You are on page 1of 15

NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY

Published in: Journal of the Washington Academy of Sciences,


vol. 90, No. 2, Summer 2004, 45–60.

Although science in the modern sense is, strictly speaking, a creation of Greek
thinkers, it is becoming increasingly clear, with the gradual recovery of data from the ancient
Near East and China, that science has far earlier antecedents. The impulse to gather data and
to anticipate the future is at least as old as civilization itself, and must have had survival
value in pre-historic hunter-gatherer and early agricultural societies. The earliest forms of
scientific reasoning are found in the practice of divination.1 Ancient science of all periods
emphasized the empirical: numerous archives have been discovered that contain records of
diviners’ observations, whether from sheep entrails or planetary observations in
Mesopotamia, or from oracle bones in Shang China, or from physicians’ clinical observations
in Egypt. Some texts (much fewer) offer explanations, attempting to connect the sign with
the thing signified by it. The Hippocratic Collection2 of medical texts (beginning 5th C. BC)
is among the earliest that adopts a full-fledged empirical stance toward observation while
attempting to derive naturalistic causal explanations. An oft-cited example is the treatise “On
the Sacred Disease”3, the first recognizably scientific attempt to explain epilepsy. Seizures
are not due to the actions of gods or demons, the text explains, but are a perfectly reasonable
consequence of natural physiological factors. Occasionally we find texts in which the author
shows us how he derives the theory from the data. These texts are of special interest, since
they show stages in the development of scientific reasoning.

1 See, for example, Jean Bottéro, “Divination and the Scientific Spirit,” Chapter 8 of Mesopotamia: Writing,
Reasoning, and the Gods. Translated by Zainab Bahrani and Marc van de Mieroop. Chicago: University of
Chicago Press, 1992; and Wolfram von Soden, “Sumerian and Babylonian Science”, Chapter 11 of The Ancient
Orient: An Introduction to the Study of the Ancient Near East. Translated by Donald G. Schley. Grand Rapids,
MI: Eerdmans, 1994.
2 The historical Hippocrates dates to the 5th Century BC. However, most of the texts attributed to him were

probably written by his followers subsequent to that time. See Lloyd, G. E. R. “The Hippocratic Question.” In
Methods and Problems in Greek Science, 194-223. Cambridge: Cambridge University Press, 1991 (1975); and
Jouanna, Jacques. Hippocrates. Translated by M. B. DeBevoise. Baltimore, London: The Johns Hopkins
University Press, 1999 (1992).
3 De morbo sacro, “The Sacred Disease”, in Hippocrates. Hippocrates Volume II. Translated by W. H. S. Jones.

Edited by G. P. Goold. Vol. 148, Loeb Classical Library. Cambridge: Harvard University Press, 1923. Reprint,
1992, pp.129-183.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 1 of 15
The treatise On the Critical Days4 by the physician Galen of Pergamon (AD 129-
c.216), who was the most important follower of Hippocrates, is such a text. The naturalistic
explanation he offers for the recurring seizures in certain illnesses—the motion of the moon
and the influence of its phases—would not, of course, be considered valid today. But his
account is consistent with a scientific tradition of invoking celestial causes that extends back
to Aristotle (384-322 BC).5 Galen’s scientific method is the rational scaffolding around
which he builds a theory of the critical days. The critical days were one of the Hippocratic
rational schemes for prognosis. There are frequent references to them in the Hippocratic
literature, namely in the Prognostic6, the Aphorisms7, and the Epidemics8; however, these
treatises include no explanation of how to use them or why they were supposed to work.
(The Epidemics are among the first medical casebooks, and have long been admired, even
today, for their dispassionate and scientific style of observation).
The steps of Galen’s derivation of his theory can be fruitfully compared with modern
scientific method. This shows that there were examples of recognizably “modern” scientific
reasoning is much earlier than is usually thought. This surprising fact may be obscured by
the nature of Galen’s scientific explanation: his approach resembles astrology to a degree
perhaps unsettling to us. Yet, the outline of his reasoning is familiar, as I shall show
presently. The early affinity of science with divination should not worry us. Medicine (and
weather forecasting) retains traces of their ancient divinatory origins: the term prognosis
itself (“foreknowledge”) evokes a sense of the divinatory or prophetic—the physician and the

4 De diebus decretoriis. The Greek text of this treatise is found in C. G. Kühn’s edition of Galen, volume 9,
pages 769–941. (Throughout this article, I shall refer to this treatise with the following shorthand: for example,
K.9.843.2 refers to Kühn’s Galen edition, (volume 9), page 843, line 2.) I have prepared an English translation
from the Arabic version, included in my dissertation: Glen M. Cooper, Galen’s “Critical Days”: Greek
Medicine in Arabic. Edition, translation, and commentary of Hunayn ibn Ishaq’s Arabic translation of Galen’s
De diebus decretoriis; Ph.D. dissertation, Columbia University, 1999.
5 In Galen, this idea is not as astrological as it would later become, when the notion of celestial influence was

generalized to embrace all of the planets. For example, in the medieval practice of medicine, this scheme
assisted the physician in knowing when to bleed or test the patient’s urine.
6 Prognostic 20.1-16, in Hippocrates. Hippocrates Volume II. Translated by W. H. S. Jones. Edited by G. P.

Goold. Vol. 148, Loeb Classical Library. Cambridge: Harvard University Press, 1923. Reprint, 1992, pp.42-43.
7 Aphorisms 2.23.1-2 and 2.24.1-5, in Hippocrates. Hippocrates Volume IV. Translated by W. H. S. Jones.

Edited by G. P. Goold. Vol. 150, Loeb Classical Library. Cambridge: Harvard University Press, 1931. Reprint,
1992, pp.112-15.
8 Epidemics: references to the critical days occur throughout. Books I and III, the most important of all, are

found in: Hippocrates Volume I. Translated by W. H. S. Jones. Edited by G. P. Goold. Vol. 147, Loeb Classical
Library. Cambridge: Harvard University Press, 1923. Reprint, 1984.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 2 of 15
seer have a common heritage. In this article, I will show how Galen’s method anticipates the
familiar scientific method. In order to achieve this, I must first describe the critical days and
place them in their historical context within ancient Greek medicine.
Anciently, as now, the physician was expected to give a prognosis. The Hippocratic
treatise On Prognosis explained the ideal.9 According to that treatise, prognosis served
several practical purposes. The physician could determine whether a case is hopeless, and if
so, then by refusing to treat the patient he could avoid damaging his reputation. The
physician was expected to infer the present and prior conditions of the patient from direct
observation, without being told. Through a comprehensive prognosis, he could gain the
confidence of the patient, which is extremely important if the patient is to follow the doctor’s
orders. Furthermore, through subsequent prognoses, the physician could determine whether
the patient has followed his regimen, and thus be exonerated if the prescribed treatment fails.
Clearly, this is a more comprehensive notion of prognosis than the modern understanding.
Prognosis, therefore, was the most dramatic means by which to impress and attract potential
clients, through demonstrating one’s knowledge and his power of reason10. In fact, Galen
himself wrote a treatise detailing his most spectacular successes, probably with the intention
of spreading his reputation11.
Prognosis involved reading signs in the patient and comparing them with a prognostic
scheme or theory. The scheme of the critical days was one such theory, central to the
argument of the Prognostic. In a society where exploratory surgery was prohibited,12 the
physician had to rely on external indicators in order to infer the interior condition of the
patient’s body, to a much greater degree than today. The most conspicuous and useful of
these indicators were bodily effusions, such as urine, excrement, perspiration, etc. By

9 See several articles in: Edelstein, Ludwig. Ancient Medicine: Selected Papers of Ludwig Edelstein. Baltimore
and London: The Johns Hopkins University Press, 1967. Specifically, “Hippocratic Prognosis”, pp.65-85, and
“The Hippocratic Physician”, pp.87-131.
10 See Barton, Tamsyn S. Power and Knowledge: Astrology, Physiognomics, and Medicine under the Roman

Empire. Ann Arbor, Michigan: The University of Michigan Press, 1994.


11 Nutton, Vivian. Galen on Prognosis: Edition, Translation, and Commentary. Berlin: Akademie-Verlag, 1979.

12 Cf. “The Oath”, in Hippocrates. Hippocrates Volume I. Translated by W. H. S. Jones. Edited by G. P. Goold.

Vol. 147, Loeb Classical Library. Cambridge: Harvard University Press, 1923. Reprint, 1984, pp. 291-301. Cf.
also “The Hippocratic Oath: Text, Translation, and Interpretation”, in: Edelstein, Ludwig. Ancient Medicine:
Selected Papers of Ludwig Edelstein. Baltimore and London: The Johns Hopkins University Press, 1967, pp.3-
63.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 3 of 15
themselves, the qualities of the patient’s excrement or urine were simple facts, devoid of
implications for the outcome of the illness. Nevertheless, within the context of the theory of
the critical days, such facts became prognostic indicators. In his treatise On Crises,13 Galen
describes in detail how specific effusions and their characteristics can be used to infer the
conditions of corresponding regions of the body, i.e. sputum indicates the condition of the
chest and lungs, excrement, of the abdominal cavity, and so forth.
According to Hippocratic medical theory, patients with acute diseases tend to suffer
recurrent dramatic symptoms, called crises,14 which signify an important change in the
illness, toward either better or worse. Health was thought to be a balance of four humors, or
fluids: blood, phlegm, yellow bile, and black bile.15 Certain types of illness were thought to
be caused by a corruption of the humors. Treatment consisted of the physician coaxing the
patient’s body to “cook” and expel the noxious humors. This “cooking” was a kind of
“digestion” whereby the body’s heat destroys the noxious humor, and it is prepared to be
expelled from the body via one of the effusions already mentioned. The resulting crisis could
be severe and dangerous, and the strain on the body could even kill the patient. The
physician, therefore, who must be skilled in assisting nature in healing,16 must also prepare
the patient to be strong enough to bear the crisis when it comes. It was nature, however,
properly disciplined by skilled medical technique, that effected the actual healing, in the view
of the Hippocratics.
In the best estimation of modern scholars, the regularity of the occurrence of seizures
on critical days suggests that they were the symptoms of cyclical fevers, such as malaria.17
The Mediterranean climate and its seasonal weather patterns are consistent with this

13 The Greek text has been edited by B. Alexanderson: Galenos Peri Kriseon: Überlieferung und Text. Vol. 23,
Studia Graeca et Latina Gothoburgensia. Göteborg: Elanders Boktryckeri Aktiebolag, 1967. That volume
contains no translation; however, I have an English translation in preparation.
14 The Greek term krisis means a “judgment”: the patient was thought to be judged by the gods, and either

condemned to die or to be saved.


15 Cf. the Hippocratic treatise On the Nature of Man, in Hippocrates Volume IV. Translated by W. H. S. Jones.

Edited by G. P. Goold. Vol. 150, Loeb Classical Library. Cambridge: Harvard University Press, 1931. Reprint,
1992, pp.1-41; and Galen, On the Natural Faculties, Translated by A. J. Brock. Edited by G. P. Goold. Vol. 71,
Loeb Classical Library. Cambridge: Harvard University Press, 1916. Reprint, 1991.
16 The word “physician” still preserves a trace of the original conception of the doctor as the ally of nature in

healing. The word for nature in Greek is physis, and physician is physikos.
17 See Burke, Paul F. “Malaria in the Greco-Roman World: A Historical and Epidemiological Survey.” Aufstieg

und Niedergang der Römischen Welt II.37.2 (1996): 2252-81.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 4 of 15
hypothesis. There are two variant schemes of critical days in the Hippocratic writings, and
Galen provides a third, although Galen’s scheme alone is supported by rational argument. A
critical day occurred roughly every fourth day, but the series was counted in an idiosyncratic
manner, to make three weeks fit within twenty days. For example, in counting seven days,
the fourth day is counted twice.18 This odd method of counting is justified by noting that the
most prominent critical days, namely the days on which the most crises were thought to
occur, were the seventh, fourteenth, and twentieth days. The intermediate days, i.e. the
fourth, eleventh, and seventeenth days were treated as “indicator” days. Physicians could use
the condition of the patient on these intermediate indicator days to determine the nature of
the next crisis.
The scheme of the critical days represents the ideal situation, i.e. if the patient and
those who care for him do all that the sage doctor commands. In this case, the crises will fall
on the defined series of critical days. Any departure from the prescribed regimen may cause
the crisis to come prematurely, i.e. before the noxious humors have been completely
concocted, or too late, thus allowing the noxious humors to damage more bodily tissue. The
critical days assisted the physician in this effort by providing a scheme of days on which
crises supposedly occurred most often. The physician could calculate backwards from the
upcoming critical day, and compare what was supposed to happen (according to the scheme)
with the way things were apparently going to happen, based on the patient’s condition.
Knowledge of the critical day scheme could be part of the physician’s esoteric
accoutrements: the patient, without the sophisticated knowledge of an educated physician,
would have no clue how the physician performed the magical prognosis. Galen considered
the critical day scheme important enough to devote two treatises (i.e. On the Crises and On
the Critical Days) to describing the method of diagnosing crises and prognosticating using

18This method of counting has similarities to the method of counting musical intervals, a convention that
originates with Greek music theorists. This connection is strongly indicated by the fact that the terminology is
identical. I argue elsewhere that Galen’s theory of the critical days has many affinities with a harmonic
conception of the natural world.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 5 of 15
the critical days.19 Both of these texts were summarized into the ancient equivalent of a
“Cliff notes” and studied as part of the standard medical curriculum in late antiquity.20
In the steps Galen took to construct his theory of the critical days, we can see a hint of
the later scientific method. Galen states that his theory is based on two components:
experience and reason. These concepts, long since accepted as cornerstones of scientific
method, had never before been so thoroughly and soundly combined as they were in this
treatise. In presenting his theory, Galen shows how to reconcile two major trends in the
development of Greek medicine. Prior to Galen, Greek medicine had been dominated by
either Rationalists or Empiricists. Rationalists insisted that theoretical knowledge was all
that is necessary in treating patients. Empiricists, on the other hand, avoided all theory,
eschewed rational inference except in well-defined situations and of a specific type, and
based their treatments on specific details of each case.21 Galen perspicaciously recognized
that neither reason nor experience alone is sufficient for medicine. His treatises on the
subject are landmarks in the development of scientific thought.22 Yet, his work shows that
the notion of “experience” had not yet been precisely defined as (controlled) “experiment.”
To compare Galen’s method with the modern scientific method as usually
understood, I shall describe what Galen does, and then assess it in light of contemporary
scientific practice. The modern scientific method is generally recognized as having several
stages, involving observation, questioning, formulating hypotheses, and testing them. Greek
thinkers were accomplished at several of these stages, especially observation and formulating
hypotheses. With the requirements that one perform carefully controlled experiments and
that one’s results be verified by other researchers, however, Greek science, which was never
standardized nor institutionalized, differs significantly from our own.

19 I am currently preparing editions and translations of the On the Crises and the On the Critical Days in Arabic,
and, in the case of the On the Critical Days, in Greek also. These volumes will contain extensive historical
commentary.
20 A. Z. Iskandar, “An Attempted Reconstruction of the Late Alexandrian Medical Curriculum,” Medical

History 20 (1976): 235-58. In late antiquity, the mastery of sixteen medical books summarized in this way was
required for the medical degree. These summaries survive only in Arabic.
21 A third school developed, the Methodists, who proposed a simpler system of medicine, perhaps in response to

these polar approaches. Galen devotes much polemic and scorn to this group.
22 The most important of these treatises are found in: Galen. Three Treatises on the Nature of Science. Trans.

Michael Frede and Richard Walzer. Indianapolis: Hackett, 1985. They are: On the Sects for Beginners; An
Outline of Empiricism; and On Medical Experience.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 6 of 15
For convenience of discussion, I have divided Galen’s procedure into eight steps. I
shall list these steps and then consider each one in the context of Galen’s discussion of the
critical days.
1) Begin with a question, or set of questions, or an anomaly that calls for an
explanation;
2) Obtain a set of data;
3) Reduce the data to a workable subset;
4) Look for patterns in the (reduced) data;
5) Find correlations with other better-known phenomena;
6) Postulate a causal link between the data and the other phenomena;
7) Exploit the hypothetical connection to infer features of what is to be explained on
the basis of the better-known phenomena;
8) Answer the question: How are anomalies in the data to be accounted for,
according to the new perspective?

In steps 1–4, Galen employs inductive reasoning, extracting from pure experience as
much as he can. On the basis of observed patterns in the data, he inductively assumed that
the crisis would occur with regularity. These steps occupy most of the treatise, comprising
all of Books I and II. Galen intended these two books to be sufficient instruction in the
method of using the critical days for prognoses for the practical physician who does not
require theory. In steps 5–8, Galen employs reason to hypothesize causes for the
phenomena. Galen provided this theoretical discussion in Book III, where he invokes
planetary influence to arrive at celestial causes, for the sake of the physician who desires to
understand why the technique works. Each of the above steps will be discussed in view of
material from the On the Critical Days treatise.

1) Begin with a question, an apparent anomaly that stands out, calling for an
explanation.
Galen had three major questions about the critical days from his experience as a
doctor for which he sought answers. The first had dogged him since his youthful medical
GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 7 of 15
studies: Why is it that three medical weeks come to twenty days and not twenty-one days as
with normal weeks23? He found passages in the Hippocratic writings that suggested that the
critical days cannot be counted in terms of whole days24. The second question: Why does
the crisis occur on some days significantly more often than others do? (It is curious that
Galen looks at the days, not the patient, for the cause.) The third question: Is there a
relationship between the individual critical days; namely, can some days be used as
indicators of later days, thus being useful for prognosis? As a mature physician, Galen wrote
the On the Critical Days treatise to answer these questions, after he had acquired sufficient
medical experience, as well as a mastery of scientific reasoning.

2) Obtain a set of data.


The next step is to obtain a set of data that all agree is valid. In this case, Galen has
chosen to use the Hippocratic data from the Epidemics, since most everyone in his time
agreed that it was canonical. Were he to employ data from his own experience, he might be
accused of fabricating it to support his own theoretical agenda.25 Galen, however, does not
strictly observe this principle, citing his personal experience,26 where he mentions 400
patients whom he observed in one summer to have experienced a crisis on the seventh or
ninth days of their illnesses. In another summer, all of as many patients had the crises on the
fourteenth or twentieth days.27 Still, the basis of his scientific argument is founded on the
Hippocratic data—personal experience is cited merely for emphasis.
There is a significant difference here between Galen’s approach and that of modern
science. Modern science would insist on obtaining new data through clinical trial, under
controlled conditions. Historical cases would be considered only as a check on the
experimentally derived theory or hypothesis. Here Galen relies almost completely on the
data of another researcher.

23 Walzer, Richard. Galen on Medical Experience: First Edition of the Arabic Version with English Translation
and Notes. Oxford: Oxford, 1947, Chapter 21, p.127 (Frede edition: Three Treatises on the Nature of Science,
page 84).
24 Prognostic 20, ll.15—18; pp.42-43, Loeb edition.

25 K.842.4—6 and K.867.15—17

26 K.873.6—8.

27 K.873.8—10. Note: Galen restricts the data to Epidemics I and III. Books II, IV, and VI he rejects, since

their authorship was disputed (K.859.15—18), even though he claims that they tend to support his conclusions.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 8 of 15
3) Reduce the data to a workable subset.
The next step is to limit the data, or to reduce it to categories of similar cases. Galen
proceeded systematically through the patients’ cases described by Hippocrates in Epidemics I
and III, and carefully evaluated the days on which crises were reported. Certain reports of
crises that occurred rarely on a specific day Galen considered anomalous, and discounted
them. He offered an explanation as to why these few data diverge from the main trend, when
he discusses the effects of external environmental factors.28
Galen’s approach is sophisticated in that he recognizes that unaccountable factors can
upset the natural trend in data. Yet, since he is using another’s data, already at least 500
years old, and not under his control, we are suspicious that he favors his own preconceptions
in selecting the data. Modern science, of course, requires the researcher to define carefully
the experiment, its scope and controls, and to present this along with the data derived from it.
4) Look for patterns in the (reduced) data.
The next step is to examine the reduced data looking for patterns. The Hippocratic
data cluster around specific days, which is to say that crises were observed with greater
frequency in some days, of which the seventh, fourteenth, and twentieth days are the most
prominent. Galen’s implicit assumption is that, if there were no natural cause at work to
produce this pattern, crises ought to be observed on every day of the illness with similar
frequency, given a large sample of patients. It is hard for us not to see Galen’s bias toward
these days enter in at this point. He began with the question: Why are the principal critical
days on the seventh, fourteenth, and twentieth days? Now, he finds this very pattern in the
data.
5) Find correlations with other better-known phenomena.
The next step passes beyond simple induction, as Galen begins to seek causes by
correlating the data with other better-known phenomena. In order to win our confidence in
his astronomical knowledge, Galen accurately describes the lunar phases and eclipses as due
to the relative positions of the sun and moon. He shows familiarity with the works of the
astronomer Hipparchus (fl. latter half of 2nd C. BC), from whom he may have derived his
28 1.11.11 (K. 818.1 – 818.7); 1.11.14-5 (K. 823.17 – 825.2).

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 9 of 15
figures for the lengths of the different kinds of month.29 The popular poet Aratus of Soloi
(c.315–240 BC) is also quoted30, to show how the lunar phases can be used to forecast the
weather: farmers and sailors are said to be most familiar with this technique. The moon was
known to exert influence on terrestrial beings. Galen describes how the moon causes the
tides and how its rays cause madness in those who sleep in the open where they are exposed
to them. He suspects that the moon may exert a tidal effect on the fluid humors of the
patient’s body. Furthermore, Galen noticed a similarity in period between the moon and the
critical days. The moon, he observed, has quarter monthly periods less than seven complete
days. (He considers only the period when the moon is actually visible, which is less than 28
days). When three of these lunar weeks or “seasons” are summed, one arrives at
approximately twenty solar days. Galen devotes much space to refining the figure, to make it
fit the data even closer. We might accuse him of manipulating the data: it is clear from the
text that he was already convinced by the astrological connection, and wanted to make the
data correlate with it as closely as possible. Such a connection seems more plausible since
all of the phenomena and causes are part of the same system of nature, so the inherent natural
harmony guarantees that there will be a connection. Modern science, of course, has built-in
steps to minimize effects of bias, such as the requirement that the experiments be repeatable
by anyone, no matter what their biases may be.
6) Postulate a causal link between the data and the other phenomena
For Galen, to postulate causal links between entities is the paradigm of how
experience and reason work together in scientific theory. Galen makes the connection
between the lunar motion and the patient’s illness explicit, through employing analogical
reasoning. We might call what he does scientific modeling. Reasoning by analogy is to take
something that is more familiar, and to use it to explain something similar, but less familiar.
The more familiar entities in this case were the heavens: mostly self-evidently unchanging
and systematic, they seemed to present an excellent causal source for the less-familiar
physiology of cyclical illnesses.

29K.907.14—908.1.
30K.909.3—13. Translation is found in the Loeb edition, by A. W. Mair and G. R. Mair, Edited by G. P. Goold,
Loeb Classical Library. Cambridge: Harvard University Press, 1969.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 10 of 15
In Book III, Galen presents us with an intricate explanation of why the series of
critical days seems to feature multiples of four and seven days. Since Aristotle, ancient
scientists looked toward the “super-lunary” heavenly motions for the explanation of cyclical
phenomena in our world. Aristotle had postulated that the sun’s apparent annual motion
across the sky was the cause of the seasons on Earth generally, and of genesis and demise of
individual beings in particular.31 Galen finds the cause for the regular occurrences of crises
to lie in the moon’s monthly period, an idea he says he learned from the Egyptians, perhaps
during the time he spent in Egypt as a youth.32 In addition, he probably noticed the fact that
the month is approximately four weeks long, with each lunar “season” being seven days,
which is also the most significant period of the critical days. Then, remembering the
statement of Hippocrates in the Prognostic that the critical days cannot be counted according
to whole numbers of days, Galen sought a fractional week. Therefore, for three weeks to
sum to 20 days, the “medical week” must be slightly smaller than a whole seven days.
The period between full moons (approximately 29 1/2 days) is too large for this, since
a quarter of this period is greater than seven days. Accordingly, Galen considers only that
part of this lunar period in which the moon is actually visible, believing that the moon’s
influence is proportional to the amount of the sun’s light it reflects at a given time (i.e. the
full moon is more powerful than the crescent moon). That period is, according to Galen,
roughly three days less, or 26 1/2 days. In addition, there is another kind of month, the
sidereal, i.e. the period of the moon’s motion from a given Zodiac Sign and back again, that
is approximately 27 1/3 days. Galen thinks that these two effects, one due to the lunar
phases, and the other to the moon’s position along the Zodiac Signs, combine to influence the
patient, but in a mixture that only the gods know.33 He is content, however, to calculate an
average period between the two, to arrive at a medical week of 6 35/48 days. When

31 De generatione et corruptione II.11 (338a 19—338b 6); English translation is found in: Aristotle. On
Sophistical Refutations; On Coming-to-Be and Passing-Away; On the Cosmos. Translated by E. S. Forster and
D. J. Furley. Edited by G. P. Goold. Vol. 400, Loeb Classical Library. Cambridge: Harvard University Press,
1955. Reprint, 1992, pp.327-29.
32 K.911.14—18. See Nutton, Vivian. “Galen and Egypt.” In Sudhoffs Archiv (Beiheft) : Galen und Das

hellenistische Erbe Verhandlungen des IV. internationalen Galen-Symposiums Veranstaltet vom Institut für
Geschichte der Medizin am Bereich Medizin (Charité) der Humboldt-Universität zu Berlin 18-20 September
1989, edited by J. Kollesch and D. Nickel, 11-31. Berlin, 1993.
33 K.932.2—3.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 11 of 15
multiplied by three, this gives a three-week period of 20 3/16 days, sufficiently close to 20
days for his purpose. Thus, Galen is content that his first two questions have been answered.
He now knows why three medical weeks are about twenty days, and why certain days are
critical and others are not: these facts can be deduced from the features of the moon’s period.
7) Deduce Features of the Original Subject from the Hypothesis
Once the analogical connection is accepted, Galen feels free to argue in reverse:
because of the parallel between the critical day scheme and the lunar period, and the
postulated causation, specific features of the lunar motion permit him to conclude specific
things about the patient’s illness. In modern terms, this means to make a fresh deduction
from the tentatively accepted hypothesis. This, by the way, is the basic inference pattern of
astrologers, where precisely we moderns reject astrology: It is one thing to notice the
obvious connection between the solar motion and the changing of the seasons, and the lunar
motion and the tidal cycle (as Galen takes care to note), but it is quite another to push the
analogy farther than warranted and say that specific features of a particular planet or of its
motion have specific earthly effects. Now he has the answer to his third question: individual
days are related to one another by appearing at specific times in the lunar cycle. By knowing
the lunar cycle in detail, he can infer features of the critical days.
8) Account for anomalies in the data
Having derived a scheme that explains the ideal pattern, Galen asks: “What causes a
crisis to occur on a non-critical day?” And, “Why does an expected crisis not occur?” In
other words, what factors perturb the scheme? Alternatively, what causes the anomalous
data that he discounted from the Hippocratic data in the earlier stage? Galen does not give
these factors a name, but they were later known as “secondary causes,” i.e. factors that
interfere with the smooth operation of the primary causes. The primary causes were the
moon and its influence. In Galen’s discussion, these factors interfere with the smooth
operation of nature, in transmitting the causes and working their effects in the patient. Galen
observed that patients who suffered anxiety or lack of rest did not heal as quickly as others,
or at all. He thus concludes that perturbing factors are whatever upsets or disturbs the
patient. In an ideal situation, if the patient were isolated from external factors that cause him
anxiety and restlessness, such as fire, robbery, torrential thunderstorms and leaky roofs,
GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 12 of 15
quarreling neighbors and barking dogs,34 and furthermore, if he is under the exclusive care of
a physician of Galen’s caliber, and not being treated by ignorant but well-meaning loved-
ones, then the patient’s crises will occur according to the perfect scheme of the critical days.
These external factors affect the patient adversely, corrupting the scheme of nature, and
frustrate the healing process. Since the practical effect of such factors is to ruin the
prognosis, and possibly also his reputation, the physician, therefore, is enjoined to avoid,
where possible, any of these deleterious factors, and to minimize the effects thereof.

In conclusion, Galen’s method has shown sophistication in anticipating parts of later


scientific practice. He was adept at reasoning, and recognizing what constituted a workable
set of data, as well as with deriving a theory. As I mentioned, however, he does not follow a
method of gathering data or experiment that would be considered sound today. Still, his
efforts are an important piece of the puzzle of how science developed into the powerful tool
that it is today, and his neglected treatise ought to be further examined with a view toward
the light it can shed on that question.
Further research into Galen’s scientific method could follow several paths. The most
compelling might be a comparative study of the scientific method of Galen’s most famous
contemporary, the Alexandrian astronomer Ptolemy (fl. 2nd C. AD). The Almagest35 is a
much larger scale attempt to derive a scientific theory from a collection of astronomical data.
One researcher referred to this process as “Saving the Appearances.”36 The planetary models
presented there are well known and have been well studied over the past two millennia since
they were first devised. Not long ago a controversial study appeared that called into question
Ptolemy’s integrity in handling his data—in effect accusing him of fabricating data to fit his
theoretical models.37 Research along the line I propose would address those concerns within
a discussion of the honest challenges of reducing a data set to a collection useful for scientific

34 K.825.17—826.3.
35 Ptolemy. Ptolemy’s Almagest. Translated by G. J. Toomer. Princeton, NJ: Princeton University Press, 1998.
36 See the influential monograph Duhem, Pierre. To Save the Phenomena: An Essay on the Idea of Physical

Theory from Plato to Galileo. Translated by Edmund Doland and Chaninah Maschler. Chicago: University of
Chicago Press, 1969.
37 Robert R. Newton, The Crime of Claudius Ptolemy, The Johns Hopkins University Press, Baltimore and

London, 1977. See the critical review by Kristian Peder Moesgaard in Journal for the History of Astronomy,
vol. 11 (1980), pp.133-135.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 13 of 15
modeling. The investigation could also include other ancient theories, such as the same
Ptolemy’s investigations in musical harmonics,38 in order to be as comprehensive as possible.
Another line of research would be to investigate in detail the debt Galen’s causal
analysis owes to earlier thinkers, such as Plato and Aristotle. For example, we know that
Galen was an avid student of Plato, since he sought to imitate Platonic arguments, and he
tried to reconcile Platonic and Hippocratic doctrines.39 One could examine the Timaeus,
Plato’s grand cosmological myth, for any connection it may have with Galen’s theoretical
scenario.40 Galen wrote a commentary on the Timaeus,41 and that would have to be carefully
examined in light of this question. In the case of Aristotle, one would compare Galen’s
postulation of celestial causes with Aristotle’s theory of change as discussed in the Physics.
A study of that scope would be a key to understanding an entire world-view of cyclical
causes, celestial influences, and a purposeful nature. Lastly, one could examine the
influence, if any, the On the Critical Days may have had, either in the Hellenistic, the
Islamic, or the Latin scientific traditions. The precise influence of Galen’s treatise may have
been obscured by that of Ptolemy’s Almagest, since the latter was a more studied and more
influential work. One could, nevertheless, examine the medical authors who treat data and
derive theories therefrom. It is known, for example, that medieval physicians employed the
critical day scheme for timing bloodletting or for observing the symptoms of fevers. In
addition, there existed at least one version of the On the Critical Days in Latin,42 and some
medical practitioners possessed tables of planetary ephemerides for use in prognosis.
Although Galen’s text is only just becoming known once more, much still remains to
be investigated.

38 See Solomon, Jon. Ptolemy Harmonics: Translation and Commentary. Leiden, Boston, Köln: Brill, 2000.
See also the following study of the Harmonics: Barker, Andrew. Scientific Method in Ptolemy's Harmonics.
Cambridge: Cambridge University Press, 2000.
39 de Lacy, Phillip. Galen on the Doctrines of Hippocrates and Plato, 3 volumes. Berlin: Akademie-Verlag,

1984.
40 For a classic translation and commentary on this influential work, see: Cornford, Francis Macdonald. Plato’s

Cosmology. London: Compton Printing Ltd, 1966.


41 Galen, In Platonis Timaeum commentarii fragmenta, ed. H. O. Schröder, Corpus Medicorum Graecorum,

Supplementum, vol. 1. Leipzig: Teubner, 1934.


42 O’Boyle, Cornelius. Medieval prognosis and Astrology: A Working Edition of the Aggregationes de crisi et

creticis diebus with Introduction and English Summary. Cambridge Wellcome Texts and Documents No. 2.
Cambridge: 1991.

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 14 of 15
Glen M. Cooper
Independent Scholar

GLEN M. COOPER, NUMBERS, PROGNOSIS, AND HEALING: GALEN ON MEDICAL THEORY Page 15 of 15

You might also like