Go to Arachnion nr. 2 - contents or to Arachnion - home page

Written in the Stars:
Poetry and Philosophy in the Phaenomena of Aratus[*]

di RICHARD L. HUNTER (Cambridge)

Note from the Editors: for best results set your browser to «load the images»: here you have a lot of greek citations.


1. Introduction

The Phainomena of Aratus, probably composed in the period c. 280-260 B.C.[1], is an account in 1154 hexameters of, first, the fixed constellations and their conjunctions and, secondly, of weather-signs in both heavenly bodies and other natural phenomena[2]. It draws extensively upon two prose sources which we can reconstruct with some confidence. For the constellations Aratus was very heavily indebted to the prose Phainomena of the pioneering astronomer Eudoxus, written perhaps as much as a century before Aratus' poem; the debt was conclusively identified by the second-century B.C. astronomer Hipparchus, whose extant commentary (exégesis) on the works of Eudoxus and Aratus preserves many fragments of the former's treatise[3]. Although Hipparchus — whose rhetoric has, of course, its own very specific agenda of self-advertisement — alleges that Aratus' debt to Eudoxus was generally doubted before his own work[4], there is no reason to believe that Aratus intended to conceal the debt nor to doubt that many, at least, of Aratus' first readers fully understood the work's genesis, as indeed all that we know of the standard techniques of third-century poetry suggests they were expected to do. The second part of the poem, on weather-signs, is clearly indebted to a lost treatise of the fourth-century, perhaps by Theophrastus, which is known principally from a surviving prose version probably later in date than Aratus (= Theophrastus fr. VI Wimmer)[5]. Aratus' use of written sources, in a poem whose stylistic and literary techniques clearly announce it as an 'élite' text[6], raises general questions about 'didactic' poetry and the place of learning in the Hellenistic period, and to these I shall return. Two prominent features of the Phainomena must first be noted, as it is in the intersection of differing modes of presentation that the most characteristic, and perhaps most interesting, literary problems arise in this text; these will form the subject of this paper. What I shall not attempt here — but what is clearly a major desideratum — is what might be termed a 'modern Hipparchanism', that is, a detailed examination of how Aratus' account of the heavens exploits and/or misunderstands contemporary 'science'. Such an examination is vital, apart from anything else, if we are to see precisely how Aratus understood the rôle of the poet to differ from that of the technician, a subject to which I shall return in general, rather than specific, terms [7]. The present essay is concerned rather to sketch out a way of reading and understanding the Phainomena as very much a text of its time, and one which forces us to be clear about what we can and cannot know about ancient reading and writing practices. Aratus presents his poem as a re-writing of Hesiod's Works and Days[8]. The opening 'Hymn to Zeus' (vv. 1-18) is replete with Hesiodic phraseology[9], and the echoing and chiastic style of the first four verses not merely enacts their message — that the world is a kind of echo-chamber resounding with 'Zeus' — but is also an 'updating' of the famous opening of the Works and Days (Phain. 1-4):

Greek text
From Zeus let us begin. We men never leave him unspoken: full of Zeus are all the streets, all the meeting-places of men, full is the sea and the harbours; everywhere it is Zeus whom we all need.

(WD 1-10, trans. M.L. West):

Greek text
Muses from Pieria, who glorify by songs, come to me, tell of Zeus your father in your singing. Because of him mortal men are unmentioned and mentioned, spoken and unspoken of, according to great Zeus' will. For easily he makes strong, and easily he oppresses the strong, easily he diminishes the conspicuous one and magnifies the inconspicuous, and easily he makes the crooked straight and withers the proud — Zeus who thunders on high, who dwells in the highest mansions. O hearken as thou seest and hearest, and make judgement straight with righteousness, Lord; while I should like to tell Perses words of truth.

The bipartite structure of Aratus' poem is presumably indebted to Hesiod's 'double-headed' construction[10], and it has long been observed that the gliding transitions between subjects within the Phainomena, which avoid a systematic sectioning such as is familiar from Latin didactic, must be an attempt to reproduce the archaic manner. A remarkable similarity in the metrical technique of Hesiod and Aratus, involving a deviation by Aratus from the tendencies of his age, has even been noted[11]. More important for our present purposes is the relationship between the subject-matter of the two poems. There is, of course, much in the Works and Days about the use of nature and the movements of the stars to regulate one's life, both in farming and sailing, and it is to farmers and sailors, although not to them uniquely[12], that Aratus directs the didaxis of his poem ( vv. 7-9, 42, 758ff etc.)[13]. Moreover, the most familiar section of the Phainomena, the myth of Parthenos-Dike (vv. 96-136), is a close re-writing of Hesiod's 'Myth of Five Ages'; both passages serve to explain the situation in which man currently finds himself. As much as anything, it was the aetiological focus of the Works and Days which so commended it to Hellenistic and later poets.

Finally, and by way of transition to the second crucial feature of the Phainomena, we may note an obvious and pointed contrast between the ethos of the two poems. The Works and Days presents us with an all-powerful and all- seeing Zeus (cf., e.g., 267-9) who is concerned with justice, but whose mind (nóos) is changeable and hard-to-know (483-4), and who has hidden from men the means of a life free from toil (42 krýpsantes...)[14]. The themes of concealment and hiddenness are, of course, most prominent in the myths of Pandora and the Five Ages. The Zeus of the Phainomena, however, while also being all-seeing and concerned with justice, openly assists mankind through the omnipresence of 'signs' (Phain. 10- 13):

Greek text
Zeus himself set signs in heaven, marking out the constellations, and for the whole year he thought out which stars should most of all give men signs of the seasons, so that all things should grow without fail.

Much remains hidden and further 'progress' depends upon Zeus' benevolence (vv. 768-71, quoted below), but the situation is much more promising than that which Hesiod offered (Phain. 771-2):

Greek text
For Zeus openly brings aid to the race of mortals, appearing on every side, and everywhere revealing his signs.

The open visibility of the sky above us carries its own persuasive force. In a sense, no argument is needed to support Aratus' exposition: we must merely look around[15]. Zeus in fact actually 'speaks' (légein) to men [16]; the Phainomena itself is a 'sign' of Zeus' benevolence [17], — it comes in fact as a tekmérion ('evidentiary sign') from the Muses (v. 18) — whereas the Works and Days presents itself as the necessary product of hard times.

The importance of Hesiod for Aratus is, therefore, not in doubt[18]. The second determinative influence which has been identified in the Phainomena is early Stoicism. The extant Lives make Aratus a pupil of Zeno and a contemporary at the court of Antigonos Gonatas of Zeno's pupil, Persaios, and it is clear from the scholia that a Stoicising interpretation of the poem set in early [19]. An all-pervasive and beneficent Zeus, who can be identified with nature itself, and whose stars function as helpful signs for mankind seems clearly related to the Stoic cosmic principle. Here, however, modern interpretation must tread carefully. We can hardly speak of a firm body of 'Stoic dogma' at a date as early as that normally supposed for the Phainomena[20], and the danger of reading later theory back into the poem is thus a very real one. Nevertheless, the risk is worth taking, not merely because a Stoicising reading proves (I believe) a fruitful hermeneutic strategy, but also because it is to some extent unavoidable. A complex intertextual relationship, such as that which Aratus sets up with Hesiod's Works and Days, is always to some extent an act of appropriation, of making the earlier text 'speak' in certain ways. There are thus clear affinities between this widespread technique of Hellenistic poetry and the manner in which the Stoics used the evidence of archaic poetry, 'accommodating' it to, or seeing in it mythical foreshadowings of, Stoic theory[21]. Aratus' use of Hesiod does not, therefore, have to be specifically 'Stoic' in inspiration (although it is difficult to explain away all the circumstantial evidence); to 'read' Hesiod's Zeus, for example, as a foreshadowing of intellectual patterns familiar in Aratus' own day was a move of a type common in the poetic tradition — there are very clear examples in, say, Attic tragedy. A Stoicising reading, however, allows us to make the best use of the comparative texts still extant. It is a lucky chance that the 'Hymn to Zeus' of Cleanthes, who succeeded Zeno as head of the Stoic 'school', has survived to allow us to see how a nearly contemporary Stoic 'accommodated' the traditional language and forms of hymnic poetry to a new world- view[22]. Aratus' 'accommodation' is of a different type, but the subject-matter of his poem invites us to see it in the same tradition, and this is what I shall attempt in this paper. Finally, from the poetic point of view, Aratus' apparent 'combination' of Hesiod with Stoicism would be analogous to other standard techniques of mimesis; in particular, Hellenistic poets and their Roman successors constantly echo both an earlier passage of poetry and the (real or 'constructed') source of that earlier passage. Viewed from this perspective, Aratus 'reads' Hesiod not merely as a forerunner of the Stoics, but as the seed from which they grew.

2. Cosmic Poetry

For the Stoics, the universe (kósmos) is order (kósmos). The word re-echoes through Cleanthes' hymn[23]. The earliest poets too were believed to have written about the development of the universe, and 'cosmogony' was generally held to be the first subject of poetry[24]. It is, for example, the subject of the first song of Orpheus, traditionally the first poet[25], in the Argonautica of Apollonius, an epic poem probably composed not long after the Phainomena. Aratus' poem is not cosmogonical in the true sense, but it is certainly cosmological, and to this extent Aratus evokes the originary voice of the archaic theológos, 'speaker about the gods', while writing in a very new mode. Similar literary combinations abound in the poetry of the third century. Moreover, from a very early date, notions of kósmos were closely bound up with the idea of poetry and, particularly, truth in poetry.

In a famous passage of Odyssey 8, Odysseus praises the Phaeacian bard Demodocus:

'Demodocus, I admire you beyond any man; either it was the Muse who taught you, daughter of Zeus himself, or else it was Apollo. With what utter rightness (katà kósmon) you sing of the fortunes of the Achaeans — all they achieved and suffered and toiled over — as though you yourself were there or had talked with one who was! Come, change now to a later theme — the wooden horse and its fashioning (kósmon)... If you recount all this for me in the fashion it deserves (katà moîran), then I will tell the world forthwith how the god has blessed you ungrudgingly with the gift of inspired song'. (Od. 8. 487-98, trans. W. Shewring)

The notion of kósmos in this passage has been much discussed[26], but 'proper order' and 'sequence' are certainly of primary importance, as they are also in the phrase katà moîran, literally 'part for part'. This is even clearer in the following passage from the Homeric Hymn to Hermes in which the young god first sings for Apollo:

'He sang of the immortal gods and the dark earth, how they came to be in the first place and how each one was allotted his portion (moîra). First of the gods he honoured Mnemosyne in his song, the mother of the Muses, for she had received the son of Maia in her share. Then the glorious son of Zeus hymned the immortal gods according to their age and he told how each was born; he told everything in due order (katà kósmon) as he played the lyre in his arm'. (vv. 427-33)

Here, again in an originating cosmogony, the young god sings first of the division (moîra) of powers, and then of each god in turn, katà kósmon. Notions of partition and of sequential order are of course very closely related, and as the passage from Odyssey 8 suggests, it is in that proper sequential order that the guarantee of truth lies[27]. These guarantees are a result of the sense that the kósmos of the song reflects the kósmos of the 'real world'. This has been well described by George Walsh: 'kósmos ... an order of the world to which song must correspond as representation ... kósmos in the song and kósmos of the world should not differ. The song viewed as an articulation of parts stands for one viewed as a representation of serially ordered facts, for the true song must reflect the world's articulation with its own'[28]. In singing of the stars and of nature — for the Stoic a perfectly ordered kósmos — Aratus' poem is its own guarantee of truth. A poem 'about the kósmos' must be a poem katà kósmon. Moreover, in the Works and Days Aratus had an authorising model which also foregrounded order and sequence, as Hesiod sets out for us the passage of the year.

The centrality of kósmos in its manifold senses sheds particular light on two important passages of the Phainomena. The first is the acrostic passage discovered by J.-M. Jacques (Phain. 783-7)[29]:

Greek text
If the moon is thin and her light pure on the third day, there will be fine weather; if thin and her light very red, there will be wind; if, however, she is on the large side and her horns are dull and her light weak on the third and fourth nights, she is being dulled by the approach of the South Wind or of rain.

Some years after Jacques' article, W. Levitan [30] identified a further acrostic in the same area of the poem (vv. 803-6 pâsa) and, more tentatively, a third (partial) acrostic immediately afterwards (vv. 808-12 semee, suggesting sêma and cognate verbs); recently, Michael Haslam has refined Levitan's insight by noting mése, 'mid-way', split between the openings of vv. 807-8[31]. Whereas Jacques and others [32] connected the lepté acrostic solely with the famous leptótes of 'Callimachean' poetry, i.e. had seen it as imitation of Homer (cf. Iliad 24.1-5) and a programmatic marker of style and nothing more, Levitan rightly sought to make sense of it in terms of the central concerns of the poem. He noted that it followed very closely upon a passage which seems to invite us to look for such things (Phain. 768-72)[33] :

Greek text
For not yet does Zeus allow us to know all things, but much remains hidden; if he wishes, Zeus will grant us this too presently, for he openly brings aid to the race of mortals, appearing on every side, and everywhere revealing his signs.

Moreover, the successful searching out of acrostic patterns recreates the activity of the anonymous 'discoverer' of the constellations who perceived the usefulness of joining together those stars which would make meaningful figures (vv. 373-82)[34]. Just as this 'discoverer' revealed patterns which had always been there, and were 'put there by god', so a reader discovers meaningful 'signs' in the apparent randomness of the first letters of a succession of hexameters. The notion of kósmos now allows us to carry Levitan's important insight further. The words which are used to describe the heavenly bodies — leptós 'fine', katharós 'pure' (e.g. 383, 783) pach´ys 'fat' — seem to have been used as descriptions of poetic style in contemporary literary debate; this is most famously attested in the literary polemics of Callimachus[35]. By this device, poetry and its subject are seen to be symmetrical, illustrating the reciprocal kósmos which I have already discussed. Even if we wish to deny that Aratus' choice of language carries a programmatic charge in the context of contemporary poetry, — and both chronology and a dearth of other comparative evidence make the matter at best uncertain[36] —, the acrostic shows us how the pattern of the universe is reflected in the pattern of the poem. The stars are literally in the poem, and vice versa. Manilius too employs a related strategy in his Astronomica, a massive Stoic poem in which the regularity of heavenly movements is a central theme and in which there is an important, if shifting, relationship between the ordo visible in the skies and the ordo inscribed by the poet in his poem.

The second passage which I wish to discuss under this head is Aratus' explicit refusal to give an account of the planets. The passage forms a transition between the description of the individual fixed stars and that of the four celestial circles (Phain. 454-61):

Greek text
Mixed with them are five other stars, in no way like them as they whirl all through the twelve figures [of the Zodiac]. Not by looking at other stars could you mark the paths of these, since all move about. Long are the periods of their revolutions, and very far apart the signs of their conjunctions[37]. Of them I have no longer confidence: may I be competent to tell of the circles of the fixed stars and the signs in heaven.

In this praeteritio Aratus alludes to a notoriously difficult astronomical problem[38], the discussion of which would certainly not be in keeping with the style and level of the rest of the poem. There is, of course, no reason to see here a serious and rather embarrassing admission of astronomical incompetence[39]: the planets (excluding the sun and the moon) are not in fact relevant to an account of star- and weather- signs intended (at least notionally) to be of use to farmers and sailors[40]. This, however, does not explain why Aratus chose to call attention to his 'omission' in such a prominent and striking way. Part at least of the explanation, I suggest, lies in the notion of kósmos. Although the planets are, of course, as much a part of the universe as are the fixed stars — indeed they are much more influential 'signs' according to certain ancient views — Aratus stresses their 'uncertainty' in order to emphasise the fixed certainty of what he actually does describe. Put very loosely, the planets lack kósmos.

The passage on the planets is introduced by verses which stress eternal and regular recurrence in the face of moving time (Phain. 451-3):

Greek text
These stars you can see returning in orderly succession as the years pass, for all these images are very firmly fixed in the heaven through the moving night.

You cannot, however, use the other stars to find the planets because the latter are all metanástai, 'vagrants'. Not only can the poet and his readers not be 'confident' (v. 460) where to look to see the planets, but the same verse also suggests that they are not the kind of signs which inspire confidence that we can read them[41]. We may compare vv. 1142-4 where we are told that the congruence of two weather-signs brings 'hope' (elporé) and of three 'confidence' (thársos). It is, after all, god's benevolence that both gives us signs and allows us to read them and to act upon our reading. Confidence comes from the repeated pattern (kósmos) of successful 'sign-reading'; in this, the argument runs, the planets fail us. Such an interpretation anchors the passage firmly within the limited bounds of Aratus' 'didactic program'; poets, after all, only use an explicit praeteritio for subjects they do not wish to pass by. It is striking that in his 'Hymn to Zeus' Cleanthes too contrasts the fixed kósmos of nature with the mad changeablilty and rush of the kakoí, the 'bad men':

' [Reason, logos] is shunned and neglected by the bad among mortal men, the wretched, who ever yearn for the possession of goods yet neither see nor hear god's universal law, by obeying which they could lead a good life in partnership with intelligence. Instead, devoid of intelligence, they rush into this evil or that, some in their belligerent quest for fame, others with an unbridled bent for acquisition, others for leisure and the pleasurable acts of the body ... <But all that they achieve is evils,> despite travelling hither and thither in burning quest of the opposite. (Cleanthes, Hymn to Zeus 22-31, trans. Long & Sedley)

The kakoí are, of course, very different from Aratus' planets, but the two play structurally related roles in their respective poems; both carry the rejected weight of change and disorder. For an astronomer, the difficulty of explaining the movement of the planets was not a serious threat to the idea of an ordered universe, but within the rhetoric of his poem Aratus gives their 'quirkiness' a particular importance.

Finally, the passage on the planets also finds an instructive parallel in the archaic poem which we have seen to be central for the Phainomena, namely Hesiod's Works and Days. The catalogue of 'days' concludes with a brief glance at the days which have not been mentioned (WD 822-5, trans. M.L. West):

Greek text
These are the days that are of great benefit for men on earth. The rest are days of changeable omen, doomless, with nothing to offer. Different people commend different sorts of day, but few know that among those ones 'sometimes a day is a stepmother, sometimes a mother' .

Interpretation of these verses is disputed, but there seems to be a central contrast between days about which something certain can be said, i.e. the days of Hesiod's catalogue (vv. 768-81), and days which 'give an uncertain sound', which may turn out good or bad, and are therefore not suitable material for 'didactic poetry'. Hesiod's poem ends (WD 826-8) with an affirmation of the power of knowledge to overcome uncertainty, an uncertainty that is a central principle of men's lives (WD 483-4). That knowledge, and the power to offer it to others, is precisely what the poet claims for himself[42]. Aratus' poem carries this claim further by eliminating uncertainty not only from the poem, but also from the world itself.

3. Authority and Truth

How and if poets spoke 'the truth' is a central concern of both the explicit and the implicit poetics of the Greek world from the earliest days. The history of this concern has often been written, and need not be repeated here. Certain key periods in the history of this debate must, however, be identified, in order to allow Aratus' exploitation of this tradition properly to be appreciated.

The Muses who appeared to Hesiod famously told him (Theog. 26-8, trans. M.L. West),
Greek text
'Shepherds that camp in the wild, disgraces, merest bellies:
we know to tell many lies that sound like truth,
but we know to sing reality, when we will'

but more important here than the explicit declaration of the Hesiodic Muses is the way in which the poet presents himself in the Works and Days. The authority of what Hesiod says in that poem stems from the authority invested in the traditional poetic form which he employs[43], from the autobiographical mode of the poem — Hesiod knows these things, because he has experienced them — and from the Muses[44]. All three come together in a perhaps unexpected way when Hesiod turns to instructions about sailing:

'When you want to escape debt and joyless hunger by turning your blight- witted heart to trade, I will show you the measure of the resounding sea — quite without instruction as I am in seafaring or in ships; for as to ships, I have never yet sailed the broad sea, except to Euboea from Aulis, the way the Achaeans once came when they waited through the winter and gathered a great army from holy Greece against Troy of the fair women. There to the funeral games for warlike Amphidamas and to Chalcis I crossed, and many were the prizes announced and displayed by the sons of that valiant; where I may say that I was victorious in poetry and won a tripod with ring handles. That I dedicated to the Muses of Helicon, in the original place where they set me on the path of fine singing. That is all my experience of dowelled ships, but even so I will tell the design of Zeus the aegis-bearer, since the Muses have taught me to make song without limit.' (WD 646-62, trans. M.L. West)

Here is an explicit claim for the didactic role of the poet in areas where the poet has no personal expertise[45]. The idea, of course, is not far from the blind Demodocus' ability to tell well events at Troy, 'as though he himself had been there or had heard from another who had' (Od. 8.491), but the Hesiodic passage marks an important statement of program, particularly when viewed from the perspective of later ages. Plutarch indeed excised vv. 650-62 from Hesiod's poem as containing 'nothing of value' (oudèn chrestón) [46], and Horace amusingly carried the Hesiodic position to its 'logical' conclusion in the Ars Poetica, where he undertakes to teach poetry, though he himself cannot be a poet because he is sane and therefore lacks the principal prerequisite of the poet (vv. 301-8)[47]. Although Aratus does not emulate the Hesiodic importance attached to the autobiography of the poet as an authorising mode — perhaps because this would ill suit the Stoicising stress on the centrality of the fixed order of nature in which no individual is particularly important — it will become clear that the Hesiodic tradition is here, as everywhere, crucial for the Hellenistic poet. In fact, the gap which the Hesiodic verses opened allowed philosophers and 'experts' eventually to drive poets from the field.

Whereas Parmenides and Empedocles had continued in the Hesiodic mode of using hexameters to offer access to truth and the avoidance of deceit[48], the twin developments of prose writing and the idea of intellectual specialisation have always been seen as sounding the death knell for truly 'didactic' poetry. It was not, of course, that the use of metre alone was sufficient to condemn a text to the realm of 'fanciful poetry'; the substance of what was written remained crucial. Thus Plato (Theaetetus 152e) can make a distinction between hoi sophoí (Parmenides, Protagoras, Heraclitus, Empedocles) and hoi poietaí (Epicharmus, Homer), a distinction which clearly does not depend upon the use of metre, and Aristotle denied metre almost any significance in determining who was and was not a poet (Poetics Chapter 1, on Homer and Empedocles)[49]; for Aristotle, 'didactic poets' were either theológoi (e.g. Hesiod) or physiológoi (e.g Empedocles) or even (Met. N4, 1091b 8) 'partly theológoi', depending on the mode of discourse used. Nevertheless, the claim of 'metrically defined' poetry to convey 'useful information' came under attack from a number of directions, most famously from Plato's assault on the lack of knowledge of first the rhapsode and then the poet himself (Ion, Republic 10). It was the pleasure and sense of being charmed that we derive from poetry, as defined metrically, that made its lies so dangerous for Plato, and he ingeniously demanded that 'the friends of the poets' should defend its usefulness 'without the aid of metre' (Rep. 10.607d-e). It is both interesting and amusing to find Hipparchus precisely blaming the charm (cháris) of Aratus' verses for their general success in convincing people of their truth (1.1.7); this 'Platonic' tradition in the attack upon 'didactic poetry' is what Lucretius turns on its head when he uses the simile of the honeyed cup proudly to advertise the use of poetic charm for serious philosophic purposes (DRN 1.935-50 = 4.11-25). Manilius too gives the old charge a new twist in introducing the very technical third book of his Astronomica (Astr. 3.38-9) :

impendas animum; nec dulcia carmina quaeras:
ornari res ipsa negat contenta doceri.
Apply your mind to understand and seek not poetry that beguiles: my theme of itself precludes adornment, content but to be taught. (trans. G.P. Goold)

From roughly the same time as Plato's attack on the poets survives a series of prose 'handbooks', téchnai, which shed important light upon developing notions of written authority[50].

I wish here to focus upon two issues in these handbooks which are important for a consideration of the Phainomena, namely the problem of the author's knowledge, and the scope of a 'didactic' work.

Whereas the so-called Rhetoric to Alexander [51] and Xenophon's Hipparchikós (On the Cavalry Commander) begin in medias res and authority rests presumably both upon the nature of the work itself and the known position of the author, [52] Xenophon begins his Hippiké (On Horse-riding) with an appeal to his long experience of the art and his desire to pass on the fruits of this knowledge to his phíloi[53]. Here we can sense a desire to avoid the persona of 'professional specialist', a desire confirmed by the claim at the end of the work that it is directed towards the needs of the ordinary rider, the idiótes [54]. Xenophon is keen to set his work within the traditional frame of aristocratic philía, a frame which could be represented as threatened by newer, more 'professional', sources of knowledge. This analysis is, I think, confirmed by the closing sections of the same writer's Kynegetiké (On Hunting). At the end of Chapter 12 Xenophon adopts what may reasonably be called a Hesiodic voice:

'For among the ancients the companions of Cheiron to whom I referred learnt many noble lessons in their youth, beginning with hunting; from these lessons there sprang in them great virtue ( arete), for which they are admired even today. That all desire Virtue is obvious, but because they must toil if they are to gain her, the many fall away. For the achievement of her is hidden in obscurity, whereas the toils inseparable from her are manifest. ... But in the presence of Virtue men do many evil and ugly things, supposing that they are not regarded by her because they do not see her. Yet she is present everywhere because she is immortal, and she honours those who are good to her, but casts off the bad. Therefore, if men knew that she is watching them, they would be impatient to undergo the toils and the discipline by which she is hardly to be captured, and would achieve her' (On Hunting 12.18- 19, 21-2, trans. Marchant).

After this passage, which looks back to Hesiod's Areté, 'in front of which the immortal gods have placed sweat, and a long, steep path, rough at first, leads to it' (WD 289-91), but also seems to anticipate Aratus' watching Díke, the concluding chapter of the work is an attack upon sophistaí who, while professing to teach areté, write books on silly subjects full of 'empty pleasures' but no areté; if they teach anything, it is bad. Xenophon then goes on:

'I am a layman (idiótes), but I know (oîda) that the best thing is to be taught what is good by one's own nature, and the next best thing is to get it from those who really know something good instead of being taught by masters of the art of deception. I daresay that I do not express myself in the language of a sophist (sesophisménos); in fact, that is not my object: my object is rather to give utterance to wholesome thoughts that will meet the needs of readers well educated in virtue. ... For I wish my work not to seem useful, but to be so, that it may stand for all time unrefuted. The sophists talk to deceive and write for their own gain, and do no good to anyone' (On Hunting 13.4-5, 7-8, trans. Marchant, modified).

Xenophon solves the problem of how to present his 'technical' knowledge of hunting, a knowledge he is keen to display, without incurring the odium of being a 'professional', by distinguishing between kinds of art. He can be an 'expert' in a 'good art' and remain an idiótes who does not write sesophisménos. The use of the words has changed, but the Hesiodic inheritance is manifest. Both the care taken by Xenophon's rhetoric and the obvious vulnerability of his didactic position point to the importance of the issues involved. To anticipate somewhat, one of the most telling aspects of the Phainomena is the complete elision of this traditional didactic rhetoric of truth and deceit. Aratus imposes a total system upon us; deceit is not a possibility because what is described istà phainómena[55]. How different are both the task and the rhetoric of Lucretius who must argue from the seen to the unseen and appeal time and again to naturae species ratioque (DRN 1.148)[56]. The second issue of importance which the handbooks raise is that of completeness. The Rhetoric to Alexander begins by sub-dividing the topic to be discussed, a strategy which becomes standard in such works and is obviously designed to convey a sense of the completeness of the knowledge being offered[57]; the author then undertakes to discuss the sub-divisions one- by-one (kath'hèn hékaston). In On Horse-riding Xenophon notes that he will cover much the same ground as an existing treatise by one Simon, but he will also fill in all the gaps ('I shall attempt to illuminate all that he has omitted', 1.1), and in On Hunting, he undertakes to give a full account of each piece of equipment needed (2.2). In On the Cavalry Commander, however, he finds it necessary to deny both the wish for and the possibility of 'omnicomprehensiveness':

' [the cavalry commander] must always act upon what chance offers[58], and with his eyes on the current situation, work towards what is advantageous. To give a written account of all he must do is no more possible than to know all of the future ...' (On the Cavalry Commander 9.1-2).

Here, as before, we can see Xenophon caught between the conflicting demands of the old and the new. Elsewhere in the same treatise he follows general hypomnémata ('pieces of advice') with the full details (hékasta), quite in the manner of the rhetorical treatises[59]. So too in his Constitution of the Spartans, which begins 'unprofessionally' with a kind of chance intellectual curiosity ('Once upon a time it occurred to me ... and I fell to wondering ...')[60], order, sequence and completeness are overtly exhibited[61]. There is, then, a 'rhetoric of completeness' in the handbooks, which may be compared with Thucydides' claim (5.26.1) to give a fully detailed account of the Peloponnesian war, hos hékasta egéneto, a claim which is clearly of a piece with the presentation of his work as 'serious history' in comparison with the entertaining display-pieces of others (1.22). It should not now surprise us to find Hipparchus using precisely the same 'rhetoric of completeness' in the introduction to his account of the failings of Eudoxus and Aratus (1.1.9-11).

Aratus uses a form of praeteritio to explain the lack of comprehensiveness in his account of weather-signs (Phain. 1036-7):

Greek text
Why should I tell of all the signs available to men?

There are so many signs in nature that it would be impossible to tell them all, and in any case Aratus wants us to observe for ourselves and to find our own 'signs'. As so often, however, we may connect a feature of the Phainomena both with Aratus' philosophical project and with the tradition of 'didactic poetry', for lack of completeness has always presented a problem for anyone wishing to understand poetic didaxis in any simple way. Thus, for example, Malcolm Heath [62] has adduced the 'astonishingly lacunose' information in the Works and Days as an argument against seeing it as seriously 'intended to instruct'. This argument is, I believe, open to many objections[63], but the observation of poetry's 'incompleteness' does throw light upon the poetic didactic mode, if we connect the 'rhetoric of completeness' in Thucydides and the handbooks with Aristotle's distinction between history and (mimetic) poetry. History deals with individual details in a chronological sequence, ta kath' hékaston (Poetics 1451b 6-7). Mimetic poetry, however, deals with the general, tà kathólou, and is 'more philosophical than history', not merely because it tells of events linked by a causal nexus of necessity or probability, but also because we can extrapolate from, say, the fate of Oedipus to analogous possible occurrences in our own lives. Expanding upon Aristotle, we can note that the prominent gnomic element in poetry, which enables verses to be lifted out of their original context and applied to wholly new situations or taken to express general truths rather than specific, character-bound attitudes, encourages such extrapolation; this is indeed what lies at the base of ancient and modern practices of anthologising. Such transfer is in fact explicitly encouraged in Plutarch's treatise How to Study Poetry, where the Stoic Chrysippus is in fact cited as a precedent (and Hesiod is the poet concerned)[64]. The way in which we should approach the question of comprehensiveness in (non- mimetic) 'didactic poetry' is, I suggest, analogous to Aristotle's treatment of mimetic poetry. If a poet tells us how to make a plough, it would be foolish to believe that we can extrapolate from this to the detailed carpentry necessary for a wagon. On the other hand, the plough, the work involved in making it, and the moral conditions which make it necessary, can stand, pars pro toto, as exemplary of the total working conditions of the farmer. 'Didactic poetry' does not have to be comprehensive to be 'didactic'. It gives us examples, exemplary signs, from which we will be able to take our starting-point. The handbook, on the other hand, seeks to offer us a complete téchne; in using the handbook, we give up active participation in the acquisition (or confirmation) of knowledge and entrust ourselves to the guidance of an expert[65]. Aratus was not, as Eudoxus was, an expert astronomer. I have noted before that there is no good reason to think that Aratus ever sought to conceal his debt to the scientist. Indeed, in an age when not only the specialisation of knowledge had triumphed, but — and this was an important factor in that triumph —'knowledge' was now contained in books and catalogued in libraries, the use of expert written sources was the only way in which the poet could satisfactorily meet the 'How do you know?' challenge. An Eratosthenes could be both an expert and (in a small way) a poet, but that was exceptional. Aratus, on the other hand, was an 'expert' or 'professional' poet, and part of his expertise lay in knowing where to find things out; like Callimachus, he can still appeal to the Muses (vv. 16-18), but the Muses were now to be found in libraries[66]. Plato might not have been satisfied with Aratus' solution, but Hipparchus acknowledges that it might not be fair to attack Aratus for what were Eudoxus' mistakes (1.1.8). The 'versifying' of prose treatises is not inherently an idle game, but is at base a serious response to a crucial question of poetics. Aratus, of course, was not alone. Whereas Apollonius' Orpheus is given the characteristics of an archaic theológos, his blind prophet Phineus speaks the language of didactic as he gives an account of Pontic geography, and he naturally relies upon written, prose sources[67]. Callimachus went further in the Aitia and sometimes named his prose sources[68], a difference which is perhaps as much one of genre as of poetic judgement, but which throws a final light upon Aratus' procedure. The constellations were 'discovered' and named long ago by a nameless man of high intellect (vv. 373-82). In these purely informational terms, Aratus has nothing of his own to contribute, not because he knows nothing of astronomy, but because the truth of what he sings is self-evident, once someone has taught us where to look. This is Aratus' other response to the 'How do you know?' challenge: everyone knows — or could do, if they were willing to read god's signs.

4. The Stars Look Down

Stars are inescapable, and as such collect an extensive mythology. In many cultures they have been regarded as the eyes of god[69], and at an early date Greeks adopted catasterism as an explanation for the presence of some stars at least. A remark by a character in Aristophanes' Peace, 'Isn't it true what they say, that we become stars in the sky when we die?' (vv. 830-2) points to a popular folklore belief; perhaps it is what mothers told their young children. In any event, the idea that the psyche escaped to the upper air after death was a familiar one at least as early as the fifth century, [70] and the popular imagination presumably contrived many aetiologies for these objects of wonder[71]. The idea that the stars are 'heaven's eyes' is prominent in an epigram which may be roughly contemporary with Aratus, though it is ascribed to Plato[72]:

Greek text
'You look at the stars, my star. Would that I were the heaven, so that I could gaze at you with many eyes'.

At a quite different intellectual level, the Platonic Socrates sees the purpose of formal astronomy as the pursuit of 'truth', not (or not importantly) to provide weather information for 'farmers, sailors and generals'[73]. Thus, at an early date, the existence and regularity of movement of the heavenly bodies became a standard argument in philosophy for the existence of god or gods — we find it used later by the Stoics[74] — and it is to philosophical texts of the fourth century that we owe important elaborations of these ideas about the stars. Two Platonic texts are particularly relevant. In the Epinomis, a work emanating from 'Platonic circles', we find the notion of stars as living beings which distinguish the good and the bad among us and which report to the gods everything that happens on earth (985a-b)[75]. In the Timaeus, the fixed stars are eternal divine beings (40b), and to each was assigned by the demiurge a soul which rides upon it 'as on a chariot' (41e). After death the just man returns to his own star, whereas the bad man goes through successive degenerative metamorphoses until he finds the right path again (42b-d). Here, then, as in the proverbial 'eye of Justice' attested already in the fifth century[76], the existence of the stars acts as a protreptic to Justice, an idea which we shall find elaborated in Aratus. Finally, as the later culmination of these developments, we may note an account in Plutarch of the thoughts of Arion as he is carried to safety by the dolphins:

'At the same time, observing that the sky was dotted with stars, and the moon was rising bright and clear, while the sea everywhere was without a wave as if a path were being opened for their course, he bethought himself that the eye of Justice is not a single eye only, but through all these eyes of hers God watches in every direction the deeds that are done here and there both on land and on the sea'.(Moralia 161e-f, trans. F.C. Babbitt)

It is along the line that leads from Plato to Plutarch that Aratus is to be situated.

The most familiar reflection of these ideas in a non-technical context is the prologue of Plautus' Rudens, an adaptation of a play of Diphilus from (?) the late fourth century[77]. The prologue is delivered by Arcturus who explains the rôle of the stars to us (Rudens 6-21) :

noctu sum in caelo clarus atque inter deos,
inter mortalis ambulo interdius.
at alia signa de caelo ad terram accidunt:
qui est imperator diuom atque hominum Iuppiter,
is nos per gentis alios alia disparat
qui facta hominum moresque, pietatem et fidem
noscamus, ut quemque adiuuet opulentia.
qui falsas litis falsis testimoniis
petunt quique in iure abiurant pecuniam,
eorum referimus nomina exscripta ad Iouem;
cottidie ille scit quis hic quaerat malum:
qui hic litem apisci postulant peiiurio
mali, res falsas qui impetrant apud iudicem,
iterum ille eam rem iudicatam iudicat;
maiore multa multat quam litem auferunt.
bonos in aliis tabulis exscriptos habet.

An attempt to distinguish what is Greek and what Roman here is not germane to the present enquiry; moreover, the 'sinners' of vv. 13-14 suggest both the complaints of Hesiod about his brother and the 'bribe-devouring basileîs' and the world of the Roman law-court. What rather is crucial is the link, found already in Plato, between the stars and the maintenance of justice. In a famous passage of the Phainomena Aratus tells a story, a lógos, in which Dike is literally a star, and so it is worth going back to the Works and Days, the central Greek text on dike and the central model for Aratus, to see whether we can see here a further case where Aratus has run Hesiod and later thought together.

Hesiod's strongest argument for the practice of justice is the impossibility of escaping Zeus' eye:

'You too, my lords, attend to this justice-doing of yours. For close at hand among men there are immortals taking note of all those who afflict each other with crooked judgements, heedless of the gods' punishment. Thrice countless are they on the rich-pastured earth, Zeus' immortal watchers (phýlakes) of mortal men, who watch over judgements and wickedness, clothed in darkness, travelling about the land on every road. And there is that maiden Right (Díke), daughter of Zeus, esteemed and respected by the gods in Olympus; and whenever someone does her down with crooked abuse, at once she sits by Zeus her father, Kronos' son, and reports the men's unrighteous mind, so that the people may pay for the crimes of their lords who balefully divert justice from its course by pronouncing it crooked.' (WD 248-62, trans. M.L. West)

Díke is not one of the 'countless immortal guardians', but such an idea is at least not a very bold move from the Hesiodic text. Moreover, if Díke is to be made a star, as in Aratus' myth, then perhaps Aratus 'read' or 'constructed' Hesiod's 'countless immortal watchers clad in air' as the countless stars of heaven[78]. If this suggestion is correct, it would not mean that this is what Hesiod actually meant[79] , or that Aratus necessarily understood Hesiod in this way; rather Hesiod is read in such a way that his authority reinforces a later conception of Justice and the Stars. The older text is read as foreshadowing the later. Hesiod's Myth of Ages presents a five-stage progression (or regression) towards the present misery which will result in the abandonment of men to their fate by Aidós and Némesis. The five ages are structured, as Jean-Pierre Vernant has most fully demonstrated[80], by a reciprocal alternation between díke and h´ybris, and the long-noticed similarities between life in the Golden Age and the blessedness of the city in which men practise justice (vv. 112-19/225-37) make Hesiod's myth a protreptic to justice, like Plato's account of the stars in the Timaeus. Aratus' vision has taken over Hesiod's account[81], and with it the gradual retreat of Dike; in Aratus this retreat becomes a gradual physical removal of personified Dike from contact with men — first, in the silver age, to the mountains (vv. 115-28), and then finally, in the bronze age, to the stars. Like Hesiod's Aidós and Némesis, Dike abandoned men to their fates. Unlike her Hesiodic counterparts, however, we can still see Dike when we look up in the night-sky.

Aratus reduced Hesiod's five ages to three, not principally to avoid the potentially embarrassing 'juster and better' age of heroes which succeeded the Bronze Age, but in part to fit his scheme of a kósmos shaped and decided 'long ago'. The age of the Trojan and Theban wars, Hesiod's heroic age, was not of the necessary antiquity; like the myths of Plato, Aratus' myth of Dike must be set before 'recorded history'. Moreover, the catasterism of Dike also carries exemplary force for the creation of the kósmos.

We have already noted that in Plato's Timaeus the just man returns after death to the star which the demiurge had allotted to him; there he leads a 'blessed and congenial life'[82]. The catasterism of Dike is obviously parallel to this, and here again, I think, we may see a blending of Hesiod with later philosophy. In the Works and Days, the men of the Golden Age — with whom Aratus' Dike is so closely associated — after death become 'divine and revered spirits on the earth, good spirits, protectors from evil, watchers (phýlakes) over mortal men, givers of wealth' (WD 122-3, 126). They become daimones who 'guard, keep their eye on' mortal men; there is no reason to believe, as Wilamowitz did [83], that Hesiod identified these daimones with the 'countless immortal watchers/guardians' of WD 252-5, but the similarity of wording might suggest this easily enough, and indeed WD 254- 5 seem to have been interpolated back into the passage on the daímones. Just as, therefore, Aratus may have constructed Hesiod's 'countless guardians' as the stars, so I think the Dike myth shows us Aratus reading Hesiod's Golden Age as the origin of the stars. To modern scholars such a reading may seem perverse, and to avoid misunderstanding I should perhaps stress again that there is, of course, no question of seeking to establish what Aratus thought Hesiod 'really meant'; the earlier text is there to be exploited, not defended. Hellenistic didaxis is at base the interpretation of prior texts; as such, it is merely a special instance of the most prominent feature of the poetry of this period as a whole[84]. The purpose of a poet's 'interpretation' of a predecessor is only rarely to establish what that predecessor 'meant'.

What would a Stoic have made of Aratus' myth? We are told that Chrysippus held that 'men are changed into gods' and that the stars are gods[85], but at least one recent analysis has noted that this myth hardly seems a model of Stoic pronoia and has labelled it 'a foreign body in the otherwise optimistic Phainomena'[86]. What such an analysis misses is the kind of optimism which Aratus promulgates. It is an optimism based on the benevolence of the guiding cosmic principle, which hymnal style calls Zeus. This is a benevolence evidenced by the signs which god offers to man as a help, not by a particularly 'optimistic' view of man's current situation or of human morality. We should all do the best we can and use what god offers us, but without particular expectation (cf. Phain. 1101-3). Aratus' readers, ancient and modern, live in Hesiod's Fifth Age. If for the Stoic 'all human beings are, and inevitably remain, bad and unhappy'[87], then, when allowances have been made for the different meaning of moral terms, Hesiod and the Stoics to some extent come together, or — and this is crucial for Aratus — can be read as coming together.

Let us consider again the passage of Cleanthes' 'Hymn to Zeus' (vv. 21-31) which has already been considered in the context of Aratus' reference to the planets[88]. In a version of the familiar 'types of life' catalogue[89], Cleanthes condemns the 'disordered' pursuit of (political) reputation, (commercial) profit and (bodily) pleasures. Both Stoic doctrine and the traditional catalogue form suggest that 'the bad among mortal men' (v. 22) covers pretty much all of mankind — except for the occasional Stoic sage — and it is hardly fanciful to see here a debt to the language and ideas of Hesiod. The spoudè dyséristos 'belligerent quest', of v.26 evokes Hesiod's double éris (WD 11-26), but rewrites it:

' [The good éris] rouses even the shiftless one to work. For when someone whose work falls short looks towards another, towards a rich man who hastens (spéudei) to plough and plant and manage his household well, then neighbour vies with neighbour as he hastens (spéudont') to wealth: this strife is good for mortals. So potter is piqued with potter, joiner with joiner, beggar begrudges beggar, and singer singer. (WD 20-26, trans. M.L. West)

What for Hesiod is spoudé provoked by the good éris is for Cleanthes a vain attempt to escape the cosmic lógos. What both poets share, of course, is the condemnation of the 'mindless' pursuit of kérdos, profit (e.g. WD 323-4) and the failure to understand the workings of justice. Cleanthes, then, uses Hesiod's description of the current situation as a 'poetically valid' account of life on earth. This too is what Aratus expects his readers to do. We live in corrupt times, but nature works towards what is good and we must seek to discover that and to live in accordance with it. Knowing about the stars and weather-signs can only help us; neither stars, nor weather-signs, nor the myth of Dike, however, offer any kind of guarantee.

5. Poetry and Philosophy

Only the most technical science can avoid using the stars' 'inherently double aspect as both points of light and conventional figures'[90]. This is particularly the case with an author more concerned with the individual stars than with the overall movements and relationships in the heavens[91]. Even if the whole apparatus of aetiological catasterism is abandoned in favour of 'pure science', traces of the former are likely to linger in the less technical parts of any discussion. If we ask where Aratus drew distinctions within his poem between different kinds of material, then no absolutely clear answer will emerge, but certain suggestive patterns can be established.

The very first catasterism of the poem is that of the Bears (vv.26-44), and their 'myth' is introduced by the familiar qualification ei eteòn dé, 'if the story is true'. This phrase can, of course, mean many different things[92], and is here probably serving more than one purpose — marking, for example, an unusual version of the myth, a version which gives prominence to the benevolence of Zeus. In the context of Aratus' didactic poem, however, it is not simply that the phrase 'serves to enhance the objective tone proper to this kind of poetry[93], but rather that it stands here prominently to mark the first introduction of what cannot be seen, but must be narrated. This distinction lies at the very heart of the Phainomena, the 'truth' of which, as I have often stressed, is guaranteed by the evidence of our own eyes. We may find it helpful to label the story of the Bears as 'myth' rather than 'astronomy', but it is more important to see how Aratus uses this distinction in those parts of his poem where it is applicable; in the weather- signs, of course, there is almost no 'narrative' in this sense because the aetiological apparatus is no longer relevant.

Many of the most prominent 'myths' of the poem are indeed marked as such by qualifications such as 'there is a story' or 'men say' (cf. vv.98, 100, 163, 216, 637, 645). Even where this is not the case, there are often markers of some change. Thus the group of Cepheus, Cassiepeia, Andromeda, Pegasus and Perseus is introduced as the 'long-suffering family of Cepheus, of the race of Io' Kephêos mogeròn génos Iasídao (v. 179). Not only the affective adjective but also the stress on descent and the family group, with its suggestions of Attic Tragedy, point to material of a 'poetic', 'mythic' kind. Within this grouping further attention is called to the 'mythic' status of the account when, after a description of how to identify Cassiepeia, the poet adds 'you would say that she was mourning her child' (v. 196)[94]. The evocation of 'the myth' and the explicit refusal to countenance it as an explanation is here characteristic of a central tension in Aratus' poem. Another common device for moving from the visible to evocation of the non-visible is the use of ekeînos, 'the one you know about' to activate the reader's knowledge of 'myth' [95]. Thus, for example, Ariadne's crown is 'that crown you know about/have seen' (v.71), and Andromeda is 'that dread image' (v. 197). This small point of technique reveals Aratus' concern with distinguishing between the visible evidence of the stars and the inherited body of mythical knowledge which he could assume in his readers. There are, however, two important qualifications which prevent us from seeing the operation of very rigid distinctions within the poem. One is the fact that some 'mythical' material is introduced without apparent qualification: a case in point is the Lyre which Hermes set in heaven (vv. 268-71). More important perhaps is that throughout the poem the movements and appearance of the constellations are described in terms which appeal to their myths. Thus Ketos 'rushes' towards Andromeda (v. 354), the Hare is hunted (v. 384), the limbs of Andromeda are 'weary' (v. 704), and so forth; further illustration is unnecessary to make the point that the poem makes extensive use of the 'drama of the heavens'[96]. Nevertheless, the distinction which I have outlined does inform much of the shape of the poem, and in one place Aratus may comment directly upon it. When Aratus comes to the Pleiades he acknowledges that there is a difference between 'myth' and the evidence of our eyes (Phain. 257- 263)[97] :

Greek text
Men tell of the seven Pleiades, though only six are visible to our eyes. No star has disappeared from Zeus' sky without a trace in all the time of which we know, but the story is told. By name those seven stars are Alcyone, Merope, Celaeno, Electra, Sterope, Taygete and revered Maia.

The translation offered follows the standard interpretation of these difficult verses, which goes back to the scholia. According to this view, Aratus explicitly denies the truth of the story of the loss of the seventh Pleiad — a story which the scholia tell us he himself treated in a now lost poem[98] — by asserting again the fixed pattern of the kósmos established by Zeus; ek Diòs does not just mean 'from the sky'. If this is correct, then there will be a contrast drawn between 'the tales of men' and the better evidence needed to establish truth. Erren, however, removes this contrast by understanding the reference of v. 259 to be to the missing Pleiad: 'the star was not lost without a trace', that trace being precisely the tales of men[99]. A full discussion would be beyond the bounds of this paper[100], although it will be clear that the standard interpretation fits well with what we have seen to be a central concern of the poem. As with the passage on the planets, therefore[101], Aratus here probably uses an authorial comment on the subject-matter of his poem to foreground a crucial issue. The passage on the Pleiades both explains and explains away potential problems that arise throughout the description of the constellations: it is not always easy to match what you see to what you have been told.

This question of the distinctions Aratus draws within his material is one prominent instance of the phenomenon with which I have been concerned throughout this paper, namely Aratus' conscious manipulation of the multiple traditions which lie behind his poem. In a now outmoded form of criticism this might have been expressed as the opposition of 'the poetry' and 'the message'; the history of, say, Lucretian criticism is of course full of such things. Rather than this, I think, we should speak in terms of different modes of organising experience. One mode is that of the (didactic) poet, another is that of philosophy, which after Plato and Aristotle tends towards systematisation and completeness, thereby increasing the gap between the two modes[102]. As a final study of the didactic, poetic mode, I wish briefly to note certain aspects of Aratus' account of shipwreck, which offers an interesting case study in many of the things we have been considering.

The passage on shipwreck (vv. 408-35) stresses the role of 'ancient Night' in setting out signs warning of storms because of her pity for men at sea (vv.408-11, 419, 433-4). Night as the agent responsible for heavenly signs occurs elsewhere in the poem[103], but the prominence of this feature here cannot be adequately explained on a purely Stoicising reading[104]. She does act as an agent of the cosmic principle, but it is not as a manifestation of that principle that she weeps (v. 409). A close parallel is again from the prologue to the Rudens, where Plautus (and probably Diphilus) combines the idea of Arcturus — a storm-sign with which the shipwreck passage in Aratus is closely connected — as a servant of Jupiter with the traditional notion that stars not merely act as weather-signs, but actually cause the weather (Rudens 67-9) :

ego quoniam uideo uirginem asportarier,
tetuli et ei auxilium et lenoni exitium simul:
increpui hibernum et fluctus moui maritumos.

By acting 'off his own bat' in the interests of justice, Arcturus is like 'Night herself' (Phain. 419) who gives men signs. In considering Aratus' passage the parallel from Comedy points away from the systematisation of philosophy, but also perhaps helps us to see how the philosophical project informs the poem. Aratus' description of shipwreck makes no distinctions between the fates of 'just' and 'unjust' men, although shipwreck was a notorious instance in which the punishment of the unjust often involved the suffering of the innocent. For the Stoic, however, the future is determined, and popular conceptions of the moral status of those on board are neither here nor there; when viewed from this perspective, we see the planning in Aratus' selection of material.

The lucky men are those whose prayers are answered (Phain. 426-9):

Greek text
If by their prayers they happen upon Zeus' assistance, and he shows his lightning from the side of the North Wind[105], then after many sufferings they yet catch sight of each other on the boat.

Zeus is not a traditional saviour from shipwreck, a function that poetry normally gives to his sons, the Dioskouroi, and it is unclear whether it is to Zeus that the sailors are imagined to pray (which would again be unusual). Here also, then, the prominence of Zeus is a product of Stoicising 'monotheism' in which each part of nature works in harmony as part of a single organism. On the other hand, a simple Stoicising reading will not account for certain important details of the passage. What does paranissoménoio mean? In what sense could Zeus be said 'to pass near them' (Martin), '[come] to their aid' (Mair), 'gegenwärtig erscheint' (i.e. through his lightning, Erren). Moreover, týchosin, particularly next to the humorous phrasing of v. 425[106], might suggest a randomness of success hardly in keeping with Stoic determinism. Some help is gained from considering the 'focalisation' of the verses: 'Zeus' might be the word of the Stoicising poet, but 'passing by' or 'arriving' is rather the notion of an ordinary man thinking in anthropomorphic terms, and týchosin reflects the sailors' experience of the chanciness of the open sea. That the rhetoric of poetry is thus different from that of philosophy is confirmed by the 'pathetic fallacy' of Night's weeping, a fallacy enacted in language which describes heavenly phenomena in terms applicable to the struggling sailors themselves (kymaínonti néphei pepiesménon 416, thlíbet' 417); here nature moulds herself to man's plight. Night, the 'kindly' time, euphróne, acts in accordance with her name.

Analysis of this kind allows us to see that whereas systematic philosophy and science seek to close down options — to this extent they become paradoxically like 'history' in the Aristotelian scheme — the poetic, didactic mode offers multiple readings which draw on diverse traditions. To this extent 'didactic poetry' often shares some of the diversity which characterises less 'scientific', more popular belief-systems. Modern discussion of ancient didactic poetry has largely focused upon 'the author's intention' rather than on the reception of the poem by different readers, much as some ancient readers were bothered as to whether Aratus was a poet or an astronomer[107]. Obviously, the Phainomena will be read differently by a convinced Stoic and an un-philosophical reader; ancient theory which saw poetry as propaideutic to philosophy recognised that, as did the standard distinction in allegorising commentaries and technical handbooks between 'specialist' and 'non-specialist' audiences[108]. It is the choice of the didactic, poetic mode which precisely advertises the importance of the audience's role. Aratus' poem, unlike the work of Eudoxus or Hipparchus, is not merely about the universe, but is also universal in the sense that it presents itself as available to all, farmers, sailors, literary scholars. The poem continues in a new mode the age-old position of the poet as communal repository of wisdom. Aratus' project is to make us all see what we have always seen[109], to 'teach' us what we have always known. In the works of Eudoxus he found a systematisation of knowledge and experience, which was not only itself a sign of god's benevolence, but matched the ordered nature all around him; the poetic mode in which to express this cannot have been a difficult choice.

[*] This paper will subsequently appear in an Oxford University Press book on didactic poetry, edited by Don Fowler and Alessandro Schiesaro. I am grateful to Marco Fantuzzi, John Vallance and an audience at the Oxford Philological Society for instructive criticism of earlier versions of this paper. Due to the kindness of Professor D.A. Kidd, I was able (after the substantial completion of my own work) to see a draft of his forthcoming commentary on the Phainomena; I have added references to it where appropriate. The following works are referred to by author name only: M. Erren, Die Phainomena des Aratos von Soloi. Untersuchungen zum Sach- und Sinnverständnis («Hermes» Einzelschrift. 19), Wiesbaden 1967; N. Hopkinson, A Hellenistic Anthology, Cambridge 1988); A.A. Long & D.N. Sedley, The Hellenistic Philosophers, Cambridge 1987; W. Ludwig, Die Phainomena Arats als hellenistische Dichtung, «Hermes» 91, 1963, 425-48. I cite the text from the edition of J. Martin (Florence 1956), and refer occasionally also to that of M. Erren (Munich 1971).

[1]The (admittedly flimsy) basis for this dating is the patronage of Antigonos Gonatas which Aratus enjoyed. For chronological arguments based on the similarity of the opening verse of the Phainomena to Theocr. 17.1 cf. Gow ad loc. and M. Fantuzzi, «MD» 5, 1980, 163-72.

[2]The structure of the poem has been much discussed, but I will be concerned with this only tangentially; a major break after v. 732 was diagnosed very early in its reception, (a papyrus text of the first century A.D. places a coronis before v. 733). For surveys of this problem and the evidence cf. Erren 227- 33; Ludwig 429-39; id., «Gnomon» 43, 1971, 353; id., RE Suppl. 10. 30-1; J. Martin, Les Phénomènes d'Aratos. Étude sur la composition du poème, in L'astronomie dans l'antiquité classique, Paris 1979, 91-104.

[3]For a brief and helpful account of Hipparchus' work cf. J. Martin, Histoire du texte des Phénomènes d'Aratos, Paris 1956, 22-9. The standard edition is the Teubner of C. Manitius (Leipzig 1894).

[4]Cf. 1.2.1. These 'many' presumably include the 'many others' who, according to Hipparchus (1.1.3), wrote commentaries on Aratus' poem before him. The story in the Lives that Antigonos Gonatas 'told' Aratus to versify Eudoxus' work is presumably a post-Hipparchan fiction (cf. Knaack, RE 2.393), although the king's 'bon mot' eudoxóteron poieîs tòn Eúdoxon might just be a contemporary joke.

[5]For a discussion of the problems cf. O. Regenbogen, RE Suppl. 7. 1412-15. Aratus' combination of astronomy and weather- signs is unlikely to have been an original conception; a parallel prose text was published by C. Wessely, Bruchstücke einer antiken Schrift Über Wetterzeichen, «SWAW» 142.1, 1900, cf. O. Neugebauer, über griechische Wetterzeichen und Schattentafeln, «SÖAW» 240,2(1962).

[6]For Aratus' 'typically Hellenistic' use of Homer cf. the (rather limited) surveys by A. Ronconi, Arato interprete di Omero, «SIFC» 14, 1937, 167-202, 237-59, and A. Traina, Variazioni omeriche in Arato, «Maia» 8, 1956, 39-48.

[7]Cf. below.

[8]About any debt to the Astronomia or Astrologia ascribed in antiquity to Hesiod (frr. 288-293 MW) we can say nothing. It would, however, be very surprising if the Phainomena did not contain echoes of that poem. Callimachus' epigram on Aratus' poem (27 Pf. = 56 GP) may have both this poem and WD in mind; I wonder whether trópos in v. 1 evokes astronomical tropaí.

[9]The facts have often been documented, cf., e.g., H. Schwabl, Zur Mimesis bei Arat, in Antidosis. Festschrift für Walther Kraus zum 70. Geburtstag, Vienna/Cologne/Graz 1972, 336-56 and, more briefly, Hopkinson 138-40. The most striking echoes are: the repeated Día- diá-Diós (WD 2-4, Phain. 1-4), árretonPhain. 2/WD 3-4, Phain. 3- 4/WD 2-4, 101, Phain. 6/WD 20, Phain. 15/WD 822 (from the epilogue to Hesiod's poem); on protére genée (Phain. 16, WD 160) cf. below. If árreton in v. 2 plays on the poet's name, then we may compare the 'play' on Zeus' name at WD 2-3 and Hesíodon at Theog. 22. The 'pun' was, to my knowledge, first suggested in print by W. Levitan, «Glyph» 5, 1979, 68 n.18 and then (presumably independently) by D.A. Kidd, «CQ» 31, 1981, 353. Cf. further P. Bing, A pun on Aratus' name in Verse 2 of the 'Phainomena', «HSCP» 93, 1990, 281-5.

[10]Cf., e.g., J. Farrell, Vergil's 'Georgics' and the Traditions of Ancient Epic, New York/Oxford 1991, 163-4. West (p. 136 of his edition) notes that the title Works and Days is first attested in Lucian, but 'was no doubt established a good deal earlier'.

[11]Cf. H.A. Porter, Hesiod and Aratus, «TAPA» 77, 1946, 158-70.

[12]Cf. P. Bing, Aratus and his audiences, «MD» 31, 1993, 99-109.

[13]At one level there is here an obvious debt to traditional modes of self-presentation by poets, cf. the related trope at Call. H. 3.170-82, with the remarks of P. Bing, «ZPE» 54, 1984, 1-8. Observe the difference from the proem of Nicander's Theriaca in which the poet tells his addressee that, after instruction from Nicander, he will be respected by 'the ploughman, the oxherd, and the woodcutter' (three likely victims of snakebite). Nicander's three levels, instead of the Hesiodic and Aratean two, mark the 'professionalism' and specialisation of the knowledge which Nicander possesses; cf. his repeated rhetoric of access to privileged information (oîda, Ther. 805, 811, 818, 829) which stands in sharp contrast to Aratus. It is in details such as this that Nicander's true difference from the 'Hesiodic' tradition is to be seen; that the three levels are not always consistently maintained (cf. B. Effe, Dichtung und Lehre. Untersuchungen zur Typologie des antiken Lehrgedichts, Munich 1977, 58 n. 6) does not diminish the programmatic significance of the opening.

[14]The classic discussions of the theme of 'hiding' in Works and Days are those of J.-P. Vernant; cf., e.g., R.L. Gordon (ed.), Myth, Religion and Society, Cambridge 1981, 43-79.

[15]For the case where the evidence of our eyes and that of tradition conflicts cf. below.

[16]Cf. vv. 7-8 (in programmatic position), 732; weather-signs, as part of Zeus' system, also 'speak' (vv. 1048, 1071).

[17]Cf. below.

[18]The ancient Lives and the Suda, in fact, preserve traces of a scholarly dispute as to whether Aratus was more a zelotés of Homer or of Hesiod; the main thrust of the dispute will have been stylistic (cf. esp. Vita II, p.12 Martin). For what it is worth, Menecrates of Ephesus, a grammarian whom the Suda makes Aratus' teacher, seems to have written 'didactic' poems à la Hesiod (SH 542-50).

[19]Cf., e.g., pp. 40-1, 49-50 Martin. The most thoroughgoing, and important, modern Stoicising reading of the poem is that of Erren; I am much indebted to this work throughout.

[20]For a helpful orientation cf. the remarks of D. Sedley in M. Schofield et al. (eds.), Doubt and Dogmatism, Oxford 1980, 4-7.

[21]See A.A. Long, Stoic readings of Homer in R. Lamberton and J.J. Keaney (eds.), Homer's Ancient Readers, Princeton 1992, 41-66. Whether or not Stoic criticism was allegorising in the full sense is not crucial to the present argument; for much valuable information cf. G.W. Most, Cornutus and Stoic allegoresis: a preliminary report in ANRW II.36.3 (1989), 2014-65.

[22]For Cleanthes' hymn cf. Hopkinson 131-6; Long/Sedley 1. 326-7, 2. 326-7; A.W. James, The Zeus hymns of Cleanthes and Aratus, «Antichthon» 6, 1972, 28-38; K. Sier, Zum Zeushymnos des Kleanthes, in P. Steinmetz (ed.), Beiträge zur hellenistischen Literatur und ihrer Rezeption in Rom, Stuttgart 1990, 93-108; R. Glei, Der Zeushymnus des Kleanthes, in L. Hagemann and E. Pulsfort (eds.), "Ihr alle aber seid Brüder". Festschrift für A.Th. Khoury zum 60. Geburtstag, Würzburg 1990, 577-97.

[23]Note vv. 7, 19, 28.

[24]Cf. esp. P.R. Hardie, Virgil's Aeneid: Cosmos and Imperium (Oxford 1986) Chapter 1. On the history of the word kósmos cf. G.S. Kirk, Heraclitus: The Cosmic Fragments (Cambridge 1954) 311-12, and H. Diller, Der vorphilosophische Gebrauch von KOSMOS und KOSMEIN, in Festschrift Bruno Snell, Munich 1956, 47-60.

[25]Cf., e.g., Ar. Frogs 1032, Hor. AP 382. On Orpheus' song (Arg 1.496-511) cf. my The Argonautica of Apollonius. Literary Studies, Cambridge 1993, 148-50, 162-3 (with bibliography).

[26]I have found the first chapter of G.B. Walsh, The Varieties of Enchantment (Chapel Hill/London 1984) particularly helpful. Cf. also S. Goldhill, The Poet's Voice, Cambridge 1991, 57-9.

[27]I suspect that order and sequence are also relevant to the morphé of words at Od. 8.170 and 11.367 (where truth and falsehood are explicitly involved). I do not, of course, mean to suggest that katà kósmon and ou katà kósmon always involve notions of truth and untruth, cf. further A.W.H. Adkins, «CQ» 22, 1972, 12- 14, although Gorgias' assertion that 'truth' is the kósmos of lógos (Helen 1) relies, I think, upon a well-established, rather than a paradoxical, notion. Relevant also is the assertion of the absurd Ion, hos eû kekósmeka tòn Hómeron (Pl. Ion 530d), a claim which reverses the traditional language of poetic kósmos, and thus underlines Ion's stupidity.

[28]Op. cit. 8-9.

[29]Sur un acrostiche d'Aratos ('Phén.' 783-787), «REA» 62, 1960, 48-61. On acrostics in general cf. E. Vogt, Das Akrostichon in der griechischen Literatur, «A&A » 13, 1967, 80-95 and (with good bibliography) E. Courtney, Greek and Latin acrostichs, «Phil.» 134, 1990, 3- 13.

[30]Plexed artistry: Aratean acrostics, «Glyph» 5, 1979, 55-68.

[31]Hidden signs: Aratus Diosemeiai 46ff., Vergil 'Georgics' 1. 424 ff., «HSCP» 94, 1992, 199-204.

[32]Cf. Vogt art. cit. 83-7 and Courtney art. cit. 10-11. See now R. Scarcia, L'isopsefo di Arato, in R. Pretagostini (ed.), Tradizione e innovazione nella cultura greca da Omero all'età ellenistica. Scritti in onore di B. Gentili, Rome 1993, III 971-80.

[33]That the acrostic remained concealed until 1960 merely confirms Aratus' words. On this passage see the helpful exegesis of Erren 255-7. The verses leave room for progress in theoretical and empirical science, because although the 'signs' are already present in the kósmos, they do not yet 'signify' because men have not yet discovered the sign.

[34]On the problems of this passage cf. D.A. Kidd, The pattern of 'Phaenomena' 367-385, «Antichthon» 1, 1967, 12-15, and M.L.B. Pendergraft, On the nature of the constellations: Aratus, 'Ph.' 367-85, «Eranos» 88, 1990, 99-106. Pendergraft's interesting discussion rather overstates the 'inconsistency' between the namer's rôle and that of Zeus.

[35]For leptós cf. fr.1.24 (leptaléos); for katharós, H. 2.111; for pachýs fr. 398.

[36]The very 'untechnical' vagueness of these terms of approbation and disapproval is crucial to the teasing way in which Callimachus uses them, cf. my Apollonius of Rhodes, Argonautica Book III, Cambridge 1989, 37.

[37]On this difficult verse see Kidd's note ad loc.

[38]For Eudoxus' solution cf. Arist. Met. L 1073b 17-32 (= Eudoxus D6 Lasserre).

[39]Aratus seems in fact to have dealt with the planets in another work (probably a poem), the lost Kanon (= SH 90), cf. E. Maass, Aratea, Berlin 1892, 219-20, Kidd's note on v. 460.

[40]Cf., e.g., Ludwig 439-40, followed by Effe, Dichtung und Lehre (n. 14 above) 41 n.8. For a rather different emphasis cf. Erren 155-6.

[41]The unusual genitive after tharsaléos (cf. Erren 302) helps the double sense here. Kidd compares sophós kakôn at Aesch. Suppl. 453; there, however, áidris in the same verse is a crucial influence.

[42]Who better fits the prescription of the final three verses of WD than the poet himself? Notice the echo of these Hesiodic verses in Archestratus' 'didactic' culinary poem (SH 169.4-5).

[43]See the survey in West's edition pp. 3-25.

[44]For Hesiod's 'autobiography' cf. G.W. Most, Hesiod and the textualisation of personal temporality in G. Arrighetti and F. Montanari (eds.), La componente autobiografica nella poesia greca e latina, fra realtà e artificio letterario, Pisa 1991, 73-92; for Hesiod and the Muses in the context of later 'didactic' poetry cf. now A. Barchiesi, Il poeta e il principe. Ovidio e il discorso augusteo, Roma-Bari 1994, 171-5.

[45]R.M. Rosen has recently, and to my mind unconvincingly, interpreted these verses 'programmatically', in the context of Hesiod's creation of his own poetic space vis-à-vis Homer, cf. Poetry and sailing in Hesiod's 'Works and Days', «CA» 9, 1990, 99-113.

[46]Plutarch, fr. 84 Sandbach, from the scholion to the passage.

[47]AP 307-8, unde parentur opes, quid alat formetque poetam, / quid deceat, quid non, quo uirtus, quo ferat error, are distinctively 'didactic' in style and serve as a generic marker, cf., e.g., Verg. Georg. 1.1ff, R.D. Brown, «HSCP» 93, 1990, 315-21.

[48]Cf., e.g., Parmenides fr. 10, Empedocles frr. 1.9, 17.14, 23.9-11, 111, 114. For a brief (and incomplete) account of Aratus' use of Empedocles cf. A. Traglia, Reminiscenze empedoclee nei 'Fenomeni' di Arato , in Miscellanea di studi alessandrini in memoria di Augusto Rostagni, Turin 1963, 382-93.

[49]Cf. S. Halliwell, The 'Poetics' of Aristotle, London 1987, 71; 'Empedocles' verse writings [are cited] to represent the use of language for directly affirmative purposes — any use of language, that is, which purports to offer true statements or propositions about some aspect of reality.' Cf. below poetry's generalising power.

[50]The standard discussion is M. Fuhrmann, Das systematische Lehrbuch, Göttingen 1960, which is, however, not particularly helpful on the issues discussed here. On specifically rhetorical handbooks see T. Cole, The Origins of Rhetoric in Ancient Greece, Baltimore-London 1991.

[51]I exclude the spurious prefatory letter.

[52]This in fact becomes the standard form, cf., e.g., the second-century grammatical téchne of Dionysius Thrax; a related form is found in the fragmentary technical treatises of Theophrastus.

[53]Ovid uses precisely the same rhetoric in his didactic mode, cf. AA 1.29, usus opus mouet hoc: uati parete perito, 3.791-2. The identity of Xenophon's young philoi is disputed (cf. K. Widdra, Xenophon. Reitkunst, Berlin 1965, 72; Delebecque, Budé edition, pp. 10- 12), but it is the language of philía which is important for present purposes.

[54]This final sentence is omitted in one of our two best witnesses to the text, and was condemned by Wilamowitz («Hermes» 40, 1905, 146-7), but is defended (though for different reasons) by both Delebecque and Widdra. The considerations adduced here may also be thought to tell in its favour.

[55]Cf. the remarks of Erren 152-3.

[56]It will be clear that I cannot agree with the characterisation of Aratus' Phainomena in the opening chapter (devoted to Lucretius) of G.B. Conte, Generi e Lettori (Milan 1991 = Genres and Readers, Baltimore 1994). For Conte, Aratus' poem is 'un brillante gioco tecnico- artistico' (p. 19) in which the style and virtuosity of the text work against its contents; he contrasts the (undoubted) 'integralità di significazione' in Lucretius' poem. As the previous section of this paper has, I hope, demonstrated, Aratus has his own 'integralità di significazione', different though it is from Lucretius'. The virtuoso Hellenistic style is itself a manifestation of control and ordering, of kósmos. Conte's view of Aratus is, of course, the standard one; cf., e.g., Farrell ( above n. 11) 328, '[the Phainomena] makes its philosophical subject the paradoxical foil for its poetic qualities'.

[57]Cf. Aristotle's criticism of earlier writers of technai for their lack of completeness, Rhet. 1. 1354a 11-16. The 'apotheosis' of this strategy is to be found in the Rhetorica ad Herrenium which impresses its omnicomprehensiveness upon us throughout the work; cf. the excellent survey of E. Rawson, The introduction of logical organisation in Roman prose literature, «PBSR» 46, 1978, 12-34 (= Roman Culture and Society. Collected Papers, Oxford 1991, 324-51).

[58]Or 'come up with what happens to meet the situation', reading ennoeîn (Madvig) for poieîn.

[59]Cf. 1.9, 3.1.

[60]Contrast the Poroi which begins with an unapologetic announcement of systematic enquiry. Clearly, the Xenophontic corpus contains different kinds of work, which must be distinguished (by audience?).

[61]Cf. 1.3, 2.1, 5.1, 11.1, 12.7 (the fact that the Spartans overlook nothing in military matters accounts for the length of Xenophon's work, which also by implication 'overlooks nothing').

[62]Hesiod's didactic poetry, «CQ» 35, 1985, 245-63.

[63]Not least on the grounds of what is meant by 'instruction', cf. below.

[64]Mor. 34b. It would be rash to assume that Chrysippus' concerns here were the same as Plutarch's; was the philosopher attempting a joke?

[65]Here again (cf. above) we can see an important difference between Aratus and Nicander who promises completeness (Ther. 837), even if that promise is unfulfilled (cf. Effe op. cit. 61 n. 12).

[66]On the importance of books to third- century poetry in general cf. P. Bing, The Well-Read Muse, Göttingen 1988.

[67]On Phineus cf. my The Argonautica of Apollonius. Literary Studies, Cambridge 1993, 90-5. For the sources of this passage cf. U. Hoefer, Pontosvölker, Ephoros und Apollonios von Rhodos, «RhM» 5, 1904, 542-64; L. Pearson, Apollonios of Rhodes and the old geographers, «AJP» 59, 1938, 443-59; P. Desideri, Studi di storia eracleota, «SCO» 16, 1967, 366-416. In view of the considerations raised here, it is perhaps tempting to associate Phineus' disclaimer of 'comprehensiveness' (Arg. 2.311-16, 388-91) with the didactic mode; for the other considerations operative there cf. Hunter loc. cit.

[68]Frr. 75.54 (Xenomedes) and (?) 92 (Leandros). Fr. 75.55 and 76 show Calllimachus playing with the 'truth' of his source.

[69]Cf. R. Pettazzoni, The All-Knowing God, London 1956, Index s.v. stars; W. Deonna, Le Symbolisme de l'oeil, Paris 1965, 258-70.

[70]Cf. Euripides, Suppl. 532-6, Helen 1013-16; Alexis fr. 163 K-A; E. Rohde, Psyche, Tübingen 1898, II 384 n .2, 387 n. 1; R. Lattimore, Themes in Greek and Latin Epitaphs, Urbana 1962, 31-4.

[71]Note that Phain. 382 'no more does any star rise to our amazement (hypò tháumati)' is deliberately paradoxical: all stars rise 'to our amazement', but not in the specialised context which Aratus evokes.

[72]Anth. Pal. 7.669 = 'Plato', Epigram 1 Page.

[73]Rep. 7.527d-e.

[74]Cf. Cicero, ND 2.12-15 (= Long/Sedley 54 C [Cleanthes]).

[75]On this text cf. M.P. Nilsson, The origin of belief among the Greeks in the divinity of the heavenly bodies, «HTR» 33, 1940, 1-8.

[76]Cf. Soph. fr. 12 Radt; Livrea on Cercidas fr. 1.12. Although 'god's eye' is normally singular in Greek, Aeschylus offers four exceptions (Ag. 520-1, 776-8, Suppl. 812-13, Eum. 970-1 with Sommerstein's note).

[77]On the relevant verses see E. Fraenkel, The stars in the prologue of the Rudens, «CQ» 36, 1942, 10-14 [= Kleine Beiträge II 37-44].

[78]Relevant also may be Theogony 901-3 where the Horai, Eunomia, Dike and Eirene, érg' oréuousi katathnetoîsi brotoîsi. The verb is something of a mystery, but the ancients glossed it as phyláttein (cf. West ad loc.), and this might aid the idea of Dike as a 'guardian' or 'watcher'.

[79]As suggested by Pettazzoni op.cit.146.

[80]Cf. Myth and Thought among the Greeks, London 1983, 3-72.

[81]The introductory lógos ... állos (v.100) picks up héteros ... lógos which introduces Hesiod's myth (WD 106). My account differs considerably from that of Erren 37-9, who, if I understand his argument, sees Aratus' myth as an exemplary story to show how human 'Gottesbegriff' has changed.

[82]Timaeus 42b, cf. above.

[83]Hesiodos Erga, Berlin 1928, 70, 140.

[84]I have deliberately omitted the problem of the meaning of the invocation in v. 18 to Zeus autòs kaì protére geneé, both because I do not know what it means (although I have obvious sympathy with attempts to link it to the Golden Age), and because I think that my analysis of how Aratus has used the Hesiodic Myth of Ages stands without it. Recent contributions include: Erren 28-9; D.A. Kidd, «CQ» 31, 1981, 356-7, and (with a different view) his note ad loc. (both unconvincing, to my mind); B. Effe, Protére geneé — eine stoische Hesiod-Interpretation in Arats 'Phainomena', «RhM» 113, 1970, 167-82; G. Luck, Aratea, «AJP» 97, 1976, 213-34. Effe revives and refines Pasquali's view, expressed in Das Proömium des Arat, in Chárites Friedrich Leo zum sechzigsten Geburtstag dargebracht, Berlin 1911, 113-22, that the reference is to the primal wise men and inventors whom we know from later Stoicising theory (Seneca, Epistle 90 = Posidonius fr. 284 Kidd).

[85]Cf. SVF II 810-11, 813-15, 1076-7.

[86]E. Pöhlmann, Charakteristika des römischen Lehrgedichts, ANRW I.3 (1973) 813-901, at p. 883.

[87]F.H. Sandbach, The Stoics, London 1975, 44.

[88]Quoted above.

[89]See Nisbet and Hubbard's introduction to Hor. c. 1.1.

[90]G.O. Hutchinson, Hellenistic Poetry, Oxford 1988, 216-17. Hutchinson's discussion of Aratus is a sympathetic account of certain stylistic features of the work with which I am not concerned in this paper. Hutchinson is rightly much concerned with the confrontation of 'the astronomical' and 'the mythical', but he does not consider how Aratus suggests we distinguish the two. For Aratus' poetic imagination cf. also Pendergraft art. cit. 104-6 and R. Caldini Montanari, Illusione e realtà nel cielo dei poeti, «Prometheus» 19, 1993, 183-210.

[91]For this distinction cf. the helpful remarks of A. Stückelberger, Sterngloben und Sternkarten. Zur wissenschaftlichen Bedeutung des Leidener Aratus, «MH» 47, 1990, 70-81.

[92]Cf. in general T.C.W. Stinton, "Si credere dignum est": some expressions of disbelief in Euripides and others, «PCPS» 22, 1976, 60-89 [= Collected Papers 236-64].

[93]Stinton, Collected Papers 240.

[94]For other related uses of pháies ken in Hellenistic poetry cf. Hunter op. cit . p. 132-3.

[95]Cf. LSJ s.v. I 2.

[96]Good remarks on this aspect of the poem in Hutchinson op.cit.

[97]Hipparchus points out that Aratus is actually wrong about this: you can see seven Pleiades if you try hard enough (1.6.14).

[98]This was a consolatory poem 'To Theopropos' (SH 103), cf. E. Maass, Aratea, Berlin 1892, 233-4.

[99]'Immerhin ist der Stern nicht ohne Nachricht aus dem Haus des Zeus verlorengegangen', p.21 of his edition.

[100]Crucial are the meaning of geneêthen, which Erren implausibly understands of how the Pleiads arose, and the precise force of mal' háutos, which we might expect to be strongly intensive (cf. vv. 21, 180, 452).

[101]Cf. above.

[102]On poetry's rejection of systematisation cf. above.

[103]Cf. vv. 470, 695, 755.

[104]For a helpful account of such a reading cf. Erren 67.

[105]On this verse see Kidd ad loc.

[106]That verse looks to the shipwrecked Odysseus at Od. 5. 319.

[107]Cf., e.g., Scholium Q to vv. 96-7 (Martin p.124).

[108]Cf. already the distinction in the 'Derveni Papyrus' (Appendix to «ZPE» 47, 1982) between 'the many' and 'those who know rightly' (Col. XIX 1-2). Again we find this topos in the introduction to Hipparchus' commentary: 'I do not think any great intellectual effort is required to expound the meaning of the poem. For the poet is straightforward and concise, and can be clearly understood even by those with a moderate background in the subject. On the other hand, to understand what he has said about heavenly phenomena, to know which parts of his work agree and which disagree with the phenomena, this one might think is what is most useful and requires mathematical skill' (1.1.4).

[109]V. 733 'Do you not see? Whenever ...' draws attention to this concern. Such a question would normally refer to a one-time event (cf. Martin's parallels ad loc.), but here the verb evokes both physical sight and mental understanding. This 'oddity' pertains both to the problem of reading a text about seeing, and to the mental effort required to 'visualise' what Aratus is talking about, a visualisation which must of course be based on visual experiences and memories.

Go to Arachnion nr. 2 - contents or to Arachnion - home page

Last technical revision August, 31, 1995.

This document (http://www.cisi.unito.it/arachne/num2/hunter.html) is part of «Arachnion - A Journal of Ancient Literature and History on the Web» (http://www.cisi.unito.it/arachne/arachne.html). The editors are Maurizio Lana and Emanuele Narducci. The journal is distributed by the host of CISI - Università degli Studi di Torino, Via Sant'Ottavio 20, I-10124 Torino .
Quoting this document, please remember to mention the original paper edition, if any, and the electronic edition of Arachnion (in the form: Arachnion. A Journal of Ancient Literature and History on the Web, nr. 2 - http://www.cisi.unito.it/arachne/num2/hunter.html). If you'd like to access this document through a WWW page, please create a link to it, not to a local copy.