Prediction, precision, and practical experience: the Hippocratics on techne.

Mann, Joel E.

It would be difficult to overestimate the thematic importance of techne in classical Greek philosophy. In particular, the nature of techne, art or expertise, (1) as well as the question of what counts as a techne, was of paramount importance to Plato, who in the Gorgias and elsewhere waged a war against rhetoric on the grounds that it was not a techne at all but a stochastic competence derived from empeiria, that is, from practical experience. Isocrates, a teacher of rhetoric and Plato's rival heir to the Socratic mantle, defended his profession's claim to technical status by arguing for the adequacy of empeiria and ridiculing the search for precise knowledge of virtue. Further complicating matters, traces of classical disputes over techne are evident also in the Hippocratic Corpus, especially in works such as On the art, Ancient medicine, Nature of man, and Regimen in acute diseases. (2) Scholars are in general agreement about the situation of these medical writers in the burgeoning debate over techne. The Standard Story runs as follows. Just as Plato demoted rhetoric because it did not meet certain 'rationalist' requirements, so, too, did some anonymous antagonists seek to dismiss medicine or certain medical theories from the canon of technai for the failure of physicians to acquire comprehensive knowledge of the domain in which they professed expertise. The Hippocratics are thus envisioned as the Gorgiases of medicine, defending the dignity of their techne against charges that it is a pseudoscience by embracing empeiria and the conception of an imprecise techne championed by Isocrates. (3)

With the present essay, I will reverse or qualify the key points of the Standard Story. Neither Platonic 'rationalism' nor sophistic 'empiricism' is adequate to capturing the Hippocratic philosophy of techne. (4) Instead, I propose to divide the battle over techne into three camps, each with its distinctive position on the relation between technical theory and practice and the role of experience in mediating between them. The conflict may be conceived as a disagreement over how best to relieve the inconsistency in the following theses. For some agent, A, of suitable natural aptitude in some specified domain of action, P:

i. A's reliable success (under varying circumstances) in P is necessary and sufficient for A's technical expertise in P;

ii. A's practical experience in P is necessary and sufficient for such success in P; and

iii. in order to acquire technical expertise in P, it is necessary for A to have knowledge of P over and above practical experience in P.

This analysis of the techne problem is more than a mere heuristic; it represents a genuine crisis in the ancient conception of techne. Thesis i would have been an analytic truth by most Greek standards. When Aeschylus' Prometheus boasts of his generosity to the human race, he describes his gift primarily as a set of practical procedures useful for effecting some beneficial end, namely, the curing of disease. So he says

 when you hear the rest of my story, you will be even more
 amazed at the kinds of technai and resources I devised.
 The greatest is this: if ever anyone fell sick,
 there was no remedy--neither food, unguent, nor drink--but
 because drugs were lacking, they wasted away, until I
 showed (edeixa) them how to mix the soothing cures
 with which they drive away all diseases. (476-83)

Moreover, the techne for which Prometheus takes credit would have been transmitted from master to apprentice through training in the manual procedures of the discipline, that is, through practical experience, or empeiria. (5) Such training involves the observation of and hands-on exposure to particular cases of injury or disease and their attempted cure. It results in a doctor with manual dexterity and a general sense for the circumstances under which his dexterity is to be applied. It does not result in comprehensive knowledge of facts or events in the physician's technical domain. Indeed, Prometheus does not promise such knowledge of anything, whether the inner workings of the human body or the natures of his materia medica. He offers only pharmaceutical know-how, and nothing else, but that is enough to guarantee reliable success in curing disease (thesis ii).

At the same time, there was the sense, especially among those who laid claim to expertise, that true techne required not just practical experience or success in a domain, but genuine knowledge, especially of the future (thesis iii). The Hippocratic treatise Prognostic begins from the premise that foreknowledge of patient outcomes is essential to medicine.

It seems to me best for a physician to try to understand things in advance. For if he knows them in advance and predicts in the presence of his patients what is happening, what has happened, and what will happen (to the extent that these things were left out of the patient's original account), the more he will be believed to know his patients' real situations, so that people will dare to place themselves in the doctor's hands. What's more, he will give the best treatment if he knows in advance what will happen on the basis of present symptoms. (I1-10 Jones)

Two points are of special interest. First, there is the supposition that patients will be suitably impressed by the physician's foreknowledge, as though such knowledge were a reliable indicator of his ability to successfully heal them. Second, there is the (unsupported) assertion that foreknowledge of the disease's development is a sufficient condition for excellent treatment. That is not to say that, without such knowledge, success is utterly impossible. But the best doctors--both in public opinion and in reality--will be those who 'know the natures of the diseases,' and knowing their natures will allow the physician to predict with greater certainty which therapeutic measures will be effective (I 19-20 Jones).

My intent is to demonstrate how and why the three theses given above came to define the ancient philosophical dispute over techne. That they defined the dispute is evident from the fact that the prominent classical theorists of techne felt compelled to take positions on each. Plato, for example, rejects i. Isocrates, the representative of empiricism, rejects iii. We will discover that, despite their theoretical and rhetorical diversity, the methodologically self-conscious Hippocratics mentioned above tend to reject ii while accepting i and iii. The Hippocratics share with the empiricists a conception of techne as successful practice; nonetheless, they join Plato in denying technical status to any practice or inquiry based solely in experience. But they do so for decidedly different reasons than does Plato, and it is these differences that set them apart within the Greek philosophical tradition and render them worthy of attention in their own right.

I Platonic rationalism

First, let us review briefly Plato's position on techne. Most familiar are his attacks in the Gorgias on the claim to technical expertise made by popular teachers of rhetoric. Plato's objections to rhetoric are many, and I shall not rehearse them all here, but arguably his most consistent criticism focuses on rhetoric's failure to develop an explanatory understanding of its subject matter. Socrates contends that rhetoric, as practiced by the sophists, is not a techne (technen de auten ouk einai)

 because it has no account (logos) of the nature (phusis) of
 whatever things it applies by which it applies them, so that it's
 unable to state the explanation (aitia) of each thing. And I refuse
 to call anything that lacks such an account a techne. (465a)

A techne, correctly conceived, is rational in the following sense: its directives can be given an aitia, or explanation, that refers to the phusis of the domain in question. (6) Causal explanation is for Plato a necessary, though not sufficient, condition of technical legitimacy. Moreover, it is a condition that standard rhetorical practice fails to meet, and so it fails to be a techne despite the success and celebrity enjoyed by Gorgias and others.

Indeed, most rhetoricians--including Gorgias--have not techne but empeiria, which Plato unflatteringly characterizes as a knack for demagoguery derived from practical experience.

There's a practice that's not technical (technikos), but one that a mind given to making guesses (psuches stochastikes) takes to, a mind that's bold and naturally clever at dealing with people. I call it flattery, basically. I think that this practice has many other parts as well, and pastry baking, too, is one of them. This part seems to be a techne, but, in my account of it, it isn't a techne but experience (empeiria) and routine (tribe).7 (Gorgias 463a-b)

Various parts of flattery, including pastry baking, seem to be technai because their practitioners really are competent at achieving relatively reliable results in some domain. Plato does not question that the pastry chef can be counted on to serve up exceptional croissants and eclairs, even if he can't explain why his recipes and procedures are successful. Plato accepts--quite reasonably, it seems to me--that experience brings with it practical success.

Moreover, Plato does not deny that experience is a necessary component of success in applying techne. In the Phaedrus, Socrates endorses an account of technical proficiency in which practical experience plays a key role. (8)

 If you have a natural ability for rhetoric, you will become a
 famous rhetorician, provided you supplement your ability with
 knowledge (episteme) and practice (melete). To the extent that you
 lack any one of them, to that extent you will be less than perfect.
 (269d)

In addition to his theoretical knowledge, the adept orator, like any technician, must

 put his theory into practice and develop the ability to perceive
 (aisthesis) each kind clearly as it occurs in the actions of real
 life. Otherwise he won't be any better off than he was when he was
 still listening to those discussions in school. He will now not
 only be able to say what kind of person is convinced by what kind
 of speech; on meeting someone he will be able to perceive
 (diaisthanesthai) what he is like and make clear to himself that
 the person actually standing in front of him is of just this
 particular sort of character he had learned about in school--so
 that he must now apply speeches of such-and-such a kind in this
 particular way in order to secure conviction about such-and-such an
 issue. (271e-2a)

According to Plato, technical reasoning starts from universal truths about the natures of things, whether bodies (in the case of medicine) or minds (in the case of rhetoric). Such knowledge is supplemented by particular judgments based on perceptual experience to yield a technical instruction tailored to the circumstances. So, for example, the able orator must know first that all curious, romantic minds are persuaded by elegant speeches on fantastic subjects. But, in order to successfully persuade Phaedrus, he must also be able to perceive that Phaedrus has just such a mind.

The expert in rhetoric or any other techne requires practical experience to become adept at determining the means useful for achieving the desired end. Insofar as he must determine what is useful, the expert must predict what will happen if certain measures are applied. As Socrates puts it in the Theaetetus,

 one might put a question about the whole class of things to which
 "what is useful" belongs. These things are concerned, I take it,
 with future time; thus when we legislate, we make laws that are
 going to be useful in the time to come. This kind of thing we may
 properly call "future". (178a)

There is an objective fact of the matter about what will happen in the future, Socrates insists. If a doctor and a layman disagree about whether the patient will take a fever, one of them will be wrong (178c). The key difference between the expert and the layman is that the expert's predictions are correct. This applies to anyone who makes a claim to wisdom in a given domain, including Protagoras, the self-proclaimed wise man, who would 'predict (prodoxazein) better than any layman about the persuasive effect that speeches in a law court will have upon any one of us' (178e). Importantly, the expert's judgments, even if better than the layman's (they need not be perfect), are not characterized as knowledge, and given Plato's remarks on rhetorical proficiency in the Phaedrus, it is easy to see why this should be the case. For the accuracy of the expert's predictions are contingent upon his natural ability and level of experience, both of which restrict his ability to convert his knowledge of universal natures into certainty about particular practical outcomes.

But if Plato accommodates practical experience in his theory of techne, he is unyielding in his decision to banish empiricists from the ranks of technical experts. With practiced perceptual abilities but no theoretical structure on which to rely, the empiricist, unlike the rationalist, will face the future armed only with a mixed bag of particular experiences. But this will not satisfy Plato; the merely experienced practitioner still lacks true technical expertise, since he fails to comprehend the universal natures of things in his domain of interest. During his confrontation with Callicles in the Gorgias, Socrates restates his contention that pastry baking is not a techne, but

empeiria, whereas medicine is a techne. I said that the one, medicine, has investigated both the nature of the object it serves and the aitia of the things it does, and is able to give a logos of each of these. The other, the one concerned with pleasure, to which the whole of its service is entirely devoted, proceeds toward its object in a quite untechnical way, without having at all considered either the nature of pleasure or its aitia. It does so completely irrationally (alogos) and "reckons naught," as they say. Through routine (tribe) and empeiria it merely preserves the memory of what usually happens, and that's how it also supplies its pleasures. (Gorgias 500e-1a)

While the empiricist will, through the accumulation of individual experiences, develop operational competence in virtue of remembering which measures have been successful in past cases, he will fall short of truly technical expertise. But what's wrong with relying on memory?

Such a memory underwrites, at best, a kind of guesswork about what will happen, which Plato indicates with the Greek verb stochazein and its cognates. (Recall Socrates' derisive description in the Gorgias of the orator's 'stochastic mind'. Cf. 464c.) Plato criticizes stochastic reasoning most sharply at Philebus 55d and following, where the necessary connection between empirical techne and stochasticism is made most manifest. There, Socrates wonders whether some of the productive (demiourgikai) technai (as opposed to the merely edifying, peri paideian kai trophen) are more or less epistemically 'pure' than others.

If someone were to take away all arithmetic, measuring, and weighing from the technai, the rest would be cheap (phaulos), so to speak.... All we would have left would be the comparison of particular cases (eikazein) (9) and the training (katameletan) of our senses (aisthesis) through empeiria and tribe. We would have to rely on our ability to make the educated guesses (stochastikos) that many people call techne, once it has acquired some proficiency through practice (melete) and effort (ponos). (55e-6a)

Any productive techne will involve a perceptual, and therefore impure, element, since its aim will be to effect some specific change in the perceptible world of particulars. Some of these technai are redeemed, at least in part, through their incorporation of mathematical principles and procedures. But the practitioners of others are doomed to muddle through on perceptual skill alone, relying on the hunches drawn from practical experience, without any recourse to the necessary and precise truths of mathematics. (10) Almost anyone could manage this sort of technical achievement, given enough experience and practice.

The empirical technai have natural limits on their potential to be successful, since, ceteris paribus, reasoning from exceptionless general principles will give more reliable results than will hunches based on experience. Thus, Socrates divides the productive technai into those with less precision (akribeia), such as music, and those with more, such as building (56c). (The former are 'what the many call technai' (55e-6a), leaving open the possibility that Plato, for his part, would withhold the title.) But the precision he has in mind here is not primarily the practical precision exemplified in a high success rate. Plato takes only incidental interest in the practical accomplishments of any techne as such. Comparing dialectic to rhetoric, Socrates dismisses the latter's claim to superiority on the basis of its utility. What matters, rather, is 'whether [dialectic] is by its nature a capacity in our soul to love the truth and do everything for its sake' (58d). Indeed, to suppose that Plato's critique proceeds from practical concerns is to misunderstand his project. His task is to show why, despite their successes, certain practices such as rhetoric should be held in low epistemic regard.

Accordingly, while a rational techne is potentially more precise in practice than its empirical counterpart, its success rate may not serve as a reliable index of its epistemic value. Given the ineliminable role of experience in technical practice, the success of any productive techne will be limited by, inter alia, the perceptual skills of the practitioner and the complexity, variability or opacity of particular circumstances. So in the Statesman, Plato suggests that the legislator, like the trainer, will be incapable of prescribing accurately (di' akribeias) for each person what is appropriate, since he is forced by circumstances to prescribe for the entire group (295a). Similarly, the doctor's prescriptions may turn out to be wrong when circumstances--in particular, climactic conditions--defy expectations (295d). Such lapses of precision in practice, occasioned as they are by obstacles to the application of general knowledge in particular cases, can't reasonably be held against a techne. However, Plato demands precision in the principles of a techne, that is, in the account of the natures of those things over which an expert claims to have influence. As Socrates in the Phaedrus says of a truly technical rhetoric, 'it is clear that someone who teaches another to make speeches as a techne will demonstrate precisely (akribos) the essential nature of that to which speeches are to be applied' (270e; cf. 271a).

For Plato, the issue of empirically derived competence in a domain is to be kept separate from the issue of technical status. The latter is to be decided exclusively on the character of the principles grounding practice. The truly expert doctor may not be any more precise in practice than the empiricist. But when his diagnoses and prognoses are correct, they are based on knowledge, not stochastic guesswork derived from practical experience. While experience is sufficient for some measure of practical success (thesis ii), practical success does not a techne make: Plato rejects thesis i. For techne is not just a system of manual procedures accompanied by a sense for when to apply them, but a body of knowledge, and experience alone is inadequate for discovering the precise principles that elevate techne above the level of mere knack (thesis iii).

II Sophistic empiricism

Not all share Plato's dim view of empeiria, however. Gorgias' associate, Polus, comes to the beleaguered sophist's aid by insisting that empeiria is the key to technical expertise in many, if not all, domains: 'many among men are the technai experientially devised by experience (empeiria), Chaerephon. Yes, it is experience that makes our lives proceed according to techne, whereas inexperience (apeiria) makes them proceed according to chance (tuche)' (448c). (11) The world is teeming with different kinds of technical ability, suggests Polus, because expertise is cheap. Its price is practical experience, which produces in the practitioner a competence exceeding what we would expect from pure guesswork alone. This in itself is good enough to confer the title of technician upon Gorgias and many others.

The fullest extant exposition of sophistic empiricism comes to us courtesy of Isocrates, the student of Gorgias and Protagoras. His early speech, Against the sophists (c. 390 BCE), is especially instructive: 'for power (dunamis), whether in speech or in any other activity, is found in those who are well endowed by nature and have been trained by practical experience (tois peri tas empeirias gegumnasmenois) (14). Later Isocrates would use even stronger language in praise of practical experience in rhetoric: 'everyone obtains here [in Athens] that practical experience (ten empeirien) which more than any other thing imparts ability to speak ...' (Antidosis 296). These remarks make clear that power and control over one's environment, not knowledge, is the chief criterion of technical achievement for Isocrates, and practical experience is the rhetorician's key to this power. Here we find no mention of the sort of formal instruction recommended to the student of rhetoric by Plato in the Phaedrus, though elsewhere Isocrates acknowledges that such instruction could play a role. The technical aspirant 'must, in addition to having the right sort of nature, learn the forms of speech and practice their application (peri de tas chreseis auton gumnasthenai). The teacher must go through the forms as precisely (akribos) as possible so that he leaves out nothing that is teachable' (17). Precision in the principles of a techne is valuable, but only in virtue of its consequences for precision in practice. Isocrates doesn't insist upon perfect precision; the 'forms' (of speech, that is) need be articulated only as precisely as they allow, since a technical principle may still have practical value even when it admits of exceptions. Moreover, education in such principles does not improve upon empeiria so much as provide a more efficient way of acquiring what could be gained through unsupervised experience: 'formal instruction (paideia) makes such men more technically adept and better equips them for their investigation, teaching them to take from a readier source what they otherwise stumble upon by chance (entunchanousi planomenoi)' (15).

As its title suggests, Against the sophists is less a defense of Isocrates' views than an attack on rhetorical and philosophical competitors. One of his main complaints is that the sophists put far too much weight on formal instruction in the acquisition of techne and far too little on experience. They 'do not attribute any of this power (dunamis) either to the practical experience (tais empeirias) or to the native ability of the student, but undertake to transmit the techne of discourse as simply as they would teach the letters of the alphabet' (10). (12) But Isocrates does not take issue with the sophists simply because they pass on technical rules to their students; rather, he is critical of the precise formulation such rules are given.

 But I marvel when I observe these men setting themselves up as
 instructors of youth who cannot see that they are applying the
 analogy of a techne with hard and fast rules to a creative process
 (poietikou pragmatos). For, excepting these teachers, who does not
 know that the art of using letters remains fixed and unchanged, so
 that we continually and invariably use the same letters for the
 same purposes, while exactly the reverse is true of the art of
 discourse? For what has been said by one speaker is not equally
 useful for the speaker who comes after him.... (12)

The sophists allege that techne is a matter of applying exceptionless general principles in particular situations. Isocrates, by contrast, denies that such principles can be formulated for the domain of rhetoric, thereby embracing a doctrine of metaphysical imprecision, namely, that the phenomena within the given domain exhibit a variability that frustrates description with precise, universal generalizations. (13) His techne will be stochastic, that is, it will involve some measure of guesswork in its predictions, because precise knowledge of the domain is simply impossible given its metaphysical imprecision. (14) The targeted sophists, like Plato, are much more optimistic about achieving precision in technical principles. But, unlike Plato, the sophists hold both that they possess knowledge of these principles and that such knowledge is necessary and sufficient for technical success.

Thus it is the sophists' 'attempt to persuade our young men that if they will only study under them they will know what to do in life and through this knowledge (dia tautes tes epistemes) will become happy and prosperous' that so raises Isocrates' ire (3). But rhetoric (or philosophy--Isocrates does not recognize a difference where Plato would have) cannot deliver the promised goods. 'I should have preferred above great riches that philosophy had as much power (dunasthai) as these men claim ... But since it has no such power, I could wish that this prating might cease' (11). The knowledge claims of the sophists are empty--not only because some particular facts within the domain of rhetoric resist assimilation to universal generalizations but also because human faculties are inherently limited.

 For I think it is manifest to all that foreknowledge of future
 events is not vouchsafed to our human nature, but that we are so
 far removed from this prescience that Homer, who has been conceded
 the highest reputation for wisdom, has pictured even the gods as at
 times debating among themselves about the future--not that he knew
 their minds but that he desired to show us that for mankind this
 power lies in the realms of the impossible. (2)

The appeals to poetic authority and mythology constitute neither argument for nor explication of Isocrates' position and so are ultimately unsatisfying. Nonetheless, we may extract from this passage an epistemological thesis that multiplies the uncertainty already implied by metaphysical imprecision: there are facts about the world that human beings, in virtue of the inadequacy of their faculties, cannot know. These unknowable facts include facts about the future that are essential to determining correct rhetorical practice and so preclude the sophist from acquiring any genuine knowledge of what should be said in a particular case. It should be noted that since all facts about future events are unknowable, Isocrates' attack on sophistic rhetoric functions implicitly as an attack on any techne that lays claim to foreknowledge of events in its domain.

Moreover, Isocrates argues, if the sophists possessed a reliable method for acquiring the knowledge they advertise, then two consequences would follow. First, there would be general agreement on correct rhetorical procedure among practitioners. Second, these agreed-upon procedures would meet with regular success. But neither of these is the case.

 [When laymen observe that] those who make use of their opinions are
 more apt to agree and are more often correct than those who profess
 to have knowledge (epistemen), I think they are likely to look down
 upon them and believe that such activities are chatter and
 triviality, not care of the soul. (8)

As a result, Isocrates ultimately concludes that 'there does not at all exist such a techne (holos men gar oudemian hegoumai toiauten einai technen)' (21), by which he means that the sophists are incapable of achieving the objectives that define their professional practice. Isocrates and the sophists are engaged in what is essentially a factual dispute. While they agree that reliable success in a given domain is a necessary and sufficient condition for techne, they disagree over whether the sophists are actually successful. If they are not, their actual practices will not satisfy the description of their techne. But if there were something that satisfied the description, there would be no question that it would warrant the title of techne. It is worth contrasting this with Plato's judgment at Gorgias 465a that rhetoric is not a techne (technen de auten ouk einai). The question there is not whether there are any successful sophists, that is, whether there is something that satisfies the sophist's description of his endeavor (supposing he could manage to give a coherent one--another of Plato's criticisms). Plato would readily concede this point. Consequently, Plato could not claim that rhetoric was completely nonexistent. But he could, and did, wonder whether it deserved to be called a techne, since the sophist's success was not accompanied by--much less caused by--genuine knowledge. This is essentially a conceptual, as opposed to factual, dispute.

As D.S. Hutchinson has suggested, the roots of this dispute reach back at least as far as the sophist Protagoras (30). In his Antidosis, Isocrates incorporates elements of Protagoras' Great Speech (Protagoras 320d-8d) into the defense of his own techne. There Isocrates tells a story of human progress that mirrors the one attributed to Protagoras by Plato (253-5). According to both Protagoras and Isocrates, it is their political techne of rhetoric that is responsible for humankind's salvation from a state of nature that was, to borrow a Hobbesian expression, nasty, brutish, and short. Further, just as Protagoras distanced himself from Hippias and other teachers of technai such as geometry, astronomy, music, poetry and mathematics (318e), Isocrates cautions against prolonged engagement with such studies. Of geometry and astronomy, he does not

think we ought to apply the term "philosophy" to a practice that does not benefit us in the present, either in speech or action, though I do call it a training of the soul and a preparation for philosophy. It is certainly more suitable for adults than what schoolboys do, though it comes close in many respects. For even those schoolboys who have labored through grammar and music and the rest of their education show no improvement in speaking or deliberating about practical matters, though they are well prepared to study the greater and more serious subjects. Thus, I would advise the young to spend some time on these studies, but not to let their education waste away on these things, nor to run aground upon the theories of the ancient sophists ... (Antidosis 266-8)

Isocrates preaches the Protagorean ethos without any trace of shame: the point of philosophy is to 'manage well both one's private household and the public assets of the community' (cf. Protagoras 319a) in contrast to those who 'don't care at all about the practical necessities but love the tall tales (teratologias) of the ancient sophists' (285). Such subjects have no intrinsic value, unless it is entertainment value. They are useful only as a kind of preparatory training for success in practical engagements. Isocrates is reluctant even to call them technai; they are more like fantasies, void of epistemic and practical value.

Thus, Isocrates follows Protagoras in conceiving of techne as essentially practical. If Plato's reproduction of the Great Speech in the Protagoras is historically accurate, we may be inclined to assign Isocrates' empiricism a Protagorean pedigree. In that dialogue, Protagoras accepts that nature and training (especially conditioning through punishment and reward) are significant factors in determining a person's technical facility (323dff.). Evidence from other sources indicates that this is no Platonic invention. An independent fragment of Protagoras, recorded in the Anecdota Parisiensia, explicitly states that 'teaching (didaskalia) requires nature (phusis) and practice (askesis)' (DK 80 B3), and the sophist is recorded elsewhere as having written that 'techne without practice (aneu meletes) and practice without techne are nothing' (DK 80 B10). Practical experience is necessary for acquiring a techne, including the political techne. But, assuming a student had suitable natural ability, it is less clear that Protagoras would have thought experience more or less sufficient. He seems to leave room for formal instruction, which he may regard as distinct from practical experience. But did Protagoras conceive of such instruction as the transmission of knowledge, as Plato would have it, or rather as the transmission of experience in a more compact form (as rough-and-ready rules of thumb, for example), as Isocrates insisted?

The historical evidence is inadequate to give much in the way of a positive answer. But there is good reason to think that Protagoras would have rejected the idea that teaching amounted to the transmission of specialized knowledge, since by all accounts Protagoras was intensely skeptical about such knowledge. In a famous fragment, he professes agnosticism about the gods' existence on the grounds of their adelotes, or non-evidence.

 Concerning gods I am able to know neither that they are nor that
 they are not, nor what sort they are in form, since there are
 considerable obstacles to my knowing, namely their non-evidence
 (adelotes) and the brevity of human life. (DK 80 B4)

Protagoras does not use adelotes simply to signal that the matter is difficult to judge; on such an interpretation, the fragment would turn out to be little more than a tautology. Surely Protagoras intends to convey more than that he finds it difficult to settle on a judgment about the gods because the matter is difficult to judge. Protagoras is generally taken to mean that his uncertainty about the gods--whether they 'are' or 'are not'--is rooted in the fact that he has himself never seen or otherwise directly perceived them.

If this interpretation of the fragment is correct, then we can draw the following conclusions about Protagoras' attitude toward ta adela, 'the non-evident things'. First, consider that Protagoras refuses to affirm sentences that contain references to the non-evident. He does not believe that he has or can have any certain knowledge about things that he has not directly observed. (15) But his unwillingness also to deny sentences that refer to the non-evident is telling. He concedes that such sentences could be true without being known, that the non-evident could exist but simply might not yet have been perceived by him. (16) The shortness of human life, therefore, is critical because it recognizes a temporal limit on our ability to observe phenomena. The god that Protagoras never witnessed during his lifetime could make an appearance after his death in such a way that he would have attained some measure of certainty. He cannot rule out such a future appearance, for future events, too, are adela.

Protagoras, like Isocrates, was highly skeptical about claims to knowledge. Human beings are incapable of knowing ta adela. The consequences of his view for many technical experts who purported to have knowledge of their domains would have been devastating. Aristotle's Metaphysics offers compelling evidence that Protagoras himself put his epistemological skepticism to destructive ends. There, Aristotle writes that

 perceptible lines are not those of which the geometer speaks (for
 nothing perceptible is straight or round in this way--the circle
 does not touch a straight-edge at a point, but as Protagoras said
 it did in refuting the geometers.) (Meta. III 2, 997b-998a4 = DK 80
 B7)

The phrase 'as Protagoras said it did' must refer to the fact that, considered purely perceptually, a circle touches its tangent at more than one point. A straight stick, for example, will seem to come into physical contact with a metal hoop along a discrete segment of its length, not at a single, infinitesimal point. How does this refute the geometers? Quite simply, Protagoras understands the perceptual evidence to contradict their basic principles. The experts say that tangents touch at a point, while he maintains, in accordance with sense experience, that they do not.

Protagoras is thus involved in a dispute over the nature of circles and tangents. But the dispute does not consist in variant claims about the way that lines and circles appear to observers. Protagoras, to the best of our knowledge, held that such disputes cannot be adjudicated with any absolute authority (a point to which I shall return). If the wind seems hot to me and cold to you, it is not the case that one of us is correct at the expense of the other (see Theaetetus 152bff.). In any case, geometers would not deny that perceptible lines and circles behave just as Protagoras describes. Rather, they will argue that the lines and circles of concern to geometry are intelligible, not perceptible.

This is just the sort of maneuver that Protagorean skepticism blocks. Protagoras does not commit himself to the view that the gods, for example, 'are not'. Instead, it is non-evident to him whether the gods exist. He simply denies that he can have knowledge of such matters. Likewise, in contradicting the geometers Protagoras cannot be taking sides in a substantive dispute over geometric principles per se; the dispute is a byproduct of his skepticism about the epistemic status of the particular principles espoused by geometers. The subjects of these principles are purely intelligible and so, in Protagorean terms, adela, non-evident. Thus, they are unverifiable and may not count as knowledge.

There is evidence, then, that Protagoras composed works attacking technical experts and that the basis for at least some of these attacks was the technical reliance upon non-evident principles. In fact, in the Sophist Plato tells us that the scope of his critique was not limited to a single techne.

STRANGER: And what about the laws and all other civic matters? Don't they promise to make people capable of debating about them?

THEAETETUS: In a word, nobody would talk to them if they didn't promise this.

STRANGER: Indeed, the points one ought to raise in a debate against each craftsman himself (pros hekaston auton ton demiourgon), concerning all technai in general and each techne specifically, have been put down in writing and published somewhere for the benefit of anyone who wants to learn.

THEAETETUS: You appear to me to be talking about the works by Protagoras on wrestling (peri pales) and the other technai.

STRANGER: And on many other subjects (kai pollon ge), too, my good man.... (232d-e)

Scholars have often been confused about the import of this passage, specifically with respect to two points emphasized in the above translation. First, it is easy to mistake auton ton demiourgon for the accusative subject of anteipein and to take pros hekaston as an independent prepositional phrase. (17) This would have Protagoras defending the technai. However, the context rules this out, and the clearly correct approach is to treat pros hekaston auton ton demiourgon as a whole. Plato fingers Protagoras as a critic of the technai, publishing pieces on how to dispute with the practitioners themselves (as opposed to attacking them indirectly, e.g., by persuading their clients to discontinue service).

Second, the stranger's final response in the above excerpt (kai pollon ge) has sometimes been understood as a reference to other sophists (in addition to Protagoras) who published critiques of the technai. (18) However, examination of the immediately preceding sentence of the dialogue reveals that kai pollon ge must fall under the scope of Theaetetus' peri and thus must refer to other works by Protagoras himself. Given the grammatical constraints, in order to mention the works of other sophists, Plato would need to have written kai ta pollon ge.

So not only did Protagoras develop a theory of techne that emphasized both the importance of practical success in a domain (thesis i) and the role of practical experience in achieving that success (thesis ii), but he used his skepticism about the non-evident to attack the epistemic claims of geometers, and it is conceivable that he adverted to the same strategy in his attacks on other technai. If so, it is likely that he would have rejected thesis iii as well. Indeed, none of the available evidence portrays Protagoras as making knowledge either a precondition or byproduct of technical achievement. (19) It is difficult to imagine what role he could have reserved for knowledge in the acquisition and practice of many technai, given that human knowledge is restricted to statements about what has been directly observed by a particular person, perhaps even at a particular time (cf. Theaetetus 167a-b). What the expert needs is the practical experience that produces good judgment, not knowledge, and good judgment (euboulia) is precisely what Protagoras promised to instill in his pupils (Protagoras 319a). When Isocrates, in his works Against the sophists and Antidosis, praises judgment and excoriates pretenders to knowledge, he is drawing directly upon this Protagorean empiricist tradition of technical theory and criticism.

III The Hippocratics

It is to attacks launched from within this empiricist tradition that the Hippocratic writers respond. (20) In the medical corpus we find a group of treatises concerned with refuting the public accusation that the medical techne 'is not' (ouk estin). The author of Regimen in acute disease complains that 'in fact the whole art is the subject of considerable slander before the public, so that medicine does not seem to be (me dokein ... einai) at all' (VIII 3-5 Jones). In Ancient medicine, the author observes that 'some craftsmen are poor, but others are extraordinary, which would not be the case if the art wholly were not (ei me en ietrike holos) and there had been no investigation into it nor discoveries made' (I 2 Schiefsky). (21) Those who give reductive accounts of medicine are especially contemptible, he says, because they are mistaken 'about a techne that is (amphi technes eouses)' (I 1 Schiefsky). Later, he portrays his project as that of demonstrating 'that medicine is' (II 2 Schiefsky). (22) Likewise, the sophistically inclined On the art opens with a rebuke of those who demean the technai and follows with a defense of techne in general and of medicine in particular. (23) The defense begins with the counter-assertion that 'there is no techne that is not' (techne einai oudemia ouk eousa), and the author prides himself on having shown that medicine 'has being and power' (hoti eousa te esti kai megale) (II 1, V 3 Jouanna). That the Hippocratics frame the problem in such terms indicates that they are taking up the factual dispute, familiar from Isocrates' Against the sophists, over whether there exists a practice that satisfies the definitional description of their techne. They do not occupy themselves at any point with the charge that medicine is not a techne, that is, with the conceptual challenge to conventional notions of what it is to be a techne.

Fittingly, then, the Hippocratic treatises, especially On the art and Ancient medicine, concentrate on demonstrating that medicine does in fact succeed in its efforts to heal the sick and suffering. As G.E.R. Lloyd has noted, the Hippocratics defend against predominantly practical objections to medicine, whereas Platonic objections to ostensible technai are exclusively epistemological (1991, 252). This is partly due to the fact that they regard practical success at healing as the ultimate criterion of technical ability (thesis i). However, we will see that Lloyd's observation requires further qualification. The Hippocratics do indeed take up certain epistemological challenges to medicine. Given the character of empiricist attacks on techne, we would expect arguments over the practical success of medicine to have epistemological consequences. For Isocrates argued that practitioners with knowledge ought to meet with great practical success; he claimed, further, that they were generally unsuccessful, proving that they did not possess the knowledge they professed. In response to similar concerns, the author of Ancient medicine protests that

we ought not reject the ancient art as not being (hos ouk eousan) and as having been poorly investigated (oude kalos zeteomenen) just because it is not precise (akribeia) in all respects. Rather, since it was able through reasoning to achieve near perfect accuracy (atrekestatos) from a state of considerable ignorance, much more do I think we ought to admire its discoveries insofar as they were discovered in a praiseworthy and correct manner and not by chance. (XII 2 Schiefsky)

The precision alluded to here is practical precision. (24) It is admitted that physicians do not have a perfect success rate in all areas of medicine. More importantly, this lack of perfect success is viewed as a threat not just to the knowledge claims of physicians but also to the methods by which they claim to acquire their knowledge. The Hippocratic defense of medicine thus takes on the further task of justifying its methodology before its detractors, who, it has generally been assumed, are rationalists of some stripe. Schiefsky, for example, is convinced that Ancient medicine is a response to critics from within the medical profession who insist that any rational, systematic physician must adopt a stringently reductionist approach to the techne (51). But Schiefsky is at pains to explain the author's sustained concern for precision and error--he labels chapters 9 through 12 of the treatise a digression--since no reductionist physician would be in a position to question the practical success of competing physicians (35). However, if we suppose the critic to have been an empiricist from outside the profession, then the attention paid to precision and error makes perfect sense, since the empiricist would have taken issue with any methodology that purported to use general causal understanding to secure practical success in healing patients.

While we might appreciate the spiritedness of medicine's defense, we ought also to sympathize with the critics of medicine. Some Greek physicians were wont to make grandiose epistemic claims with little attention to methodological concerns. The author of the treatise Regimen, who can 'in advance of a patient's falling ill from excess, discern the course that the condition will take' (I II 59-61 Jones), has also

discovered the power to discern the prevailing factors in the body, whether exercise prevails over food, or whether food prevails over exercise, how one ought to cure completely each condition and secure health at the outset so as to stave off disease, provided few serious errors are made. (Regimen III LXVII 17-23 Jones)

Those who wish to follow suit

must before all else know (gnonai) and discern human nature. He must know from what a human being is composed in terms of a first principle (ex arches), and he must perceive by what constituents it has been mastered. For if he doesn't know the composition in terms of a first principle he will be unable to know what comes to be through these components. If he doesn't know what is dominant in the body, he will be incapable of applying beneficial treatment to the body. (Regimen I II 1-10 Jones)

We could not hope for a clearer, more confident statement of the physician's epistemic and technical aspirations. By seeing through to the non-evident natures of things in the past and present, including the internal structures and processes of the body, he can foresee those future events that remain obscure to the unskilled. Moreover, he can influence outcomes with a nearly perfect rate of success--but only after he has grasped the underlying composition of the human body. Regimen happily provides all this and more: the human body, just like the universe as a whole, is composed of fire and water, and it is through the interaction of these that human beings (only apparently) come to be and pass away.

It is difficult to square the theoretical and practical enthusiasm of Regimen with the obvious fact that patients sometimes die despite the best efforts to diagnose and cure them. The author of the Hippocratic On the art frames the critic's objection thus: 'the techne is now blamed because not all regain their health; and those who denigrate it claim that, because some patients succumb to disease, even those who escape do so by chance and not because of techne' (IV 1 Jouanna). If, as the author of Regimen claims, medicine really offers exhaustive knowledge of the body's constituents and their powers, as well as how to cure disease in every case, how are medical failures to be explained? Both On the art and Ancient medicine respond by distinguishing, at least implicitly, between general medical knowledge and knowledge in particular cases of disease. Physicians have knowledge of human anatomy and physiology (On the art X); they know, broadly speaking, what constitute correct and incorrect procedures (V 5-6). But the exigencies of particular cases bedevil the application of this general knowledge. A patient may sabotage himself by disobeying a doctor's orders (VII). He may die before a physician has time to fully diagnose his condition (XI 6). Likewise, a physician may complete his diagnosis and even administer a few remedies, but the patient may be suffering from a disease for which there is no cure (VIII).

Neither of these Hippocratic authors dispute that medicine's principle task is to discover the causes of the various conditions and diseases that affect human beings (Ancient medicine XIX 3 Schiefsky; On the art XI 4 Jouanna). Causes, according to the author of Ancient medicine, are just 'those things that, when they are present, a specific condition arises, but, when their combination changes, the condition ceases,' and all valid medical explanations are to advert to such causes (ibid.). The author's analysis of causation is precise in its principles in the sense that it requires the physician to express necessary relations between theoretical terms. A cause according to Ancient medicine is an (in itself) insufficient but necessary component of a constellation of conditions that is as a whole necessary and sufficient for the effect. So, for example, a predominance of acrid mucus in the nose is cited as the cause of its inflammation, since the inflammation correlates to the described conditions (XVIII 1-2 Schiefsky).

In other ways, however, the Hippocratics reject the Platonic ideal of precision in the general principles of a techne. The above examples show that the author of Ancient medicine does not formulate his general explanatory principles in mathematical terms. This has practical consequences for medicine.

And if it were simple, as has been claimed, so that whatever was strong was harmful and whatever was weaker helped and nourished both the sick and healthy person, then medicine would be an easy affair. For one would need only direct the patient to what was weakest, and this could be done with great certainty. However, it is no less an error, nor does it hurt the person less, if one prescribes what is lacking and deficient in what is sufficient.... For this reason, things are much more complex and require more precision (akribeia). For one must aim at (stochazein) some measure (metron), but there is no measure, neither a number nor any other standard by reference to which you will attain precise knowledge (eisei to akribes), nor could you find one other than bodily sensation (tou somatos ten aisthesin). For this reason, it is a chore to make such a precise observation (katamathein akribos) that only small errors are made here and there. (IX 1-3 Schiefsky)

At first glance, this passage may read like an endorsement of medicine as an essentially stochastic practice. However, the author does not mean that medicine is stochastic either in the Platonic sense that it relies on guesswork where precise knowledge is called for, or in the Isocratean sense that it operates in a metaphysically imprecise domain. (25) The verb stochazein is used here to convey that any goal-directed procedure will aim to apply some specific therapeutic measure between two extremes, that is, at some measure in the sense of 'mean'.

How does the physician determine the correct measured procedure? He knows what the goal is: to achieve a balance of humoral powers in the body. The optimal state of health is one in which none of the phenomenal-chemical properties (eg., sweet, bitter, acid, insipid, and so on) dominates (XIV 4; XX 6). The physician's task will be to compare the patient's current, unbalanced state to the optimal state and deduce therefrom the correct therapy. How will the physician evaluate the patient's current state? Not by using mathematical measurements of the kind a carpenter or shipwright might make. The levels or strengths of these properties cannot be measured except in the loose sense that they may be detected by the diagnosing physician's perceptual faculties. (26) These faculties are not inherently limited or imprecise, as Plato believed. They may, with some effort, become sensitive enough to gauge the patient's condition accurately, though the complexity of variables and the perceptual unavailability of the body's internal systems will sometimes pose obstacles to 'measurement.' However, when the author admits that 'perfect accuracy is rarely seen' (IX 4), this at least concedes that perfect accuracy is achievable, and elsewhere he boasts that 'many areas of medicine have achieved such precision,' thus allowing that therapeutic perfection is possible, at least in principle (XII 8-9).

All the Hippocratic writers under consideration hold that success in medicine requires causal knowledge of the body and its pathologies. Accordingly, their attitudes with respect to theses ii and iii will be determined by their view on whether they deny, like Plato, that such knowledge can be acquired through practical experience. Both On the art and Ancient medicine answer this question decisively, and their conclusions, if not their rationales, are identical: experience alone cannot give a person the sort of knowledge he needs to become a successful physician. Thus, the author of On the art concedes to his critic that 'even those who do not consult a physician can chance upon (perituchein) medicine' (V 2). Accidental cures are possible, but 'this does not actually result in knowing which medical procedures are correct or incorrect, but rather in hitting upon by chance (epituchein) the very treatments that would have been applied had they consulted a physician' (ibid.). A layman may, after the fact, know that fasting cured his condition, but he cannot explain why it did so. His knowledge of medicine, if causal in any sense, will be so only insofar as he can make accurate singular causal statements about particular cases. The physician, on the other hand, uses his causal knowledge to foresee what will result from the application of different remedies and formulates the course of treatment accordingly: 'medicine clearly has and always will have being, both in virtue of things that happen 'because of something' and in virtue of the things it knows in advance' (VI 4). He knows to orthon and to me orthon--the correct and the incorrect (V 6). The author uses articular constructions here to convey that the physician's knowledge is completely general in character. Moreover, experience is inadequate to discover the causes of disease because many diseases cannot be directly encountered. Knowledge of most internal diseases the layman cannot hope to achieve without substantial background knowledge of the body's internal processes and the intellectual capacity to reason from evident signs to underlying conditions (XII 3). But, if he had such background, he would not be a layman.

In short, On the art allows the experienced layman some medical knowledge while insisting that it is useless for the purposes of prescribing successful treatment. Ancient medicine will not allow the experienced layman even this much.

I know that, if the patients happen to have done something atypical around this day [of their decline in health], either bathing or walking around or eating something strange (supposing it had been better to do all these than not), nonetheless many doctors, just like laymen, put the blame (aitia) on one of these things and remain ignorant of the cause (to aition), canceling what treatment might have proven most beneficial. Instead, they should know what an untimely bath does or what fatigue does. For the distress associated with either one of these is never associated with the other, nor with those that originate from repletion or from such and such a food. So whoever does not learn how these things stand in relation to a human being will be able neither to know their results nor to use them correctly. (Ancient medicine XXI 2-3 Schiefsky, emphasis added)

One must have grasped the general underlying causal relations that govern the human body's interaction with its environment before one is qualified to say that some event was the cause of a particular case of illness. The mere spatiotemporal contiguity of an atypical occurrence is not sufficient to judge its causal significance. Experience does not just make it plain somehow that the man who stays out late carousing and comes down with malaria the next day does so because of his carousing. And as the author makes clear, the same point holds for all examples in which a layman would claim to have stumbled upon the cure for a disease by chance. In fact, he could not know that he had done so unless able to cite the underlying necessities in virtue of which his treatment brought about the patient's recovery.

But if practical experience in itself is inadequate for developing a causal theory of health and disease, we must ask the Hippocratics, who are so confident of their own causal knowledge, to supply a coherent alternative. This returns us once again to questions of medical methodology, questions that are only complicated by the failure rates of physicians. The Hippocratics, however, have plausible responses at the ready. The success of every art is sensitive to the obstacles posed by particular circumstances (On the art XI 7 Jouanna), and some doctors really get better results than others (Ancient medicine I 2 Schiefsky), even if the vast majority are incompetent (IX 5). Still, practical success shows only that a physician has applied the correct treatment to his patient. It does not by itself show that he knows why this treatment was successful. Of course, he will have a theory that he believes will explain everything--if the Hippocratic Corpus is a reliable indication, there was no shortage of speculative medical theories in ancient Greece--but this is just the point. As Isocrates might have put it, if physicians had genuine knowledge of their field, they would agree more often than they do.

In order to show that a medical techne exists, then, the Hippocratics must do more than find a physician with a high success rate; at minimum, they must also demonstrate the truth of his particular causal theory to the exclusion of its competitors. Thus, the Hippocratic texts that show some sensitivity to the problem of theoretical diversity develop ultimately into polemics--attacks on physicians who endorse competing medical methodologies. The stated raison d'etre of Regimen in acute diseases, for example, is to criticize and correct the mistakes made by those who wrote the medical treatise Cnidian sentences (I 1-10 Jones). His chief complaint is methodological: the Cnidians classified diseases according to symptoms instead of according to their underlying causes (III 12-16). After remarking that medicine has a bad reputation among laypeople, the author explains that 'if in acute diseases practitioners differ from each other to such an extent that whatever one doctor thinks best to apply are considered bad by another, it is likely to be said in such cases that the techne is similar to divination' (VIII 5-11 Jones). The author of Ancient medicine laments the proliferation of theories that 'reduce the explanatory principle of disease and death to be the same in all cases, hypothesizing one or two things' (I 1 Schiefsky). Similarly, Nature of man opens with an attack on philosophical types who claim that a human being is composed of only one substance--air, fire, water or earth--though they cannot agree on what that is, and each of them 'adduces evidence and proofs that are not (estin ouden)' (I 15-19). But doctors have the same problem, he admits, 'some saying that a human being is blood, others claiming bile, and still others phlegm' (II 2-4).

The polemics in Ancient medicine and Nature of man criticize competing theories on strikingly similar methodological grounds. Both wash their hands of reductionism in the explanation of health and disease, and they do so because reductive theories involve claims that have little, if any, connection to empirical observation. 'Whoever is accustomed to hearing people speak concerning the human nature outside of its relevance to medicine,' begins Nature of man, 'to him will the present discourse prove of no interest. For I say neither that a human being is completely air, nor fire, nor water, nor earth, nor anything else that is not apparent (phaneros) as a constituent of a human being' (I 1-8 Jones). (27) In contrast with such theories, the author proposes that health and disease are explained by perturbation of the proper proportion of the body's four humors, namely blood, phlegm, yellow bile and black bile. In practice, of course, this humoral theory will fare no better than the speculative theories the author rejects, but it does have the merit of consistency in that its explanations contain references only to entities that are observable constituents of the human body. (28)

Ancient medicine takes a similar but more sophisticated approach to the problem of non-evidence. Instead of defending the role of non-evident entities in medical theory and practice, the author champions a techne in which explanations rely on causal principles verified by observation, thereby distancing his theory from those that depend on nonevident postulates.

Medicine does not need any new-fangled hypothesis as do non-evident (aphanes) and puzzling matters, concerning which it is necessary to use a hypothesis, should someone try to say something of significance, e.g., investigations concerning the things in heaven and below the earth. If someone says he has knowledge that these are thus and so, it would not be clear either to him who said it or to his listeners whether these things were true or not. (I 3 Schiefsky)

The author struggles to distinguish his brand of medicine from that which asserts that 'it is impossible for someone to have knowledge of medicine who does not know what a human being is' (XX 1). Such questions are best left to natural philosophers like Empedocles, who, it turns out, won't be able to answer them without abandoning their theoretical postulates for the empirical method advocated in Ancient medicine (XX 2). (29)

But given the obvious technological limitations of Greek medicine, all ancient theories of human health and disease inevitably make reference to internal processes, structures, entities or events that cannot be directly observed, making it difficult, perhaps impossible, to verify their truth. This potential objection demands an answer from the Hippocratic defenders: how can medicine come by a sound explanation of health and disease when their causes are often unobservable? The problem was not lost upon the author of On the art, who, following a hasty sketch of the human body's internal anatomy and a crude general pathology, takes up the epistemological challenge with zeal.

Of course, a person who sees only with his eyes cannot know any of the things just mentioned. For this reason, I have referred to them as non-evident (adela), and so they have been judged by the art. Though they are non-evident, these diseases have not prevailed. Rather, they have been prevailed over where possible. (XI 1 Jouanna)

In fact, the problem of non-evident diseases occupies the author for the remainder of the text. How can we determine the internal causes of a particular patient's disease--or disease in general--when we cannot perceive them with our senses? The author's solution: we use our intellects (gnome), which provide us with a kind of second sight (XI 2). (30) Reason allows the physician to draw inferences about the internal structures and processes of the body on the basis of the signs and symptoms he observes (XII 2).

We have seen that the authors of Ancient medicine and Nature of man demonstrate a desire to distinguish themselves from theorists who make use of reductive, non-evident postulates. In On the art we find a defense of medicine's claim to have knowledge of entities and processes that cannot be directly observed. This common concern with the non-evident and its implications for medical theory suggests that all are responding to an attack on medicine's methodology that used non-evidence to question whether physicians could really know all that they claimed to know about the body. This appears yet more probable in light of Protagoras' documented criticisms of the techne of geometry based on the non-evidence of the theoretical entities--specifically, intelligible circles--postulated by geometers. Indeed, we should expect empiricists about techne to be favorably disposed toward the observable considering the inherent difficulty of gaining practical experience of what cannot be experienced at all. The Hippocratic reaction is remarkably uniform: non-evidence may be an obstacle to knowledge, but it is not insurmountable, so long as medicine is practiced correctly.

IV The New Narrative

Contrary to the Standard Story, according to which the Hippocratic defenders of techne came under attack from rationalist critics, there is ample evidence to suggest that the Hippocratics shared with Plato an appreciation for the role of causal theory in the development of techne. In this sense, Plato is more than justified when, in the Phaedrus, he compares his own theory of techne to the method espoused by Hippocrates.

 Consider, then, what both Hippocrates and true argument say about
 nature. Isn't this the way to think systematically about the nature
 of anything? First, we must consider whether the object regarding
 which we intend to become experts is simple or complex. Then, if it
 is simple, we must investigate its power (dunamis): What things
 does it have what natural power of acting upon? By what things does
 it have a natural disposition to be acted upon? If, on the other
 hand, it takes many forms, we must enumerate them all and, as we
 did in the simple case, investigate how each is naturally able to
 act upon what and how it has a natural disposition to be acted upon
 by what. (270c-d)

The Hippocratics considered in this paper are, like Plato, fully invested in giving causal accounts of their technical domain. Hippocratics like the authors of Ancient medicine and Nature of man differ sharply from Plato, however, in their belief that precise causal accounts of nature can be arrived at through the direct observation of phenomena without relying heavily on claims about unobservables. (31)

Moreover, and again in contrast to Plato, the Hippocratics are committed to causal knowledge not because it has any special value in itself but because they believe it is the only way to succeed regularly at healing human beings. Mere experience cannot provide such knowledge, and so, contra Plato and Isocrates, it cannot be at all successful. Practical success, for the Hippocratics no less than for the empiricists, is the mark of a true techne. This explains why the Hippocratics exhibit such acute sensitivity to questions of practical precision. Frequent failure suggests that a doctor is not expert, and that is bad enough. But if no doctor has a respectable success rate, then the entire causal-theoretical approach to the techne is threatened.

If Isocrates' attack on sophists is any clue, empiricists did indeed question the general effectiveness of experts who made knowledge claims about their domain. Moreover, they were skeptical about the knowledge claims themselves. Isocrates doubted whether the future could ever be known, perhaps in part because the world is metaphysically imprecise. Protagoras doubted whether anything perceptually non-evident could be known. Either way, the objectives and methods of Hippocratic medicine faced a challenge from empiricism that no rationalist would have posed. And, if I am largely correct in my analysis of the ancient evidence, no rationalist ever did.

Moreover, the Hippocratics met the empiricist challenge not by accommodating the empiricist's relaxed attitude toward precision in the principles of a techne. They maintained instead that the domain of medicine is subject to causal necessity, governed by precise regularities that allow at least for the possibility of precision in practice. There is no question that knowledge of the body's nature is their goal, even if they fail--as they most surely did--to meet it. Moreover, though medicine often failed also to achieve its practical goals, the Hippocratics did not take this as a sign that the techne of healing was a different, inherently less precise kind of techne. Medicine differs from other technai only in degree of difficulty. The doctor's job is more complex than the shipwright's or the carpenter's, with much more potential for slip between cup and lip. That is why, as the Hippocratic defenders of medicine would no doubt agree, the successful physician deserves the highest respect, not the harshest criticism. (32)

Works Cited

Burnyeat, M. F. 1982. 'Idealism and Greek philosophy: What Descartes saw and Berkeley missed'. Philosophical Review 91:3-40.

Cooper, John M. 1997. Plato: Complete Works. Indianapolis: Hackett.

--. 2004. 'Method and Science in On Ancient Medicine'. In Knowledge, nature, and the good: essays on ancient philosophy. Princeton, NJ: Princeton University Press.

Dean-Jones, Lesley. 2003. 'Literacy and the charlatan in ancient Greek medicine'. In Harvey Yunis, ed., Written texts and the rise of literate culture in ancient Greece. New York: Cambridge University Press.

Diels, H. and W. Kranz. 1951. Die Fragmente der Vorsokratiker. 6th ed. Zurich: Weidmann.

Gagarin, Michael. 2002. Antiphon the Athenian: Oratory, Law, and Justice in the Age of the Sophists. Austin: University of Texas Press.

Gomperz, Theodor. 1910. Die Apologie der Heilkunst. Eine griechische Sophistenrede des funften vorchristlichen Jahrhunderts. 2nd ed. Leipzig: Viet & Comp.

Hankinson, R.J. 1992. 'Doing without hypotheses: the nature of Ancient medicine'. In J. A. Lopez Ferez, ed., Tratados hipocraticos (Estudios acerca de su contendo, forma y influencia) Actas del VII colloque International Hippocratique. Madrid: Universidad Nacional de Educacion a Distancia.

--. 1995. 'Experience and expertise: Aristotle's conception of Techne'. Unpublished photocopy.

Hutchinson, D. S. 1988. 'Doctrines of the mean and the epistemology of skill in ancient medicine and ethics'. In R. J. Hankinson, ed., Method, medicine, and metaphysics: Studies in the philosophy of ancient science. Edmonton, AB: Academic Printing & Publishing.

Jouanna, Jacques. 1988. Hippocrates V. Collection des universites de France, Publiee sous le patronage de l'Association Guillaume Bude. Paris: Societe d'edition 'les belles lettres'.

Jones, W. H. S. 1923-1931. Hippocrates I-IV. Loeb Classical Library. Cambridge, MA: Harvard University Press.

--. 1946. Philosophy and Medicine in Ancient Greece. Baltimore, MD: The Johns Hopkins Press.

Lloyd, G.E.R. 1991. 'The medical techne in the fifth and fourth centuries'. In Alan C. Bowen, ed., Science and philosophy in classical Greece. New York: Garland Publishing, Inc.

Mackie, J.L. 1975. 'Causes and Conditions'. In Ernest Sosa, ed., Causes and Conditionals. Oxford: OUP.

Norlin, George. 1980. Isocrates I-III. Loeb Classical Library. Cambridge, MA: Harvard University Press.

Roochnik, David. 1996. Of art and wisdom: Plato's understanding of techne. University Park, PA: Pennsylvania State University Press.

Schiefsky, Mark J. 2005. Hippocrates' On Ancient Medicine. Leiden: Brill. Schumacher, Joseph. 1963. Antike Medizin: die naturphilosophischen Grundlagen der Medizin in der griechischen Antike. Berlin: De Gruyter.

Vlastos, Gregory. 1969. 'Reasons and Causes in the Phaedo'. Philosophical Review 78: 291325.

Woodruff, Paul. 1985. 'Didymus on Protagoras and the Protagoreans'. Journal of the History of Philosophy 23: 483-497.

--. 1994. 'Eikos and Bad Faith in the Paired Speeches of Thucydides'. Proceedings of the Boston Area Colloquium in Ancient Philosophy 10: 115-45.

(1) I will render the Greek word techne either as the transliterated form techne or as the English 'expertise', except where conventional usage demands an alternative, as in the title of the Hippocratic treatise On the art. For the adjectival form, I will use the English 'technical', and so on for adverbs and other parts of speech.

(2) In a fine account of the disagreement between Plato and Isocrates (and ultimately Aristotle) over technical education and the epistemic aims of techne, Hutchinson discusses key contributions of pre-Socratic, sophistic and medical thought to the techne question. The current essay draws heavily (but not uncritically) on Hutchinson's conclusions for inspiration in its examination of the Hippocratic Corpus.

(3) Roochnik seems to endorse this version of the story in general terms (42-57). Hutchinson sees in Ancient medicine the very model of anti-Platonic stochasticism (43). More recently, Schiefsky has suggested that the author of Ancient medicine is responding to criticisms originating with the causal reductionists he criticizes, as though the author were worried about distinguishing himself from traditional non-rational medical practices (12-13; 51; 53-4).

(4) I place the terms rationalism and empiricism in cautionary quotes. As will become clearer, rationalism and empiricism in relation to the theory of techne differ significantly from rationalism and empiricism in the modern philosophical sense.

(5) For a useful treatment of apprenticeship and training in the medical techne, see Dean-Jones 108-11.

(6) The exact senses of these terms, in the Platonic Corpus and elsewhere, are a matter of considerable debate. For present purposes, I will assume that in demanding causal knowledge Plato is requiring of a technical expert that he be able to articulate the causal powers that distinguish one kind of thing from another (cf. Phaedrus 270 c-d).

(7) Plato often distinguishes between experience and routine, that is, between the epistemic and manual components of practical experience, while my tendency will be to collapse them. This is justified insofar as the experience of particular cases is obviously necessary for manual training in a field.

(8) This is the so-called 'tripartite' theory of techne described in Hutchinson.

(9) It is difficult to settle on a definitive translation of the verb eikazein and its cognates, including the common adjective eikos. Eikos is often rendered as 'probable', though the Greeks surely didn't use it to signal genuine statistical probability as we conceive of it today. Woodruff 1994 has suggested that an argument from eikos is one based on a defeasible inference, but Plato here seems to be playing on the root meaning of eikazein, 'to represent by an image'. The idea is that the empiricist must proceed by comparing the present case to past cases he has seen; hence Plato's charge in the Gorgias that the empiricist relies on memory, not knowledge. See also Gagarin 2002 for the use of eikos in classical rhetoric, especially in Antiphon (112-18).

(10) Plato includes medicine among the inherently stochastic procedures. To what extent this denies technical status to medicine is a substantial question, especially given Plato's frequent use of medicine as a paradigmatic techne in dialogues such as the Gorgias, Republic, Phaedrus, Statesman, and Laws, not to mention his special treatment of the subject in the Timaeus. I have yet to discover in the critical literature a completely satisfactory resolution of this apparent tension in Plato's views. I hope to treat the problem myself in a future paper.

(11) Techne and tuche are routinely contrasted in Greek thought. For a helpful recent discussion of the antithesis, see Schiefsky 5-12.

(12) Writing was sometimes held up as the paradigm of a highly beneficial and scrupulously precise techne (Plato, Protagoras 326d; Aristotle, NE III 3, 1112b1-3). At the other end of the spectrum lay navigation, a highly beneficial yet frustratingly imprecise techne, with which medicine was often paired (Ancient medicine IX 4-5 Schiefsky; Plato, Philebus 56a-b; Aristotle, NE II 2, 1104a10).

(13) In paragraph 2, Isocrates asserts that knowledge of the future is impossible for human beings given our nature, which seems to place epistemological, in addition to metaphysical, limitations on our potential to formulate universal generalizations over technical domains. But he adds that even the gods dispute the future, which suggests that the world is structured so as to be fundamentally unknowable.

(14) Hankinson 1995 draws the helpful distinction between technai that are epistemically stochastic and those that are ontologically stochastic (16). Isocrates himself never uses the term.

(15) Cf. Anaxagoras DK 59 B21a: opsis adelon ta phainomena (the perceptible is a sight of the non-evident).

(16) This is consistent with Burnyeat's view that Protagoras (at least as presented in Plato's Theaetetus) could not have been an idealist because he does not hold that the objects of perception are mind-dependent (1982, 4-14).

(17) As did Gomperz in the first edition of his study of On the art (1910, 30ff).

(18) Most notably, Jouanna has reinforced this reading in his introduction to the Bude edition of On the art (1988, 174).

(19) It is perhaps significant that, in the Theaetetus, the imaginary Protagoras fends off Socratic attacks on the homo mensura by characterizing technical expertise not as an achievement in knowledge but rather as practical success at changing appearances (166d-7d).

(20) The relevant Hippocratic texts are generally agreed to date anywhere from the last quarter of the fifth century BCE (Schiefsky 2005, 63-4). Certain about particular dates of composition is largely impossible and in any case will not affect the present argument.

(21) Schiefsky's notion that the author of Ancient medicine is taking sides in a dispute internal to the medical community can't explain his concern with the peculiar charge that medicine 'is not' or doesn't exist at all (ibid.). Certainly no doctor would make such a claim.

(22) Reading hoti with A and Jouanna instead of ho ti.

(23) Compare against Plato's report at Sophist 232d (cited earlier) that Protagoras composed antilogies against 'all technai in general and each techne specifically'.

(24) Schiefsky distinguishes between the author's use of akribeia and to atrekes (34). The former, he argues, is used to refer to the quality of the doctor's knowledge, while the latter is used to qualify the accuracy of his prescriptions. This passage suggests the Hippocratic author perceives the meanings of the two terms as more fluid.

(25) Of course, if we observe the letter of Plato's law laid down in the Philebus that any non-mathematical techne is stochastic, then perhaps medicine would count as stochastic. But if we adhere to the Platonic spirit that the principle of a techne ought to be precisely articulated and allow for the possibility of general precision in practice, then surely the Hippocratic author believes medicine to pass the test.

(26) Schiefsky breaks with the mainstream interpretation to argue that the bodily sensation at issue here is the patient's (esp. 186-92; 196-200). In short, Schiefsky claims that the doctor's only way of knowing with precision the patient's physical condition before treatment is by observing his reaction to the treatment. I will not react here to any specific arguments made by Schiefsky in his prolonged defense of this view, though I will point out that this would make the physician's 'measurements' not just inherently imprecise (as Schiefsky claims contra the clear evidence in this passage) but downright impossible to characterize as measurements in any meaningful sense.

(27) Bizarrely, the author's allergy to 'philosophical' medicine is so severe that he rejects two elements, air and water, which the available perceptual evidence would suggest play an important role in physiological processes.

(28) This supposes that we give the author the benefit of the doubt in the case of black bile, which does not correspond obviously to any known bodily fluid.

(29) For different interpretations of the role of hypothesis in medicine as the Hippocratic author understands it, see Hankinson 1992, Cooper 2004, and Schiefsky, especially pages 118-126.

(30) Cf. Democritus DK 68 B11.

(31) Hankinson 1992 and Cooper 2004 discuss Ancient medicine's attempt to develop a theory of medicine built on and faithful to empirical observation. Plato is notoriously pessimistic about the value of observation (for example, at Phaedo 99d-e), hoping instead to discover a general teleological principle from which knowledge of nature can be deduced (Phaedo 97c-e; Timaeus 46d-e).

(32) I would like to thank Andrew Cross, Matt Evans, Dorothea Lotter, Sam Hyle, Getty Lustila, and the anonymous referees at Apeiron for their help in refining the ideas presented in this paper, which developed out of work done under the supervision of Lesley Dean-Jones and Jim Hankinson.

Joel E. Mann

Department of Philosophy

St. Norbert College

100 Grant Street

De Peve, WI 54115-2099

USA

joel.mann@snc.edu