[In the following excerpt, Levitan identifies three hidden acrostics in the Phaenomena that, he concludes, suggest the concepts of “subtlety, totality, and signification” which inform the literary aesthetics of Aratus.]

ϕύσιs κρύπτεσθαι ϕιλεῖ

Heraclitus

Well, we have frankly enjoyed
more than anything these
secret workings of nature. …

Finnegans Wake

The aesthetic revolution that gave rise to the literature of the Hellenistic period (and ultimately to the literature of Rome) was occasioned at least in part by the growing awareness that the medium of literary expression had suffered a change in orientation from spoken to written language, and that the new emphasis on the visual aspect of literature had opened new possibilities and made new demands on both the writer and his audience. The most dramatic evidence of this re-orientation can be found in a number of figurative poems in the shape of wings, an egg, an altar, an axe, the so-called technopaegnia conveniently assembled at the end of Gow's edition of the Bucolici Graeci;¹ but its most subtle and revealing manifestation, I think, is in the acrostics of Aratus of Soli.

Despite the prestige and influence of his work in antiquity, Aratus (fl. 270 b.c.) has all but disappeared from the ranks of familiar classical authors. Even among classicists his most important poem, the Phaenomena, is not as much read as it is known by its traces—translations into Latin by Cicero and Germanicus Caesar among others, quotation and paraphrase by centuries of ancient writers, preeminently Vergil. Ovid's prophecy, then, is today better known than its subject: “cum sole et luna semper Aratus erit.”² Though disappointing, this is not altogether surprising. Hellenistic literature in general discourages a wide audience. Its elegant, allusive texture, dense and often difficult, does not easily lend itself to the attractive translation that is essential today for any but the smallest readership.³ And, when the poems themselves make explicit statements of aesthetic value, the recurring expressions of purity, delicacy, restriction, and refinement betoken a literature exclusive by design as well. In this Aratus is an author typically Hellenistic.

The Phaenomena (Starry Sphere or literally—and more appropriately—Appearances) is a didactic poem of less than twelve hundred lines on the subject of natural signs. Its greater part (lines 19-732) is an analytical description of the heavens that closely follows a prose treatise of one of the most famous early astronomers and mathematicians, Eudoxus, a pupil of Plato. Aratus's poem, however, is not a Platonic work but is thought to reflect Stoic influence. A second section, often considered separately as the Diosemiai or Weather-Signs, discusses the meteorological significance of atmospheric, terrestrial, and even biological events; the whole poem becoming thereby a kind of grand semiological compendium.

The Phaenomena was designed for manuscript and, more, conceived in terms of a matrix particularly appropriate to a papyrus roll: a rough rectangle of text, its horizontal dimension proceeding from an axis at the left-hand margin to the right for a length determined by the individual hexameter line, its vertical dimension proceeding from top to bottom for a length determined by the column of text on the roll—the prototype, simply, of the ordinary modern page. Such a matrix, of course, facilitates the introduction of acrostics but is not necessarily dominant in an acrostic poem: Gow includes in his collection, for example, one poem in the shape of an altar that is the product of two interfering matrices, one which uses a central axis to make the figure apparent, a second which yields an acrostic.⁴ In the case of the Phaenomena this rectangular matrix facilitated the introduction of three acrostics.

The first, rediscovered by J.-M. Jacques,⁵ occurs at lines 783-787:

ΛΕΠΤΗ μἐν καθαρή τε περὶ τρίτον ἠμαρ ἐοῦσα
Εὔδιόs κ'εἴη(.) λεπτὴ δἐ καὶ εὑ μάλ'ἐρευθὴs
Πνευματίη(.) παχίων δἐ καὶ ἀμβλείῃσι κεραίαιs
Τέτρατον ἐκ τριτάτοιο ϕόωs ἀμενηνὸυ ἔχουσα
Η νότo ἄμβλυνται e ὕδατοs ἐγγὺs ἐόντοs.

[(If the moon's crescent is) slender and clear on the third day, it means fair weather; if slender and with a red cast, wind; if thick and faint on the third and fourth day, then her horns are blunted by a south wind or rain.]

Jacques understands the acrostic here as a reminiscence of word play imagined in the first lines of Iliad 24 whose initial letters spell λευκή (“white”):

Λῦτο δ'ἀγών, λαοὶ δἐ θοὰs ἐπὶ νῆαs ἔκαστοι
Εσκίδναντ' ἰέναι(.) τοὶ μἐν δόρποιο μέδοντο
Uπνου τε γλυκεροῦ ταρπήμεναι(.) αὐτὰρ 'Αχιλλεὺs
Κλαῖε ϕίλου ἑτάρου μεμνημένοs, οὐδέ μιν ὕπνοs
Ηρει πανδαμάτωρ, ἀλλ' ἐστρέϕετ' ἔθα καὶ ἔνθα.

The Homeric acrostic must be fortuitous; Hellenistic scholars, however, themselves living comfortably with manuscripts and producing manuscript editions of earlier poets (Aratus is known to have edited the Odyssey), in many ways projected manuscript conventions onto Homer. But Jacques also points out that the key to Aratus's acrostic, the word λεπτή (“subtle, slender, slight”) is a term of value particularly important for Hellenistic aesthetics:⁶ here is both a bow to Homer and an assertion of modernity.

As far as I know, there is no Homeric inspiration for the other two acrostics. The passage containing both begins at line 802:

πάντη γὰρ καθαρἠ κε μάλ'εὔδια τεκμήραιο(.)
ΠΑNΤΑ δ'ἐρευθομένῃ δοκέειν ἀνέμοιο κελεύθουs(.)
Αλλοθι δ'ἄλλο μελαινομένῃ δοκέειν ὑετοῖο.
σήματα δ'οὐ μάλα πaσιν ἐπ'ἤμασι πάντα τέτυκται(.)
Αλλ'ὅσα μἠν τριτάτη τε τεταρταίῃ τε πελονται,
μέσϕα διχαιομένηs, διχάδοs γε μἠν ἄρχιs ἐπ'αὐτὴν
σΗΜΑΙNΕΙ διχόμηνον, ἀτὰρ πάλιν ἐκ διχομήνου
Εs διχάδα ϕθιμένην(.) ἔχεται δέ οἱ αὐτίκα τετρὰs
Μηνὸs ἀποιχομένου, τἠ δἐ τριτάτη ἐπιόντοs.
ΕΙ δέ κέ μιν περὶ πaσαν ἀλωαὶ κυκλώσωνται
Η τρεῖs ἠἠ δύω περικείμεναι ἠἠ μί'οἴη. …

[When the moon is entirely clear, forecast fair weather; when red over all, wind; when spotted, rain. The signs are not all for each separate day: but the signs of the third and fourth day show the weather up to the half moon; those of the half moon up to midmonth; those of midmonth up to the waning half moon; the signs of the fourth-to-last day follow those of the waning half moon, followed in turn by those of the third day of the new month. If halos encircle the whole moon, triple or double or even one set alone. …]

The acrostic beginning at 803 departs somewhat from the earlier model. Instead of being repeated in identical form down the left-hand margin, the key word πάντα (“all”) is reflected by its grammatical variants, first by πas the masculine form of the adjective and then, as the initial alpha of the next line is considered, by its feminine form πaσα; the adverb πάντη beginning line 802 introduces and, with πaσσιν and πάντα in 805, augments the sophisticated pattern. Using one of the poem's most important words with unabashed precision—ringing the changes on “all”—this complex rhetorical figure, polyptoton in two dimensions, expresses in grammatical terms one of the poem's important concepts, plenitude and perfection. Compare, for example, the poem's opening lines:

'Εκ Διὸs ἀρχώμεσθα, τὸν οὐδέποτ' ἄνδρεs ἐῶμεν
ἄρρητον(.) μεσταὶ δἐ Διὸs πaσαι μἐν ἀγυιαί,
πaσαι δ'ἀνθρώπων ἀγοραί, μεστὴ δἐθάλασσα
καὶ λιμένεs(.) πάντη δἐ Διὸs κεχρήμεσθα πάντεs.

[Let us begin with Zeus whom we men never leave unnamed: all the streets are filled with Zeus, and all the market places of men; the sea is filled and the harbors. In all ways do we all have need of Zeus.]

The third acrostic also plays on an important concept, perhaps the most crucial idea of the poem, the idea of signification; and also departs from the simpler acrostic model, here more dramatically. The key word σημαίνει (“to show by a sign, to mean”) in line 808, suffers a change in the acrostic from a verb form to the plural noun σημεῖα (“tokens, signs, constellations”); and the acrostic is incomplete, or if complete, misspelled, teetering on the verge of nonsense. At this point in the poem, at least, the reader cannot remain passive: the signs are there to be deciphered.

I realize that such a brief account alone is insufficient to dispel the uncomfortable feelings the notion of a misspelled acrostic can excite; and further discussion is certainly called for.

Aratus's may not be the only distorted acrostic in Hellenistic literature. Some years ago, E. Lobel published a piece that should become a model for classical articles: it is three sentences long, the first two being introductory, and is worth quoting in full.

It is unusual for Greek poets to sign their work. Nicander ends the Theriaca καί κεν ‘Ομηρείοιο καὶ εἰσέτι Nικάνδροιο μνῆστιν ἔχοιs, and the Alexipharmaca καί κ'ἔνθ' ὑμνοπόλοιο καὶ εἰσέτι Nικάνδροιο μνῆστιν ἔχοιs. In addition, he embeds his name in the body of his verses; in the Alexipharmaca at 11. 266-274:

σ ὺν δἐ καὶ ἀμπελόειs ἔλικαs ἔνθρύπτεο νύμϕαιs
Ι σωs καὶ βατόεντα περὶ πτορθεῖα κολούσαs.
Κ αί τε σὺ γυμνώσειαs εὐτρεϕέοs νέα τέρϕη
κ αστηνοῦ, καρύοιο λαχυϕλοίοιο κάλυμμα,
N είαιραν τόθι σάρκα περὶ σκύλοs αὐον ὀπάζει
Δ υσλεπέοs καρύοιο, τὸ Καστανὶs ἔτρεϕεν αἔα.
Ρ εῖα δἐ νάρθηκοs νεάτην ἔξαίνυσο νηὸὺν,
Ο s τε Προμηθείοιο κλοπὴν ἀνεδέξατο ϕώρηs.
σ ὺν δἐ καὶ ἔρπύλλοιο ϕιλοζώοιο πέτηλα,

though one can imagine the heat with which this conclusion would have been contested, if at Theriaca 345-353,

N ειμάμενοs κασίεσσιν ἐκὰs περικυδέαs ἀρχὰs
Ι δμοσύνηι νεότητα γέραs πόρεν ἡμερίοισι
Κ υδαίνων(.) δὴ γάρ ῥα πυρὸs ληίστορ' ἐνιπτον
Α ϕρονεs(.) οὐ γὰρ τῆs γε κακοϕραδίηs ἀπόνηντο.
N ωθεῖs γὰρ κάμνοντεs ἀμορβεύοντο λεπάργωι
Δ ῶρα(.) πολύσκαρθμοs δἐ κεκαυμένοs αὐχενα δίψηι
Ρ ώετο(.) γωλειοῖσι δ'ἰδoν ὁλκήρεα θῆρα
Ο ὐλοὸν ἐλλιτάνευε κακῆι ἐπαλαλκέμεν ἄτηι
σ αίνων(.) αὐτὰρ ὃ βρῖθοs, ὃ δὴ ῥ'ἀνεδέξατο νώτοιs,

he had not had better luck.⁷

Surely Lobel is not serious in ascribing the distorted acrostic in the Alexipharmaca to bad luck: acrostics, easy enough to produce in English, are easier still in an inflected language like Greek whose word order is much more fluid. But if it is the result of design, as I believe it is, it is an exceptionally curious design. There have been attempts to cushion the shock. An easy emendation has been proposed, νῦν δἠ καὶ in the first line,⁸ that would restore the initial letter of Nικανδροs and make the acrostic easier to recognize but without completely addressing the problem in the passage. More helpful is the suggestion of E. L. Brown⁹ that the missing alpha in Nicander's name has become redundant, following as it does the initial kappa in the third line of the acrostic—καί—and so insistently in the fourth—καστηνοῦ, καρύοιο, κάλυμμα. Aratus also exploits the initial syllable of a word for an acrostic (in line 811) and Brown points out instances of this procedure in other poems.¹⁰ This suggestion saves Nicander from the charge of bad spelling and dismisses the imputation of bad luck, but still it avoids the central question. If it is presumed that a poet has the means to do what he likes in his poem, what kind of acrostic is a distorted acrostic? What kind of pattern is a broken pattern? What kind of success is there in the near miss? What kind of game is it that obscures its own rules?

This at least is familiar ground. While the type of word play exemplified by the acrostic, the treatment of words as plastic objects, is perhaps yet to emerge as a central theme in twentieth-century writing—the return of the concrete poem is but one indication—, the revival of concern with games in literature and with literature itself as game is certainly widespread.¹¹ Many readers (and I emphatically include myself), though, are somewhat at a loss to appreciate the frisson of joy that mathematicians and a growing number of writers find in word play; but there is no difficulty in finding a modern master of literary gamesmanship who may help us articulate the subtleties of such pleasure and put us in a better position to intuit the sensibility behind the near miss.

You will recall an episode in Pale Fire: John Shade—nominal author of the 999 verses that begin the novel—John Shade, obsessed throughout most of his life, but particularly since the suicide of his daughter, with thoughts of death and the riddle of an after-life, suffers a cardiac arrest. During his collapse, which he understands without question as his death, he has a distinct vision of a tall white fountain. He is revived and naturally continues to be fascinated with this clue until he reads a magazine interview with a woman he calls Mrs. Z. whose heart had stopped on the operating table and like Shade had been revived. Like Shade too, she saw a tall white fountain—confirmation! He visits the woman but finds her personally so repulsive that he cannot bear to risk the intimacy a direct question and confession would imply; so he leaves her and goes to speak with the journalist who wrote the original interview.

He took his article from a steel file:
“It's accurate. I have not changed her style.
There's one misprint—not that it matters much:
Mountain, not fountain. The majestic touch!”
Life Everlasting—based on a misprint!
I mused as I drove homeward: take the hint,
And stop investigating my abyss?
But all at once it dawned on me that this
Was the real point, the contrapuntal theme;
Just this: not text, but texture; not the dream
But topsy-turvical coincidence,
Not flimsy nonsense, but a web of sense.
Yes! It sufficed that I in life could find
Some kind of link-and-bobolink, some kind
Of correlated pattern in the game,
Plexed artistry, and something of the same
Pleasure in it as they who played it found.
It did not matter who they were. No sound,
No furtive light came from their involute
Abode, but there they were, aloof and mute,
Playing a game of worlds, promoting pawns
To ivory unicorns and ebon fauns;
Kindling a long life here, extinguishing
A short one there; killing a Balkan king;
Causing a chunk of ice formed on a high-
Flying airplane to plummet from the sky
And strike a farmer dead; hiding my keys,
Glasses or pipe. Coordinating these
Events and objects with remote events
And vanished objects. Making ornaments
Of accidents and possibilities.(12)

We have to be careful how seriously we can take this insight; after all it is Nabokov I've been quoting, and the fictional Shade himself withdraws from this certainty a few lines later in the poem. But the passage does, I think, shed light on our problem. The near miss as confirmation, as evidence, dim but irrefutable, that a whole pattern of sense does exist; a secret sign to those with a private longing for coherence: now they see as through a glass, darkly, but they do see, excited witnesses present at the moment when design, trembling, just emerges from chaos. The sensibility of the near miss is not everyone's: procul, O procul este, profani. It is the joy of hermeneutics and by nature esoteric, aristocratic in a way, even mandarin. The connoisseur of its delicate delights avoids the straightforward with the same energy John Shade displays leaving Mrs. Z., and for the same reason, loathing of vulgarity. The pleasure is not in bright lights but the subtle iridescence of sense and nonsense. Obliquity, esotericism, and at all costs nothing vulgar: these may not be, should not be our terms, but they are Nabokov's and they are the Hellenistics'.

Nabokov, curiously, has found his way into the mainstream of modern literature (or perhaps, curiouser, the mainstream has found him.) There is a literary enterprise, however, more marginal than Nabokov's that is in some ways closer to the Hellenistic sensibility, the work of the Ouvroir de Littérature Potentielle, OuLiPo, the Paris-based society that has devoted itself to sophisticated and serious word play.¹³OuLiPo continues the ancient traditions of word play, the acrostic, the anagram, the palindrome (one was produced of over 5,000 letters),¹⁴ the snowball sentence, bilingual puns, etc.; and it invents new modes. One book by the society's cofounder, the late Raymond Queneau, that has received some notoriety consists of ten sonnets, each sliced into strips, one verse per strip, so that the reader can turn the strips instead of pages and find himself confronted, if he has (literally) nearly infinite patience, with 10¹⁴ sonnets, all making sense: the book is called One Hundred Thousand Billion Poems.¹⁵ An experiment I find especially intriguing involves the transformation of a text by replacing its words with their dictionary definitions, transforming the new text in the same manner, and again ad lib. What results plainly is on the threshold of nonsense though by the rules, algebraic if not linguistic, the semantic integrity of the text has not been lost: there is method in the madness if you know where and how to look for it. A group of similar projects takes all the words of one part of speech in a familiar text and replaces them with others of the same part of speech chosen according to some specified rule: each noun, for example, replaced by the noun that comes seventh after it in the dictionary, or the first noun of a text replaced by the last, the second by the next-to-last, and so on.¹⁶ Again, sense is tortured but never obliterated because the system of syntax is preserved. In fact behind this playing tag with chaos lies the presupposition, the necessary presupposition of system, of an order total and finely articulated: not only the iron rules of the game but the larger systems as well—logical, algebraic, linguistic—in which there is consummate sense. The gamesmen of OuLiPo are not precisely seekers after meaning like John Shade. They are more like his enigmatic gods, the remote imps who control the world, making ornament—κόσμοs—of possibilities, and hiding the keys. Once privy to their secret methods, we can see against the strange darkness of their results the twinkle of grand cohesion. This is what the gamesmen of OuLiPo and all literary gamesmen, I think, savor: the sensibility of the near miss is, from a different perspective, the sensibility of total coherence.

Belief in the web of sense, that the universe is in fact thoroughly designed is, of course, widespread and multifarious. It is, for example, the scientist's faith and the sorcerer's. Clinically, it has been called paranoia, as readers of Thomas Pynchon know and more than one visitor to Thomas Jefferson's Monticello have suspected. Religiously, philosophically, it has many names, not the least apt of which is Stoicism, generally considered to be the intellectual matrix of Aratus's Phaenomena. But the approach to the sensibility of total coherence has to be different for those of us whose trust in the order of the universe is limited by the suspicion that while the cotillion of neutrinos, pi-mesons, and Joycean quarks, charmed or otherwise, may dance its spastic dance to the sweetest of rhythms—the Heisenberg Fling?—it still will never sing the song of human passion. Points on a graph, like stars in the sky, may be fixed; a bit of the curve that connects them may be sketched or an asterism identified; but the equation that describes the whole curve we expect will remain unknown, the great constellation unnamed.

For many of us, attempts to map the universe inevitably fail. Art can offer momentary illumination, a tentative pattern; and this is not trivial: we all know of books of great scope and importance that are very satisfying and true by any sensible criterion—the Iliad, to name only the greatest and truest. But this is not all we can expect:

cum canerem reges et proelia, Cynthius aurem
vellit et admonuit, “pastorem, Tityre, pingues
pascere oportet ovis, deductum dicere carmen.”(17)

There are works of guardedly unambitious proportions and intent that make available even to skeptics the sensibility of total coherence. These are works that gladly embrace the limits of their own finite resources—a fixed vocabulary, an unyielding syntax, a very small number of letters—, self-conscious works that, though circumscribed, are complete, that are in themselves thoroughly designed—in Latin perfecta—and they are works that deserve some attention.

Aratus's Phaenomena is a poem about the manifestation of design. The casual star-gazer experiences the sky directly and with true insight into what it is, a busy tableau of white specks randomly arranged against a dark background. With instruction and practice he learns to read the heavens, to group stars into the arbitrary patterns of constellations which, once named, are irradicably fixed. For Aratus, though, there is nothing random or arbitrary about the sky. Rather he finds there immanent order in and of Zeus, the god understood doubly as an individual divinity and as Sky itself, that is, as both designer and design. And not order alone but significance:

αὐτὸs γὰρ τά γε σήματ' ἐν οὐρανῳ ἐστήριξεν,
ἄστρα διακρίναs, ἐσκέψατο δ' εἰs ἐνιαυτὸν
ἀστέραs οἵ κε μάλιστα τετυγμένα σημαίνοιεν
ἀνδράσιν ὡράων, ὄϕρ'ἐμπεδα πάντα ϕύωνται.

(lines 10-13)

[For it is Zeus who set the signs in the heavens, who demarked constellations and devised for the year the chief stars that give to men the signs of the seasons, so that all things might prosper.]

The constellations are both signs of god's benevolence to be interpreted by men and visible tokens of his coherent design. The signs may be hard to discern or their interpretation obscure, but if the design appears unfinished, it is Aratus's trust that it is complete nonetheless. A passage just before the first acrostic reads

                                                                                                    πάντα γὰρ οὔπω
ἐκ Διὸs ἄνθρωποι γινώσκομεν, ἀλλ' ἐτι πολλὰ
κέκρυπται, τῶν αἴ κε θέλῃ καὶ ἐσαυτίκα δώσει
Zεύs(.) ὁ γὰρ οὑν γενεὴν ἀνδρῶν ἀναϕανδὸν ὀϕέλλει,
πάντοθεν εἰδόμενοs, πάντη δ' ὅ γε σὅματα ϕαίνων.

(768-772)

[We mortals do not yet know all things from Zeus, but much is still hidden which he will grant as he wills; for he aids the race of men openly, manifesting himself on all sides and revealing his signs everywhere.]

As an individual or even a teacher, Aratus remains like other men, the exegete of the heavenly text. But as a didactic poet, he becomes responsible for a text of his own, Zeus-like¹⁸ in his own domain. His resources are a systematizing intelligence and ultimately the system of his language. Of this he is well aware; understanding, for example, the ancient link between the recognition of pattern and the first act of language, the process of naming. His description of the work of the primeval astronomer is to the point:

                                                                                τά τιs ἀνδρῶν οὐκέτ' ἐόντων
ἐϕράσατ'ἠδ'ἐνόησεν ἄπαντ' ὀνομαστὶ καλέσσαι
ἤλιθα μορϕώσαs. οὐ γάρ κ'ἐδυνήσατο πάντων
οἰόθι κεκριμένων ὄνομ'εἰπεῖν, οὐδἐ δαῆναι.
πολλοὶ γὰρ πάντη, πολέων δ'ἐπὶ ἐσα πέλονται
μέτρα τε καὶ χροιή, πάντεs γε μἐν ἀμϕιέλικτοι.
τῶ καὶ ὁμηγερέαs οἱ ἐείσατο ποιήσασθαι
ἀστέραs, ὄϕρ'ἐπιτὰξ ἄλλῳ παρακείμενοs ἄλλοs
εἴδεα σημαίνοιεν. ἄϕαρ δ'ὀνομαστὰ γένοντο
ἄστρα, καὶ οὐκέτι νῦν ὑπὸ θαύματι τέλλεται ἀστήρ.
ἀλλ' οἱ μἐν καθαροῖs ἐναρηρότεs εἰδώλοισιν
ϕαίνονται(.) τὰ δ' ἐνερθε διωκομένοιο Λαγωοῦ
πάντα μάλ'ἠερόεντα καὶ οὐκ ὀνομαστὰ ϕέρονται.

(373-385)

[Someone, long dead, noted the stars and understood to call them all by one name as he grouped them in figures. He could not know all of them singly nor name them, there are so many stars all over, many of the same brightness and color, and all wheel in circles. So he grouped the stars together, so that, set in order, in relation, they might form figures. Thus the constellations got their names, and now no star rises above the horizon, a lone marvel.

The other stars shine grouped in clear figuration, but those beneath the constellation the Hare are all in mist and without names.]¹⁹

A star, isolated and nameless, is obscure and without meaning; significance is born of pattern, and pattern of connection. The last lines of the poem are a warning against the uselessness of an event in isolation:

τῶν μηδἐν κατόνοσσο(.) καλὸν δ'ἐπὶ σήματι σῆμα
σκέπτεσθαι(.) μaλλον δἐ δυοῖν εἰs ταὐτὸν ἰόντων
ἐλπωρὴ τελέθοι, τριτἁτῳ δέ κε θαρσήσειαs. …
τῶν ἄμυδιs πάντων ἐσκεμμένοs εἰs ἐνιαυτὸν
οὐδέποτε σχεδίωs κεν ἐπ'αἰθέρι τεκμήραιο.

(1142-1144, 1153-1154)

[Ignore nothing. It is proper to look for sign confirming sign: when two coincide, that gives you hope; when three, confidence. …

Watch all the signs together throughout the year and never will you guess at random.]

The three acrostics of the Phaenomena taken together also bear out Aratus's linguistic self-consciousness. Their key words, as I have mentioned, are all crucial terms for his poem and aesthetic—subtlety, totality, and signification; and the acrostics themselves exploit and underscore three basic aspects of the linguistic system he uses—style, grammar, and orthography. But the visual aspect of his language allows him perhaps a greater privilege, to arrange constellations of his own, here black on white but, as bold as Orion or as shy as the hazy Pleiades,²⁰ still with a clarity analogous to Zeus's.

For Aratus, sight and sense are closely related. The gods can show but their signs must be mediated by speech; they can speak but what they mean must be shown to men. At a critical transition in the poem, the movement earthward from celestial to atmospheric phenomena, Aratus makes this evident by a neat juxtaposition:

πάντη γὰρ τά γε πολλὰ θεοὶ ἄνδρεσσι λέγουσιν.
οὐχ ὁράᾳs;

(732-733)

[For on all sides many things do the gods speak to men. See!]

This theme, its articulation in the Phaenomena and in Hellenistic and Roman literature, could be explored at much greater length, but that exploration lies beyond the scope of my intentions.

Notes

1. A. S. F. Gow, Bucolici Graeci (Oxford: Clarendon Press, 1952).

2. “Aratus will live as long as the sun and the moon,” Amores 1.15.6.

3. There does exist an excellent translation of Aratus into English verse: Stanley Frank Lombardo, Aratus' Phaenomena: an Introduction and Translation (Ph.D. diss., University of Texas, 1976). Until this is published, however, the graceless Loeb edition of G. R. Mair (London: W. Heinemann, 1921) is the most easily available to the Greekless reader. All translations in this article are my own.

4. The “Besantinou Bomos” of Roman date, Gow, Bucolici Graeci, p. 184. While acrostic poems are not rare in Greek literature, Latin authors seem to have taken to them with more enthusiasm; from the hoary father of Latin literature Ennius, whose acrostic signature—Quintus Ennius fecit—Cicero records (de Div. 2.111), through a contemporary of Constantine, Publilius Optatianus Porfyrius, who bore prodigies: see, for example, his eighteenth poem.

   For Vergil's recreation of one Aratean acrostic, see Georgics 1.429-433, and E. L. Brown, Numeri Vergiliani, Collection Latomus 63 (1963): 102-4.

5. “Sur un Acrostiche d'Aratos,” Revue des Etudes Anciennes 62 (1960): 48-61.

6. Jacques's remarks on the programmatic content of this whole passage are excellent. See also the valuable comments of E. D. Francis, “Aratos” ΑΓΡUNΟσ and Vergil (Georg. 1,424-437),” forthcoming in Hermes. I am grateful to Prof. Francis for permission to read and cite this paper.

7. “Nicander's Signature,” Classical Quarterly 22 (1928): 114.

8. W. C. Helmbold, American Journal of Philology 76 (1955): 110. Helmbold's conjecture was frankly made to avoid a defective acrostic; he confessed himself unable, however, to remedy the difficulty in the fourth line. J.-M. Jacques, in “Les ‘Alexipharmaques’ de Nicandre,” Revue des Etudes Anciennes 57 (1955): 20, conjectures ναὶ μήν for σὺν δἐ καὶ, and ἀσκηροῦ for καστηνοῦ: if he is correct, the acrostic is complete. A. S. F. Gow and A. F. Scholfield, Nicander: The Poems and Poetical Fragments (Cambridge: At the University Press, 1953) print the initial letters as read by Lobel.

   Perhaps understanding is needed rather than correction.

9. Brown, Numeri Vergiliani, p. 109.

10. Ibid., p. 102-14.

11. The literature on literary games is now of great volume and disheartening sophistication. A reader might begin with the lucid first chapter of Elizabeth Sewell, The Field of Nonsense (London: Chatto and Windus, 1952); or, in a different vein, with Game, Play, Literature, Yale French Studies 41 (1969); or, best of all, with Francis Huxley, The Raven and the Writing Desk (London: Thames and Hudson, 1976).

12. V. Nabokov, Pale Fire (New York: Putnam, 1962), p. 62-63.

13. See Paul Fournel, Clefs pour la littérature potentielle (Paris: Editions Denöel, 1972); Oulipo, La littérature potentielle: Créations re-créations récréations (Paris: Gallimard, 1973); Harry Mathews, “Oulipo,” Word Ways 9 (May 1976): 67-74; Martin Gardner, “Mathematical Games,” Scientific American 236 (February 1977): 121-26.

14. By Georges Perec, in La littérature potentielle, p. 101-6.

15. Cent mille milliards de poèmes (Paris: Gallimard, 1961).

16. An example concocted by Martin Gardner, “Mathematical Games,” p. 123: the first sentences of Moby Dick transformed according to the Noun-plus-7 algorithm as it's called—

    Call me islander. Some yeggs ago—never mind how long precisely—having little or no Mongol in my purulence, and nothing particular to interest me on shortbread, I thought I would sail about a little and see the watery partiality of the worriment.

    Or the first noun/last noun exchange using Melville's first chapter—

    Call me air. Some hills ago—never mind how long precisely—having little or no phantoms in my whale, and nothing particular to interest me on processions, I thought I would sail about a little and see the watery soul of the purpose.

    Beautiful, whatever it is.

17. “When I was singing of kings and battles, Apollo plucked my ear and said, ‘A shepherd, Tityrus, should let his sheep grow fat but keep his poems slender.’” Vergil, Eclogues 6.3-5; a translation of the advice given Callimachus (Aetia 1.23-24):

              ἀοιδέ, τὸ μἐν θύοs ὅττι πάχιστον
    θρέψαι, τὴν Μοῦσαν δ'ὠγαθἐλεπταλέην.

    in which deductum represents λεπταλέην, a version of λεπτή.

18. This equation may be suggested early in the poem:

    'Εκ Διὸs ἀρχώμεσθα, τὸν οὐδεποτ'ἄνδρεs έῶμεν
    ἄρρητον

    (ll. 1-2)

    If APPHTON conceals APATON, then the true author of pattern is indeed “not unnamed.”

19. John Hollander has cited this passage as a note to his poem “The Great Bear.” See Spectral Emanations: New and Selected Poems (New York: Atheneum, 1978), pp. 237-38. It is not surprising that Hollander, the foremost living practitioner of the figurative poem (among other distinctions), should have recognized Aratus.

20. Aratus's description of this group (ll. 254-267) as an incomplete constellation one of whose stars is missing provides a neat analogue to the third acrostic.