Dreaming, Vol. 7, No. 2, 1997

Coleridge, Wordplay, and Dream

L. R. Kennard., Ph.D.

University of Calgary
Calgary, Alberta



Wordplay is creative sign-making: it deploys patterns of identity or resemblance among signifiers, allowing linguistic form to create fresh, unique meanings. In broad agreement with a definition that includes but also exceeds punning, wordplay occupies a central position in Freudian dream theory, participating in the dream-work as a form of condensation. Coleridge, interested like Freud in both dreams and wordplay, directs his interest towards poetry and poetics. Analysis of two canonical poems of 1797-8, "Kubla Khan" and "Frost at Midnight," shows that in both wordplay flourishes within a context of reverie and dream. The evidence suggests that Coleridge, aware like Freud of the hermeneutic tradition which linked wordplay and dream, made use of the connection to produce highly-condensed works that satisfy his own poetic ideal of "untranslatableness." For Coleridge, dream-states legitimate wordplay because, for him as for Freud, wordplay is the language of dream.

KEYWORDS: poetics, wordplay , dream poetry, condensation

Coleridge, Wordplay, and Dream


In his theoretical model of dream development Freud emphasizes the connection between wordplay and dream. The model given in The Interpretation of Dreams makes wordplay a key component of condensation, the process of compression which occurs, as an essential part of the dream-work, in the transformation of latent dream thoughts to manifest dream content. In Freud's view,

[t]he work of condensation is seen at its clearest when it handles words and names. It is true in general that words are treated in dreams as though they were concrete things, and for that reason they are apt to be combined in just the same way as presentations of concrete things. (330)

Many of the examples of verbal condensation that Freud gives consist of charade-like combinations such as "Maistollmütz" (331), a word that, in Freud's reading, condenses "Mais" (maize), "toll" (mad), "mannstoll" (nymphomaniac) and "Olmütz" (a town in Moravia). However, while he emphasizes puns and punning combinations, Freud also points to the occurrence of a more distributed, or "expansively-condensed," kind of verbal play which involves repetition or resemblance rather than the pure condensation of the pun. He provides an example of such a "verbal bridge," as he calls it, when he cites a dream in which, in the course of a pilgrimage (pélerinage), the dreamer visits a chemist named Pelletier, who gives him a shovel (pelle) (91-2). And elsewhere Freud compares the verbal connections made in the course of the dream-work with alliteration (91-2) and poetic rhyme (375-6). As a result, "the whole domain of verbal wit is put at the disposal of the dream-work" (376).

Freudian wordplay then covers a range of devices, extending from puns (principally homophones), "charades" like Maistollmütz, antanaclasis (punning repetition of the same signifier), through to various forms of paronomastic play which depend upon resemblance, rather than identity, between signifiers. Paronomastic play consists of more than just paronomasia itself (close resemblance between two signifiers that suggests a punning relationship) in this definition; it also includes devices such as alliteration, consonance, assonance and rhyme. Conceptually there is a distinction between homophonic play, which depends upon phonetic identity, and paronomastic play, the play of resemblance. In practice, however, the two may be combined, as in many charades. Both kinds, we should note, forge a connection between sound and sense. Since tropes such as metaphor, simile and metonymy do not usually involve, or rely upon, such a connection, there is a conceptual difference between wordplay and tropological figures: in the first case phonetics and semantics are connected; in the second they are not. Once again, however, the two may be combined, as in a punning metaphor (in which the vehicle is remotivated so as to produce a pun). Wordplay, in the dual contexts of Freudian dream theory and poetic language, then represents more than just the play of quibbles, puns, and near-puns, while not being so broad a category as to be synonymous with all figurative language. The account given here emphasizes phonetic wordplay, as Freud does. However, Freud also mentions anagrams in dreams (335), and it is evident that a comprehensive definition should also encompass instances where, as in anagrams and other kinds of "letter-play," the connections among signifying elements are not purely phonetic.

The Freudian unconscious has a playful linguistic turn which manifests itself in dreams and jokes; the poet, with varying degrees of conscious intent, exploits a similar linguistic turn, using many of those same devices that, for Freud, characterize the dream-work and joke-work.1 The convergence suggests that the connection between wordplay and dream is part of a larger pattern of mental activity, in which associative "games of the underground," as Arthur Koestler calls them, both generate and participate in creative processes in a range of different fields. This is the position that Koestler takes in The Act of Creation: "[t]he capacity to regress, more or less at will, to the games of the underground, without losing contact with the surface, seems to be the essence of the poetic, and of any form of creativity" (319). Punning has the ability to forge "bisociative" links between seemingly unrelated images or ideas, links that are often prohibited or suppressed in rationally-controlled, "serious" discourse (65-6, 179, 317-19). So wordplay is creative sign-making - or rather, creative sign-making and sign-using, a kind of praxis: it can generate new meanings and give pleasure at the same time, thanks, in the Freudian account (1960, 19-47; 1965, 312-39), to its economy or condensation.

In the account that he gives in The Unconscious before Freud, L.L. Whyte mentions many writers who, implicitly or explicitly, referred to the unconscious, or to unconscious mental processes, prior to the twentieth century. Of all these writers, Coleridge was possibly the best equipped, not just to explore the connection between wordplay and dream, but also to point out and reflect upon the wider convergence which I have mentioned. He had a natural inclination towards the pun which found a ready expression in the jokes and conundrums with which he entertained correspondents in his letters. He combined this jocular practice with more serious enquiries into the subject; his "intended Essay in defence of Punning" (Notebooks 3: 3762), characteristically never written, was to have discussed wordplay in theoretical terms. On the other hand, his interest in dreams and dreaming was also wide-ranging, extending, somewhat analogously, from the personal and particular - the recording and interrogation of his own dreams - to the consideration of larger theoretical issues which, notably, included the relationship between dreaming and poetic creativity. Did Coleridge connect these interests in his own mind? And did he then ground the link between poetry and dream - a link that was almost a commonplace in his era - in language theory? The extensive record of his thought that we have in the form of conversation, lectures, letters, and other private and public writing supplies few answers. I will argue, however, that when we examine key early poems there is evidence that Coleridge did indeed connect wordplay and dream, and that his understanding of the connection furnished him with a poetic technique that remains a key to the acknowledged success of that poetry.

In making such a claim it is clearly relevant to ask how Coleridge might have conceptualized wordplay himself. The notebook entry of 1810 that mentions the intended essay on punning supplies a possible answer:

N.B. - In my intended Essay in defence of Punning . . .

to defend those turns of words,

che L'onda chiara

E L'ombra non men cara

in certain styles of writing, by proving that Language itself is formed upon associations of this kind, . . . that words are not mere symbols of things & thoughts, but themselves things - and that any harmony in the things symbolized will perforce be presented to us more easily as well as with additional beauty by a correspondent harmony of the Symbols with each other. (Notebooks 3: 3762)

The passage adumbrates a theory of linguistic harmony: harmonious sounds can be deployed to imitate harmonious "things" (as in Freudian dream, words have a material existence). So for Coleridge, one might suggest, wordplay has a quasi-religious function, figuring the recovery of an original, God-given unity which is lost in everyday language or sense-experience. In this "metaphysical" aesthetic, the poet uses "strange power of speech" to minister recuperative grace to a fallen world. It is also very pertinent that Coleridge intended to include a discussion of paronomastic wordplay (rather than homophones) within an essay "in defence of Punning." Evidently he understood the close relationship between puns and paronomastic devices, as Freud did later. Moreover, as James McKusick points out, Coleridge's defence of wordplay has other important bearings.2 It promotes a poetry that counters the doctrine of linguistic arbitrariness (which implies that there is no necessary connection between words and their meaning); and, because it makes sense depend partly upon sound, it passes what Coleridge described as "the infallible test of a blameless [poetic] style; its untranslatableness in words of the same language without injury to the meaning" (Biographia 2: 142). So the linguistic economy (I use the word in the double sense of frugality and system) of dream, which relies heavily upon wordplay, satisfies Coleridge's own criterion of poetic excellence. Plainly, it is one thing to point out this agreement, and quite another to suggest that Coleridge would have stressed it himself. However, it is intriguing to find such congruence between languages of dream and "blameless" poetry; perhaps Coleridge was struck by this congruence too.


The examination of the poetry which follows will focus primarily on two poems, "Kubla Khan" and "Frost at Midnight." Prior to this it will be useful to consider three key contexts: the tradition of dream interpretation that Coleridge inherits, the connection between wordplay and dream in Shakespeare, and, as a direct introduction to the poetry, the characteristic way in which Coleridge uses wordplay in letters written to friends.

Like Freud, Coleridge is heir to an older tradition of dream interpretation which involves wordplay. In a letter written to C.A. Tulk in 1824 he demonstrates his familiarity with this existing tradition. The postscript to the letter begins: "of the Writers who should be consulted previously to the Understanding Mr. Hartley's scheme I would suggest Achmetes, Artemidorus, and the Oneirocritici generally -" (Collected Letters 5: 326). In linking Hartley with the Oneirocritica of Artemidorus of Daldis, Coleridge relates eighteenth-century associationism to the notions of the most famous of the dream theorists of antiquity, while demonstrating that he is also familiar with the works of other theorists or dream-interpreters in the same ancient tradition. As Marjorie Garber notes in Dream in Shakespeare (6), Artemidorus' method of dream interpretation relies heavily on principles of association. Coleridge's reference to Hartley is evidence that he is aware of this fact, having presumably read Artemidorus, either in Latin or Greek, himself.3 He would then also have known that Artemidorus, like some other writers in this early tradition, emphasizes the role of wordplay and puns in dreams and their interpretation. Artemidorus relates one of the most famous examples of the use of wordplay in the ancient tradition of symbolic dream interpretation, when he refers to Alexander's dream before the siege of Tyros. Alexander dreamed of a satyr (in Greek: satyros). Breaking the word into its syllabic components, sa and tyros, it means "your" "tyre", or "tyre is thine", a prophecy (Garber 7). Wordplay functions much as it does in Freud's charades. As Douglas Wilson points out (2), Artemidorus' methods make the authority of dream problematical; does it rest with the dreamer, or is it shared between dreamer and interpreter?

Artemidorus draws upon an older tradition himself. The oldest known work on the interpretation of dreams, the Egyptian Dream Book, contains many examples of dreams interpreted through puns or near-puns. Seeing a penis stiff in a dream is a bad omen: it means victory to one's enemies. Entering a room with wet clothing is bad because it means fighting. Both these examples, given by Napthali Lewis in a list that contains others of the same type (Lewis 7-19), are dependent on wordplay. The words "stiff" and "victory" sound alike, and "wet" and "fighting" are also similar (101, n4, n6). These dreams, like Alexander's, are taken to be portents, and this connects them with another tradition that Coleridge was doubtless familiar with, that of oracular prediction. The priestess at Delphi, speaking on emerging from a trance, frequently gave her replies to enquirers in ambiguous, punning language. Domine, stes ("Lord, stay") might also mean Domi ne stes ("don't stay home") (Redfern 35). Again, ancient literature gives us, and both Coleridge and Freud, the association between wordplay and the unconscious mind. And in this example, we note, it also has the imaginative role (in Coleridgean terms) of uniting opposites.

If Coleridge knew of the ancient tradition connecting wordplay and dream from a familiarity with Artemidorus, he would surely have been equally aware that the connection exists in Shakespeare. In A Midsummer-Night's Dream, to cite a well-known example, Bottom plays upon his own name in trying to relate his "most rare vision": "man is but an ass, if he go about to expound this dream . . . it shall be called Bottom's dream, because it hath no bottom; . . . " (MND 4.1.213-14, 222-3). And Marjorie Garber directs us towards Coleridge's comments on Mercutio, who combines imaginative insight into the psychology of dreams with a lively facility for wordplay. Lecturing on Romeo and Juliet in 1812, Coleridge found Mercutio to be a kindred spirit: "a man possessing all the elements of a Poet: high fancy; rapid thoughts: the whole world was as it were subject to his law of association" (Lectures on Literature 1: 307). As Garber says of Mercutio:

His words have the authentic ring of the oracle who speaks in riddle to hide his meanings and to reveal them. Mercutio's language, in short, is in the finest sense poetical. At the same time, in its shifting patterns of association and double meaning, its reliance on puns and serious wit, it is the language of dream. (36)

Garber claims that the language of dream, which relies upon wordplay, is "in Shakespeare the essential fabric of the dream state; it is the paradigm of transformation, and transformation is at the heart of dream" (8). Coleridge, as I will show, uses the overdetermined ambiguity of homophonic wordplay to produce "Shakespearean" effects of transformation in "Kubla Khan."

Coleridge's early letters are important because they show a pattern of wordplay that relates both to the traditions of dream interpretation (ancient as well as Freudian) discussed above and, lending credence to my own reading, to some of the wordplay in his own poetry. In July 1796 he writes to John Estlin:

I would write Odes and Sonnets Morning & Evening - & metaphysicize at Noon - and of rainy days I would overwhelm you with an Avalanche of Puns and Conundrums loosened by sudden thaw from the Alps of my imagination. (Letters 1: 223)

Coleridge associates wordplay and poetry even as he keeps them apart - might the avalanches not occur on sunny as well as rainy days and then invade the odes and sonnets? Following the quoted statement he immediately gives an example which typifies the punning conundrums and jokes in his letters.4 He asks, "Why Satan sitting on a house-top would be like a decayed Merchant? - Answer. Because he would be imp-over-a-shed." Like both Freud's "Maistollmütz" and Alexander's satyros, the solution depends upon wordplay in which, as in the game of charades, individual syllables are treated as homophones. Puns are an occasional, rather than a persistent feature of the letters; when they occur they frequently do so in clusters, as in an 1821 letter to James Gillman in which Coleridge jests of "the punarhoea now on me" (Letters 5: 185). The punning metaphor is significant, suggesting as it does that Coleridge viewed these dis-easing attacks with a mixture of pride and disgust. Associative excess both attracted and repelled him, apparently.

In the late 1790's, immediately prior to the composition of "Kubla Khan" and "Frost at Midnight" (1797-8), Coleridge had a lively penchant for puns and charades and, quite probably, a knowledge of the traditional association between wordplay and dream. These two strands of evidence do not themselves confirm the hypothesis that Coleridge, aware of the connection between wordplay and dream, used it as a poetic technique. Rather, they render it plausible. To be able to say more than that we need to examine the poems.


In turning to "Kubla Khan" my intention is not to produce a new reading. Rather, I use a commonly accepted interpretive framework for the poem - it figures its own creation in a way that is synecdochic of the art of creative making in general5 - in order to investigate the function of wordplay and dream within it. I also assume, against the prefatory note of 1816, which I take to be ironic - that "Kubla Khan" was originally written as an artificial dream-poem, and make use of David Perkins' listing of the major characteristics that might have allowed contemporary readers to identify such a poem.6 A number of the elements that Perkins mentions - vivid imagery, exotic content, rapidity, economy, formal fragmentation, and hints of a deeper, unified meaning (Perkins 104) - are present in "Kubla Khan." Perkins makes no mention of wordplay, presumably because it is primarily in the oneirocritical, rather than the poetic, tradition of dream.

Both types of wordplay can, however, help to produce the characteristics of a dream-poem. Homophonic play effects verbal economy, and frequently hints at deeper meaning or unities; paronomastic play suggests a unifying resemblance which may be independent of, supportive of, or ironically opposed to, the primary sense of the words involved. While both types are present in "Kubla Khan," I pay more attention to homophonic devices, which have a stronger traditional connection with dream, than to paronomastic ones.

To begin with homophonic play, I examine four key words: measure, device, air, and Khan (or Kubla Khan).7 Considering them briefly in turn, I highlight their function in the context of a dream-poem. The first two stanzas play upon measure, contrasting it with measureless:

In Xanadu did Kubla Khan

A stately pleasure-dome decree:

Where Alph, the sacred river, ran

Through caverns measureless to man

Down to a sunless sea.

. . . . . . . . . . .

The shadow of the dome of pleasure

Floated midway on the waves;

Where was heard the mingled measure

From the fountain and the caves.

It was a miracle of rare device,

A sunny pleasure-dome with caves of ice. (1-5, 31-6)

Measure acts as a verbal junction, a condensed centre that radiates punning connections to a host of other words and phrases in the poem: the Khan's "decree," the precise, measured geometry of the paradise-garden (6-11), "dancing rocks" (23), and, not the least, the "mingled measure" of "Kubla Khan" itself, a poem of irregular, mixed, metrical form. Measure has the transforming power that Garber attributes to Shakespeare's language of dream, converting visual geometry into aural music, thus figuring the very see-change through which, in the composition of the dream-poem, images become words. And, set against the measureless caves of the underworld (4, 27), it has bearings that reach out to the fundamental underlying antithesis of the first two stanzas (lines 1-36), production by measured artifice versus the more uncontrolled, "measureless" energy of the natural world. In the context of dream the contrast converts into the antithesis between two kinds of composition: unconscious, non-volitional composition in dream itself, and the volitional creativity of the totally-conscious, wakened mind.8 By implication, the creative artist must be able to fuse the energies of both processes. So the artificial dream-poem is not so much a hierarchical composition, the conscious will totally controlling unconscious, associative impulse, as it is a balance of the two creative modes.

Device introduces the unifying conceit that concludes the second stanza:

It was a miracle of rare device,

A sunny pleasure-dome with caves of ice! (35-6)

Among the wide range of dictionary meanings that fit the context, three are sufficiently distinct to qualify device as a pun - physical shape, mental construct (as intent, trick or plan), and written device or conceit. The three may refer to the signifying triad of referent or thing, signifier (idea or concept), and signified (words themselves). Device can then figure the rapid compression of dream-composition itself, a collapse of signification in which external "things" metamorphose almost magically into ideas and language. In addition, as de-vice, a charade in the style of Maistollmütz, satyros, or impoverished, it hints at a homophonic link with the "sinuous rills" (8) of the opening stanza, as if (in further verbal play) the poem is another version of the Fall.9

The transition between the second and third stanzas has all the abruptness associated with a dream-narrative, yet there may be a latent connection, via the suppressed pun on mouth - the river dies at its mouth and is resuscitated, synaesthetically, as oral song. "I"/"he," the entranced (and entrancing) poet who speaks in the third stanza, refers to a prophetic dream or "vision":

A damsel with a dulcimer

In a vision once I saw.

It was an Abyssinian maid,

And on her dulcimer she played,

Singing of Mount Abora. (37-41)

The context of dream that he refers to then seems to trigger further wordplay, as he declares:

I would build that dome in air.

That sunny dome: Those caves of ice!

And all who heard should see them there,

And all should cry. Beware! Beware!

His flashing eyes, his floating hair! (46-50)

Air doubles as both atmosphere and melody or song. "And all who heard should see them there" echoes back to Bottom's synaesthetic account of his "most rare vision": "[t]he eye of man hath not heard, the ear of man hath not seen . . . what my dream was" (MND 4.1.211, 219-21). In keeping with the oracular tradition linking wordplay with visionary speech, it is quite proper that Coleridge's poet-prophet, like Bottom, should speak polysemously.

Kubla Khan condenses the poem, its chief protagonist, and, by projective identification, the poet of the third stanza. Within a poem that is subtitled "A Vision in a Dream. A Fragment" (in the 1834 version), Kubla Khan seems, via wordplay, to be a fragment of himself/itself. If "Huge fragments" (21) can exist within the fragment, what might constitute a whole? The fragmented form of the incomplete dream, foregrounded in the prefatory note to the 1816 text, thus makes the very notion of fragmentation problematical. To Estlin, Charles Lamb, and other members of Coleridge's inner circle, the first line, "In Xanadu did Kubla Khan," must have seemed like another homophonic conundrum: Question - Xan-a-du? Answer - Kubla Khan (as the rhyme-scheme implies, and as Coleridge may have done as a west-countryman, we may pronounce Khan as can). The line is oracular, providing a visionary or prophetic answer to the conditional "could I revive within me / Her symphony and song." (42-3) of the third stanza. The notion that "Kubla Khan" is an incomplete fragment or dream is already ironized, via some very Coleridgean wordplay that draws upon his own comic conundrums as well as the oracular and oneirocritical traditions of antiquity.

Wordplay, once recognised, encourages further wordplay. If device and Xanadu work as charades, might not the same hold for delight (44),10 Abora, and symphony (41, 43), given their visual/oral/aural context. And what of the conclusion?

For he on honey-dew hath fed

And drunk the milk of Paradise. (53-4)

Is not Paradise, when contrasted with the very visible materiality of the pleasure-garden of the first stanza, quite beyond ice, or eyes? Other words are not so much puns as punning metaphors, or condensed verbal nodes which generate punning connotations, in much the same way that measure does. Within the "deep romantic chasm" of the second stanza,

A might fountain momently was forced:

Amid whose swift half-intermitted burst

Huge fragments vaulted like rebounding hail, . . . (19-21)

Vaulted radiates geometrical-architectural connotations - subterranean, domal, paradisal - even as, via fragments, it may image its own unifying function, or the associative mental leaps made first by the dreamer and, subsequently, by the oneirocritical reader. As with other examples, vaulted achieves condensation, not so much, as many words and phrases do, by activating connections that reach outside the immediate, manifest content (connections which involve the vast intertextuality explored by J.L. Lowes and others)11 but by forging unifying links among the "fragmented" elements of the manifest content itself.

Paronomastic wordplay, in the wide sense defined earlier, works almost subliminally to realize similar unifying effects. Since these effects are very widespread, and have been noted by others, there is little need for detailed comment.12 However, examples include the repeated a-, ab-, re- and de- prefixes, the elaborate rhyme-scheme, the chiasmic pattern of alliteration, assonance, and consonance that marks the first line, and the prosodic and phonetic echoes that link Xanadu, Kubla Khan, pleasure-dome, honey-dew, and Paradise; these all contribute to a complex network of paronomastic interconnection that ironizes the apparent lack of unity characteristic of the dream-sequence, while it concurrently suggests the hypnotic power of a trance - inducing spell or charm. If the poet of the final stanza is "tranced," he would in turn trance his readers. Paronomastic effects can also support the homophonic wordplay already noted: the inverted resemblance between Xanadu and Kubla Khan makes the reading of the first line as a reversed question and answer more plausible, for example.

Through its use of wordplay "Kubla Khan" imitates the verbal strategies of dream, as the latter have been formulated in both Freud and the older tradition which he and Coleridge inherit. As Marjorie Garber notes (7), the traditional stress on the ambiguity of dreams tends to shift semantic authority away from the objective or manifest content of the dream itself, vesting it in subjective interpretation. Ambiguity results from compression, and keeps language alive and mobile, as it were, so that, as Freud noted of dream, meaning is never closed, and "it is impossible to determine the amount of condensation" (313). The sheer compression of the dream-poem draws us in as readers and makes us its co-creators. As a result, the poem becomes overdetermined and destabilized by signifying excess. In Kathleen Wheeler's reading, it is the prefatory note (or preface) of 1816 that destabilizes the poem: "the mixing up of Coleridge and his ironic persona in the preface breaks down the distinction between poet and reader" (30) and the notion of the text as a thing-in-itself independent of its reader becomes problematic as a result (29-30). But one can argue that "Kubla Khan"'s extraordinary compression and overdetermination has already achieved this goal, independently of the preface.

In a letter which Coleridge wrote in October 1797 he advised his friend John Thelwall, who had sent him a poem: "[a] little compression would make it a beautiful poem. Study compression" (Letters 1: 351). E.L. Griggs, in an editorial note, connects the letter's language with "Kubla Khan," and reckons that the initial composition of "Kubla Khan" took place close to the date of the letter.13 Perhaps Coleridge was particularly keen to achieve poetic compression himself at the time.


The reading given in the preceding section depends on, and may help to confirm, various assumptions: that "Kubla Khan" is an artificial dream-poem, that Coleridge was aware of the traditional connection between dream, prophecy and wordplay, and that the context of dream and trance gave him license, against the governing sense of literary decorum of the age, to use wordplay freely to achieve poetic compression. The assumptions, if accepted, give us one reason why no other poem of Coleridge's contains the same wealth of wordplay that "Kubla Khan" does - he never attempted to write a similar artificial dream-poem. While this is not to say that similar effects cannot be found elsewhere - Coleridge's predilection for the pun is evident in a number of other poems14 - it may be true that elsewhere he never exploits the connection between wordplay and dream so freely. Yet it is worth examining the connection in the second poem, "Frost at Midnight," written in 1798. "Frost at Midnight" invokes the context of dream more indirectly than "Kubla Khan" does. The poem begins by emphasizing the enigmatic stillness of condensation:

The Frost performs its secret ministry,

Unhelped by any wind. The owlet's cry

Came loud - and hark, again! loud as before.

The inmates of my cottage, all at rest,

Have left me to that solitude, which suits

Abstruser musings: save that at my side

My cradled infant slumbers peacefully. (1-7)

Paronainastic onomatopeia heightens the sense of secrecy, linking the spontaneous, self-acting frost with the abstruser musings of the speaker. The musings are obscure, but also profound; their half-open secret is, of course, that they nod towards poetic creation, now underway. Already, the focus has shifted to the speaker's state of mind. Carried forward by the governing motif of silence, the meditation moves from the external world, "[i]naudible as dreams!" (13), to the film that flutters on the grate,

Whose punny flaps and freaks the idling Spirit

By its own moods interprets, every where

Echo or mirror seeking of itself,

And makes a toy of Thought. (20-23)

Giving his mind the freedom to interpret playfully, the speaker abandons full volitional control of his own thought-processes. In this way, while the mind is not dreaming, it moves towards a state of consciousness not unlike that experienced in reverie or day-dream, one that may be characterized by idling association.

At the same time, the speaker's language continues to hint at wordplay. Tim Fulford points out the punning nuances of infant (7), signifying both "child" and, via the Latin, "speechless". He also notes that Spirit (20), via the Latin spiritus ("breeze", "breathing", and even "inspiration"), echoes back to the poem's opening, in which "[t]he Frost performs its secret ministry, / Unhelped by any wind."15 Spirit is also linked to the context of dream through the familiar meaning of "ghostly apparition." Alone with his slumbering child, the speaker's solitude is contrasted with the "numberless goings-on of life" outside his cottage. The goings-on are numerous, but also "unmusical," presaging (again by contrast) music to come.

The film on the grate, shifting casually to and fro, emblematizes the idling, reverie-like thoughts of the speaker who watches it. Again, it is appropriate that the film, which helps to make "a toy of Thought," should trigger wordplay. The film,

Still flutters there, the sole unquiet thing.

Methinks its motion in this hush of nature

Gives it dim sympathies with me who live,

Making it a companionable form,

Whose puny flaps and freaks the idling Spirit

By its own moods interprets, . . . (16-21)

Still ostensibly means "yet", but also flutters playfully between rest and motion, silence and sound. Sole means "only" but also "underlying," alluding to its own doubleness, while it also winks at Spirit. And puny, "undersized" and "petty," hints at the practice of wordplay itself, and the opprobrium conventionally attached to it.

The apostrophe at the opening of the second stanza marks an associative turn to the past:

But O! how oft,

How oft, at school, with most believing mind,

Presageful, have I gazed upon the bars,

To watch that fluttering stranger! and as oft

With unclosed lids, already had I dreamt

Of my sweet birth-place, and the old church-tower,

Whose bells, the poor man's only music, rang

From morn to evening, all the hot Fair-day,

So sweetly, that they stirred and haunted me

With a wild pleasure, falling on mine ear

Most like articulate sounds of things to come! (22-33)

An accompanying note inserted by Coleridge explains stranger: "In all parts of the kingdom these films are called strangers and supposed to portend the arrival of some absent friend." In Shakespeare's Hamlet there is an allusion to the same folk tradition, when Hamlet advises Horatio of the ghost, "[a]nd therefore as a stranger give it welcome" (1.4.165). Initially a metonym for the fluttering film, stranger later takes on the folk meaning of absent friend, as the speaker recalls the dreams - wishful day-dreams and real dreams - of his schooldays:

So gazed I, till the soothing things, I dreamt,

Lulled me to sleep, and sleep prolonged my dreams!

And so I brooded all the following morn,

Awed by the stern preceptor's face, mine eye

Fixed with mock study on my swimming book:

Save if the door opened, and I snatched

A hasty glance, and still my heart leapt up,

For still I hoped to see the stranger's face,

Townsman, or aunt, or sister more beloved,

My playmate when we both were clothed alike! (34-43)

Via the play between the metonymic and folk senses of the word - an example of antanaclasis - the stranger, as a portent, becomes the fulcrum on which the poem turns. It leads the speaker from a state of solipsistic isolation to one of community with other "strangers," from an idling, associative state of mind to one that is, in Coleridge's own sense of the word, fully imaginative. As in "Kubla Khan," wordplay appears in a context of dream and "presageful" (another charade?) omen. To brood (36) is to ponder or to sit hatching, in gestation. So a past act of thought becomes proleptic of future acts of creation, acts that might be poetic or, given the presence of one more stranger, his "cradled infant" (7), even procreative. "Therefore all seasons shall be sweet to thee, . . . " (65) the speaker closes, addressing his infant son,

Whether the eave-drops fall

Heard only in the trances of the blast,

Or if the secret ministry of frost

Shall hang them up in silent icicles,

Quietly shining to the quiet moon. (70-74)

The metaphorical play on trances (lapses, skips or dances, passages, states of trance or reverie) again invokes the twin contexts of dream and prophecy, as it harks back to the opening lines:

The frost performs it secret ministry,

Unhelped by any wind. (1-2)

Imaginative consciousness can have its origins in dreamlike, "idling" association, and has no need of external inspiration.

As in "Kubla Khan," paronomastic wordplay has a unifying role in "Frost at Midnight." The chiasmic return of "the secret ministry of frost" is part of a larger pattern of consonantal play, principally on the letters s, c, and t, which lends an underlying sense of sibilant/sybilline stealth to the whole poem. Somewhat overdeterminedly, these letters also reinscribe the initials of the author, Samuel Taylor Coleridge. 16 A fragmented man of letters, he is now variously disseminated within the poem - in the "secret ministry," the "owlet's cry," "articulate sounds of things to come," the "stern preceptor," "cloisters dim (52)," "trances of the blast," "silent icicles," and even in the infant who "sleepest cradled at my side" (44). This widespread letter-play is by no means gratuitons: if, as Richard Rand notes (310), it ironizes authorial intentions (how do we authenticate the true presence of a playfully distributed signature?) it also serves to link two kinds of consciousness - idling and imaginative; it implicitly insists that, while the two are different, they are surely (if secretly) related, and that the ambiguous, transforming ministry of frost presides over both.

Frost, in this reading, condenses a meteorological metaphor and a cryptic allusion to the author. But, as we know, frost crystals are already a form of condensation, so that the secret ministry may, among other things, be that of wordplay (or poetry) itself, in which a fixed or frozen pattern of signifiers can intimate "magical" metamorphoses, or generate "unfrozen" interpretive choices.17 Paradoxically, the stilled condensation of signifiers mobilizes and multiplies signifieds.

Metaphors of ice, frost, freezing, and condensation serve to link "Kubla Khan" and "Frost at Midnight" with "The Rime of the Ancient Manner." As Arden Reed has suggested, there is good reason to read "Rime" as a pun; the distinctive orthography triggers the double sense of "rhyme" or "hoar-frost" far more effectively than "rhyme" itself can do (Reed 149). The Mariner rhymes with great facility, but as one "[w]hose beard with age is hoar" (line 619) he is also rimed, or frosted with age. Rime, a homophone that refers to the paronomastic-play of rhyme, fuses the two kinds of wordplay; consequently, in Reed's words, the poem "combines both techniques of condensation and expansion" (158).

Can we then demonstrate a relationship between wordplay and dream in the "Ancient Mariner"? Two circumstances are well known: the poem originated, according to Wordsworth, in a dream of Coleridge's friend John Cruikshank,18 and when the poem was published in the 1800 edition of Lyrical Ballads, Coleridge changed the title from "The Rime of the Ancyent Marinere. In Seven Parts" to "The Ancient Mariner. A Poet's Reverie." Quite plausibly, J.L. Lowes suggested in The Road to Xanadu (280) that the alteration was a defensive move on Coleridge's part, a retreat from full authorial responsibility in the face of Wordsworth's criticism of the poem in the preface to the same edition. Alternatively, Patricia Adair argues that in calling the poem "a reverie" Coleridge probably meant "that the images and rhythm should be allowed to do their work at a partly subconscious, dream-like level; . . . " (58). To images and rhythm we should surely add the full range of paronomastic devices - rhyme, alliteration, assonance, consonance, and paronomasia itself, which are deployed so effectively in the poem. If we accept this extension of Adair's argument, then the "Ancient Mariner," like "Kubla Khan" and "Frost at Midnight," exploits the connection between wordplay and dream - or dream-like mental processes - for poetic effect.


After writing "Kubla Khan," "Frost at Midnight," and the "Ancient Mariner" in 1797-8, Coleridge continued to mention dreams and use wordplay, particularly in the wider, paronomastic sense of the term, in his poetry. "Christabel" (1798-1800), for example, contains a dream, a trance, and marked paronomastic effects. But as a generalization it is true to say that in succeeding years he never deployed either homophonic or paronomastic wordplay in so concentrated or effective a manner as he did earlier.19 The reasons for this lapse are no doubt bound up with the various factors that led towards, and accompanied, his decline as a poet. What is quite evident is that Coleridge's main focus shifted from the making of poetry to criticism and theory. Most important for present purposes, the literary lectures that he gave in the years from 1808 to 1819 contain a defence of Shakespeare's wordplay that is relevant both to his own poetry and, I will argue, to the connection between wordplay and dream.

In a notebook entry made in 1811, prior to delivering a lecture on Love's Labour's Lost, Coleridge discusses Shakespeare's wordplay:

Sometimes connecting [disparate] thoughts by means of resemblances in the words expressing them - a thing in character in lighter comedy especially that kind in which Shakespeare delights, the purposed display of Wit, but sometimes too disfiguring his graver scenes - but more often doubling the natural connection or order of [logical] consequence in the thoughts by introducing an artificial & sought for resemblance in words (as in the third line of L. L. Lost) -

And then grace us in the disgrace of Death: (Notebooks 3: 3113)

Two kinds of wordplay are distinguished: comic wordplay, in which the effect is almost entirely dependant on resemblance in sound, and a preferred serious type of play in which a resemblance or connection of ideas is quite deliberately and artificially doubled in language.20 While both types are "purposed" or "sought for," sound association tends to govern the association of ideas in the first type, whereas it does not do so in the second type. Perhaps the note comments indirectly on Coleridge's own use of poetic wordplay, on its "artificial" use in poems such as "Kubla Khan" and "Frost at Midnight." Structurally, the play on grace/disgrace is very similar to Coleridge's measure/measureless, and, had he chosen to, he could no doubt have defended his own figurative language in the same way that he defends the Shakespearean example. Yet the denigration of purely associative punning is telling. In a frequently quoted notebook entry of 1804, Coleridge had characterized poetry as "a rationalized dream," going on to ask, "What is the Lear, the Othello, but a divine Dream / All Shakespeare, & nothing Shakespeare" (Notebooks 2: 2086).21 Now, insofar as it employs wordplay, poetry is to be more logical or rational than dreamlike. Moreover the very notion of "doubling" seems to favour mimesis over creativity, just as the theory of linguistic harmony (discussed earlier) seems to do. Does language imitate preexisting relationships between things, or does it shape or create the very relationships it purports to re-present? Does the wordplay in "Kubla Khan" simply double certain preexisting ideas, or does it partly create them? The argument is pertinent whenever language becomes homophonic or paronomastic - in the rhetoric of everyday speech, advertising, Freudian dreams, and poetry. Consider an example from the "Ancient Mariner":

Instead of the cross, the Albatross

About my neck was hung. (141-2)

How can the full sense of the couplet, which nudges us towards metaphorical identity, be expressible without the fortuitous doubling of rhyme? Not to belabour the point, the untranslatableness of wordplay would seem to enable poetry to be particularly creative, giving it (as Koestler notes) the ability to forge fresh, distinctive ideas. But concepts of doubling and harmony downplay this aspect of figurative language in favour of mimesis. Wordplay is often thought to undermine assumptions of representational fidelity or linguistic transparency, but Coleridge's notion of "doubling" tends to support them, against the more creative alternative.22

Referring in the twelfth of the lectures given in 1811-1812 to Gaunt's deathbed punning on his own name in Richard II, Coleridge considers Gaunt's mental state. "Who knows", Coleridge argues (in John Payne Collier's report),

the state of deep passion must know that it approaches to that state of madness which is not frenzy or delirium, but which models all things to the one reigning idea: still to stray in complaining from the main subject of complaint and still to return to it again by a sort of irresistible impulse. (Lectures on Literature 1: 380)

For Coleridge, "the excess of fancy is delirium, of imagination mania."23 Gaunt's state of mind may be akin to mania, or excessive imagination. Wordplay, then, can be imaginative, but it is in this case an imaginative aberration, involving an irrational loss of control. Mania, one might think, differs significantly from dream, since the latter seem closer to delirium.24 However, Collier's shorthand notes, which must in general be faithful to Coleridge's own words (despite a tendency to elide in order to record), may alter our view of the matter. The shorthand account reads:

He who knows the state of deep passion by this must knows [sic] it approaches to that state of madness not frenzy or delirium but to trance [my italics] idea modelling all to rest still to depart from the main subject of grievance. (Lectures on Literature 2: 453)

According to this account Coleridge did not compare Gaunt's state of mind to mania but to trance - although the context of "deep passion" and the antithesis to delirium both continue to suggest mania. Among the relevant range of dictionary meanings of trance are: "a cataleptic or hypnotic condition," "absorbtion, exaltation, rapture, ecstasy" and "an intermediate state between sleeping and waking, half-conscious or half-awake condition" (Oxford English Dictionary) - the latter state being very much like that of reverie. There may be a distinction here between trance, which "models" association, via passion, to a ruling idea or motif, and reverie, a state of less-focussed, idling association. Taken together, though, they are clearly connected with the constructed authorial personae of "Kubla Khan," "Frost at Midnight," and the "Ancient Mariner." The Mariner (who later falls into a trance himself) attempts to "trance" the wedding guest by putting him into a state of fixed seizure when he "holds him with his glittering eye - " (13); the visionary poet of the final stanza of "Kubla Khan" is both tranced and trancing; and the speaking subject of "Frost at Midnight" emerges from a state akin to reverie. But trances (traditionally and in Coleridge's lecture) produce wordplay, which may be rime or frost. A complex web of association then links meteorological and linguistic condensation with the psychological condition of trance. But the primary message that trance bears is that Coleridge did indeed connect wordplay with dream, within a context that links both to the imagination: Gaunt's trance-like condition produces, and legitimates, his manic punning.

Collier's shorthand notes have more to reveal. As Coleridge continues to defend Shakespeare's wordplay the notes report:

We find these plays upon words in the best of ancients and the most beautiful parts of Sh[akespeare] . . . He has condensed [my italics] and has gained far more by the figures of speech than the moderns have by abandoning them. (2.453)

In Collier's finished version of the lecture the last sentence reads: "Shakespeare has gained more admiration by the use of speech in this way than all the moderns have acquired by abandoning them . . " (1: 381). In suppressing condensed Collier managed to rob the sentence of much of its metaphorical, condensed force, as well as its vital connection to Coleridge's own poetry. A condensed word that images and represents the act of condensing, condensed also figures, or reaches out to, all the other emblems of "stilled" meteorological/linguistic condensation in the poetry - frost, fog, mist, rime, icicles and so on. And this is done, as we have seen, within the psychological context of trance, or stilled seizure, and reverie. What is impressive about the Gaunt passage is not that it foreshadows a key aspect of Freudian dream theory; Freud is, after all, a very belated figure in the history of dream interpretation. Rather, we should appreciate how Coleridge, in defending Shakespeare, also defends and justifies his own poetic practice, albeit in a stealthy, displaced manner. The defence of Gaunt's punning relies upon psychologically-based notions of dramatic decorum. Similarly, Coleridge legitimates the use of dream-devices by creating an authorial persona, situated within the poem or its introductory apparatus, who either dreams the poem ("Kubla Khan") or creates it from a state of reverie or trance ("Kubla Khan," "Frost at Midnight," and the "Ancient Mariner").

The passages from the lectures show that Coleridge qualifies his defence of Shakespeare's wordplay with a distrust of purely associative punning, a distrust which goes hand in hand with a fear of letting associative play govern or shape thought, as stranger seems to do in "Frost at Midnight," Wordplay may be allowable in psychologically-abnormal states, but otherwise reason and conscious volition should be in charge. Ideally reason and "deep passion" might work together, but such harmony may be difficult to maintain. In 1803 Coleridge had related associative mental activity and dreams to the origin of evil:

I will at least make the attempt to explain to myself the Origin of moral Evil from the streamy nature of Association, which thinking - Reason, curbs & rudders / how this comes to be so difficult / Do not the bad Passions in Dreams throw light & show of proof upon this Hypothesis? - (Notebooks 1: 1770)

Given this fear of dreamlike association and the loss of volitional control that accompanies it, the fact that Coleridge never again exploited the relationship between wordplay and dream as he had done so successfully in 1797-8, despite the connection, via passion, with imaginative creation, is not so surprising. His increasing emphasis on individual volition, combined with a loss of passion itself (as suggested in the "Dejection" ode), may have been responsible.

If one accepts a dualistic model of mental processes, then conscious processes are primarily concerned with relationships at the level of the concept or signified, while unconscious ones, as Freud pointed out in much detail, often play associatively with signifiers, treating them "as though they were concrete things."25 In "Kubla Khan" Coleridge confounds this dualism, by collapsing referentiality and self-referentiality. Device refers, among other things, to its own status, as a conceit or device. Measures are decrees, measurements, dances, or patterned arrangements of words such as measure. Kubla Khan is a poem, a statement about the poem, and a protagonist within the poem. Signifiers refer to signifieds, but those refer back again to signifiers. Freud's parallel barriers between unconscious/conscious and signifier/signified, the sites of repression in The Interpretation of Dreams, are both breached in "Kubla Khan." Scandalously, the poem is "all Coleridge and nothing Coleridge." We shall never know just where the "conscious" and "unconscious" contributions to its composition meet, begin or end. Similarly, we can take device or measure to be referential, self-referential, or both at once, ambiguously and creatively.

For Coleridge, the conscious/unconscious balance was extraordinarily difficult to achieve or maintain. For the balance demanded that poetic composition become an epic journey in which the conscious, volitional mind must be prepared, figuratively speaking, to descend to the very depths of "streamy," unconscious association. Almost inevitably, the heroic struggle with associationism that figures so prominently in the early chapters of Biographia Literaria upset this balance, tilting it towards conscious volition, against dreamlike association and, concurrently, against wordplay. No wonder the defence of Shakespeare's wordplay is apologetic and defensive in tone. If Coleridge fails to examine the relationship between wordplay and dream in detail in his later lectures or prose writings, it may be because the same barriers of repression that Freud emphasized inhibited him from doing so. In the Biographia Literaria the poetic imagination is presented as a "magical power" (2: 16). Rather less magically, the imaginative poet can employ specific techniques to achieve unifying and creative effects. In Coleridge's own case one of the most efficacious of these techniques was the use of the associative, dreamlike devices of wordplay. In terms of authorial psychology, he seems to have felt that the use of such techniques demanded the fabrication of a poetic persona in, or emergent from, a state of reverie or dream. The context of dream then gave him full licence to produce poetry that satisfies two criteria of poetic excellence: his own test of untranslatableness, and the proposition, to which William Empson gives qualified assent in Seven Types of Ambiguity, that "all good poetry is supposed to be ambiguous" (xv). As a result (and to the degree that the two criteria are valid) the continuing canonicity of the poems that I have referred to may owe much to their exploitation of the connection between wordplay and dream. Wordplay is essentially creative: its secret ministry is not just to freeze, but to freeze and unfreeze language. Condensation makes us creative readers. Coleridge knew this, and refashioned the "games of the underground" as techniques of poetic compression. To bring together the various ludic techniques I have mentioned - imitative harmony, manic punning, charades, Coleridgean letter-play and other condensing devices - is to appreciate that as a group they make up little less than a "poetics of wordplay." As we know, Coleridge never articulated the theory behind this poetics in a comprehensive, fully-argued way. However, the poetics of wordplay survives. Licensed by the presiding contexts of dream, reverie, trance, and vision, it lives on as creative dream-language, in the highly-condensed poems of 1797-8 which I have examined.


Works Cited

Adair, Patricia. The Waking Dream. London: Arnold, 1967.

Attridge, Derek. Peculiar Language. London: Methuen, 1988.

Beer, John. "The Languages of Kubla Khan." Coleridge'sImagination. Ed. Richard Gravil et al. Cambridge: Cambridge UP, 1985. 220-262.

Coleridge, Samuel Taylor. Biographia Literaria. Ed. James Engell and Walter Jackson Bate. 2 vols. Princeton: Princeton UP, 1983.

- - - . Biographia Literaria. Ed. J. Shawcross. 2 vols. Oxford: Oxford UP, 1907.

- - - . Collected Letters. Ed. E.L. Griggs. 6 vols. Oxford: Clarendon P, 1956-71.

- - - . Lectures 1808-1819: On Literature. Ed. R.A. Foakes. 2 vols. Princeton: Princeton UP, 1987.

- - - . The Notebooks of Samuel Taylor Coleridge. Ed. Kathleen Coburn. 5 vols. Princeton: Princeton UP, 1957 - .

- - - . Poetical Works. Ed. E.H. Coleridge. Oxford: Oxford UP, 1912.

Culler, Jonathan. On Deconstruction. Ithica: Cornell UP, 1982.

Empson, William. Seven Types of Ambiguity. New York: New Directions, 1966.

Freud, Sigmund. The Interpretation of Dreams. Trans. James Strachey. New York: Avon, 1965.

- - - . Jokes and their Relation to the Unconscious. Trans. James Strachey. London: Routledge, 1960.

Fulford, Tim. Coleridge's Figurative Language. New York: St. Martin's P, 1991.

Garber, Marjorie. Dream in Shakespeare. New Haven: Yale UP, 1974.

Kennard, L.R. "Kubla Can: Wordplay in Coleridge's Poetry". The Wordsworth Circle. 26.1 (1995):8-12.

Koestler, Arthur. The Act of Creation. London: Pan Books, 1966.

Lewis, Napthali. Dreams and Portents. Toronto: Samuel Stevens, Hakkert, 1976.

Lowes, John Livingston. The Road to Xanadu. Boston: Houghton Mifflin, 1927.

McKusick, James C. Coleridge's Philosophy of Language. New Haven: Yale UP, 1986.

Perkins, David. "The Imaginative Vision of Kubla Khan: On Coleridge's Introductory Note." Coleridge, Keats and the Imagination. Eds. J. Robert Barth and John L. Mahoney. Columbia, Miss.: U of Missouri P, 1990. 97-108.

Rand, Richard A. "Geraldine". Untying the Text. Ed. Robert Young. Boston: Routledge, 1981. 280-316.

Reed, Arden. Romantic Weather. Hanover: UP of New England, 1983.

Redfern, Walter. Puns. Oxford: Blackwell, 1984.

Schneider, Elizabeth. Coleridge, Opium and Kubla Khan. New York: Octagon Books, 1966.

Smith, Eric. Some Versions of the Fall. Pittsburgh: U of Pittsburgh P, 1973.

Shakespeare, William. Complete Works. Ed. W.J. Craig. London: Oxford UP, 1943.

Stillinger, Jack. Coleridge and Textual Instability. New York: Oxford UP, 1994.

Wheeler, Kathleen. The Creative Mind in Coleridge's Poetry. Cambridge, Mass.: Harvard UP, 1981.

Whyte, L.L. The Unconscious Before Freud. New York: Anchor Books, 1962.

Wilson, Douglas B. The Romantic Dream: Wordsworth and the Poetics of the Unconscious. Lincoln: U of Nebraska P, 1993.



1 In Freud's account the joke-work and dream-work use wordplay in a very similar manner. See Jokes and the Relation to the Unconscious (36-9, 42-7, 169-70).

2 See the discussion of the relationship between wordplay, linguistic arbitrariness and "untranslatableness" in McKusick (32), to which I am indebted.

3 E.L. Griggs mentions a Latin version of Artemidorus, dated 1603, as a possible source for Coleridge's reference (Letters 5: 326 n3). For further discussion of Artemidorus, see Garber (6-8) and Wilson (1-4).

4 For further examples of wordplay in Coleridge's early correspondence, much of it similar to the example given, see the Letters 1: 134, 147, 182, 242, 262, 295, 351, 354, 406, 436, 536-7, 564.

5 For one recent account among the many that assume such an interpretive framework, see Wheeler, Chapter 1.

6 Both Perkins (106) and Wheeler (170 n10) suggest that "Kubla Khan" was composed as an artificial dream-poem.

7 For a somewhat different approach to the wordplay in "Kubla Khan," which emphasizes self-referentiality rather than the context of dream, see my article, "Kubla Can: Wordplay in Coleridge's Poetry."

8 Patricia Adair suggests that "the river flowing through Kubla's garden is consciousness, while 'the caverns measureless to man' are the unconscious" (118).

9 Eric Smith, in Some Versions of the Fall, cites "Kubla Khan" as a rendering of the Fall myth (144-8).

10 For a detailed explanation of the way that delight works as a charade, in the context of Coleridge's own ideas, see Kennard (10).

11 For these connections see particularly Lowes and Beer. Intertextual echoes are almost a kind of wordplay. As Beer notes, they act as a form of condensation, overdetermining the poem in the Freudian sense (253).

12 For a detailed examination of such paronomastic effects, see Elizabeth Schneider's Coleridge, Opium and Kubla Khan (262-277). Schneider's conjecture that Coleridge used "the 'technique' of the daydream" in "Kubla Khan" (90) accords with my own approach.

13 Among recent commentators both Wheeler (168 n2) and Stillinger (73) agree with Griggs in citing October 1797 as the most likely date of original composition, thus lending strength to my claim that the remark "[s]tudy compression" is directly relevant to the genesis of "Kubla Khan."

14 For instances of wordplay in a number of other poems besides those discussed here, see Kennard (8-9).

15 See Fulford (46-51).

16 Coleridge's play upon his own initials is discussed by Rand (310-12) and Fulford (28-32). Rand emphasizes the fact that such signing problematizes authority (310-12).

17 Arden Reed argues in Romantic Weather that "frost in 'Frost at Midnight' may not be simply a 'natural' or innocent image, but may . . . figure frozen language" (180). Wheeler, without mentioning condensation per se, connects the image of frost with the remark "study compression" (110).

18 Wordsworth's account of the origin of the "Ancient Mariner" is given in Lowes (203).

19 For two exceptions to the generalization, both of which involve the context of dream, see "Recollections of Love" (particularly the final two stanzas) and "A Day-dream," which plays upon the initials of family friends in the Coleridge/ Wordsworth circle with cryptic ingenuity. See Kennard (8-9).

20 For a more detailed (and less judgmental) discussion of the notebook entry, see McKusick (107-9).

21 Poetry a rationalized dream dealing [? about] to manifold Forms our own Feelings, that never perhaps were attached by us consciously to our own personal Selves. - What is the Lear, the Othello, but a divine Dream / all Shakespeare, & nothing Shakespeare. (Notebooks 2: 2086)

In the note Coleridge represents poetic creativity as an ideal balance: the plays are at once consciously composed ("all Shakespeare") and yet, paradoxically, given as in a dream ("nothing Shakespeare"). In the same note Coleridge complains that "the combining Power, the power to do, the manly effective Will, that is dead or slumbers most diseasedly - . . . " The gendering of will (versus passive association) suggests an incipient tendency to disturb the conscious/unconscious balance and make it hierarchical.

22 Jonathan Culler discusses the way in which wordplay threatens the ideal of linguistic transparency in On Deconstruction. The pun, he notes, is "a sin against reason itself . . . we treat the pun as a joke, lest signifiers infect thought" (91-2). See also the very lucid treatment of "the insubordination revealed by the pun" in Derek Attridge's Peculiar Language (188-195).

23 Engell and Bate discuss Coleridge's distinction between delirium and mania in a note to the Princeton Biographia Literaria (1: 84 n2). See also McKusick (103-4).

24 J. Shawcross points out the connection between fancy, delirium, and the state of dreaming in a note to his edition of Biographia Literaria (1: 225).

25 I have chosen to associate the parallelism between unconscious/conscious and signifier/signified with Freud rather than Jacques Lacan because, despite the fact that it is Lacan rather than Freud who discusses it in the terminology of structural linguistics, the parallelism is clearly indicated by Freud's emphasis on the unconscious origins of much wordplay.

  Copyright ©2003 Association for the Study of Dreams. All Rights Reserved