Jump to ContentJump to Main Navigation
Universe and Inner Self in Early Indian and Early Greek Thought$

Richard Seaford

Print publication date: 2016

Print ISBN-13: 9781474410991

Published to Edinburgh Scholarship Online: January 2018

DOI: 10.3366/edinburgh/9781474410991.001.0001

Show Summary Details
Page of

PRINTED FROM EDINBURGH SCHOLARSHIP ONLINE (www.edinburgh.universitypressscholarship.com). (c) Copyright Edinburgh University Press, 2020. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in ESO for personal use. Subscriber: null; date: 28 March 2020

Soul chariots in Indian and Greek thought: polygenesis or diffusion?

Soul chariots in Indian and Greek thought: polygenesis or diffusion?

(p.149) 10 Soul chariots in Indian and Greek thought: polygenesis or diffusion?
Universe and Inner Self in Early Indian and Early Greek Thought

Paolo Magnone

Edinburgh University Press

Abstract and Keywords

This chapter discusses the similarity between the allegories of the soul as chariot in Plato's Phaedrus and the Katha Upanishad. It begins by investigating the methodological assumptions underlying such cross-cultural comparison in the absence of pertinent historical documentation. Then the congruences and discrepancies between the two texts are reviewed. The allegory is integral to Upanishadic thought in a way that is unparalleled in Greek thought, and this supports the conjecture of diffusion in a westward direction. The paramount difference between the two texts is the idle passenger, absent from the Phaedrus but central to the allegory in the Katha Upanishad. This difference is significant as a watershed between Upanishad-based Indian and Plato-influenced Greek philosophy.

Keywords:   Soul, Chariot, Plato Phaedrus, Katha Upanishad, Diffusion

Not only the classicist, but even the layman with a casual interest in Greek philosophy is familiar with the allegory which Plato employs in the Phaedrus to describe the nature of the soul in terms, as he says, that are ‘within human power’:

Let [the soul] be likened to the composite inborn power of a pair of winged horses and of a charioteer … (246a)

Both classical scholars and cultivated laymen alike, on the other hand, have seldom been aware of a strikingly similar allegory occurring in one of the most celebrated works of the final period of Vedic literature, the Kaṭha Upaniṣad:

Know that the Self is the rider in a chariot, and the body is the chariot; and know that the intelligence is the charioteer, and the mind is the bridle. They say that the senses are the horses, and the sense objects are their lanes …

(KaU 1.3.3)

For their part, indologists have taken due notice of the puzzling similarity from early on, albeit with differing assessments. Already a century ago, in connection with the Kaṭha passage, Keith observed that ‘the contrast with the Platonic metaphor of the Phaidros is as obvious as the parallel’, further on passing his judgement that in spite of the interesting parallelism ‘the details of the two [metaphors] are perfectly distinct, for Plato uses the conception to illustrate the struggle between the rational and the irrational elements in the soul, and his distinction of θύμος‎ and ἐπιθυμία‎ has no real parallel in the Upaniṣads’.1 On the other hand, Belvalkar and Ranade evidently did not share his caution, as they enthusiastically aver that ‘the extraordinary resemblance of the two descriptions down to the smallest details staggers us, and we must confess we do not know how to account for it’.2 Almost (p.150) in between there is Radhakrishnan’s opinion that ‘in spite of difference in details, the Kaṭha and Plato agree in looking upon intelligence as the ruling power of the soul … and aiming at the integration of the different elements of human nature’.3 More recently McEvilley, who must be credited with the first serious attempt to posit with amplitude and lucidity the question of possible reciprocal influences between early Greek and Indian thought in his path-breaking essay on The Shape of Ancient Thought, confines himself to observing that ‘the similarity in imagery is intriguing’ but answers Friedländer’s4 wondering whether the figure might have travelled from the Far East to Plato with the milder suggestion of a possible common Indo-European heritage.5

This last remark introduces us to our subject: are the two metaphors linked by a process of westward diffusion, or did they originate independently,6 possibly as independent developments of a common Indo-European stock? This paper purports to show that the former hypothesis (i.e. the hypothesis of westward diffusion) appears to be the more plausible one. However, because assessments of this sort, in the lack of direct proof, are to such a great extent influenced by theoretical assumptions, I will premise some considerations of a general methodological nature.

Methodological Considerations

A major stumbling block for any comparative enterprise investigating possible influences between ancient Greek and Indian thought has always been the historian’s (p.151) unwillingness to concede anything in the absence of sound historical documentation. To me this is just a particular instance of the scholars’ partiality for their own favourite method of going about a subject of research – a partiality which is understandable, but nevertheless a bane to the progress of science. If historical documentation were always indispensable, whole disciplines like folklore would ipso facto simply dissolve into thin air. Folklorists are hardly ever able to document the historical vicissitudes of transmission of folk tales, customs, rituals, etc., which has not prevented them from developing significant and fruitful discourse about their subject matter.7

What is needed to satisfy the historian, in my view, is merely to establish that the historical conditions were actually there which could have made cross-cultural contacts possible. Being myself no fully-fledged historian, I will just refer to the opinion of one such, who, writing a booklet with the explicit intent of debunking what he regarded as (in his own words) the ‘inflated, doubtful, simplistic and misleading’ claims of the proponents of the ‘Indian hypothesis’, had nevertheless to concede:

Historical circumstances appear to favor lasting exchanges between India and the East in the ancient world. Phoenician traders were carriers of Indian commodities for centuries. Persia imposed a long rule on north-western India, bridging east and west [emphasis added]. Alexander the Great’s hellenising crusade … brought in his wake artists, philosophers, historians and naturalists, thus preparing ground for the remarkable Greco-Bactrian kingdoms of the following century. The powerful empire of the Kushans was in touch with the masters of Rome, and the Roman eagle adorned trading posts on the coast of Coromandel … Persian, Greek and Roman officials, civilians and military, had direct experience with India for about seven hundred years.8

So much for the viability of the hypothesis of influences between early Greek and Indian thought.9 Since, however, we cannot progress further on this ground (p.152) by way of direct historical verification of actual contacts and transmissions, we are left with the only option of arguing from similarities that could bespeak a common origin or a filiation. I submit that there is nothing inherently unsound in this endeavour. On the contrary, it is part and parcel of the method of other disciplines, such as, again, folklore, or, to add a fresh example, philology. Philologists are usually unable to reconstruct the historical vicissitudes of the copying of manuscripts, but are guided in postulating stemmatic relationships by coincidences of details (errores coniunctivi). Of course, care must be exercised: all coincidences of scribal errors are not suitable to postulate dependencies, for trivial mistakes related to common scribal propensities could have arisen independently.

In much the same way, all similarities between cultural constructs are not apt to establish connections. Stunkel summarises the argument from similarities as ‘the postulation of a mechanism of intellectual and spiritual transmission beyond historical verification, but which one must adopt and respect if parallels [allegedly] “not easy to ascribe to chance” are to be explained’, lamenting that ‘the diffusion of religious attitudes and ideas is construed as tenuous, piecemeal, unconscious, fragmented in time and unspectacular in the act of transmission and reception’.10 This is not the place to examine the import of each of these charges in detail: suffice it to say that in my view, some are correct, some are not, indeed, they find fault with what I actually regard as conditions of effectual comparisons.

As to the charge of tenuity, I agree that parallels of a very general nature, not substantiated by similarities of details, are inconclusive, for similar general ideas could very well have cropped up independently from shared psychological and cultural grounds. This is akin to the aforesaid scribal proneness to mistakes due to homoioteleuton, haplography, dittography and the like. On the other hand, as to the charge of fragmentariness, I actually regard it rather as a requisite of fruitful comparison, in the same way as the more haphazard scribal errors are, the more telling they are and the more apt to function as errores coniunctivi to establish stemmatic dependencies.

Let me clarify with a well-known example taken from the more congenial field of comparative mythology and folklore. As is well known, the deluge myth has enjoyed a wide diffusion all over the Eurasian continent and beyond,11 prompting the inevitable question of polygenesis versus monogenesis and diffusion. All versions agree on but few fundamental points: mankind is swept away by a deluge, except for one or more people surviving in a vessel, who are entrusted with the task of the renewal. However, we should hardly be justified (p.153) in grounding any presumption of a common origin on the strength of such similarities: after all, flood is one of the few fitting ways to end the world which experience suggests to primitive folks; to survive a flood you need a vessel of sorts; and in the aftermath survivors must find a way to repeople the empty world. On the other hand, more particular coincidences of casual details not integral to the overall structure of the myths make a strong case for diffusion. For instance, within the fold of the biblical and near-eastern versions of the deluge myth, the correspondences in the episode of the birds sent out as scouts are too precise to admit of an independent origin. When the existence of a Chaldean account of the deluge myth was first brought to the attention of the world by George Smith’s famous lecture in 1872, the sheer force of such coincidences was enough to unsettle the cherished certitudes of confessional biblists about the absolute originality of the Bible, whose dependence on Sumero-Accadian archetypes was established even in the utter lack of historical documentation and has never been called to question ever since.12

In much the same way, I contend that there may not be much to be gained for the comparatist interested in investigating possible cognations by taking stock of wide-ranging similarities between ancient Greek and Indian doctrines, like, say, the Parmenidean, or rather Melissian, principle: οὐδὲν ἐκ μηδενός‎ (Melissus, fr. 1), and its intriguing counterpart (although lacking all the speculative elaboration) in the Upaniṣadic master Uddālaka’s rhetorical question: katham asataḥ saj jāyeta, ‘how could being be born from non-being?’ (BU 6.1.2). After all, it is only reasonable that overarching conceptions like the metaphysics of Being may have developed independently in different cultures, possibly (in this case) against the common Indo-European backdrop of shared linguistic categories. On the other hand, the concurrence in details of lesser import, such as are not required either by logical necessity or psychological inclination does require an explanation, to the extent that the concurrence is too specific to be ascribed to mere chance.

One such instance is, in my opinion, the parallel occurrence of the allegory of the soul chariot in Plato’s Phaedrus and in the Kaṭha Upaniṣad. Let us briefly review the passages in question.

The chariot allegory in Plato’s Phaedrus

The Phaedrus is a very complex dialogue, whose chief aim is to establish the superiority of philosophical discourse, caring for truth, over rhetorico-sophistic (p.154) discourse, geared only to success. Socrates runs into Phaidros on his way back from listening to a lecture by Lysias, in which the celebrated rhetor claimed to demonstrate the extravagant assumption that it be better for the beloved to please the unloving rather than the loving one. In order to defuse the young man’s enthusiasm and win his admiration, Socrates undertakes to improvise a lecture in the same strain, but soon professes to repent the ‘sacrilege’ against the god Eros, and feels obliged to pronounce in atonement a second discourse of retraction, after the manner of Stesichoros’ famous Παλινῳδία‎. In this second discourse Socrates impugns Lysias’ thesis according to which the unloving one should be preferred on account of being more reasonable: the lover’s presumed folly is no human foolishness, but divine μανία‎: a superhuman possession aroused when the sight of human beauty awakens the dim memory of divine beauty, which the immortal soul once beheld in the region ‘above the heaven’ before incarnating in a mortal body.

Against this backdrop, the chariot allegory is immediately preceded by a proof of the immortality of the soul (Phdr. 246ff.), after which Socrates introduces the theme of its ‘aspect’13 (ἰδέα‎). To define its essence, namely what the soul really is (οἷον μὲν ἐστι‎), is beyond the reach of human capabilities, says Socrates, but it is indeed possible to say what it is like (ἔοικε‎). Let us read Plato’s words:

Let [the soul] be likened to the composite inborn power of a pair (ζεῦγος‎) of winged horses and of a charioteer. However, both the horses and the charioteers of the gods are all good, and of good descent; but as for those of the others, it is a mixed affair; and first of all our driver leads an ill-assorted pair (συνωρίς‎), and secondly one of the horses is himself noble and of like descent, but the other is quite the opposite, and of opposite descent: so that difficult indeed and troublesome is of necessity the driving for us [mortals].

(Phdr. 246a–b)14

It is worth noticing, with Robin,15 that although ζεῦγος‎ is the word usually employed for a pair of horses, in applying the metaphor to the human soul Plato makes use of the word συνωρίς‎ instead (which I have accordingly translated as ‘ill-assorted pair’) to signify that the human horses are not really paired, or ‘on the same par’, so to speak, but they are extrinsically conjoined (συν-ωρίζω‎) in spite of their different natures.

(p.155) Plato does not specify the other constituents of the chariot; since, however, he says further on, while distinguishing between the mortal and immortal living beings, that the whole soul is the one who ‘takes care’ (ἐπιμελεῖται‎) of the soulless, viz. the body, the supposition lies near at hand that the chariot represents the latter. Although the tenor of the Platonic text is not quite clear at places, it seems that both the souls of the gods and those of men are joined to motionless bodies which acquire the appearance of self-mobility in the union, thanks to the inherent moving power of the soul. Nevertheless, there are differences in either case, namely: in the first place, the souls of the gods never lose their wings; and secondly, they are forever joined to their respective bodies. So much for the souls of the immortals, as Plato himself does not wish to belabour the matter, relinquishing it to godly placet.

As for ourselves, beings called ‘mortals’ were originated in consequence of the loss of the wings and the ensuing fall of the soul from heaven to earth, where it clung to a solid body in which it established its dwelling. But how did the loss of the wings come about? It pertains to the nature of the wings to lift what is bound downwards, and for that reason they are the most godlike among corporeal things. Therefore, as the divine is beautiful, wise and good, the wings are nourished and grown by similar things, but they are consumed and destroyed by the opposite. Now the gods led by Zeus revolve in the heaven in their winged chariots governing the universe in good harmony, followed by whoever so wishes and is able, for gods know no jealousy (φθόνος‎). But when they convene to a banquet, they proceed towards the culmination of the interior cusp of heaven (ὑπουράνιος ἁψίς‎) and they pass to the other side, where the plain of truth (ἀληθείας πεδίον‎) stretches out in the region above the heaven (ὑπερουράνιον τόπον‎). There the charioteers sate themselves on the sight of the substances that really are and feed on the pure science which is not affected by becoming and variability, but only pertains to true being. And once they are sated, they revert to their seats under the vault of heaven, unharness their horses and feed them on nectar and ambrosia. But mortals have a hard time trying to follow the gods, because the horses are difficult to control, so that the charioteers can only take a glimpse of some of the true substances, or of none at all; and in the ensuing flurry and in the collision of chariots vying for the precedence some horses are maimed and lose their wings. And the charioteers who do not get a view of the pure science of reality have to feed on opinion, which in turn entails the loss of the wings.

The chariot allegory in the Kaṭha Upaniṣad

The Kaṭha Upaniṣad in its present form consists of six chapters termed vallīs (‘creepers’) in two sections of three chapters each, but the original core may (p.156) have comprised only the first two (or three) vallīs,16 dealing with the teachings imparted by the death god Yama to the boy Naciketas, who had reached while still living the abode of the dead through his father’s curse. To acquit himself of faulty hospitality Yama grants the boy three wishes, and Naciketas chooses as third the solution of his doubts about the continuation of existence beyond death. After a few vain attempts to elude the question by offering various boons in exchange, finally Yama praises the boy’s steadfastness and reveals to him the doctrine of the knowledge of the Self (ātman) which rescues the wise from the wheel of rebirths and of reiterated death.17 The high-flown celebration of ātman which concludes the second vallī links up with the third, which purposes to show the path to reach the supreme abode of the ātman through the practice of yoga, introduced precisely by the chariot allegory:

Knowers of brahman … call Shadow and Light the two, drinkers of Truth (ṛta) in the world of their deeds,18 [the one] installed in the cave of the heart, [the other] in the supreme region beyond. … Know that the Self (ātman) is the rider in a chariot, and the body is the chariot (śarīra); and know that the intelligence (buddhi) is the charioteer, and the mind (manas) is the bridle. They say that the senses (indriya) are the horses, and the sense objects (viṣaya) are their lanes (gocara). The wise call Enjoyer (bhoktṛ) the yoke (yukta) of Self, senses and objects. But as for him who has no understanding (vijñāna) and whose mind is ever unyoked (ayukta), his senses are unrestrained, like bad horses for a charioteer; on the other hand, for him who has understanding and whose mind is ever yoked (yukta), his senses are (p.157) indeed restrained, like good horses for a charioteer. He who has no understanding, is mindless and ever impure, does not reach that region, and incurs rebirth in the flux of existence (saṁsāra); but as for him who has understanding, is mindful and ever pure, he does indeed reach that region whence he is not born again. A man whose understanding is his charioteer and whose mind his bridle reaches the end of the road, that supreme region (parama pada) of Viṣṇu.

(KaU 1.3.1; 1.3.3–6)19

In this short passage there occurs three times the word yukta, from root yuj, ‘yoke, join’, which, beside its literal meaning referring to the action of yoking or harnessing draught animals, is also employed in the figurative meaning of ‘subjugating’ passions and the like. The latter meaning lies at the foundation of the name of Yoga, one of the six classical schools of Indian philosophy, in so far as the proper object of Yoga is precisely the subjugation of psychical functions.20 As a matter of fact, the passage quoted above is immediately followed by a rudimentary sketch of the way of introversion advocated by Yoga, which (sketch) is considered one of the most ancient documents of the proto-history of that school of thought.21 According to the retroversion procedure elsewhere imaginatively styled as pratiloma22 (i.e. literally ‘against the hair’), which lies at the core of Yoga, each of the psychical faculties, which normally act outwardly, must be made to flow back inwardly to its source, or, as our Upaniṣad says, it must be ‘curbed’ by the faculty which stands higher in a hierarchy comprising, from low to high, senses and their objects, mind, intellect, the ‘great Self’ (ātmā mahān) and the unmanifest (p.158) (avyakta), to finally end up in the awareness principle (puruṣa) which is the final goal (parā gatiḥ) of the process. It is worth observing that the term that I have translated as ‘curb’ is a form of root YAM, ‘restrain’, which once again is frequently employed in connection with riding and draught animals.

These details of lexical usage manifest the close relationship between the chariot imagery and the beginnings of Yoga in the Kaṭha Upaniṣad. We must not overlook that what we have here is something more than just an allegory of the soul and the physio-psychic complex. The reining of the horses by the charioteer through the reverberations of the lexical choices distinctively hints at the process of Yoga, and the final goal of the journey, the supreme region of Viṣṇu, represents the final resolution in the awareness principle termed puruṣa which is the goal of the yogic process.

Actually, the technical term puruṣa (literally ‘man’ in the sense of ἀνήρ‎), which designates the awareness principle in the Sāṃkhya-Yoga philosophy, has evolved from very ancient roots in the cosmogonic hymn Ṛgveda 10.90 (end of II millennium BCE?), and on the religious side has acquired strong connections within the fold of Hinduism with the aspect of supreme Godhead named Viṣṇu. This explains why the resolution into the puruṣa is intimated within the allegory by the attainment of the supreme region (parama pada) of Viṣṇu. The parama pada is likewise well-known from the most ancient Vedic hymns in connection with Viṣṇu in his capacity as Trivikrama, i.e. the ‘Thrice-Strider’, his three strides supposedly representing the three stations of the sun in its diurnal course. According to the lexical values of the word, which include ‘step, footstep and place’, the parama pada connotes at one and the same time the ‘supreme place’ carrying the footprint of the third step of Viṣṇu as the sun traversing the horizon.

We shall presently turn to the convergences and discrepancies evidenced in the allegories as they appear in each text, in order to evaluate whether the similarities occur in traits that are arbitrary enough to enhance the probability of borrowing, according to the criterion laid down previously.

In case borrowing should be deemed likely, its direction would still remain to be determined. Should Plato have borrowed the allegory from the Kaṭha or the other way round? As always, chronology does not help in an Indian setting, since the dating of Indian texts is, at best, aleatory and highly speculative.23 The utmost that can be said is that the Kaṭha may very well be anterior to Plato, or roughly contemporary, without being able to absolutely rule out the possibility of the contrary.

(p.159) But one additional criterion may be resorted to, which finds ready application in the present circumstance, namely, the criterion of integrality, that might be conveniently expressed thus: the more an element appears to be organically integrated in its surrounding structure, the lesser the probability of its having been borrowed from somewhere else.

Now the allegory of the soul chariot is not particularly called for in the context of the Phaedrus: it works, but as far as I can see there is no compelling reason that makes it specially suitable to express the ruling function of the intellect over the other psychic faculties in the way Plato understands it, so that it is not unthinkable that he could have borrowed it from an external source. On the other hand, as we have seen, the allegory of the soul chariot in the Kaṭha Upaniṣad, with its attendant imagery of yoking and bridling, is conjured up by the very name of the Yoga doctrine it is meant to illustrate, so much so that it would be utterly unreasonable even to suppose that it might be of alien origin.24

The persuasiveness of this conclusion is reinforced if we extend the application of the criterion of integrality to the whole of the Greek and Indian scenarios. I have reviewed occurrences of the soul chariot allegory in both literatures elsewhere,25 and space constraints do not permit to go over it now. I must be content with stating the conclusions: leaving hints of a questionable nature aside, a fully-fledged soul chariot allegory is unknown in Greece before Plato (with the possible exception of the locus Parmenideus, the allegorical nature of which, however, is moot).26 On the other side, soul chariot allegories occur in India in at least three works belonging to the Upaniṣad genre – i.e. the Mahaitareya, the Chāgaleya and the Kauṣitaki – all of them probably antedating the Kaṭha; and two more – the Śvetāśvatara and the Maitrāyaṇīya – almost certainly more recent, but interesting nevertheless, in that both uphold Yoga as a means of deliverance and so corroborate the structural link obtaining between Yoga and the chariot allegory.

(p.160) Comparative assessment

The foregoing scrutiny has evidenced the following areas as most amenable to a comparison: in the first place, the correspondence of the distinct components of the chariot with certain psychic faculties; secondly, the discipline of the driving; and thirdly, the use of the chariot for the journey to the world beyond. Let us begin with the last one.

The chariot as a vehicle for the world beyond

Both in the Homeric epics and in the Vedic hymns the chariot is sometimes envisaged as a means of communication between the world of men and the world of gods;27 likewise, in both literary traditions chariot imagery is employed as a metaphor of the poetic word granting the poet access to other planes of consciousness.28 Because it is so, it is unnecessary to suppose external influences in this connection, as the Phaedrus and the Kaṭha allegories may very well have drawn on elements belonging to their respective traditions. Nevertheless, it is worthwhile pointing out some striking parallels in the two conceptions of the final goal of the journey:

  1. a. The ὑπερουράνιος τόπος‎, the region above the heaven, is conceptually identical to the parama pada, the supreme region of Viṣṇu, for we have seen that the latter connotes at once the footprint (pada) of the third step (pada) of Viṣṇu ‘whose eye is the sun’; at the same time, as representing the highest step, the parama pada corresponds to the zenith, or the ὑπουράνιος ἁψίς‎ which is the threshold to the world beyond.

  2. b. Besides, it may be no more than a baffling coincidence that the ‘region above the heaven’ is also qualified as the ‘plain of truth’, i.e. (ἀληθείας‎) πέδιον‎, where the word πέδιον‎ is linguistically cognate to pada in parama pada.

  3. c. According to the Phaedrus, in the ὑπερουράνιος τόπος‎ the gods and their followers apply themselves to the contemplation of the true essences. In like manner in Ṛgveda 1.22.20 the parama pada is said to be perpetually contemplated by gods (or sages).29

  4. (p.161) d. In the ὑπερουράνιος τόπος‎ the gods and their followers enjoy the pleasures of the banquet; likewise in Ṛgveda 8.29.7 the gods and elsewhere (1.154.5) their worshippers rejoice in the parama pada, where flows the fountain of honey: ‘O that I could reach that dear place where men sacred to the gods rejoice; for there, in Viṣṇu’s supreme region, is the fountain of honey!’30

  5. e. After reaching the ὑπερουράνιος τόπος‎ the chariots come to a stand and are carried round by the revolution of the heavenly sphere; this reminds one of an (admittedly rather obscure) passage where the lord of the parama pada is said to have ‘set his steeds in swift motion … like a turning wheel’ (RV 1.155.6).

On the other hand, in the Phaedrus the ὑπερουράνιος τόπος‎ can only be attained thanks to the horses’ being provided with wings, and this undoubtedly constitutes the major discrepancy with respect to the image of the Kaṭha. Of course, winged hybrids are well-known to Greek mythology, so that we must look no further for an explanation.31

The steering of the chariot

Both the Phaedrus and the Kaṭha agree in stressing the need for disciplined steering of the chariot in order to reach the journey’s destination. The notion of discipline is conveyed in the Kaṭha through the metaphor of the ‘subjugated’ (yukta) horses, and we have already drawn attention to the close lexical, semantic and conceptual relationship obtaining between the terms employed for subjugating and restraining the horses on the one hand and some key concepts of the burgeoning school of Yoga as a method for subjugating and restraining psychic faculties on the other. The seamless integration of the chariot imagery in the conceptual array of proto-Yoga, which is unparalleled in the Phaedrus, speaks for the native status of the metaphor in the Kaṭha, as we have already remarked.

However, we may perhaps discern some faint echo of it in the lexical usage of the Platonic dialogue. The Sanskrit term yoga, literally meaning a ‘yoke’, is linguistically cognate to the Greek ζεῦγος‎ which designates the pair of divine horses harnessed to the chariot. Although the figurative meaning of ‘subjugation’ (p.162) is ostensibly absent in the Platonic passage, it may not be devoid of significance that, as we pointed out, Plato employs a different word (i.e. συνωρίς‎) to designate the unruly pair of human horses of opposite temperaments; so that by implication the word ζεῦγος‎ seems to acquire the additional value of connoting the divine horses as unanimous and obedient to the charioteer: that is to say, ‘subjugated’ in the same sense as yukta.

But the most notable point of similarity with respect to the steering of the chariot is without doubt the one concerning the difficulty caused by the opposition between good and bad horses, although such opposition wears quite different aspects in either case, for in the Phaedrus one horse is congenitally good and the other the reverse, whereas the horses of the Kaṭha do not admit of an internal disparity, but they are only susceptible of being, all of them, well-behaved, or else ill-behaved. This divergence stems from the different symbolic function of the horses, and above all from the paramount difference in the underlying ontology, as we shall presently see; nevertheless, even the coincidence of the mere idea of the antithesis is worthy of note.

The chariot as an allegory of psychic functions

We finally come to the most important congruence, i.e. the application of the chariot imagery as a sustained allegory for the psychic functions. We have already observed that, although the motif of the chariot as a vehicle for a journey to the other world had been common to both literary traditions since hoary antiquity, on the contrary the motif of the ‘soul chariot’ appears to be more or less specific to the Indian literary tradition, at least in the earliest period, being attested in several Upaniṣads, whereas it is virtually unknown to the Greek literature before Plato (with the possible if controversial exception of Parmenides).

The table given in Appendix I outlines a synopsis of the correspondences between the single parts of the chariot and the constituents of the physio-psychic complex in some of the most meaningful texts, while the bottom line highlights the several goals of the journey in the same texts. Items appearing in all texts, albeit possibly in different connections, are marked in ALL CAPITALS; whereas correspondences of a more opinable nature by reason of not being expressly declared in the text, but only inferred from the context (or, in the case of Parmenides, from Sextus’ interpretation) have been enclosed in square brackets.

What impresses one on first perusing the table is that, to start with, the correspondences are much more articulate in the Indian texts than in the Phaedrus, irrespective of their variability in the detail. The chariot is everywhere found to represent the body (albeit not explicitly in the Greek texts). The charioteer, on the other hand, severally corresponds to the mind (manas), the intellect (buddhi) or (p.163) the soul in different texts both Indian and Greek. The alternation between manas and buddhi may be ascribed to a terminological indetermination characteristic of the older Upaniṣads. As for the equation of the charioteer with the soul, it occurs on the Indian side only in the Chāgaleya, which is peculiar in many respects, and seemingly cut off from the line of development connecting the Mahaitareya to the Maitrāyaṇīya through the Kaṭha.

What interests us more is that both in the Kaṭha and in the Phaedrus the charioteer represents the rational faculty: buddhi/vijñāna in the Indian text and νοῦς‎/διάνοια‎ in the Greek one, viz. the intelligent (or ‘intelligible’, in Scholastic parlance) aspect of the soul (τὸ λογιστικόν‎, according to the psychology of the Republic).

The bridle, which represents the mind in the Kaṭha, is not expressly mentioned in the Phaedrus, but is implied in the Greek word for ‘charioteer’, which is ἡνίοχος‎,32 i.e. ‘he who holds the reins’ (ἡνία‎).

As for the horses, their correlates are totally different in India and Greece, for in the Indian tradition they stand for the indriyas, i.e. the ‘faculties’ without distinction (in the earliest period), and later, at the time of the Maitrāyaṇīya, when the rising Sāṃkhya cosmo-psychology had started distinguishing between sense organs and action organs,33 the latter ones. On the Greek side, in the Phaedrus the two horses represent the irrational aspects of the soul, which would later be called in scholastic parlance the irascible and the concupiscible (τὸ θυμοειδές‎ and τὸ ἐπιθυμητικόν‎ according to the psychology of the Republic). In the proem of Parmenides according to Sextus’ interpretation, the mares likewise stand for desires and irrational impulses of the soul.34

The distinction of three aspects of the soul – τὸ λογιστικόν, τὸ θυμοειδές‎ and τὸ ἐπιθυμητικόν‎ – finds no equivalent in the ancient Upaniṣadic psychology. However, the rudiment of an analogous conception may perhaps be seen in the stereotyped pair kāma and krodha (‘desire’ and ‘anger’) which occurs several times in the Bhagavad Gītā35 in the capacity of arch-enemies of jñāna (‘knowledge’). For example, in a passage which calls to mind the chariot allegory through the use of certain words and images, Kṛṣṇa admonishes Arjuna as follows:

(p.164) This is desire (kāma), this is anger (krodha), fostered by the element of ardour (rajoguṇa) … as fire is enveloped by smoke and a mirror is clouded by dust … in the same way is knowledge obfuscated by this relentless opponent of the knower … the senses, the mind and the intellect are its abode, as they say, and through them it deludes the embodied [soul] by obfuscating knowledge. Therefore, restrain (niYAM) the senses in the first place, and then kill that iniquitous destroyer of knowledge and science.

(Bhagavad Gītā 3.37–41)36

Admittedly, the Bhagavad Gītā properly speaking is no Upaniṣad, being embedded in the Mahābhārata epos, datable after the close of the Vedic period, and for that reason must on all likelihood be ascribed to a later time than the Phaedrus. Nevertheless, the couple of kāma and krodha already occurs in one of the most ancient Upaniṣads, in a passage enumerating as components of the (world-immanent) universal Self, in addition to the faculties and the elements, also kāma-krodha and their opposites (BU 4.4.5).

In general terms, it may be observed that the Indian tradition is more interested in articulating the physio-psychic complex in its entirety, in order to account for the ordinary, world-affirming sensory experience as well as for its opposite, the extraordinary, world-negating practice of sensory restraint (yoga) leading to the suprasensory. For its part, the allegory of the Phaedrus only contemplates the nature of the soul with its essential components, the intelligible, the irascible and the concupiscible, represented by the joint agency of the charioteer and the pair of horses.

Against the backdrop of all the varying degrees of similarity between the allegorical correlates examined above, one item of the allegory has been left unreviewed thus far, which appears in one way or another in all Indian texts, but is conspicuously absent in the Greek ones:37 namely, the idle passenger on the chariot. In all of them (except for the odd Chāgaleya) its regular correlate is the soul: for the soul, according to the standard Indian view, coincides neither with any of the several psychic functions signified by the different parts of the chariots, nor with their joint agency (as is the case with the Phaedrus).

Indeed, here lies the paramount disparity between the Greek and Indian versions of the chariot allegory, which is rooted in the widely differing ontologies of Plato and of the school of Sāṃkhya-Yoga at its dawn in the Kaṭha. Those ontologies diverge essentially with respect to where they set the boundary line between the respective pertinences of body and soul. According to Plato, the soul is tripartite (p.165) in its functions, this tripartition being reflected in the image of the charioteer and the pair of horses; but, according to the same image, it is up to the rational faculty to oversee the other two. On the other hand, in the Indian texts one meets the distinct figure of the rathin, that is to say, literally, the ‘owner of the chariot’, or he who makes use of the chariot as an instrument, while remaining distinct and detached with respect to it. The reason for this is that, according to the dualistic psychology of Sāṃkhya-Yoga, there exists a radical opposition between the soul (puruṣa), which is the pure luminosity of awareness as the horizon of the appearance of objects, and nature (prakṛti), which is the physical substrate of the outer world as well as of the inner physio-psychic complex, inclusive of the rational, volitional and desiderative faculties. To put it succinctly, the intellect is part (indeed, the best part) of the soul, according to Plato, whereas it is non-soul, but merely a part of the body, according to Sāṃkhya-Yoga.


The above tentative analysis, albeit needing further refinement, has brought to light both similarities and differences between the chariot allegories in the Kaṭha Upaniṣad and in the Phaedrus, which can be summed up as follows. We may count as points of similarity:

  1. a. the overall idea of the chariot as an allegory of the psychic functions (for which there are more and still earlier instances on the Indian, but not on the Greek side);

  2. b. the identification of the charioteer with the rational faculty;

  3. c. the analogous characterisation of the parama pada and, respectively, the ὑπερουράνιος τόπος‎/ἀληθείας πέδιον‎ as the final goal of the journey;

  4. d. the general notion of the necessity of a ‘subjugation’ of the psychic faculties in order to reach the journey’s end;

  5. e. the broad concept of an opposition between good and bad horses (however differently declined in either case).

On the other hand, we can count as major discrepancies:

  1. a. the different symbolic significance of the horses; and, above all,

  2. b. the absence of the separate figure of the idle traveller in the Phaedrus;

both of them grounded in the crucial diversity of the underlying ontology of the two works.

It lies with the reader to judge for himself of the plausibility or otherwise of the considerations propounded; as for my own assessment, I am of the opinion that the similarities evinced may be momentous enough to justify the supposition, (p.166) pending further research, that Plato might have been acquainted (either directly of indirectly) with the chariot allegory and the attendant doctrines of the Kaṭha Upaniṣad, and might have drawn inspiration from them, while contextually adapting them to the theoretical frame of his own ontological thought.

Aitareya Āraṇyaka(2,3,8)

Chāgaleya Upaniṣad

Kaṭha Upaniṣad(1,3)

Maitrāyaṇīya Upaniṣad(2,3.6; 6,28)








[the ‘mortal who knows’]








(body parts: tendons, bones, blood, skin)

(belts, supports, oil, hood)



– [the maiden ‘immortal CHARIOTEERs’ ἀθάνατοι ἡνίοχοι‎] – [Dike]

CHARIOTEER (ἡνίοχος‎)





sense organs eyes

pair of HORSES




[sun daughters]

ears sides (wings)


action organs





natural disposition

whip (karma)


sensible experiences


irrational impulses

[MARES] pair (ζεῦγος/συνωρίς‎) of winged HORSES


destruction of

worldly desires


Viṣṇu’s ‘highest footstep/abode’ (parama pada) to revolve (like a potter’s wheel)

the Goddess’s abode

– the ‘region above the heaven’ (ὑπερουράνιον τόπον‎) – the ‘plain of truth’ (ἀληθείας πεδίον‎)


(5) McEvilley 2002: 185. He goes so far as to suggest that ‘the allegory of the Self [may be] a development of the Homeric chariot hero on the one hand and of the hero of the Bhārata war on the other’, which is scarcely tenable, as the Mahābhārata epic is generally believed to be later than the Upaniṣads.

(6) Both Schlieter and Forte-Smith, in their contributions on the same subject to the present volume, subscribe to the hypothesis of independent origination, albeit on different grounds. While our respective appraisal of the many aspects involved may be different, I appreciate Schlieter’s perspective – stressing the primacy of cognitive metaphors in shaping conceptual thought and positing a shared metaphorical ground as a precondition for the independent (in his view) origination of the philosophical chariot metaphors (or rather allegories) in the Phaedrus and the Kaṭha Upaniṣad – as capable of throwing complementary light on my own. On the other hand, while I also appreciate Forte and Smith’s stress on intertextuality as a useful, nay, indispensable key to a proper understanding of texts – of which they offer many apt examples (if at places a bit far-fetched, in my opinion) – I think they rely too one-sidedly on it to make sense of any given text. I believe there is more to Parmenides’ proem or the Kaṭha Upaniṣad than can be made out by falling back on their Homeric and Vedic intertexts – but of course here we border on the domain of fundamental philosophical options: I do not defer to Derrida’s différance, but rather side with Vedantic (and Parmenidean) astikatva.

(7) For a detailed examination of the various flavours of the comparative method as employed in anthropology and folklore see, e.g., Dundes 1986.

(9) Forte and Smith object that cross-cultural contacts concerning the secret lore of the Upaniṣads would be virtually impossible, because the persons involved ‘would have had to be Vedic initiates and committed to memory the corpus of Kaṭha texts before instruction in an Upaniṣad would be permitted’. Of course, this is true, in principle. But the very fact that the Upaniṣads repeatedly inculcate the prohibition to divulge certain doctrines to unfit persons (examples quoted in Deussen 1906: 11) speaks for the actual, if sporadic, occurrence of such divulgations, which could involve just those particular rahasyas, detached from the embedding corpus. After all, the same strict requirements for Vedic transmission had been in force at least up to the inception of Western Indology, which did not prevent Anquetil Duperron from coming by the mysterium tegendum of the Oupnek’hat (on its own account already the fruit of a disclosure to a Muslim prince – hardly a worthy recipient of the sacred Vedic lore!).

(11) For an ample survey see Dundes (ed.) 1988.

(12) On the other hand, the utter lack of any such correspondences, together with other considerations, in my view provides a good reason for establishing the originality of the Indian flood myth vs. the Classical and Near-Eastern versions. See Magnone 2000, 2004.

(13) Thus translates Velardi (2002: 185), on account of the fact that here, as Plato himself says, we are dealing with a metaphoric image of the soul.

(14) ἐοικέτω δὴ συμφύτῳ δυνάμει ὑποπτέρου ζεύγους τε καὶ ἡνιόχου. θεῶν μὲν οὖν ἵπποι τε καὶ ἡνίοχοι πάντες αὐτοί τε ἀγαθοὶ καὶ ἐξ ἀγαθῶν, τὸ δὲ τῶν ἄλλων μέμεικται. καὶ πρῶτον μὲν ἡμῶν ὁ ἄρχων συνωρίδος ἡνιοχεῖ, εἶτα τῶν ἵππων ὁ μὲν αὐτῷ καλός τε καὶ ἀγαθὸς καὶ ἐκ τοιούτων, ὁ δ᾽ ἐξ ἐναντίων τε καὶ ἐναντίος: χαλεπὴ δὴ καὶ δύσκολος ἐξ ἀνάγκης ἡ περὶ ἡμᾶς ἡνιόχησις‎. (All translations are mine).

(16) The first three, according to Deussen (1897: 264); or just the first two, according to Olivelle (1998: 372). However, as Deussen (1897: 278) remarks, the last two strophes of the third vallī look like the epilogue of the original Upaniṣad.

(17) According to Forte and Smith there is ‘no self-realization or arcane knowledge about brahman’ taught in this Upaniṣad, nor any philosophical doctrine of monism or of whatever other description; indeed, their paper purports to be ‘an attempt to demonstrate that the chariot in the Kaṭha Upaniṣad still refers to the sacrifice and constitutes a metaphysical theorisation of the operations of the fire altar’. This is indeed an astounding proposition, which runs counter to all the commonly held lines of interpretation of the Upaniṣad both Western and Indian. Of course, it is impossible to go into its merits in the space of a footnote; I shall be content with pointing out what I regard as the fundamental flaw in Forte and Smith’s analysis, namely that it begs the question, in that it only deals with such passages as are amenable to be read in connection with the ritualistic brahmanical background (the first vallī up to the grant of the second boon; and the beginning of the third vallī, up to [and obviously including] the chariot metaphor), whereas it omits to take stock of those very passages (the concluding part of first vallī with the request of the third boon, of paramount importance; and the whole of the second vallī, devoted to the praise of the path of knowledge, culminating in the realisation of the ātman, as Yama’s fulfilment of the request) where the Upaniṣad really comes into its own, pouring, as it were, the new wine of the metaphysics of brahman-ātman into the old wineskins of ritualistic thought.

(18) Reading with Śaṅkara svakṛtasya; otherwise sukṛtasya ‘of good deeds’.

(19) ṛtaṃ pibantau svakṛtasya loke guhāṃ praviṣṭau parame parārdhe | chāyātapau brahmavido vadanti … || … || ātmānaṃ rathinaṃ viddhi śarīraṃ ratham eva tu | buddhiṃ tu sārathiṃ viddhi manaḥ pragraham eva ca || indriyāṇi hayān āhur viṣayāṁs teṣu gocarān | ātmendriyamanoyuktaṃ bhoktety āhur manīṣiṇaḥ || yas tv avijñānavān bhavaty ayuktena manasā sadā | tasyendriyāṇy avaśyāni duṣṭāśvā iva sāratheḥ || yas tu vijñānavān bhavati yuktena manasā sadā | tasyendriyāṇi vaśyāni sadaśvā iva sāratheḥ || yas tv avijñānavān bhavaty amanaskaḥ sadāśuciḥ | na sa tat padam āpnoti saṁsāraṃ cādhigacchati || yas tu vijñānavān bhavati samanaskaḥ sadā śuciḥ | sa tu tat padam āpnoti yasmād bhūyo na jāyate || vijñānasārathir yas tu manaḥpragrahavān naraḥ | so ‘dhvānaḥ pāram āpnoti tad viṣṇoḥ paramaṃ padam.

(20) Cf. Yoga Sūtra 1.2: yogaś cittavṛttinirodhaḥ.

(21) Under their presumption that the Kaṭha Upaniṣad be nothing more than a ‘poetic recreation’ of its ‘source [Brāhmaṇa] text’, Forte and Smith quickly dispose of this passage in a couple of sentences as ‘definitions of ātman and puruṣa contoured by Vedic ritual reveal[ing] that the text describes the [ritual] transmission of the terrestrial self to the heavenly self’. It might be so – but here again, one would expect a bit more of careful reasoning when calling into question the unanimous interpretative tradition both Western and Indian, which envisages this very text as the foundation stone of the classical Yoga darśana.

(22) As explained in the Rājamārtaṇḍa commentary at Yoga Sūtra 1.2: ‘The mental organ is an evolute of pure sattva and the psychic functions are its secondary modes. Yoga consists in their inhibition, namely in their dissolution in their source by stopping their extroversion by means of a process “against the hair” of introversion’ (cf. Magnone 1999: 25).

(23) As one recent Upaniṣadic scholar puts it rather bluntly: ‘any dating of these texts that attempts a precision closer than a few centuries is as stable as a house of cards’ (Olivelle 1998: 12).

(24) Richard Seaford alerts me to the concurrent option that it might be precisely the (preexisting) terminology of the Yoga doctrine that made the expression of the doctrine receptive to a possible influence of a chariot metaphor borrowed from elsewhere (e.g. Greece). Irrespective of the problems posed by chronology, here I subscribe to the tenet of the conceptual metaphor theory as advocated by Schlieter, according to which abstract thought is inherently metaphorical, and cognitive metaphors precede and structure (or are at least conterminous with) abstract thought. In other words, the verbal notion of yoga, having a literal meaning in the domain of activities concerning draught animals, could not have acquired its figurative meaning in the domain of activities concerning the human psyche, before the metaphor connecting both domains as source and target was in place.

(26) On the subject see Latona 2008.

(27) E. g. Iliad V, 355ff. (Ares and Aphodite); V, 720 ss. (Hera and Athena); VIII, 41 ss. (Zeus); VIII, 381 ss. (Hera and Athene); XIII, 23 ss. (Poseidon); on the Indian side, see Macdonell (1981: 18) and in the entries of individual deities about gods riding in chariots; Sparreboom (1986: 18) about chariots as vehicles for the world of heaven.

(28) On the Greek side, see, e.g., Simpson 1969; on the Indian side, Sparreboom 1986: 20f., also Magnone 2012: 102ff.

(29) RV 1.22.20: tád víṣṇoḥ paramáṃ padáṃ sádā paśyanti sūráyaḥ. Sūri may mean a god, a lord or a sage.

(30) RV 1.154.5: tád asya priyám abhi pā́tho aśyāṃ náro yátra devayávo mádanti/urukramásya sá hí bándhur itthā́ víṣṇoḥ padé paramé mádhva útsaḥ.

(31) Nevertheless, the idea of wings as means to reach the supreme region is not unknown even in India: in the strophe immediately before the last quoted one, it is said that no one dares to approach Viṣṇu’s third step, not even the birds flying with wings: tṛtī́yam asya nákir ā́ dadharṣati váyaś caná patáyantaḥ patatríṇaḥ. (It is just a noteworthy curiosity that the word patatrin, literally ‘winged’, hence ‘bird’, is also attested in post-Vedic times in the meaning of ‘horse’).

(32) Phdr. 246a, 247b, 247e, 248a, etc. (ἡνίοχος‎); 246b (ἡνιοχέω, ἡνιόχησις‎). Elsewhere the charioteer is termed ἄρχων‎ (‘commander’, 246b) and κυβερνήτης‎ (‘pilot’, 247c).

(33) The psychology of classical Sāṃkhya reckons, in addition to the five usual sense organs (buddhīndriya), also five action organs (karmendriya): speaking, grasping, going, ejaculating and evacuating.

(34) See Latona 2008.

(35) E.g. BG 2.62–3; 16.18; 16.21; 18.53, and especially the passage quoted below.

(36) The strophe immediately following overtly quotes KaU 3.10, which speaks for the likelihood that the author had in mind the Kaṭha chariot allegory when he composed the passage in question.

(37) This difference is also discussed by Schlieter in this volume.