"Kubla Khan": Genesis of an Archetype

"Kubla Khan" has been so often cited as a dream specimen that a curious belief has grown up—accepted even in scholarly quarters—that the work now floats in a vast sea of psychoanalic interpretation. But in fact scarcely more than an occasional Freudian or Jungian spray ever reaches the poem. (Norman Fruman, Coleridge : The Damaged Angel, 394)

I

The circumstances of the composition of "Kubla Khan" (1797) are well known, at least if we trust Coleridge’s reminiscences, some nineteen years afterwards (1816), though many questions arise as to his trustworthiness.¹

     For the time being, however, it is simpler to retain the general picture drawn by the poet in his introductory account, for, in the end, all we have at our disposal is a text, a piece of discourse, and the question of knowing whether it is the result of a deliberate intention or not can in no way lead us to a fuller understanding of "Kubla Khan." The particular circumstances described by Coleridge evoke the idea of what takes place when a dreamer, on awakening, reports his or her dream. The 1816 annotation gives a good, although general, description of the process : while reading—and here we already have an example of what psychoanalysis calls day’s residue--, the poet fell asleep, and in a state of "profound" sleep composed, or rather reports he composed, from two to three hundred lines ("he could not have composed less than[. . .]"), nevertheless prudently adding: " if that indeed can be called composition." On awakening, he seems to remember those lines and starts putting them down on paper until, unfortunately, he is disturbed by a caller. When he resumes his task, an hour or so later, the "vision" has disappeared "with the exception of some eight or ten scattered lines and images[. . .]"²

     Can we take Coleridge at his word and accept that "Kubla Khan" is only a fragment of what could have been a much longer poem had he not been disturbed? I do not think so, and it is my opinion that we should simply take the said "fragment" as the complete narrative of the mentioned dream, and therefore as a complete poem; in other words, there is no reason to imagine it could have had a sequence. Indeed, dreaming one composes a complete, and generally splendid, piece of work is a well-known experience, a fantasy in fact which is never—at least to my knowledge—followed by actual realization. As it is, then, what we have is a poem written after a dream, remarkably, therefore, like the account of such a dream, that is to say of what remained of the "actual" dream that took place while the poet was asleep.

     I know that Coleridge insisted that what he dreamt was the poem itself, the text as we have it, with words, lines and stanzas complete, but however fascinating a remark this may be it needn’t have any bearing on the nature of our object, as I have just remarked above. Indeed, no one can tell what was really dreamt: the images of which the poem is made and which the poet would have put into words, or these words themselves, these lines even, organized as we can see them in the final text, complete with rimes, alliterations and assonances? But this, in the end, is of little importance to us since—as is the case with all dreams, let me repeat this—what is left of what the dreamer experienced during sleep is only what he or she remembers afterwards. In fact, our hesitations when it comes to defining our object are just an effect of the literary nature of "Kubla Khan" and nothing else. When confronted with a text like Coleridge’s poem, our doubts as to the very nature of the material under analysis form part of our own reading response to it, that is to say of our literary experience. From the world of dreams,³ the poem inherited a lack of coherence between its three parts⁴—or at least a coherence that is not easy to establish--, and some imprecision in the scenes described or in the characters mentioned: "pleasure-dome," "demon-lover," "the shadow of the dome of pleasure," "a vision," "a miracle of rare device." Naturally, these terms and images can be found in the works of others, Milton being the most prominent example, but they are nonetheless strongly reminiscent of what one may oneself have dreamt: "caverns measureless to man," "chasm," not to mention some of the adjectives: "sacred," "holy and enchanted." Thus, although "Kubla Khan" should not be taken for a piece of pre-surrealist automatic writing (for Coleridge dreamt, and wrote, a romantic poem), the enterprise resulted in the description, the evocation rather, of a universe that one might think closer to[. . .]"truth"—whatever that is--, a world, let us say, exceeding the limits of simple concrete perception, more authentic perhaps, beyond appearances certainly, and with no ordinary status. And from this point of view, I think we must admit the enterprise was successful.⁵

     Except that for the psychoanalytic critic, the mysterious, strange world referred to in the poem is not so mysterious and can lead us to a more precise, and satisfactory, description of the way humans function. I hope it will be understood that I make a distinction between the pleasure of the reader, of myself as a reader, and what is involved in the task of the analyst, of the scientist who questions our object: however connected, the two enterprises are definitely of a different nature; this is not always understood or taken into consideration; and yet, the same distinction can be made between what the dreamer experiences when he or she dreams and what is at work and is being experienced in the analysis of the dream later on.⁶

     It is this difference in nature between the two operations which always prompts me to insist that what the psychoanalytic critic unearths from texts in no way diminishes the pleasure of the reader, although this reassurance doesn’t generally suffice to convince "readers." All this to repeat that the ambiguous and difficult task of the analyst is to try and find out what unseen forces were at work in the production of a text, try to see—and hear might be more appropriate here—beyond what we can consider as the "first manifest meaning" of literary discourse. There shouldn’t be anything surprising in such an expression, but one may fear it equates dream and literature too easily and without further ado. Let’s us therefore for the time being consider it as our hypothesis (a well-known if not well accepted proposition, although the idea that literature has a symbolical dimension is always readily accepted). Quite psychoanalytically—and again there is nothing new in this--, our hypothesis distinguishes various levels of meaning in a text: what we first encounter on reading is akin to a surface, an architectonic layer, but also a screen (and this last point is what distinguishes psychoanalytic criticism from, let us say, standard literary criticism). This surface constitutes a necessary factor or component of literary pleasure but it should not be taken for the sole and ultimate source from which the text proceeds; for indeed for psychoanalysis, unconscious desire is the real "architect" of literary discourse, its prime promoter.

     Which amounts to saying that between dream and literature, and this means discourse, any discourse, there is a similarity of structure: first, an unconscious "core," and then, afterwards, the representation of it. This, naturally, corresponds to the structure of the dream described by Freud, but it may not be so obvious that it can also apply to literature. Here is what we have learnt from Freud, as is no doubt well known: the dream proper, what happens during sleep, must be carefully distinguished from the memory we have of such an unconscious experience. It may be that the original terms, Traumgedanken and manifester inhalt, were somewhat awkwardly chosen, particularly the first term, but the distinction is quite clear: what takes place during sleep, what "visits" the dreamer—the dream as we shall never know it⁷--, corresponds to the "dream-thoughts," Traumgedanken, while what follows, what is reminisced as best we can is labeled "dream-content," Inhalt or, to be quite specific, manifester Inhalt,( which reinforces the opposition with what was only "latent" and unknown before analysis). In the end, and Freud is adamant on this point, the account of a dream is not the dream, already it is a construction, the result of a first transformation. Perhaps it is now clear why I consider "Kubla Khan" to be such a perfect model for the demonstration that between dream and literature there is no difference from the point of view of structure. For whether there was something intentionally (consciously?) written or not in the lines we have, the outcome is remarkably like the "manifest content" (Freud’s meaning) of a dream which Coleridge may well have experienced. (I cannot at this point use the poet’s 1816 presentation to support my argument for it remains too ambiguous, possibly to preserve the idea that images and words were dreamt. The passage deserves to be mentioned, all the same, because of the clear distinction it makes between images and words: "[. . .]all the images rose up before [me] as things, with the parallel production of the corresponding expressions[. . .]" What of course the presentation does not say is that the words followed the images in time, which would have been a way to speak of latent and manifest contents.) The automatic aspect of the enterprise, then—of part of it at least—, resulted in a piece of discourse which simply rendered manifest what had visited the poet in his dream. It is a "content," let me repeat this, which is the result of the transformation of "latent" images and words, whatever their origin. This origin, we shall call unconscious core, and we shall only be able to make an hypothesis about it thanks to the traces left us by the poet in his narrative.

     The model does apply to any text, of course, but the particular circumstances of the composition of "Kubla Khan," that is to say its close relationship with a dream—irrespective of what that relationship was—, and perhaps the spontaneity with which it was put to paper, constitute a factor of great interest for the analyst. And at this point a clarification is needed: can indeed a piece of writing which did not proceed from a decision to compose (the poet didn’t seem to have known what he intended to "communicate"), a piece of writing which was as if directly dictated to the author by the dream he had just experienced, which arrived "all armed" with pleasure-dome, gardens, caverns, chasm and the Abyssinian maid be considered as a text like any other text?

     What I have so far emphasized is the structural similarity between dream and literature but I must be more specific. Let us examine the structure of the dream again:

What is illustrated here is the relationship between unconscious desire and its representation, a representation which is the result of a transformation which followed the path from a to b : this is quite precisely the structure of the metaphor. Freud called this transformation "the dream-work," Traumarbeit, insisting—and here lies the heart of one of his discoveries—that unconscious desire could not accede to consciousness without such a transformation, disguise in fact, screen or mask. Hence the apparently incomprehensible, strange aspect of our dreams; all this is well known. The formula I use to describe this movement of our psyche—which, it is important to note, is the fundamental structure of representation, and by this I mean language, the lacanian parole—insists on the fact that our dreams, and discourse generally, at the same time carry and disguise unconscious desire: la parole porte et masque à la fois; which is another way of repeating what I have just quoted from Freud: representation cannot come to our knowledge without a transformation. The Freudian bar naturally occupies the center of the structure with a "below" and an "above," or an "inside" and an "outside," a cause and an effect, and it is this which, by definition and quite naturally, we tend to forget, insisting on the autonomy of consciousness. At this point, obviously, Freud’s useful concept of "screen memory" comes to mind; but we are going to see that when it comes to literature there is a difference. The fundamental structure I am describing—an unconscious "core" and its transformation/representation--, we find in all texts, but between dream and literature here ends the analogy and we must, when considering literature, add a factor to the general picture described above. In short, with "literary" texts a further transformation takes place. With "literature," which we read and which we consider intelligible,⁸ the transformation—which can be represented as following a path of the same structural nature as in the dream, with a clear division between what is conscious and what is not--has another function to fulfill and can be said to amount to a second transformation. The function of this second moment in the transformation process consists in providing unconscious desire with a representation that can be accepted by consciousness: it thus acts as a screen—while still continuing to "carry" desire, to bear its mark "concealed"--, and its efficiency comes from the fact that it has received a second, specific, autonomous function which is enough in itself to justify its existence, a function of communication for instance, or of description or narration. Thus, one way or another, does literature become legible, more easily decipherable, we find, than our dreams: of a nature in the end that can be valued for itself. The place in the structure of this second transformation, which belongs to the domain of what I have called "communication," can be illustrated as follows:

Following a first transformation--I use "first" and "second" for the sake of a theoretical presentation, for the path followed by our second transformation may well be parallel to that of the other--, which is the result of what Freud calls Traumarbeit, dream-work, and which occurs every time language is produced ( the slip of the tongue has the structure of the dream and is for us a perfect example of this: it cannot be explained until it is decoded, the unconscious desire it expresses has already been submitted to the first transformation I am alluding to, parole vraie but nevertheless disguised ), following a first transformation, then, comes a second one which results from what we might call travail d’écriture, writing-work, whose function it is to render legible what is being produced, to give unconscious desire an appearance that can be, "in other words," accepted. That this "mask," as I pointed out above, may well have an interest in itself, there is no doubt, but on no account should this make us forget the other function of discourse. For I must not oversimplify our model for the sake of clarity, and when I speak of the two functions of language it must be understood that these two functions cannot be separated and form part of a single movement, since discourse, as we saw, at once carries and conceals unconscious desire. The simple and plain act by which I name a person or a thing, that is to say represent, replace a referent by a sign—and here I use the terms and the model of the linguist--, at once carries in itself my desire that the word should be the thing and implicitly refuses to acknowledge the distance between the two since the illusion that the distance can be bridged is necessary to the success of the operation of language. Language as a denial of absence, this is what, in the end, we should be able to read in "Kubla Khan." It is this double function, directly tied with unconscious desire, and at its service in fact, which must be acknowledged as the fundamental characteristic of discourse and of literature, desire being recognized as the prime cause, the actual architect of the whole edifice. For indeed, as the analysis of Coleridge’s poem will show, I hope, without unconscious desire—the expression being taken in its most general, structural, sense--, there would not be any language (while we couldn’t, on the other hand, speak of unconscious desire if there were not language, no representation).⁹

     One will have recognized in what precedes an allusion to the double nature of language: when we speak, or write, we communicate, but we also symbolize; in the words we pronounce or write there is more than meets the eye, or the ear. It is perhaps when looking at poetry—which never worried much about meaning or communication—that the model I am discussing can best be understood. In the case of poetry,¹⁰ the communication which can be valued in itself, and which I take to have also the function of a screen, is reduced to a minimum or does not exist, even. Which brings us back to "Kubla Khan."

     No mediation here, no screen between the dreamer and his product other than the first transformation to which all dreams are submitted, with the consequence that the reader will come to the conclusion that the poem is devoid of meaning in the ordinary sense of the word. No meaning, that is, other than what we have when confronted with the manifest content of a dream, pure poetry, it seems!

     What "Kubla Khan" tells us, indirectly, is that one should be very careful--in the case of literature at least--when speaking of communication, and this not only because the notion points to an ideal never reached, but because in Coleridge’s poem this dimension is lacking: such an absence should incite us to refrain from taking the said function of communication for the sole dimension of language, thus obliterating its symbolical dimension. In the model above, the Freudian model, things are quite clear: on one hand a function of communication and on the other the function of representation of unconscious desire. But the diagram is theoretical only, and words come to us in one piece. It is because the two functions mentioned are indiscernibly, indistinguishably fused into one that the analysis of the symbolical dimension of language and of literature remains so difficult. What psychoanalysis has taught us is that the natural attitude of the reader (dreamer)—only concerned with his or her hallucinatory pleasure—is antagonistic to that of the analyst (even though the latter will never attain any certainty and will have to accept to progress asymptotically only). This is to say that the analyst, or critic, cannot content himself or herself with remaining a reader only, however talented, and should be careful not to reduce our object to one of its parts only. To mistake what I have called the "mask," Freud’s "screen"—which in literature has its own importance, there is no doubt about this—for the whole of the work amounts to renouncing to a complete understanding of what was finally at stake when it was produced. For discourse is the product of desire: not only does it "carry" it, but is its cause. Meaning is desire. In the same way as in the dream, but with more complexity since a second transformation has taken place, literary discourse is produced by the individual subject for no other reason than to satisfy unconscious desire, that is to say represent—in hallucination-- the satisfaction of such desire.

     Does all this mean that I am not giving enough importance to the complexity mentioned above, and perhaps even overlooking it? Of course not. The double, ambiguous, nature of representation dispenses us from having to choose between one aspect of the literary object and the other, both should be taken into consideration. No one can deny that the reception of literature largely depends on what I have called the "second transformation," that part of the writing process, écriture, whose function is to render legible works of literature. But this is not a reason to remain blind to what happens when discourse is produced. In other terms, I am not forgetting the nature of this transformation, the physical aspect of literature, its "quality," its rhythms, sounds, musicality, and even the form and colors of its images, but we must also realize that it is precisely the other function of this literary quality to conceal the desire it carries.

     This is no doubt why "meaning" is such a difficult notion to handle, and at this point T.S.Eliot’s reply when answering a question about The Waste Land comes to mind: "It means what it says." Indeed, we should rather ask what it does to us as readers, and what it did—probably--to its author, rather than: "What does it mean?" Put in this manner, the question then becomes a question on affect and on desire, it is a direct question about what was very likely implied in the production of a text, it amounts to a question on representation.

     What remains, though, is that the whole process was started by a dream, and if "Kubla Khan" cannot be considered as a piece of automatic writing, or as the straight account of a dream, it still retains the traces of what brought about its production. What we are dealing with in this case is the articulation of the original (unconscious) dream-thought with secondary elaboration (and perhaps with what we can call a "literary" elaboration on secondary elaboration, since, to be specific, secondary elaboration is simply the immediate account of the dream).

     "Kubla Khan" is made of words, and while their "meaning" in the poem may be difficult to decipher, what they represent--because they bear the traces of unconscious desire--may not be so impossible to discover.

     That these words came from "somewhere," that they were first received by the poet, cannot be questioned: this is a common-place, the words we use were once given us. And at this point, history comes into play--as does intertextuality--, which however still permits me to continue using the Freudian model we have been working with so far. That desire, in dreams as in literary works, cannot be made manifest without the assistance of recent memories or impressions that will be "woven into its texture" simply testifies of the strength of the "censorship imposed by the resistance." (Freud, The Interpretation of Dreams, 563.)

     We know indeed that the material with which the dreamer constructs his or her dream is borrowed from waking life and most often comes from what took place during the day which preceded the occurrence of the dream. The term Freud uses is "day’s residues," and he insists that whereas these recent impressions form a necessary ingredient in the formation of dreams they should in no way be considered as the motive force which caused the dream:

Written over a hundred years ago, Freud’s monumental, groundbreaking study of dreams started a debate on the concept of "desire"—he mostly used Wunsch--that still goes on today. He himself went on and added notes and remarks to his original texts for several decades, sometimes completing previous comments. We may, now that the ground has been cleared and a fundamental structure set up, consider the incidence of unconscious desire on our lives with less hesitation and caution than he seemed to have done at first.¹¹ What I have called, I hope not imprudently, "Freud’s hesitation" can perhaps best be observed in a passage of The Interpretation of Dreams, pages 560-562, where a paragraph begins with: "I am now in a position to give a precise account of the part played in dreams by the unconscious wish. I am ready to admit that there is a whole class of dreams the instigation of which arises principally or even exclusively from the residues of day-time life [. . .] But the worry [over my friend’s health (which) had persisted from the previous day] alone could not have made the dream. The motive force which the dream required had to be provided by a wish; it was the business of the worry to get hold of a wish to act as the motive force for the dream."

     The whole long section in Chapter VII of The Interpretation of Dreams, "Wish-Fulfilment," which deals with the relationship between conscious waking life and the construction of dreams, remains the basis for any discussion on desire,¹² but the relationship between what is conscious and what is unconscious in us, the dynamics of "representation," is essentially the one we find in his model. What has changed in our views today, in mine in any case, is that the motive force required for the production of a dream is an unconscious wish and proceeds from it to residues of day-time life and not the contrary. What we must not forget is that both factors are needed in the end. The formula is quite simple: no representation without desire, but no desire—accessible to analysis, that is—without representation. Which brings us back to "Kubla Khan" and to the words which compose Coleridge’s poem. They were "received" by the poet, we noted--as all our words are--, but more specifically, here, they were prompted to him by what he had just been reading before falling asleep: Purchas’s Pilgrimage.

     In 1927, in his impressive and exhaustive study of the books Coleridge had read, John Livingston Lowes set out on a quest to understand the imaginative and poetic processes:

The Road to Xanadu provides us with an almost complete list of Coleridge’s readings and constitutes an indispensable source for any student of Coleridge’s poetry; Robert F. Flissner’s more recent Sources, Meaning, and Influences of Coleridge’s Kubla Khan, helpfully completes the list and is also a necessary tool.

     Lowes’s approach could seem Freudian—from the words to the dream--, but in considering that the images of the dream "‘flashed’ from words" (which, therefore, Coleridge had found elsewhere in the first place), he confused the material of which the dream was made with the motive force which brought it into existence and organized this material into its final shape.¹³

     Naturally, what started the dreaming process and, in the case of "Kubla Khan" the poetic process, can only be found in the words of which the poem is made—as in a dream, this is indeed our only material--, but we must then remember, as I have just pointed out, that these words have two sides and must not be taken at their apparent face value: the words of the poem, yes, but with the knowledge that there is more in them than meets the eye or ear in as much as they carry what with Freud we can call unconscious desire. Here is our architect. Admittedly, we could not speak of this "force" if it had not left traces of its existence in our dreams, and in discourse more generally, but we shall not mistake the material it chose to organize into a dream or poem with the "primary" cause at the origin of such a choice. True, Coleridge "tells us what was before his eyes at the instant he fell asleep, and the poem begins with the actual words on which his eyes had closed" (Lowes, 358), but this leaves open the question of the reason why the poet was so impressed by the words he had just read. What must have happened—as in the case of all dreams in fact--was that a given sign, image(s) or word(s), signified or signifier, was "chosen" to give expression to a particular desire. For indeed, and this perfectly applies to literature, what characterizes language is that the reason for its production can be read in it. We have already met with the "formula:" the mask, though acting as a screen, also carries unconscious desire: porte/masque.

     I realize that such an affirmation may seem to modify Freud’s presentation of the dynamics involved in the formation of a dream—since for him day’s residues can now and then be considered as having an active role to play at the outset --, but, as I have already tried to explain, the above formula does remain faithful to what is fundamental in the model he left us: the basic structure is the same, and this has a bearing on the way we can look at Lowes’s enterprise today. It may be that Freud, although he always insisted that the role played by day’s residues was secondary, gave them too much importance in spite of all, but his complete description of the forces at work in the construction of a dream amounts in the end to illustrating the encounter of an unconscious desire and of a sign. Several passages in The Interpretation of Dreams leave no doubt about this:

Another illustration by Freud of the relationship between desire and representation is the well known "analogy" which describes the relationship between the entrepreneur and the capitalist in the same section of The Interpretation of Dreams:

In my own formulation, Freud’s capitalist becomes the "architect," and I wouldn’t even say that his entrepreneur, from the point of view of psychoanalysis, has any initiative at all. "From the point of view of psychoanalysis" is a necessary restriction, for the simple reason that the sociological dimension of literature cannot be overlooked. In this perspective, day’s residues come to be valued for themselves and can be considered as the representatives of history. A research on the historical significance of "Kubla Khan" is not our debate here, but there is no doubt that such an undertaking would have to start with Lowes’s colossal enterprise. All that belongs to the romantic tradition in "Kubla Khan" is well known: this is where history played its part. Indeed, while carrying/concealing the desire of an individual subject, literature also bears witness of the time which has produced it: straight and easily legible or complex and metaphoric,¹⁵ it has its own sociological and historical meaning, a testimony which should not be confused with what discourse can tell us about the individual producer.

     And what we can learn about the poet as the individual producer of a text, but also about the imaginative process, is to be found in the very words the author "chose" to construct his work with. As I have tried to show throughout, it does not make much difference whether Coleridge consciously chose the words of his poem or simply wrote what the dream dictated to him, what he remembered from it, because in both cases these words had already been submitted to a transformation, a transformation which had taken place within him, the dreamer-poet. It is such a transformation—necessary to the construction of a dream--that Freud managed to analyze, to untie and translate, revealing what must have preceded not only the account of the dream but the dream itself and beyond this the motive force at the source of the whole process of representation.

     The psychoanalytic hypothesis—so often verified in each of us--is well known; I only mention it here to emphasize what relationship there exists between dream and language: just as our dreams, our discourse, in its ambiguity, retains the traces of the desire which caused it to appear. This leads to the question we can now put to "Kubla Khan" at last: "behind" the words, what desire? Can a hidden core of feelings and affects be unveiled which will give the poem the coherence it seems to lack?

II

The words,¹⁶ then, and to begin with, those which recur most often in Coleridge’s poetic narrative. And instead of wondering where they seem to come from, as reminiscences of what the poet had no doubt read, let us try to understand why they were selected among many others. I have in the preceding pages given enough reasons—convincing arguments, I hope—not to have to justify the psychoanalytic interpretation on which we are now more specifically about to embark. Should some extra reassurance be needed, however, none other than John Livingston Lowes would provide it:

As is well known by now, in "dream psychology" the theory is that the analyst listens to the words of the patient with what Freud called "suspended, or evenly poised, attention" (gleichschwebende Aufmerksamkeit),¹⁷ and then waits for whatever associations these signs may give rise to in the patient’s mind or memory. In the case of a text written several centuries ago, however, there is no patient on the couch, and the psychoanalytical critic is left with a piece of discourse which, as we saw, "means what it says" but no more. Unless, of course, some logic can be found, some pattern which will give a new coherence to the text or, as in the case of "Kubla Khan," replace an apparent lack of coherence, and help to account for the specific construction of the discourse under analysis.

     Hence, the words: from the words of the poem to their likely origin in the (opium) dream. The analyst, this time, shall endeavor to discover what "links" may have drawn "together a chaos of elements into" what became "Kubla Khan," (Lowes, 351)¹⁸ and try to establish what the poet’s dream could have been.

     A rapid word-count gives the following results:

If we now consider the narrative proper and try to listen to it as a "dream psychologist" should, we might get a hint as to what to do with the recurrences in the vocabulary just pointed out above, particularly as they occur at different times in the narration (dome and gardens, chasm, dome and shadow). This narration, which develops in time, can be said to have a dramatic structure, and this causes us to begin with a discussion of the way the poem is divided into parts or stanzas. For when we look at the presentation of the text on the printed or written page, we find out there are several versions of the poem, at least as far as the number of parts are concerned. The 1816 version, which is the one which first appeared in print, has three, whereas the Crewe manuscript, supposedly of 1797, has two, and we know there have been editors who divided the poem into four stanzas.

     The interest of the division into two parts, as in the Crewe manuscript, resides in the fact that it introduces a difference of status between the two parts, the first looking more like what "visited" the poet during his dream, while the rest, slightly shorter, could be taken for a conscious addition, though not really an interpretation. This indeed might well be the significance of such a division. But we could also have a "dream" in two parts, or two Acts, and then the more or less conscious addition. In fact, there is no way to tell and we must be content with a first long part—which I shall divide into two dramatic "moments"—and a sequel of twenty four lines which either formed part of the dream proper or were the result of an association to it, an afterthought directly related to the dream and therefore made of the same unconscious material. As for the degree of consciousness involved in the composition, the psychoanalytic critic needn’t take part in the quarrel since he is mainly concerned with the words of the poems and its general movement. The only thing we can be certain of is that there is an obvious difference between the first two "moments" of the narration and the last twenty four lines, if only because the subject of the narration appears in that third part, active, at least in his wishes or regrets, and no longer an observer only.

     If we examine the temporal structure, then, I take it we can distinguish three moments in the dream-poem, three tableaux representing the memory of what was dreamt or the associations it brought about, roughly: the palace and its gardens; the romantic chasm and the mighty fountain; the reconstruction, reproduction from another angle, of the dome of pleasure with the damsel and her dulcimer and "he" who has "drunk the milk of Paradise."

     Between these three main movements, themselves susceptible to be sub-divided, what relationship? If one accepts the hypothesis that a secret, invisible, thread runs through the various parts of which "Kubla Khan" is made, it can be the task of the analyst to discover what coherence unites these apparently disconnected images or tableaux, thus getting nearer to an understanding of the fantasy that was at the source of the actual dream, or even reverie, which in the end resulted in the written poem.

     It is not too difficult to identify the first eleven lines of "Kubla Khan" as a representation of paradise: a palace of pleasure is being built and magnificent gardens complete the picture; on this all critics agree, if only because the poem ends with the word "Paradise," a more than probable testimony of the influence of Milton’s Paradise Lost on Coleridge’s poem.¹⁹

     With the hope that the exotic figure of the oriental ruler can later find a suitable place in the overall structure of the poem, I shall for the moment leave the interpretation of the opening line, and start with a feature I find easy to analyze. Kubla Khan’s palace, then, but Coleridge wrote "dome," which is not so frequent to describe a palace or mansion, and then he added "pleasure" to it, which resulted in "pleasure dome." My own spontaneous reading, focussing on "dome" or taking "dome" and "pleasure" as if it constituted a single unit, a single sign, saw a woman’s breast in the image and it is true that the idea of curved lines conveyed here may lead to such an interpretation: a dome susceptible to give me pleasure. Such a reading is an obvious one and has been made by some critics (House, 118; Flissner,10), although others have deplored the simplicity of the interpretation (Schneider (1975), 9-10 ; Fruman, 396):