Alkman's Cosmogony

De Gruyter

ALKMAN'S COSMOGONYAuthor(s): John L. PenwillSource: Apeiron: A Journal for Ancient Philosophy and Science, Vol. 8, No. 2 (November,1974), pp. 13-39Published by: De GruyterStable URL: http://www.jstor.org/stable/40913347 .Accessed: 20/06/2011 10:03

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .http://www.jstor.org/action/showPublisher?publisherCode=degruyter. .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact [email protected].

De Gruyter is collaborating with JSTOR to digitize, preserve and extend access to Apeiron: A Journal forAncient Philosophy and Science.

http://www.jstor.org

13.

ALKMAN 'S COSMOGONY

The 1957 volume of Oxyrhynchus papyri brought to light an interesting document which has now found its way into Page ' s Poetae Melici Graeci as Alkman fragment 5. The only readable portion of this papyrus (fr. 2, cols, i and ii) provides the remains of a commentary on two of Alkman fs poems; the first is significant for the glimmer of light it sheds on the notoriously obscure history of early Sparta,2 while the second (col. i.22ff.) contains , according to the commentator, the fruits of the poet's excursus into the realms of nat- ural philosophy. 3 if this fragment is to be taken at its face value it shows that a cosmogony was composed in seventh century. Sparta of a wholly different order from the normal run of Hesiod-style cosmic genealogies current in this early period, in that it employed the concept of a divine demiurge. It would thus be the earliest known use of the "craftsman" analogy in a Greek cosmogony,4 and as such should be of great interest to students of early Greek thought .

Unfortunately, however, one quick reading of the extant part of the com- mentary is sufficient to show that it is quite impossible to take it at face value. Not only is the exegesis couched in blatantly peripatetic terms (uAn, apxn» xe'Aos abound), but from col. ii.2X onwards the commentator succeeds in destroying any real confidence we might have felt in him through giving an explanation of xou Tpixos axd-ros which his own quotation from the original text of the poem a few lines later (ii. 25f.) shows to be nonsense. 5 Further- more, West ha3 demonstrated by reference to the scholiast on Hesiod Theogong 116 that late commentators are quite capable of avowing the presence of a cosmic demiurge in an early cosmogony despite the total absence of such a concept in the original work.6 Hence it might seem quite reasonable to argue that the poem with which our commentary is concerned was not a cosmogony at all, and that" all we have here is an example of the type of exegesis that Theagenes of Rhegium and his successors applied to the theomachy in Iliad 20, or Herakleitos Homerikos to the forging of Achilles1 shield in Iliad 18.7

Such an idea might well appear to be supported by the silence of Aris- totle, who says nothing about Alkman when dealing with the cosmogonical theories of Homer and Hesiod and of the old poets and mythographers generally. This, however, proves nothing. There is in fact only one reference to Alkman in the whole of Aristotle's extant works, and this is merely to record that g "they say Alkman the poet and Pherekydes of Syros" both died of morbus pedicularis, which reads more like information gathered from some medical treatise or collec- tion of anecdotes than from a first-hand knowledge of the poet's life and works. Moreover, this passage contains the first known reference to Alkman in the whole of Greek literature; and he is not met again until we find him being annotated by Hellenistic grammarians. Thus the most likely explanation of Aristotle's silence (if one is needed) is that he had not read Alkman - a suggestion per- haps supported by the total absence of any treatment of the lyric genre in the Poetics. Indeed, it is unlikely that Alkman became at all well known before the establishment of comprehensive libraries in the Hellenistic period and the rise of literary scholarship as a full-time profession; his style is of such obscurity (v. infra) that he would tend to become a specialist's poet.

We may thus dispose of this last objection, but the others still remain. We must therefore look for some external evidence to support the commentator's contention that Alkman is here describing the coming-to-be of the world.

Apeiron Vol. VIII (1974) No. 2.

14.

Such evidence is fortunately available. A glance through the surviving fragments of Alkman shows that he was interested in matters well outside the range of the average Greek lyric poet. He was notorious among later writers for introducing outlandish and fantastical tribes into his poetry; so much so that Alexander Polyhistor wrote a whole book itep! t&v nap' 'AAxyavi totux&s etpnyevwv (Fr. Gr. Hist. 273 nos . 95, 96). 9 He has a penchant for learned allusions; witness his naming of various species of birds^O and the mention in the Louvre Partheneion (fr. 1 Page) of three different and exotic breeds of horse. ^ This predilection for the obscure and the unusual with respect to the constituents of the geographical and biological world suggests wide- ranging interests and scholarly learning on the part of Alkman. And there is another striking feature of his poetry which should be noted. The language he employs is obscure in the extreme; of the second part of the Louvre Parth- eneion (which is almost wholly free from mutilation) Bowra was led to remark that it "presents difficulties of interpretation so formidable that almost every single sentence has been disputed."12 This is surely deliberate on Alkman 's part; the obscurity in expression goes hand in hand with obscurity and pedantry in nomenclature to create a specific effect. Alkman is the poeta doctus of seventh century Sparta. It should not, therefore, surprise us too much to learn that he composed a cosmogony at some time in his career; a cosmogony, moreover, which, as I hope to show, displays a highly individual character both in terminology and in the motif it employs.

Such general considerations of course prove nothing; but it is possible to go further. First, we already have a fragment of Alkman that could quite easily be an excerpt from a cosmogony. 13 i tefer to fr. 20.

Spas 6' eanxe TpeTs, Se'pos xal xe^yot xinwpav ip i xav xai xexpaxov to /np, oxa odXXei ye'v, eadi'nv 6* a6av oux eaTL.

This passage is most naturally taken as an account of the creation of the seasons, which would indicate at least a passing interest in the origins of the present world order. I shall have more to say about this fragment later.

Secondly, one of the few things of which we can be sure with respect to fr. 5 is that Alkman spoke of an entity called Poros. Now it happens that Poros figures in another passage of Alkman, and in a highly significant con- text. Line 14 of the Louvre Partheneion is unfortunately mutilated, leaving us with the single word ] YepcaxdToi : but the scholiast at this point provides the interesting exegesis otl tov n<5pov eipnxe tov auxov t$ oitb tou 'HaLo6o(u) yeuodoXoynyevv x«ei,on the basis of which editors have quite legitimately res- tored ndpos to the beginning of the line. The preceding line ends ] ap hZoo. TtavTfov: and the inescapable conclusion is that Alkman considers Aisa and Poros to be "the oldest of all ..." Editors for the most part take ucxvt&v to be adjectival rather than substantive, and insert the necessary genitive plural noun at the beginning of line 15. The word that gives the best sense is %euv (or, to use Alkman 's dialect, aiffiv); thus the Loeb edition gives lines 13-15 as:

xpaxnoe (speculative) y]ap Alaa itavx&v xai Ildpos] yepaitdTOi ^ A aia)V'0ni]e6iAos aXxa. ̂ A

15.

However, it is not necessary to rely on the conjectural a i&v to see that Poros in these lines is given the status of an c/r-figure - either "oldest of all gods" or "oldest of all things": and as I believe it is legitimate to assume that what Poros is in one of Alkman 's poems he or it will be in another, it tran- spires that fr. 5 is in fact dealing with one of Alkman fs UrgStter or Urwesen; one moreover whom a different commentator on a different poem identifies with the first entity that appears in Hesiod's cosmogony (rh.116). The case for fr. 5 actually being a cosmogony is thus considerably strengthened.

Thirdly, if we can believe anything our commentator says, it would appear that he is not the first to give a cosmogonical interpretation of Alk- man's text. Towards the beginning of his exegesis he writes (i.26-8) [

' ev 6]e Ta'5iTj xrj 4)6 [^ 'AXjxyav (pug [loXo(yeT) • e]-K%r'o [d].ye$a 6e [xa 6]oHo0vxa ntiuv y]exa xas tcov Xoiic&Iv'iceilpas : "In this ode Alkman discourses on nature: we shall set ^ forth what appear to us to be the facts following the attempts of the rest." This can only mean that the commentator is consciously drawing on what has been said about this poem by previous scholars. These would presumably in- clude the Athenian Philochoros, of whose work on Alkman only the title sur- vives ( Fr. Gr. Hist. 328 no. 24) , and the Laconian Sosibios, an expert on Spartan affairs, whose itep' 'AXxyavos comprised at least three books ( Fr. Gr. Hist. 595 fr. 6) and appears from the one remaining fragment to have taken the form of an exegetical commentary.16 If these learned authorities also inter- preted Alkman fr. 5 as a cosmogony, then there is much more likelihood that it was so in fact.

In view of these considerations I believe there is no case for writing off the present conimentary as an imposition of a cosmogony on to a non-cosmog- onical poem. Of course we cannot accept many of the details contained in the exegesis; the only safe way to a reconstruction of Alkman 's own views is to base our arguments on the poet's words rather than those of the commentator. I shall commence therefore by listing what it is reasonably certain that Alkman himself wrote in this fragment.

The commentator quotes the following lemmata:

1. col. i.22f. - Mo)]aa Xiaaoyai x [e ai]&v ydXiaxa

2. col. ii.3 - a 6e tG) % [

3. col. ii.20 - icpeayfus

4. col. ii.21 - xa' xpi'xog axdxos 1 7 18

5. col. ii. 25ff. - 5yap] xe mcu aeXdva xa' xpi'xov . axdxos <eu)$> xds (or xSs) yapyapuyds (or - as)

We may also be quite sure that the words IIOPOZ, TEKMfiP and GETIZ figured in the original text, since it is their presence that the commentator is trying to explain. This is the sum total of our knowledge as to Alkman fs own words that can be gained from a preliminary reading of the papyrus, and must there- fore form the starting-point of any discussion.

16.

Let us begin with the first lemma. Since this appears immediately after the conclusion of the commentary on the previous poem, it must be taken from the opening lines of the one we are discussing. Hence we may assume that Alkman began his odef as was commonly done, with a prayer to the Muse. However, it seems that this particular Muse was somewhat out of the ordinary, for the commentator, after noting the composition of the chorus who performed the ode20 and declaring his intention to set forth xa 6oxo0vxa nytv, continues with the remark (i.28f.) r ns [yev] Mo'5aa[s] Suyaxepas u>s Miyvepy[os .] Tag eye- [veaXdynae. .This statement receives support from Diodorus Siculus (4.7.1 = Alkman fr. 67) : oXiyoi xfov TtoinT&v, ev ols eaxi xa! 'AXxydv, duyaxepag ontocpaivovxai (sc. xas Mouaas) Oupavou xa! Tns. However, in the two other extant instances where Alkman invokes the Muse he addresses her as "daughter of Zeus". 21 Now it is possible that both the commentator and Diodorus were wrong; but it does not seem likely that Diodorus or his source22 would make much a specific state- ment regarding Alkman unless there were some grounds for doing so. I suggest that the basis for this assertion was provided by our present poem. Although it is perhaps dangerous to draw conclusions from the minutiae of the comment- ator's form of expression, it nevertheless remains that he does not say, e.g. M&aa Aiaaoyai xe.oiwv ydAiaxa] oxt 6 'AXxyav $uyaxepas aicotpai vexai xas Mo'5aas Oupavou xa' Tns, "'Muse, of the gods 'tis thee I most invoke1: because Alkman believes that the Muses are daughters of Heaven and Earth"; rather he is saying (in a sentence not immediately following the lemma) , " (Notice that Alkman here makes) the Muses daughters of Earth, as Mimnermos too gives their ancestry." That is, Alkman goes on himself to say exactly who this Muse is that he addresses "especially, among the gods" - to wit, the daughter of Earth; and the commentator is merely citing a parallel instance of such a belief about the Muses' (or some Muse's) ancestry in another poet. Alkman thus commences his song with a prayer to a Muse more primeval than her whom he elsewhere invokes; and it is not unreasonable to suppose that he chooses her "especially" precisely because he wants the aid of a goddess old enough to tell "how Heav'n and Earth/Rose out of Chaos. "23

Lemma no. 2 introduces us to the crucial passage of the papyrus; and it is most unfortunate that it has been so mutilated. All that remains of it it ex 6e xuj ic[; and any reconstruction must be extremely hypothetical.2 4 The only certainty is that in this or some preceding lemma lost in the break between cols, i and ii Poros, Tekmor and Thetis made their appearance; the figures who, according to the commentator, fulfil the chief roles in the cosmo- gony. As Alkman 's own version of their activities is lost to us, we must see if anything more can be extracted from the commentary than these three names. I begin with a summary of the sequence of events given by the comment- ator (in the section beginning at X[eyei] o3v 6 'AXxydv ii.8ff.):

A. In the beginning there is n uXn itdv[xu>v xexa]payyevn xa! cncdnxos.

B. The coming-to-be of 6 xaxaaxeudtcoov] udvxa.

C. The coming-to-be of Poros.

D. Tekmor arrives after Poros has "passed on" (napeXSdvxos ii.13).

E. The coming-to-be of Thetis; Poros and Tekmor become apxn and xe'Xos respectively, and Thetis acts upon the uXn as a bronze- smith upon bronze.

I shall now consider these points one by one to see what may be gleaned therefrom.

17.

A. fiAn, as has been pointed out by others, cannot have been used by Alkman in the sense of "matter" or "material cause". 25 Moreover, it is largely lost labour to try and discover what, if anything, in Alkman 's text could have been interpreted as uXn by the commentator. His remark is best taken as a "rationalisation" of Alkman's account, in the best Aristotelian tradition, to bring it into the framework of his own thought pattern. Alk- man himself may or may not have chosen to describe the pre-creation state of things (Hesiod certainly felt no need to do so) ; but even if he did, it would not have been in the abstract philosophical terminology of the commentator.26 (For further remarks on this matter see below, n. 34.)

B. The identity of 6 xaxaaxeuaccov ndvxa is something of a puzzle. Modern scholars who have dealt with this fragment have identified him with Thetis, despite the gender.2? But it seems inherently unlikely that even this commentator would (a) make such an elementary grammatical mistake, and (b) mention in extremely vague terms (efxa[Yeve]adai xivct cpnatv xov >taxaaxeua[covxa] uavxa) at line 11 a figure he introduces by name in line 15. Doubtless in the commentator's eyes he corresponds to Aristotle's Prime Mover, without whose influence none of the later events could take place; he may therefore be a pure invention. However, it is possible that he is based on one of the figures in Alkman's poem; I shall say more about this later {infra p. 29 ).

C. Poros. We have already noted that this entity figures elsewhere in Alkman's poetry, in a context which confirms its early appearance on the world scene. There has been some dispute as to how one should interpret this figure and what its coming signifies. The usual English translation of Poros in line 14 of the Louvre Partheneion gives "Device" (so e.g. Edmonds, Bowra) , while Herzog-Hauser (P.-W. s.v. Poros) renders "Betriebsamkeit" . Since the discovery of the present fragment, other suggestions and elaborations have been made: thus Bowra on fr. 5, "(Poros) is the way of contriving things and sets them going . . . [elsewhere] ndpos is closely related to Aiaa and seems to stand for initiative as opposed to destiny" ( glp 26); and now on fr. 1.14, "Just as somewhere else he makes [Poros] a shaping power in his cosmology, and evidently means it to signify 'Device1 in the sense of the intelligence which shapes situations, so here he must have done something of the same kind" ( ibid . 40) . Similarly Lobel (55) gives "way of contriving things or beginnings," and is followed by his reviewers Barrett (689) "the way or means of doing things?" and Page (20) "the way of contriving"; Schwabl (1467) offers "Findigkeit." West in his first article gave "'provision' with the accessory idea of 'apportionment'" (1.155); in his second he is less specific, saying merely that Poros is required to cure the primeval substance of being aitopos (2.2)." The interpretation of Frankel is "6ffene MSglichkeit (oder Zuganglichkeit) " (290) ;29 his reviewer Burkert gives simply "Weg" (827).

It seems, therefore, that, although the actual meaning and significance of Poros may be in dispute, it is nonetheless generally agreed that it con- stitutes a hypos tasis of some abstract idea. However, a study of the instances of itdpos as an ordinary word down to the end of the fifth century suggests that this is most unlikely.

Etymologically , the word is part of the neipa) group (v. Boisacq s.h.v.), which also includes nepdvn* rcepovav, Ttdpitn9 itopi'ca), uopeuu), itop^yds. netpa) itself carries the basic meaning "pierce"; in Homer it is most commonly used of spitting or skewering meat ( Jl. 1.465 = 2.428 = Od. 3.462 =12.365 = 14.430, Jl. 7.317 = 24.623 = od. 19 .422, Jl. 9.210, Od. 3.33, 14.75, H. Merc. 121) , but also of harpooning

18.

fish ( Od. 10,124;, of wounding with a spear (JI. 16.405, 2.0.479, 21.577), of boats "cleaving" a passage through water (od. 2.434), of heroes fighting their way through a seething mass of warriors ( II. 24,8 = od. 8.183 = .13.91 = 13.264; , and, figur- atively, of one "pierced" with pain ( II. 5.399;. itdpos is formed from netpw in the same way as itdxos from iceixu>, andpos and cntopd from auetpw, cpddpos and cpdopd from (pdeipa), and ayopa from ayeopa): and as e.g. cpSdpos ("ruin, destruction") is the result of the action of the verb <p$etpu) ("destroy") and iyopd ("assembly") the result of the action of dyetpw ("assemble"), so itdpos will be the result of the action of iteipw ; literally, the "hole" or "passage" left after whatever now contains the icdpos has been iceitapyevov. Let us now examine how the word is used in writers down to the end of the fifth century.

(1) Literal uses.

(a) "Hole through a solid object". The only author of the period who certainly employed the word in this sense is Empedokles: fr. 100.17 of the mouth of the klepsydra; fr. 3.12 of the passages or "pores" in the body by means of which sense-perception is possible, itdpos in this literal sense appears to have been an important item in Empedokles1 philosophy: v. Plato Meno 76c, Theophrastos de sensu 12, et al.

Tcdpos is used by later writers referring to "holes" in the presocratics - e.g. the "passages" in Anaximander ' s circles through which the stars and moon appear (Hippol. Ref. 1.6.4f = DK 12 A 11), the "channels of perception" (aiadnTixoi Ttdpoi) in Herakleitos by which men derive intelligence from to itepie'xov ( Sext. adv. math. 6,129f. = DK 22 A 16), and the "passages" which carry sensations to the brain in Alkmaion (Thphr. de sensu 26 = DK 24 A 5) ; but of course it is by no means possible to say whether the philosophers themselves used the word.

(b) "Passage" in the sense of "way through" or "across" some obstacle.

(i) "Ford" (i.e. the "way across" a river - the predominant sense in Homer). So II. 2.592, 14.433 = 21.1 - 24.692; H. Apoll. 423; H. Merc. 398; Pi. 01. 1.92, 2.13, 6.28, 10.48; Soph. Trach. 564; Eur. Pho. 730, 825; Thuc. 7.78, 80.

(ii) "Strait" (i.e. the "way through" a land mass for ships). So Pi. Nem. 4.53 ('idviov ndpov = straits of Otranto) , 9.41 (? perh. should be under (i)), fr. 189; Aesch. Pers. 747, 875, Supp. 546; Hdt. 7.176, 8.15; Eur. I.T. 253; Thuc 6.48 (of the straits of Messina, though perh. better here interpreted as "the way across from Italy to Sicily").

(iii) "Passage through" shallows. So Hdt. 4.179.

(iv) Used of Darius1 bridge over the Danube Hdt. 4.136, 139, 140, 7.10.

(v) Used of Xerxes1 artificial causeway over the Hellespont: Aesch. Pers. 722; Hdt. 7.34, 35, 36, 8.111, 115, 117, 126, 9.120.

(c) More generally, "path", though still retaining the idea of traversing an obstacle, so that the division between the uses listed here and those under (b) is somewhat artificial.

19.

(i) Of the "paths" of the sea (cf. the use of iteipw at Od. 2.434 noted above p. 18 ). So Od. 12.259 (itopous aXds) ; Hes. Th. 292 (6ia3as itdpov 'ftweavoto); [Arion] fr. 1; Aesch. Supp. 844, Pers. 367, 453; Soph. Ajax 412; Eur. fr. 304.2 (by which time simply = "journey across the sea"), Hyps. fr. 64.103.

(ii) Of the "path" formed by a river on its journey to the sea; hence "river-bed" or simply "river". (Common in Aeschylus but only two instances elsewhere.) So Bacchylides 9.42; Aesch. Pers. 493, 501, 505, 864, Sept. 378, Prom. 532, 806, Cho. 72, 366, Eum. 293, 452, fr. 69.4; Eur. H.F. 839.

(d) Simply "path", "fairway", "course", "route". So Aesch. Prom. 280, Supp. 546, Aga. 910, 921 (the last two of the purple cloth on which Agamemnon walks), Eum. 770; Soph. Phil. 704; Eur. I.T. 116, 1325, 1388, Hyps. fr. 64.85; Hdt. 7.183, 8.76.

(e) As an extension of <c) (i) above itdpos is found in Euripides meaning simply "sea"; so And. 1262, Hel. 130, Tr. 82.

(2) Metaphorical uses.

(a) "Path of life". Pi. Isthm. 8.16 (eXiaawv 3iou itdpov) - life itself being viewed as the "obstacle" to be traversed; the usage probably derives from (1) (c) above.

(b) "Paths of the mind". Aesch. Supp. 94 (ddonioi xe Teivouaiv itdpot, sc. Atos itpaiu'6a)v) . If this is indeed metaphorical, it presumably derives from (1) (c) or (l)(d); but it could well be literal, in which case it belongs rather under (l)(a) - cf. R.B. Onians, The Origins of European Thought /Cambridge, 1951, p. '29.

(c) "Path of song". Emp. fr. 35.1 (eXe'5aoyou es itdpov uyvanO- again most likely deriving from (l)(d).

(d) "Way out of" (sc. trouble, which is regarded as an obstacle to be traversed; this usage based on (l)(b)). So Lyr. ad. 1019.7 Page (tu - s.c. T'5xa - 6* ayaxavi'as itdpov eXdes ev aXyeai) ; Aesch. Prom. 59 (aynx<*vu>v ndpov) ; Eur. Ale. 213 (itdpos xaH&v) , I.T. 897 (absolute); At. Knights 759 (otynxavwv itdpoos).

(e) "Way to achieve" - viewing the obstacle to be traversed in this case as that which lies between one and the desired object; again based on (l)(b). So Aesch. Supp. 806 (etyepuyas ito'pov) , Prom. Ill (x^xvrIS itdpos); Hdt. 2.2 (itdpov to'5tou aveupetv) , 3.156 (itdpos aXtifoios); Eur. Med. 260 (itdpos avTiTtaaaSai) , Pho. 984 (xpnycrrujv itdpos) , Ale. 1162 (a6oxn'TU)v rcdpov) = And. 1287 - Bacch. 1391 - Hel. 1691 - Med. 1418, H.F. 80 (itdpov aurrnp i'as) , Supp. Ill (xPnyctxcav itdpou); Ar. Peace 124 (udpos o6ou) , Eccl. 653 (uyaxLcav itdpos) .

(f) Hence absolute: "way", "method", "device". So Aesch. Prom. 477; Eur. I.T. 875, I. A. 356; Ar. Wasps 308 (a parody of Pi. fr. 189 - see (l)(b) (ii)) , Thes. 769.

(g) Absolute in the sense of "profit", "revenue": Ar. Frogs 1465.

20.

This survey makes it clear that the use of ndpog absolutely to mean "device" was not particularly common even in the fifth century, and that it developed from the figurative use with genitive or infinitive meaning "way of achieving" which first appears in Aeschylus. Moreover, the one instance in Aeschylus where the word is used absolutely - Prom. 477 oias xe'xvas xe xal itdpous eynadynv - suggests a deliberate metaphor on the part of the poet, who feels the need to illuminate his unusual use of rcdpous by adding the commoner, readily compreh- ensible xe'xvas. Thus I think this passage clearly shows that in the earlier part of the fifth century itdpos on its own would not readily be taken as meaning "device". It is not until Euripides that such a usage is possible; and by his time the word had become less specific in other respects as well (cf . (1) (e) and (1) (d) , where Euripides1 use of ndpos = "path" is more more matter-of- fact than Aeschylus1). On the other hand, the earliest uses of the word are all concrete; in Homer it signifies either "fordM ( (1) (b) (i) ) or the "paths" of the sea ((1) (c) (i) - the sea-routes which result from the action of neipw as used at Od . 2.434 navvuxi'n yev n ye (sc. vnos) xal nw/rceipe xeXeuSov). Moreover, of the eight uses in Pindar seven are concrete ("ford" or "strait") and only one figurative, this last being a conscious metaphor and having nothing to do with the meaning "device" (see (2) (a)); the single instance in Bacchylides is concrete ("river-bed"); and of the twenty-nine places where the word is found in Aeschylus it has a concrete meaning in twenty-four and a figurative one in only five (including that cited under (2)(b)). This evidence regarding the use of udpog in archaic and early classical times strongly indicates that we must seek a more concrete interpretation of Alkman fs Poros than those which have hitherto been supplied. 3^

The fact that early uses of the word very often refer to waterways and means of crossing them may at 'first sight be thought to give some support to the "waste of waters" which West now believes to have been the primeval state of the world in Alkman's cosmogony. 31 Citing the ndpous aXds of od. 12.259 as a parallel, we might then say that the coming-to-be of Poros signifies the opening up of some path across these waters, which could serve as a locus standi for the presumed demiurge Thetis - herself, as West points out,J^ a sea- goddess. Unfortunately this interpretation of Poros must fail before the principle enunciated above (p. 15) , viz. that what Poros is in one of Alkman's poems it will be in another. "Path through the primeval waters" may look well enough here, but it surely cannot mean this at fr. 1.14. "Passage", "path", or "hole" is indeed the sort of translation we want, but across what or through what is it to be?

Let us put this question in a different way, having regard to the derivation and basic meaning of udpos; what, in Alkmanfs view, is likely to have been the entity which required itapnvcu as the first event in the coming-to-be of the cosmos? The answer, I think, lies in the following often-quoted fragment of Euripides (fr. 484 N.) :

xou'» eyos 6 y'3$os otXX' eyns ynxpbs Ttdpa is oupavds xe yald x' ?|v yopcpn yi'or enel 6' £xwpio%r)oa.v aXXnXoov 6i'xa xi'xxouai itdvxa xave6u)xav eis cpdog, 6ev6pn9 itexeivd, Snpag, ous %' aXyn xpecpeu yevog ie Svnxuiv.33

It is the yopcpn yua oupavou xal yns that is neTcapyevn in Alkman; and the resulting Ttdpos is none other than the "gap" or "passage" that now exists between Heaven and Earth; as the scholiast on fr. 1.14 so rightly says, Alkman 's Poros is the same as Hesiodfs Chaos.34 It is becoming increasingly clear that the motif of

21.

the separation of earth and heaven as the first stage of cosmogony played a large part in Greek non-philosophical speculation about the origin of the universe. It appears in Hesiod in two guises (see Kirk-Raven loc.cit.): first of all in the bald statement?! tol yev upuhioxa Xaos y^t* (Th. 116), and later in the myth of the castration of Ouranos by his son Kronos (rh. 154 ff.). We find it also in the fragment of Euripides cited above and at Diodorus Siculus 1.7.1 ( xcxxa yap xhv e£ apxns t&v oXujv auaxaaiv yi'av ex£tv i6eav oupavdv xe xai yTW, yeyeiyyevns aux&v xns cpuaews* yexa 6e xauxa 6iaaxdvxu)V x&v atDydicav a*' aXXnXwv xbv yev xdayov itepiXa$etv Sitaaav xnv ipcayevnv ev auxij auvxa£iv); while in the "Orphic" tradition we. have Ap. Rhod. 1.496 ff. (= DK 1 B 16):

n'ei6ev (sc. Orpheus) 6' is yoaa *at oupavbs r)6e SaXaaaa xo npiv en' aXXnXoici ytq ouvapnpoxa yopcprj veiMeos e£ oXoolo 6iewpt§ev aycpis exacxa* n6* is eyne6ov atev ev atdepL xexyap exouaLV aaxpa aeXnvoa'n xe xal neXioio xeXeudot, kxX.

and the account of an Orphic cosmogony given by Athenagoras (18, p. 20 Schwartz = DK 1 B 13) :

oSxos 6 *HpaxXfjs eyewT'ozv uitepyeye^es (J)dvs o auynXr)po'5yevov unb 3tas xoO ysY^^k^os in Tiapaxpt3ns ets 6uo eppayn. to yev o5v naxa xopucpnv auxou Oupavbs elvai exeXea^n, xb 6e kcxxo) evex^^v rfj .

The motif is of course a common one all over the world, representing mythically and symbolically the process of the differentiated world being produced from the undifferentiated. ^5

If this is what Alkman means by the appearance of Poros in his cosmogony, we must now examine whether a similar interpretation will fit Poros in fr. 1. As far as it can be reconstructed, the extant portion of this latter poem opens with a list of the Hippocoontids who were slain by Herakles. Then comes the Poros passage (13 ff.):

] ap hZoa. navxujv hou ndpos] yepoaxaxoi* ''5%r' 6' om]e6iXos aXxa* yn xls avd]p(j5ftu)v Is ipavbv itoxTfa^a) yn6e itri]pTixu) yayfjv xdv 'A(ppo6i'xav ... (the following lines name other goddesses)

In his second article, West puts forward a similar proposition in respect of this passage as he does on Poros and Tekmor in the cosmogony (cf. supra p. 17): i.e. "Strength without wisdom, airopos, e£aiaios, fails; its failure is the triumph of ndpos and Alaa. "36 Others, as we have seen (ibid p. 10), prefer something like "Device", despite the fact that, as we have also seen, such a rendering is most improbable for this early period. But if we accept Poros as meaning the "gap" between earth and heaven here too, it makes very good sense in the context. In the previous lines, Alkman has been enumerating slain heroes; line 13 in fact is the beginning of a new section of the ode, in which the poet moralises upon their deaths, "Aisa and Poros, oldest of all beings, (defeated them)37f»» he says, "their strength, shoeless, 38 collapsed; let no man fly to heaven, nor attempt to marry Aphrodite.11 The prohibitions are an expression of the common Greek maxim that man should not try to over- reach himself; but the form of the expression here is interesting. What in

22.

fact stops man from "flying to heaven and marrying goddesses?" Alkman himself gives the answer. First, Aisa: it is not the lot of man to mix with divin- ities. 39 Secondly, Poros: the "path" that now divides earth from heaven is incapable of being negotiated by man. 40 Thus for the "Doom and Device" which the Loeb translation gives for Aisa and Poros, we should substitute the equally alliterative pair "Portion and Passage".

D. Tekmor. "Poros having passed on" says the commentator1 s paraphrase of Alkman, "Tekmor follows. "41 As the language employed to describe this event (which is given, moreover, in oratio obliqua) conflicts somewhat with the commentator's equation of Poros with apxn and Tekmor with xeXos in the next clause (oratio recta) , it may be that we have here something fairly close in meaning if not in actual phraseology to what Alkman himself said. I think it is legitimate to infer, therefore, that Tekmor appears subsequently to Poros (just as Gaia etc. appear subsequently to Chaos in Hesiod) ; i.e. it is a figure that appears upon the cosmic stage only after Heaven and Earth are separated.4 2 we must now attempt to determine what his coming signifies.

Modern interpretations fall broadly into two categories, according to whether or not they connect Tekmor in the cosmogony with Aisa, who is coupled with Poros in the Louvre Partheneion. * The majority do not: Lobel (55) gives "boundary or end", and is followed by Page (20) and Bowra (26); West in his first article renders "principle of differentiation11, adding that "it represents a principle or potentiality rather than a specific event (sc. in the cosmogonical process)",43 while in his second he gives the basic meaning of Texjjoap as "boundary -mark or sjlgn" , saying again that "(it) is recognisable as a principle of differentiation", but this time adding (as a parallel to his second definition of Poros - cf. supra pp. 17, 21 ) that Tekmor is necessary to save the world from being axexyapxos. ̂ West is followed by Vernant (49 - cf . supra n. 28) , who also makes much of the later astronomical significance of xe'xyap. Barrett (689) gives "Final consummation?" and Schwabl (1467) "Erfullung". Representative of the other school of thought is Frankel, who interprets Tekmor as a "Variante zu Schicksal . . . ' (bindende) Festlegung1 " , saying elsewhere of Alkman that "er verfiigte so aktiv 'iber die spekulative Denkweise, dass er den Namen und Begriff des einen Partners frei variieren konnte."45 His reviewer Burkert, however, counters this interpretation - "Dagegen spricht jedoch schon das Genus von Texyop , das nie gleich alaot mit iccJpos ein Paar bilden kann" - and instead gives rWegzeichen" ("doch gehort xe'xywp zum 'Weg1 ( ndpos) ... als Hilfe und Bestatigung" ) . 46 However, it is surely better to agree with Frankel: if in one passage Alkman calls Poros and Aisa "oldest of all beings", and then in another gives the pride of place to Poros and Tekmor, it seems not unreasonable to suppose that Aisa and Tekmor stand for a similar idea. With this in mind, let us now revert to the method of inquiry which has already proved successful in the case of Poros; i.e. an examination of the occurrences of the word xexywp (or xe'xyap, to give it its more usual form4 7) in writers down to the end of the fifth century.

(1) The word xe'xyap outside Homer normally carries the meaning of

"sign" f" clue", "that by which something may be shown or proved", thus correspond- inq to the post-Homeric use of the cognate verb xexyoa'poyca (v. L.S.J. s.h.v. II) to mean "judge from signs and tokens". It is often merely a poetic equivalent of xexynptov.

So Hes. fr. 273 (Merkelbach and West) = Mousaios B 7 (DK) (f)6i> 6e xou ̂xo Tio-deoSai, oaa dvnTOiaiv eveiyav/aSavaxoi , 6eiAS>v xe xat eadXwv xe'xyap evapyes:

23.

"Sweet is it to know this too, what things the immortals dealt out unto mortals, a clear sign (that distinguishes) the noble and the worthless"^**) ; Vl.Nem. 11.44 (to 6' ex Albs avdpwitois aacpes oux eitexai/xexyap); Aesch. Supp. 483 (is l6o)ql xna6* aq>i£eu>s xexyap), Prom. 454» Aga. 272, 315 (coupled with a'5y3oXov) Cho. 667, Eum. 244 (of Orestes1 tracks), fr. 530.30 Mette; Hp. Mul . 2.123 (aXXncu 6e &X'j) icn xe'xyap taxexai, = "symptom"): used of heavenly bodies Horn. Hymn. 32.13 (xe'xywp 6e $poxots xai an'ya xexuxxai, of the Moon) and Eur. Hec. 1273 (xuvbs TctAaivns afjya, vauxi'Xois xe'xyap).

(2) Homer's xexywp, however, appears to mean something different. True, "siqn"* could be applied to XI. 1.526 - but this would be the only instance in nine uses, and it seems safer to assume that another interpretation should be given here too. In fact, as Leaf says on II. 7.30, xe'xywp is closely con- ^ nected with the Homeric use of xexyaipoyai in the sense of "ordain", "foretell". Leaf therefore gave the basic sense as "a thing established"; a better render- ing, however, bearing in mind the Homeric xexyaipoyai, would be "ordained thing", "ordained end" , or simply "ordinance" .

Thus when Zeus says of his nod (II. 1.525f.) xouxo YaP e£ eye§ev ye yex* adavaxoicu yeyiaxov/Texyuip , he is emphasising that that to which he gives his approval is irrevocably ordained to happen; when Apollo says to Athene (il. 7.30f.) uaxepov a'5xe yaxn'aovx', eis 5 xe xexyiap/'lXiou eupwaiv, he means "they will fight on until they find the 'ordained end' of Ilion (i.e. the time when Ilion is fated to fall)" - the phrase xexyu>p 'IXlou occurs on three other occasions in this sense (II. 9.48, 418, 685); when Poseidon travels to Aigai, he takes three paces to 6e xexpaxov I'xexo xexyajp/Atyas (II. 13.20f.) - i.e. he reaches his "ordained goal", the place he himself has determined as his destination (it being a god's prerogative xexyai'peadai) ; and when Eidothea comes to the rescue of Menelaos and exclaims (Od. 4.373) is 6n 6n$* evl vn'a^ ep'5xeai% ou6^ xt xexyoop/eupeyevai 6'5vaaoa, she means "you cannot discover the appointed end (sc. of your troubles)" - these lines being repeated by Menelaos himself in the first person a little later (4.466). The final Homeric example is a little difficult; when in II. 16 the trace-horse of Patroklos' chariot is killed by Sarpedon's spear and threatens to cause chaos in the rest of the team Patroklos' squire Automedon comes to the rescue - xoto ybv Auxoye6cov 6oopixXuxsos eupexo xexycop (472) - by cutting the harness. -The normal practice here is to translate xe'xywp by "remedy" or "solution"; it can, however, be made to conform to its normal Homeric signification if we view the whole situation as some sort of puzzle, the solution to which is already "determined" or "ordained" by the compiler (here probably yolpa, whose influence on the events which culminate in Sarpedon's death at 503 is stressed in lines 431-61).

Texuwp in this sense is not solely confined to Homer. It is found (in the form xexyap) in two places in Pindar: Pyth. 2.49 (Seos aitav en! eXiu'6eaai xexyap avtfexai, "God accomplishes his every end upon the thought") and fr. 168 Bowra (iao6ev6pou xexyap ataivos Xaxotaai, "getting as their ordained span the lifetime of a tree", of the Nymphs). It also occurs in Orpheus' song of creation at Ap. Rhod. 1.499 - cf. supra p. 21 and n. 47.

We are thus faced with a choice between two distinct possible inter- pretations of Alkman's Tekmor, both of which correspond to known early uses of the word. However, it seems fairly clear that the Homeric sense is the one required here, and for the following reasons. First, to define Tekmor in this cosmogony as "sign" or "clue" is in itself quite meaningless; nor is West's extension of this (which a little sophistry may support by reference to Hes. fr. 273) to "principle of differentiation" much better; differentiation has already appeared, a fact expressed in mythic terms by the coming-to-be of Poros (supra

24.

p. 21 and n. 35). Secondly, the fact that Alkman chose to use the Homeric form of the word suggests that he had the Homeric rather than the non-Homeric meaning in mind. 5° Thirdly, xexywp in the sense of "ordained end" obviously has far more connection with alaa than it would in the sense of "sign" or "clue", which at once renders it a more attractive interpretation (cf. supra p. 22)-

We may, in fact, go further in establishing a correspondence between t ex ywp and alaa. If we look at the Homeric expression xe'xywp 'IAiou, which the Achaians, it is said, will eventually "find" or "discover", we are immediately struck by the fact that tcxvudp here is practically identical with alaa - for the "fate" or "lot" of Ilion and the "ordained end" of Ilion can hardly be called dissimilar conceptions. xexyajp in this sense has strong affiliations with yol'pa, the im- portance of which in earliest Greek cosmo logical thought was well illustrated by Cornford.51 I suggest, therefore, that the arrival of Tekmor in Alkman 's cosmogony signifies the arrival of that determinative power which presides over world order; that power which elsewhere in Alkman fs poetry prevents man from ascending to heaven and marrying goddesses,, that which in Homer controls the actions even of the gods,52 that which in Anaximander "necessitates" the pay- ment of retribution by encroaching opposites,53 that which in Herakleitos pre- vents the sun from "over-stepping his measures".54 Altaian's Tekmor will thus join the well-known company of Aisa, Moira, the Erinyes, Nemesis, Necessity and the rest as yet another manifestation of the force which determines that the order present in the world shall not be disrupted.

The appearance of a determinative power of this type at the dawn of creation has parallels in both Greek and non-Greek material. In Damaskios1 account of "the Orphic cosmogony according to Hieronymos and Hellanikos", the primordial union of earth and water produces "Ageless Time" accompanied by'Avdvxn and *A6pdaT£ia: the parallel version given by Athenagoras tells that after the egg generated by Herakles broke in two to form Ouranos and Ge (cf . supra p.. 21) , the first offspring of this pair were the Fates Clotho, Lachesis and Atropos55 - a sequence of events which shows a marked similarity to Altaian's Poros followed by Tekmor. Along with this we have the references in Orphic literature and Pherekydes of Syros to the primacy of Chronos , Time, who may well himself have been viewed as a sort of determinative power; for in Anaximander • s fragment time appears as the "assessor of damages" in the litigation between the cosmic opposites.56 In non-Greek sources, one is particularly struck by the parallel provided in the Mesopotamian myths of creation which incorporated the idea that nothing could be said to exist until its fate had been determined. 57

It is at once apparent that if the interpretation of Tekmor given here is correct, then Alkman 's cosmogony embodies a significant divergence from that contained in Hesiod's Theogony. For in this latter account, after Chaos, the Gap, has appeared, revealing Earth and Tartaros "in a recess of earth", the ̂ next arrival on the scene is Eros, whereas there is no mention of the Moirai ̂

until nearly a hundred lines further on (217) , where they figure among the children of Night.58 Eros appears very early (as far as we can tell) in the theogony of Akousilaos, also. 5$ But these authors are among these for whom cosmogony and theogony are the same thing: i.e. their account of cosmogony is one of a succession of divine matings and parturitions. Eros's presence is therefore essential as the driving force behind this process; hence his early appearance. But for Alkman, as we shall see, the universe was not self-created in this way; so for him Eros would be superfluous at this primeval stage. What is needed, and what he gives us, is a power which will regulate what happens next.

25.

E. Thetis. First Poros , then Tekmor, and now, seemingly, Thetis, to whom our commentator assigns a demiurgic role. This has occasioned express- ions of wonder among modern scholars: "rather surprising" (Lobel) , "astonish- ing" (Barrett), "ratselhaft" (Erankel) are typical. 60 Certainly we should not have expected to find a Nereid cast as the demiurge from anything we read about these goddesses in the rest of Greek literature. Has this document, therefore, given us something quite new?

Burkert and West seem to think that it has. Noting that Thetis else- where in Greek mythology is primarily a sea-goddess, they deduce that Alkman's was a water cosmogony: i.e. that it belongs to that class of creation myths in which the world arises from a primeval waste of waters. Thus Burkert: "Thetis als die beruhmteste der Nereiden konnte dann vielleicht die Urflut reprasentieren"61; and West: "The clearest indication of water is the presence of Thetis. "62 Later in this same article West summarises Alkman's cosmogony thus: "It begins ^ 3

... with a waste of waters, characterised as trackless and g, featureless, ^ 3 ana the appearance of a deity who can make something out of it." The theory is, however, most improbable. It depends to some extent on a mis- interpretation of Poros and Tekmor (cf . supra ) ; but an even bigger drawback is pointed out by West himself in the very next sentence: "Nothing similar to this is found in Greece. "65 now it is true, as West also notes, 66 that there are traces of a water-cosmogony - or rather, of an Urflut - in Homer and Thales. But in both these cases the Urflut has no need of any agent to act upon it, being perfectly capable of commencing generation from itself. Thus in Homer we find that Okeanos is termed the "begetter of all", 67 which, if it is to be taken as a serious cosmo-/theological statement, merely means that Okeanos was the first parent in some cosmic genealogy; 68 while Thales1 primordial water took on the nature of a Milesian apxn, "the original source of all existing things, that from which they first come-to-be and into which they are finally destroyed, the substance persisting but changing in its qualities" as Aristotle describes it (Metaph. A3, 983b 8) , and was itself in some way alive69 - certainly no question here of inert matter needing some divine agent to work it into shape. West is thus forced to go for his parallels to the Near East, citing Genesis 1 and Enuma elish, and concludes that "in view of these parallels it would appear that a Semitic-type cosmogony lies behind Alkman.""^ Burkert, moreover, cites Alkman's supposed Lydian origins to suggest that any Near Eastern influence we find in his work "cannot be very surprising."'1 But it seems inherently unlikely that Alkman, whatever his birthplace,72 would compose for performance at a Greek religious festival an ode whose material was culled from an alien mythological tradition. 73 Moreover, in the case of what we have so far discovered about Alkman's cosmogony we have not had to go beyond Greece for parallels; indeed, on the interpretation offered above, Poros and Tekmor belong to the mainstream of early Greek cosmological thought. We should therefore attempt to interpret Thetis1 appearance also without recourse to the hypothesis of Near Eastern influence.

Little can be gathered from an examination of Thetis1 already known deeds. She was a Nereid, a daughter of the Old Man of the Sea (II. 1.358, etc.), and normally lived in the sea with her father. Like her father, she was capable of sel f -metamorphosis , a power in which she indulged to resist the attentions of Peleus.74 Homer's Hephaistos terms her 6eivn xcu ou6oin (II. 18.394), and goes on to describe how she and Eurynome rescued him after his mother Hera had cast him out of heaven because he was lame. 75 Rescuing the gods from tricky situ- ations seems to have been one of Thetis1 specialties; for she is also recorded as having given sanctuary to Dionysos from the anger of Lykourgos,76 and as _- having saved Zeus from a plot hatched against him by Hera, Poseidon and Athene. But possibly the most well-known story about her is that she was destined to

26.

bear a son who would be greater than his father; Zeus, being warned in time by Metis (or Prometheus or Themis as other versions have it) , ceased his attempts to seduce her and gave her instead to the mortal Peleus, whereupon she became the mother of Achilles. 78 Later she returns to the sea (though her husband Peleus is still living - il. 18.433f.), which is where we find her at the time of the Trojan War.

There is nothing in this to suggest that Thetis was anything more than a mermaid with friends in high places. But there is a fact concerning her wor- ship that at first sight appears as though it might be significant with respect to Alkman. Pausanias tells the following story regarding the shrine of Thetis at Sparta (3.14.4, Loeb trans.):

The sanctuary of Thetis was set up, they say, for the following reason. The Lacedaemonians were making war against the Messenians, who had revolted, and their king Anaxander, having invaded Messenia, took prisoners certain women, and among them Cleo, priestess of Thetis. This Cleo the wife of Anax- ander asked for from her husband, and discovering that she had the wooden image (Sdavov) of Thetis, she set up with her a temple for the goddess. Leandris did this because of a vision in a dream, but the wooden image of Thetis is guarded in secret.

Much is uncertain about early Spartan chronology, but the Messenian Revolt is generally given as c.650, and Alkman 's floruit as 630. Thus this shrine might have been set up quite recently when Alkman composed his cosmogonical ode; there may be those bold. enough to suggest that this very ode, in which Thetis appears to play such a vital role, was written for performance at its dedication.

Such a suggestion would be pure speculation; in any case, I think it must be fairly obvious by now that we have been following a false trail. There is not a jot of evidence anywhere in Greek literature to connect the Thetis of myth- ology with any kind of demiurgic activity whatever - not even, as we shall see, among those later commentators who insist on turning the Homeric poems into a cosmological allegory. Nor do we receive any guidance by examining her cult, since our knowledge of this is virtually confined to the fact that there were shrines dedicated to her in various parts of Greece. The truth of the matter is that we have been led into a blind alley by our present commentator. We have already had occasion to note (e.g. supra p. 13) that his reading of Alkman's text was none too careful, and that his critical acumen is, to say the least, open to doubt. It need occasion no surprise, therefore, if we find that he has made yet another blunder here; and this, I believe, is precisely what he has done. When he saw the word GETIC (or CETIC) in the original poem, he immediately jumped to the conclusion that Alkman was talking about the Nereid Thetis; a mistake which modern editors and scholars have helped perpetuate by printing the word with a capital 8.79 The answer, as in the case of Poros and Tekmor, lies in an examination of the etymology of the word; the fact that this word is identical in the nominative case with the name of a well- known sea-goddess will then be seen to be quite irrelevant.

80 The clue was spotted by West in his first article, where he pointed out, in connection with the possible etymological associations of the actual name Thetis, that besides the suffix -tic; (gen. -tu6os) denoting a female agent (e.g. xaTaiBaTis, fern, of xaTaiftdxns) "there is another suffix -xis, gen. -tios, which generally developed to -cis. and was used extensively for forming abstract nouns." Unfortunately he was more interested in the rare examples where this second suffix also has an agent signification, being concerned to show that "Alkman

27.

could certainly have interpreted the name Thetis as 'she who sets1," But it is surely more natural to assume that if a form * Sens for the normal Seats existed in Alkman's dialect, then it would have the same signification that Se'ats has elsewhere; i.e. it would be the abstract noun derived fromxt'Snyt with the basic meaning of "action de poser" (Boisacq s.v. Seats). The suffix -a is j gen. -atos, employed to form abstract nouns in Greek comes from the I.-E. *-tis, which appears as -ti-h in Sanskrit and as -tio(n) in Latin; thus for the Greek otolo is we have in Sanskrit sthitih and in Latin statio, and for Greek Baals* Sanskrit gatih and Latin (in-)ventio.' Seats itself corresponds to the Sanskrit -dhitih (found in compounds) and the second element of the Latin con-ditio?^ Now it is true, as Buck points out (preek Dialects §61) f that the intervocalic -t- of I.-E. * -tis words shows an early change to -a- in all dialects; there are, however, two notable examples of the survival of the original suffix in the words cpdxts (the abstract noun from cpnyt, retained alongside its des- cendant <paai"s to denote a special sort of "action de parler" - cf. L.S.J. s.v.) and units (whose parent verb is lost in Greek - corresponds to Sanskrit matih "mesure, connaissance exacte"). Furthermore, Doric tends to retain inter-" vocalic -T- in other cases, such as the verbal endings -it and -vxt, the preposition Tcoxt, /txaxt for etxoat, (xpta)- xaxtQL etc. for -xdatot, and IIoTet6dv: it also retains initial t- in xu, xe'., xet. "Le dorien conserve en g^ndral t devant t," says Bourguet.^3 Hence it is by no means impossible that the -its form of g4 abstract nouns was still current in the Laconian dialect when Alkman was writing; so that when he wrote Se'xts eye'vexo or whatever, he meant neither more nor less than "a Seats took place. "85

Consequently we may now proceed to an interpretation based not on the activities of Thetis the Nereid but on the instances of the word Seats in early literature. Of these there are but seven, as the word is fairly rare before the fourth century philosophers; and they fall, broadly speaking, into two categories.

(1) Literal, "action de xtSe'vat". (A)

(a) "Creation, production". Alkaios fr. 133 Bgkv ' from Et. Mag. 319.30: "ESrixe anyat'vet 6i5o, xb ipoxaxe'Srixev f' eitot'naev ... acp

' oS xat Seats f' not'nats nap a 'AAxattj). (Cf. Alkaios fr. 204.6 Lobel-Page: J.atSeats)

(b) "Order, disposition", from xt'Snyt in the sense of "dispose, ordain" (L.S.J. s.v. A.VIII.l). Soph. Ichn. 275 [6 yev (sc. Hermes) axa]x[ds Y* gt] ' eaxl xou Ttaxpbs Seaet.

(2) Trans ferative, a result of the action of xtSe'vat.

(a) "Position, setting" of words in verse, Pi. 01. 3.8 cpopytyy^ T^ itotxtAoYapuv xat Boav/aoXwv hniuv xe Seatv ... (auyyet£at ) .

(b) "Place, position" of town or city. So Hippokrates Airs Waters Places 6 (11.24.13 Littre) avdtYwn xauxas xas noXtas (sc. xas itpbs xas 6v5atas xedyevas) Seatv xe'eaSat voaeptDxaxnv. This meaning also occurs at Thuc. 1.37.3 and 5.7.4.

(c) "Deposit" to be paid in respect of a law-suit, Ar. Clouds 1191 tv' at Se'aets ylyvolvto xrj vouynvta (this usage made possible by Snaetv xa itpuxaveta of line 1180).

28.

Before proceeding further , let us examine the implications of the Alkaios fragment quoted under (1) (a) . With respect to the first part of the entry in the Etymologicon Magnum, other late grammarians also assert that "among the ancients" Setvai signified noifjaai: so Asklepiades of Myrlea ap. Ath. 11.501c to noinaai SeTvai itpbs t&v apxawov eAeyexo and Helladios ap. Phot. 533 A. 18 Bekker xal yap xo'is TiaXatots Xe'yetv edos to e^nxev ercl too ti 6pav. One sense in which this may be said to be so is the apparent use of Tt'Snyt with noun + other noun or adjective to mean "render something so-and-so": e.g. Jl. 1.290 el 6e yiv atxvnTnv eSeoav deoi, Od. 5.135f. tov yev ... eq>aoxov 'hfaeiv a^dvaTov xai ayTfpwv. However, it does not seem to be this apparent synonymity**** that the grammarians are reporting? it has little to do with Helladios 's statement that eSnxe referred to "doing some- thing" nor with Alkaios 's use of dec is to mean lotnais. Rather these statements reflect the fact that TiSriyi in the early poets often seems to denote the action of what one versed in peripatetic terminology would call the efficient cause , or to tcolouv. 87 So e.g. Jl. 9.547 (Artemis) ayq>* aoT<|> Sfjxe rcoXuv xe'Xa6ov xal auTnv. But in fact it is a mistake to suppose that xrtmii in these passages means anything different from xidnvi elsewhere. "Set" or "place" is still the correct trans- lation, for we are still operating in the scheme of thought which is unable to distinguish spatial and non-spatial activity,88 and consequently causality is expressed in concrete spatial terms. Thus the ^ opening^ lines of the Iliad ynviv aei6e, ded, ITTiXriid6eu) 'AxiXfjos/ouXoyevnv, fi yupt* 'Axaiots

^ aXye* ednxe, which we

interpret as "the wrath ... which caused a host of trouble to the Achaians", are literally (and in a way which correctly reproduces Homer's way of thinking) rendered "the wrath ... which placed a host of troubles amongst the Achaians"; and so with the other passages.89 Thus Alkaios ' sweats will have been used in the basic sense of "action de poser", but in a context where a later writer, who possessed a notion of abstract causality, could interpret it as "action de causer", "creation", "production". It is in fact an "action de causer" in- volving space.

If we transpose this meaning of deais into a cosmogonical context, then we need look no further for an interpretation of Alkman's Se'xis. It is nothing less than the creation of the world expressed in terms of a spatial ordering. Corroboratory evidence for this assertion is supplied by fr. 20, which I quote once again:

Spas 6' £or'nc xpeis, ftepos xai xetya xinwpav Tpixav xal TSTpaTOV To/np, oxa adXXei yev, ea$uiv 6* a6av oux eaxt.

Lloyd-Jones has already suggested that this piece comes from the same poem as that with which our commentary is dealing,90 and this certainly seems very likely. Not only is the metre, as West points out, "similar to that suggested by some of our lemmata";91 but in this fragment we have an account of the creation of the seasons, which would fit very well into a cosmogonical ode* The striking thing about fr. 20, however, whether it comes from our^poem or not, is that the verb used to describe this creation is none other than eanxe ( =

e%nxe), the very word whose action is implied by d£xis. Nor is Alkman the only one to have employed xidnvi in a cosmogonical context. Herodotos, citing as his source the priests of Dodona, gives us the following account of a Pelasgian attempt at etymology: deous 6e upoawvdyotadv (sc. ot neXaayoi) a<peas onto too toio'5tou otl h6o'u$ SevTes Ta ndvTa rcpnyyaTa xal itdaas voyas elxov. 9^ These facts combine to suggest that the interpretation of Sexes in Alkman *s cosmogony offered above is the correct one.

29.

The final lemma of fr. 5, [ aydp] tc xal aeXava xai tpi'tov oxotos (ews) xas yapyapuyas, tells us some of the fruits of this cosmic Sexis, and if we append fr. 20, the seasons will become some more. "Day, moon, darkness" and pre- sumably others are mentioned before the word yap yapuyas appears; the cosmos is beginning to take shape.93 yapyctpoyas itself is a little difficult; "flashings" as produced by gold is its meaning in Bacchylides (3.17 Adyitei 6* uitb yapyctpoyaCs 6 xpuads) and elsewhere (e.g. Plato, Critias 116c); it is also used of "seeing stars" (Hipp. Prog. 24 , Plato Rep. 518a), while in the Odyssey (8.265) the expression yapyapuyou no6a)v corresponds to our "flashing feet". It is best taken here as a poetic representation of the stars. 9^

There remains one unsolved and possibly insoluble question raised by the above interpretation of dexis in Alkman: who performs it? The fact that a cosmic dexis took place implies the existence of a Se'xns of some sort; more- over, if fr. 20 does belong to this poem we need to find a subject for eanxe in line 1. It is possible that this figure is reflected in the xaxaaxeodcujv Ttavia who precedes Poros in the commentator's version of the cosmogony, but this gives us no idea as to his actual identity. Moreover, it seems unlikely, on the Hesiodic parallel, that Alkman would believe any god to have existed bet ore the separation of Heaven and Earth. The concept of deity eternal with respect to the past as well as to the future is not attested until Pherekydes of Syros, and probably only appears in this author as a result of the influence of Mil- esian philosophy .95 Either, therefore, the commentator has changed Alkman's order in an attempt to bring him into line with orthodox peripatetic doctrine, or else his xaxaaxeudcwv TEdtvxa is a pure invention (cf. supra p. 17). In any case, the words in which he is mentioned in the commentary are so vague that nothing can be made out of them with respect to the identity of Alkman 's xoayodeins. Consequently it F.ust be stressed at the outset that the following discussion is highly speculative.

There seem to me to be three possible candidates for the role. (i) Poros itself,, which may be said to "create" the world by causing the partition between heaven and earth; however, there is no evidence provided either in Alkman or in any parallel Greek instance to connect it with the later Seats, and furthermore the figure is said by the commentator to "pass on" - a state- ment which, as we have already seen (supra p. 22) , is likely to reflect Alkman 's own account of the event. (ii) Ouranos. The commentator does not mention this ur-sky-god, but his existence at the dawn of creation is implied by the coming-to-be of Poros to separate Heaven and Earth (Ouranos and Ge) . The concept of a sky-god who sets the world in order and then departs from the scene leaving the running of his creation to his successors is a universal one;96 it is also the case that Ouranos f s Vedic counterpart, Varuna, is as- cribed a demiurgic role.97 With respect to Greece, there is a possible parallel in Pherekydes1 Zas, whose demiurgic activity is described in the papyrus fragment (DK 7 B 2) as an adjunct to his marriage with Chthonie; in this passage Zas is obviously a personification of Heaven, as Chthonie is of Earth. 98 Also, Aristotle lists Ouranos as one of the "first figures" mentioned by the "ancient poets" - though here, as in the case of Okeanos whom he mentions in the same passage (cf . supra p. 25 and n. 68) , he is presumably referring to Ouranos as one of the primary pair in a cosmic genealogy. (iii) Tekmor. This seems to me the most likely candidate. We have already seen that Alkman fs Tekmor be- longs to the Moira-Aisa group; that is, it is a force primarily responsible for cosmic order. It is also the case that we know of no other Urwesen in Alkman fs Weltanschauung than Poros and this controlling force that appears as Tekmor in the cosmogony and Aisa in the Louvre Partheneion; to postulate a third entity, such as Ouranos or some other god, to fill the role of demiurge is dangerous in

30.

the absence of any kind of concrete evidence. It is a simple enough thought- process to view the force which controls the order in the universe as being also the originator of this order. We need feel no surprise if an abstract entity such as xe'xywp is made the subject of the concrete verb xiSnyt: a parallel for this has already been quoted , viz. the opening lines of the Iliad, where the subject of this same verb is a noun, y?Wis, which is nothing if not abstract. On this point it is also relevant to note that yotpa in Homer is often made the subject of concrete verbs, as in the following examples.

J2. 22.5 f. "Exxopa 6* auxou yelvat oAoin yotpa Tte6naev 'IX lou Ttporcapoide icuXawv xe Ixaidoov.

Od. 3.236ff. ou6e Seoi icp xal cpiXq) av6pi 6i5vavxai aXaAxeyev, oimoxe xev 6n ybtp ' oXofi xafteXTjai xavriXeyeos §avdxoio.

II. 24-.209f. xtf 6* (Ss no§i yoTpa xpaxain Ytyvoyevy enevnae Aivcj)

Even more significant with respect to what we have postulated as Alkman "s views on Tekmor's function in the universe are the following expressions of Parmenides :

8.29ff. xauxo*v t* ev xauxijj xe yevov xad1 eauxd xe xetxat Xouxtos eyite6ov a3^i y^vef xpaxepn Y^P 'Avdyxn netpaxos ev 6eayoXaiv ixei 9 xd yiv aytpls tepyei , ouvexev oux axeXe'5xnxov xb i6v deyts elvat.

8.36ff. ou6ev yhp^ri > eaxtv n eaxat aXXo ndpe^ xov edvxos, itiei xo ye MbZp enednaev oSXov axtvrixdv t' eyevai.

10.5ff. ei6rfo*£Ls 6e xal oupavov aycpls exovxa evdev e<pu xe xai £s ytv ayoua' eue6naev 'Avdyxn ne i pax1 exeiv aaxpwv.

Here we see the operation of Ananke and Moira in Parmenides1 world, the first two passages being from the way of Truth and the third one from the Way of Seeming. It is the second of these three quotations that forms the most precise para- lell to Alkman; for it is here stated that the present state of the world is due to a past act of Moira: "for there is not, nor shall be, anything else besides what is, since Moira fettered it to be entire and immovable."" So, I believe, with Alkman: the present world is an orderly place, a cosmos, pre- cisely because it was "set in order" by Tekmor/Aisa at the beginning 10^

of time - or, to use Alkman's own word, at the Se'xis of the universe. 10^

John L. Penwill Monash University

Notes :

1. Oxyrhynchus. Papyri XXIV, ed. E. Lobel. The papyrus with which we are here concerned is no. 2390 fr. 2, which appears on pp. 52-53; Lobel fs commentary (cited hereafter as "Lobel11) occupies pp. 54-55.

The following works form the relevant literature on the subject. (a) Editions. Besides the editio princeps of Lobel the fragment has been edited twice by D.L. Page,

31.

first in Poetae Melici Graeci (Oxford, 1962) pp. 22-24, and secondly in the O.C.T. Lyrica Graeca Selecta (Oxford, 1968) pp. 12-14. (b) Exegesis. (In the case of each work I give in brackets the method of citing it hereafter.) (i) D.L. Page's review of Ox. Pap. XXIV in C.R. n.s. 9 (1959) - the cosmogony is dealt with on pp. 20-21 ("Page"), (ii) W.S. Barrett's review of the same in Gnomon 33 (1961) - on the cosmogony see p. 689 ("Barrett"). (iii) CM. Bowra, Greek Lyric Poetry, 2nd ed. (Oxford, 1961), pp. 25-26 ("Bowra GLP") . (iv) H. Frankel, Dichtung und Philosophie des fruhen Griechentvms 9 2nd ed., (Munich, 1962), pp. 184 and 290-291 ( "Frankel11 ). (v) W. Burkert's review of Frankel 's book in Gnomon 35 (1963) pp. 827-828 ("Burkert") . (vi) M.L. West, "Three Presocratic cosmologies", C.Q. n.s. 13 (1963) pp. 154ff. ("West (1)"). (vii) M.L, West, "Alcman and Pythagoras", C.Q. n.s. 17 (1967) pp. Iff. ("West (2)"). (viii) H. Schwabl, art. "Weltschopfung", in P.-W. Realencyclopadie der classischen Altertumswissenschaft Suppl. IX (Stuttgart, 1962), p. 1467 ("Schwabl"). (ix) M. Treu, "Licht und Leuchtendes in der archaischen griechischen Poesie", Studium Generale 18.2 (1965) pp. 83ff. ("Treu"). (x) A. Garzya, "Idee cosmogoniche e morali in Alcmane", P. & I. 4 (1962) pp. 247ff. ("Garzya"). (xi) J.-P. Vernant, "Thetis et le poeme cosmogonique d'Alcman", Coll. Latomus no. 114 (1970), pp. 38ff. ("Vernant"). (xii) M.L. West, Early Greek Philosophy and the Orient (Oxford, 1971), pp. 206-208.

2. See F.D. Harvey, "Oxyrhynchus Papyrus 2390 and early Spartan history", J.H.S. 87 (1967) pp. 62ff.

3. Col. i. 25f. [ev 6]e xauxri xrj <j>6[ij *AX]xyav cpua [ioXo(yeT )] .

4. Mythical and protophilosophical accounts of cosmogony fall broadly speaking into two categories, depending on the .analogy employed. Primitive man knew only two methods of creation; one was by biological reproduction, the other by plastic art. Consequently when an account was sought for the origin of the world, an analogy was drawn with either one or the other of these creative methods. The commonest analogy used by the early Greek speculators was, as is well known, that of biological reproduction, so that an instance of the "craftsman" analogy in this early period would be especially interesting.

5. ii.21ff. -Mai xpi'xos axoxos* 6ia to yn6ena) yn'xe nXiov yn'xe ae[X]nvnv yeYovevou <*XX* ex i a6i,axpiT [o] v etvai xhv uXnv, while the next lemma reads (25f.) [Syap] xe xal aeXava xou Tpi'xov choxos. Page's comment in the apparatus to fr. 5 in Poetae Melici Graeci p. 24, "hariolari mihi visus hie commentator: scilicet xotL xptxos axdxos, quod poeta (25 seq.) prope 5yap et aeXdvav tertio loco posuit, hie a contextu evolsit, ad Porum Tecmorque quasi socium transtulit", gives the best explanation of this muddle. Cf. Harvey (above n. 2) p. 62: "Whether or not he (sc. the commentator) was an intelligent man is a question on which it is better not to dogmatise." On the fact that one lemma has xpi'xos and the other xpi'xov, Barrett comments "Presumably xpi'xos is right and xpi'xov the common modernisation of axdxos into a neuter noun." (609 n.5).

6. West (2) p. 4.

7. Cf. Theagenes of Rhegium ap. Schol. B ad II. 20.67 (= DK 8 2) ; Herakleitos Horn. All . 43. For the whole question of cosmogonies and cosmologies extrapolated from Homer by later writers see F. Buffiere, Les mythes d'Homere et la pensie grecque (Paris, 1956) 9 pp. 155-186.

8. H.A. 557a 1. Cf . however the following fragment of another commentary on Alkman which suggests that Aristotle said somewhere that the poet came from Lydia: Fr. 13.8 (Page) dXX* eoixe Au6bv au[xov voyi,']Cei<v 6 xe '

ApiaxoxeXns hcu, [ . . . . a'5] pc^ncpo l ontaxnSevxeg [ ] "avhp aypetos xxX" (fir. 16). But this gives no indication that Aristotle had actually read Alkman,

32.

9. Cf. Alkman frr. 148 (Exicko6es ), 149 (AiyiaXi's), 150 (' Avvi'xwpov) , and the following frr. down to 157, It may or may not be relevant to note that Hesiod was another poet fond of introducing such remote peoples into his poetry - v. frr. 150-153 Merkelbach- West.

10. Frr. 26, 39, 82. Cf. fr. 40, where he says, /^oi6a 6* opvt'xwv vo'yws icavxajv.

11. Viz. the Enetic, the Ibenian and the Kolaxaian (lines 51 and 59). The last two were of such obscurity that they defied certain interpretation even in Hellenistic times, as is well attested by the confusion in the scholiasts. See G. Devereux, "The Kolaxaian horse of Alkman's Partheneion", C.Q. n.s. 15 (1965) pp. 176ff., and "The Enetian horse of Alkman's Partheneion" , Hermes 94 (1966) pp. 129ff.

12. GLP 39. Cf. also G. Devereux's article on the Kolaxaian horse (above n. 11) p. 176: "Alkman fs Partheneion contains so many obscure details that even the elucidation of a single point seems worth while."

13. So Lloyd-Jones ap. West (1) p. 156. Cf. infra 28.

14. Page suggests as an alternative for 14-15 [6atydvwv] yepaixdxoi. 1 am not happy with either atwv or 6cuydvu)v, since it does not seem to me that Alkman would have considered either of these beings as divine. I would therefore leave itavxwv as absolute and insert a verb at the beginning of line 15 of which onc]e'6iXos aXxd would become the sub- ject. Xudn 6' would fit very nicely.

15. Following the translation of Fage p. 20.

16. In addition, the following authors are mentioned by the scholiasts on the Louvre Partheneion: Aristophanes, Aristarchos, Pamphilos, Sosiphanes, and Stasikles.

17. Syap is restored from the exegesis 5yap ou ((aXuJs aXXa auv nXiq>.

18. Cf. Barrett's comment quoted above n. 5.

19. <eo>S> suppl. Page. yctpyapuyots must be a word that appeared in Alkman's poem: (a) it is much too obscure to be a gloss; (b) it is difficult to imagine what it could possibly be intended to "explain"; and (c) it does occur elsewhere in early Greek poetry (Od. 8.265, H. Ap. 203, Bacchylides 3.17). Cf. infra 29.

20. Which shows that this poem, like the Partheneion , was written for public performance, presumably at some religious festival.

21. Frr. 27 (Mu>a* aye KaXXtdita $i5yaxep Aids), 28 (M&aa Aibs duyaxep Xiy' otei'aoyai wpaviacpl.) Cf. also fr. 3.1 (['OX] uyTua6es icepi ye cppevas.)

22. Aristarchos? Cf. Schol. Pi, Nem. 3. 16b: 6 yev 'Apiarapxos Oupavou duyaxepa xnv Mouaav 6e6exTai xadditep MoOaav 6e6exxaL xaSaitep Muyvepyos xat 'AXxyav CaxopoOaLV.

23. This is also the view of Frankel, who says that Alkman here makes the Muses daughters of Heaven and Earth "sodass sie dem Dichter eine mehr authentische Auskunft liber die ersten Begebnisse der Weltwerdung geben konnten" (291 n. 4). On the idea that there were in fact two generations of Muses, cf. Paus. 9.29.4: Muyvepyos 6e eXeyeua es rhv ydxnv nounaas xnv Eyupvauwv rcpbs F'5ynv xe xal Au6o'5s. (pnaiv ev x$ Tipootyu^ Suyaxe*pas Oupavov) xas apxatoxepas Mo'5aas, xov5xwv 6e aXXas veu)xe*pas eCvat Atbs uau6as. Alkman probably held a similar view to that of Mimnermos.

24. ex 6e xw it[dpaj xb x£xya>p« xb 6e xe] xywp eyevexo x [$ udptp axdXoudov Page (P.M.G. p.24j/

33.

ex 6e to) Tt[peayus Tldpos Texyoop xe» xe]xyoop West (1) p. 154.

25. Bowra GLP p. 26, Lobel p. 55, West (2) p. 3.

26. West's idea, propounded in his second article, that the original state of the world was "a rude mass that is anopov xai axexyapxov" (p. 2) - which in any case sounds more like Anaxagoras than Alkman - consequently sheds no light on the subject. Treu sees in the commentator's reference to f' uXn udvxoov an early expression of the Eleatic doctrine that "what is cannot come-to-be from what is not": "Schon um 650 v. Chr. hat sich demnach der Standpunkt der cpuaixoi durchgesetzt, dass alles Entstandene aus etwas e:>t- standen s*ein muss und demnach 'alles1 bereits im Urzustand prakonzipiert zu denken i.«t" (p. 86). This he supports by reference to Alkaios fr. 320 L.-P. xai x' ou6ev ex 6evbs ye'voixo. There is no evidence, however, to suggest that Alkaios was here enunciating a universal law of physics any more than Shakespeare was when Lear tells Cordelia "Nothing will come of nothing" (King Lear 1.1.90).

27. So Lobel p. 55, Page p. 20, Bowra GLP p. 26, West (2) p. 2, Treu p. 86.

28. So also Vernant p. 49: "Dans l'obscurite (axo'xos) du ciel et des eaux originellement confondus, il [sc. Poros] introduit des voies dif f e'rencie'es , rendant visibles sur la voute ce'leste et sur la mer les directions diverses de l'espace, orientant une e*tendue d'abord de'pourvue de tout trace' et point de repere, arcopov xai axe'xyapxov."

29. Elsewhere, and more interestingly, he gives "eine Brucke liber den Abgrund" (p. 184).

30. As West rightly points out ((1) p. 155), the appearance of Poros in Plato's Symposium (203b) as the father of Eros is entirely irrelevant as regards any attempt to interpret Poros in Alkman. By Plato's time the meaning "device" had become common, and it is of this that his Poros is a hypos tasis. One can see the ghost of Plato's Poros lurking behind the common interpretation of Alkman' s as "Device"; it is even more manifest in West's second article, when he declares that Alkman 's original world-mass was aitopov xai axexyapxov and therefore needed Poros and Tekmor to "bring it into shape" ((2) p. 2), for Plato says that the result of Poros being the father of Eros is that oux anopel "Epoos Ttoxe (Symp. 203e) . Vernant (pp. 44ff.) seeks to demonstrate that Plato is

drawing from an older myth which had Eros forming a connecting link between Poros, the

symbol of differentiation, and Penia, symbolic of undifferentiated uXn. The evidence he adduces is very late and cannot really be used to substantiate such a theory; in

any case, as will become apparent, it is not the key to interpreting Poros in Alkman.

31. West (2) pp. 3-6. So also Vernant, who speaks of "l'obscurite' totale qui regne dans la Nuit des eaux primordiales" (pp. 50f .)

32. West (2) p. 3, Burkert p. 828. Cf. infra p. 25.

33. "(This fragment) is particularly important as explicitly describing the separation of sky and earth as being passed on from mother to child, i.e. as a popular and traditional account." G.S. Kirk in Kirk-Raven, The Presocratic Philosophers (Cambridge, 1957), p. 33,

34. There can surely now be little doubt that Chaos in Hesiod means the gap between earth and heaven. See Cornford, From Religion to Philosophy (New York, 1957) pp. 66f., Principium Sapientiae (New York, 1965) pp. 194f . , and the most convincing argument of Kirk in Kirk-Raven op.cit. pp. 27ff . As regards Alkman, the interpretation offered here fits fairly well with what the commentator says at ii.7f. d)s yap np£otxo n v'r' xaxaaxeua [a$fjvai] eyevexo ndpos xts oiovel apxn: and if Alkman himself used words such as yopcpn yja oupavou xa' yns to commence his account of cosmogony (though it should be noted that Hesiod certainly felt no need to do so) , then it is presumably this that the commentator terms uXn.

34.

35. That is, polarities make their appearance, binary oppositions being a major feature of the world state. Cf . in this regard E.R. Leach, "Genesis as myth" in J. Middleton (ed.) Myth and Cosmos (New York, 1967) pp. 1-13, where the creation account in Genesis is analysed in terms of the coming-to-be of successive polarities (pp. 5f.)

With respect to the separation of sky and earth in Greek thought, something of the sort very likely lies behind the garbled account that remains of Anaximander f s cos- mogony, where the sphere of flame surrounding the air round the earth "like bark round a tree" is "broken off" and "shut off in certain circles" (DK 12 A 10) .

36. West (2) p. 8.

37. Reading [xpaxnae ya]p in line 13.

38. cnt]e6"rXqs seems to be the only possible restoration for the papyrus reading ]e6eiXos. The various interpretations that have been put upon it are discussed by West ((2) pp.7f.); of these, I prefer that of Lloyd-Jones (ap. Bowra GLP p. 42) that it means "lacking firm foundation", for the context must require it to carry some notion of weakness - it is the aXxa of defeated heroes that is being spoken of, after all. Inserting X'5$n 6' at the beginning of the line as I suggested in n. 14 would bring this out clearly; though clarity, as already observed, is something which does not appear to bother Alkman in this poem.

39. Cf. Pi. Pyth. 10.27 and Aesch. Prom. 894-6 cited by Garzya p. 249, nn. 10, 11.

40. Fr. 102 (XeTuxot 6' axapitos avnXeihs 6* avayxa) may well be another expression by Alkman of a similar pair of conceits*. In terms of myth, the appearance of the gap between earth and heaven also signals the formation of the gulf between men and gods; it is, in fact, another facet of the concept of man's "fall". That the gap is considered "out of bounds" to man is suggested also by Aesch. Prom. 280 (oa§epa $' ayvov ndpov olojvwv - aither is the Syvos Ttdpos which man cannot pass) and Ibykos fr. 28 (itOTaxai ev aAXoipuj) xaei - chaos is "another's province").

41. ii.l2ff. too [6e- 7td]pou icapeXdovtog eTtaxoXoo^n [acu] texyiop.

42. It is as well to note in passing that Poros and Tekmor do not form any type of pair. They are neither the primary cosmic parents, nor are they the primary polarity; in

fact, as will become apparent, they have practically nothing to do with one another.

43. West (1) pp. 155, 156.

44. West (2) p. 2.

45. Frankel p. 184 and p. 291 n. 2. This second remark is quoted with approval by Garzya p. 251.

46. Burkert p. 827.

47. In view of the fact that the Homeric xexyiop and the later xexyap normally have quite distinct meanings, it could perhaps be maintained that they are in fact two distinct words. However, this theory is ruled out by the two passages in Pindar where xexyap is used in the Homeric sense, and in the case of Pyth. 2.49 at least is required by the metre. In later poets, such as Apollonios, xexywp and xexyap are used indiscriminately as dictated by metrical considerations. (Cf. Frankel 's note on Ap. Rh. 1.499 [Noten zu den Argonautika des Apollonios , Munich, 196 8] : xexyap exouao hatte ich angezweifelt, weil in 3.1002 das Sternbild ein xe'xywp nicht 'hat' sondern 'ist', und weil es in dem [spaten] horn. Hymnos 32.13 vom Vollmond heisst xexywp 6e $poxois xoa afjya xe'xuxxai.

35.

Jedoch scheint in der neuen Alkmanfragment 5...Texyu>p in einem Sinn vorzukommen der hier zu dem (iberlieferten Text passen konnte." Texyap here does in fact come close to meaning yotpa, which is the sense I shall suggest it has in Alkman; it may well be significant that it occurs in Apollonios, too, in a cosmogonical context [the passage is quoted above, p. 18] - one, moreover, which embodies the motif of the separation of sky and earth.)

48. i.e. "possession of knowledge regarding divine dispensation is a mark of nobility", taking to as a demonstrative pronoun referring to the whole clause 5aa ... addvaToi, Texyap as being in apposition to nu$ea$ai, and 6etXwv xou eaSXftv as meaning "worthless and noble persons." Texyap here comes close to the meaning of "dividing line", and it may have been this passage that Aristotle had in mind when he wrote (Rhet. 1357b 7) to yap Texyap xal rcepas tcxutov zcti xaTa ttiv apxatav yXuvrTav.

49. The passages where Texaipoyai occurs are II. 7.70, 6.349; Od. 5.317, 11.112, 12.139; and H. Apoll. 285.

50. Alkman was in fact familiar with the Ionian epic tradition; cf. fr. 81 Ze') idTep, at yap i'ibs itoais ein which recalls Nausikaa's ai yap eyoi T0t,da6e ndais xexXnye'vos eiT) (Od. 6.244), and fr. 77 Av5aiapis, aivdrcapts, xaxov 'EXXd6i Burriaveipa the beginning of which looks very much like ari imitation of Homerfe Ai5anapi, el6os aptaTe (J2. 3.39, 13.769). See Bowra GLP pp. 20ff.

51. From Religion to Philosophy, ch. 1 (pp. 1-39).

52. Cf. the Sarpedon episode in .the Iliad (16.431ff.), where Zeus is unable to rescue his son from a fate TtdXai Tteupcoyevov (441); see also Od. 3.236ff. quoted below, p. 30.

53. xaTa to xpe^v: the whole fr. is quoted below, n. 56.

54. DK 22 B 94 aHXios yap o^X uiiepBriaeTai yeTpa? el 6e yn, 'EptviJes ytv Atxns enixoupoL e£eoprfaouat,v.

55. Dam 123 bis auvelvcu 6e a'JT(p (sc.Xpdvy) Tnv 'Avdyxnv, cp'5aiv oSaav ttiv auTnv xal 'A6pdaTeLav, daaiyaTwv 6 iwpyut wy evnv ev iravTi xcp x6o'u$9 twv icepaTwv auTou e(paitToyevnv: Ath. 18 p. 20 Schwartz Oupavos 6e Ft} yix^els yevvqt driXetas yev KXwda) AdxTiciv "Axponov. (DK 1 B 13)

56. For Chronos in Orphic belief see W.K.C. Guthrie, Orpheus and Greek Religion (New York, 1966), pp. 85ff. With respect to Pherekydes, cf. DK 7 B 1 Zas yev xal X pdvos fiaav ae' xal XSovi'n, on which Probus (ad Verg. Eel. 6.31 - DK 7 A 9) comments "Znva inquit xai X^c5va xau Kpdvov, ignem ac terram et tempus significans, et esse aethera qui regat, terram quae regatur, tempus in quo universa pars moderetur." Cf. also Herm. irr. 12, quoted below n. 98. For Anaximander cf. DK 12 B 1 *Ava^uyav6pos ... apxnv ... eJpnxe t&v ovtojv to aiceupov ... eB, Sv 6e n yeveaus eaTU tous o5au, xa', ttiv (pdopav eus Ta'3Ta yuveadat xaTa to xP£^v* 6u6dvau yap auTa 6txnv xa't tlolv aXXr^XoLS ttIs a6uxuas xaTa Tnv toO xpdvou Ta^tv.

57. So in the Sumerian account of the creation of man (S.N. Kramer, Sumerian Mythology, Philadelphia, 1944, pp. 59-62), after the clay has been fashioned into human shape, Enki instructs Nammu "0 my mother, decree its fate"; while later on in the same myth Enki and Ninmah engage in competition, Ninmah fashioning deformed human creatures and defying Enki to nominate a "fate" for them, which, however, Enki manages to do. In later Babylonian theology, the gods (Anunnaki) are referred to as "the great Anunnaki, who decree the fate" (J.B. Pritchard (ed.)# Ancient Near Eastern Texts, Prince ton, 1950, p. 114), while in Enuma elish the accession of Marduk to the supreme position in the pantheon is signified by his taking from Kingu "the tablets of Fate, not rightfully his (sc. Kingu's)" (4.121).

58. They also seem to come at a somewhat later stage in Epimenides1 theogony: cf. DK 3 B 19 ex tou (sc. Kpdvou) xaXXixoyos ye'veTO XPua^ *Aq>po6iTn/Mo£pat t* addvaTou xal *Epuvues aLoXd6ujpOL.

36.

59. According to Damaskios1 report (DK 9 B 1), Eros is the child of Erebos and Night, who are the first two beings to appear after Chaos. (Cf. also Plato Symp. 178b ■ DK 9 B 2.) One may also note the cosmogony given in Ar. Birds 693ff., where Eros is the first being to come from the wind-egg produced by Night.

60. Vernant, however, professes to find nothing surprising about it: "l'e'tonnant est qu'on ai pu s'etonner du role devolu a la fille de Ne*ree?"(p. 41). He associates Thetis with Metis, another sea-nymph who shows a great similarity to Thetis - see Cook, Zeus III.l p. 745. The name Metis appears in Orphic cosmogonies as one of the many names given to Phanes (Guthrie, above n. 56, p. 97); from this Vernant deduces that a sea-nymph could easily play an important role in cosmogony. Such reasoning does not appear to me very convincing; it relies on a number of tenuous associations, and surely in the case of Metis it is the etymological association of the word (= "Wisdom, Counsel") that is important with respect to identification with Phanes, not the fact that she was a sea-nymph.

61. Burkert p. 828.

62. West (2) p. 3.

63. i.e. auopov xai axexyapxov.

64. West (2) pp. 5-6.

65. Ibid. p. 6.

66. Ibid. p. 3.

67. II. 14.201 ' ftxeavdv xe Sefov yeveaiv xai ynxe'pa Tr^uv and 246 pe'edpa/' ftxeavoO, os nep

yeveois itdvxeaai. xexuxxai.

68. Okeanos is mentioned by Aristotle as one of the "first beings" of the "ancient poets" (Met. N4 1091b 4 ot 6e TtounTau ou apxotCou xauxrj oyouws, 5 BaauXeueLV xai apxeuv (paatv ou xous upcoxous otov Nuxxa xau Oupavbv f' Xdos n 'ftxeavdv, aXXa xbv Aua), probably basing this supposition on the passages cited in the previous note.

69. Cf. Aetius 1.7.11 = DK 11 A 23 8aXfjs vouv xoO x day ou xov $edv, xo 6Ne nav eyc|>uxov apa xai 6aiydvu)V TtXnpes* 6tnxeLV 6e xa'i 6ta xou axoixc1'^00^ uypoO 6uvayiv deiav xtvnxixViv auxou.

70. West (2) p. 6.

71. "Sofern die kosmogonische Deutung zutrifft, liegt das legitime Vergleichsmaterial in den altorientalischen Kosmogonien vor, die von Hesiod bis Thales, Anaximandros und dariiber hinaus griechische Spekulation beeinflusst haben, deren auftreten bei Alkman von Sardes gar nicht so liberraschen kann." (Burkert pp. 82 7f.)

72. On this question I agree with Page "that, on the available evidence, no judgement can be passed" (Alcman: The Partheneion, Oxford, 1951, p. 167). Such evidence as there is is reviewed in Appendix II of this work (pp. 167-170); and see also P. Janni, "II

problema dell'origine di Alcmane" in La Cultura di Sparta Arcaica I (Rome, 1965) pp. 96-120.

73. It is true that the evidence for Near Eastern influence on early Greek myth and phil- osophy is strong (see e.g. M.L. West, Early Greek Philosophy and the Orient, Oxford, 1971), but there is no parallel for a wholesale importation such as this. The two instances that might be cited as such are the mutilation/succession myth of Hesiod fs

Theogony which bears a marked similarity to the Hittite Kumarbi epic (Pritchard, op.

37.

cit.f pp. 120-121), and Thales1 choice of water as his first principle, which may be based on Egyptian cosmological beliefs. But with respect to the first, it is quite wrong to say that Hesiod's account is derived directly from the Hittite; as Kirk points out (Kirk-Raven, op.cit., p. 37), "there was a widely diffused common account, with many local variants, of which the Hittite tablet gives one version and Hesiod another.11 Both, in fact, display the Indo-European penchant for dividing things into threes (cf . C. Scott Littleton, The New Comparative Mythology, California, 1966, pp. 151 and 208-209), which accounts for the appearance in each of three sets of gods, while the son's mutiliation/murder/dethronement/castration of the father is a mytho- logical motif found all over the world. Hence the question here is one of common heritage rather than direct influence. With respect to Thales, it may indeed be as- serted that water owes its place in his cosmology to Egyptian influence (though the Okeanos passages of Homer cited above n. 67 could well have a more direct bearing in this case), but his final scheme, so far as we can piece it together, is far different from any he may have come across in the temples of Heliopolis.

74. Pi. Nem. 4.62ff.; Soph. frr. 150, 618 Jebb; Eur. fr. 1093 Nauck; PI. Rep, 318d; and in art Paus. 5.18.5 (chest of Kypselos); Baumeister, Denkmaler, 1797, 1799. Vernant finds this power of self-metamorphosis somehow significant: "Une des raisons pour lesquelles ces divinites marines etaient aptes a jouer, a l'origine du monde, ce role cosmogonique , c'est leur pouvoir de metamorphose. Elles contenaient en quelque sorte par avance a l'interieur d'elles-memes, les dissimulant puis les reVeUant tour a tour a la lumiere, toutes les formes susceptibles d'apparaltre dans le cours du devenir." (Vernant p. 42.) This seems somewhat far-fetched to me; furthermore, the evidence he adduces is late and refers to the Orphic Metis (cf. above n. 60), not Thetis.

75. Cf . if. Apoll. 319ff . The coupling of Thetis with Eurynome might appear significant to those who consider Alkman is talking about the Nereid (so e.g. Vernant pp. 41f.); for in the cosmogony sung by Orpheus in Book 1 of Apollonios1 Argonautica (495ff. = DK 1 B 16 - cf. above p. 21 ) , Ophion and Eurynome are stated to be "the first to exer- cise power over snow-clad Olympos" (503f.), being subsequently overthrown by Kronos. Thus Thetis may have a tenuous link with the Urgotter - but certainly not sufficient to elevate her to the status of cosmic demiurge.

76. II. 6.132ff.; Stesichoros fr. 67 Page.

77. II. 1.393ff.; Ion fr. 2 Page.

78. Melanippides fr. 9 Page; Pi. Isthm. 8.28ff.; etc.

79. I am not sure, in fact, whether it is really correct to give Poros and Tekmor capital letters, either, since there is no proof that Alkman actually thought of them as 6ouyoves (cf. above n. 14), Homer's yoipa is just as powerful as Alkman's T^xywp, but one does not usually spell it with a capital y. However, I have followed the normal procedure in Alkman fs case.

80. West (1) p. 155. So also Frankel p. 290 n. 2: "Vielleicht darf man vermuten, dass Alkman selbst nicht von Oexts , Gen. 6eTi6os, gesprochen hat, sondern von lakon. Sex is (oder aexis), mit dem Genitiv (der aber in Original text nicht vorgekommen ware) §ctios = attisch deais, im Sinne von 'Setzung, Ordnung1 (vgl, das haufige ednxe, lautlich = fecit, !machte!).fl

81. Taken from Boisacq s. w.

82. CD. Buck, Greek Dialects (Chicago, 1955) p. 223.

83. E. Bourguet, Le dialecte laconien (Paris, 1927) p. 146.

38.

84. It is true that we find tiois at fr. 1.36 and yadn'aios at fr. 125; but this does not invalidate the argument , as rare laconisms would tend to be "corrected" by copyists unfamiliar with the dialect into more usual forms, especially if it were only necessary to change one letter. Cf . Page, Paaetheneion, p. 103, and the citation from Lobel thereon. x before i is preserved in Alkman at fr. 38 oooou 6e itat6es ay&ov/evxj,', xbv xidapiaxav/ouveovxi and fr, 136 (nxi for net)- Page's conclusion about Alkman's dialect (Partheneion p. 163) is that it is "basically and preponderantly the Laconian vernacular" with, however, some features of Epic dialect occurring in places.

85. It is interesting, although possibly not relevant to the argument, to note that later writers who turn Homer into a cosmological allegory never make Thetis into a female Sexns, but always an allegorical reference to a cosmic S£ais« T^ie following passage of Eustathius is especially noteworthy (122.45, on II. 1.397ff.): n 6e aXXnyopia "Hpav yev xal evxaO%a xbv ae*pa voeT ... 8£xtv 6e xhv xoo itotvxbs Se'aiv, xauxbv 3y<pa> fiyouyevoi xnv xe Se*xiv xou xnv §eaiv, waitep xal (pdxiv xoU cpdaiv. Cf. 1150.2 (on II. 18.394); cf. also Schol. B ad II. 1.399, cited by Lobel p. 55 n. 1.

86. That it is only an apparent synonymity see n. 89 below.

87. Aristotle Phys, 194b 29 exi odev n apxh tfis yexafloXns n itpwxn n ttIs npeyn'aews, olov ... to TtotoOv xou noiouye'vou: and later in the peripatetic tradition cf. Aetius1 strictures on Anaximenes (1.3.4. - DK 13 B 2) ayapxdvei 6e xou o5xos it ontXoo xoi yovoei6ous aepos xai nve'5yaxos 6oxwv auveaxdvai xa c$oi' a6uvaxov yap apxnv yiav xnv uXnv x25v ovxojv UTtoaxfivat , aXXa xa' xb tioloOv afxiov XP^I uicoxid^vat* olov apyupos oox apxei icpbg xb exicwya yevea$oti , eav yn xb itoiouv ̂, icouxeaxtv 6 apyupoxdnos.

88. Cf . Cornford, From Religion to Philosophy (above n. 34) p. 140.

89. Cf. also Od. 9.235, 24.546, II. 3.321, 4.83, 21.524, etc. This applies even to those passages where xi'$nyi appears to mean "render"; again, it is spatial causality that is being expressed, and the proper translation will be something like "set x (in the world) as y".

90. Above n., 13.

91. West (1) p. 156. The metre of fr. 20 is iambic dimeter xaxa axi'xov. "If this conjec- ture is right, 5ydp xe xal aeXdva is a catalectic line. The occurrence of the word yapyapuyds or -Ss is a difficulty, but we might assume that Alcman scanned it yctpyap'3yas (cf. dyapOyn)... In the initial words of the poem, col. i.22f. [ae Ma3]aa Xuaaoyat ic[..«]u)v ydXtaxa^ Lobel's suggested supplement navx&v would have to be discarded in favour of an iambic word; for example, Barrett's Xuaqoyott x[e might be continued with aifov (monosyllabic, cf. fr. 56. 2. "(West loc. cit.) West's suggestions have been adopted by Page in his second edition of the papyrus (Lyrica Graeca Selecta p. 13), which is the reading I have followed when quoting the first lemma.

92. Hdt. 2.52. Cf. also Empedokles fr. 111.6ff: drfaeis 6fe£ oy$poio xeXouvoO xcu'piov auxyov/avdpwitois, Enacts 6'e xou e£ auxyoio depeiou/pe'5yaxa 6ev6peddpeitxa.

93. One assumes that earth and heaven are already in position as a result of the icopos which began the cosmogony; cf. the Orphic cosmogony given by Athenagoras above p. 21. Vernant (p. 39) cites col. ii. 24-25 eyevovxo oSv utco. [.]... itdpos xai xe'xyajp xa' axdx[os] to try and show that Poros, Tekmor and Darkness form some sort of primeval triad. But as this final lemma makes quite clear, axdxos is third not after Poros and Tekmor but after Day and Moon. Exdxos refers not to the darkness of pre-creation primordial Night but to the darkness of phenomenal night, which follows the day as the moon follows the sun (cf. ii.27f. Xyotp ou (|aXS)s aXXa auv nXup). It is obviously better to use Alkman's

39.

own words as evidence for the content of his verse than the oratio recta "exegesis11 of the commentator.

94. Of the other two lemmata, Tcpeayfus is most likely to be an epithet attached to Poros (so West (1) p. 154 etc.), while the first xca xpixos awd-ros takes a phrase from the final lemma and treats it out of context (cf. above n. 5).

95. "His (Pherekydes*) triad of primal forces presupposes the philosophical conception of the apxn, which he has merely fused with the genealogical principle." W. Jaeger, The Theology of the Early Greek Philosophers, tr. E.S. Robinson (Oxford, 1967) p. 68.

96. As e.g. in many African religions. "Thus, for instance, the Yorubas of West Africa believe in a sky god called Olorun (which means literally 'Owner of the Sky1), who having started to create the world handed over the finishing and the governing of it to a lower god, Obatalla." M. Eliade, Patterns in Comparative Religion, tr. R. Sheed (London, 1958) p. 47.

97. "Wise truly and great is his (Varuna's) own nature, /Who held asunder spacious Earth and Heaven. /He pressed the sky, the broad and lofty, upward, /Ay, spread the stars and spread the earth out broadly." Rig-Veda 7.86, as given in Cornford, From Religion to Philo- sophy (above n. 34) p. 67.

98. Cf. Herm. irr. 12 (DK 7 A 9) $epewu6ns yev apxots eTvai 'eyuv Zfjva xai XSovunv hou, Kpdvov Zfiva ]iev tov audepa, X^ovltiv 6e tViv ynv, Kpdvov 6e xov xpovov, 6 yev au^hp to tiouoOv, n 6e yfi T0 rccxcxov, 6 6k XP°*V°S ^v V ™ Y^vo*yeva«

99. Tr. J.E. Raven in Kirk-Raven op.cit. p. 277.

100. I would like to thank Dr. G.E.R. Lloyd of King's College, Cambridge, for advice and encouragement received during the preparation of this paper.