¹ I borrow this term from modern philosophy, but I must warn the reader that the ancient doubts about other peoples mental or psychological states should not be confused with the so-called problem of Other Minds as it appears in post-Cartesian philosophy. The formulation of the problem of Other Minds relies on the Cartesian assumption that physical things and mental entities belong to different ontological realms and that the latter may exist and be conceived independently from the former. It consists primarily in the ontological problem whether there exist minds other than my own. In its classical version it asks the question whether bodies other than my own, but resembling my own, which I perceive to move, make intelligible noises, change facial expressions, and, in brief, behave in the way I perceive myself behaving, are each animated by a mind and experience mental states. A host of epistemological and semantic issues are appended to the ontological question. Predominant among them are arguments purporting to show or to refute the validity of the inference from the behaviour of a body, which is public, to the existence or to the content of mental or psychological states, which, were it the case that other bodies are animated by minds, would be private. The whole modern issue of Other Minds rests on the presupposition that there is something mysterious and peculiar about mental states, as opposed to bodily or physical states, which guarantees to the percipient direct cognitive access to their content. Although there are similarities between the ancient doubts about our access to other peoples mental states and the modern issue described above, the differences between them are crucial and parallels between the two involve risks of misunderstanding.

² For adequate ways of speaking about the mind and mental states in the context of a philosophical discussion of Hellenistic texts see the comprehensive study of Annas, J., Hellenistic Philosophy of Mind (Berkeley, Los Angeles, and Oxford, 1992).Google Scholar

³ The translations of Greek and Latin passages cited in this paper are my own, unless it is indicated otherwise.

⁴ For a philosophical interpretation of that claim, see Tsouna-McKirahan, Voula, ‘The Cyrenaic theory of knowledge’, Oxford Studies in Ancient Philosophy 10 (1992), 161–92 (henceforth cited as Tsouna 1992).Google Scholar

⁵ See Caujolle-Zaslawsky, F., Qu'est-ce que la methode sceptique? (Paris, 1982), These de Doctorat dEtat, vol. 1, p. 196.Google Scholar

⁶ An interesting argument to that effect might be made regarding the first part of Platos Theaetetus, which discusses the definition of knowledge as perception (see, especially, 153e-154a, 159e-160a, 166c), and I am grateful to A. Hobbs for pointing this out to me. However, it seems to me that, rather than raising questions about other minds, Protagorean relativism implies that such questions are meaningless. If whatever I perceive is true for me and if this kind of truth (i.e truth relative to a particular perceiver at a particular time) is the only truth there is, it makes no sense to ask how I know that another perceiver who says that he perceives something red really perceives something red. For if it appears to him that he perceives something red, it is true for him that he perceives something ted, and this is all the truth there is about his perception. And if it appears to me that it appears to some other perceiver that he sees something red, this appearance is true for me although not true for any other perceiver at a given time, including the person whom I perceive as seeing something red. I discuss this topic in detail in an essay entitled Remarks about other minds in Greek philosophy, Phronesis (forthcoming).

⁷ The evidence about ancient Physiognomy is collected in Foerster, R., Scriptores Physiognomonici (Leipzig, 1893).Google Scholar

⁸ This treatise will be often cited as ps. Arist.

⁹ On Loxus date and doctrine, see Misener, G., ‘Loxus: physician and physiognomist’, Classical Philology 18 (1923), 1–22.CrossRefGoogle Scholar

¹⁰ The Greek paraphrase of Polemos work was composed by Adamantius of Alexandria in the fourth century.

¹¹ For a brief discussion of the nature and contents of these treatises, see Evans, E., ‘Physiognomies in the ancient world’, Transactions of the American Philosophical Society N.S. 59 (1969), 6–17 (henceforth cited as Evans).CrossRefGoogle Scholar

¹² See ps.–Arist. 806a, which stresses that not all affections (pathemata) are of interest to the physiognomist, and that not all bodily characteristics can be used as physiognomic signs.

¹³ On the zoological method, see ps.–Arist. 809aff.: the characteristics of the main animal types reappear in the list of qualities related to physiognomic signs in human beings. On the ethnological method, see ps.–Arist. where the Corinthians and the Leucadians are taken to exemplify the marks of the small–minded or petty man; see also ps.–Arist. where the ethnological observation that men living in the North are brave and stiff–haired while men living in the South are cowardly and soft–haired is preceded by zoological observations supporting the thesis that soft hair is a sign of timidity, while stiff hair signifies courage. On the methods of the pathognomists, see e.g. ps.–Arist.

¹⁴ It has been suggested that the philosophical method of physiognomical analysis presented in the first part of the pseudo–Aristotelian handbook may be related to the method described in Arist. An. pr. 70b6–39. On this, see Evans, p. 8. However, I do not think that this is plausible, for it is explicitly stated in the pseudo–Aristotelian handbook that nobody has practised the philosophical method of Physiognomy; on the other hand, the method that Aristotle describes and the problems that he raises clearly concern the traditional ways of engaging in Physiognomy.

¹⁵ Some of the philosophical doctrines mentioned below are briefly discussed in Evans, pp. 17–28.

¹⁶ See especially the treatises On the Nature of Man and Airs, Waters, Places.

¹⁷ However, Plato does not defend or discuss these doctrines in detail.

¹⁸ On this, see Evans, pp. 21ff.

¹⁹ Galen studied under the physician Pelops in Smyrna, a centre for rhetoric and a city that held Polemo of Laodicea in high regard. Apparently, one of Polemos compositions was a handbook on Physiognomy intended mainly for professors of rhetoric. In relating the rules of oratorical delivery with physiognomical principles, Polemo moves within a tradition that can be considered to start with Aristotle (Rhet.).

²⁰ On Galen's appeal to Hippocrates, Plato, and Aristotle in Quod animi mores in support of the thesis that the faculties of the soul follow the mixtures of elements in the body, see Lloyd, G. E. R., Scholarship, authority and argument in Galen's Quod animi mores, in Manuli, P. and Vegetti, M. (edd.), Operepsicologiche di Galeno (Napoli, 1989), pp. 11–42; see also G. E. R. Lloyd, Galen on Hellenistics and Hippocrateans: contemporary battles and past authorities, in Methods and Problems in Greek Science (Cambridge, 1991), pp. 398–416. As Lloyd points out, there are obvious weaknesses in Galens arguments. Many of them are caused by his concern to present himself as a faithful follower of Hippocrates and to trace his own doctrine and medical art back to that ideal.Google Scholar

²¹ According to several sources, it was Pythagoras, not Hippocrates, who was the father of Physiognomy.

²² Evans stresses that Xenophon differs from other historians of the fifth and fourth centuries in that in his historical works he uses descriptions of physical appearance for purposes of characterization, while they habitually do not (see Evans, p. 46). Xenophons emphasis on the characterological significance of iconistic portraiture may be partly due, I suggest, to his awareness of the works of other Socratics as well as to his ethical interests.

²³ Polemo concentrated on their study and probably added some types to the standard Peripatetic lists. See chapters 53–70 of his treatise, comprising studies of the daring man, the timid man, the intellectual, the man close to the end of his life without being ill, and so on.

²⁴ One may recall Seneca's description of the angry man in his treatise De ira and the way in which he exploits physiognomical elements for ethical purposes. See the remarkable analysis of Senecas discussion of anger by Anderson, W. S., Anger in Juvenal and Seneca, University of California Publications in Classical Philosophy (Berkeley and Los Angeles, 1964), pp. 127–95, and especially pp. 149–73.Google Scholar

²⁵ See Frischer's, B. stimulating (although controversial) discussion in The Sculpted Word: Epicureanism and Philosophical Recruitment in Ancient Greece (Berkeley, Los Angeles, and London, 1982). For the evidential value of statues see also Cameron.Google Scholar

²⁶ The epistemological doctrine of the Cyrenaics was probably developed by Aristippus the Younger, who may have been born around 380–370 B.C. If so, the doubts of the Cyrenaics concerning the knowledge of other minds may be placed towards the middle of the fourth century B.C. Theodosius argument about other minds was probably formulated much later, after the revival of the Pyrrhonist movement by Aenesidemus in the first century B.C.

²⁷ The sources attest that, according to the Cyrenaics, the pathe are related to motions in the flesh or in the soul (Cic. Defin. 2.39; Sext.Emp. PH. 1.215; D.L. 2.90). Some authors appear to construe this relation as an identity (D.L. II 86, Suda II SS3, 4f.). However, there is reliable evidence that the Cyrenaics in fact distinguished between the physical motions associated with the pathe and experiences, such as individual feelings of pleasure and pain (D.L. 2.85, Euseb. Prep.Evang. 14.18.32).

²⁸ On this point, see Tsouna 1992.

²⁹ Some examples of reductionist positions are Lewis, D. K., ‘An argument for the identity theory’, Journal of Philosophy 63 (1966), 17–25;CrossRefGoogle ScholarSmart, J. J. C., Philosophy and Scientific Realism (London, 1963); D. M. Armstrong, A Materialist Theory of the Mind (London, 1968); and D. C. Dennet, Content and Consciousness (London, 1968). A forceful argument to the effect that reductionist theories miss what makes mental phenomena unique and uniquely difficult to account for is presented by T. Nagel in his widely influential article What it is like to be a bat?, Philosophical Review 83 (1974), 435–50.Google Scholar

³⁰ Aside from the physiognomical handbooks, such comments also occur in historical and literary writings. On misleading physiognomical signs, see e.g. Suet. Tib. 68.3: Augustus points out that Tiberius arrogant features are a fault of nature, not of mind, and that they belie the reality within. See also the satirists, e.g. Juvenal 2.8–16; 9.1.

³¹ See e.g. ps.–Arist.814b: generally speaking, clearest are the signs that appear in the most favourable position and that occur in those regions of the body (topol) in which there is greatest evidence of the intelligence (phronesis) of the individual.

³² See e.g. Loxus views concerning the colours of the eyes: Loxus says that three colours of the eyes are best. The first kind of eyes that he mentions are very light blue (charopoi), which, he claims, fall between black eyes and grey eyes. The second kind he identifies as goat–eyes, which, he says, are closer to grey but whiter, and the third kind he identifies as nearly black. He says that greasy eyes are gentle and prone to lust and tears. He proclaims dry eyes to be shameless; and he says that excessively grey eyes lack courage, but are marked by shamelessness (Anonym. Physiog. ch. 81).

³³ See Aristotles remarks in An. pr. 70b and the criticisms in ps.–Arist. 805a–806a.

³⁴ See Adamantius, Physiogn. A, ch. 3 (reporting the views of Polemo): the combinations of signs make a very big difference to the art. For the majority of the dispositions and thoughts of people are understood from the combination of signs.

³⁵ See e.g. Aristotles discussion of the physiognomic clarity of the eyes (Hist. anim. 491b) and his remark concerning the hairy ears: although moderately hairy ears are the best for hearing, they carry no physiognomic significance (Hist. anim. 492a). These statements anticipate remarks in the physiognomical handbooks from the third century B.C. onwards.

³⁶ The freedom with which physiognomical ideas allow the move from the physical to the mental, and also from the mental to the physical, is well illustrated in areas marked by the influence of physiognomical principles. Take an example from Hellenistic art (which owes much to Physiognomy and to Peripatetic typology). The spectator of a portrait is expected to move from the body to the mind and to read on the external features of a portrait the character of the person depicted. But in creating the portrait, the artist often follows the opposite procedure: the Elder Philostratus attests that a painter who had never seen the beautiful Pantheia but had only read Xenophons description of her character depicted her as from her soul he divined her to be.

³⁷ This is not the only physiognomical angle from which the bodily features of Socrates have been examined. According to Aristoxenus, a contemporary of Socrates attested that the figure, lips, and voice of Socrates reveal his superb power of persuasion. On this, see Evans, p. 51.

³⁸ This seems to have been the central theme of Phaedo's dialogue Zopyrus. See my article on Phaedo in the Encyclopedia of Classical Philosophy.

³⁹ On predictions concerning the duration of ones life, see Plin. Nat. 11.274. On metoposcopy, See Plin. Nat. 35.88.

⁴⁰ This is the predominant goal of Physiognomy in the Middle Ages and in the Renaissance. See C. Camden, The minds construction in the face, Renaissance Studies in Honor of Hardin Craig (Palo Alto, 1941), pp. 208–20. Camden aptly connects the influence of Physiognomy in that period with the belief of the average Elisabethan in the results of the new learning and with the craving to acquire short cuts to the absolute, back–stairs approaches to certainty, get–rich–quick methods of acquiring the truth about the people that one deals with in everyday life.

⁴¹ My discussion of the Cyrenaic doubts about other minds and of the ways in which they differ from the modern problem of Other Minds owes much to the comments of Myles Burnyeat. I presented earlier versions of this paper at the meetings of the American Philological Association in Atlanta (December 1994) and at the University of Warwick, and I learnt much from the audiences remarks on both occasions. David Blanks suggestions forced me to rethink the arguments in section HI of the paper and to give them a tighter focus. Richard McKirahans comments on the philosophical way of practising Physiognomy and on several Aristotelian passages helped me clarify methodological issues. I should like to warmly thank them all.