1

The Tetradic Circle

THE CREATION of every work of art necessarily involves four elements: the world, the artist, the work, and the audience or spectator. The world consists of both the natural world and the cultural world in which every individual lives, and although no two individuals perceive and experience exactly the same world, we may posit one “world” for all, for without a common world no communication is possible. The interrelations among the four elements involved may be illustrated by the following diagram:

Here, 1 represents “world,” 2 “artist,” 3 “work of art,” and 4 “audience” or “spectator.” The world naturally affects the artist, who reacts to it, and this self-world interaction constitutes this artist’s lived world or Lebenswelt. By exploring this lived world as well as other, possible worlds, the artist then creates an imaginary world in a work of art. When a spectator perceives the work of art, it will affect him in certain ways, and he will react to it in certain ways. Of course, he does not come to the work with a blank mind, but with all his previous experiences of the world and of art. After his experience of the work of art, his interaction with the world will be modified to some extent and in some ways. At the same time, by reacting to the work, the spectator comes into contact with the artist’s mind and recaptures the latter’s interaction with the world. That is why the arrows point in both directions in the diagram.¹

The same tetradic circle can apply to every speech act, which necessarily involves the world, the speaker, language, and the hearer, for whenever something is said, the following conditions must be present: what is said must be about something (and the world is the sum total of “somethings” about each of which something can be said); there must be someone saying it; there must be some language to say it in; and there must be someone hearing it, even if it is the speaker himself, who cannot help hearing what he is saying. If we substitute “speaker” for “artist” at position 2, “language” for “work” at position 3, and “hearer” for “audience” at position 4, then the diagram will illustrate the situation of any speech act.

It should be obvious that the diagram will also apply to poetry, since poetry is both an art and a kind of speech act. The assumption that poetry is an art does not seem to need any justification. As for poetry as a speech act, I disagree with those who consider poetry to be an imitation of a speech act² or a quasi-speech act (shades of Aristotle!): for what basic difference is there between saying “I love you” orally and writing a love poem, or between writing a love poem to an actual person and writing one to an imaginary lover? Are Elizabeth Barrett’s Sonnets from the Portuguese a more genuine speech act and therefore less poetic than Emily Brontë’s “Cold in the earth—and the deep snow piled above thee” because the former were actually addressed to Robert Browning and the latter was addressed to a product of childhood fantasy? Nor do I find it particularly helpful to apply the Austinian distinctions between locutionary, illocutionary, and perlocutionary acts to poetry.³ For one thing, it is often difficult to say what kind of illocutionary act a poet is performing in a given poem. For example, when Wang Wei (701-761?) writes,

what kind of illocutionary act is he performing? We could, I suppose, say that he is “describing,” but for whose benefit, his own or ours? If it is for his own benefit, why should he find it necessary to tell himself in words, and in an antithetical couplet to boot, what he must have known perfectly well he was perceiving? If it is for our benefit, can we be sure that he had an audience in mind when he wrote the lines? In any case, to say that a poet is “describing” or “representing” the world or that he is “expressing an emotion” is not to say very much about a poem. As for any perlocutionary effects a poem may have, these are generally irrelevant to its artistic value. For my own part, rather than asking what kinds of illocutionary and perlocutionary acts poetry may perform, I would simply observe that poetry differs from other speech acts in function but shares with them the same linguistic structure (although the linguistic structure of every individual poem is unique in the sense of not being identical with that of any other poem), while it differs from other arts in structure, its medium being language, but shares with them the same basic function. In other words, poetry may be conceived of as the overlapping of linguistic structure and artistic function. This conception of poetry does not oblige us to maintain the distinction between “poetic language” and “ordinary language,” a distinction that has recently been shown to be untenable,⁵ nor does it force us to the unhappy conclusion that poetry does not exist.⁶ According to this conception, “poetic language” is not a special kind of language with identifiable features, but simply language that fulfills the artistic function. (What this is will be discussed below.) At the same time, the fact that there has never been a perfect and universally accepted definition of “poetry” does not mean that poetry does not exist, for the same is true of many other words, such as “religion” and “cat.”⁷ What I am attempting is not a perfect definition of “poetry” but a heuristic hypothesis that may help us turn the hermeneutical circle into an open-ended spiral of infinite reinterpretations and revaluations. I shall elaborate on this point in chapter 4.

In saying that poetry is the overlapping of linguistic structure and artistic function, I do not imply that it cannot fulfil other (i.e., non-artistic) functions, such as moral, social, and political ones, which the author may have consciously intended his work to fulfil. However, these are not what make it a work of art, just as Shang bronzes or Greek vases may have fulfilled utilitarian functions as containers, but these are not what make them works of art. Conversely, a poet may not have been conscious of the artistic function of his work, but this should not prevent us from discussing it.

What, then, is the primary artistic function of poetry? This I conceive to be twofold: extension of reality through the creation (on the part of the poet) and re-creation (on the part of the reader) of imaginary worlds, and satisfaction of the creative impulse of both poet and reader. To elaborate this concept of the artistic function of poetry, we can now examine it in the framework of the tetradic circle, placing “poet” (or “author” or “writer”) at position 2, “poem” (or “work”) at position 3, and “reader” at position 4. The diagram will illustrate how the poet creates an imaginary world out of his own Lebenswelt (which is constituted by his interaction with the “real” world) as well as other, possible worlds, and how the reader re-creates the world of the poem and thus recaptures the poet’s interaction with the world. It should be realized that the world of the poem, which does not exist and has never existed in the real world, exists first in the poet’s consciousness and, once created, exists potentially outside time and space, to be re-created by the reader’s consciousness. Therefore, the world that emerges from the linguistic structure of the poem is not the same as the author’s Lebenswelt, which merely provides the occasion for the creation of a work:⁸ it is his Lebenswelt fused with and transmuted by the creative experience. Whether the author’s Lebenswelt is knowable to us is open to question, but we may assume its existence, just as we may assume the existence of another person’s “being,” even if it can never be fully known to us. In any case, if an author’s Lebenswelt is indeed unknowable, it will render impossible the task of the biographer, not that of the critic, whose concern is with the created world of the work.

Since the world of the poem has never existed in the real world, it is an extension of reality. This statement differs from both Wang Guowei’s and my own earlier opinions. Wang wrote, “There are some [poets] who create worlds [jingjie] and others who describe worlds. This is the origin of the distinction between Idealism and Realism.”⁹ I disagreed with this and suggested that it was not so much a distinction between Idealism and Realism as one between great and lesser poets.¹⁰ However, I now believe that every poem that is a genuine work of art has its own created world, which is an extension of reality, just as every painting that is a genuine work of art has its own created world. Even an apple in a trompe l’oeil painting exists only in the imaginary world created by the artist, with its own dimensions of space and time, and not in the real world. It therefore constitutes an extension of the reality of apples or “appleness.” Of course, the existence of the apple as part of the imaginary world of the painting is not to be confused with the physical existence of the paint on the canvas, any more than the existence of a tree or a person in a poem is to be confused with the physical existence of the print on the paper in the real world.

The concept of the created world of the literary work of art is nothing new. As M. H. Abrams has demonstrated, the concept of the poem as heterocosm originated in the eighteenth century and has become a commonplace in Western criticism.¹¹ However, what is being proposed here does not imply, as did the Romantic concept of the heterocosm, an analogy between God the Creator and the artist as creator. Rather, the word “created” is used in contradistinction to “made,” the difference between the two having been clearly shown by R. G. Collingwood.¹² Thus, to speak of the poet as a creator does not mean “raising the function of art to the level of something divine or making the artist into a kind of God.”¹³ It is worthwhile comparing our concept of the created world with Roman Ingarden’s “represented” or “portrayed” world (dargestellte Welt) and Mikel Dufrenne’s concept of the world of the aesthetic object, as well as with some Chinese antecedents.

Ingarden’s description of the world of the literary work of art at times betrays traces of the mimetic concept of art, such as when he speaks of the “representation function” (Repräsentationfunktion) of the “represented” or “portrayed” objects (dargestellten Gegenstande), or when he remarks, “persons ‘appearing’ in literary works do not merely carry such names as ‘Julius Caesar’, ‘Wallenstein’, ‘Richard II’, etc., but are also in a sense supposed to ‘be’ the persons who were once so called and who once actually existed.”¹⁴ To me, however, whether a character in a literary work is an authentic historical figure like Julius Caesar, or a shadowy semihistorical figure like Macbeth, or a purely fictitious character like David Copperfield, or even a fabulous creature like Ariel, is irrelevant to its artistic nature and value, just as whether the subject of a painting is an authentic historical figure like Henry VIII, or a shadowy semihistorical figure like Homer, or a purely fictitious character like Adonis, or even a fabulous creature like the dragon killed by St. George, is irrelevant to its artistic nature and value. I do not deny that our previous knowledge about a historical person will affect our response to the work in which he appears, but I think such knowledge is not indispensable. Even someone who has never heard of Julius Caesar will not fail to see the historical significance of his assassination from Shakespeare’s play itself, and presumably Marlowe’s audience had no knowledge of the historical Timur, nor Voltaire’s of the historical “orphan of China.” How far the portrayal of a historical person in a literary work resembles the real person (assuming, for the sake of argument, that this is knowable) is a question about its historical value, which is a separate question from that about its artistic value, just as how far a portrait is a “good likeness” of the sitter is a question about its practical or sentimental value, which is a separate question from that about its artistic value, even if we are in a position to know what the person portrayed did look like, which we seldom are. In brief, I believe that although the created world of a literary work of art bears a necessary but variable relation to the real world, its artistic nature and value do not depend on the degree to which it resembles the real world.

However, it seems that Ingarden’s basic conception of literature is not really mimetic, for he states that in every literary work there is a “more or less determined background, which, along with the represented objects, constitutes an ontic sphere.”¹⁵ This “ontic sphere” appears to be similar to what Wang Guowei called jingjie and what I call the “created world” of a poem. Furthermore, Ingarden took pains to point out that the character of “reality” possessed by the “represented objects” is “not to be identified with the ontic character of truly existing real objects,” and yet “it would obviously be a mistake to assert that the represented objects possess no character of reality at all or that perhaps they take on the character of another ontic mode (i.e., that of ideal existence).”¹⁶ In plainer English, what is presented in a literary work is not “real” in the sense of actually existing, nor “unreal” in the sense of nonexisting, nor yet “ideal” in the sense of existing in some transcendent world. This seems to me much closer to the truth than Sartre’s insistence on the unreality of aesthetic objects. Sartre’s comparison of the experience of watching a play to that of dreaming is misleading,¹⁷ because a play is sharable but a dream is not. Although Hamlet and Ophelia are not real people, anyone can see them in the play, but no one can see another’s dream. If a dreamer can describe his dream so vividly as to make me feel as if I could see it, I still cannot share his dream but only his description of it, since I am in no position to know whether what he describes is the same as what he actually dreamed. The dreamer, or rather the teller of the dream, has in fact become an artist. That is why I speak of the created world of a poem as an extension of reality, not as unreality.

As for Dufrenne’s concept of the world of the literary work of art, it is certainly not mimetic, for he remarks, “the writer strives not so much to describe or mimic a pre-existing world as to evoke a world re-created by him.”¹⁸ Further, he conceives the world of the aesthetic object as being constituted by the “represented world” (monde representé) and the “expressed world” (monde exprimé), the latter being like the soul of the former, which is its body. In his account of the world of the aesthetic object, Dufrenne shows affinities with certain traditional Chinese critics. First, when he speaks of the “world atmosphere” (atmosphère de monde), which is “a certain quality which words cannot translate but which communicates itself in arousing a feeling,”¹⁹ he reminds one of Chinese critics who speak of qixiang, which may be translated as “atmosphere.” For example, Jiang Kui (ca. 1155-ca. 1221) remarked, “In general, poetry must have its own atmosphere [qixiang], countenance [timian], veins [xuemo], and tone [yundu].”²⁰ Yan Yu mentioned “atmosphere” as one of the five principles of poetry and praised ancient poetry for its atmosphere, which is integral, holistic, and not discernible in isolated lines.²¹ Wang Guowei, who used jingjie or “world” as the key term in his criticism, occasionally also spoke of qixiang or “atmosphere.”²² Although he did not explain the relation between the two, it seems reasonable to infer that the “atmosphere” of a poem emanates from its “world,” and that the former is an indescribable quality that characterizes the latter. Further, Dufrenne stresses the subjective aspect of the world of the aesthetic object. To him, the aesthetic object is a “quasi-subject,” and it expresses a personal vision of the world, a Weltanschauung, which “is not a doctrine but rather the vital metaphysical elements in all men, their way of being in the world which reveals itself in a personality.”²³ In somewhat similar fashion, some Chinese critics also stressed the personal mode of sensibility, which reveals itself in an ineffable tone in poetry. Thus, Jiang Kui said, “The poetry of each master has its own flavor [fengwei], just as each of the twenty-four modes of music has its own tone [yunsheng], which is where the music comes to rest.”²⁴ Wang Shizhen, who quoted this remark with approval, advocated shenyun or “spirit and tone,” which, as I suggested elsewhere, involves not only intuitive apprehension of reality and intuitive control over the artistic medium but also a personal tone.²⁵ However, we should realize that this personal mode of sensibility or artistic persona, which is inherent in the work, is not to be confused with the historical, empirical person who created it. Or, to use the terminology of Jacques Maritain, the “creative Self” is not to be confused with the “self-centered ego.”²⁶ Therefore, when Dufrenne identifies the world of the aesthetic object with the world of the author, we should understand by the “world of the author” a reference to the world that the author has created in his work or works, not one to his actual Lebenswelt. It also seems advisable to make a distinction between the world of an individual work and the total world of an author’s whole corpus, the latter being the conglomeration of all the individual works that he has created.

With regard to creativity, although this is not as often stressed in Chinese criticism as in Western criticism, we do find some expressions of a concept of creativity in Chinese poetry and criticism. For example, the poet and critic Lu Ji (261-303) described the process of writing as, inter alia,

The “demonic” poet Li He (790-816), in praising his senior contemporaries Han Yu and Huangfu Shi, wrote,

The critic Xie Zhen (1495-1575) echoed Lu Ji when he wrote,

Whenever we climb high and let our thoughts roam, we communicate with the ancients in spirit. [Our thoughts] reach far and near, and we feel sorrow or joy accordingly. These things give rise to one another in a chance manner, thereby causing forms to appear where there was no trace of anything, and echoes to be evoked where there was no sound.²⁹

Similarly, the critic Ye Xie (1627-1703) wrote,

When a poet is touched by something and his inspiration rises, his ideas, words, and lines all come out of the blue: they all come into being from nothingness. He takes them from his mind wherever he finds them, and expresses them as “emotions” (qing), “scenes” (jing), and “events” (shi) ³⁰

To return to the artistic function of poetry from the author’s point of view: the process of creating an imaginary world is one of verbalization, or verbal incarnation, which involves an exploration of the possibilities of language as an artistic medium and the creation of a unique verbal structure (unique in the sense, already indicated, of not being identical with the structure of any other poem, rather than that of possessing unique linguistic features, just as every individual is unique in the sense of not being identical with any other individual, rather than that of possessing unique physical or mental attributes). It is this process that, when successfully carried out, satisfies the author’s creative impulse and fulfils the second part of the artistic function of poetry for the author. This creative impulse is distinct from any ulterior motive—social, political, moral, or monetary—that the author may have, for such motives cannot account for the choice of the artistic medium and are often absent in any case. Why does anyone want to write a poem, then? Perhaps we can answer this by reversing the answer, made famous by Sir Edmund Hillary, to the question why anyone should want to climb Mount Everest: “Because it is not there.” It seems only natural that a human being should want to create something and to feel a special kind of satisfaction on seeing that what he has created is good. A child who has succeeded in drawing, for no particular reason, what it considers a perfect circle, feels the same kind of satisfaction that Leonardo da Vinci may be supposed to have felt when he finished painting the Mona Lisa, the differences in quality between the two “works” notwithstanding. Similarly, when a writer has created a verbal structure to his own satisfaction, the act will have fulfilled his creative impulse; whether the work has any artistic value depends, among other things, on whether it enables us to satisfy our creative impulse. This leads us to a consideration of the artistic function of poetry from the reader’s point of view.

As already suggested, a literary work of art exists potentially and awaits actualization (or concretization, in Ingarden’s terminology) by a reader, so that rather than saying, with Archibald MacLeish, “A poem should not mean, but be,” perhaps we shoud say, “A poem should not be, but become.” The process of a poem’s becoming, from the reader’s point of view, is paradoxically both a reversal and a revival of the creative process. Insofar as the reader re-creates a world from the verbal structure of a poem, the process is one of deverbalization, which is the reversal of verbalization, yet insofar as the reader follows the words that make up the verbal structure, reading is an approximate reenactment of writing. I say “approximate,” since, obviously, the reader’s experience cannot be literally identical with the writer’s, yet the two must be similar. As John Dewey put it,

For to perceive, a beholder must create his own experience. And his creation must include relations comparable to those which the original producer underwent. They are not the same in any literal sense. But with the perceiver, as with the artist, there must be an ordering of the elements of the whole that is in form, although not in details, the same as the process of organization the creator of the work consciously experienced.³¹

In re-creating the world that the author has created, the reader extends his own Lebenswelt and his perception of reality, and thus the poem fulfils the first part of its artistic function for the reader. Paul Ricoeur appears to be expressing similar ideas when he writes,

For us, the world is the ensemble of references opened up by the texts. Thus, we speak about the “world” of Greece, not to designate any more what were the situations for those who lived them, but to designate the nonostensive references which outlive the effacement of the first and which henceforth are offered as possible modes of being, as symbolic dimensions of our being-in-the-world. For me, this is the reference of all literature: no longer the Umwelt of the ostensive references of dialogue, but the Welt projected by the nonostensive references of every text we have read, understood, and loved. To understand a text is at the same time to light up our own situation or, if you will, to interpolate among the predicates of our situation all the significations which make a Welt of our Umwelt. It is this enlarging of the Umwelt into the World that permits us to speak of references opened up by the text—it would be better to say the references open up the world.³²

If I understand him correctly, the World or Welt of which he speaks is constituted by the individual worlds that we have re-created from all the literary works that we have read, understood, and loved, while the world re-created by each reader from an individual work is an instance of the realization of the possibilities opened up by its nonostensive references. A word of caution may be necessary here: the expression “open up” should not be taken too literally. A work will only “open up” a world to a competent reader actively responding to it; obviously, it will not open up any world to someone ignorant of the language in which it is written or someone who is a competent reader but is not being attentive. Epiphany will come only to the deserving.

The world re-created by the reader is not to be mistaken as the author’s Lebenswelt, since the work no longer refers ostensively to the author’s actual situation or environment but nonostensively to all similar situations that may possibly exist. Even if a work contains references to actual persons, events, places, dates, and so forth (and Chinese poems are particularly replete with such seemingly ostensive references), it cannot revive the actual situation that provided the occasion for the creation of the work. What it can do is to let us experience an imaginary world that resembles the original situation. Poetry is not a time machine that can bring the past back to life; it is, rather, a magic carpet that can transport us into a world that exists outside time. By transcending the author’s actual situation at the time of writing, poetry escapes from time into timelessness.

This is not to say that poetry deals only with the universal, but that it reveals the universal through the particular, or, to adopt the distinction made by E. D. Hirsch, Jr., between “meaning” and “significance,”³³ that poetry transcends its local meaning and acquires universal significance. That is why I wrote elsewhere that “the world of a poem is the concrete embodiment and individualization of a theme,”³⁴ and also, perhaps, why a poem can be a “concrete universal,” as described by W. K. Wimsatt, Jr.³⁵

Just as the reader’s experience of re-creating a work is similar to, but cannot be identical with, the author’s experience of creating it, so is the world re-created by the reader similar to, but not identical with, the one created by the author, or at least the two cannot be proven to be identical. How far they overlap or intersect each other depends on various factors (apart from the question of how far the author has succeeded in creating a world in a verbal structure in the first place), such as the reader’s linguistic competence, knowledge about the author’s cultural world, sense of affinity with the author’s temperament, and employment of one’s powers of perception, imagination, understanding, and reflection. We shall have occasion to return to this question of overlapping or intersecting worlds later.

Let us now consider further the second part of the artistic function of poetry from the reader’s point of view. When we read a poem, whether aloud or silently, we have to say to ourselves certain words, and no other, in a certain order, and no other. In so doing, we are repeating, to some extent, the author’s experience of putting these particular words in this particular order. The possibility of revision does not affect the issue, since in that case we can be said to repeat the author’s experience of writing the final version. When we finish reading, if it is a successful poem, we shall realize that these are just the right words in just the right order, and this realization will give us a feeling of satisfaction comparable to the author’s feeling of satisfaction at finishing the poem and seeing that it was good. It is this feeling of satisfaction, together with the experience of repeating the author’s creative experience, that fulfils the reader’s creative impulse. This feeling of satisfaction is not necessarily “pleasure”: even if a poem shocks us or horrifies us, we are still “satisfied” that the author has employed the right strategy to produce the desired effect, just as a judge is “satisfied” that the defendant did commit the crime, without taking pleasure in the fact. Moreover, when we feel shocked or horrified by a poem, the feeling is not the same as if what is described were actually taking place: to read about someone’s death is not the same as actually seeing someone die. The psychological distance involved ensures our aesthetic satisfaction.

Some readers may doubt the statement that in reading we repeat the author’s process of writing. That this is so can be shown from even the simplest sentences. When we read the sentence “I am cold,” even if we do not say it aloud, in which case we say the words mentally, we are repeating the act of saying these words, and not merely receiving the “message” that someone was cold. In other words, to understand what is said we have to temporarily put ourselves in the speaker’s position. If reading were indeed merely a matter of decoding a message—if, for example, reading the line “Shall I compare thee to a summer’s day” merely gave us the message that Shakespeare wondered whether he should compare his love (Lord Southampton, perhaps?) to a summer’s day—of what possible interest could such a message be to us? I think that we read, and enjoy reading, the line because, first, it enables us to enter the imaginary world of this sonnet and experience the speaker’s (not necessarily the author’s) state of mind when he is imagined as saying these words, and, second, it gives us a feeling of satisfaction to repeat these words that the author has placed there, and to realize that these are just the right words in just the right order. I do not mean that we enjoy the words purely for the beauty of their sound apart from their meaning; on the contrary, our satisfaction is partly due to the fact that the sound reinforces the meaning. For instance, the long vowels and diphthongs compel us to slow down, as if we (identifying with the speaker while we read) were trying to slow down the passing of time and make this short “summer’s day” last a little longer. What I have just said is, of course, also true of Chinese poetry, as well as music and perhaps even some kinds of visual arts. The satisfaction we get from looking at Chinese calligraphy and some Chinese paintings is due primarily, I think, not to the spatial relations among the lines and brushstrokes but to the temporal experience of repeating in our mind the movement of the artist’s brush. In short, I believe that aesthetic experience is a kind of creative experience by proxy: to read a love poem is not a substitute for being in love but an approximate substitute for writing a love poem, just as looking at a painting of apples is not a substitute for eating or even looking at apples but an approximate substitute for painting them.

Having considered the artistic function of poetry, we should, logically, turn our attention to its linguistic structure. However, I do not intend to discuss this in detail, for several reasons. First, our discussions so far have in fact involved some structural considerations, because the relationship between structure and function is a dialectical one: it is the linguistic structure of a poem that enables it to fulfil its artistic function (as well as any nonartistic functions it may also fulfil), and it is the artistic function, whether consciously intended by the author or not, that (together with any nonartistic functions it may also fulfil) determines its linguistic structure. Second, since the linguistic structure of each poem is unique, it is more profitable to discuss the linguistic structure of an individual poem in relation to its artistic function than to discuss the linguistic structure of all poetry in general terms. Finally, even if we do wish to do so, Roman Ingarden has described the structure of the literary work of art in such detail that it would be difficult to add much to his description or to do it full justice with a brief summary. I venture, nonetheless, to raise two questions about terminology, not to register disagreement but to avoid any misunderstandings that may arise from Ingarden’s use of certain terms.

In the first place, Ingarden describes the structure of the literary work of art as a stratified one, consisting of four strata: the stratum of linguistic sound formations, the stratum of meaning units, the stratum of represented objects (or objects projected by the state of affairs, the intentional correlates of the sentences), and the stratum of schematized aspects (or aspects under which these objects appear in the work). The expression “stratified” could give the misleading impression that the structure of a literary work of art is a static one, when in fact it is really a dynamic one, the four strata (which I would prefer to call “elements”) interacting simultaneously. Ingarden himself did discuss the temporal dimension of the literary work of art, but this aspect of his theory has often been overlooked. I still think the word “polyphonic” would be a better description that “stratified,” pace the reviewer of one of my books, who thought it “a pretty metaphor but fairly useless as far as theory is concerned,”³⁶ especially, as Ingarden himself used it, although in a slightly different context, as mentioned before.

Second, it seems to me the “represented objects” and the “schematized aspects” do not belong to the linguistic structure of the work but rather to the created world that emerges from that structure. To be sure, Ingarden does not specify that the structure he describes is purely linguistic, but in the interest of greater clarity it may be better to speak of the “created objects” (rather than “represented objects”) and the aspects under which they appear as supralinguistic elements, which are constitutive of the created world, rather than as strata of the linguistic structure.

The concept of poetry described above implies that a poem, as a literary work of art, is both referential and self-referential, both signifier (signifiant) and signified (signifié), both centrifugal and centripetal (in Northrop Frye’s sense).³⁷ In other words, the linguistic structure of a poem both transcends itself and draws attention to itself. In transcending itself, it yields a created world, which is an extension of reality; in drawing attention to itself, it satisfies the creative impulse of both author and reader. This concept of poetry can be applied, with modifications, to fiction and drama, but in the following pages we shall confine our discussions to what is generally considered “poetry.”