In our own time, syllabic representations of animal calls are most familiar in birders’ field guides and in stories and verse for children.¹ Teaching a single iterable version of what each kind of animal says became a function of children’s literature in the first half of the nineteenth century, as did the syllabic representation of birdsong in field guides.² As we saw in Chapter 1, the earliest field guide to have used this device, Sarah Waring’s Minstrelsy, was published in 1832, and it combines the two genres. Waring explicitly addresses her work to young readers in its dedication and in an introductory poem.³ Minstrelsy’s aim is to teach its readers to identify the species of birds they encounter in the wild, and though she models her work on Bewick and Cuvier, Waring’s use of onomatopoeias as an aid to identification appears to be her own innovation and makes an intervention into what I will argue in this chapter is a Romantic-era debate about animal utterance, automatism, and species identity.

Besides Bewick and Cuvier, one of the work’s other acknowledged sources is Gilbert White’s Natural History of Selborne. White’s letters on the natural history of his parish in Hampshire were first published in 1788 (dated 1789), and by 1827, they had reached a third edition. Birdsong is one of this work’s recurring topics; for White, it forms part of his larger project of investigating “the life and conversation of animals” to supplement the “bare descriptions” of “foreign systematics” like Linnaeus.⁴ Unlike Waring, White does not transcribe bird calls; nor does he view them as a tool for identifying birds by species. He is rather interested in interpreting them, and his treatment of them as meaningful amounts to a sustained rebuke to the Cartesian and Lockean philosophical tradition of denying non-human animals the power to speak:

White’s extensive discussions of birdsong thus characteristically view them as communicative, and seek to explain their meaning and social function:

Even when Natural History of Selborne aims to represent bird calls rather than interpret them, it stresses their variability, and never makes recourse to syllabic transcription. When White gives measurements of birdsong’s pitch, he concludes that they vary within species:

In the nineteenth century, though, species identity becomes associated with forms of automatism like the syllabified call. The growing acceptance of onomatopoeia as a tool for identifying birds appears in Natural History of Selborne’s publication history; during the Victorian period, the work was repeatedly reissued, and in many of these reissues, Gilbert White’s descriptions of birds were supplemented with notes providing syllabic transcriptions of their calls.⁸

Waring’s Minstrelsy does not seem to have been widely read, so the emergence of this kind of transcription can’t be attributed to its influence. It does make clear, though, the technique’s source in poetry. As we have seen, the book is a compendium of poems about birds, including original species lyrics like “The Song of the Wood-Grouse,” in which a bird endowed with speech describes its own identifying traits, and uses onomatopoeia to quote a stereotyped version of its call. Another lyric on the same plan is given to the African Honey-Bird, whose refrain runs “Give heed to my note, so shrill and clear: / Come follow the honey-bird cheer, cheer, cheer.”⁹ The species lyric of the stormy petrel announces that it has no song to quote: “No song-note have we but a piping cry / That blends with the storm when the winds rise high.”¹⁰ The cuckoo’s lyric thematizes the incorporation of its song in human language, cheerfully terming it a mockery:

In syllabifying and transcribing bird calls, Minstrelsy adopts an approach to identifying birds by species that doesn’t have a precedent in earlier works of natural history. The precedents for such transcriptions are to be found elsewhere. Some are traditional and have come along with other onomatopoeias to form a regular part of English as such. The names of the owl and the cuckoo, for instance, both come into English from other languages, but are onomatopoeic in origin.¹² In the eighteenth century, the practice of giving birds names that transcribe their calls was adopted as a specific response to encounters with hitherto unknown fauna of the New World. The English name of the whippoorwill, a species of goatsucker, is first recorded in 1709 in John Lawson’s New Voyage to Carolina, which writes of the “Whippoo-Will, so nam’d, because it makes those Words exactly.”¹³ The author of Minstrelsy is well aware of the precedent for her project in the travel literature of the Americas, and includes her own chapter on the whippoorwill, which quotes at length from Charles Waterton’s Wanderings in South America, the North-West of the United States, and the Antilles (1825):

Waterton elsewhere makes explicit the identification between the nocturnal whippoorwill and the enslaved Indians and Africans of the Caribbean and the American South that overdetermined its English name:

This identification disappears from Minstrelsy, however, where the author’s own transcriptions of bird calls typically have no semantic content, and where the meanings of the names and transcriptions she takes over from elsewhere are normally ignored.¹⁶ Contrast the poem she writes for the “Willy-come-go” with Waterton’s account, where its call is said to speak for enslaved people’s ghosts:

Minstrelsy forgets Waterton’s identification of the bird and the Indian, and thus strips the social meaning Waterton had given them from its fear of humans and its call. Even where the poetry in Minstrelsy gives versions of bird calls that include English words, it tends to ignore or diminish their semantic content.

***

Minstrelsy is an early but by no means unique example of nonce transcriptions of birdsong in the nineteenth century. Not every such transcription was linked to species poetics. In a version of the of the Romantic aesthetic of particularity, syllabic transcriptions of a bird’s song could be used to distinguish the utterance of a particular bird at a given time and place from all others. The German forester and naturalist Johann Matthaus Bechstein made a transcript of a nightingale’s song in the late eighteenth century; according to a sceptical British writer in 1836, it begins “Tiuu tiuu tiuu – spe tiu zqua” and continues for thirteen lines of print, ending “gaigaigai gaigaigai–Quior ziozio pi.” This writer refers to Bechstein’s transcription as a “nursery ditty,”¹⁸ and argues that it is in general not “of much use to … describe the notes of birds in writing.”¹⁹ John Clare’s transcription of a nightingale’s song in his 1832 poem “The Progress of Ryhme” [sic] was not published in his lifetime; it fills nine lines of verse and makes an implicit rebuke to versions like “jug jug” that present themselves as normative or typical of the species. In his Why Birds Sing, the musicologist David Rothenberg praises the musicality of Clare’s transcription and connects his work and Bechstein’s to the Surrealist poetry of Kurt Schwitters.²⁰

These experiments notwithstanding, most syllabic transcriptions of bird calls in the nineteenth century represent what is purportedly the call or song of a species.²¹ In some cases, the transcriptions, like Waring’s, are made up of nonsense words. In 1837, for instance, an article in the first issue of The Naturalist describes the song of the sedge-reedling: “it is a strange medley … combining a characteristic chiddy, chiddy, chit, chit, chit, with a very sparrowlike chou, chou, chou, and an occasional peet-weet, reminding one forcibly of the chimney-sparrow.”²² In others, the renderings use words, but always with the implication that the bird’s call conveys a meaning of which it is unaware. Also in 1837, William MacGillivray writes of the chaffinch’s song that “the people in the south of Scotland most unpoetically imagine [it] to resemble the words ‘wee, wee, wee, wee drunken sowie,’ to which no doubt it bears some resemblance.”²³ Birders in North America seem especially given to this type of syllabification of birdsong: the well-known rendering of the white-throated sparrow’s song as “Old Sam Peabody, Peabody” is first found in print in 1871.²⁴ The New York state naturalist and writer John Burroughs wrote about his visits to Britain in 1889’s Fresh Fields, where he says of the song-thrush that “it is easy to translate its strain into various words” and fancies that to a boy out walking with a girl it might say, “kiss her, kiss her; do it, do it; be quick, be quick; stick her to it, stick her to it; that was neat, that was neat; that will do.”²⁵ For the yellowhammer’s song, he gives one of each kind of transcription: “Sip, sip, sip, see-e-e-e; or If, if, if you please-e-e-e.”²⁶

The lore of birdsong transcriptions is rationalized by Charles Louis Hett in his 1898 Dictionary of Bird Notes. This volume contains an alphabetical list of what it terms “spellings” of bird notes, keyed to the name of the appropriate birds, followed by a list of birds, each with its note. Verso pages are left blank for the book’s users to add their own notes. Hett frequently gives alternate “spellings” for a bird’s note – as in the entry on the chaffinch’s song: “‘tol-de-rol, lol, chickweedo’ [as we will see, this rendering had become conventional] or ‘tol, lol, lol, kiss me dear’ with occasional ‘wee, wee,’ or snore of a drunken man.”²⁷ The song of the male capercaillie is given as “peller-peller-peller”; of the female as “gock-gock-gock.”²⁸

Transcriptions of this kind were an important part of the culture of amateur naturalists in the nineteenth and twentieth centuries; they play little role in official science. They are also lacking from works catering to the large number of nineteenth-century fanciers of cage birds. Throughout the century, there was an extensive trade in what we would today think of as field birds, which were kept at home by fanciers. Species kept as pets or trained for competition included crows, ravens, many members of the families of larks, finches (including the canary), and thrushes (especially the nightingale), as well as exotic birds still widely domesticated today such as parrots. With the partial exception of corvids (who were sometimes taught to speak), all of these species were valued above all for their abilities as singers, and training these abilities was one of the fancier’s principal tasks.

As this implies, fanciers did not view bird species as limited to a single characteristic song. On the contrary, guides to the management of cage birds give extensive advice on how birds can be taught both music composed by humans and the songs of other bird species. Samuel Beeton thus describes how thrushes may be taught tunes played on the flute or tin whistle, or whistled by the mouth.²⁹ The bullfinch likewise, according to Bechstein, can “learn all kinds of songs, airs and melodies,” which they sing “in a soft, pure, round, flute-like note, which becomes the more agreeable if they are taught by means of a flute or of the mouth.”³⁰ The skylark too can be trained by whistling.³¹ Many species will also learn music from other birds. The linnet, Bechstein writes, will learn the songs of other birds like the nightingale, the lark, or the chaffinch if kept near it.³² The reed thrush’s song is greatly improved when it is “disciplined” by proximity to a nightingale.³³

The literature of bird fanciers is full of anecdotes about birds whose capacity to learn and reproduce the sounds they hear around them has humorous or uncanny consequences. Beeton relates how “once, on passing a bird fancier’s, my attention was arrested by a most extraordinary noise coming from a bird that was hung up for sale. It was a bullfinch, and as near as I can express it in writing this was his lay, ‘Chuff-chuff-chuff-chuff-creak-creak-chuff-chuff-chuff,’ and then a sort of ‘cluck, cluck,’ such as a nurse makes to a baby.” The seller cannot account for “this extraordinary music,” but on further inquiry Beeton finds that the bird’s previous owner had lodged at a tobacco manufacturer’s at Bermondsey and the bird’s

When writers like Beeton do give transcripts of a species’ characteristic song, it is so they can be trained up to sing it, rather than so they can be recognized in the wild. In a discussion of chaffinch singing matches, on which he tells us “much money is lost and won,” Beeton opens by noting that chaffinches of different counties sing different songs, distinguishing between those on the Essex and the Kent sides of the Thames estuary. The Essex song is “Toll-loll-loll-chickweedo,” and it is this song that the birds must sing when competing in London’s East End.³⁵ Beeton quotes an informant on the conditions of competition, which are “that the bird who delivered the greatest number of notes in fifteen minutes was to be declared the winner … A perfect note, according to bird-keeping phraseology, is a perfect and complete toll-loll-loll-chickweedo: if this phrase is a syllable short, it counts for nothing.”³⁶

For fanciers, syllabic transcriptions of birdsong serve the formation rather than the recognition of species identity. The malleability of species traits implicit in songbird fanciers’ practice of teaching birds of different species to sing one another’s songs is also apparent in their fondness for biological hybrids. Crosses of the canary with other species were a particular favourite; Beeton describes canary–goldfinch crosses as the most beautiful and the most likely to succeed,³⁷ but canaries were also crossed with linnets and other species of finches and with sparrows.³⁸ The connection between hybridity of song and biological hybridity is made explicit by a naturalist cited in Bechstein, who, holding the consensus view of naturalists in the period that birdsong is normally a means by which male birds attract a mate, suggests that though birds in confinement “will learn the songs of birds they are constantly kept with,” in the wild, “the peculiar note of each is an unerring mark for each to discover its own species. If a confined bird had learned the song of another … and was set at liberty, it is probable that it would never find a mate of its own species.”³⁹ By the mid-nineteenth century, birdsong and species identity are inextricably related – even for writers like Darwin, though he views the relation in a more complex way than Bechstein. In The Descent of Man, Darwin indeed represents song as a means by which male birds compete for mates – but recognizes that even in the wild, birds of different species imitate one another’s songs.⁴⁰ Moreover, he argues that cross-species sex results from the aesthetic preference of some female birds for masculine traits – including song – of other species, leading to comedies of mésalliance of the kind he relates involving a blackbird and a thrush, as well as different species of geese, ducks, grouse, and pheasants (Descent, 465–6).

***

Poets know the technique we have been discussing as onomatopoeia, and as we return to poetry, I will argue that a submerged debate about onomatopoeia carried out in poetry throughout the century has species identity and the relation between humans and other animals as its implied topic. In the remainder of this chapter, I will consider two different kinds of animal onomatopoeia in Romantic poetry. In the examples to be discussed, I will argue that onomatopoeia is treated as a poetic error, incorporating animal utterance in the poem only to reject it, or to make the animal itself an object of sacrifice.

Onomatopoeia’s foundational theorist was Quintilian, in whose late first-century Institutio Oratoria the term refers – as its etymology implies – to any new creation of a word. The examples Quintilian gives, however, are all of what would be termed onomatopoeias in our time, words coined to refer to a sound while also mimicking it: they are “mugitus, lowing, sibilus, a hiss, and murmur.”⁴¹ For the Elizabethan rhetorician George Puttenham, onomatopoeia is “the New namer”;⁴² Puttenham is explicit that the newly minted word should be “consonant” to the nature of its object, and gives a list of onomatopoeic representations of sounds as illustrations: “as the poet Virgil said of the sounding of a trumpet, ta-ra-tant, taratantara, or as we give special names to the voices of dombe beasts, as to say, a horse neigheth, a lion brayes, a swine grunts, a hen cackleth, a dogge howles, and a hundred mo.”⁴³

It is the trope of onomatopoeia in this sense that Minstrelsy adapts from poetry to the purposes of natural history. As Puttenham implies, animals whose calls are rendered by onomatopoeia are viewed as “dumb” – that is to say capable only of making sounds that can be named or mimicked, not of uttering speech that might be translated. In its use of onomatopoeia, as we have seen, Minstrelsy makes a contrast with one of its sources, White’s Natural History of Selborne, which without using onomatopoeias represents birds as conversing in an unknown language.⁴⁴

In what follows, I will view onomatopoeia as a trope that incorporates the utterances of non-human animals into language while also rendering them mute. Either by transcribing an animal call in nonce syllables – such as “he-de-he-de-he” – or by giving it a name, as in the examples from Quintilian and Puttenham, onomatopoeia incorporates animal sounds into human speech, while excluding any possibility that they might belong to language in their own right.⁴⁵ In this view, the long history of onomatopoeia in poetry and – more recently – in nature writing constitutes a counter-tradition to the equally long literary history of talking animals. In The Open, Giorgio Agamben avers that “up until the eighteenth century, language – which would become man’s identifying characteristic par excellence – jumps across orders and classes, for it is suspected that even birds can talk.”⁴⁶ As the passage from White I have just mentioned shows, this history is too simple. Romantic-era nature writing revives the idea that non-human animals have their own languages, and as we will see in the next chapter, the Darwinian critique of anthropocentrism rejected theories of language that exclude non-human animals. In what follows, I will argue that a revival of onomatopoeia in Romantic poetry stages the problem the animal–automaton poses for a poetry of nature.

Both the onomatopoeic animal and the talking one are represented in Romantic-era poetry, though animal speech is in fact surprisingly rare.⁴⁷ I begin, however, with texts from slightly later showing the contradiction between the two kinds of representation. Minstrelsy, for example, indulges a taxonomist’s dream by endowing birds with language so that they may describe in words their own wordless song. The contradiction between representations of non-human animals as speaking and as nonspeaking is still more acute in Lewis Carroll’s Alice books, where Jacques Derrida aligns it with what he takes to be the difference between poetic and philosophical views of the animal. At the end of Through the Looking Glass, when the newly wakened Alice asks her kitten whether it had a role in her dream, she gets no answer: it only purrs. “It is a very inconvenient habit of kittens … that, whatever you say to them, they always purr,” writes Carroll, and goes on to ask, “how can you talk with a person if they always say the same thing?” (344). Derrida identifies the prose realm to which Alice awakes at the end of her dream with philosophy: Carroll’s complaint about the kitten, he breezily notes, is “exactly like Descartes.”⁴⁸ In dreams and in poetry, though, Carroll and Derrida both imply, the animals can speak; for the latter, “thinking concerning the animal,” while excluded from philosophy, “derives from poetry [la pensée de l’animal, si il y’en a, revient à la poésie].”⁴⁹

The first animal to speak in Carroll’s Alice books is the White Rabbit – who says “Oh dear! Oh dear! I shall be too late” as he hurries by on his way to a rabbit hole. Alice does not think his utterance is anything “very remarkable”; only when the Rabbit takes “a watch out of his waist-coat pocket” and looks at it before hurrying on does she start to her feet and follow down the hole to Wonderland (Carroll, 24). The talking animal is one who doesn’t know the time, and we will see that this conjunction in Carroll obeys a logic previously established by poetry. To understand this logic, let me recall Descartes, who does not in fact quite say that an animal always says the same thing. Rather he argues that animal utterance is automatic, unlike that of human beings, which he takes to be the expression of rational thought:

The philosopher’s repeated analogy in his discussion of the animal’s automatism is between its action and that of a clock; here is the most extended version of this analogy:

Descartes’s examples of animal automatism – the swallows returning in spring, the bees’ labour, and the calls of cranes in flight – all refer to well-established poetic topoi.⁵² Far from differentiating philosophy from poetry, Descartes’s examples of animal behaviour have a specifically poetic origin.

Since ancient times, as well as transcribing animals’ calls and birdsongs, poets have observed the synchronization of their behaviour with the hour and the season. Such transcriptions are especially common in late medieval and early modern poetry – as in the Middle English round “Sumer is icumen in, / Lhude sing cuccu!” which re-entered the canon in 1801 with the second edition of Ellis’s Specimens of the Early English Poets.⁵³ In Elizabethan lyric, the cuckoo, the cock crowing at daybreak, and the nocturnal owl are stock figures. Two of them are paired at the end of Shakespeare’s Loves Labour’s Lost, in songs contrasting “Hiems, Winter” and “Ver, the Spring; the one side maintained by the owl, th’ other by the cuckoo” (5. 2. 891–3). This song is also an example of the characteristically Elizabethan device of making the formula for a bird’s song into a pun, by which it is arbitrarily endowed with meaning: The owl’s “tu-whit, tu who” thus sums up the play’s theme: to wit, to woo; while the cuckoo’s song hails the cuckold, becoming a “word of fear, / Unpleasing to a married ear!” (5. 2. 901–2).⁵⁴

This poetic tradition is eclipsed during the seventeenth and eighteenth centuries, but reappears in new contexts during the nineteenth. As we have seen, poetry for children takes on the function of teaching stereotyped versions of animal sounds. Works of natural history begin to use syllabic transcriptions, especially of bird calls, as a taxonomic device; this way of identifying birds presumes that, as Carroll said of kittens, they are animals who always say the same thing. The revival of lyrics that record bird calls and use animals to mark time was not limited to the nursery and the field guide. Other genres of poetry recover the convention to define themselves by contrast with animals’ supposed character as automata and mimics. The device of punning on the bird’s song and the convention of treating it as a natural clock or calendar both reappear as objects of parody in a poem from 1830 where Alfred Tennyson rewrites Shakespeare’s song from Love’s Labour’s Lost. Like its model, Tennyson’s song has two parts. In the first, the owl sits silent as the cock crows and the day begins; in the second, the poet recalls the owl’s nocturnal call, which neither part of the poem directly reproduces. The inability to do so is the second song’s topic:

Even as it echoes Shakespeare and repeats his pun, Tennyson’s poem disavows echo and breaks the identity between language and the cry of the owl on which the pun depends. “Thy tuwhit” becomes only the name of the owl’s call, rather than a representation or echo of it.

Slight as it is, this jokey piece of Tennysonian ephemera belongs at the centre of a certain nineteenth-century poetic canon. For in the form of a Shakespearian song it revisits the topic and some of the key terms of one of William Wordsworth’s best-known autobiographical poems, eventually incorporated into The Prelude, but published separately in the 1800 edition of Lyrical Ballads as “There was a boy.”⁵⁵ In this poem, Wordsworth describes how a child – a version of himself – in the Lake District, at night, by the water, “with fingers interwoven, both hands / Press’d closely palm to palm and to his mouth / Uplifted … / Blew mimic hootings to the silent owls / That they might answer him” (7–11). At first the owls answer his call, “With quivering peals, / And long halloos” which are “redoubled and redoubled” by echoes across the water (13–15). But sometimes pauses of silence “mock’d his skill” (17) – and, at those moments, the boy would experience an access of consciousness – “a gentle shock of mild surprise” that would carry awareness of himself and his surroundings deep into his heart (19). In this access of consciousness, there is also an intimation of mortality – for the poem goes on to tell us of the boy’s death “when he was ten years old” (32). As Paul de Man notes, the adult poet, who stands “mute” over the boy’s grave, is a double of the silent boy, waiting in silence by the lake for an answer from the owls.⁵⁶ The poem is thus legible as a parable of the poet’s coming into being through the contemplation of death, in a moment of rupture with the animal world.

By combining in a single text allusions to Shakespeare’s song and to Wordsworth’s narrative, Tennyson’s poem rewrites this parable and shifts its setting from the life of the individual to the history of poetry. In his version, when Shakespeare’s stereotyped “tu-whit, tu whoo” – which does not appear in “There was a boy” – fails to mimic the owl, the poem stages its own specifically linguistic character, recalling the Elizabethan identification of poetry and birdsong only to make a break with it. Further, by making this break in a mash-up of Shakespeare and Wordsworth, Tennyson implies that his poem’s relation to early modern lyric is not unique, but rather characteristic of his own period in literary history, with Wordsworth chosen as its exemplary figure.

Let me for a moment take Tennyson’s parody seriously as a guide to the relation between Wordsworthian lyric and its models. Although the model for Lyrical Ballads was ostensibly to be found in the printed and oral ballad traditions, there are few examples to be found there of birdsong. The nightingale’s jug-jug, the cuckoo’s song, and the owl’s tu-whit-tu-woo, all to be found in Wordsworth and Coleridge’s collection, come from early modern lyric, and there is evidence that in their dialogue as they assembled their collection, the two poets had these conventional representations of birdsong on their minds as a kind of running joke. In the spring of 1798, about a year before the likely date of “There was a boy,” Wordsworth wrote “The Idiot Boy,” a poem whose narrative opens with the “lonely shout, / Halloo! Halloo!” of the young owl (5–6); the owls’ calls accompany the entire story of Johnny’s midnight ride, whose happy conclusion the narrator introduces by summing up: “The owls have hooted all night long, / And with the owls began my song, / And with the owls must end” (444–6). The last word, however, belongs to Johnny himself, who tells what has happened to him in only two lines: “The cocks did crow tu-whoo, tu-whoo, / And the sun did shine so cold” (460–1). Johnny’s narration of his adventures both repeats and parodies the stereotyped representation of the owl’s cry, along with its traditional time-telling function; as he transposes the names of the cock and the owl, Johnny also reverses night and day. In this poem too, as in “There was a boy,” the celebration of poetic power accompanies a mimetic or nominative failure.⁵⁷ The same serious joke appears at the opening of Coleridge’s “Christabel,” written around the same time as “The Idiot Boy.”⁵⁸ This passage displays a mimetic excess that like Johnny’s narrative has the effect of prying apart the links between a bird, the stock representation of its call, and the time of day or night: “Tis the middle of the night by the castle clock, / And the owls have awakened the crowing cock; / Tu-whit! – Tu-whoo! / And hark, again! The crowing cock, / How drowsily it crew” (1–5).⁵⁹

A chain of allusions to cocks crowing at the wrong time stretches out from “Christabel” into later poetry. Walter Scott’s version of “Clerk Saunders” in The Minstrelsy of the Scottish Borders has the cocks crowing at midnight (stanza 21), in a line that Tennyson acknowledged as a source for the cock’s crowing “an hour ere light” in “Mariana” (27).⁶⁰ We also find other reworkings of the matter of “There was a Boy” contemporaneous with Tennyson’s. For instance a sonnet by John Clare, published in the Englishman’s Magazine in August 1831, also uses stock representations of bird calls to pry apart their attachment to a single type of bird or time or place. Here a boy mistakes the black-cap’s song for the nightingale’s, leading him to wonder if May has come in March:

***

For Keats, the connection of error with animals and with poetry is a recurrent theme. In a journal letter from the spring of 1819 he wonders how the “erroneous” reasonings of human beings would appear to a “superior being”: “though erroneous they may be fine.” This qualified fineness, he goes on, “is the very thing in which consists poetry; and if so it is not so fine a thing as philosophy – For the same reason that an eagle is not so fine a thing as a truth.”⁶² On one side of the binaries organizing this passage lie philosophy and truth; on the other are the half-truths of poetry and animals. In the spring odes of 1819, contemporaneous with this letter, Keats picks up the topic of the animal voice that sounds at the wrong time from the earlier romantics, though his more subjective poetics tends to transform the animal’s error into an error of its hearer.

Neither in the odes nor elsewhere in his poetry does Keats write out syllabic transcriptions of animal utterances, like those Wordsworth and Coleridge revive from Renaissance lyric. His work, however, is rich in onomatopoeia; as we will see two of the odes turn on canonical examples of the genre taken from Quintilian: “murmur” and “low.” In “To Autumn,” Keats develops techniques from late eighteenth-century works such as Collins’s “Ode to Evening” and Gray’s “Elegy.” Like the earlier Romantics, though, Keats adopts a critical relation to his models. Where for Wordsworth and Coleridge, animal voices are excluded from poetry, in Keats, they are incorporated under what we will see to be the sign of melancholy.

The “Ode to a Nightingale” has birdsong as its occasion, and the temporal dislocation the song induces is one of its major topics. The poem is framed by inextricable confusions of day with night and waking with sleeping. It is set in the springtime month of May, when the nightingale arrives, but the bird sings “of” summer (10), and its song precipitates in the poet a reverie on the coming summer flowers, concluding with the “musk rose, full of dewy wine, / The murmurous haunt of flies on summer eves” (49–50).⁶³ This line, unlike any of the references to the nightingale’s song itself, mimics the sound to which it refers. The poem’s recourse to onomatopoeia occurs at the moment where it represents the poet as most thoroughly transported by the nightingale’s song, a moment corresponding to the “false surmise” of grieving nature that Milton summons up and then rejects in the form of a flower catalogue in “Lycidas.”

Onomatopoeia marks an even more categorical poetic error in the “Ode on a Grecian Urn.” Since the ode is addressed to a silent artifact and takes silence in the form of negation – as in unheard melodies – as a major topic, the representation of a voice makes a major break in its decorum. As with the “Ode to a Nightingale” and its representation of summer, though, the dialectical logic of this poem eventually leads it to represent the very thing whose absence set it in motion. In the ode’s penultimate stanza, the images on the urn are given, by means of rhetorical questions, the very thing that up to this point they, like the urn itself, have lacked – an origin and a destination: “To what green altar … ?” (32); “What little town … ?” (35). At this moment, the poem’s silence is broken, and Keats sums up the illusions about the urn that the poem has entertained – the illusions of consciousness, of change, of speech – in the onomatopoeic utterance of an animal: a heifer, “lowing at the skies” (33).

Even more markedly than the murmur of flies in the “Ode to a Nightingale,” the heifer’s low breaks into the “Ode on a Grecian Urn” as a violation of its conventions and functions to signal a deepening of illusion, a moment of maximal enchantment that precedes the poem’s turn and its disenchanted close. In the last of Keats’s odes, a similar turn away from illusion also constitutes the main drama and, as in the earlier odes though in a more complex way, this turn is mediated by animal sounds.

Keats wrote “To Autumn” on September 19, 1819, and more than any of his other writing it has affinities with the tradition of closely observed writing on English natural history White inaugurated in Natural History of Selborne. The poem shows White’s influence and that of the column on the calendar and rural life that Keats’s friend and mentor Leigh Hunt published in his paper The Examiner.⁶⁴ A possible source for “To Autumn” is White’s attention to what he calls the language of birds and to its social function around the time of migration:

Keats’s swallows twittering in the skies as they gather to migrate echo White’s, as more generally does the way “To Autumn” in its last stanza gives its animals, birds, and insects utterances proper to the time and season when they occur.

The poem’s opening, however, represents animals that, like others we have seen in Romantic poetry, mistake the time. As in Keats’s other odes, animals here appear as projections of a poetic false surmise, though here indeed the error is not even expressed in sound, but only imputed by the poem. The first stanza of “To Autumn” consists of an incomplete sentence apostrophizing autumn as the “bosom-friend” of the sun. In spite of this apostrophe, though, the stanza’s main point is that autumn is difficult to recognize. She is veiled in mists, and even her sex is hard to determine: partnered with the presumably masculine sun, she appears to be a goddess of the harvest. But Keats’s term “friend” is non-committal. Even more crucially, in the exceptionally warm September of 1819, in the first stanza of this ode addressed to her, autumn does not appear as herself, but as summer. Together with the sun, she loads the trees with fruit and nuts, and sets “budding more, / And still more, later flowers for the bees, / Until they think warm days will never cease, / For summer has o’erbrimmed their clammy cells” (8–11). The bees’ deception is another one of the animals’ mistakes about time that I have argued open a space for poetry in Romantic and post-Romantic writing. Unlike other animals we have discussed, though, Keats’s bees are silent. Rather than making a sound at the wrong time, the bees are endowed with thoughts about the wrong season.

Like the murmur of flies in the “Ode to a Nightingale” and the heifer’s low in “Ode on a Grecian Urn,” neither of which properly belongs in the poem where it appears, the bees’ thought forms part of a fabric of illusion whose dissolution constitutes the poem’s action. The entire first stanza consists of misrecognitions, among which autumn’s appearance as an extended summer is only the most basic. Grammatically, the stanza is a sentence fragment that addresses autumn without telling or asking her anything, like an envelope without a letter inside. Together with the sun, she is personified as something like a divinity: the two of them collaborate, breathing together to shape the destinies of crops, animals, and human beings. Helen Vendler sees in them the hazy outlines of a “sky-god impregnating a goddess.”⁶⁶

But the poem’s address to autumn as a goddess is no more to be relied on than the season’s apparent offer of a perpetual supply of more and later flowers and fruits. In the second stanza, the poem’s work of demystification replaces the supposed divinity with representations of human beings. Now autumn appears in succession as a reaper, a thresher, a gleaner, and a cider-maker. Autumn in this stanza is not made by gods, but by human labour. Like the divinities of the first stanza, however, the labourers of the second don’t actually do anything. As generations of the poem’s critics have observed, the stanza represents a thresher who doesn’t thresh, a reaper who doesn’t reap, and a gleaner and a cider-maker who are caught, motionless, respectively in mid-stride and between the last drops of cider falling from the press.⁶⁷ The poem’s opening stanza, lacking a verb, shows autumn and the sun conspiring to bring about a fruit harvest that, in the stanza’s tenseless time, cannot occur. Though the second stanza has verbs, it nonetheless represents the harvest only under the sign of negation. In the motionless scenes the stanza shows us, the harvest is the thing that is not happening.

In the first two stanzas of “To Autumn,” the personifications of the season conspire to hide it by presenting an autumn of endless warm days from which labour and even movement are excluded. More accurately, the trouble with the scenes where autumn appears in these stanzas is less what they exclude than a kind of excess in her self-presencing. Bringing surfeit and over-fullness, she is not recognized as herself. Personified as a reaper, she is unable to reap because the late-flowering poppies overcome her with their scent. At the end of the second stanza, the cider-press crushes out the juice with which the fruit had been filled to bursting in the first, in a metapoetic moment that announces the poem’s will to wring out its own figural excesses and confront autumn in its truth.

Of these excesses, the most important is the personification of autumn itself. This figure organizes the poem’s first two stanzas and determines its genre as an ode. It provides the medium for the poem’s rejection of theology as it shifts from a representation of the season as made by gods to one in which it is made by human beings. But in the last stanza, to experience autumn as actually present – in the “now” first specified in line 31 – requires giving up the idea that the season is made at all, or that there is any agency responsible for it.

As it moves from its first to its second stanza, I have argued, the poem abandons its personification of autumn as a goddess and personifies her anew as a human being. In the course of the last stanza, though, it abandons personification altogether, and indeed more broadly the figure of metaphor that enables us to identify autumn as anything at all, whether human or divine. While the poem therefore does not represent autumn as animal, in its last stanza it does complete a trajectory begun with the shift from divine to human persons by representing an autumn scene populated only by non-human animals. These animals make sounds, breaking the silence that has prevailed in the poem until now, and replacing the static scenes which have hitherto been its object of representation with a soundscape constituted in time.

By making sound the medium in which autumn appears, Keats recognizes the season as an essentially temporal phenomenon. But why animal sounds? These animals don’t sing out of time to announce a supernatural visitation or as part of a poetic reverie. The animal sounds that suffuse Keats’s poem are of the Cartesian type: they mark time. They are all proper to a warm September evening in the south of England, and their conjunction in Keats’s soundscape endows it with a determinate location in time and space.⁶⁸ Keats’s animals function together as a clock, and he adopts the Cartesian model of animal behaviour precisely because, as we have seen, the main point of the poem’s final stanza is to produce a representation of autumn from which agency and thought have been emptied out. Nature here is observed with the eyes and ears of a natural historian, as the stanza suggests by its echoes of Gilbert White.

Though the sounds that the poem represents in its final stanza are all animal sounds, the nature in which these sounds occur is not one from which human beings are excluded. It is a nature that includes culture, though unlike the nature of the first two stanzas, it is not entirely given over to it. Human beings shape this landscape, raise animals and crops in it, and use them for food. This is the significance of the stubble-fields over which the sun sets. The flowers that overcame the reaper in stanza two are gone now, and the day’s labour has been completed. Nonetheless, autumn’s presence in this final stanza includes absence, for instance in the after-image of flowers in Keats’s description of the fields and clouds at sunset: “barred clouds bloom the soft-dying day / And touch the stubble-field with rosy hue” (25–6).

The synthesis of antithetical pairs, like nature/culture and presence/absence, is indeed the governing formal principle of this representation of autumn. Its paradigmatic image is that of the gnats rising and falling on the intermittent breeze, “borne aloft / Or sinking as the light wind lives or dies” (28–9). Not only life and death come together in this liminal moment: the birdsong that suffuses it represents the final conjuncture of two species that the advancing year will separate. The swallows overhead are gathering to migrate, but the robins in the garden will remain to endure the winter.⁶⁹ The evanescence of Keats’s autumn evening is the evanescence of the last moment in which these two songs will be heard together.

A similar but somewhat grimmer conjunction of opposites appears in Keats’s soundscape with the reference to “full-grown lambs” that “loud bleat from hilly bourn” (30). The phrase “full-grown lambs” has sometimes been read as an oddly periphrastic reference to sheep.⁷⁰ In fact, however, it does not properly refer to a living animal at all, but to meat – or, given the plural, to animals destined for slaughter. In the annual cycle of sheepherding, lambs are born in the spring. In flocks of dairy and wool sheep, which only require a small number of rams as sires, the superfluous males are normally slaughtered in the fall.⁷¹ Their meat was consumed as “full-grown lamb”; outside Keats’s poem, the phrase appears in the nineteenth century only in contexts informed by this practice.⁷² Its use here, in the plural, is a kind of catachresis, in which meat appears still on the hoof, at once bleating and bleeding.

By introducing into his representation of sheepherding this reference to its sacrificial logic, Keats differentiates his poem from pastoral, as he had also done by his references to reaping and to the grain harvest. I will return to the significance of this reference to sacrifice and to the consumption of meat; before doing so, let me note another instance in Keats’s autumnal soundscape of apparently posthumous animal life. In lines I have already cited, Keats begins his catalogue of autumn’s sounds by referring to the gnats among the willows at the water’s edge: “in a wailful choir the small gnats mourn / Among the river sallows” (27–8). Like “bleat,” “twitter,” “whistle,” and probably even “sing” later in the stanza, the word “mourn” is an onomatopoeia, even though unlike them it does not name the act of making a sound. Unlike these onomatopoeias, where the mimetic and semantic functions coincide, in “mourn,” they are split. Mimetically, the word represents the sound of gnats’ wings in flight; semantically, it refers to an affective state or psychic process. Even more acutely than other onomatopoeias, with its double reference, “mourn” poses the problem of how an animal noise – the sound of gnats’ wings – becomes a word.

I have read Keats’s poem as dramatizing the de-personification of autumn; as it unfolds, its figures of the season as produced by divine or human agency are replaced by the representation of nature as an automaton. In the context of this reading, it is impossible to take at face value the last stanza’s endowment of its gnats with the capacity to mourn proleptically their own impending deaths. Like most of the poem’s readers, I take the stanza’s description of the autumn landscape and its dwellers to express a consciousness of the passage of time in the observing poet. The “bloom” of the soft-dying day on the stubble-fields at sunset appears in this reading as a projection of the poet’s memory of actual blooms that had once been there. In this reading, the gnats that “mourn” at evening would be an example of what John Ruskin was later in the century to call pathetic fallacy. But regardless of where we locate the mourning consciousness in this line, pathetic fallacy cannot be the sole determinant of “mourn.” Rather, Keats’s poem asserts, the word also functions as a transcription of the actual sound of the gnats’ wings. Rising and falling with the wind rather than originating as a breath, the hum of the swarm in flight exemplifies the automatic, unconscious, and collective dimensions of animal life. But the word that represents it also signifies what philosophy for a long time took to be the most inward, profound, and characteristically human form of consciousness, the consciousness of death.⁷³ In poetry, as we saw, Wordsworth identifies this consciousness with a break between the human and the non-human animal. To interpret Keats’s poem, we should ask what principle links the two referents it attaches to the word “mourn.”

That principle appears later in the passage: it is the principle of sacrifice. The analogy the poem supplies for reading in “mourn” at once a sound and a state of consciousness is that of seeing full-grown lambs at once as live animals and as meat.⁷⁴ The gnats themselves appear in “To Autumn” to serve as swallow-food. In this poem, to make an animal’s life-process into a concept or into a word is to eat it. We know that it was written as Keats abandoned his ambition to write an epic on the Hyperion story. Some critics have seen it as a farewell to poetry in general, and Keats certainly wrote little or no verse after September 1819. If, as I have argued, “To Autumn” compares making poetry out of animals to eating them, it is a suggestive conjunction that on his return to London the month after the poem was written, he became a vegetarian.⁷⁵

***

In this chapter, my point of departure was the poetic basis of the philosophical and natural historical traditions in which the sounds animals make are understood as innate and automatic expressions of species identity. The first field guide to use onomatopoeias did so in verse. Descartes’s original examples of animal automatism – including the migration of swallows, the flight of cranes, and the social behaviour of bees – derive from poetry. The incorporation into philosophy and amateur naturalism of poetic versions of animal behaviour and animal calls, seems in the nineteenth century to have motivated a poetic counter-tradition, in which animals fail to behave like automata and poets insist that they can’t echo animal voices. This counter-tradition historically originated in the Romantic era, but it would be an error to identify it with Romanticism as such, which also includes much natural historical writing, including poems where the animals do not miss their proper time, and works, like many of Clare’s bird poems, that contain elaborate phonetic renderings of animal sounds. As we have seen, “To Autumn,” the last of Keats’s odes, ends by representing a chorus of animal sounds, all heard at their right time and season. I have argued that here, the incorporation of animal automatism in poetry, rather than being the object of demystifying critique or mockery, occasions melancholy. Keats identifies animals that tell time with animals killed for food.

Along with many other allusions to Romantic poetry, Carroll includes in the Alice books a recollection of the gnats of “To Autumn.” In Through the Looking Glass, as Alice travels by train towards the fourth square of the chessboard where the story unfolds, she becomes aware of “an extremely small voice, close to her ear” (Carroll, 222). This voice belongs to a gnat, whose predominant characteristic is that it is “very unhappy” (223), much of its speech being accompanied by sighing. This unhappiness receives no explanation in Carroll’s text; it is motivated by the recollection of Keats’s onomatopoeic form of the sound of gnats in flight: “mourn.”

Rather than explaining its unhappiness, the gnat’s discourse is largely concerned with the two topics at issue in the last stanza of Keats’s poem: naming and eating. In its survey of looking-glass insects the creatures the gnat describes turn out to be animated words – typically, words for food.⁷⁶ Alice is thus introduced to the snapdragon-fly – whose “body is made of plum pudding, its wings of holly-leaves,” and whose “head is a raisin burning in brandy” – and to the Bread-and-Butterfly – “whose wings are thin slices of Bread-and-butter, its body … a crust, and its head … a lump of sugar” (227). Whereas in Keats the animal’s becoming-language makes it into food, in Carroll’s looking-glass world the transformation of words into animals makes them inedible: foodstuffs become insects. And as is very often the case in the Alice books, looking-glass insects are also jokes about what should be eaten and what shouldn’t, and about who eats and who is eaten. As the novel vomits out the language of food in insect form, the words not only become inedible animals, but also starving ones: when she learns that the Bread-and-Butterfly requires weak tea with cream for food, Alice recognizes a difficulty:

A full account of the Alice books and their claims about poetry would need to discuss their representations of food and animals, as well as of language, and to explain the system by which Carroll’s narrative allows for any of these terms to be transformed into either of the others. Sometimes these transformations seem unremarkable – like the walrus and the carpenter’s consumption of live oysters – and sometimes they are very strange, like the lobster’s complaint that he has been baked too long. Such an account would bring into view a Victorian reading of Romanticism as a question of diet.⁷⁷ Alice – who very much resembles some protagonists of the Lyrical Ballads – needs to be instructed in what she can eat and when. The readings in this chapter have described two kinds of Romantic language. One, in Wordsworth and Coleridge, is characterized by the non-incorporation of animal voices. This language opens out an interior space to which the animal remains a stranger. The other, which I have identified with the onomatopoeic last stanza of Keats’s “To Autumn,” incorporates the voices of non-human animals into human utterance. In so doing, it opens out an interior space that is alimentary as well as psychic.⁷⁸