SOURCE: Forsyth, Neil. “Gaston Bachelard's Theory of the Poetic Imagination: Psychoanalysis to Phenomenology.” In The Quest for Imagination: Essays in Twentieth-Century Aesthetic Criticism, edited by O. B. Hardison, Jr., pp. 225-53. Cleveland, Ohio: The Press of Case Western Reserve University, 1971.

[In the following essay, Forsyth provides an overview of Bachelard's critical approach to the concept of imagination, asserting that his development progressed “from the objectivity of psychoanalysis to the subjectivity of phenomenology.”]

When Gaston Bachelard died in 1962, he was probably best known to students of the philosophy of science for his work in the French post-Cartesian tradition. But at present he is also becoming well known for his studies of the literary imagination and the theory of criticism. He continued to teach the philosophy of science at the Sorbonne until his retirement in 1955, always maintaining his early interest in the rational intellect. But gradually he became interested in the problem of the non-rational imagination, partly as a result of his studies of rationalism, but also through the influence of such writers as Bergson and Husserl. While insisting upon the radical separation of the intellect and the imagination, he ultimately became chiefly concerned with the imagination. As his interest in it grew, he first attempted to study it objectively in the same way he had approached the intellect, and he developed a psychological theory of the imagination based on the schema of the four alchemical elements. Eventually he became dissatisfied with this approach, and adopted a position best understood as a variant of twentieth-century phenomenology. The central course of his development, then, is from the objectivity of psychoanalysis to the subjectivity of phenomenology.

Bachelard begins his book L'Air et les songes with a quotation from Joseph Joubert that is significant not only for the development of his own critical theory but for much contemporary French criticism: “The poets should be the chief study of the philosopher who wants to understand man.” As he admits in the introduction to the same book, he is studying literature not as an end in itself, from an aesthetic position, but because it can lead to something else, to an understanding of the human psyche.

As Blake said, “Imagination is not a state, it is human existence itself.” One will be more readily convinced of the truth of this maxim if one studies, as I am going to do in this work, the literary imagination, the spoken imagination, that which, keeping to language, forms the temporal tissue of spirituality, and which consequently separates itself from reality.¹ The principle themes of Bachelard's work are sounded here. He is making a study of the imagination, and in particular of the literary imagination, since that presents the best evidence; it is the “temporal tissue of spirituality.” And, far from adopting a scientific approach to reality, he is studying that which “separates itself from reality.”

Bachelard's study of the literary imagination is in fact a study of the subjective consciousness; it evolves gradually into the completely subjective method of his final period, the phenomenology of the imagination. There is the same distinction here that we find Georges Poulet (himself heavily influenced by Bachelard) setting up in his series of essays on what he calls “the interior distance,” a distinction between an objective and a subjective approach to literature.

Objectively, literature is made up of formal works, the contours of which stand out with greater or lesser clarity. They are poems, maxims, plays, and novels. Subjectively, literature is not at all formal. It is the reality of a thought that is always particular, always anterior and posterior to any object; one which, across and beyond all objects, ceaselessly reveals the strange and natural impossibility in which it finds itself, of ever having an objective existence.²

If we relate this obscure statement to Bachelard's thought, it may become a little clearer. Poulet is approaching in his own way what Bachelard was tackling—the subjective consciousness as it can be studied in literature.

Bachelard too had ignored the formal, objective aspect of literature. He set out to understand not the formal imagination, but the material imagination. It is in this context that we should view the statement he made in a tape-recorded interview: “A poem without unity but with ten beautiful images, well that suits me. You see, my role is very modest, and I don't claim to be a professor of literature. I don't have enough culture for that.”³ With Bachelard we are watching the very first stages of the imagining process, we are back at the roots of being; we are studying not the formal synthesizing imagination but the imagination of matter, which is always fresh. We are ultimately studying the imagination rather than literature. But why does he use literature for his study at all? He attempts to explain this in his introduction to his two books on the terrestrial imagination. His interest in literature is more particularly an interest in the literary image, and we shall watch that interest later as it grows into the central concern of his philosophy. For Bachelard it is through the essential novelty of the poetic image that we are led back to the root of things.

This novelty is evidently the sign of the creative power of the imagination. An imitative literary image loses its animating power. Literature has to surprise. Of course, literary images can exploit some fundamental images—and my work in general consists in classifying these fundamental images—but each image which comes under the writer's pen must have its own differential of novelty. A literary image says what will never be imagined twice. There might be some merit in copying a painting. There is none in repeating a literary image.⁴

We need to remember here that the most powerful tradition in twentieth-century French poetry, in strong contrast to contemporary Anglo-American trends, has been toward a poetry that puts its emphasis on the revelatory power of the image itself, on the fresh, surprising quality of the poetic image, the poetry of Surrealism. Bachelard at times sounds a lot like the philosopher of Surrealism, but he is not bound by its limits, since he is using the surrealistic image for his own purposes. He has this to say in the Introduction to La Terre et les rêveries de la volonté:

To reanimate language by creating new images, that is the function of literature and poetry. … Every new literary image is an original use of language. To feel its action, it is not necessary to have the knowledge of a linguist. The literary image gives us the experience of a creation of language. …

We are now in a century of the image. For better or worse, we are more than ever undergoing the action of the image. … In the fire and brilliance of literary images, the ramifications multiply; words are no longer simply terms. They do not terminate in thoughts; they have the future of the image. Poetry makes the sense of a word ramify by surrounding it with an atmosphere of images. It has been demonstrated that most of Victor Hugo's rhymes excite images; between two rhyming words plays a sort of metaphorical bond: thus images are linked together thanks simply to the sonority of words. In a freer kind of poetry like Surrealism, language is a full ramification. The poem is a cluster of images.

(pp. 6f.)

Of translations, two are particularly significant: La Poétique de l'espace, is from his last period,⁵ and the other, La Psychanalyse du feu,⁶ represents the transition from his first to his middle period, from his strictly philosophical concerns to the literary studies that will fill the last years of his life. There is a considerable difference between these two books, indeed they represent almost the opposite ends of Bachelard's development—objective and subjective optimism—although both reflect the constant presence of optimism. In the first one he is optimistic about the possibility of objectivity; in the later he is tracing the completely subjective path by which he discovers his own happiness.⁷ When we look more closely at La Psychanalyse du feu later we shall see that the transition from one kind of thinking to the other is already in evidence there.

I

The original concern of Bachelard's thought was a scientific one, a search for pure objective knowledge of the external world. This search involved a careful examination of our own attitudes to whatever is the subject of our investigation. In order to advance in the knowledge of an object, we must guard ourselves, it seems, from knowing it. The assumption is that final and pure objectivity of knowledge is going to be something strange, and we need to keep a careful watch that we do not attribute the least recognizable trait to the object. The search is for pure, external, and impersonal truth, and it is consequently a denial of all subjectivity. His chief work on this subject is La Formation de l'esprit scientifique, published in 1938, but as he himself was to say later, it was badly named;⁸ the subtitle is more significant: “A Contribution to the Psychoanalysis of Objective Knowledge.” In it he is concerned about demonstrating this problem: “The scientific mind must align itself against Nature, against what is in us and outside us, the impulse and instruction of Nature.”⁹

La Psychanalyse du feu grew out of this epistemological concern, and it was published in the same year. In it he makes a very similar statement:

Scientific objectivity is possible only if one has broken first with the immediate object if one has refused to yield to the seduction of the initial choice, if one has checked and contradicted the thoughts which arise from one's first observation.

(P.F. [La Psychanalyse du feu], 1)

The important word here is perhaps “seduction,” and we should notice that it is the imagination which has this seductive power. The aim of this scientific method is to detach the observer from the object, to remove all the affective and imaginative qualities from the mind, to achieve the purely objective functioning of the intellect, to attain to “the perfect limpidity of a thought that, because it has been thoroughly psychoanalyzed, has become exclusively objective.”¹⁰ It is a sifting, purifying process; separating objectivity from all the impurities of subjectivity. The surprising result of this attempt becomes clear in La Psychanalyse du feu: when you psychoanalyze the process of objectivization you also psychoanalyze the subjective life. If you remove all the elements of subjectivity from objectivity, you don't just have pure objectivity, you have pure subjectivity as well. The implications of this fact are important for Bachelard's development.

Bachelard recognizes that if one eliminates from every mental phenomenon the appearance of objectivity that it offers to the contemplator, a spiritual substratum remains. It is this which is going from now on to occupy his attention: no longer an integral objectivity which discovers itself to the mind only after every trace of subjectivity has disappeared, but on the contrary an interior life whose significance can only be understood when the mind, having, conditionally or not, abandoned the search for objective truth, contemplates the subjective reality in itself, in its substantial simplicity and in itself.¹¹

From this article by Poulet on Bachelard we may notice in passing how much like his own studies does this analysis of his mentor sound. But he is expressing something that is central to Bachelard's thought. It might help to understand it if we use an analogy from the smelting process, which is also rather appropriate in view of the paradox we will discover in Bachelard's discussion of the relation between fire and the purifying process. Smelting separates out all the impurities from iron and leaves you with a molten mass of pure ore; but you are also left with a pure heap of impurities. It is just a matter of which pile you direct your attention to. Bachelard moves more and more from the study of one pile, the activity of the intellect, to the other, the activity of the imagination, always keeping the two entirely separate.

La Psychanalyse du feu represents the transition between these two stages of Bachelard's thought. It grew out of his concern with establishing a method for achieving this scientific objectivity, and thus primarily deals with exposing the fallacies of our convictions about fire. Yet it is, he says,

no longer the axis of objectivization but that of subjectivity that I would like to explore in order to illustrate the double perspectives that might be attached to all problems connected with the knowledge of any particular reality.

(P.F., 3)

By the time he had completed the book he had become aware of the importance of the subjective power, the power of the imagination, so that he could write a conclusion looking forward to the possibility of creating a new kind of literary criticism, one which would study the true source of psychic and poetic production, the imagination, through the images it creates.

In the larger context of Bachelard's entire oeuvre we can see La Psychanalyse du feu as the book which looks both backward and forward, and which established the distinctive pattern of the rest of his philosophical work. It is not so much that he made radical revisions in his theory as that he expanded it enormously. His consciousness now has two faces. Instead of trying to eradicate all the imaginative fallacies that cluster around scientific observation, he begins to look at the imagination as an important activity of the human psyche. The polarity of the intellect and the imagination remains central to his thinking, but the relationship between them is reversed.

After La Psychanalyse du feu he can say:

We need to examine all the desires we have to leave what one sees and what one says in favor of what one imagines. That way we shall have the chance to give back to the imagination its role as seducer. By the imagination we abandon the ordinary course of things. To perceive and to imagine are as antithetical as presence and absence. To imagine is to absent oneself; it is to start out toward a new life.

(A.S. [L'Air et les songes], 10)

The imagination is still the seducer, but it is no longer the corrupter. It is worth studying in its own right, it is even healthy: “A being deprived of the function of the unreal is a neurotic just as much as one deprived of the function of the real” (A.S., 14).

La Psychanalyse du feu, then, begins the middle period of Bachelard's writings. He is no longer so concerned with the need for scientific thinking to be objective, though he still remained a philosopher of science, which he taught at the Sorbonne until his retirement.¹² His primary interest now is with the four kinds of material imagination, more especially as they manifest themselves in literature. The four elements of ancient science and its philosophy have no objective value for modern science, but they are important to the imagination. They are a priori archetypes of the mind, rooted in the human unconscious; all imaginative activity is conditioned by them. When we apply this to the poetic imagination in particular, we arrive at a theory of the four humors or temperaments, a theory of imaginative polarization, so that every poet may be inscribed in one of these four categories, according to which of the larger images he prefers. Bachelard summarizes the theory in L'Air et les songes:

I am considered to have initiated the notion of a law of the four material imaginations, a law which attributes necessarily to the creative imagination one of the four elements: fire, earth, air, and water. Of course, several elements can intervene in the make-up of a particular image; there are composite images; but the life of images has a more demanding purity of relation. As soon as the images present themselves in order, they designate one material as first, one basic element. The physiology of the imagination, even more than anatomy, obeys the law of the four elements.

(A.S., 14)

If we look back to La Psychanalyse du feu, we can see the theory in its initial stages, apparently when it first occurs to Bachelard. Fire had originally been chosen as the object he would use for his study, not because he had any idea of a theory of the four elements, but because it is the most obviously double of the objects that produce images: as well as being something objectively and in itself, it is very obviously beset with subjective fallacies, as Bachelard had discovered in his studies of alchemy. But as he studied our reactions to fire he became aware of the possibilities. The entire passage deserves to be quoted since it stands at the beginning of the efforts that were to mark the next ten years of his life.

If our present work serves any useful purpose, it should suggest a classification of objective themes which would prepare the way for a classification of poetic temperaments. We have not yet been able to perfect an overall doctrine, but it seems quite clear to us that there is some relation between the doctrine of the four physical elements and the doctrine of the four temperaments. In any case, the four categories of souls in whose dreams fire, water, air, or earth predominate, show themselves to be markedly different. … Reverie has four domains, four points from which it soars into infinite space. To surprise the secret of a true poet, of a sincere poet, of a poet who is faithful to his original language and is deaf to the discordant echoes of sensuous eclecticism, which would like to play on all the senses, one word is sufficient: “Tell me what your favorite phantom is. Is it the gnome, the salamander, the sylph, or the undine?” Now—and I wonder if this has been noticed—all these chimerical beings are formed from and sustained by a unique substance: the gnome, terrestrial and condensed, lives in a fissure of the rock, guardian of the mineral and the gold, and stuffs himself with the most compact substances; the salamander, composed all of fire, is consumed in its own flame; the water nymph or undine glides noiselessly across the pond and feeds on her own reflection; the sylph, for whom the least substance is a burden, who is frightened away by the tiniest drop of alcohol, who would even perhaps be angry with a smoker who might contaminate her element, rises effortlessly into the blue sky, happy in her anorexia.

(P.F., 89ff.)

This theory does not of course mean that people are rooted in a particular substance, but that “these primitive images orient psychological tendencies; these were the sights and impressions which suddenly aroused an interest in what is normally devoid of interest, which gave an interest to the object” (P.F., 90). For that “normally” we might read “objectively” and thus we are back to our original distinction between the subjective and the objective activity of the mind. The theory of the four material imaginations is a theory which attempts to classify the subjective projections that we make onto the outside world. For example Hoffman and Poe, writers that one might imagine to be alike in the character of their imagination, are in fact very different; their attitude toward alcohol is entirely dissimilar:

The alcohol of Hoffman is the alcohol which flames up; it is marked by the wholly qualitative and masculine sign of fire. The alcohol of Poe is the alcohol that submerges and brings forgetfulness and death; it is marked by the wholly quantitative and feminine sign of water.

(P.F., 91)

We are to assume that objectively alcohol is the same substance both for Hoffman and for Poe; but subjectively it is entirely different. There it is conditioned by the humor, by that a priori category of the human psyche which conditions all our imaginative life.

II

Two doctrines begin to emerge here in Bachelard's middle period that are going to be of central importance, and which he carries to their extreme in his final books. The first is a discovery that the central concern of his literary studies should be the poetic image; that his theory of the four material imaginations is taking us back to the roots of our being, and that what is found can be studied as it reveals itself in the image. The second is the notion, which we see only in embryo in La Psychanalyse du feu, that the state of mind in which the imagination is freest to create these images is “reverie.”

The image is to be studied in its dynamic freshness. This is the century of the image, and Bachelard makes the most of that fact: “The poem is essentially an aspiration towards new images” (A.S., 8). He considers the dynamics of the image, that is, the way the image echoes in the psyche, gathering nuances. “The value of an image is measured by the extent of its imaginary halo” (A.S., 7). This dynamism can most clearly be seen in relation to aerial images, since they are the most obvious. The image and, in particular, the aerial image, has a liberating function. I quote from the Introduction to L'Air et les songes, where he claims that the aerial imagination is the best example of the imagination of movement, of this dynamic imagination.¹³ It is interesting to note how far he has already (1943) come from the position of La Psychanalyse du feu; indeed, he begins to sound like Sartre in attributing such value to the unrealising function of the mind:

Like many problems of psychology, studies of the imagination are impeded by the false light of etymology. One always wants the imagination to be the faculty of making images. But it is more often the faculty of unmaking the images furnished by the perception. It is above all the faculty of freeing us from the initial images, of changing images. If there is no changing of the images, an unexpected union of images, there is no imagination, there is no action of imagining. If an image that is present does not make us think of an image that is absent, if one occasional image does not set going a whole host of wandering images, an explosion of images, there is no imagination. There is perception, a memory of perception, a familiar memory, the customary colors and shapes. The basic word which corresponds to imagination, is not image but imaginary. The value of an image is measured by the extent of its imaginary halo. Thanks to the imaginary, the imagination is essentially open, escapist. Within the human psyche it is the experience of opening, the experience even, of novelty.

(A.S., 7)

And it is the poem which responds to our need for novelty, for the freshness “which is a characteristic of the human psyche.”

These dynamics only take place properly when the image is entirely fresh, and at this stage in Bachelard's thought there is an important reason for this demand for novelty in the image. He is not studying metaphors, which are standardized images and therefore lifeless, lacking in dynamism; nor is he studying those images that have been developed in myth and become domesticated, downgraded into what Heidegger calls the Gerede.¹⁴

There is a big difference between a literary image which describes a beauty that has already been realized, a beauty that has reached its full form, and a literary image that works in the mystery of matter and which wants more to suggest than to describe.

(T.R.V. [La Terre et les rêveries de la volonté], 8)

The mystery of matter. It is only by the freshness of the individual poetic image that the imagination gets back to matter, that a harmony is established between the self and the thing, a harmony that implies an innocence, a happiness that is very dear to Bachelard, and which lies at the root of his theory of reverie.

Several problems begin to arise here. First, although Bachelard wants each image to be new and fresh, he establishes categories for his images that are as old as the human psyche itself, that are embedded in the human unconscious as archetypes. He comes to discuss this problem in his later period, and indeed this is perhaps the central issue that is forcing him toward the phenomenological method. We should now extend the discussion to include that final period, since we are dealing with the problems that he attempts to face there.

He attacks the problem of the freshness of the image at the beginning of the long and important statement of his poetics which serves as an introduction to La Poétique de l'espace. He talks here of the non-causal relation of the image, as it emerges, to the unconscious archetypes. The image is “a sudden salience on the surface of the psyche” (P.S. [La Poétique de l'espace], xi); but the onset of the image is not caused by the poet's unconscious. “The poetic image is not the result of an inner thrust. It is not an echo of the past.” It would apparently be taking a very simplistic view of the workings of the mind if one regarded the unconscious as the cause of phenomena in the conscious mind. We can see in this attitude the increasing concern of Bachelard not to explain away the phenomena of the imagination as sublimations of certain facts of the unconscious. This is the anti-reductionism that is a common feature of the phenomenological reaction against classical psychoanalysis, a reaction that we discover as early as La Psychanalyse du feu. If anything, Bachelard may be suggesting here that the causal relation is the other way round; if the image is not caused by the past, if the fact of its newness cannot be explained by relating it to an archetype, it is still true that it is only by means of this newness that its archetypal nature is revealed. The image, in fact, may cause reverberations that are the echoes of the past: “… through the brilliance of an image, the distant past resounds with echoes” (P.S., xiii). At any rate, we can see what is involved in the movement of Bachelard's thought. He is becoming increasingly aware of the inadequacy of previous theories to explain the observed facts of mind, and in particular to explain the surprising nature of the poetic image; he is moving away from his studies of the material imagination to a study of the image in all its uniqueness, a movement from psychoanalysis toward phenomenology.

If we find that the problem of the poetic image is pushing Bachelard toward his later method, even more so does the closely related doctrine of the state of mind in which the poetic image emerges, the state of reverie. Indeed, his last major work, La Poétique de la rêverie, is devoted to the subject, and represents the culmination of a trend toward the subjective method; it is almost a reverie on reverie. Let us look at two important statements in which the method of this last period is announced, and which repudiate his earlier approach:

In my earlier works on the subject of the imagination I did in fact consider it preferable to maintain as objective a position as possible with regard to the images of the four material elements, the four principles of the intuitive cosmogonies; and faithful to my habits as a philosopher of science, I tried to consider images without attempting personal interpretation. Gradually, this method, which had scientific prudence on its side, seemed to be an insufficient basis on which to found a metaphysics of the imagination. The prudent attitude itself is a refusal to obey the immediate dynamics of the image.

(P.S., xiv)

As well as restating his position on the utter polarity of the two aspects of the consciousness, the intellect and the imagination, Bachelard eventually adopted entirely different methods for studying them.

Far be it from me to try to weaken this polarity … I have already found it necessary to write a book in order to exorcise the images which, in a scientific culture, pretend to engender and sustain concepts. … But on the other hand, in stating my faithful devotion to images, I would not dream of studying them from the point of view of concepts. The intellectual criticism of poetry will never get to the seat [foyer] where poetic images are formed. … The image can be studied only by the image, dreaming the images as they gather in reverie. It is nonsense to pretend to study the imagination objectively, since one only really receives the image if one admires it. Even in comparing one image with another, one risks losing one's participation in its individuality.¹⁵

The reference here is back to his early period, to La Formation de l'esprit scientifique, out of which La Psychanalyse du feu grew. We can see now how far he has come. He has established two opposite poles in the activity of the psyche; one is the objective knowledge he was searching for in his first books, knowledge that is to be achieved by the heightened activity of the rational intellect; the other is the subjective “living” of the poetic image in the moment when it is formed, in the state of reverie.

The notion of reverie has developed considerably from the time of its introduction in La Psychanalyse du feu. If we look back for a moment to that book, we can see that the idea changes even there. It begins simply as the state of the mind observing the fire, the state which he is trying to overcome. It is the fireside reverie that he calls the “futility of reverie” (P.F., 4) because it engenders all those false subjective notions about fire. But he soon realizes its importance in the study of poetry. It becomes the poetic state of mind, and we find Bachelard looking upon his book as a basis for determining the objective conditions of reverie, and thus establishing an “objective literary criticism” (P.F., 109). From the simple state that the mind invariably adopts before the fire, it becomes the whole state of the poetic imagination.

In La Psychanalyse du feu Bachelard differentiates reverie rather casually from the night dream, largely on the basis that reverie is always more or less centered on one object. He is only dimly aware of a distinction that was to become central to his later poetics: the fact that reverie is a consciousness, whereas the night dream is unconscious. In La Psychanalyse du feu Bachelard first becomes aware of the problem:

As it happens, one of the advantages of the psychoanalysis of objective knowledge that we are proposing to carry out seems to be that we are examining a zone that is less deep than that in which the primitive instincts function; and it is because this zone is intermediary that it has a determinative action on clear thought, on scientific thought. …

Since we are limiting ourselves to psychoanalysing a psychic layer that is less deep, more intellectualised, we must replace the study of dreams by the study of reverie, and more particularly in this little book we must study the reverie before the fire.

(P.F., 12-14)

As the phenomenological method develops, Bachelard comes to insist more and more on this distinction between dream and daydream that we see here in embryo. There is a vital and absolute line to be drawn between the two, the line between consciousness and the unconscious, between the presence or absence of a cogito. It is a distinction which ultimately comes to be the difference between phenomenology and psychology.

In brief, it boils down to this: to determine the essence of reverie, one should come back to the reverie itself. And it is precisely by phenomenology that the distinction between dream and daydream can be elucidated, since the fact that consciousness is involved in reverie is the decisive mark.

(P.R. [La Poétique de la rêverie], 10)

The theory has now become the center of Bachelard's poetics, and we need to trace briefly some of its implications in the final period of his life.

Reverie implies solitude, the state where man finally encounters things in their true reality. It is creative and not passive; in other words, it is not the state where man is gradually lapsing into the unconscious world of dreams. It is the supremely creative state in which the poetic image is formed (though we should beware of using the passive voice here, since Bachelard often says that “image” is the subject of the verb “to imagine”). All the senses awaken and harmonize in poetic reverie. It is this polyphony of the senses that the reverie listens to, and which the poetic consciousness must register. Poetic reverie writes itself, or at least promises itself to write. Bachelard's philosophical ambition is large, and he admits it: “to prove that reverie gives us the world of a soul, that a poetic image bears witness to a soul as it discovers its world, the world where it would like to live, where it is worthy of living” (P.R., 14).

The reverie operating poetically maintains us in a space of intimacy which stops at no frontier—a space uniting the intimacy of our dreaming being to the intimacy of the beings we dream. … All the being of the world amasses itself around the cogito of the dreamer.

(P.R., 140)

In reverie, there is no longer a “non-I”; the word “no” has no meaning any longer. We are plunged into an inside that has no outside. In reverie, the purest state of the imagination, we lose the distinction of subject and object.

The act of the creative consciousness [of the reader's as well as the poet's, since in reverie the reader repeats the creative act of the poet] must be systematically associated with the fleeting product of that consciousness, the poetic image. At the level of the poetic image, the duality of subject and object is iridescent, shimmering, unceasingly active in its inversions.

(P.S., xv)

Reverie becomes a cosmic reverie, a completely unified existence. “The simplest hearth enfolds a universe” (P.R., 166).

It is in this state of reverie that the poetic image is formed. Bachelard's phenomenology concentrates on the emergence of the image in the individual psyche. The image has no past, it is a totally new event in the consciousness. And yet the strange fact is: it is communicable, it has a quality of trans-subjectivity. This communicability of the image is the subject that Bachelard later investigates. It comes to mean that the creative act of the poet's imagination, its intentionality, is repeated in that of the reader. The phenomenologist studies this fact of communication, but not as if he were in the world of science, the “real” world composed of external elements, perceived by the senses and coordinated by the intellect; he is in an unreal world, acting as a subjective power. That is, he is trying to fabricate within himself the world of the imagining man. One can only understand this imagining man by sharing his imagining, by reenacting the same act of imagination. The phenomenologist is, in fact, acting as the perfect reader of poetry—he becomes a poet himself. He has to keep himself from objectifying his own thoughts, his own act of consciousness. This is the problem, the paradox that we discover in the development of Bachelard's thought, and that we find in La Psychanalyse du feu—the problem that, in studying the subjective power objectively, you destroy its validity. And yet, how can you study something subjectively? In his phenomenological method, Bachelard answers the problem by attempting to become the subject, to live the subject.

Bachelard admits candidly that this involves leaving out of account all the “labor of composition.” We noticed before that this was because of his distinction between the formal and the material imaginations, that he was not studying the synthesising power, but the image in its initial dynamism. An interesting development appears now in his later work. His study of the material imagination has become both concentrated and expanded. He is now studying the individual image as it emerges; and relating it to more categories than simply the four elements. But the old distinction reappears under a new guise. It is not the material imagination, but the “soul”; the formal imagination is now the “mind.” He is studying the poetic image as it appears in the poet's soul, not as it is acted upon by the mind. We can see here, then, the development of Bachelard's earlier categories. He relates the notions of “soul” and reverie as essential to the poetic act.

To specify exactly what a phenomenology of the image can be, to specify that the image comes before thought, we should have to say that poetry, rather than being a phenomenology of the mind, is a phenomenology of the soul. …

The dialectics of inspiration and talent become clear if we study their two poles, the soul and the mind. In my opinion, soul and mind are indispensable for studying the phenomena of the poetic image in their various nuances, above all, for following the evolution of poetic images from the original state of reverie to that of execution. … To compose a finished, well-constructed poem, the mind is obliged to make projects that prefigure it. But for a simple poetic image, there is no project; a flicker of the soul is all that is needed.

(P.S., xviff.)

A flicker of the soul. He quotes the poet Pierre-Jean Jouve to help make his point clearer: “Poetry is a soul inaugurating a form.” Inaugurating it, not bringing it to its completion—that is the job of the talent, of the “mind,” of what he used to call the formal imagination. In La Poétique de l'espace Bachelard promised to write a book on poetic reverie that would be a phenomenology of the soul, and that is how we must interpret La Poétique de la rêverie, his next book. It is in the soul, during reverie, that the poetic image emerges.

III

We have been following the development of the concept of reverie from its beginnings to its final enshrinement at the center of a poetic theory. We are going now to look back to that beginning, to a study in which the notion of reverie, and its accompanying idea of soul, seems false or at least an annoying interruption of the subjective psyche onto the objective world. As we look back, we move from the subjective study of the soul to the objective study of the rational intellect, but now, with hindsight, we can discern the fallacy that is at the basis of the earlier study. In viewing La Psychanalyse du feu we see the paradox that Bachelard became aware of himself only much later, yet which opened the way for the development of his future theory. This paradox exists in all the ways in which fire is linked to knowledge.

Although Bachelard was to interpret it as such later in his life, La Psychanalyse du feu is neither a study of the material imagination nor of the poetic image; those are the studies that grow out of it. It sets out, in fact, to be a contribution to the science of epistemology that had concerned Bachelard so much in his early period—a study of the general phenomena he had isolated in La Formation de l'esprit scientifique. The fact that the book does not finally come out like that is something for which aesthetics can be grateful, since it opened the way for the studies we have been outlining. There are, in a sense, two introductions in this book. One is his original project, the actual introduction which describes how he plans to test the particular phenomena of fire for subjective errors and eliminate them. By psychoanalyzing our convictions about fire, “we shall have many opportunities to show the dangers that first impressions, sympathetic attractions, and careless reveries hold for true scientific knowledge” (P.F., 3). But the other introduction is what, in retrospect, seems to be the true direction of the work. For he also sets himself another, very different, task, apparently rejecting the very psychoanalysis which had been his original objective.

Petitjean was able to write that the imagination eludes the determinations of psychology—psychoanalysis included—and that it constitutes an autochthonous, autogenous realm. We subscribe to this view: rather than the will, rather than the élan vital, imagination is the true source of psychic production. Psychically we are created by our reverie—created and limited by our reverie—for it is the reverie which delineates the furthest limits of our mind. Imagination works at the summit of the mind like a flame. … We must then find the way to set ourselves at the place from which the original impulse is directed into various channels, doubtless led astray by its own anarchical tendency, but also impelled by the desire to charm others.

(P.F., 110f.)

The rest of Bachelard's work is, in a sense, attempting to do just that—to find a way to set ourselves at the place at which the original impulse is formed.

We may notice here another important development or movement within the book, reflecting the overall trend we have been analysing. It sets out to be a study of the errors of the imagination; the tone is confident, ironic. Bachelard psychoanalyzes the subjective convictions related to the knowledge of fire phenomena, or “more briefly, a psychoanalysis of fire” (P.F., 5). He is attempting to establish “the secret persistence of this idolatry of fire,” and consequently he examines documents from close to his own time. Since his purpose is to expose error, the most useful documents will be from pre-scientific literature, that is, documents which purport to be scientific but which, because of ignorance of modern scientific methods, are full of subjective illusions. But many of his examples are taken from poetry, initially for the purpose of exposing the subjective error more completely. But the literature of alchemy and the literature of poetry can only be lumped together as error like this when the focus is not on them so much as on what is left after they have been removed—the pure objectivity that Bachelard is pursuing. As soon as the focus shifts, as soon as he becomes more interested in what he is removing than what he is removing it from, he begins to realize that the poetic image is unique. There is a movement from considering literature as error to considering it in itself, from considering imaginative works as examples of something else, to considering them in themselves. He begins to turn his attention to the “pure heap of subjectivity” that we talked about earlier. This quite evidently is an important change of focus, and Bachelard is only partially aware in this book of its implications for a study of poetry. The poetics of this theory are to come later. But in the movement from pre-scientific to poetic literature, we can see its beginnings; we can see reflected here the movement from an objective pursuit of objectivity to the objective pursuit of subjectivity which we have been discovering in Bachelard's overall development. That is partly why the book is such a hodge-podge, since it represents the transition between the two interests.¹⁶

Bachelard groups the subjective values that we attribute to fire under separate heads that he calls “complexes,” but in view of his anti-Freudian viewpoint we should consider them as something different from, though closely related to, the categories of classical psychology. They are networks of chain reactions, complexes of events within the mind that are set in motion merely by the contemplation of fire. There are several of these complexes, but only three of them need concern us here. In the theory of complexes are the seeds of what he is later to discuss as the “reverberation” of the poetic image. We imbue fire with values and attributes that it obviously does not have objectively. Among all phenomena, for example, it is the only one to which we can impute so definitely the opposing values of good and evil. “It shines in Paradise. It burns in Hell” (P.F., 7). The associations that we have with fire are both archetypal and personal, and there is no real problem here of the relation between those two aspects as there was to be later.

Let us look at the first of these complexes—the Prometheus complex. This is the basic complex associated with fire; and it is one that is closely linked with childhood. We avoid the fire because our father will rap our fingers if we get too close to it; thus the respect that we show to fire is originally a social respect, and it is our first general knowledge of fire. The Prometheus complex is an outgrowth of this and leads to our first particular knowledge of fire. In order to obtain it, we have to disobey our father, to outwit him. This is in fact a fundamental characteristic of man; it is the will to intellectuality. We can see already Bachelard's anti-reductionism here. Instead of explaining our desire to know about fire by the usefulness of the knowledge we are going to obtain, he looks at it in itself, as a will to know as much as our fathers do for its own sake. It may not be as basic an instinct as the Freudian Oedipus complex, but psychology must also study “minds of a rarer stamp” (P.F., 12). “The Prometheus complex is the Oedipus complex of the life of the mind” (P.F., 19). We are going to return later to this basic complex of our associations with fire, but it is noticeable that the linking of fire and knowledge is already apparent.

IV

There is an opposite complex associated with fire, and in this connection we can take account of Bachelard's more specifically literary interests. This complex is called the Empedocles complex, since it is the instinct for dying—the death wish. If the Prometheus complex is the wish to seize fire, an active wish, the Empedocles complex is the wish to be seized by the fire, an ultimately passive wish. Bachelard cites as an example a long and very Gothic fragment by George Sand, which is rather poor literature. This is significant because in bad literature the basic errors of the subjective mind that he is looking for are going to be more easily recognizable. At this point, Bachelard does not quote literature because he likes it, so much as because it helps him to expose the imaginative errors. But he follows this passage with an explicit comment about successful poetic works, and though we are here at the very beginning of Bachelard's literary criticism, we see a tendency that is going to come much more into evidence later.

When one has recognized a psychological complex, it seems that one has a better and more synthetic understanding of certain poetic works. In point of fact a poetic work can hardly be unified except by a complex. If the complex is lacking, the work, cut off from its roots, no longer communicates with the unconscious. It appears cold, artificial, false.

(P.F., 19)

Bachelard is talking here not exactly about form itself, but about a certain formal unity of the total poetic work, instead of what he comes later to concentrate on, the individual image. Even an unfinished work by Hölderlin, Empedocles, has a certain unity simply because it is grafted onto the Empedocles complex.

Secondly, the sole criterion here for the success of a poetic work seems to be the presence or absence of a complex, of some relation to the unconscious. Even when Bachelard narrows his theory, as it becomes more specifically literary, to a concentration on the poetic image itself, the same criterion seems to apply. The passage above sounds very similar to the previous citations from Bachelard's middle period, showing how he differentiated between image and metaphor; even in his final period, he still uses the same criterion—indeed, it is the only basis on which he presumes to judge literature.

Let us look forward for a moment to La Poétique de l'espace. Bachelard has just been analyzing what he calls “intimate immensity” in Baudelaire. It is one of the fundamental images of the psyche, and in the work of a great poet like Baudelaire, one can hear its call. This suggests, for one thing, that the purpose of poetry is to communicate these fundamental images (even with Baudelaire, who preceded “the century of images”), and also that we must judge poetry by whether it succeeds in doing so. In order to make his point clearer, he gives a passage from Taine's Voyage aux Pyrénées as an example of immensity that is not successful because it is not in touch with the fundamental image.

“The first time I saw the sea,” writes Taine, “I was most disagreeably disillusioned. … I seemed to see one of those long stretches of beet fields that one sees in the country near Paris, intersected by patches of green cabbage, and strips of russet barley. The distant sails looked like homing pigeons and even the outlook seemed narrow to me; painters had represented the sea as being much larger. It was three days before I recaptured the feeling of immensity.”

Beets, barley, cabbages, and pigeons in a perfectly artificial association! To bring them together in one “image” could only be a slip in the conversation of someone who is trying to be “original.” For it is hard to believe that in the presence of the sea, anyone could be so obsessed by beet fields.

(P.S., 199)

Two points need to be made here. First that the passage from Taine is treated a little like a statement by a client on the existential psychoanalyst's couch, and second, that there seems to be an implicit assumption that the presence of the sea, by itself, will produce the right associations. Thus any literature which does not reproduce those associations is not merely fiction, it is dishonest and therefore bad. There is a criterion of truth implied in the judging of literary works that we find almost explicitly stated in an earlier book, La Terre et les rêveries de la volonté:

For me, the argument I want to discuss about the primitiveness of the image is immediately decided because I attach the true life of images to the archetypes whose activity psychoanalysis has demonstrated. Imagined images are sublimations of archetypes rather than reproductions of reality. And since sublimation is the most normal dynamism of the psyche, we shall be able to show that images come out of a true human depth.

(T.R.V., 4)

Disregarding the difference, for the moment, between a work of Bachelard's middle and a work of his late periods we can see clearly enough the criterion implied, and the ultimate purpose of the study of literature. “I thought it possible, in the simple examination of the literary image, to discover an emergent activity of the imagination” (T.R.V., 7). Yet how can the critic be expected to achieve this task if the writer of literature lies about that “emergent activity of the imagination?”

Whether this is a justifiable reason for dividing literature into “good” and “bad,” instead of some other category such as “useful for my purpose” or “not applicable to my purpose,” is a complicated question. It involves deciding, for one thing, whether there is any better reason for dividing literature up in this way. The basis for the distinction is: does this image work or not? It is the same distinction, in effect, that Coleridge made between the fancy and the imagination: a good image for Bachelard, one of Baudelaire's for example, would be the product of the imagination in Coleridge's terms; a bad one, like that by Taine, would come solely from the fancy. There is, in fact, a good case for regarding Bachelard's work as an extension of Coleridge's theory. Whereas Coleridge was content to make such a distinction and examine a few favorite images in terms of it, Bachelard, with the advantage of twentieth-century psychological discoveries, and ultimately by the use of the phenomenological method, tries to explain in extreme detail why the Coleridgian category of the imagination produces great poetry, while the fancy can only associate objects like Taine's beets, barley, cabbages, and pigeons. In order to demonstrate the “why” of poetry, Bachelard examines the faculty that Coleridge associated with it, the creative imagination. Just as Coleridge was driven ultimately to metaphysical theories, so Bachelard gradually moves toward his later phenomenology, which he sometimes calls a metaphysics of the imagination. He states this explicitly in the final chapter on the phenomenology of roundness of La Poétique de l'espace, claiming that we have to “de-psychoanalyse” ourselves.

Some five or ten years ago [that is, at about the time of the above quotation from La Terre et les rêveries de la volonté, in which images were specifically related to unconscious archetypes] in any psychological examination of images of roundness, … we should have laid stress on psychoanalytical explanations for which we could have collected an enormous amount of documentation, since everything round invites a caress. Such psychoanalytical explanations are, no doubt, largely sound. But they do not tell everything, and above all they cannot be put in the direct line of ontological determinations. When a metaphysician tells us that being is round, he displaces all psychological determinations at one time. He rids us of a past of dreams and thoughts, at the same time that he invites us to actuality of being. [Bachelard has already changed Jaspers' statement “Jedes Dasein scheint in sich rund,” to “das Dasein ist rund,” in order to abolish the doublet of being and appearance, when we mean the entire being in its roundness.] It is not likely that a psychoanalyst would become attached to this actuality enclosed in the very being of an expression. From his standpoint such an expression is humanly insignificant because of the very fact of its rarity. But it is this rarity that attracts the attention of the phenomenologist and encourages him to look with fresh eyes, with the perspective of being that is suggested by metaphysicians and poets.

(P.S., 236f.)

That last sentence is perhaps as close as we can get to an understanding of the reason for the poetic quality of Bachelard's final metaphysical period, and we should notice how the poet's and the metaphysician's approach are equated. Both are an attempt to look with fresh eyes from a new “perspective of being.”

If this was to be the final direction of Bachelard's thought, to explain the “why” of poetry in metaphysical terms, we can see why the method of La Psychanalyse du feu was inadequate. In retrospect, we can see this growing apparent.

V

After examining the Prometheus and the Empedocles complexes, La Psychanalyse du feu goes on to isolate several other complexes around the central archetype of fire: the associations of sex, alchemy, and alcohol. But for our examination of the central paradox of Bachelard's work, the most interesting is the last complex he studies, in which fire is seen as a purifying force, and ultimately as purity itself. We are back at the beginning again. The whole book is based on the assumption that in order to know something about fire we must first know something about the mind that claims to know fire, so that the problem of the knowledge of fire is a true problem of psychological structure. When we see how fire becomes a symbol of purity, how this becomes another of the complexes of fire, we assume some structure of the mind that makes it so, an impulse to purity. Freudian psychology is thus wrong, as Max Scheler had pointed out (P.F., 99), to reduce all love for spiritual things to a shrewdly sublimated sexual love; that is confusing the kernel and the shell, reducing the word “love” to a literal, not a metaphorical meaning. We should consider it rather as a true impulse to purity, and consequently as the impulse to repress impurities.¹⁷ In fact, Bachelard justifies this kind of repression in very un-Freudian fashion; he talks about repression as a joyful activity, as the joy in accepting limitations that is inherent in all learning. The fire symbol has now come full circle in the course of the book; just as it began as the Prometheus complex—the impulse to intellectuality—so it now becomes, at the end, the symbol of joy in learning. It is now time to draw a conclusion from this close connection between fire and knowledge.

Bachelard does not fully realize the implications of his discoveries for his own epistemology. True, he does relate this learning pleasure to his own work:

In our own field of study, through the application of psychoanalytical methods to the activity of objective knowledge, we have arrived at the conclusion that repression is a normal activity, a useful activity, better a joyful activity. There can be no scientific thought without repression. Repression is at the origin of concentrated, reflective, and abstract thought. Every coherent thought is constructed on a system of sound, clear inhibitions. There is a joy in accepting limitations inherent in all joy of learning. It is insofar as it is joyful that a well-founded repression becomes dynamic and useful.

(P.F., 100)

In his joy at this new idea, Bachelard even goes on to apply it to methods of psychiatry. “The truly anagogical cure does not consist in liberating the repressed tendencies, but of substituting for the unconscious repression a conscious repression, a constant will to self-correction.” But he was going beyond himself, and soon realized his mistake; in fact, he says exactly the opposite of this in L'Air et les songes (see A.S., 8). We are back with the original paradox, where the smelting analogy depicted the separating of objective from subjective. The problem can be stated as follows.

All the other subjective values attributed to fire have been dismissed as errors in the pursuit of objective knowledge, of the purification of the object from all subjective impurities. Yet now we find that the pursuit itself of purity, of objective knowledge, is a subjective value attributed to fire; it is our impulse to repress the subjective errors we make in the process of learning. Bachelard's book has been an objective study of subjectivity, that is, he has been treating subjectivity as an object. He has been using the psychoanalysis of fire as a means to expose the subjective errors that the seductive imagination is liable to lead the rational intellect into. But he has now discovered that the impulse to pure objectivity itself is an aspect of the imagination. Now if he has been clearing away as error all the other subjective and imaginative responses to fire should he not now also clear away this impulse to repress impurities as itself subjective, and therefore error? It would seem inevitable that he is led to this conclusion, yet he does not realize it until long afterwards, when he finally adopts the subjective method. Ironically, Bachelard seems to have been aware of this fact, even in La Psychanalyse du feu, though he did not make the connection with his own method. In his conclusion he refers to what later became the central statement of his phenomenology, the need to keep from objectifying oneself: “It is impossible to escape this dialectic: to be aware that one is burning is to grow cold; to feel an intensity is to diminish it; it is necessary to be an intensity without realizing it. Such is the bitter law of man's activity” (P.F., 112).

VI

We have been dividing Bachelard's work into periods, but now the rigid distinctions between the three periods begin to blur. One period grows out of another, there is no sudden reversal. Just as we saw the gradual transition from one study to another as we moved through La Psychanalyse du feu, so we can now see the confession of subjectivity that he makes at the outset of his final period (in the Introduction to La Poétique de l'espace) adumbrated in his second book on the material imagination in 1942, which is called, significantly, L'Eau et les rêves, and not La Psychanalyse de l'eau. He claims to be a rationalist, but if we listen carefully, as Bachelard would always have us do, we can hear the notes of his final period:

Rationalist? I am trying to become one, not only in our culture as a whole but in the detail of our thoughts, of our imaginations. By a psychoanalysis of objective knowledge and of “imagified” knowledge, I became a rationalist in my view of fire. Sincerity obliges me to confess that I have not achieved the same objectivity with water. The images of water I live again synthetically in their first complexity, in often giving them my irrational adhesion.¹⁸

Between the beginning and the end, indeed, between the study of the intellect and the study of the imagination, there is an absolute polarity. From his last period, he looks back on his first period, the psychoanalytical period, with aversion. It is called “mere psychoanalysis,” that criticism which makes a man out of a poet, but does not explain how to make a poet out of a man (P.R., 9). Nor does it explain what we have been calling the “why” of poetry.

So we were oversimplifying when we divided his three periods into the objective study of objectivity, the objective study of subjectivity, and the subjective study of subjectivity. But there is this basic movement from one extreme to the other, and we must see it as a progressive attempt to be honest about poetry. It is a movement towards phenomenology, because phenomenology is the only method which truly faces the reality of the poetic experience. It refuses to become objective to itself and thus destroy the unique validity of the poetic essence.

We have been examining the treatment of fire in Bachelard's objective method. By way of contrast, we can look at an example of how he talks subjectively about a literary fire in his last major work, La Poétique de la rêverie.¹⁹ Bachelard has just quoted a beautiful passage, describing a fireside reverie, from the novel Malicroix by his favorite author, Henri Bosco. Though the method of this final period is subjective, we need to see that Bachelard still emphasises the communicability of the subjective experience.

What other pages does one need to contemplate in order to understand that fire inhabits the house? In the utilitarian fashion one would say that fire makes the house habitable. An expression like that belongs to the language of those who do not know the reveries of the word “inhabit.” Fire transmits its friendliness to the whole house, and thus makes the House into a Cosmos of warmth. Bosco knows that, says that: “The air expanding in the heat, filled all the corners of the house, pressing against the walls, the earth, the low ceiling, the heavy furniture. Life flowed around from the fire to the closed doors and from the doors to the fire, tracing invisible circles of heat that floated across my face. …”

Someone will object, perhaps, as he reads this page, that the writer is not telling his reverie, but that he is describing his well-being in a closed room. But let us read better, let us read and dream, let us read and remember. It is of ourselves, dreamers, of ourselves, that the writer is speaking. With us too, the fire has kept company. We too have known the fire's friendship. We communicate with the writer because we communicate with the images we keep in the depths of our selves. We return to dream in those rooms where we knew the friendship of the fire.

(P.R., 167f.)

The difference between this and his earlier approach, both to fire and to literature, is obvious. Far from dismissing the fireside reverie as the source of subjective errors which get in the way of our objective knowledge about fire, he is treating it as an example of the supreme poetic experience—the activity of the poetic imagination. Reading the literary description of a reverie, Bachelard partakes in that reverie. He describes it to us, and urges us to take part in it too, by the activity of our own imaginations. He wants us to share the happiness that comes from this kind of unified reverie. One is no longer lonely, no longer suffering existential Angst, or the sense of division within the self. “Outside time, outside space, before the fire, our being is no longer chained to a Dasein.” In reverie we flow with the well-being of the world, a well-being that “teaches me to be the same as myself” (P.R., 166).

This new kind of fireside reverie duplicates the poetic experience. It gets its importance partly from the sense of cosmic union, where “the soul is no longer ensconced in its corner of the world. It is at the center of the world, at the center of its world. The simplest hearth enfolds a universe” (P.R., 166). Thus Bachelard claims that the state of poetic reverie is healthful partly because it abolishes the distinction between subject and object, between the self and Nature. That distinction is a product of the intellect. He sounds very like Coleridge here, with the “coalescence of subject and object” through the activity of the imagination. What Bachelard is doing is taking over one of the central notions of the phenomenological movement (which itself owes much to the Kantian tradition which Coleridge drew upon): reality is an aspect of phenomena, it does not subsist in a division of the perceiver and the perceived. We need to notice how different this is from the Cartesian methods of his early period, which involved arriving at objective knowledge by a ruthless dismissal of all imaginative responses to the object. Where the initial desire was to separate subject and object, this final period praises their union. The former was a destructive attempt, but the latter describes the creative state.

One could not say of the man of reverie that he is tossed into the world with a jolt. The world is entirely friendly to him, and he himself is the principle of friendship. The dreamer is the double consciousness of his well-being and of the happy world. His cogito is not divided into a dialectic of subject and object.

The correlation of the dreamer to his world is a powerful one. … To doubt the worlds of reverie one must stop dreaming, come out of one's reverie. The man of reverie and the world of his reverie are very close, they touch, they penetrate each other. They are on the same level of being. … I dream the world, therefore the world exists as I dream it.

(P.R., 134f.)

Even here Bachelard still takes account of that Cartesian suspicion, that doubt which one might feel outside the poetic reverie. But that doubt is an activity of the intellect, not of the imagination. Bachelard still wants to separate the two, but the emphasis is reversed. Now, in this subjective period, he wants to stop the intellect intruding upon the proper field of the imagination. This is the complete opposite of his earlier interest. His final denial of the objective method of literary criticism is closely related to this desire to keep the intellect out of the field of imaginative experience. The intellect functions properly only when it is being truly objective; thus its opposite, the imagination, can only truly realize itself in absolute subjectivity.

Evidently Bachelard recognized the impossibility of achieving either ideal, but he still felt the importance of trying to maintain the polarity of intellect and imagination. He tried not to make any comparative value judgments between them and he was happy in either pursuit, but in view of this kind of cosmic happiness which can be the result of imaginative experience, one can understand why he came to concentrate more and more on the studies of the imagination. The last years of his life were devoted to pursuing the happiness of the imagination, and in the quality of the two kinds of happiness we can see his preference for imaginative activity. The joy in learning which he had talked about in La Psychanalyse du feu was a repressive joy, it narrowed itself to meet the purity of its object. But the joy of La Poétique de la rêverie is an expansive joy, it flows outward to become as large as the cosmos. It has no object, for in this world there is no “non-I.” The only way to be faithful to this experience of poetic union is to be a poet oneself, to repeat the subjective activity of the poet through the methods of phenomenology. In order to understand the unique world of the poet, you have to live that world yourself.

Notes

1. Gaston Bachelard, L'Air et les songes (Paris, 1943), p. 7. Future references are included in the text, and abbreviated as A.S. All translations are my own except for citations from the two works already translated into English (see below).

2. Georges Poulet, The Interior Distance, trans. Elliott Coleman (Ann Arbor, 1964), p. viii.

3. Cited in C. O. Christofides, “Bachelard's Aesthetics,” Journal of Aesthetics and Art Criticism, XX (1961-62), 268f.

4. Bachelard, La Terre et les rêveries de la volonté (Paris, 1948), p. 6. (Abbreviated as T.R.V. in future.)

5. Bachelard, La Poétique de l'espace (Paris, 1957). References are to the English translation by Maria Jolas, The Poetics of Space (New York, 1964), and abbreviated as P.S.

6. Bachelard, La Psychanalyse du feu (Paris, 1938). References are to the English translation by Alan Ross, The Psychoanalysis of Fire (Boston, 1964), and abbreviated as P.F.

7. This optimism is a dominant characteristic of Bachelard's thinking. Though there are times when his ideas sound a little like Sartre's, especially “the function of the unreal,” he has none of Sartre's horror at what he takes to be a totally meaningless existence. Bachelard's reality is warm and friendly, like the fireside reverie which becomes so important for him.

8. In the tape-recording cited note 3.

9. Bachelard, La Formation de l'esprit scientifique (Paris, 1938), p. 23.

10. Georges Poulet, “Bachelard et la conscience de soi,” Revue de Métaphysique et de Morale, LXX (1965), 4. (I have drawn considerably on this article for my argument.) Bachelard actually wrote a book called La Philosophie du non (Paris, 1940).

11. Poulet, “Bachelard et la conscience de soi.”

12. He had wanted to teach literature, but was not allowed to since he held the chair of the philosophy of science, and the department could not have its man teaching literature.

13. We need not be concerned that the imagination of movement is also the dynamic imagination. Similarly, the material imagination is also the imagination of matter. Bachelard means exactly what he says—the materialization of the imagination. He is talking about a world in which there is no distinction between subject and object, nor between subjective and objective genitives.

14. See Mikel Dufrenne, “Gaston Bachelard et la poésie de l'imaginaire,” Etudes philosophiques, N.S. XVIII (1963). This would mean that poetry has nothing to do with dead metaphors or the vulgar commonplaces that occur in everyday conversation. Bachelard would exclude, with Mallarmé, “les mots de la tribu.” Compare this passage from La Poétique de la rêverie, p. 137: “What a relief the poetic image is in our language! If only we could speak this lofty language, climb with the poet into that solitude of the speaking being which gives a new sense to the words of the tribe, then we would be in a realm where the active man cannot enter, the man for whom the world of reveries is ‘just a dream.’” There are many important affinities between Bachelard and symbolism; he too is talking about an image which both elicits and requires the activity of the reader to complete it.

15. La Poétique de la rêverie (Paris, 1960), p. 45f. (Abbreviated to P.R. in future references.)

16. Bachelard recognized its incompleteness later, in the tape-recording previously cited. He called it “ragged.”

17. The word I am translating here as “true” is “propre,” not “vrai,” but I think the implication is clear enough.

18. Bachelard, L'Eau et les rêves (Paris, 1942), p. 3.

19. Fire remained one of Bachelard's chief interests, and his final work, La Flamme d'une chandelle, began as an attempt to write a poetics of fire, and thus return to the subject which, by alerting him to the importance of reverie, had begun his literary studies so long ago.