generis nomological-intui-
tive explanation, although supported by Husserl’s terminology, construes
transcendental phenomenology as explanatorily weak. Recall that the
force of D-N explanations relies on the universality of the laws of nature
and the certainty of deductive inferences. But phenomenological laws do
not appear to be universal, nor are phenomenological claims deductive;
they are rather the result of the reductive methodology enabling essential
seeing. Moreover, phenomenological lawlike generalizations have limited
predictive power unless they are specified and applied to empirical phe-
nomena, which is hard to do when one applies transcendental reduction,
Models of Explanation in Cognitive Science 125
suspending causal relations. So a condition to make testable phenomeno-
logical predictions is to downplay the transcendental dimension, just as it
is downplayed in phenomenological psychology.
According to the phenomenological agenda, a phenomenologist sus-
pends claims about the causal structure underlying mental phenomena.
However, in some cases, we clearly have to refer to the sub-personal level
in order to fully explain someone’s behavior. The addiction mechanisms
discussed earlier or mental diseases such as depression are clear examples
of such cases. Generally speaking, it seems, on the basis of the debate
surrounding the integration of cognitive science, that in most cases a
complete explanation of a mental phenomenon requires addressing sev-
eral levels including the phenomenal level of experience, psychological
level of cognitive capacities and functions, and the level of neural mecha-
nisms. Clearly phenomenology cannot deliver such a complete explana-
tion alone. It is a result of integrating different frameworks and various
research fields. That said, it is also clear that phenomenology gives us an
insightful account of our mental lives and a general understanding of
the structures of consciousness that can be informative to other research
fields. But rather than a genuine explanation, phenomenology offers us
a specific clarification or way of understanding conscious phenomena.
Now, what kind of understanding does phenomenology provide, and
how is it related to insights from other disciplines?
It is often argued that understanding is inseparable from explanation,
that understanding is the main goal of explanation, or that understand-
ing why a certain phenomenon occurred is part of the explanation of that
phenomenon. However, according to Peter Lipton (2009), understanding
and explanation are distinct: understanding concerns the cognitive ben-
efits that an explanation provides but is not identical with an explanation
as such. Lipton’s line of reasoning is as follows: a good scientific explana-
tion provides cognitive benefits related to four kinds of understanding
or knowledge—of causes, of necessity, of possibility, and of unification
(Lipton, 2009, p. 43). But these epistemic benefits can also be provided
by different types of cognitive activities.
First, according to Lipton, causal information often comes from tacit
knowledge different from explanatorily explicit representations. For
example, seeing a physical model or graphical representation of a system
may give us an understanding of causal information without providing a
genuine explanation. And manipulation, for example, physical interven-
tions on parts of mechanisms, can be very informative with respect to the
causal relations between parts and their organization, but we would not
say that the resulting understanding provides us with a full explanation.
Second, an understanding of necessities may emerge from thought experi-
ments, but we would not say that thought experiments are explanatory.
For example, Galileo’s thought experiment concerning the acceleration
of massive objects shows us that acceleration is necessarily independent
126 Integrating Phenomenology
of mass. He tells us to imagine two falling objects connected with a rope,
one heavier than the other. If acceleration depended on mass, then the
lighter object would slow down the heavier one. But when we consider
these objects together as one mass, they should apparently fall faster than
they would separated. Thus, thinking of acceleration as dependent on
mass leads to contradiction, and so we have to accept that they are inde-
pendent. The thought experiment conveys some worldly necessity, but
it does not tell us why it is the case. We may know that events happen
necessarily, but we can still not know why. Third, as Lipton (2009) argues
further, “we can gain actual understanding from merely potential expla-
nation. As in the case of thought experiments, the kind of understand-
ing gained is modal, but here what is gained is knowledge of possibility
rather than of necessity” (p. 49). Potential explanations give us an under-
standing of how things might be related, but they are not yet complete
explanations. Consider, for example, how-possibly mechanistic models;
they help us in understanding how a target phenomenon may possibly be
realized. Despite that, they are incomplete and merely possible explana-
tions; they give us an actual understanding by, for instance, showing the
type of actual mechanism responsible for the target phenomenon. Fourth,
unification, that is, understanding how different phenomena fit together,
can be achieved through sharing exemplar mechanisms in the scientific
community, which organize research practice, for example, allow for for-
mulating new problems or evaluating proposed solutions but that are not
explanatory.
I think that phenomenology should be considered in an analogous
manner—as an enterprise providing us with a better understanding of
mental phenomena but not complete explanations. The cognitive benefit
that phenomenological understanding provides concerns modal knowl-
edge of mental phenomena. Eidetic analyses give us insight into the nec-
essary conditions for an occurrence of an experience of a certain type,
and these necessities are often expressed in lawlike statements. Genetic
analyses provide us with an understanding of how a studied experience
may possibly be produced by the underlying passive processes of syn-
thesis, as, for example, in the case of the perception of a temporal object
such as a melody elucidated by functions of retention, protention, and
ur-impression (Husserl, 1991).
The phenomenological method of imaginative variation, which enables
“essential seeing” can be conceived of as analogous to thought experiments
which help us understand the necessities governing the world (as I noted
in Chapter 1, Husserl himself considered imaginative variation a form of
phenomenological experiment; see, e.g., Husserl 1980, pp. 44–45). A phe-
nomenologist varies experience in imagination in order to see that certain
modifications lead to contradiction or a change in the experienced object’s
identity, thus demonstrating what the necessary conditions for a certain
type of experience are. Such a procedure is not an explanation or a formal
Models of Explanation in Cognitive Science 127
proof but rather a “see it for yourself” strategy of providing understanding.
Notice that phenomenological thought experiments allow one to answer
some counterfactual questions regarding our mental lives. For example,
seeing an object from a certain perspective and imagining seeing it from
a different perceptual perspective allows me to answer what part of the
object I would see, what kind of perceptual experience I would have, if I
moved and saw it from that different perspective. That kind of prediction is
only possible in virtue of understanding the structure of perception.
Drawing on Michael Wheeler (2013), I call phenomenological under-
standing a constitutive understanding. Constitutive understanding con-
cerns “the identification, articulation and clarification of the conditions
that determine what it is for a phenomenon to be the phenomenon that
it is” (Wheeler, 2013, p. 143). According to Wheeler, phenomenological
analysis is a form of constitutive understanding; it reveals conditions
relevant to the target phenomenon and its key properties. Importantly,
this type of understanding is distinct from enabling understanding,
which
reveals the causal elements, along with the organization of, and the
systematic causal interactions between, those elements, that together
make it intelligible to us how a phenomenon of a certain kind could
be realized or generated in a world like ours.
(p. 143)
Notes
1 In the rest of this book, I use the expression absent the neo– prefix to refer to
the contemporary version of mechanism—as the mechanistic model of expla-
nation or simply mechanistic explanation.
2 Mutual manipulability and its application to mechanism is a controversial
topic (e.g., see Baumgartner & Gebharter, 2016). For the sake of clarity, I
draw on Glennan (2017, p. 44) and understand MM as an epistemic criterion
that serves as a useful instrument for delineating boundaries between relevant
and irrelevant components.
3 Contrary to the epistemic view, Craver (2007) endorses an ontic version of
mechanistic explanation, which acknowledges that “explanations are objec-
tive features of the world” (p. 27).
4 In mechanistic philosophy, the term phenomenal typically refers to natural phe-
nomena that are observable and describable from a third-person perspective.
References
Abraham, R. H., & Shaw, C. D. (1992). Dynamics: The geometry of behavior
(2nd ed.). Addison-Wesley Publishing Company.
Abramova, E., Slors, M. V. P., & Rooij, I. J. E. I. (2017). Enactive mechanistic
explanation of social cognition. In G. Gunzelmann, A. Howes, & T. Tenbrink
(Eds.), Proceedings of the 39th annual meeting of the cognitive science society
(pp. 45–50). Cognitive Science Society.
Andreasen, N. (1999). A unitary model of schizophrenia: Bleuler’s “fragmented
phrene” as schizencephaly. Archives of General Psychiatry, 56, 781–787.
Baddeley, A. (2007). Working memory, thought, and action. Oxford University
Press.
Bangu, S. (2017). Scientific explanation and understanding: Unificationism recon-
sidered. European Journal for Philosophy of Science, 7(1), 103–126.
Barker, M. J. (2019). Eliminative pluralism and integrative alternatives: The case
of SPECIES. The British Journal for the Philosophy of Science, 70(3), 657–681.
Barrett, D. (2014). Functional analysis and mechanistic explanation. Synthese,
191(12), 2695–2714.
130 Integrating Phenomenology
Baumgartner, M., & Gebharter, A. (2016). Constitutive relevance, mutual manip-
ulability, and fat-handedness. The British Journal for the Philosophy of Science,
67(3), 731–756.
Bechtel, W. (1998). Representations and cognitive explanations: Assessing the
dynamicist’s challenge in cognitive science. Cognitive Science, 22(3), 295–317.
Bechtel, W. (2008). Mental mechanisms: Philosophical perspectives on cognitive
neuroscience. Taylor & Francis Group.
Bechtel, W. (2009a). Explanation: Mechanism, modularity, and situated cogni-
tion. In P. Robbins & M. Aydede (Eds.), The Cambridge handbook of situated
cognition (pp. 155–170). Cambridge University Press.
Bechtel, W. (2009b). Looking down, around, and up: Mechanistic expla-
nation in psychology. Philosophical Psychology, 22(5), 543–564.
doi:10.1080/09515080903238948
Bechtel, W. (2019). Resituating cognitive mechanisms within heterarchical net-
works controlling physiology and behavior. Theory & Psychology, 29(5), 620–
639. doi:10.1177/0959354319873725
Bechtel, W., & Abrahamsen, A. (2005). Explanation: A mechanist alternative.
Studies in History and Philosophy of Science Part C: Studies in History and
Philosophy of Biological and Biomedical Sciences, 36(2), 421–441.
Bechtel, W., & Abrahamsen, A. (2010). Dynamic mechanistic explanation:
Computational modeling of circadian rhythms as an exemplar for cognitive
science. Studies in History and Philosophy of Science Part A, 41(3), 321–333.
Bechtel, W., & Richardson, R. C. (2010). Discovering complexity: Decomposition
and localization as strategies in scientific research. MIT Press.
Bechtel, W., & Wright, C. (2009). What is psychological explanation? In P. Calvo
& J. Symons (Eds.), Routledge companion to philosophy of psychology (pp.
113–130). Routledge.
Beer, R. D. (2000). Dynamical approaches to cognitive science. Trends in Cognitive
Sciences, 4, 91–99.
Beer, R. D. (2003). The Dynamics of active categorical perception in an evolved
model agent. Adaptive Behavior, 11(4), 209–243.
Bickle, J. (2003). Philosophy of neuroscience: A ruthlessly reductive approach.
Kluwer Academic Publishers.
Block, N. (1980). What is functionalism? In N. Block (Ed.), Readings in philoso-
phy of psychology (Vol. 1, pp. 171–184). Harvard University Press.
Bermúdez, J. L. (2000). Personal and sub-personal; A difference without a distinc-
tion. Philosophical Explorations, 3(1), 63–82.
Campbell, J. (2002). Rationality, meaning and the analysis of delusion. Philosophy,
Psychiatry and Psychology, 8, 89–100.
Chalmers, D. J. (1996). The conscious mind: In search of a fundamental theory.
Oxford University Press.
Chemero, A. (2000). Anti-representationalism and the dynamical stance.
Philosophy of Science, 67(4), 625–647. doi.org/10.1086/392858
Chemero, A. (2011). Radical embodied cognitive science. MIT Press.
Churchland, P. S. (1986). Neurophilosophy: Toward a unified science of the
mind-brain. MIT Press.
Clark, A., & Chalmers, D. (1998). The extended mind. Analysis, 58(1), 7–19.
Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future
of cognitive science. Behavioral and Brain Sciences, 36(3), 181–204.
Models of Explanation in Cognitive Science 131
Colombo, M. (2013). Constitutive relevance and the personal/subpersonal dis-
tinction. Philosophical Psychology, 26(4), 547–570.
Colombo, M., & Heinz, A. (2019). Explanatory integration, computational phe-
notypes, and dimensional psychiatry: The case of alcohol use disorder. Theory
& Psychology, 29(5), 697–718.
Cosmides, L., & Tooby, J. (1995). From function to structure: The role of evo-
lutionary biology and computational theories in cognitive neuroscience. In M.
S. Gazzaniga (Ed.), The cognitive neurosciences (pp. 1199–1210). MIT Press.
Cummins, R. (1975). Functional analysis. The Journal of Philosophy, 72(20),
741–765. doi:10.2307/2024640
Cummins, R. (2000). “How does it work” versus “what are the laws?”: Two
conceptions of psychological explanation. In F. Keil & R. A. Wilson (Eds.),
Explanation and cognition (pp. 117–145). MIT Press.
Craver, C. F. (2006). When mechanistic models explain. Synthese, 153, 355–376.
Craver, C. F. (2007). Explaining the brain: Mechanisms and the mosaic unity of
neuroscience. Clarendon Press.
Craver, C. F., & Darden, L. (2013). In search of mechanisms: Discoveries across
the life sciences. University of Chicago Press.
Craver, C. F., & Kaplan, D. M. (2020). Are more details better? On the norms
of completeness for mechanistic explanations. The British Journal for the
Philosophy of Science, 71(1), 287–319.
Darden, L., & Maull, N. (1977). Interfield theories. Philosophy of Science, 44(1),
43–64.
Dennett, D. C. (1969). Content and consciousness. Routledge and Kegan Paul.
Dennett, D. C. (1981). True believers: The intentional strategy and why it works.
In A. Heath (Ed.), Scientific explanation (pp. 53–75). Oxford University Press.
Dennett, D. C. (1991a). Consciousness explained. Little, Brown.
Dennett, D. C. (1991b). Real patterns. Journal of Philosophy, 88(1), 27–51.
Dennett, D. C. (2003). Who’s on first? Heterophenomenology explained. Journal
of Consciousness Studies, 10(9–10), 19–30.
Dennett, D. C. (2005). Sweet dreams: Philosophical obstacles to a science of con-
sciousness. MIT Press.
Drayson, Z. (2014). The personal/subpersonal distinction. Philosophy Compass,
9(5), 338–346.
Dupré, J. (1995). The disorder of things: Metaphysical foundations of the dis-
unity of science. Harvard University Press.
Eliasmith, C. (2001). Attractive and in-discrete: A critique of two putative virtues
of the dynamicist theory of mind. Minds and Machines, 11, 417–426.
Engel, G. (1977). The need for a new medical model: A challenge for biomedicine.
Science, 196(4286), 129–136.
Felleman, D. J., & van Essen, D. C. (1991). Distributed hierarchical processing in the
primate cerebral cortex. Cerebral Cortex, 1(1), 1–47. doi:10.1093/cercor/1.1.1
Fodor, J. (1965). Explanations in psychology. In M. Black (Ed.), Philosophy in
America (pp. 161–179). Cornell University Press.
Fodor, J. (1974). Special sciences (Or: The disunity of science as a working
hypothesis). Synthese, 28(2), 97–115.
Fodor, J. (1983). The modularity of mind. MIT Press.
Friston, K. J. (2010). The free-energy principle: A unified brain theory? Nature
Reviews Neuroscience, 11(2), 127–138.
132 Integrating Phenomenology
Gallagher, S., & Brøsted Sørensen, J. (2006). Experimenting with phenomenol-
ogy. Consciousness and Cognition, 15(1), 119–134.
Gallagher, S., & Zahavi, D. (2012). The phenomenological mind (2nd ed.).
Routledge.
Glennan, S. (2017). The new mechanical philosophy. Oxford University Press.
Godfrey-Smith, P. (2014). Philosophy of biology. Princeton University Press.
Golonka, S., & Wilson, A. D. (2019). Ecological mechanisms in cognitive science.
Theory & Psychology, 29(5), 676–696.
Grobler, A., & Wiśniewski, A. (2005). Explanation and theory evaluation. In W.
R. Festa, A. Aliseda, & J. Peijnenburg (Eds.), Cognitive structures in scientific
inquiry: Poznan studies in the philosophy of the sciences and the humanities
(pp. 299–310). Brill | Rodopi.
Haken, H., Kelso, J. A. S., & Bunz, H. (1985). A theoretical model of phase transi-
tions in human hand movements. Biological Cybernetics, 51, 347–442.
Halina, M. (2017). Mechanistic explanation and its limits. In S. Glennan & P.
Illari (Eds.), The Routledge handbook of mechanisms and mechanical philoso-
phy (pp. 213–224). Taylor & Francis Group.
Hempel, C. G. (1965). Aspects of scientific explanation and other essays in the
philosophy of science. The Free Press.
Hempel, C. G., & Oppenheim, P. (1948). Studies in the logic of explanation.
Philosophy of Science, 15(2), 135–175.
Hensel, W. M. (2013). On reduction and interfield integration in neuroscience. In
M. Miłkowski & K. Talmont-Kamiński (Eds.), Regarding the mind, naturally:
Naturalist approaches to the sciences of the mental (pp. 167–181). Cambridge
Scholars Publishing.
Hodgkin, A. L., & Huxley, A. F. (1939). Action potentials recorded from inside a
nerve fibre. Nature, 144, 710–711.
Hohwy, J. (2013). The predictive mind. Oxford University Press.
Hurlburt, R. T., & Schwitzgebel, E. (2007). Describing inner experience?
Proponent meets skeptic. MIT Press.
Husserl, E. (1969). Formal and transcendental logic (D. Cairns, Trans.). Martinus
Nijhoff.
Husserl, E. (1973). Experience and judgment: Investigations in a genealogy of
logic (J. S. Churchill & K. Ameriks, Trans.). Routledge & Kegan Paul.
Husserl, E. (1977). Phenomenological psychology (J. Scanlon, Trans.). Martinus
Nijhoff.
Husserl, E. (1980). Phenomenology and the foundations of the sciences: Third
book of Ideas pertaining to a pure phenomenology and to a phenomenological
philosophy (T. E. Klein & W. E. Pohl, Trans.). Martinus Nijhoff.
Husserl, E. (1982). Ideas pertaining to a pure phenomenology and to a phenom-
enological philosophy; First book: General introduction to a pure phenomenol-
ogy (F. Kersten, Trans.). Springer.
Husserl, E. (1989). Ideas pertaining to a pure phenomenology and to a phenom-
enological philosophy; Second book: Studies in the phenomenology of consti-
tution (R. Rojcewicz & A. Schuwer, Trans.). Springer.
Husserl, E. (1991). On the phenomenology of the consciousness of internal time
(1893–1917) (J. B. Brough, Trans.). Springer.
Husserl, E. (2001). Analyses concerning passive and active synthesis: Lectures on
transcendental logic (A. J. Steinbock, Trans.). Springer.
Models of Explanation in Cognitive Science 133
Husserl, E. (2002). Philosophy as rigorous science (M. Brainard, Trans.). New
Yearbook for Phenomenology and Phenomenological Philosophy, 2, 249–295.
Husserl, E. (2005). Phantasy, image consciousness, and memory (1898–1925)
(Vol. 11; J. B. Brough, Trans.). Springer Science & Business Media.
Hutchins, E. (1995). How a cockpit remembers its speeds. Cognitive Science, 19,
265–288.
Jack, A. I., & Roepstorff, A. (2002). Introspection and cognitive brain mapping:
From stimulus–response to script–report. Trends in Cognitive Sciences, 6(8),
333–339.
Kaplan, D. M. (2017). Integrating mind and brain science: A field guide. In D.
M. Kaplan (Ed.), Explanation and integration in mind and brain science (pp.
1–28). Oxford University Press.
Kaplan, D. M., & Craver, C. F. (2011). The explanatory force of dynamical and
mathematical models in neuroscience: A mechanistic perspective. Philosophy
of Science, 78(4), 601–627.
Kelso, J. A. S. (1995). Dynamic patterns: The self-organization of brain and
behavior. MIT Press.
Kendler, K., Zachar, P., & Craver, C. (2011). What kinds of things are psychi-
atric disorders? Psychological Medicine, 41(6), 1143–1150. doi:10.1017/
S0033291710001844
Komorowska-Mach, J. (2019). Introspection – one or more? Pluralism about self-
knowledge. Filozofia Nauki, 27(1), 5–25. doi.org/10.14394/filnau.2019.0001
Lakatos, I. (1980). The methodology of scientific research programmes.
Cambridge University Press.
Lingis, A. (1986). Phenomenological explanations. Martinus Nijhoff Publishers.
Lipton, P. (2009). Understanding without explanation. In H. W. de Regt, S.
Leonelli, & K. Eigner (Eds.), Scientific understanding: Philosophical perspec-
tives (pp. 43–63). University of Pittsburgh Press.
McDowell, J. (1986). Singular thought and the extent of ‘inner space.’ In J.
McDowell & P. Pettit (Eds.), Subject, thought, and context (pp. 137–168).
Oxford University Press.
McDowell, J. (1994). The content of perceptual experience. Philosophical
Quarterly, 44, 190–205.
Millikan, R. G. (2002). Biofunctions. In A. Ariew, R. Cummins, & M. Perlman
(Eds.), Functions: New essays in the philosophy of psychology and biology.
Oxford University Press.
Miłkowski, M. (2016). Integrating cognitive (neuro)science using mechanisms.
AVANT. The Journal of the Philosophical-Interdisciplinary Vanguard, VII(2),
45–67.
Miłkowski, M. (2017). Mechanisms and the mental. In S. Glennan & P. Illari
(Eds.), The Routledge handbook of mechanisms and mechanical philosophy
(pp. 74–88). Taylor & Francis Group.
Miłkowski, M., Clowes, R., Rucińska, Z., Przegalińska, A., Zawidzki, T., Krueger,
J., Gies, A., McGann, M., Afeltowicz, Ł., Wachowski, W., Stjernberg, F.,
Loughlin, V., & Hohol, M. (2018). From wide cognition to mechanisms: A silent
revolution. Frontiers in Psychology, 9. doi.org/10.3389/fpsyg.2018.02393
Miłkowski, M., & Hohol, M. (2020). Explanations in cognitive science:
Unification versus pluralism. Synthese. Advanced online publication. doi.
org/10.1007/s11229-020-02777-y
134 Integrating Phenomenology
Mitchell, S. D. (2002). Integrative pluralism. Biology and Philosophy, 17(1),
55–70.
Nagel, E. (1961). The structure of science: Problems in the logic of scientific
explanation. Brace & World.
Newell, A., & Simon, H. A. (1972). Human problem solving. Prentice Hall.
Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can know: Verbal
reports on mental processes. Psychological Review, 84, 231–359.
Oppenheim, P., & Putnam, H. (1958). Unity of science as a working hypothesis.
Minnesota Studies in the Philosophy of Science, 2, 3–36.
Oullier, O., De Guzman, G. C., Jantzen, K. J., Lagarde, J., & Kelso, J. A. S.
(2008). Social coordination dynamics: Measuring human bonding. Social
Neuroscience, 3(2), 178–192.
Piccinini, G. (2003). Data from introspective reports: Upgrading from common
sense to science. Journal of Consciousness Studies, 10(9–10), 141–156.
Piccinini, G. (2010). How to improve on heterophenomenology: The self-mea-
surement methodology of first-person data. Journal of Consciousness Studies,
17(3–4), 84–106.
Piccinini, G., & Craver, C. (2011). Integrating psychology and neuroscience:
Functional analyses as mechanism sketches. Synthese, 183(3), 283–311.
Port, R. (2003). Meter and speech. Journal of Phonetics, 31, 599–611.
Prinz, J. (2012). The conscious brain. Oxford University Press.
Putnam, H. (1960). Minds and machines. In S. Hook (Ed.), Dimensions of mind
(pp. 138–164). New York University Press.
Raja, V., Biener, Z., & Chemero, A. (2017). From Kepler to Gibson. Ecological
Psychology, 29(2), 146–160.
Rodemeyer, L. (2020). Levels of embodiment: A Husserlian analysis of gender
and the development of eating disorders. In C. Tewes & G. Stanghellini (Eds.),
Time and body: Phenomenological and psychopathological approaches (pp.
234–255). Cambridge University Press. doi:10.1017/9781108776660.019
Rosenthal, D. (2005). Consciousness and mind. Clarendon Press.
Roth, M., & Cummins, R. (2017). Neuroscience, psychology, reduction, and
functional analysis. In D. M. Kaplan (Ed.), Explanation and integration in
mind and brain science (pp. 29–43). Oxford University Press.
Roux, S. (2017). From the mechanical philosophy to early modern mechanisms.
In S. Glennan & P. Illari (Eds.), The Routledge handbook of mechanisms and
mechanical philosophy (pp. 26–45). Taylor & Francis Group.
Ruiz-Mirazo, K., & Moreno, A. (2004). Basic autonomy as a fundamental step in
the synthesis of life. Artificial Life, 10(3), 235–259.
Salmon, W. C. (1984). Scientific explanation and the causal structure of the world.
Princeton University Press.
Salmon, W. C. (1989). Four decades of scientific explanation. University of
Pittsburgh Press.
Sass, L. A. (2014). Explanation and description in phenomenological psychopa-
thology. Journal of Psychopathology, 20, 366–376.
Schacter, D. L. (2012). Constructive memory: Past and future. Dialogues in
Clinical Neuroscience, 14(1), 7–18.
Scoville, W. B., & Milner, B. (1957). Loss of recent memory after bilateral hip-
pocampal lesions. Journal of Neurology, Neurosurgery, and Psychiatry, 20,
11–21.
Models of Explanation in Cognitive Science 135
Scriven, M. (1975). Causation as explanation. Nous, 9, 3–16.
Schwitzgebel, E. (2012). Introspection, what? In D. Smithies & D. Stojlar (Eds.),
Introspection and consciousness (pp. 29–48). Oxford University Press.
Simon, H. A. (1962). The architecture of complexity. Proceedings of the American
Philosophical Society, 106(6), 467–482.
Smith, J. (2010). Seeing other people. Philosophy and Phenomenological
Research, 81(3), 731–748.
Smith, L. B., & Thelen, E. (2003). Development as a dynamic system. Trends in
Cognitive Sciences, 7(8), 343–348.
Sober, E. (1983). Equilibrium explanation. Philosophical Studies, 43(2), 201–210.
Sowa, R. (2012). Eidetics and its methodology (E. A. Behnke, Trans.). In S. Luft
& S. Overgaard (Eds.), The Routledge companion to phenomenology (pp.
254–265). Routledge.
Stich, S., & Nichols, S. (2003). Folk psychology. In S. Stich & T. A. Warfield (Eds.),
The Blackwell guide to philosophy of mind (pp. 235–255). Blackwell Publishing.
Walmsley, J. (2008). Explanation in dynamical cognitive science. Minds and
Machines, 18(3), 331–348. doi:10.1007/s11023-008-9103-9
Watkins, J. W. (1957). Historical explanation in the social sciences. The British
Journal for the Philosophy of Science, 8(30), 104–117.
Weiskopf, D. A. (2011). Models and mechanisms in psychological explanation.
Synthese, 183, 313–338. doi:10.1007/s11229-011-9958-9
Wheeler, M. (2013). Science friction: Phenomenology, naturalism and cognitive
science. Royal Institute of Philosophy Supplement, 72, 135–167.
Williams, H. (2020). Husserl on personal level explanation. Human Studies, 43,
1–22. doi:10.1007/s10746-020-09537-4
Wimsatt, W. C. (1986). Forms of aggregativity. In A. Donagan, A. N. Perovich, &
M. V. Wedin (Eds.), Human nature and natural knowledge: Essays presented
to Marjorie Grene on the occasion of her seventy-fifth birthday (pp. 259–291).
Springer.
Woodward, J. (2003). Making things happen: A theory of causal explanation.
Oxford University Press.
Woodward, J. (2017). Explanation in neurobiology. In D. M. Kaplan (Ed.),
Explanation and integration in mind and brain science (pp. 70–100). Oxford
University Press.
Wright, L. (1976). Teleological explanations. University of California Press.
van Gelder, T., & Port, R. F. (1995). It’s about time: An overview of the dynamical
approach to cognition. In R. F. Port & T. van Gelder (Eds.), Mind as motion:
Explorations in the dynamics of cognition. MIT Press.
van Gelder, T. (1998). The dynamical hypothesis in cognitive science. Behavioral
and Brain Sciences, 21(5), 615–628.
Ylikoski, P. K. (2017). Social mechanisms. In S. Glennan & P. Illari (Eds.), The
Routledge handbook of mechanisms and mechanical philosophy (pp. 401–
412). Taylor & Francis Group.
Zahavi, D. (2007). Killing the straw man: Dennett and phenomenology.
Phenomenology and the Cognitive Sciences, 6(1–2), 21–43.
Zednik, C. (2011). The nature of dynamical explanation. Philosophy of Science,
78(2), 238–263. doi:10.1086/659221
Part II
Phenomenology and
Mechanism
In Search of Constraints
4 Phenomenology and Functionalism
4.1 Introduction1
Functionalism is recognized as one of the most influential approaches to
mind and cognition—one usually perceived as incongruent with phenom-
enological approaches descending from the Husserlian tradition. I think
this is misleading, and there are arguments for connections and similari-
ties between these traditions. In philosophy of mind, functionalism may
be approached from a number of different perspectives (for an overview,
see, e.g., Block, 1980). First, it may be approached as a metaphysical
theory of mental states, which argues that to be a mental state is to be
a functional state individuated by causal relations to other mental states
and inputs and outputs. There are different types of such functionalist
theories of mind, including the most popular one: computation-repre-
sentation functionalism. The key idea behind this version of function-
alism is that mental states can be understood as Turing machine table
states (e.g., Putnam, 1960) and the mind as a sort of computer program.
In this fashion, Dreyfus (Dreyfus & Hall, 1982) and McIntyre (1986)
proposed computational-functionalist readings of Husserl. Second, func-
tionalism can be approached as an explanatory strategy called functional
analysis, which is applied in psychology and cognitive science. It con-
sists in decomposing a cognitive system into component processes and
capacities and describing its functional organization (Cummins, 1975;
Fodor, 1968). Functional analysis may differ depending on what type of
functionalism one is committed to, and it does not have to necessarily
entail computationalism (Piccinini, 2010). A proponent of computation-
representation functionalism will decompose a target system into a set
of specialized computational modules processing input information and
returning outputs and will produce an explanation in terms of program
execution and data manipulation. But Cummins-style functional analy-
sis, which relies on a more general systemic notion of function as a causal
role, will decompose the system into a set of sub-capacities that contrib-
ute to the system’s functioning. These sub-capacities are then described in
dispositional-behavioral terms, not in computational terms.
DOI: 10.4324/9781003035367-7
140 Phenomenology and Mechanism
In this chapter, I propose a functionalist reading of Husserlian phe-
nomenology and argue that such a reading opens a way toward inte-
grating phenomenology with cognitive science, in particular with the
mechanistic explanatory framework. First, I critically discuss a func-
tional-computational interpretation of Husserl coined in the discussion
between Dreyfus and McIntyre. I outline the deficiencies of Dreyfus’s and
McIntyre’s proposals, one of which is not addressing the notion of func-
tion in Husserl. Next, I propose shifting perspective and approaching the
relation between phenomenology and functionalism from a methodolog-
ical perspective. Methodological similarities were already indicated by
McIntyre (1986), and more recently, Paul Livingston (2005) has argued
that there is continuity between the phenomenological and functional
methods of conceptual and logical analysis of our experience. My argu-
ment is different. The key claim is that in Husserlian “functional phe-
nomenology,” as introduced in the first book of Ideas (Husserl, 1982),
we find an original notion of intentional function as well as a method
of decomposition that can be understood as analogous to the explana-
tory strategy of functional analysis. Finally, I argue that the proposed
functionalist reading of phenomenology opens a new approach to the
integration of phenomenology with cognitive sciences. Accordingly, phe-
nomenology provides constraints concerning the functional structure of
the phenomenon in question that can be applied to explanatory models.
I demonstrate this on two examples: vision studies and the evolutionary-
developmental account of first-person perspective.
Not only with regard to the difficulties which it arrives at, but also
with regard to the ranking of problems from the standpoint of the
idea of an absolute cognition, [pure hyletic phenomenology] obvi-
ously stands far below the noetic and functional phenomenology
(both of which, moreover, are properly not to be separated).
(Husserl, 1982, para. 86)
4.4.1 Phenomenological Decomposition
One of the key ideas in phenomenology is that in the naïve attitude we
encounter objects in the world which are, so to speak, ready-made. The
task of phenomenology is to break down this naivety and show how
objects of experience are constituted in the activity of consciousness.
Attaining the phenomenological attitude through the reduction is the
first step. What is next is the process of analyzing an experience in order
to extract its structures. Taking into account the idea of functional phe-
nomenology, I argue that it employs a method that relies on decomposing
an experience into its noematic contents and related functions involved
in the production of experience. I call this type of decomposition static,
since it is closely related to the idea of static phenomenology and later
developed into the genetic approach (see Chapter 1 for an introduction
to static and genetic phenomenology).
Phenomenology and Functionalism 151
Static decomposition is, relatively speaking, the simplest and most
intuitive way to analyze experience, and it is already present in Husserl’s
early works, in which he employs the procedure of fragmenting objects
of experience. In Logical Investigations, he writes that “the division of a
whole into a plurality of mutually exclusive pieces we call a piecing or
fragmentation (Zerstückung) of the same” (Husserl, 2001b, p. 29). In the
third investigation, he develops a general mereological theory of wholes
and parts, and specifically, he discusses piecing experiences, that is, phe-
nomenological contents.
There is no space here for a detailed discussion of this complex mereo-
logical theory, so I mention only its main ideas. The key distinction intro-
duced at the beginning of the investigation is the distinction between
independent and “non-independent” parts. A part is anything that can
be distinguished in an object (Husserl, 2001b, pp. 5, 29). An independent
part can also be called a piece (or portion), since it can be a separate
object of analysis. A dependent part is tied to other contents and cannot
be separated in reality, only in abstraction. Sometimes Husserl also calls a
dependent part a moment or an abstract part. Moreover, each part can be
composed of other parts; thus, it can also be decomposed; the whole pro-
cedure of decomposition is iterative and assumes the existence of differ-
ent levels of decomposition. Husserl (2001b) also introduces the relation
of foundation: “a content of the species A is founded upon a content of
the species B, if an A can by its essence (i.e., legally, in virtue of its specific
nature) not exist, unless a B also exists” (p. 34). The relation of founda-
tion is important because it expresses the necessary (a priori necessity
based on “material essences”) relations between parts and wholes.
Let us consider an example: perceiving an apple tree. First, we see the
tree’s shape, the colors of its trunk, leaves, and fruits. When we touch
its bark, we experience shapes, texture, and temperature. We can also
sense the smell of the fruits and hear the leaves rustle in the wind. All
these sensory contents are parts of the unitary experience of an apple
tree. Whereas some of them are independent parts and can be analyzed
as separate pieces, others are nonindependent (or abstract parts) because
they need to be analyzed in reference to one another. For instance, we can
consider the shape of a leaf separately, but a color cannot exist without a
shape, and the tactile experience of texture, without extension. When we
move around the object, the shape, the size of the parts, and the colors
vary—these are partial moments of the perceptual experience. The whole
experience of the tree can be thus divided into different aspects contain-
ing different sensory, as well as representational, contents.4
Importantly, this decomposition can be accompanied by another phe-
nomenological method, namely, imaginary variation, in which the con-
tents of experience are varied in order to apprehend the experience’s
invariant structure. After the procedure of variation, the studied experi-
ence is no longer seen as a particular experience of an individual subject
152 Phenomenology and Mechanism
but as a type of experience, for example, a type of perceptual experience.
What is especially interesting for us here is that the decomposition of
contents is supplemented in Husserl with the decomposition of noetic
functions. Units of meaning, or noemata in Husserlian terminology, iden-
tified in the decomposition of contents are related to noetic functions.
Thanks to noetic functions, sensory contents are correlated with con-
ceptual (noematic) contents, which can then be expressed in the form of
judgments. To put it differently, noetic functions play the role of repre-
sentational functions, which give meaning to sense-data by correlating it
with mental representations.
Husserl, however, was not satisfied with static decomposition and saw
that the simple dichotomy of sensory matter and noetic form did not
cover the whole process of the production of experience. He recognized
that the level of intentional functions related to the ego’s activity is just
one among several levels of constitution, including the level of passive
synthesis. As I argued in previous chapters, in order to explain the active,
as well as passive, synthesis of experience, Husserl develops the idea of
genetic phenomenology, which aims to explain the genesis of the very pro-
cess of constitution. The active level of constitution concerns all processes
related to all cognitive activities accessible to the subject’s awareness, for
example, producing judgments. Noemata understood as linguistic mean-
ings, or as mental representations, are part of this active synthesis. By the
passive level of constitution, Husserl understands processes that influence
cognitive acts and their contents but are not explicitly aware. An exam-
ple of such passive syntheses are different sorts of associations related
to affectivity and kinesthesis as well as previous experiences (see, e.g.,
Welton, 1982).
According to Husserl (1973), the static approach, which describes the
constitution of an intentional object as “the product of an objectivating
operation of the ego” (p. 72), provides “an understanding of intentional
accomplishment” and “reveals to us the graduated levels of intentional
objects that emerge in founded apperceptions of a higher level as objec-
tive senses and in functions of sense giving, and it reveals to us how they
function in them, etc.” (2001a, p. 629). But the static approach should
be supplemented by a genetic investigation that explains the passive pro-
cesses accompanying and influencing the acts of the ego. Genetic phe-
nomenology discloses that under noetic functions, there are motivational
and associative functions that also contribute to our experience.
To see how the genetic approach contributes to phenomenological
decomposition, recall what was said in Chapter 1 concerning Husserl’s
model of time-consciousness. The cognitive capacity to produce the expe-
rience of an object that extends in time can be divided into three subfunc-
tions: retention (retaining in consciousness parts of the object that are
no longer present), protention (anticipating the parts of the object not
yet present), and the primal- or “ur-impression” (sensory receptivity in
Phenomenology and Functionalism 153
the present moment). According to Husserl, we do not experience the
momentary “now” of a temporal object as sensorially stimulating (the
ur-impression is, so to speak, a product of idealization). On the con-
trary, the lived experience of “now” is a product of all these functions,
which means that the sensory core of “now” is always accompanied
by retentional-protentional formations. Husserl also divides the reten-
tional function into two subfunctions: so-called longitudinal retention
and transverse retention. The former is directed toward the “how” of the
stream of consciousness, namely, into succeeding phases synthesized into
a non-objectified continuity. The latter is directed toward what appears
in the flow, that is, a temporal object. As I mentioned in Chapter 1, the
static model was revised by Husserl in the development of genetic phe-
nomenology (see Kortooms, 2002, pp. 175–284). For example, different
functions of affectivity and embodiment, such as kinesthesis, are included
in the genetic model of time-consciousness (Pokropski, 2015). In this
way, Husserl demonstrated that “the functions of corporeality belong to
the functions of the passive pre-constitution” (Landgrebe, 1981, p. 56).
Associative functions were another addition to the model, that is, asso-
ciations with past experiences influencing the occurrent experience. Thus,
the retentional function as described earlier was called “near retention,”
because it is involved with the production of our experience of the living
present; it retains the fading phases of perception, for example, the phases
of visual adumbrations of perceptual objects—and for the associative
process, Husserl introduced a new term: “far retention” (see Rodemeyer,
2006, pp. 88–90). Far retention is not limited to a current experience and
reaches back to past experiences. It is a form of passive association con-
cerning a given perceived object. Far retention plays an important role in
object recognition; that is, when we recognize an object, its meaning is
enriched by associations with previous experiences of that object.
To sum up, Husserlian phenomenology applies an explanatory strategy
of decomposition that shares similarities with functional decomposition
as applied in functionalism. Phenomenological decomposition has two
main levels: the decomposition of experiential contents and related func-
tions. But as Husserl argues, this is merely a descriptive method, and
it should be supplemented with the explanatory genetic approach. The
genetic approach decomposes the passive processes (such as affectivity,
embodiment, associations) accompanying the activities of the ego.
On the one hand, we perceive and recognize the object of perception, and
the object’s meaning (the noematic core) remains constant. But, on the
other, the flow of adumbrations changes in accordance with bodily move-
ments. Importantly, Husserl remarks that the co-perceived adumbrations
of an object are, to some extent, undetermined, but the indeterminacy can
also be reduced by bodily movements; that is, bodily movements generate
new adumbrations, for instance, revealing a side of the object that was
previously not in view.
Phenomenology and Functionalism 159
For Husserl (1989a), the body is “the medium of all perception; it is
the organ of perception and is necessarily involved in all perception” (p.
61, emphasis in the original). Visual perception is therefore an embodied
process that is characterized by bodily spatial orientation and constitutes
the visual perspective. As Husserl (1989a) writes,
the Body acquires as the bearer of the zero point of orientation, the
bearer of the here and the now, out of which the pure Ego intu-
its space and the whole world of the senses. Thus each thing that
appears has eo ipso an orienting relation to the Body, and this refers
not only to what actually appears but to each thing that is supposed
to be able to appear.
(p. 61)
4.6 Conclusion
In this chapter, I considered two different functionalist interpretations of
Husserlian phenomenology. I argued that a strong computational reading
of phenomenology is implausible for several reasons. Another function-
alist reading of phenomenology, one that I argued in favor of, adopts
a methodological perspective and points to the similarities between the
explanatory strategy of functional analysis and the phenomenological
method of decomposition. I argued that Husserl developed an original
notion of intentional function and applied it to decompose experiences
and identify the functions responsible for the production or constitution
of those experiences. According to this interpretation, phenomenologi-
cal descriptions of the intentional structure of experience can be read as
analogous to functional descriptions. Finally, I considered the implica-
tions of such a functionalist reading for the project of naturalization. I
argued that nonreductive naturalization should be understood in terms
of an integration of research fields that relies on providing constraints.
Taking under consideration two case studies, the first-person perspective
Phenomenology and Functionalism 163
and vision studies, I have shown that functional phenomenological analy-
ses can provide general constraints concerning functions involved in the
constitution of the experience in question and thus become a part of a
chain of interlevel constraints constituting a multilevel integrative expla-
nation. Phenomenological constraints based on the analysis of first-per-
son experience can be specified on the sub-personal level in the functional
model of cognitive maps—and visual perception based on predictive cod-
ing theory. These functional models, in turn, provide constraints on the
models of underlying neural mechanisms.
Notes
1 An earlier version of this chapter appeared as “Phenomenology and
Functional Analysis: A Functionalist Reading of Husserlian Phenomenology”
(Pokropski, 2020).
2 This is not the only possible interpretation of Husserl’s noema (for an over-
view, see Drummond, 1997). For example, a different reading of noema was
proposed by Gurwitsch (2009), who argues for perceptual noema. Perceptual
noema, in short, is an adumbration of a thing seen from a certain perspective.
Gurwitsch’s noema is not abstract and imperceivable but, on the contrary, is
concrete and perceivable. Thus, perception could be conceived as more direct.
However, some argue (e.g., Dreyfus, 1982) that it is an extension of Husserl’s
notion of noema rather than Husserl’s original idea.
3 Another concept of function is the set-theoretical concept, according to which
function is an abstract relation between two sets. The set-theoretical understand-
ing of function is not present in Husserl’s works, but, interestingly, it was used by
Bolzano, whose works influenced Husserl (Benoist, 2002; Centrone, 2010).
4 A more recent example of the phenomenological decomposition of the con-
tents of experience can be found in the micro-phenomenological studies of
first-person experience conducted by Petitmengin et al. (2013, 2019). The
micro-phenomenological method is a second-person approach and begins
from a phenomenological interview. The descriptions of experience from
micro-phenomenological interviews are analyzed and decomposed in order to
extract the structure of the studied experience (see Chapter 5 for an extended
discussion of micro-phenomenology).
5 Note, however, that some (e.g., van Es, 2020) defend an instrumentalist read-
ing of predictive processing models.
References
Benoist, J. (2002). Husserl and Bolzano. In A. T. Tymieniecka (Ed.), Phenomenology
world-wide: Analecta Husserliana (pp. 98–99). Springer.
Block, N. (1980). What is functionalism? In N. Block (Ed.), Readings in philoso-
phy of psychology (Vol. 1, pp. 171–184). Harvard University Press.
Block, N. (2007). Consciousness, function, and representation: Collected papers.
MIT Press.
Centrone, S. (2010). Functions in Frege, Bolzano and Husserl. History and
Philosophy of Logic, 31(4), 315–336. doi:10.1080/01445340.2010.511795
Clark, A. (2013). Whatever next? Predictive brains, situated agents, and the future
of cognitive science. Behavioral and Brain Sciences, 36(3), 1–24.
164 Phenomenology and Mechanism
Clark, A. (2015). Surfing uncertainty: Prediction, action, and the embodied mind.
Oxford University Press.
Cummins, R. (1975). Functional analysis. The Journal of Philosophy, 72(20),
741–765. doi:10.2307/2024640
Cummins, R. (2000). “How does it work” versus “what are the laws?”: Two
conceptions of psychological explanation. In F. Keil & R. A. Wilson (Eds.),
Explanation and cognition (pp. 117–145). MIT Press.
Craver, C. F. (2007). Explaining the brain: Mechanisms and the mosaic unity of
neuroscience. Clarendon Press.
Dreyfus, H. (1982). Husserl’s perceptual noema. In H. L. Dreyfus & H. Hall
(Eds.), Husserl, intentionality and cognitive science (pp. 97–123). MIT Press.
Dreyfus, H. (1988). Husserl’s epiphenomenology. In H. Otto & J. Tuedio (Eds.),
Perspectives on mind (pp. 85–101). Reidel.
Dreyfus, H., & Hall, H. (1982). Introduction. In H. L. Dreyfus & H. Hall (Eds.),
Husserl, intentionality and cognitive science (pp. 1–27). MIT Press.
Drummond, J. J. (1997). Noema. In L. Embree (Ed.), Encyclopedia of phenom-
enology (pp. 494–499). Springer.
Evans, G. (1982). The varieties of reference. Clarendon Press.
Fodor, J. (1968). Psychological explanation. Random House.
Fodor, J. (1975). Language of thought. Harvard University Press.
Fodor, J. (1980). Methodological solipsism considered as a research strategy in
cognitive psychology. The Behavioral and Brain Sciences, 3(1), 63–73.
Fodor, J. (1983). The modularity of mind. MIT Press.
Føllesdal, D. (1969). Husserl’s notion of noema. The Journal of Philosophy,
66(20), 680–687. doi:10.2307/2024451
Føllesdal, D. (1988). Husserl on evidence and justification. In R. Sokolowski
(Ed.), Edmund Husserl and the phenomenological tradition (pp. 107–129). The
Catholic University of America Press.
Gallagher, S., & Brøsted Sørensen, J. (2006). Experimenting with phenomenology.
Consciousness and Cognition, 15(1), 119–134. doi:10.1016/j.concog.2005.03.002
Gallagher, S., & Zahavi, D. (2012). The phenomenological mind (2nd ed.).
Routledge.
Gurwitsch, A. (2009). The structure of perceptual noema (J. Garcia-Gómez,
Trans.). In J. Garcia-Gómez (Ed.), The collected works of Aron Gurwitsch
(1901–1973) (Vol. 1, pp. 185–307). Springer.
Hohwy, J. (2013). The predictive mind. Oxford University Press.
Husserl, E. (1964). The idea of phenomenology (W. P. Alston, Trans.). Martinus
Nijhoff.
Husserl, E. (1969). Formal and transcendental logic (D. Cairns, Trans.). Martinus
Nijhoff.
Husserl, E. (1971). “Phenomenology” Edmund Husserl’s article for the
Encyclopaedia Britannica (1927): New complete translation by Richard E.
Palmer. Journal of the British Society for Phenomenology, 2(2), 77–90.
Husserl, E. (1973). Experience and judgment: Investigations in a genealogy of
logic (J. S. Churchill & K. Ameriks, Trans.). Routledge & Kegan Paul.
Husserl, E. (1977). Phenomenological psychology (J. Scanlon, Trans.). Martinus Nijhoff.
Husserl, E. (1982). Ideas pertaining to a pure phenomenology and to a phenom-
enological philosophy; First book: General introduction to a pure phenomenol-
ogy (F. Kersten, Trans.). Springer.
Phenomenology and Functionalism 165
Husserl, E. (1989a). Ideas pertaining to a pure phenomenology and to a phenom-
enological philosophy; Second book: Studies in the phenomenology of consti-
tution (R. Rojcewicz & A. Schuwer, Trans.). Springer.
Husserl, E. (1989b). The crisis of European sciences and transcendental phenom-
enology: An introduction to phenomenological philosophy (D. Carr, Trans.).
Northwestern University Press.
Husserl, E. (1991). On the phenomenology of the consciousness of internal time
(1893–1917) (J. B. Brough, Trans.). Springer.
Husserl, E. (1997). Thing and space: Lectures of 1907 (R. Rojcewicz, Trans.).
Kluwer.
Husserl, E. (2001a). Analyses concerning passive and active synthesis: Lectures
on transcendental logic (A. J. Steinbock, Trans.). Springer.
Husserl, E. (2001b). Logical investigations (Vol. 2; J. N. Findlay, Trans.).
Routledge.
Husserl, E. (2002). Philosophy as rigorous science (M. Brainard, Trans.). New
Yearbook for Phenomenology and Phenomenological Philosophy, 2, 249–295.
Husserl, E. (2012). Philosophy of arithmetic: Psychological and logical investiga-
tions with supplementary texts from 1887–1901 (Vol. 10). Springer Science &
Business Media.
Jacobs, L. (2012). From chemotaxis to the cognitive map. Proceedings of the
National Academy of Sciences, 109(Suppl. 1), 10693–10700.
Jacobs, L., & Schenk, F. (2003). Unpacking the cognitive map: The parallel map
theory of hippocampal function. Psychological Review, 110(2), 285–315.
Kockelmans, J. J. (1987). HusserI’s original view on phenomenological psychol-
ogy. In J. J. Kockelmans (Ed.), Phenomenological psychology: The Dutch
school (pp. 3–29). Martinus Nijhoff Publishers.
Kortooms, T. (2002). Phenomenology of time. Springer.
Landgrebe, L. (1981). The phenomenology of Edmund Husserl. Cornell
University Press.
Livingston, P. (2005). Functionalism and logical analysis. In D. W. Smith & A.
L. Thomasson (Eds.), Phenomenology and philosophy of mind (pp. 19–40).
Oxford University Press.
Madary, M. (2017). Visual phenomenology. MIT Press.
McIntyre, R. (1986). Husserl and the representational theory of mind. Topoi,
5(2), 101–113. doi:10.1007/BF00139224
Moran, D. (2013). ‘Let’s look at it objectively’: Why phenomenology cannot be
naturalized. Royal Institute of Philosophy Supplements, 72, 89–115.
Neisser, J. (2015). The science of subjectivity. Palgrave Macmillan.
Panksepp, J., & Biven, L. (2012). The archaeology of mind: Neuroevolutionary
origins of human emotion. W. W. Norton & Company.
Petitmengin, C., Remillieux, A., Cahour, B., & Carter-Thomas, S. (2013). A gap in
Nisbett and Wilson’s findings? A first-person access to our cognitive processes.
Consciousness and Cognition, 22(2), 654–669.
Petitmengin, C., Remillieux, A., & Valenzuela-Moguillansky, C. (2019). Discovering
the structures of lived experience: Towards a micro-phenomenological analysis
method. Phenomenology and the Cognitive Sciences, 18(4), 691–730.
Piekarski, M. (2017). Commentary: Brain, mind, world: Predictive coding,
neo-Kantianism, and transcendental idealism. Frontiers in Psychology, 8.
doi:10.3389/fpsyg.2017.02077
166 Phenomenology and Mechanism
Piccinini, G. (2010). The mind as neural software? Understanding functional-
ism, computationalism, and computational functionalism. Philosophy and
Phenomenological Research, 81(2), 269–311.
Piccinini, G., & Craver, C. (2011). Integrating psychology and neuroscience:
Functional analyses as mechanism sketches. Synthese, 183(3), 283–311.
Płotka, W. (2021). A commentary on Pokropski’s functionalist reading of
Husserlian phenomenology. Frontiers in Psychology, 12. doi:10.3389/
fpsyg.2021.614842
Pokropski, M. (2015). Affectivity and time: Towards a phenomenology of
embodied time-consciousness. Studies in Logic, Grammar and Rhetoric, 41(1),
161–172.
Pokropski, M. (2020). Phenomenology and functional analysis: Afunctionalist
reading of Husserlian phenomenology. Phenomenology and the Cognitive
Sciences, 19, 869–889.
Putnam, H. (1960). Minds and machines. In S. Hook (Ed.), Dimensions of mind
(pp. 138–164). New York University Press.
Reynolds, J. (2017). Phenomenology, naturalism and science: A hybrid and hereti-
cal proposal. Routledge.
Rodemeyer, L. (2006). Intersubjective temporality: It’s about time. Springer.
Roy, J.-M., Petitot J., Pachoud, B., & Varela F. J. (1999). Beyond the gap: An intro-
duction to naturalizing phenomenology. In J. Petitot, F. J. Varela, B. Pachoud,
& J.-M. Roy (Eds.), Naturalizing phenomenology: Issues in contemporary phe-
nomenology and cognitive science (pp. 1–80). Stanford University Press.
Sowa, R. (2012). Eidetics and its methodology (E. A. Behnke, Trans.). In S. Luft
& S. Overgaard (Eds.), The Routledge companion to phenomenology (pp.
254–265). Routledge.
Thinès, G. (1977). Phenomenology and the science of behaviour. George Allen
& Unwin.
van Es, T. (2020). Living models or life modelled? On the use of models in
the free energy principle. Adaptive Behavior. Advanced online publication.
doi:10.1177/1059712320918678
Welton, D. (1982). Husserl’s genetic phenomenology of perception. Research in
Phenomenology, 12(1), 59–83.
Zahavi, D. (2003). Husserl’s phenomenology. Stanford University Press.
Zahavi, D. (2017). Brain, mind, world: Predictive coding, neo-Kantianism, and
transcendental idealism. Husserl Studies, 34(2–4), 47–61.
5 Phenomenology and Dynamical
Modeling
5.1 Introduction
In this chapter, I consider whether phenomenology could provide dynam-
ical constraints on the space of possible mechanisms and thus contrib-
ute to integrative dynamical-mechanistic explanations. Dynamical
constraints concern the dynamics of the phenomenon in question, for
instance, that it is oscillatory or consists of three subsequent phases. The
chapter is structured as follows: First, I discuss the relation between the
dynamical approach and the mechanistic model of explanation in the
context of the integration of research fields in cognitive science. I argue
that these frameworks are not opposed to one another but rather are
complementary and that dynamical-mechanistic explanation is the new
gold standard of explanation in cognitive science. Second, I consider
applications of phenomenology in cognitive science in which the dynami-
cal nature of the studied phenomena is emphasized, namely neuro- and
micro-phenomenology. In particular, I discuss neurophenomenological
studies of epilepsy and confront their results with dynamical modeling
of epileptic seizures in neuroscience. I argue that neurophenomenology
provides dynamical constraints that are mainly conceptual and thus too
weak to contribute to dynamical-mechanistic models. I conclude with
a discussion of how to revise neurophenomenology in order to provide
stronger constraints. In particular, I argue for formalizing pheno-dynam-
ical models.
DOI: 10.4324/9781003035367-8
168 Phenomenology and Mechanism
mechanisms—these are the two key explanatory strategies of the mecha-
nistic framework. On the one hand, some dynamicists (e.g., Lamb &
Chemero, 2014) argue that cognitive systems are non-decomposable;
thus, they cannot supplement the mechanistic framework, which depends
on assuming the target system’s decomposability or near-decomposabil-
ity. On the other hand, mechanists (e.g., Bechtel & Richardson, 2010)
argue that decomposability, as well as localization, are heuristic strate-
gies the results of which can be revised or falsified in the research pro-
cess. It does not follow, however, that the mechanistic approach is wrong.
For example, an initial assumption about the direct localization of visual
processing in the occipital lobe was changed in the course of research
to a composition of cognitive subfunctions localized in distributed areas
of the brain. The simple heuristic of direct localization failed and was
replaced by complex localization. But this model of visual processing is
still mechanistic (e.g., Bechtel, 2008). We can push this idea further and
argue that decomposition and localization strategies may fail in general
and leave us with cognitive models the functions of which are nonlocaliz-
able in the parts of the brain but that are nevertheless still explanatorily
useful models (Stinson, 2016).
Generally speaking, decomposability should not be understood as an
ontological category but as an epistemological one. We can decompose
a target system in many ways depending on our explanatory interests
and knowledge. Furthermore, there are different degrees of decompos-
ability, and, accordingly, a system may be decomposable, nearly decom-
posable, minimally decomposable, or non-decomposable (Simon, 1962;
Wimsatt, 1986). The decomposability and localization heuristics fail
in the case of systems that do not have parts or operations that can
be distinguished. Connectionist networks are good examples, as their
behavior emerges from the organization of relatively simple but highly
integrated elements (Bechtel & Richardson, 2010, p. 228). Network
models are not classical mechanistic models because they are non-
decomposable, yet they are in some sense mechanistic because the sys-
tem’s behavior is produced by its components and activities within the
system. What is crucial in network systems is not the contribution of
individual parts but their organization.
The heuristics of decomposition and localization may also fail when
parts and operations interact with each other in a nonlinear fashion and
when nonlinear interactions appear between parts and the behavior of
the whole system; then the behavior of the parts is system dependent and
may be called emergent. But as Bechtel and Richardson (2010) argue,
“when these two conditions are met, the systemic behavior is reason-
ably counted as emergent, even though it is fully explicable mechanisti-
cally” (p. xlvi). Such mechanistic explanation is different from the classic
picture, in which we identify and empirically study the real parts and
operations of a mechanism; it mainly relies on computational modeling.
Phenomenology and Dynamical Modeling 169
Furthermore, to capture the emergent character of the studied behavior,
such a model often involves formal and conceptual tools from dynamical
systems theory; thus, it is called a dynamical-mechanistic explanation.
According to Bechtel and Abrahamsen (2010), such explanations may be
built on research on circadian rhythms, whereby our understanding of
the underlying mechanisms was extended by computational dynamical
modeling. The variables of the computational models captured key prop-
erties of parts, their nonlinear operations, and the complex organization
of the mechanism; simulations have shown that these models are able to
produce the target dynamical phenomenon.
Another example of this kind of dynamical-mechanistic explanation
comes from network neuroscience (M. Silberstein & Chemero, 2013;
Zednik, 2019). Network neuroscience is a research field investigating how
the brain coordinates and integrates distributed processes, which often
vary in level of organization, localization, and timescale (for an overview,
see Sporns, 2011). According to Michael Silberstein and Chemero (2013),
“most of this coordination represents patterns of spontaneous, self-orga-
nizing, macroscopic spatiotemporal patterns, which resemble the on-the-
fly functional networks recruited during activity” (p. 962). As they argue,
most models which account for these self-organizing networks share
several key generalizations, including that in such network processes,
the temporal dimension is essential, and thus, they should be viewed as
dynamically evolving graphs. In this approach, which draws on graph
theory and dynamical systems theory, the basic units of explanation are
brain multiscale networks and their distributed connections, which are
represented in graphical form, having specific topological properties. For
example, it is argued that from among the various network topologies,
the so-called small-world networks are especially interesting, as such net-
works can be found in the brain (Sporns, 2011). How do we explain
these network processes? As Silberstein and Chemero claim, explana-
tions in network neuroscience are pluralistic and draw on various frame-
works including mechanistic, dynamical, and statistical-causal ones, each
of which contributes to network simulations. But the main explanatory
work is done by the graph models:
Importantly, predictions are not the only thing to be gained from network
simulations; thus, one cannot make the predictivism objection often formu-
lated against covering law and dynamical explanations (see Chapter 3 for
170 Phenomenology and Mechanism
a discussion). On the contrary, they allow us to understand why a studied
network exhibits certain properties in virtue of its dynamical and graphical
properties.
So, the mechanistic explanatory strategies of localization and decom-
position may fail, but that does not rule out the plausibility of the mecha-
nistic account. Nor is it true that all dynamical cognitive systems are
always non-decomposable. They can be decomposed in numerous ways,
but the key to doing it right is capturing how a system’s components are
dynamically coupled, that is, how they interact with each other and how
these interactions contribute to the behavior of the system. An example
of a decomposable dynamical system is the already mentioned percep-
tual agent model proposed by Beer (2000, 2003; see Chapter 3). The
two components of the system are described in the model by a sepa-
rate list of equations (16 equations in total). The fact that these compo-
nents are coupled is expressed in the equations by variables describing
one component being parameters for the other components, and vice
versa. Beer’s model shares features of both dynamical and mechanistic
explanations. A further example is related to the decomposability of eco-
logical systems. As Sabrina Golonka and Andrew Wilson (2019) argue,
the ecological approach does not have to be limited to purely dynami-
cal explanations—it can support mechanistic ones too. Accordingly, “the
ecological approach grounds psychological explanations in the types of
parts and operations that are amenable to dynamic causal mechanistic
models” (Golonka & Wilson, 2019, p. 679). The key to decomposing the
target cognitive system in an ecologically meaningful way is to choose
the relevant scale of decomposition. For Golonka and Wilson, this scale
is defined by ecological information, that is, kinematic patterns in ambi-
ent arrays which specify dynamical properties. In short, an organism uses
these kinematic patterns to coordinate action and perceive the properties
of an environment, such as affordances, in such a way that the “eco-
logical information specifying task dynamics constrains action coordi-
nation and control to the space of task-relevant solutions” (Golonka &
Wilson, 2019, p. 683). To put it differently, a cognitive system’s degrees of
freedom are limited by ecological information in order to shape actions
and achieve the goal in an optimal fashion. Furthermore, Golonka and
Wilson suggest that ecological information may also constrain neural
coordination processes. Taking into account ecological information,
they reconsider the model of rhythmic movement proposed by Geoffrey
Bingham (2004) as an example of a dynamical-mechanistic model. The
model differs from the famous Haken–Kelso–Bunz (HKB) model (Haken
et al., 1985; see Chapter 3), which was purely dynamical, in that it allows
one to make predictions about physical components of underlying mech-
anisms. The refined model is based on the assumption, supported by
empirical research, that rhythmic coordination depends on the percep-
tion of relative phase:
Phenomenology and Dynamical Modeling 171
The information variable must be kinematic, it must specify relative
phase, it must be made of state variables (to preserve autonomy), it
must break the symmetry between relative phases in a manner that
matches the phenomena catalogued above, and it must be detectable
by both vision and proprioception.
(Golonka & Wilson, 2019, p. 688)
These examples show, on the one hand, that mechanism is not blindly
committed to decomposability and localization and, on the other, that
the dynamical approach is not limited to non-decomposable and emer-
gent phenomena. If that is the case, then the mechanistic framework
and the dynamical framework should be perceived as complementary
rather than exclusive (e.g., Kaplan & Bechtel, 2011; M. Silberstein &
Chemero, 2013; Zednik, 2011). For some (Golonka & Wilson 2019),
this “hybrid” dynamical-mechanistic model is the new “gold standard” of
explanation for cognitive science. Such a model combines the advantages
of the two approaches; it demonstrates the organization and composition
of the mechanism responsible for the target phenomenon and explores
the mechanism’s functioning in simulations. In particular, according
to Bechtel and Abrahamsen (2010), dynamical modeling may contrib-
ute to our understanding of a given mechanism in several ways: (1) by
demonstrating that the modeled mechanism is able to exhibit the target
phenomenon; (2) providing a larger space of parameter values and thus
allowing one to study the behavior of the mechanism, which would be
difficult to do empirically; (3) helping identify which parts and operations
are crucial for producing the target phenomenon; (4) helping explain the
mechanism’s malfunctions; (5) revealing the conditions under which the
mechanism may be coupled with another mechanism, thereby exhibit-
ing collective coordinated behavior; (6) allowing one to explore possible
alterations in relations between multiple mechanisms. From among these,
the last two are especially closely tied to dynamical modeling (Golonka
& Wilson, 2019, p. 677) and may contribute to studies of extended and
distributed systems (Zednik, 2011).
Now, taking into account that the dynamical and mechanistic
approaches may be understood as complementary, we can consider
whether phenomenology, especially where it is read through the lens of
dynamical systems theory, might provide dynamical constraints and thus
contribute to dynamical-mechanistic models.
Figure 5.1 P
hase space of the (A) nonepileptic and (B) epileptic brain model.
Note that in the case of the nonepileptic brain the distance between
the seizure attractor and normal attractor (concentrated basin of
attraction) is large; thus, the transition does not occur. In the case of
the epileptic brain, the distance is smaller and fluctuations of parame-
ters may lead to phase transition. Adapted from “Dynamical Diseases
of Brain Systems: Different Routes to Epileptic Seizures,” by Lopes
da Silva et al., 2003, IEEE Transactions on Biomedical Engineering,
50(5), p. 542. Copyright © 2003 by IEEE.
180 Phenomenology and Mechanism
stability of the network caused by endogenous factors such as hormones
(Lopes da Silva et al., 2003, p. 543–544).
The models proposed by Lopes da Silva et al. (2003) and Piotr
Suffczyński et al. (2004) are further examples of dynamical-mechanistic
models that extend our understanding of the mechanisms responsible for
the transition of neural dynamics leading to seizures. These simulations
provide a means to study a greater range of parameters than possible
through empirical studies and show nonlinear interaction and coordina-
tion between neural mechanisms involved in epilepsy. These models are
also a good example of how dynamical models can tell us something
about mechanisms’ parts and their organization; that is, they include
parameters and values related to actual components of neural mecha-
nisms, such as membrane (e.g., Ca2+ currents), synaptic (e.g., thresh-
old of GABAA), and network properties (e.g., thalamocortical system;
Suffczyński et al., 2004, p. 483).
It is plausible that some of the model’s values could stand for experi-
ential states, such as prodromal sensations in the preictal phase, that is,
the progressive negative symptoms reported by patients in Petitmengin’s
(2007) study, or cognitive deficits in the postictal phase (Pottkämper et
al., 2020). Such modification of the model would, however, require quan-
tification of experiential categories.
Notes
1 The notion of bifurcation or phase transition has also been applied to other
types of cognitive processes (see, e.g., Spivey et al., 2009).
2 As we have seen, an aura is also a symptom of an epileptic seizure. According to
some models, epileptic seizures and CSD responsible for the migraine aura have
a common underlying mechanism related to the ion concentration in neuronal
membranes and cells’ external microenvironment (e.g., Ullah et al., 2015).
References
Bauer P. R., Thijs, R. D., Lamberts, R. J., Velis, D. N., Visser, G. H., Tolner, E. A.,
Sander, J. W., Lopes da Silva, F. H., & Kalitzin, S. N. (2017). Dynamics of con-
vulsive seizure termination and postictal generalized EEG suppression. Brain,
140(3), 655–668.
Bechtel, W. (2008). Mental mechanisms: Philosophical perspectives on cognitive
neuroscience. Taylor & Francis Group.
Bechtel, W., & Abrahamsen, A. (2010). Dynamic mechanistic explanation:
Computational modeling of circadian rhythms as an exemplar for cognitive
science. Studies in History and Philosophy of Science Part A, 41(3), 321–333.
Bechtel, W., & Richardson, R. C. (2010). Discovering complexity: Decomposition
and localization as strategies in scientific research. MIT Press.
Beer, R. D. (2000). Dynamical approaches to cognitive science. Trends in Cognitive
Sciences, 4, 91–99.
Beer, R. D. (2003). The dynamics of active categorical perception in an evolved
model agent. Adaptive Behavior, 11(4), 209–243.
Phenomenology and Dynamical Modeling 189
Bingham, G. P. (2004). A perceptually driven dynamical model of bimanual rhyth-
mic movement (and phase perception). Ecological Psychology, 16(1), 45–53.
Bockelman, P., Reinerman-Jones, L., and Gallagher, S. (2013). Methodological
lessons in neurophenomenology: Review of a baseline study and recom-
mendations for research approaches. Frontiers in Human Neuroscience, 7.
doi:10.3389/fnhum.2013.00608
Chalmers, D. J. (1996). The conscious mind: In search of a fundamental theory.
Oxford University Press.
Ciechanowski, L. (2017). Has the philosopher’s stone of the interaction between
first- and third-person data finally been found? Constructivist Foundations,
12(2), 203–205.
Craver, C. F. (2007). Explaining the brain: Mechanisms and the mosaic unity of
neuroscience. Clarendon Press.
Dahlem, M. A., & Müller, S. C. (2003). Migraine aura dynamics after reverse
retinotopic mapping of weak excitation waves in the primary visual cortex.
Biological Cybernetics, 88(6), 419–424.
Davis, L. E., King, M. K., & Schultz, J. L. (2005). Fundamentals of neurologic
disease. Demos Medical Publishing.
Foldvary-Schaefer, N., & Unnwongse, K. (2011). Localizing and lateralizing fea-
tures of auras and seizures. Epilepsy & Behavior, 20(2), 160–166.
Golonka, S., & Wilson, A. D. (2019). Ecological mechanisms in cognitive science.
Theory & Psychology, 29(5), 676–696. doi:10.1177/0959354319877686
Haken, H., Kelso, J. A. S., & Bunz, H. (1985). A theoretical model of phase transi-
tions in human hand movements. Biological Cybernetics, 51, 347–442.
Høffding, S., & Martiny, K. (2016). Framing a phenomenological interview:
What, why and how. Phenomenology and the Cognitive Sciences, 15, 539–564.
doi:10.1007/s11097-015-9433-z
Jackson, J. H. (1868). Case of convulsive attacks arrested by stopping the aura.
The Lancet, 1, 618–619.
Jedličková, L., & Müller, M. (2019). Problems and hopes of neurophenomenol-
ogy and first-person neuroscience. e-Rhizome, 1(2): 138–159. doi:10.5507/
rh.2019.008
Kałwak, W., Reuter, M., Łukowska, M., Majchrowicz, B., & Wierzchoń, M. (2018).
Guidelines for quantitative and qualitative studies of sensory substitution expe-
rience. Adaptive Behavior, 26(3), 111–127. doi:10.1177/1059712318771690
Kaplan, D. M., & Bechtel, W. (2011). Dynamical models: An alternative or
complement to mechanistic explanations? Topics in Cognitive Science, 3(2),
438–444.
Lamb, M., & Chemero, A. (2014). Structure and application of dynamical mod-
els in cognitive science. Proceedings of the Annual Meeting of the Cognitive
Science Society, 36, 809–814.
Lashley, K. S. (1941). Patterns of cerebral integration indicated by the scotomas of
migraine. Archives of Neurology and Psychiatry, 46(2), 331–339. doi:10.1001/
archneurpsyc.1941.02280200137007
le van Quyen, M., Martinerie, J., Navarro, V., Boon, P., D’Havé, M., Adam, C.,
Renault, B., Varela, F., & Baulac, M. (2001). Anticipation of epileptic seizures
from standard EEG recordings. The Lancet, 357(9251), 183–188.
le van Quyen, M. (2003). Disentangling the dynamic core: A research program
for a neurodynamics at the large-scale. Biological Research, 36(1), 67–88.
190 Phenomenology and Mechanism
Liu, Y., Guo, X. M., Wu, X., Li, P., & Wang, W. W. (2017). Clinical analysis of
partial epilepsy with auras. Chinese Medical Journal (Eng.), 130(3), 318–322.
doi:10.4103/0366-6999.198918
Lopes da Silva, F. H., Blanes, W., Kalitzin, S. N., Parra, J., Suffczyński, P., & Velis,
D. N. (2003). Dynamical diseases of brain systems: Different routes to epileptic
seizures. IEEE Transactions on Biomedical Engineering, 50(5), 540–548.
Lutz, A. (2002). Toward a neurophenomenology as an account of generative passages:
A first empirical case study. Phenomenology and the Cognitive Sciences, 1, 133–167.
Lutz, A. (2007). Neurophenomenology and the study of self-consciousness.
Consciousness and Cognition, 16(3), 765–767.
Lutz, A., Lachaux, J. P., Martinerie, J., & Varela, F. J. (2002). Guiding the study
of brain dynamics by using first-person data: Synchrony patterns correlate
with ongoing conscious states during a simple visual task. Proceedings of the
National Academy of Sciences, 99(3), 1586–1591.
Lutz, A., & Thompson, E. (2003). Neurophenomenology integrating subjective
experience and brain dynamics in the neuroscience of consciousness. Journal of
Consciousness Studies, 10(9–10), 31–52.
Milner, P. M. (1958). Note on a possible correspondence between the scotomas
of migraine and spreading depression of Leão. Electroencephalography and
Clinical Neurophysiology, 10(4), 705.
Mormann, F., Kreuz, T., Andrzejak, R. G., David, P., Lehnertz, K., & Elger, C.
E. (2003). Epileptic seizures are preceded by a decrease in synchronization.
Epilepsy Research, 53(3), 173–185.
Petitmengin, C. (2006). Describing one’s subjective experience in the second per-
son: An interview method for the science of consciousness. Phenomenology and
the Cognitive Sciences, 5(3–4), 229–269. doi:10.1007/s11097–006–9022–2
Petitmengin, C. (2010). A neurophenomenological study of epileptic seizure
anticipation. In S. Gallagher & D. Schmicking (Eds.), Handbook of phenom-
enology and cognitive science (pp. 471–499). Springer.
Petitmengin, C., Baulac, M., & Navarro, V. (2006). Seizure anticipation: Are neu-
rophenomenological approaches able to detect preictal symptoms? Epilepsy &
Behavior, 9(2), 298–306.
Petitmengin, C., Navarro, V., & le van Quyen, M. (2007). Anticipating seizure:
Pre-reflective experience at the center of neuro-phenomenology. Consciousness
and Cognition, 16(3), 746–764.
Petitmengin, C., & Lachaux, J.-P. (2013). Microcognitive science: Bridging expe-
riential and neuronal microdynamics. Frontiers in Human Neuroscience, 7.
doi:10.3389/fnhum.2013.00617
Petitmengin, C., Remillieux, A., Cahour, B., & Carter-Thomas, S. (2013). A gap in
Nisbett and Wilson’s findings? A first-person access to our cognitive processes.
Consciousness and Cognition, 22(2), 654–669.
Petitmengin, C., Remillieux, A., & Valenzuela-Moguillansky, C. (2019). Discovering
the structures of lived experience: Towards a micro-phenomenological analysis
method. Phenomenology and the Cognitive Sciences, 18(4), 691–730.
Pottkämper, J. C. M., Hofmeijer, J., van Waarde, J. A., & van Putten, M. J. A. M.
(2020). The postictal state—What do we know? Epilepsia, 61(6), 1045–1061.
Przyrembel, M., Vrticka, P., Engert, V., & Singer, T. (2019). Loving-kindness med-
itation—A queen of hearts?: A physio-phenomenological investigation on the
variety of experience. Journal of Consciousness Studies, 26(7–8), 95–129.
Phenomenology and Dynamical Modeling 191
Roskies, A. (1999). The binding problem: Special issue. Neuron, 24, 7–125.
Schott, G. D. (2007). Exploring the visual hallucinations of migraine aura: The
tacit contribution of illustration. Brain, 130(6), 1690–1703.
Silberstein, M., & Chemero, A. (2013). Constraints on localization and decom-
position as explanatory strategies in the biological sciences. Philosophy of
Science, 80(5), 958–970.
Silberstein, S. D. (2004). Migraine. The Lancet, 363(9406), 381–391.
Simon, H. A. (1962). The architecture of complexity. Proceedings of the American
Philosophical Society, 106(6), 467–482.
Smith, J. A., Flowers, P., & Larkin, M. (2009). Interpretative phenomenological
analysis: Theory, method and research. Sage.
Spivey, M. J., Anderson, S. E., & Dale, R. (2009). The phase transition in human
cognition. New Mathematics and Natural Computation, 5(1), 197–220.
Sporns, O. (2011). Networks of the brain. MIT Press.
Stinson, C. (2016). Mechanisms in psychology: Ripping nature at its seams.
Synthese, 193, 1585–1614.
Suffczyński, P., Kalitzin, S., & Lopes Da Silva, F. H. (2004). Dynamics of non-
convulsive epileptic phenomena modeled by a bistable neuronal network.
Neuroscience, 126(2), 467–484.
Timmermann, C., Roseman, L., Schartner, M., Milliere, R., Williams, L. T.,
Erritzoe, D., & Turton, S. (2019). Neural correlates of the DMT experience
assessed with multivariate EEG. Scientific Reports, 9(1), 1–13.
Ullah, G., Wei, Y., Dahlem, M. A., Wechselberger, M., & Schiff, S. J. (2015). The
role of cell volume in the dynamics of seizure, spreading depression, and anoxic
depolarization. PLOS Computational Biology, 11(8). doi:10.1371/journal.
pcbi.1004414
Wimsatt, W. C. (1986). Forms of aggregativity. In B. Donagan, A. N. Perovich, Jr.,
& M. V. Wedin (Eds.), Human nature and natural knowledge: Essays presented
to Marjorie Grene on the occasion of her seventy-fifth birthday (pp. 259–291).
Springer.
Valenzuela-Moguillansky, C., O’Regan, J. K., & Petitmengin, C. (2013). Exploring
the subjective experience of the “rubber hand” illusion. Frontiers in Human
Neuroscience, 23. doi.org/10.3389/fnhum.2013.00659
van Gelder, T. (1999). Wooden iron? Husserlian phenomenology meets cognitive
science. In J. Petitot, F. J. Varela, B. Pachoud, & J.-M. Roy (Eds.), Naturalizing
phenomenology: Issues in contemporary phenomenology and cognitive science
(pp. 245–265). Stanford University Press.
van Gelder, T., & Port, R. F. (1995). It’s about time: An overview of the dynamical
approach to cognition. In R. F. Port & T. van Gelder (Eds.), Mind as motion:
Explorations in the dynamics of cognition (pp. 1–43). MIT Press.
Varela, F. (1996). Neurophenomenology: A methodological remedy for the hard
problem. Journal of Consciousness Studies, 3(4), 330–349.
Varela, F. (1999). The specious present: A neurophenomenology of time conscious-
ness. In J. Petitot, F. J. Varela, B. Pachoud, & J.-M. Roy (Eds.), Naturalizing
phenomenology: Issues in contemporary phenomenology and cognitive science
(pp. 266–329). Stanford University Press.
Varela, F., Lachaux, J. P., Rodriguez, E., & Martinerie, J. (2001). The brain-
web: Phase synchronization and large-scale integration. Nature Reviews
Neuroscience, 2, 229–239. doi:10.1038/35067550
192 Phenomenology and Mechanism
Vermersch, P. (2012). Explicitation et phénoménologie. PUF.
Zednik, C. (2011). The nature of dynamical explanation. Philosophy of Science,
78(2), 238–263.
Zednik, C. (2019). Models and mechanisms in network neuroscience.
Philosophical Psychology, 32(1), 23–51.
6 Conclusion
Toward Methodologically Guided
Mutual Constraints
DOI: 10.4324/9781003035367-9
194 Phenomenology and Mechanism
certain limitations, and each of them is incapable of providing a complete
explanation of complex mental phenomena. Mechanistic integration of
cognitive science provides a remedy (Craver, 2007), according to which
various research fields contribute to a complete multilevel explanation
by providing constraints on the space of mechanisms. In this context,
the type of explanation phenomenology offers, if any, was also consid-
ered. The conclusion was that phenomenology does not provide genuine
explanations but rather gives us a constitutive understanding of mental
phenomena. It was hypothesized that phenomenological understanding
can provide constraints and thus contribute to multilevel mechanistic
explanations. In particular, two types of constraints were considered,
namely, functional and dynamical constraints.
In searching for functional constraints, we went back to Husserlian
phenomenology and his idea of functional phenomenology (Husserl,
1982), which introduced the notion of constitutive functions of con-
sciousness, that is, functions involved in the production of intentional
experience. It was argued that functional phenomenology relies on a
method of phenomenological decomposition, which shares similarities
with the analytical method of functional decomposition applied in psy-
chological explanations (Cummins, 1975). Much like functional analy-
sis, phenomenology can provide general characterizations of cognitive
functions constitutive for the target experience. By considering examples
of vision studies and a naturalistic account of the first-person perspec-
tive, it was shown that such a characterization can be further specified in
functional models, which, in turn, can constrain how-possibly models of
underlying mechanisms.
Considering phenomenological-dynamical constraints required reject-
ing the purported opposition between dynamical and mechanistic
approaches and instead thinking in terms of their integration. Dynamical-
mechanistic explanations address both the dynamics of the behavior in
question and the components of the underlying mechanisms (e.g., Bechtel
& Abrahamsen, 2010). Assessing the application of dynamical systems
theory in neurophenomenological studies (e.g., Lutz, 2002), in particu-
lar in studies of epilepsy (e.g., Petitmengin et al., 2007), I argued that
it is unsatisfying and limited to mere dynamic descriptions of the tar-
get phenomenon, and thus, it cannot genuinely constrain the space of
mechanisms.
The case of dynamical analyses of epilepsy providing stronger con-
straints on mechanisms was shown on the example of a dynamical-
mechanistic model of the epileptic brain (Lopes da Silva et al., 2003).
Discussing this model, I suggested that it could be modified in order to
include experiential parameters related to subjective experience of the
preictal or postictal phase. Such a phenomenological revision of the model
would be an example of mutual constraints between first-person analyses
of experience, third-person neuroscience, and dynamical modeling. The
Conclusion 195
history of studies concerning migraines and visual auras is also an illus-
trative example, one that suggests that analysis of first-person experience
may provide key information about the dynamics of the phenomenon in
question and thus lead to the discovery of underlying mechanisms.
Integrating phenomenology with the mechanistic framework can be
conceived of as a new approach to naturalization, which, as I argued,
is nonreductive and leaves integrated fields a certain dose of autonomy.
Generally speaking, mechanistic integration of cognitive science does not
offer one elegant, unified theory, but rather, it creates a patchwork of var-
ious research fields involved in the study of cognitive phenomena. Thus,
this way of naturalizing phenomenology does not assume epistemologi-
cal continuity between phenomenology and the natural sciences and does
not attempt to create a unified theory of cognition. Nor does it aim to
reduce phenomenology to other research fields, in particular to neuro-
biology. Such reductionism would run contrary to the ideas of explana-
tory pluralism, interfield integration, and the conviction that explanation
of mental phenomena requires the cooperation of multiple approaches
addressing various aspects (and levels) of the same target phenomenon.
In the proposed approach to naturalization, phenomenology is also not
merely applied to improve experimental design, as it is in front-loaded
phenomenology (e.g., Gallagher, 2010). In a nutshell, the idea of natu-
ralization through mechanistic integration has it that phenomenology
can contribute to building multilevel explanatory models by providing
intralevel constraints on the space of mechanisms, that is, by providing
information limiting the space of possibilities of mechanistic solutions.
As I argued, the character of the constraints that phenomenology can
provide is twofold: functional constraints concern constitutive functions
of consciousness involved in the production of experience and dynamic
constraints concern the dynamical landscape of experience.
6.1 Constraining Phenomenology
The earlier considerations mainly concerned theoretical and methodolog-
ical grounds for the integration of phenomenology with the mechanistic
framework, in particular how phenomenology can constrain explanatory
mechanistic models. But the relation between the mechanistic framework
of cognitive science and phenomenology is not one-directional. Thus it is
now important to consider what phenomenology gains from this integra-
tion, how it is constrained, and whether we can think of this relation as
“mutual enlightenment” (e.g., Gallagher, 2010).
To allow for this mutual relation between phenomenology and cog-
nitive science, we have to acknowledge that phenomenological claims
are not infallible. The infallibility of phenomenological theories was
often a part of the critique of phenomenology, stressing the opposition
between the phenomenological approach and the natural sciences, which
196 Phenomenology and Mechanism
is grounded in the ideas of falsifiability and corroboration. But it is a
critique of phenomenology that is unsupported. Husserl himself admit-
ted that phenomenological evidence is fallible, even if it is considered
apodictic. For example, as he writes in Formal and Transcendental Logic,
To put it differently, we can err in our eidetic intuitions about the nature
of experience, and these errors can be detected by confronting them with
the evidence of a new experience. Confirmation or falsification of eidetic
claims can also be achieved on other grounds. The phenomenological
description in question can be confirmed or falsified through intersubjec-
tive corroboration, that is, by comparing it with others’ descriptions of
the same type of experience (Gallagher & Zahavi, 2012) or by confront-
ing the description with empirical evidence (e.g., Gallagher & Brøsted
Sørensen, 2006). An interesting historical example of how a phenom-
enological theory was confirmed by empirical research is discussed by
Philipp Berghofer (2020). As he argues, phenomenological investigations
of illusory and hallucinatory experience lead Husserl to the key distinc-
tion between perceptual states and judgments. Accordingly, perceptual
experience representing an object O does not imply having a belief that
O. In other words, one can have a perceptual experience that is incon-
gruent with their judgments about the perceptual object. According to
Berghofer, this phenomenological distinction was later supported by
empirical research conducted by Vittorio Benussi, who was a member of
Meinong’s school of experimental psychology in Graz. Benussi studied
various optical illusions, such as the Müller-Lyer illusion, and how sub-
jects react to them. His conclusions were that optical illusions should be
considered inadequate representations rather than errors of judgment and
that perception, in general, is an experience of a perceptual object’s pres-
ence, which does not involve judgments, although it may lead to them.
Thus, Berghofer constates, experimental research led Benussi to conclu-
sions similar to those that can be found in Husserlian phenomenology.
Benussi provides a clear example of convergence between phenomeno-
logical claims and the results of experimental research. But as Gallagher
(2010) rightly remarks,
References
Bechtel, W., & Abrahamsen, A. (2010). Dynamic mechanistic explanation:
Computational modeling of circadian rhythms as an exemplar for cogni-
tive science. Studies in History and Philosophy of Science Part A, 41(3),
321–333.
Berghofer, P. (2020). Phenomenology and experimental psychology: On the pros-
pects and limitations of experimental research for a phenomenological episte-
mology. Journal of Transcendental Philosophy, 1(1), 85–108.
Craver, C. F. (2007). Explaining the brain: Mechanisms and the mosaic unity of
neuroscience. Clarendon Press.
Cummins, R. (1975). Functional analysis. The Journal of Philosophy, 72(20),
741–765.
Gallagher, S., & Brøsted Sørensen, J. (2006). Experimenting with phenom-
enology. Consciousness and Cognition, 15(1), 119–134. doi.org/10.1016/j.
concog.2005.03.002
Gallagher, S. (2010). Phenomenology and non-reductionist cognitive science. In
D. Schmicking & S. Gallagher (Eds.), Handbook of phenomenology and cogni-
tive science (pp. 21–34). Springer.
Gallagher, S., & Zahavi, D. (2012). The phenomenological mind (2nd ed.).
Routledge.
Husserl, E. (1969). Formal and transcendental logic (D. Cairns, Trans.). Martinus
Nijhoff.
Husserl, E. (1980). Phenomenology and the foundations of the sciences: Third
book of Ideas pertaining to a pure phenomenology and to a phenomenological
philosophy (T. E. Klein & W. E. Pohl, Trans.). Martinus Nijhoff.
Husserl, E. (1982). Ideas pertaining to a pure phenomenology and to a phenom-
enological philosophy; First book: General introduction to a pure phenomenol-
ogy (F. Kersten, Trans.). Springer.
Lopes da Silva, F. H., Blanes, W., Kalitzin, S. N., Parra, J., Suffczyński, P., & Velis,
D. N. (2003). Dynamical diseases of brain systems: Different routes to epileptic
seizures. IEEE Transactions on Biomedical Engineering, 50(5), 540–548.
200 Phenomenology and Mechanism
Lutz, A. (2002). Toward a neurophenomenology as an account of generative pas-
sages: A first empirical case study. Phenomenology and the Cognitive Sciences,
1(2), 133–167.
Petitmengin, C., Navarro, V., & le van Quyen, M. (2007). Anticipating seizure:
Pre-reflective experience at the center of neuro-phenomenology. Consciousness
and Cognition, 16(3), 746–764.
Piccinini, G., & Craver, C. (2011). Integrating psychology and neuroscience:
Functional analyses as mechanism sketches. Synthese, 183(3), 283–311.
Sowa, R. (2012). Eidetics and its methodology (E. A. Behnke, Trans.). In S.
Luft & S. Overgaard (Eds.), The Routledge companion to phenomenology
(pp. 254–265). Routledge.
Index