Robert Benchley, the actor, critic and member of the Algonquin Wits, once quipped that “There are two kinds of people in the world, those who believe there are two kinds of people in the world and those who don’t.” At the risk of murdering to dissect and conferring ontological status upon a distinction that is a mere abstraction and obviously untrue, I would like to begin by suggesting that there are exactly two things happening in every poem: what a poem says and what a poem does. Describing, explaining and evaluating these two things or actions is the job of the critic and, perhaps after more time has passed, the scholar. Further, despite heroic efforts by Roman Jakobson, W. K. Wimsatt, John Crowe Ransom, John Hollander, Terry Brogan, and before them, Pope, Johnson, and many others, it has always been the case, as it is now, that critics and scholars are likely to be better at doing one of these things – and it is always the same thing, which is to explore what a poem says – than they are at doing the other, which is to explain what a poem does.
The reductive, binary fiction I just suggested – between what a poem says and what it does – has gone by many names, such as sound and sense, form and content, and so on.  In this essay I’m going to spend some time explaining the problems with those terms, and then briefly propose a new response to the problem of the relationship betweenwhat a poem says and what a poem does, which I think has been broadly misconstrued by most critics throughout history.
Despite attempts to kill off the distinction between “form” and “content,” attempts which began with the New Critics and which have now extended through generations of linguists and literary theorists who remain far more deeply rooted in the canons of neo-Aristotelian philosophy than they generally realize, millions of schoolchildren still learn that works of art, in particular poems, have something called “form” which stands in some relation to something called “content.” The form is generally taken to be the literary prosody, verse structure and generic structures or qualities of the work. The content is purely semantic: a set of paraphrasable meanings, the poem’s rhetoric, by which I mean its argument or propositional sense.
As many have pointed out, part of the problem that the distinction between “form” and “content” creates is that it insists that each aspect of a poem is a thing, or something that is very much like a thing, a “verbal icon,” as W. K. Wimsatt put it so long ago. Some good, but probably more bad has come from this metaphor. The flaws in the metaphor of the poem as object drove Stanley Fish into such paroxysms of distraction that he wound up asserting that because poems cannot be things, they must not exist at all except as manifestations of “reader response” – but that’s another story. Here it is sufficient to point out that when the metaphor of the poem as object is taken too literally it misrepresents language, which can be stored (in writing) rather like a thing, but is actually far more like an activity – indeed may well best be understood as an activity tout court. More of that below.
The critical problems that emerge when we treat poems as things that store semantic information (what they say) become particularly acute when we seek to interpret how they work as verse (what they do). In fact, treating a poem only as a thing that stores semantic information seems to make it impossible to respond to verse as verse, and thus compromises that task of criticism. The central dilemma grows from the fact that words are the most referential materials used in any art and therefore conjure propositional sense at every turn. After all, words are what statements and critical arguments made out of. And treating statements, arguments or ideas as things that store semantic information is a technique that proceeds quite well most of the time if what one wishes to do is to analyze their propositional sense. The problem is that poems not only say things; they also do things, and what they do as verse is crucial to what they are. Therefore, subdividing words into binary conceptual entities such as “sound” and “sense, “form” and “content,” or even (as Ransom did) “logical structure” and “local texture,” presents an immediate crisis, a splitting of poetry into sense on the one hand and prosody on the other. And if prosody is expected to operate like a construable statement that means something – like a thing instead of an action – it then follows that it means…well, what? If the prosody of a poem only means exactly what the words are already saying, why bother? And does that even make sense? Is prosody merely an onomatopoetic setting of propositional sense? And if not…well, again, what does it mean?
The predicament is that our tools for close reading of the referential dimensions of poetry are quite powerful, so powerful, in fact, that they seem at times to have a life of their own. They are, after all, words applied to words, and they therefore seem to want everything that they touch to be words as well. In this capacity I think of close reading as a group of Amazon sirens calling to those who would talk about prosody as its own realm, urging them to steer the sweet ship of verse straight onto the rocks of semantics.
The poets and critics on the ship of verse, however, know that the strongest poets everywhere have always passionately insisted on prosody’s primal, generative, independent and defining significance for poetry. As A. E. Stallings suggests in “Crooked Roads Without Improvement,” verse is the primary thing poetry does that prose does not do. Prose can contain any word and any group of words in any possible order; it can tackle any subject, theme or argument; it can use any figure of speech and include any image; it can employ any and every construction of syntax and grammar; it can even employ widely different rhythms (Saintsbury wrote an entire book about them); and virtually all of these subjects are fodder for close reading. The one thing prose cannot do is … be verse, because verse by definition does something different than prose. 
The crux is to realize that despite its dependency on words, versification is not a referential function of language. In and of itself it has no diction, grammar, syntax, figuration, tropes, thematics, tone, or any kind of inherent, independent propositional sense. This maddening lack of referentiality means that critics who attempt close readings of prosody as if it were everywhere analogous to propositional sense are sailing without a chart in the fog. The tools are just not suited to the job. For close reading is powerful, beautiful and seductive, but it is useless outside a referential system. To apply it to the prosody of a poem is thus a self-sabotage from which the interpretation generally emerges capsized if not wrecked. It makes the poem look as if all of its prosody were a mere illustration of its relatively objectifiable argument, or its imagery, or, most broadly, its verbal sense and meaning. We murder to dissect.
Because poetry is crucially defined by being in verse, good poetry criticism should therefore attend to the nature and quality of those verses, including metrics, other aspects of verse-form, genre and so on. And yet, as I hope the foregoing suggests, as soon as one enters into the discussion of how that can or should happen, problems surge on all sides like treacherous currents and the going gets heavy. In most cases the critic makes the prosody of the poem analogous to the meaning of the poem and then proceeds in a circle. Influential examples are deliciously ubiquitous. In her 1965 study The Language Poets Use, Winifred Nowottny makes the case for “trans-verbal values”: feelings, or affects, built up from the words, that are apparently not in the meanings of the words themselves. For Nowottny as for others, the trans-verbal values of verse – its prosody – are at first consigned to a realm of non-semantic sound:
A formal structure [such as verse] capable of articulating, in its own terms, finer differences beyond the discrimination-level than is possible in blunt verbal terms, charges those merely verbal terms with precise values of another order of existence.
These values are different from the meanings of the words of the poem, but since they only exist in the sound of the very same words, Nowottny is forced to locate them in a nether region, “another order of existence.” The sounds of words in verse, especially in the most powerful poems, signify something that is apparently not verbal, “a kind of contact with the particularity of the (unverbal) world, which the very nature of language ordinarily denies to us.” As a result, prosody refers the reader (or listener) to a “particularity of the unverbal world” that emerges from words, but which words cannot describe. The fascinating next step is that Nowottny cannot resist the temptation – indeed, most critics cannot resist the temptation – to interpret the trans-verbal world with words, and the tools they use are almost always the wrong ones.
How so? Nowottny, with the best of critical intentions, and like legions of other sailors since the dawn of criticism, always rediscovers the meaning of the verse structure in the verbal sense of particular poems and thereby immediately sends the larger meaningfulness of the poem to its watery grave. Here come the rocks. Directly following the passage quoted above, Nowottny analyzes the versification of a Dryden poem, a song from The Spanish Fryar:
Farwell ungratefull Traytor,
Farwell my perjur’d Swain,
Let never injur’d Creature
Believe a Man again.
The Pleasure of Possessing
Surpasses all Expressing,
But ‘tis too short a Blessing,
And Love too long a Pain.
In the fifth and sixth lines the combination of the feminine rhyme with the concentration on similar vowel and consonant groups gives an effect of delightedly dwelling on the same experience over and over. But after a transition line in which only a trace of these sound-groups survives, and an l prepares for the “Love … long’ alliteration in the last line, we are suddenly deprived, at the word “Pain”, of this dancing rhyme, and at the same time “Pain” takes over the p-alliteration found in the fifth and sixth lines, dropped in the seventh, and nowhere echoed in this line except on the word “Pain”, which therefore makes a sharp antithesis to “Pleasure”.
Big problems here. Far from being understood as a “trans-verbal value,” the words of the poem have become ad hoc terms for its prosody. The idea that in the fifth and sixth lines “the feminine rhyme with the concentration on similar vowel and consonant groups gives an effect of delightedly dwelling on the same experience over and over,” is a circular interpretation. It translates the “pleasure” of the poem’s sense into a value of the verse structure (in the words “delightedly dwelling”). This is not meaningful prosodical analysis, because it confers “meaning” on the prosody of the passage by interpreting it as a mere illustrations of the sense of words. Nowottny’s terms, however, carry no weight whatsoever in terms of the linguistic categories or principles of versecraft, growing instead only out of the propositional sense of this particular close reading of this particular poem. That is why the reading is ultimately unsatisfying, giving the feeling of a tautology. Despite Nowottny’s caveats about the inability to describe all the values derived from a particular poem, she just cannot resist interpreting the poem’s prosody in terms of her close reading of the meanings of its words. It is Winters’s “fallacy of imitative form” par excellence, except practiced by a critic, not by a self-justifying poet.
Here is how Winters first defined the fallacy, in 1937, in an essay in Primitivism and Decadence:
To say that a poet is justified in employing a disintegrating form in order to express a feeling of disintegration, is merely a sophistical justification for bad poetry, akin to the Whitmanian notion that one must write loose and sprawling poetry to “express” the loose and sprawling American continent. In fact, all feeling, if one gives oneself (that is, one’s form) up to it, is a way of disintegration; poetic form is by definition a means to arrest the disintegration and order the feeling; and in so far as any poetry tends toward the formless, it fails to be expressive of anything.
Even Winters himself, however, despite his own warning, could not resist the trap of reading sound as an annex of sense on occasion. Consider how he reads this brief passage from Herrick in The Anatomy of Nonsense:
… the elves also
Whose little eyes glow
[T]he quick movement of the second line is actually amusing: perhaps in part because of some intrinsic quality of sound, however slight it may be in language; perhaps in part because there is a suggestion of the quick movement of the little creatures who are being described.
Winters defends himself against the idea that he is committing his own fallacy here, because Herrick is not attempting to express the actual subject matter imitatively, but rather is successfully expressing “his own understanding of his subject matter and the feelings properly motivated by that understanding.” The distinction is hardly meaningful; one could make the same argument for any passage one admires.
Harvey Gross’s well-known Sound and Form in Modern Poetry also relies heavily on fallaciously imitative readings of prosody. Gross makes his perspective quite clear: “A poem’s prosody cannot exist apart from its propositional sense.” Certainly true, but that doesn’t mean the prosody has to imitate that sense. Yet Gross frequently casts prosody in an imitative role, acknowledging that his position derives from Suzanne K. Langer’s ideas in Philosophy in a New Key. Langer’s central insight is that rational discourse is incapable of completely describing the meanings of symbols, meanings which exist nevertheless. Gross takes versification to constitute an aural symbol system as Langer describes it, an intellectual move quite similar to Nowottny’s. The problem is that, when turned to practical criticism, the approach inevitably and immediately re-verbalizes the symbolic structures.
As a result, Gross bases interpretations of sound and form on the meanings he finds in particular poems, just like Nowottny and even at times Winters. Gross applies this imitative approach to the graphic element of poems as well as to their aural dimension – to any and all manifestations of verbal form. Consider his analysis of poem 13 from E. E. Cummings’s No Thanks, “r-p-o-p-h-e-s-s-a-g-r”:
The poem resembles its subject; the disintegrating and reintegrating grasshopper. Actually, the poem does not so much look like the grasshopper’s action as give the feel of action. Cummings uses an elaborate technique of synaesthesia, a complex visual and aural derangement, to signify emotional meaning.
The reading is circular. The grasshopper is a represented thing in this poem, not an aspect of its versification. Other words, that do not represent hopping animals, could be arranged in exactly the same way; grasshopperity is not a principle of versification. 
Examples of imitative reading like those in Nowottny, Gross, and even Winters could be multiplied endlessly from contemporary reviews and books; the ones offered here are older, but only to show the exalted provenance of the concept. To discuss but one more influential critic who goes down the imitative path, Northrop Frye makes exactly the same move in Anatomy of Criticism. Leaving aside Frye’s Lanierite scansion and cranky ideas about what constitutes “music” in poetry, we find he eventually goes straight for fallaciously imitative readings of prosody. He argues that in Spenser’s line “‘The Eugh obedient to the bender’s will,’ the line has a number of weak syllables in the middle that makes it sag out in a bow shape.” This reading glorifies aesthetic impressions by burying them in a supposedly systematic scansion. The problem is not with the reading of the line’s sense, but rather the hypothetical relation between the sense and the versification, which Frye has approached backwards. For verse is not an imitation of a poem’s meaning, but rather one source of its power, coeval with propositional sense.
Some particularly thoughtful critics who are also particularly strong poets have realized the profound nature of the problems that arise when we talk about prosody in terms that should be reserved for close reading. The most notable of these is still probably Pope, who provides the modern locus classicus on the matter in the “Essay on Criticism.” In the famous passage, Pope qualifies his observation about the relations of sound and sense very, very carefully. He does not say “The sound must be an echo to the sense,” which would scan just fine and retain even the vowel value of the key syllable, but ratherhe says “The sound must seem an echo to the sense.” Pope’s cunning caution in that one word shows that he sensed that there was a function more complicated than simple imitation at work, although he never said what he thought it was. Most readers of the passage, of course, read it exactly if Pope had said “be” – but he does not.
Johnson also understood the problem and suggests it more analytically in a brief but acute insight. While willing to ascribe some mimetic affects to verse, he points out “It is scarcely to be doubted, that on many occasions we make the musick which we imagine ourselves to hear, that we modulate the poem by our own disposition, and ascribe to the numbers the effects of the sense.” Johnson frames the issue precisely. The problem with mimetic reading is that prosody is forced to bear the burdens of verbal sense. But if we want to understand versification in individual poems, which is one of the crucial tasks of criticism, the only solution is to investigate how meaningfulness grows out of prosody as much as meaning grows out of the rhetoric of statements, propositional sense, arguments, ideas and so on.
Among more recent strong poet-critics John Crowe Ransom has stood virtually alone in stubbornly insisting on the differences between sound and sense and then following through on the distinction. More than any other critic, he refused to give in to the critical sirens who would seduce the ship of prosodical interpretation onto the syllogistic rocks of close reading. Recognizing its ancient philosophical pedigree, Ransom labels the critical impulse to unify sound and sense “Eleatic,” taking that term from the group of Pre-Socratics founded by Parmenides who argued that all things could and should be reconciled into a monistic vision of being. In his 1941 essay “Wanted: An Ontological Critic,” Ransom counters by arguing that a poem’s peculiar properties grow out of heterogeneity, not similarity (though it’s worth noting that he reverses Nowottny’s location of tangibility and transcendent values): “As a thing of sounds it exists in the words; as a thing of meanings it exists in a world beyond the words. The heterogeneity is rather extreme.” Ransom, like Winters, thoughtfully criticizes the notion of “expressive form”: “The single word does not in fact resemble appreciably the thing it denotes … There is no point-to-point coordination between the development of the semantic structure and that of the phonetic structure.” He is then far more careful than most of the others, however, to resist the impulse to interpret verse structure as if it were construable like a sentence.
So now my argument, I hope, is becoming more clear. It is generally only the poets, or rather a group of stronger poet-critics, such as Pope, Johnson, Winters (sometimes…), Ransom and others, who have grasped the dangers of Eleatic reading and the fallacy of imitative form. They have understood that treating the prosody of poems as a manifestation of what they say – ascribing to the numbers the effects of the sense – reduces poems to art objects that merely say things, rather than speech acts that also do things. A dissociative approach, in which, following Ransom, we stubbornly seek a criticism that can somehow talk about the “meaningfulness” of verse (what Ransom calls its “ontology”) without reducing it to a mere province of “meaning,” is essential to good criticism, for otherwise a great deal of the art becomes philosophy and rhetoric’s bastard.
As I have implied throughout this essay, one way to cut the Gordian knot of the sound/sense, form/content, local structure/logical texture conundrum is to think of poems as speech acts rather than as objects. As J. L. Austin and John Searle have taught us, we have no trouble imagining and engaging in illocutionary speech acts all the time, making sentences, such as promises, vows, declarations, oaths, laws, curses, blessings, and so on, that not only say something but also do something. We can also imagine saying words that mean one thing, but do another, like irony and sarcasm, or as Searle has argued, any metaphor whatsoever (because, for the metaphor to work, it must be understood as doing something other than what it is saying; a man is not really a wolf). Perhaps one way, then, to see what prosody does in a poem, is to explore what it does, which may be quite different from what it says. Just like an illocutionary speech act, a poem says something and does something at the same time.
If the prosodical component of a poem is not merely a reflection of meaning, but is instead a meaningful speech act, then the question remains what kind of a speech act it is. And here is the key, which has been hiding in plain sight: as an organizational principle of art, prosody has nothing to do with the referential functions of language. Rather, verse draws its power from an utterly different faculty, the number sense, which orders experience not by construing it into propositions but instead by categorizing and counting, an activity that does not require linguistic syntax. Only by drawing upon a theory that treats all elements of versification as somehow dependent on the number sense, broadly understood, can we come to see more clearly the heterogeneity that Ransom intuited and perhaps find a better way to talk about it. Speech act theory is tremendously helpful here, indeed a kind of Rosetta Stone, because of the way it helps us to understand how we can do things with words other than simply make statements. When we then recognize the crucial importance of the number sense in verse, it gives us, perhaps for the first time, a sense of the disjunct faculties we synthesize in the making of and enjoyment of poems. Language and counting fuse in verse and they are equal; such an approach can help us to think about literary prosody as being in and of itself meaningful as a poetic function or activity, yet still utterly different from the semantics of propositions.
This is not a completely new idea, but it is a slippery one that needs to be perennially restated. Many earlier poets and critics who have written about verse have intuited the power of number as an organizing principle of verbal art, indeed of art in general. Indeed, Samuel Johnson, as we saw in the quotation above, refers to the making of verse as “writing in numbers,” a common observation at the time. In “The Epistle to Dr. Arbuthnot” Pope writes, “I lisped in numbers, for the numbers came” and similar comments from other poets writing in English could be multiplied endlessly, including among the moderns. “Meter,” after all, is “measure,” something even William Carlos Williams knew was essential, going to great pains in his theory of the variable foot to show that “the crux of the issue is measure.” The tradition is far older than our own poetry, however. St. Augustine when discussing versification in De Musica writes that “the science of modulating is the science of moving well, in such a way that the movement is desired for itself, and for this reason charms through itself alone.” Another way to say this is to observe that such regular and therefore measureable movement is itself a manifestation of order, even though it does not participate in construable syntax. Strong poets have always understood the number sense, indeed been obsessed with it, just as obsessed as they are with words, which is why explorations of the number sense in verse extends back to the beginning in all traditions, indeed by definition into every use of and discussion about meter whatsoever.
Again, it is the strong poet-critics who have best understood this notion of verse making as a meaningful action of counting, and who have thus carefully refused the temptation of the fallacy of imitative form. In Donne’s “The Triple Foole” he neatly ties together the ordering powers of language, counting and writing into a single critical observation:
I thought, if I could draw my paines
Through Rimes vexation, I should them allay,
Griefe brought to numbers cannot be so fierce,
For, he tames it, that letters it in verse.
There is of course much to say about this passage that has nothing to do with prosody. For our purposes, it is notable that Donne describes how counting and other aspects of prosodic artifice (“Rime,” which we should remember derives from “rhythm”) bring an order to language that language alone can never discover. To poets, this has always seemed obvious: isn’t that the whole point? Indeed, it is counting that defines verse, even free verse, which is after all still written in discrete, countable lines and is therefore not prose. The action and experience of the counting is every bit as important as the words.
Examples of strong poets who indicate their awareness of the number sense could be multiplied endlessly. The foundational importance of the number sense is a profound creative and critical truth hiding in plain sight, but one that very few critics have seen, let alone understood, as most critical terminology renders it invisible. Among contemporary poets, one who is deeply aware of the issues is W. S. Merwin. In his recent Pulitzer-prize winning volume, The Shadow of Sirius, Merwin includes a poem, “The Long and the Short of It,” that indicates how deeply he perceives the magic of counting as one source of poetry:
As long as we can believe anything
we believe in measure
we do it with the first breath we take
and the first sound we make
it is in each word we learn
and in each of them it means
what will come again and when
it is there in meal and in moon
and in meaning it is the meaning
it is the firmament and the furrow
turning at the end of the field
and the verse turning with its breath
it is in memory that keeps telling us
some of the old story about us
This wonderful poem is filled with critical puns, games, and suggestions of metrics. The title of course refers punningly to classical scansion but also suggests that such measure is a totality, an idea the poem then articulates. The poem’s 14 lines suggest the sonnet, and while only a few of the line endings rhyme (take/make, us/us), there are others that suggest rhyme in one way or another (when/moon, means/meaning), along with internal line rhymes and partial rhymes (again/when, each/means). There is a highly self-conscious use of reiteration which also of course functions as perfect rhyme: first/first, each/each, meaning/meaning, turning/turning, and of course many smaller words repeated to powerful anaphoric effect. The poem is filled with alliterations that suggest Anglo-Saxon strong stress metrics, in ways that Merwin explicitly identifies as measure in the poem: meal/moon/meaning, but also firmament/furrow/field, turning/telling and all the reiterated words as well. Merwin obviously knows the history of metrics stone cold and even the etymologies of terms such as “verse.” And, like all strong poets who have taken up the matter, and with greater insight than most, he is unequivocal about the meaningfulness of measure as an activity in and of itself both in life and in art. Note that in this poem measure is something we do, not merely the meaning of what we say: “we do it with the first breath we take / and the first sound we make” – not the first word we say, but even before that, in the first sound we make. Far from being merely a paean to the music of language, however, Merwin’s poem includes a great critical contribution when he identifies the significance of measure in language in the most spare, direct terms possible: “it means / what will come again.” Measuring with language, language as measure, a poetic function heightened and intensified by the making of verses, creates the meaningful recurrence and reiteration – “what will come again” – that is one of the essential dreams and foundations of all poetry. Indeed, in its emphasis on the meaningfulness of similarity, recurrence and reiteration, the poem reads like a far more sensuous and beautiful working out of Jakobson’s famous dictum that “The poetic function projects the principle of equivalence from the axis of selection into the axis of combination.” Notice that Merwin never says what measure may mean in and of itself in the particular words he uses: “It is there in meal and moon,” he writes, but meal and moon are not forced to mean it; rather they embody it. As he puts it, the measuring is the meaning.
And so we come to “the dark pool” of my title. There are several springs sacred to the muses in Greek mythology, notably two on Mt. Helicon, the Aganippe and the Hippocrene, the Pierian Spring in Macedonia near Mt. Olympus (from which, as Pope pointed out, it is a dangerous thing to drink a shallow draught), and the Castalian Spring, which is between the Phaedriades, 700 meter cliffs at the foot of Mt. Parnassus, near the home of the Delphic Oracle. According to legend and apparent practice, the Delphic oracle only spoke during the nine warmest months of the year, when Apollo’s spirit was present to inspire her; for the other three months, she fell silent and the sun god’s place was taken by his half-brother, Dionysus, whose tomb is supposedly housed there as well. To seize upon another problematic distinction, we might say that Apollo, a god of poetry but also of light, represents sweet clarity, and Dionysus, a god of poetry but also of wine, tragedy and perhaps inarticulate music, represents a kind of darkness, not an evil or even bitter or sour muse, but rather one that is more unruly. His pool is sweet, but it is dark. Indeed, if one tries to describe it in Apollo’s terms – terms of poetry but also of reason, philosophy, and rhetoric…it disappears. Strong poets know this and realize they must drink from both pools to make poems, for strong poems both say things and simultaneously do things. As critics, what we need to do is remember that we, too, must always honor the dark pool – the source of verse – and study how we might drink with the poets from its refreshing waters without then describing them as something that they are not. For writing in numbers may still be words, but if we treat those words as words alone, their full power to enchant will never reveal itself to us.
  Throughout this essay I use two sets of terms rather broadly to suggest an ancient dichotomy: on the one hand “meaning,” “content,” “semantics,” “sense” or “propositional sense,” “logical structure,” “argument” and “rhetoric” to denote what a poem says; on the other “structure,” “form,” “sound,” “versification,” “literary prosody” and “local texture” to denote what a poem does.
  As usual, the language is smarter than we are and suggests this distinction in the history of the words. The attested etymology of verse is the Latin versus, meaning, originally, the turn at the end of a plowed furrow; prose is a contraction of pro- +versus, i.e. “a turning forward,” or no turn at all: one endless furrow. Prose is verse that does not turn and thereby gives up what verse does. (It is also worth pointing out that the etymology suggests that verse came before prose and spawned it, which is probably true.)
  See also Gross’ reading of H. D.’s “Along the Yellow Sand,” where he argues that “[t]he subtle rhythmic shift in the last line emphasizes the suppressed excitement, the breath suddenly withdrawn, as the girls become aware of Lady Aphrodite’s physical beauty.” Gross confuses performance, versification, and the meanings of words.