Who is the Subject of Enunciation?

John Phillips

10/02/2006

 

A useful starting point when answering this question would once again be a consideration of the function of language.  The common sense response to the question of what language is for is likely to be that language functions as an instrument of communication.  This common sense response, which is not so much a matter of universal perception but more a symptom of the age, is given powerful support by the status of knowledge generally and scientific knowledge in particular, which treats language, as it treats everything else, as something that can be made into an object of knowledge.  These assumptions, that language is an instrument of communication and that it can be made the object of scientific knowledge, follow apparently quite naturally from the most pervasive attitudes that govern knowledge.  Knowledge is attributed to a knowing subject who is capable through his reason of making judgments about his objects.  One of the tools or instruments at the scientist’s disposal would therefore be his means of communication, without which this special knowledge would have no means of dissemination.

 

These assumptions are questionable on a number of grounds.  What is at stake here would be extremely powerful because, if language turned out not to be just an instrument for communication, and if it was not possible, after all, to turn language into an object of science, then we’d be faced with a discourse that was more powerful and more fundamental than scientific discourses are currently capable of comprehending.  We would need to develop an alternative attitude to knowledge. 

 

Language would be perfectly fit as an object of science only if it was reducible to the instrumental function of communication.  And, in empirical and formal terms, that is exactly what language seems to be.  We might question its efficiency and we might question whether it is entirely suited to this function.  But so long as we have an instrumental attitude towards it, language provides us with what we’re looking for: an instrument.  There is something peculiar in this fact.  The way we use language tends to encourage us to see it as essentially what we use it for.  If we use language as an instrument then it certainly seems to be an instrument.  When we communicate (I send a message which is correctly understood and perhaps acted upon) language certainly seems to be an instrument for our communication.  So perhaps, when science takes language as an object of science and establishes a theoretical or a practical linguistics, all that it is capable of studying is the way language is used.  The only object that linguistics knows under the term language is the instrument of communication that science uses in order to disseminate its knowledge.

 

A number of developments in linguistics and philosophy have made it difficult to maintain these assumptions.  In Linguistics first, Ferdinand de Saussure’s Course in General Linguistics seems to have had two rather profound long term consequences.  They correspond to a distinction that he makes early in the lecture course, between langue and parole (a distinction that will have powerful influence on the development of structuralism and semiotics).  Langue represents the “work of a collective intelligence” [l’œuvre de l’intelligence collective], which is both internal to each individual and collective, in so far as it is beyond the will of any individual to change.  Parole, on the other hand, designates individual acts, statements and utterances, events of language use manifesting each time a speaker’s ephemeral individual will through his combination of concepts and his “phonation”—the formal aspects of the utterance.  Saussure points out here that the single word “linguistics” therefore covers two radically different kinds of study.  The study of Parole would be entirely focused on individual utterances, using all the available resources of formal and empirical study to analyze—usually within a specific language—actual statements.  Whereas the study of langue would be focused entirely on language generally, which means the most basic and universal conditions of possibility for any language and any use of language whatsoever.  His Course in General Linguistics thus follows the second route in this inevitable “bifurcation,” setting out in what has become a decisive historical event in knowledge, the groundwork for all attempts to grasp the basic conditions of possibility for language and language use generally.                        

 

Firmly within this tradition, the French linguist Emile Benveniste is responsible for outlining the need to make a distinction between what he calls the subject of the énoncé and subject of the énunciation.  In two influential arguments Benveniste focuses on the role and implications of the ubiquitous first person pronoun (and its reciprocal second person), used at least implicitly in every language known to man and woman.  In “On the Nature of Pronouns” he notes that the first person, “I,” operates in a way quite unlike other pronouns because it is essentially linked to the exercise of language.  In other words, the sign I links Saussure’s two dimensions of language, the collective intelligence of langue and the ephemeral individual acts of parole: “it is this property that establishes the basis for individual discourse, in which each speaker takes over all the resources of language for his own behalf” (220).  In fact the I not only links the otherwise heterogeneous dimensions of langue and parole but it also keeps its speakers unaware of this profound difference.  What is peculiar about the signs I and you is that they are essentially empty of meaning except when they are being used.  So the reality to which I or you refers is solely a reality of discourse.  They refer to nothing but the fact that someone is speaking or has spoken (and nothing changes when we consider fictional or reported dialogue).  Benveniste states the precise definition for I as follows: “I is the individual who utters the present instance of discourse containing the linguistic instance I” (218).  By taking the always implicit and often explicit situation of “address” into account, one has the symmetrical definition for you: “the individual spoken to in the present instance of discourse containing the instance you.”  Now after Saussure we know that all signs are intrinsically empty of meaning, which is determined only in the repetitions of institutions, systems and events.  However, I and you are instances of signs that lack even the possibility of material reference.  These signs cannot be misused because they “do not assert anything, they are not subject to the condition of truth and escape all denial” (220).  The implications are far reaching.  First by indicating the situation of the speaker yet by escaping the conditions normally attributed to language (especially when it is regarded as an instrument of communication), the pronoun tells us something about the relation of the human animal to the language she speaks.  Language is not something the human subject uses (as Rene Descartes and the traditions of modernity that follow his lead had always asserted), but rather, the human subject is something only made possible by language.  In his 1958 article, “Subjectivity in Language,” Benveniste underlines this point:

 We are always inclined to that naïve concept of a primordial period in which a complete man discovered another one, equally complete, and between the two of them language was worked out little by little.  This is pure fiction.  We can never get back to man separated by language and we shall never see him inventing it … It is a speaking man whom we find in the world, a man speaking to another man, and language provides the very definition of man. (224).

We probably should be a little careful here, because when Benveniste says that language provides the very definition of man, we mustn’t assume, with theoretical linguistics, that we know what language is.  At this stage language provides us with the definition of man only because of the peculiarity of personal pronouns.  The foundation of “subjectivity” is determined, according to him, by the linguistic status of the person:

Consciousness of self is only possible if it is experienced by contrast.  I use I only when I am speaking to someone who will be a you in my address.  It is this condition of dialogue that is constitutive of person, for it implies that reciprocally I becomes you in the address of the one who in his turn designates himself as I” (224-225).

So the basis of subjectivity, if we take language as a model, would not be those aspects that constitute either its lexical content (meaning) or its formal and grammatical rules, but it would only be discoverable in the exercise of language.  It thus becomes necessary to recognize an irreducible division corresponding to that between enunciation and statement (énoncé).  The subject of the statement seems fixed in time, a snapshot of a moment that has immediately passed, already fading in its enunciation.  The speaker is already in principle out of the picture and all that remains is his representative in language.  What this means is simply that subjectivity comes into being in language alone and that, in speaking, the human subject is irreconcilably divided in himself.  A temporal disjunction between the subject speaking (enunciation) and the subject represented in speech (statement) implies that with the single pronoun I, there are always at least two subjects: a subject who is speaking and a subject represented in speech.  By focusing on one we necessarily lose sight of the other.  There are instances that bring this situation to light rather obviously.  The old paradox of the Cretan Liar provides a fine example.  When someone says “I am lying,” the I must refer to a different subject than the one who makes the statement.  When someone says “I am dead” a similar situation arises.  The I in principle (and thus in fact) lives on beyond the I who speaks.  This is easily demonstrated by the fact that the meaning of the statement is the same whether it is true or false at the moment of utterance and is destined to be true anyway independently of any individual speaker or writer.  But it is this “at the moment of utterance” that loses its anchor once we begin to focus on the modality of personal address.  Benveniste reminds us that “linguistic time is self-referential” (227).  The eternally present moment is an illusion that covers up or sutures the fundamental disjunction in language according to which a present moment (the moment of utterance) can only ever appear as a representation (the statement).

 

Benveniste’s distinction plays a decisive role in the work of Jacques Lacan, Roland Barthes, Julia Kristeva and Michel Foucault, who are some of the “names” we associate with the category of critical theory called poststructuralism.  For Lacan this distinction in language corresponds exactly to Freud’s distinction between consciousness and the unconscious.  For Lacan, since the subject comes into being through language he does so through the exercise of signifying articulation—the act of enunciation.  As soon as he comes into being he finds himself not as he is (what Lacan would call the truth of his being) but as he imagines himself to be—that is as a representation (at the level of the statement).  In order to discover the subject of the unconscious the analyst must focus on the level of enunciation (performance, expression)—in order to recognize the truth of the subject in the articulation of language—its enunciation.  Lacan puts it like this: “In order to be situated in the locus of the Other, the presence of the unconscious is to be sought in any discourse in its enunciation” (Ecrits 834).  So the relation between statement and enunciation (the said and the saying) actualizes the divided structure of the psychoanalytic subject and helps us further to grasp the difference between the imaginary (fixed and complete image of person) and the symbolic (the constitutive function of language).

 

Roland Barthes explicitly draws attention to the imaginary function of the I in classic realist fiction in his S/Z.  He draws attention to the use of the personal pronoun as character forming and rethinks the distinction énoncé/enunciation as that between a character (for traditional readings) and a figure:

In principle, the character who says “I” has no name (Proust’s narrator is an outstanding example); in fact, however, I immediately becomes a name, his name.  In the story (and in many conversations), I is no longer a pronoun, but a name, the best of names: to say I is inevitably to attribute signifieds to oneself; further, it gives one a biographical duration, it enables one to undergo, in one’s imagination, an intelligible “evolution,” to signify oneself as an object with a destiny, to give a meaning to time.  On this level, I (and notably the narrator of Sarrasine) is therefore a character.  The figure is altogether different: it is not a combination of semes concentrated on a legal name, nor can biography, psychology, or time encompass it: it is an illegals, impersonal, anachronistic configuration of symbolic relationships.  As figure, the character can oscillate between two roles, without this oscillation having any meaning, for it occurs outside biographical time (outside chronology): the symbolic structure is completely reversible: it can be read in any direction … As a symbolic ideality, the character has no Name; he is nothing but a site for the passage (and return) of the figure. (S/Z 68).

All of S/Z’s polarities can be situated on the model of énoncé/enunciation.  What is revealed, if anything, is that, above the bar (on the level of the statement) we find the sum total of determinations, institutions, codes and systematizations—the whole sedimented world of the statement and its theoretical conditions of truth and falsity.  Beneath the bar, however, we find the conditions of discourse itself in an essentially empty sign.                          

 

When a person reads a text then they adopt a position—a place or a space in a metaphorical sense—that is always partially determined but also always partly undetermined too.  The relation between the determined and the undetermined in this sense constitutes the difference between the statement and the enunciation.  Now while it is possible for me to instruct you in the determinate aspects of reading (and we have done some of this): form and content, history, biography, reader’s interpretive community, cultural coding or ideology etc., it would not have been possible for me to point to—point out—define or delimit—the indeterminacy that always necessarily accompanies all determinations (meanings, decisions, choices, relations etc.), as both their condition and their effect.  In other words I’ve been trying to draw attention to an essential component of both writing and reading that is strictly not visible or intelligible because it is best understood as a form of possibility (the most basic component of the experience of time is “future”).  A person and a text meet at the level called “subject of enunciation” but only in so far as this meeting is always “yet to come.” 

 

So if a person reads a text, then they adopt a position that has in advance been made possible.  What are the conditions of possibility for reading?  The simultaneous effects of instituted meanings (repetitions and regularized relations) and the undetermined (your possibility of reading).  Now a significant mark--granted significance by its repeatability (a fundamental source of what I called possibility just now)--must, before it is determined by system or code or decision as meaning this or that, always also possibly means nothing at all.  So the significant mark is also marked by a certain insignificance; to put it less enigmatically--its meaning is to an extent “not yet determined.”  By being “not yet determined” it is readable and our interpretation has a chance beyond instituted or established modes of interpretation. The nothing at all is an a priori condition for all language because if this was not the case then words would simply mean what they mean and that would be it.  Each determination would be a one off.  And we’d need a unique mark for each event.  As we know we can use a mark like “cat” infinitely many times and even with this very simple example it can have many possible determinations (or meanings).  These conditions are what give us the “subject of enunciation.”

 

So the subject of enunciation is strictly not determined or determinable.  Think of it as a projection of the ideal addressee.  But the ideal—an addressee who would be the final reader—the reader whose reading finishes off the text and its reserves for once and for all—is impossible because there seems to be no limit to the possibilities of a text reappearing in new contexts and meeting new addressees, possibly or even essentially accidentally, ad infinitum.

 

The Archie Bunker scenario that Paul de Man refers to in his “Semiology and Rhetoric” article (see my commentary on de Man) serves as an example that demonstrates the rhetorical underpinning of language use.  The meaning of Archie’s rhetorical question, “What’s the difference?” made in response to his wife, who has asked the actual question, “do you want your laces tied over or under?” can only be decided.   There is nothing formal or empirical about the rhetorical question that determines it as either rhetorical (there’s no difference) or literal (please tell me the difference).  No computer, no matter how sophisticated, has yet been programmed in such a way that it could decide this kind of thing.  However, neither Archie nor his wife could be said to be “subject of enunciation.”  That would rather be indicated by the conditions that the exchange demonstrates.  And neither you nor I will ever “be” subject of enunciation.  Rather, “being” is what the subject of the statement represents in its maintenance of the imaginary—radically, there just “is” no being.  The example can serve, as it does for de Man, as a relatively trivial and thus manageable illustration of how the rhetorical foundation of enunciation can subvert what we might call the standard or dominant reading of a text.  A standard reading would be the one that a text seems to be proposing for its own interpretation.  All texts can be read at the level of enunciation insofar as an at least implicit I appeals to an at least implicit you.  In the Archie Bunker case two interpretations of interpretation are proposed by the text: 1) the rhetorical reading (Archie—even though he may be an unwitting rhetorician in this case) and the literal or grammatical reading (his wife—who is also unwitting in this case).  This mutual unwitting-ness would perhaps be a component of the comedy.  So a certain narrative irony kicks in and at the level of enunciation a rhetorical chaos underlying instituted meaning is revealed.  Because of the narrative irony we are in fact constrained to see the narrator or persona or whatever as a fictional construct.  Radically, of course, such self-construction can be read as a commentary on what we all do with our “selves.”  But subject of enunciation need not always be located at the level of who speaks.  This is especially true of dramatic monologue, which usually operates by exposing the fictional persona’s own conditions in interesting ways for critical reading.  In this way literature is sometimes already doing the work of deconstruction—though each time we’d have to clarify how.  And actually more often the “framing” of the persona can reveal assumptions at the level of the text (about truth and falsity, say) that the text itself may contest.  Browning’s “My Last Duchess” frames a tyrant Duke but replicates the terms of the Duke’s tyranny in its own framing rhetoric. 

 

The moment that we act on our possibilities as readers (i.e., we determine a text as meaning this or that) we betray those possibilities.  They are what give us the power of interpretation or determination—human agency, or freedom, to use an old fashioned term.  But that betrayal is itself always immediately betrayed too by the automatic generating of further indeterminacies, which necessarily follow every determination.  If a text can be read as proposing a theory or account of interpretation (a theory of truth, for instance), then the conditions of interpretation that can be read at the level of address, if only because someone is reading a text, will be in an automatic relationship to it, which might always be the source of a profound contestation.  One way to attempt to avoid these effects might be to simply ally oneself with the police (and you might guess, correctly, that I am not proposing that we do)—those institutions which govern interpretations, meanings and truth by policing through exclusion or containment aspects of the text that threaten the lawful and legal institutions themselves.  This is what some would call the mystical foundation of authority and this is the main target of most deconstruction.

 

Deconstruction in this understanding would not be what the reader does to the text.  The conditions that we have called deconstruction would apply whenever an utterance and its enunciation occur.  The various kinds of reading that have come to be called deconstruction (or even deconstructionism for those who like to be able to identify movements and schools and who thus require the suffix of an –ism) would certainly be engaged in exposing the conditions according to which a system or ensemble sets itself up against its own conditions of possibility, its enunciation, and this activity of reading, this doing deconstruction, would thus have the effect of meddling with established horizons of thought and action at their very source (which would anyway have been an illusion). 

 

One thing that could barely be avoided, then, would be the need to establish these conditions through the practice of, or at least a meditation on the problems of, reading.  The establishment of person and the basis of subjectivity, we have learned, must be located in the oscillations of the empty signs I-You, which simply means that a potentially active (though often passive) reading position is always implied.  Christina Rossetti’s poem, “In an Artist’s Studio,” seems to be addressing us (and I use the pronoun advisedly) with statements of the kind that tie into the concerns that we started with: a consideration of the function—not of language simply this time—but of representation generally.

One face looks out from all his canvases,

One selfsame figure sits or walks or leans:

We found her hidden just behind those screens,

That mirror gave back all her loveliness.

A queen in opal or in ruby dress,

A nameless girl in freshest summer-green,

A saint, an angel—every canvas means

The same one meaning, neither more nor less.

He feeds upon her face by day and night,

And she with true kind eyes looks back on him,

Fair as the moon and joyful as the light:

Not wan with waiting, not with sorrow dim;

Not as she is, but was when hope shone bright;

Not as she is, but as she fills his dream.

But first, how do I justify singling out one poem from thousands of possible examples of discourses that address the conditions of their enunciation?  If what we have learned so far is correct, then there would not be a text without enunciation and thus no text that did not address its own conditions.  The scientific text, in its practice of deliberately leaving out all references to enunciation (“This experiment will establish … ”), nonetheless exhibits an attitude to its own enunciative modality, which we might think of as negation or denial.  However, my decision to locate what we call a literary text was informed by the nature of what we might call the institution of literature.  In an era where the dominant notions of truth imply a relation between statements and states of affairs the role of literature achieves a peculiar status.  It would not be possible to establish answers to the question: “What is literature?” without at least addressing its peculiar position amongst other discourses (the relationships that Barthes examines in “Science and Literature,” for instance).  I would guess that all positive identifications of literature are bound to fail.  Literature is distinguished fundamentally by its status in the multiplicity of different discourses and the special values attributed to it by virtually everyone, whether they read it or not and whether they are professional critics and professors or writers of it (and sometimes both at once).  The notion of “poetic license” exemplifies the situation.  The poet can (almost literally in the legal sense—she is allowed to) write statements that are free from the normal constraints of the correspondence theory of truth.  So the notion of poetic license serves as a way of legalizing (metaphorically) a certain abuse of truth.  Rather than contest these standard and dominant notions, then, literature’s truth actually supports them.  It is the standard notion that grants the license, thus revealing legalization as a kind of domestication or even pathologization of a discourse that—if it had ever managed to infiltrate the active discourses of social reality—would have created merry hell with their notions of truth and knowledge.  So whether literature is revered or despised it nonetheless nearly always plays the same role—it is allowed to say whatever it wants because this is the only institution that has been granted that license. 

 

The difference, then, between science and literature in some fairly exact ways replicates the enunciative dimension we indicated earlier that sets the subject of knowledge up over and against his object.  Literature is regarded, despite protestations to the contrary, as something subjective—in both its expression and reception—while knowledge in the true sense of the word would supposedly be objective.  The objective criterion of judgment would be the truth or falsity of the correspondence between statement and state of affairs, whereas the truth of literature would have to be regarded as that of a discourse representing subjective thoughts, feelings and experiences.  These discourses—supposedly separated and qualitatively different—merge when we turn to the dominant practices (maintained in schools and universities around the world) which must take literature as their object of knowledge.  The establishment of practical criticism at the beginning of the nineteenth century and its general deployment in Literature Departments as various kinds of formalism throughout the twentieth century demonstrates the fundamental complicity between the discourses of science (including social science) and those of the arts.  Practical criticism sets out to analyze the relationships between ideas and poetic devices, as the following statement from a late twentieth century foundation course in critical reading confirms:

It is wrong to think that the reading of literature is just a matter of basking in the emotional glow of a “great work”.  Equally misguided is the idea that literature itself is concerned only with the emotions and has little to do with thought.  Popular opinion concerning literary writing sometimes assumes the existence of a sharp distinction between emotion and judgement, feeling and thought, but this is to ignore the close relationship which exists between ideas and poetic devices.  Metaphors, similes and rhythmically organized words are not simply vessels for carrying fine feelings, but can equally serve to convey thoughts and ideas. (Critical Reading 1).

And so the first thing that undergraduate students would learn on this course is that critical reading must be a matter of establishing the relationship between content (ideas) and form (poetic devices).  The second thing they would learn is that literariness (in the sense just given) “is an inevitable presence in any writing that seeks to persuade,” including perhaps “‘scientific’” books.  Nothing here departs from assumptions that have their basis in the failure to recognize or accept the modality of enunciation: a clear distinction between the formal and ideal aspects of any text; the establishment and identification of formal devices as tools of communication; the privilege of the observer skilled in making judgments about his object; the unquestioned faith in the empirical substantiality of the object itself.  But what, in the modality of enunciation, does the course do?  Having established that poetic language can be found in any kind of discourse, thus effectively neutralizing anything that might have served to mark the literary out against other kinds of discourse, it establishes a ground for transferable skills and knowledge in the discipline of literary study, claiming the literary as a privileged foundation for critical reading skills that can be applied generally.  This simultaneous neutralization and generalization of the literary in literary departments indicates in a minor way the sad destiny of literary study.  In obvious ways, this text, which emerged in the 1990s, rescues the common sense of a practical criticism from a perceived threat represented by the word theory, which had shadowed the practices of literary study from at least the late 1950s yet now resides, safely contained within those same departments, as the “advanced” or “critical theory” modules that sometimes students are required to take.  The opening statement of the Course bears this out: “our chosen method should [my emphasis] focus intensely upon reading rather than upon more theoretical issues.”  Our own practical reading skills invite us to seize on the modal should (supported by “it is wrong to” and “misguided is the idea”) in order to ask, by what authority?  Given that forty years of literary theory preceding the publication of this book has at the very least put into suspense the form and content and author/reader models of writing and reading, their blithe perpetuation in foundation courses of literary study might justifiably be regarded as an interesting symptom or function of a less visible condition. 

 

At the level of its statement practical criticism takes the statement itself as its object of study (form and content, sense and reference), thus supporting all its own assumptions by finding support for them in the text, which in its usual generous way reflects the approach of its readers.  But on the level of enunciation practical criticism continues to support the very dimension that the difficult critical theory might have created potential problems for: the most pervasive discourses of knowledge and an increasingly commoditized culture of knowledge.  The neutralization of the literary is the neutralization of theory—which, domesticated within literary departments, poses no threat to the horizons of thought and action that govern the historicity of the age.  The idea that “reading intensely” can come before “more theoretical issues” is an old one and is supported by the ideal of the “unseen” passage—testing students to see how they “respond directly to the text itself without interference from what might be irrelevant and inaccurate preconceptions and prejudices” (Critical Reading 57).  But the principle of direct engagement between subject and object, and the principle that relates meaning to the poetic devices supposed to be their vehicle, constitute profoundly theoretical (and thus questionable) positions.  On these foundations we’ll never learn what happens when real critical reading is attempted, that is, reading that establishes its principles of reading from the untiring attempt to read.

 

So my turning to a work of literature is informed not just by historical contingencies that maintain departments of literature as neutralized or licensed zones for ungrounded knowledge, but also by the practices of those departments themselves, which fulfill that neutralization in their teaching and profession.  I turn again to “In an Artist’s Studio.”  The aim is to establish as far as is possible a relationship between the statement and its enunciation.  I say “as far as is possible” in the knowledge that this is strictly impossible.  The enunciation—the saying of the said or the process of manufacture that produces the statement—would have already passed and always be yet to come, in terms of its temporality—the division between enunciator 1 and enunciator 2 only giving rise to the perpetual present of the utterance by withdrawing from it.  Our reading inserts itself—for this is its only chance—as it were, between enunciations.  But there are certain things that we may look out for.  We can establish thanks to the access that each one of us has to our language what the reading or interpretation proposed by the text seems to be.  This aspect can be relatively loose, even ambiguous, depending on the text, but it does not require the thought of the extra-textual existence of an author.  It does require our accepting that there will be aspects of a text that lie quite outside the control of an author and which, therefore, may be in contest with the way the text proposes we should read it.  The enunciative function, found at the level of address and presupposing the figure of an I, can only manifest itself as a self-reflexive component.  It will not take a great deal of difficult training to begin to locate moments that thematize the process of interpretation itself, through all kinds of devices, whether intended or not but including those identified by practical criticism.  That way our reading can be folded over—applied to—the notions of reading thematized and probably proposed but at least implicitly suggested by the text itself.  We can also allow ourselves considerable critical reflection on the constraints on and possibilities available for our reading this text.                

One face looks out from all his canvases,

One selfsame figure sits or walks or leans:

We found her hidden just behind those screens,

That mirror gave back all her loveliness.

A queen in opal or in ruby dress,

A nameless girl in freshest summer-green,

A saint, an angel—every canvas means

The same one meaning, neither more nor less.

He feeds upon her face by day and night,

And she with true kind eyes looks back on him,

Fair as the moon and joyful as the light:

Not wan with waiting, not with sorrow dim;

Not as she is, but was when hope shone bright;

Not as she is, but as she fills his dream.

This sonnet by Rossetti establishes a situation using the resources that language makes available: person (a speaker—marked by the use of the plural personal pronoun) and deixis (those screens, that mirror) establishing a sense of space.  The speaker comments on the situation and provides what will turn out to be a reasonably elaborate interpretation of it.  We might pay particular attention to tense.  The relationship between past and present would be worth analyzing.  I’m not interested in what we can establish objectively about the relationship between what the poem “says” and how it says it.  I need no ghost of practical criticism come from the grave to tell me this.  If I speak English I know what the poem says.  The speaker observes many paintings depicting the same face and figure differently and observes that the figure in the pictures is more lovely and joyful than their single model seems now—all wan and sorrowful.  Its meaning is quite clear independently of any extra-textual knowledge that might help to illuminate it, such as the note provided in the Everyman edition of the Poems and Prose: “believed to be based on a visit to DGR’s studio, and his many portraits of Elizabeth Siddal.  In 1854 Ford Madox Brown remarked on the many ‘wonderful and lovely’ drawings of Siddal by DGR.  At the date of this poem she was wintering in the South of France for health reasons” (Poems and Prose 433).  The note is like a second text.  Oddly it requires illumination from the poem rather than vice versa.  The relevance of the model’s convalescent vacation requires the poem’s commentary on the state of the model.  Without directly proposing it, however, the note does suggest an interpretation: Rossetti is writing about her brother’s pictures of this currently sick woman and is thus suggesting a cause for her sickness in the artist’s conduct.  The note encourages an interpretation of the poem that would read it as an interpretation of the relation between the artist and his model.  So we don’t actually need to know the biographical details.  In this case they turn out to be irrelevant.  However the suggestion does take us to an interesting discovery.  The present tense (Rossetti is writing) ought really to have been put in the past tense (Rossetti was writing …).  In fact the poem creates the illusion of the present tense, which immediately prizes the text out of what we might have imagined was its biographical setting.  A basic condition of possibility for reading (and thus for writing) would be this establishment—again at the level of enunciation—of an imaginary present moment of reading.  What this means is simple: we can imagine the biographical moment or we may just as well choose to ignore it—but if it wasn’t for the latter possibility (that we remain ignorant of biographical details) we would not have been able to imagine the former (the imaginary biographical setting).  This is how the statement—and the withdrawal of enunciation—works. 

 

So the message of the poem—its statement—serves as a kind of interpretation of the paintings: they depict a sick woman as if she was joyful and lovely—either as she once was or as she is imagined to be even now by the painter, despite appearances.  At the level of enunciation it should be possible to establish the principles of interpretation that the poem proposes.  First the question of interpretation focuses on a single element, “one face … one selfsame figure … the same one meaning, neither more nor less,” which is singularly expressed or represented by a diversity of texts.  We might well ask: expressed or represented?  What is the nature of the relationship between the text and its meaning?  Not by chance are these the central questions of literary criticism.  The first four lines give us some options:

One face looks out from all his canvases,

One selfsame figure sits or walks or leans:

We found her hidden just behind those screens,

That mirror gave back all her loveliness.

The canvases depict a face that “looks out.”  The face belongs to a “her” who “we found … hidden … behind … screens.”  The pictures are “that mirror,” which provides the loveliness that is missing from what it depicts, which remains hidden yet found behind screens.  Each time the canvases are related (by similarity or contrast) to another kind of surface.  The selfsame figure that looks out of each of the canvases is hidden behind a screen.  Yet the canvases reflect the figure by providing more than mere reflection: they replace what is missing in the model.  Three choices then are set up.  The first remains enigmatic, a peculiar phenomenon, the multiple figures of a singular gaze.  These are not pictures merely to be looked at.  These are pictures that look out at you.  The second option, indicated by the function of the screen, is that the text hides its meaning behind an alternative one, its reference behind its sense.  Whether the model is absent or present—whether we find her hidden or find her hidden, the canvasses keep her hidden, for these are not her in truth.  The third option is that the canvases function as a kind of reflection but as in a distorted mirror: this is her each time but not as she is, not as we know her now.  The combination of canvas (gazing out), screen (concealing) and mirror (reflecting and distorting) provides already a complex account of the functions of a text.  A text is something that fixes you while concealing the true nature of the subject, yet reflecting the subject by adding what is missing from it.  Each time, it is worth noting, a singularity, the “one true meaning,” is enigmatic, hidden, or distorted.  The final six lines, composing a sentence, push the interpretation of these enigmatic canvases home:

He feeds upon her face by day and night,

And she with true kind eyes looks back on him,

Fair as the moon and joyful as the light:

Not wan with waiting, not with sorrow dim;

Not as she is, but was when hope shone bright;

Not as she is, but as she fills his dream.

Again the eyes in the painting now address the painter, who “feeds upon her face by day and night.”  It is the eyes which are given as the site of truth.  But these “true” eyes are not the eyes of her “as she is.”  They are true either as she was (in the past) or as he imagines her (an alternative present).  These are the options that the poem provides.  It proposes two possible theories of truth and tests them out.  The first proposal suggests a correspondence theory, according to which the statement may be judged true or false in relation to some state of affairs against which it may be tested.  In this case the canvases and what they are supposed to be representations of do not match up.  On a correspondence theory of truth the artwork fails miserably.  Well we always knew that.  But there are two possible reasons why: either the correspondence theory still applies and the artwork can be attributed with truth in the way a photograph can be: this may not be the situation now but there was a time—the time when they were painted—when they were bang on true.  Or else this is not a representation of the model but a true representation of the artist’s desire, his dreams, his imagination.  In other words it looks like the poem is suggesting that we abandon the correspondence theory of truth for artworks and accept instead an alternative truth for art: the truth of the artist’s imagination.  Except that, if that were the case, then the sense of loss, the hidden-ness and, frankly, the sickness and sorrow of the absent “her” would still need to be accounted for in some way.  This truth of the imaginary (and I would extend this now to the statement) that is attributed to the artist, who stares into the eyes that stare back at him as if he stares only into the eyes of his own truth, oblivious of the lie they represent, is arrived at to the cost of the subject of his paintings (and enunciation).  A potential critique of the male romantic ideal may be emerging.  The loveliness of (masculine) art is won at the cost of the loveliness of (feminine) life.  At this level, there’s nothing to choose between either of the proposed theories of truth.  The conclusion (if we go with it) that the pictures represent the artist’s dream (and not some living thing) could not have been arrived at without the acknowledgement of the disjuncture (canvas-mirror-screen) between the artwork and its model.  And then one extra-textual “source” (the model) is replaced by another (the artist).  At the level of enunciation a third “source” emerges: the interpretation of the “we” in its analytic reading.  So this same one meaning oscillates consistently between at least three alternative and incompatible determinations that nonetheless are parasitically dependent upon each other, like the different paintings yoked together by the same absent meaning.  The correspondence theory fails in two ways: temporally and structurally.  In terms of temporality, the disjuncture would not be unlike the difference between the subject of enunciation and subject of the statement.  Once you represent yourself in language you become an I severed from the moment of uttering, the enunciation, which withdraws immediately from the scene.  And once you represent a subject in language or art your subject looks back from a mask already no longer attached to its origin.  The origin in this sense, the source, was never actually present.  The disjuncture allows a fixing of the idea of the present that is in fact immediately divided in itself—as the disjuncture.  And structurally, the idea or dream of an origin suffers the disjunction according to which the origin is subject to a series of displacements, which we may follow part of the way.  The “her” who looks back also looks back on us as we read a text in which the subject is strictly missing.  We find her only behind screens; we find her hidden; we find that she is hidden.  Like the “we” of the statement we will find that we might not be able to come to a decision as to what the “her” that is only ever represented (and never present) actually is.  We can imagine a Rossetti in her brother’s studio.  But we don’t have to.  And that fact alone, which is the possibility of our reading the text at all, means that no text would ever give itself up to definitive interpretation.  The wider implication is that the correspondence theory of truth and all the paraphernalia of the statement—predicative logic, truth and falsity, the subject of knowledge, objective judgments, etc.—rest on grounds that no one will ever be in a position to comprehend, as a structural condition of possibility for the judgments that we do make.

 

Deconstruction Defined