"Life without Parole": metaphor and discursive commitment - 1

Polonius. My Lord, the Queen would speak with you, and presently.
Hamlet. Do you see yonder cloud that's almost in shape of a camel?
Polonius. By the' mass and 'tis, like a camel indeed.
Hamlet. Methinks it is like a weasel.
Polonius. It is back'd like a weasel.
Hamlet. Or like a whale.
Polonius. Very like a whale.
Hamlet. Then I will come to my mother by and by.
                          --Shakespeare, Hamlet 3. 2. 374-83.

Poor Polonius. The Prince is a mystery to him--and master of the non-sequitur. As Hamlet switches perceptions, dropping one to form the next, Polonius can only follow tamely behind. The dialogue is a whimsical lesson on authority but also a strong lesson on language and thought. It shows directly that figurative thought, the understanding that seeing is seeing-as, requires a capacity for unmaking perceptions as well as making them, and a capacity for irony as well as metaphor, a real-time awareness that the figures of perception are contingent and approximate, both true and not true at the same time. This is a view that differs in significant ways from the one offered by cognitive linguists in their recent work on a theory of metaphor.

A leading figure among linguists working on metaphor, and probably the most recognizable to non-linguists, is George Lakoff. In a series of books, Metaphors We Live By in 1980, co-authored with philosopher Mark Johnson; Women, Fire, and Dangerous Things in 1987; and Philosophy in the Flesh in 1999 (also co-authored with Mark Johnson), Lakoff has worked out the details and implications of a cognitive semantics that he characterizes as an "embodied" or "experiential" realism. At the core of this account is a theory of bodily experience and the image schemas and primary metaphors growing out of that experience to form the basis for conceptual thought. The strong claim throughout Lakoff's work is that thought and language are not disembodied phenomena; nor is metaphor a special case of verbal activity, unique to poetry or art; instead, metaphor is fundamental to thought across all areas of cognition.

Now, I am broadly convinced by the argument that mind, language, and metaphor are in some way embodied. But the way the argument is worked out, and the frequently large claims made for it, especially in Lakoff and Johnson's Philosophy in the Flesh, have made me increasingly skeptical. Their manner of explaining metaphor as the activation of unconscious, basic-level schemas makes it difficult to account for the activity of a cultural subject acting as an ethical, intuitive agent capable of judgment. How this is so is something I can only sketch in for now. To do this, I am going to focus on the theory of metaphor proposed by Mark Turner and Gilles Fauconnier, a theory which is an advance over Lakoff and Johnson's. And I am going to argue that the figure of the ethical, intuitive subject appears only as a ghost in the model of mind informing their theory of metaphor. An "ethical" subject, in this view, is not engaged necessarily in good ethical behavior, but is situated always and inevitably in a scene describa ble in ethical terms--a subject understood to be always at least somewhere, doing something, in relation to someone (this is the basic pragmatic premise). And an "intuitive" subject is a subject intelligently aware of being not only in but involved in a situation and in the actions and meanings in play within it. Though Turner and Fauconnier's theory requires such a figure and depends fundamentally on its capacity for judgment, the theory itself offers no account of it.

But Turner and Fauconnier's theory does important work, and metaphor is only one example of a wide range of cognitive phenomena that they seek to describe according to a general model of conceptual integration or blending. Still, metaphor is a central topic for them, as typified by Turner's account of the sentence, "Vanity is the quicksand of reason," an example of a general XYZ figure, or "X is the Y of Z," as in "money is the root of all evil," or "brevity is the soul of wit" (Turner 53). Turner analyzes the figure using his customary diagram, which looks something like a baseball diamond (65). At first and third bases are circles representing input spaces, one with quicksand, the other with vanity and reason. Between the input spaces are solid lines representing partial cross-space mapping of counterparts (traveler and reason). At home plate is a third space, the blend, again a circle. Dotted lines run from the input spaces into the blend marking paths along which the blend recruits selective features from the inputs (from the quicksand input, knowledge of quicksand as a dangerous trap, but not knowledge of travelers' usual avoidance of deserts). Also, in the blend are points unconnected to any lines, indicating what Turner emphasizes are emergent structures not present in the input spaces or traceable to them: "the traveler can be ignorant of the trap even when he is in it" (65). At second base is another circle, the generic space, which contains abstract structure supporting the cross-space mapping between the input spaces, and there are dotted lines again linking points in the input spaces with corresponding points in the generic space. For the "quicksand of reason" blend, the generic space contains an abstract structure for "action (not specified as physical or mental) intended to achieve something, and a difficulty for that action" (66). Finally, Turner's analysis specifies an organizing frame, represented as a square within the blend circle at home plate. The organizing frame provides conceptual structure specific to the activity of the blend, here a narrative agent encountering difficulties and dangers.

Altogether, this describes what Turner and Fauconnier call a conceptual integration network (Turner 64-67; Fauconnier and Turner, "Conceptual Integration Networks" 142-44).. With its visual scaffolding, it is an analytic tool that in their hands often produces admirably complex, subtle accounts of metaphors and figurative statements. (It also accounts for the baseball/network blend I've just used to explain it.) Most important is the clear, convincing way in which they show that metaphors and blends carry emergent structure that cannot be traced to their input spaces. Turner and Fauconnier rebut the view that metaphor is analyzable as a composition of its inputs alone, and they preserve the sense one often has of metaphor as some sort of mental leap. Turner and Fauconnier do not pursue this point further, relying instead on an emergentist paradigm without considering whether there might be something else at work besides a bottom-up process of inputs, frames, and blends. This constitutes a gap in their analysi s that will be considered presently.

Another important feature of their analysis is its showing that common-sense distinctions between figurative and literal meanings are an illusion. Arguably, in the "quicksand of reason" blend, reason is not "really" a traveler, and vanity is not "really" quicksand. But to hold this view seriously one must believe that there is some standpoint outside figurative language where one can think and say what vanity really is. And this is impossible, for as Turner, and Lakoff and Johnson, show convincingly, language and thought are figurative through and through (Turner 60-64, Lakoff and Johnson, Philosophy 118-29). What feels literal and what feels figurative are no more than variations on the same cognitive process (Turner 68). This too is a deeply significant point, but it loses something important in the thoroughness with which Turner neutralizes literal vs. figurative distinctions, and this too will be considered presently.

Life without Parole

I turn, now, to a recent New Yorker cartoon by Peter Steiner (fig. 1), titled "Life without Parole." Cartoons, like jokes, function very much like metaphors, calling on us to parse their often unlikely juxtapositions and combinations of elements. Here we have a domestic scene, probably an urban apartment. A man sits alone in an armchair in the living room. His plaid jacket worn against the chintz pattern of the chair cover signals a certain tastelessness. He sits upright as he watches, rather glumly, the TV in front of him. Visible through the doorway, with her back turned, a woman in an apron is busy at the kitchen counter--preparing dinner, or maybe cleaning up afterwards. The man and the woman, triangulated by the dark box in the foreground, may or may not be married. It doesn't matter; what joins them is their wordless separation. The caption, "Life Without Parole," interprets the scene as an imprisonment, the two figures sentenced to a lifetime of inert TV gazing and domestic drudgery. Touched by the cap tion, his chair suggests a kind of execution that does not kill but deadens, an execution powered by the circuit between the chair and the wired media set-up in front of it.

It is easy enough to see how this works as metaphor. Using Turner and Fauconnier's method, we can say that there are two mental input spaces involved--urban apartment living and criminal prosecution--from which details are recruited into the blend space of the cartoon metaphor. (For the sake of simplicity, no account of a generic space or frame is offered here.) Each mental space has available to it a large, complex inventory of cultural knowledge, but the blend recruits only a small subset of details from each inventory. In the drawing, the apartment space is just minimally sketched in. All that we see are the TV, the man, the woman, and their immediate domestic circumstances. On the other hand, we know the phrase, "life without parole," as a sentence typically imposed on a convicted murderer. There are no further details recruited from the domain of criminal prosecution, and, by itself, the phrase suggests nothing beyond its conventional punitive meaning. But with caption and image merged, the scene appears to be a comic judgment, as if to say: such is the life of this man and woman, as subjects of contemporary gender roles and media culture, condemned to a lifetime frozen in their positions with no time out.

Still, we might have teased a judgment like this from the image alone if it had been available to us without a caption, for it already suggests the interpretation expressed by the cartoon blend. For instance, there is the oddness of just the single armchair in front of the TV, leaving no way for the man and woman to sit together. Then there is his strangely inert posture and facial expression, the placement of the TV prominently between him and the woman, and so on. So, the image alone--without a caption--clearly suggests many of the same qualities of disconnectedness and boredom apparent in the full cartoon. Where, then, do we locate the separate elements of the cartoon blend? How do we isolate and identify them?

Turner and Fauconnier's style of analysis implies the existence of a conceptual moment somewhere prior to the published cartoon. But we have to be careful here, for there is really no "before" or "after" to the cartoon as an object of interpretation. Assuming we do gain access to the "pre"-cartoon content of the two input spaces--urban apartment living and criminal prosecution--we gain such access only as an artifact of the analytic process. And yet, according to Turner and Fauconnier, "in blending, structure from two or more input mental spaces is projected to a separate 'blended' space, which inherits partial structure from the inputs and has emergent structure of its own" (183). Their metaphors of projection and inheritance imply both generational and ontological priority for input spaces, and what is new in the blend is explained poorly or not at all by invoking an emergentist paradigm. The problem is that, once we analyze a cartoon or metaphor, decomposing it and abstracting what we take to be its primar y elements, then what is to guide us in the reverse process, in recomposing it? The metaphor is not simply greater than the sum of its parts; it is something fundamentally different, the product of a commitment to a sweeping act of judgment. To be sure, Turner and Fauconnier repeatedly emphasize the newness of conceptual structure in the blend, and their analysis is always careful and subtle but, finally, it does not account for the actual metaphor. What is new in the blend is not just new. It is a crucial feature informing the metaphor and its intelligibility, and yet the conceptual-blend theory only notes its "emergence," as if doing so satisfies the demand for interpretation. If a metaphor cannot be predicted from its inputs, then where does the new meaning come from? Do we need to rethink the concept of "input"?

These questions arise, in large part, from what seem to be basic disciplinary commitments entailed by a cognitive linguistic approach to metaphor. The commitment to describe what is regular, invariant, and generalizable across an open-ended sample of instances appears to afford no opportunity to account for particular, situated instances of language use. Such a commitment appears dramatically in the opening chapter of Lakoff and Turner's 1989 book, More than Cool Reason. A Field Guide to Poetic Metaphor. Starting with Emily Dickinson's poem "Because I could not stop for Death--," they analyze its relation to "the range of common, unconscious, automatic basic metaphors" (15) which inform our thinking about life and death:

They are DEATH IS THE END OF LIFE'S JOURNEY, DEATH IS DEPARTURE (an inference from LIFE IS BEING PRESENT HERE), DEATH IS NIGHT (from A LIFETIME IS A DAY), HUMAN DEATH IS THE DEATH OF A PLANT ... .] (from PEOPLE ARE PLANTS), and DEATH IS GOING TO A FINAL DESTINATION (an instance of CHANGE OF STATE IS CHANGE OF LOCATION).

(8, capitals in original)

What exactly is the relation of Dickinson's poem to such metaphors? The book's subtitle, A Field Guide to Poetic Metaphor, seems to promise an explanation of poets' metaphorical practice. But that is not quite what the book does. Its major claim is that such metaphors underwrite the basic intelligibility of common language use and that they are the means by which we actually think about our experience. Thinking, in this view, is fundamentally and unavoidably metaphorical, and poetry is a formal variant of thinking in language. But showing that certain "basic" metaphors underlie Dickinson's poem is not the same as showing what it actually does as a poem. If composing a poem has something to do with rhetoric, form, prosody, and the strategies of discourse, then listing such basic metaphors won't say much at all about its composition. It would be like saying that all we need is "MARRIAGE IS A PRISON" as the basic metaphor to make sense of Steiner's cartoon. In fact, by Lakoff and Turner's own argument, "DEATH IS A JOURNEY" has only a banal connection to Dickinson's poem: for what distinguishes basic metaphors is that they are "commonplace," "general," and "mundane" (1). It would appear that a cognitive approach to metaphor, in order to be true to its disciplinary commitments, must abstract itself from the particular, situated subject of discourse and from its formal and rhetorical motives, even if that leads to a theory that cannot address what is distinctive about a particular instance of metaphoric usage.

However, there is nothing necessary about this conclusion. Cognitive science is remarkable for the variety of contesting viewpoints within it. The contributions of phenomenology and ethnomethodology are strong reminders of the principle that meaning is local to the situation where it is made, and that there is nothing to prevent a cognitive approach to metaphor from joining a description of its systematic structure with accounts of particular, situated acts of meaning. A way to see this perhaps more clearly is to venture a further reading of "Life without Parole."

Language without Parole?

Peter Steiner's cartoon metaphor works in part by way of a pun on the word "life"--life as biographical concept and life as judicial concept. But there is another pun that takes our reading of the cartoon possibly to another level, and that is a cross-linguistic pun, French and English, on the word parole. I can't vouch for Steiner's intent, but parole, in its French meaning and particularly its usage in the linguistic and semiotic tradition of Saussure, virtually begs to be noticed here. Indeed the lack of parole, or speech, is precisely what marks the relationship between the man and woman in the cartoon--they don't talk with one another, and there appears to be no way to imagine them ever doing so. Saussure, in a move crucial to the twentieth-century development of linguistics as a discipline, divided language into two broad categories: langue, the systematic structure of language; and parole, people's actual utterances in everyday situations. Langue, according to Saussure, is a socio-cultural system of di fferences, a system generalized from the differential nature of the signifier in its most basic form as the phoneme. In this view, what makes the humor of Steiner's cartoon possible is the reduction of the man and woman to types frozen as signifiers in the system of differences that structure gender roles, media relationships, and urban identities. Their life is, thus, doubly a life without parole/parole. They are emblems of life as structure, as langue. Or if they do speak, they can only speak the structure, repeat it, and perform its totalizing systematicity.

Parole is the domain of language use and social discourse, and what linguists call pragmatics or, more fully, discourse pragmatics, where meaning is understood to be contingent, negotiable, and revisable in the interaction between subjects who are socially and ethically aware and capable of making judgments on the spot, thus keeping communication focused, open, and inventive. This is a perspective that requires the figure of the ethical, intuitive subject. But it is not a perspective that can be accounted for in Turner and Fauconnier's theory, or in Lakoff and Johnson's theory either.

As one of their examples, Turner and Fauconnier quote a remark that, with a compact metaphor, disparages people who complain that they are trapped in the world of low-level employment: "The Whiners' most common complaint is that they've been relegated to what Mr. Coupland calls 'McJobs"' ("Conceptual Integration and Formal Expression" 193). Turner and Fauconnier analyze "McJobs" at some length, but they never comment on the fact that "Mr. Coupland" is Douglas Coupland, Canadian author of the widely popular 1991 novel Generation X, which introduces the term, "McJob: A low-pay, low-dignity, low-benefit, no-future job in the service sector" (Generation X 5). Subtitled Tales for an Accelerated Culture, the novel is about a trio of dropouts in their twenties, refugees from the world of McJobs living on the margins of the desert retirement community of Palm Springs, California. Their lives are in part a spiritual quest, an echo perhaps of the earlier generation of the Beats, and the novel reflects on whether narrat ive and autobiography are possible in a world that is at once accelerated and futureless. For the three friends the solution, always temporary, is to savor the quirky moments that their lives now and then offer them and to tell each other stories. That is, the novel reflects directly on the possibility of anyone being the narrative subject of her own story, in a way that bears directly on the question of the situated subject of metaphorical speech.

The sentence about "Whiners" that Turner and Fauconnier quote is not from the novel, and it is unclear who utters it, though it is apparently someone who disapproves of the novel and its slacker heroes, and who uses the term "McJobs" sarcastically. In any case, Turner and Fauconnier submit the metaphor to their standard analysis: it is formed by the complex blending of two mental spaces that recruit details about the world of employment and work on the one hand, and about the world of McDonald's on the other. Straightforward mapping of the two mental spaces links the common roles:

[...] but it does not in itself provide the central inference of "McJobs." It is in the blend that specific aspects of the McDonald jobs--no prestige, no chance of advancement, no challenge, no future, boredom, a certain kind of social stigma--are blended with the more general notion of low-level service jobs. Absent this blend, we would be free to associate low-level service jobs with other prototypes: altruistic and even saintly devotion to others, Horatio Algeresque climbing the social ladder from the bottom up, small-town serenity and routine, freedom from avarice and grueling ambition, and so on.

("Conceptual Integration" 193)

Granted, "absent this blend," we are free to speculate. But the blend is not absent. Nor is the situated judgment absent that gives rise to the blend. Speculations about Horatio Alger, altruism, and small-town serenity make sense only if we deliberately step outside the discourse context of the blend, or adopt a posture in opposition to its underlying judgment--as indeed the speaker does whom Turner and Fauconnier quote. Otherwise, such speculations are irrelevant.

In their defense, it should be said that Turner and Fauconnier are concentrating here on the word "McJobs" as an example of a compact, minimal set of formal cues for a reader or listener to use in unpacking the structure of the metaphor. But, even then, there is a problem in the way they fail to acknowledge that the metaphor is not just used but quoted by the speaker, with an apparent intention not mentioned by their analysis. There is the further problem that it is not easy to see what makes the difference between straightforward mapping, which Turner and Fauconnier reject, and the action of blending, which they favor. What the blend seems to do is exercise selectivity. It is not just any features of MeWorld that are blended to form the metaphor. But that begs the question of what determines the selection, and invoking "emergence" does not advance our understanding. For it is not just a matter of picking Out the relevant features of McWorld but of having committed oneself to a judgment that colors all the fe atures within it.

The choice of negative features is the product of an ethical judgment in response to being frustrated, bored, hopeless, and stigmatized--a judgment that quickly became a cliche in the wake of Coupland's novel. And it is as a cliche that the word "McJobs" is quoted and parodied in the example. At the level of the metaphor as blend, some other speaker--not the one quoted by Turner and Fauconnier--makes a choice on the spot that Coupland's metaphor is the thing to use as a way to express an attitude about her situation. And, then, at the level of the quotation, someone else--the person directly quoted in Turner and Fauconnier' s example--chooses to repeat the term "McJobs" as a way to parody people whom he disparages as Whiners. So, there are two acts of judgment involved, and neither one finds its way into the analysis, even though they are fundamental to the interactive, situated meaning of the metaphor each time it is used.

The style with which Turner and Fauconnier write about metaphor says a lot about their aims, their disciplinary commitments, and, consequently, what counts for them as valid explanation. These are important matters, because inevitably one works within a single disciplinary discourse--or, with luck and perseverance, one can hope to work across two or maybe three disciplines. But it is the discipline that determines the kinds of questions worth asking, the methods for asking them, and even the very nature of the object to which questions are put. Turner and Fauconnier's disciplinary commitment is to understanding and describing the systematic nature of metaphoric thought, the ways in which we think metaphorically without thinking about it. Such a commitment and the general research program supporting it have defined the core discipline of linguistics, at least from the moment when Saussure established langue, not parole--the linguistic system, not language use--as the object of linguistic study.

The Blood of Children on the Street

For Saussure, the linguistic system is fundamentally a sociological system, a view that led logically to the twentieth-century recruitment of Saussure's work as the foundation of structuralist anthropology and, subsequently, to the elaboration of the wide range of post-structuralist theory and practice. Language, in this view, is understood to function as a sociological unconscious, a system of linguistic practices defining how members of a culture speak and act without thinking about it. For Lakoff, Johnson, and Turner language is also an unconscious system, but "unconscious" in a very problematic sense. Lakoff and Turner, in More than Cool Reason, speak of common, unconscious, basic metaphors that underlie thought. Lakoff and Johnson, in Philosophy in the Flesh, routinely invoke a "cognitive unconscious" as the domain of primary experiential schemas and basic metaphors that we acquire automatically, simply as a consequence of growing and living as sentient beings in the natural environment. But the concept of a cognitive unconscious, as deployed in Lakoff and Johnson in Philosophy in the Flesh, is deeply problematic on several counts.

Although Lakoff and Johnson, once or twice, use the more appropriate phrase "cognitive nonconscious," they seem to prefer the wording "cognitive unconscious," which they distinguish right away from Freud's concept (Philosophy 10). In doing so, they invite comparison and also suggest a motive not just to be different from Freud, but to displace Freud, to appropriate the concept of an unconscious, with its long twentieth-century tradition in psychoanalytic and philosophical thought, and to remake it according to their model of embodied cognition. And yet, whatever one may think of Freud and psychoanalysis, and whether or not one rejects Freud's idea of an unconscious, it is uncontroversial that Freud understood the symptom to be an interpretable sign of unconscious thought, that unconscious thought is capable of being brought to awareness by a process of interpretation. Nothing of the sort is claimed for the cognitive unconscious, a vaguely defined domain of neuro-cognitive functions and processes, and in fact nothing of the sort can be claimed for it. And yet Lakoff and Johnson are not always clear about the boundary between conscious (or unconscious) thought and neuropsychological process. Not only do they repeatedly confuse phenomenal embodiment and neural embodiment, they even re-introduce Descartes (with a little help from Adam Smith) in describing the cognitive unconscious as "the hidden hand that shapes conscious thought" (Philosophy 12)--a picture of the mind not as the familiar Cartesian theater but as a somewhat more up-to-date Cartesian free market. In spite of their best claims for embodied cognition, it is clear that the mind/body problem is still very much alive.

A basic problem, then, with the theory of embodied cognition is how one mediates between descriptions of nonconscious mental processes and descriptions of conscious, situated linguistic activity, when the possibility of interpretation between them cannot be entertained even in principle. This problem was clearly evident during a recent graduate seminar in Poetics at SUNY/Buffalo, (2) in which Lakoff lectured on metaphor. Responding to student questions, Lakoff consistently explained examples of poetic metaphor by invoking primary body schemas: life is a journey, happy is up, sad is down, problem solving is movement toward a goal, and so forth. He also used Turner-style readings of conceptual blending. As might be expected, these explanations produced a good deal of frustration. Students wanted to talk about poetic activity, the craft of writing, the making of lines, phrases, and tropes. Instead, they were told that poetry is the result of processes in the cognitive unconscious, that tropes arise from activati ons of basic schemas of embodied cognition and conceptual blends across mental spaces.

Then one student asked a question that challenged Lakoff in a very telling way. He asked Lakoff to comment on lines he remembered from a poem by Pablo Neruda: "the blood of children on the street is like / the blood of children on the Street." Neruda's lines in translation--from the poem, "I Explain a Few Things (Explico Algunas Cosas)" (Spain In the Heart 35)--are actually somewhat different from how the student remembered them: "and through the streets the blood of the children / simply flowed, like children's blood (y por las calles la sangre de los ninos / corria simplemente, como sangre de ninos)." (3) Clearly the student wanted to quote lines that deliberately prompt an expectation for metaphor while refusing to satisfy it. He closely echoed Neruda's austere phrasing and his refusal--ethical in its passion--to say that the blood of Spanish children killed and wounded by fascist bombs was like anything else in the world. Both versions, both the student's phrasing with its exactly repeated phrases, and Ne ruda's looser simile, pose a challenge: explain this metaphor if you can. Lakoff, to his credit, did not take up the challenge: he did not try to explain the metaphor in his usual manner. Conceivably, he might have addressed issues of basic-level cognition and analyzed the container/contained schema of a liquid flowing in a gutter. But he did not. Or he might have reflected on the rhetoric, the hyperbole of Neruda's picture of children bleeding such volumes of blood that it actually flowed through the streets. But he did not do that either. And the obvious reason--though it never got voiced in the seminar--was that this is not where the interest is. It is not what Neruda's deliberately crafted lines are doing. That is, the interest is not so much in what the lines are saying semantically, in what they "mean," as in what they are doing. A poem in this view is not the report of an experience. It does not "have" a "content" that it "communicates" to the listener, as if the speaker were located somewhere outside the space and time of what the poem is "about." Its language, too, its rhetoric and prosody, its figurative practice, do not exist apart from the particular here and now of its utterance, as something spoken and something heard.

Edges of Being

The example of Neruda's anti-metaphorical gesture echoes a familiar understanding of American poetry in the tradition of William Carlos Williams, a tradition whose most celebrated contemporary poet is surely Robert Creeley. Consider, for instance, the first several lines from his recent poem "Edges," composed for his wife Pen's birthday:

Edges of the field, the blue flowers, the reddish wash of
the grasses, the cut green path up to the garden
plot overgrown with seedlings and weeds --

green first of all, but light, the cut of the sunlight
edges each shift of the vivid particulars, grown large
--even the stones large in their givens, the shadows massing

their bulk, and so seeing I could follow out to another
edge of the farther field, where trees are thick on the sky's
edge, thinking I am not simply a response to this, this light,

not just an agency sees and vaguely adumbrates, adds an opinion.
There is no opinion for life, no word more or less general.
I had begun and returned, again and again, to find you finally,

felt it all gather, as here, to be a place again [...]

Edges and cuts are a frequent theme for Creeley: edges of forms, edges of light, edges of phrases cut at the line-breaks, projecting the broken rhythms and freshness of a voice finding its words and its place in the here and now of the moment of speech. The effect is both visual and verbal, in keeping with Creeley's long-time collaboration with visual artists (Jim Dine, Cletus Johnson, Robert Indiana, among others), and we have the immediate sense of a place, tentatively identifiable as a loose array of edges all gathered, at last, about the figure of a woman, "as here, to be a place again," a figure not described but addressed. If there are metaphors at work in these plain-spoken lines, they would be the cuts and edges themselves, edges of field and sky, cuts of path and sunlight. These are very much like the ordinary metaphors that Lakoff and Turner, or Turner and Fauconnier, treat routinely in their research. But they are not figures to be analyzed just as blends of inputs from separate mental spaces, or a s expressions of basic mental schemas. For their reach is ontological: speaker, words, and place are thoroughly entangled together in their being, all gathered, here, "to be a place again," shaped by the cuts and edges of voice, eye, and landscape. Creeley's explicit rejection of the role of mere observer, "an agency [that] sees and vaguely adumbrates, adds an opinion," underscores this pragmatic understanding of poetry and language.

This is not just a matter of poetic practice; it is also the nature of ordinary, everyday language use, which is always situated, always occurring in the awareness of something at stake and calling for an unavoidable commitment of judgment and action. This is a view recognizable as the legacy of Kant, Wittgenstein, and Austin. In a very recent iteration, it appears in the work of Robert Brandom. The last phrase in the title of this essay, "discursive commitment," echoes the title of Brandom's 1994 book, Making It Explicit: Reasoning. Representing, and Discursive Commitment. Brandom is concerned with understanding language more as discourse than as representation. Rather than a Cartesian concern for certainty, Brandom takes as his topic a Kantian concern for what speech or writing responds (and is responsible) to. According to Brandom, Kant's "fundamental insight is that judgments and actions are to be understood to begin with in terms of the special way in which we are responsible for them" (Making it Explici t 8, italics in original). That is, language understood not as system or structure, but understood as use, as utterance, inevitably entails commitment to judgment of one kind or another about the situation in which one acts by saying what one says. There is no neutral space, no time out, where we do not make such a commitment when we speak--even at the moment when we forcefully assert our freedom from making commitments. This is what it means to say that Neruda's deliberate--though seemingly casual--refusal of metaphor about the blood of children is an action, no more nor less than the student's question citing Neruda's lines is an action, and not just a request for information or for an explanation (though of course requests too are actions).

It will be objected, and with reason, that this is an inappropriate standard to use in judging the cognitive linguistic study of metaphor, that it misses the point of the research project that Lakoff and Johnson, and Turner and Fauconnier are engaged in. Understandably, the effort to map the systematic nature of metaphorical activity will tend to look past the particular, situated actions of speakers and writers. As we have seen, the usual practice is to test cognitive theory across a wide range of metaphor samples, usually brief and only minimally related to their original contexts, as in Turner and Fauconnier's use of the quotation about Whiners and McJobs. This allows for a strong focus on the project of developing models of invariant forms and functions at a very basic level. It is a process of picking out salient features from the linguistic data and analyzing them to the point of specifying their elemental constituents. Clearly this is what Lakoff and Johnson are attempting to do, as they seek to establ ish a theory of thought that is metaphorical througfr and through and that is grounded in the body and the neurolinguistic functions that mediate mind, brain, and body. But the deep problem with this project is that analysis is no guide for synthesis. Having broken metaphoric speech down into its elements, one loses in the process a grasp of the occasion in which such speech was formed in the first place, and the reason is that any moment of metaphorical speech--in poetic language or ordinary language--is situated in a specific here and now that is discarded in the process of cognitive analysis. It is in fact this discarding of context that makes possible the kind of analysis done by Lakoff and Johnson and by Turner and Fauconnier. Of course, this is not a new difficulty, and Mark Turner has noted its importance in his recent essay on "Figure," which he takes as the overall category for metaphor and tropes generally:

Historically, the study of figure has taken on the job of proposing abstract elements--"figures"--and giving examples but has not taken on the job of explaining the dynamism and completeness of individual examples. Typically, the study of figure attempts to isolate and exemplify partial structures that get used in the construction of meaning, but not to give a theory of that actual use. [...] The theory of basic metaphor (with which I have been associated) attempts to isolate a quite small number of elemental basic metaphors (maybe 600) that we all know, and to provide examples of each, where the examples are meant as evidence for the existence of the abstract elements of the theory.

The central danger for such partial models of conceptual construction is that they might not "scale up" appropriately. The well-known failure of attempts to scale up from partial artificial intelligence models to full models is worth remembering in this respect.

(Turner, "Figure" 80-81)

Turner states the difficulty clearly--knowing the way down gives no assurance that you will know the way up--it is not just the reverse of the way down. And yet, Turner's understanding of this problem seems to have no further effect on his analysis of metaphors and blends. Occasionally, Fauconnier and Turner speak of projection from the blend back to the input spaces (e.g., "Conceptual Projection and Middle Spaces" 6), but this still assumes an initial projection from input space to blend and does not address the effect of a prior judgment that has already selected and globally organized the content of input domains.

Peeled Apples

In Turner's essay on "Figure," he cites Gagne and Murphy's paper on discourse context and its effect on perception of emergent features in a conceptual blend. Working as cognitive psychologists with a series of controlled experiments, Gagne and Murphy found that, for a blended phrase like peeled apple, test subjects identify phrase features (e.g., the whiteness of peeled apples) more accurately than noun features (e.g., the roundness typical of all apples). That is, in Turner and Fauconnier's terms, features of the conceptual blend are more readily available to cognition than features of elements from input spaces. Turner sees this outcome as confirmation that blends display emergent properties not present in input spaces or predicted by them. But Gagne and Murphy's work challenges Turner's argument in a way that he may not have anticipated.

In each experiment, Gagne and Murphy's subjects are given one of two versions of a brief narrative before being presented with a target sentence. In one experiment, both versions of the narrative begin with the same two sentences: "Alan was a well-known chef who used fruit and vegetables as garnish. Every night he spends half an hour selecting and arranging the perfect items." Each version then has a different last sentence, one with phrases that repeat the modifier ("Tonight, Alan decided to use peeled apples and peeled carrots"), the other with phrases that repeat the noun ("Tonight, Alan decided to use peeled apples and chopped apples") (Gagne and Murphy 87, italics in original). After reading the narrative in one of its versions, subjects are asked to verify one of two sentences: "Peeled apples are round," or, "Peeled apples are white." On average, subjects verify the sentence about whiteness significantly better than the sentence about roundness, and they do so whether their version of the narrative repe ats the noun or repeats the modifier. This outcome suggests that blends are primary, that in discourse they exist prior to their input elements, and that a strictly bottom-up analysis of blends must encounter serious problems.

Given the work of Gagne and Murphy and the effect of Peter Steiner's cartoon, we can see that there are three related issues that are not adequately treated by Turner and Fauconnier's conceptual-blend model for metaphors, and for phrase and word blends generally: (1) discourse contexts and their effect on the production and comprehension of blends and metaphors; (2) the commitments of language users in specific situations; and (3) the contingent, revisable nature of a user's commitment to a blend.

(1) Both production and comprehension occur in a discourse context which already motivates the blend or metaphor and poses constraints for relevance and purpose.

The new, emergent properties of a blend are not really so new, although they must seem new if one holds strictly to a bottom-up model of interpretation, as I believe Turner and Fauconnier do. A discourse context provides the rules, top down, for the language game in play. It does not determine moves but only limits which moves make sense. By a mathematical analogy, just as there is an infinity of real numbers between 0 and 1 (or between 0 and 0.1, or 0 and 0.01 ...), so in any game there is in principle an infinite number of possible moves within the rules. But is this true, say, in chess where the number of possible moves seems clearly finite? After all, each chess piece is defined according to the limited set of moves assigned to it. And yet, once we recognize that a move is a move in a particular situation or game state--with a particular number and combination of pieces arranged in a specific pattern on the board--then the number of possible chess moves expands toward infinity. This is why it takes immens e processing power for a computer, scanning all possible moves several game states forward, to match the intuitive play of a chess master who can see in a moment the two or three really promising moves available.

Discourse is not chess, but the analogy helps to explain how new, emergent properties in a blend or metaphor are actually new, while at the same time responding to the motivation and constraint of the discourse context where they occur. But newness in this sense is not the same as the surprising clarity or humor or depth of a quick-witted phrase or rhetorical figure. That's the newness of a "McJobs," a "Life without Parole," or a "blood of children in the street." In contrast, the phrase "peeled apple" is unremarkable because it fits easily with our cultural knowledge of chefs and food preparation. Also, apples are a familiar garnish and are often peeled--unlike blueberries or cherries or grapes--which is why Mae West's sultry demand, "Peel me a grape," stands out as funny, while peeled apple" is really a stealth blend. It asks for no cognitive stretch, or, in Turner's terms, it is well "entrenched" ("Figure" 62-63)--which is to say that even the most ordinary and unremarkable detail of language use is analyz able as a blend or metaphor, a point that Turner and Fauconnier make especially well, even though they do not give enough attention to discourse context or to the commitments of language users who make blends and interpret them.

(2) A blend or metaphor is motivated by the situated commitments of a language user before it is referable to a compositional or integrative process: the whole precedes its parts.

We have already seen Robert Brandom's contention, following Kant, that semantics is about discursive commitment, about what speech or writing responds (and is responsible) to. In a more recent book, he again follows Kant in insisting that "the fundamental unit of awareness or cognition, the minimum graspable, is the judgment," and he follows Frege in claiming that "it is only to the utterance of sentences that pragmatic force attaches" (Articulating Reasons 125), which is another way of underscoring the primacy of discursive context, with the sentence considered as the minimum grammatical unit of awareness. This is not an uncontested position, and the heat of the argument is evident in Brandom's own occasional metaphor. For instance, he argues, "What makes something a specifically linguistic (and therefore [...] discursive) practice is that it accords some performances the force or significance of claimings, of propositionally contentful commitments, which can both serve as and stand in need of reasons." And then he elaborates: "[L]anguage (discursive practice) has a center, it is not a motley. Inferential practices of producing and consuming reasons are downtown in the region of linguistic practice. Suburban linguistic practices utilize and depend on the conceptual contents forged in the game of giving and asking for reasons, [and] are parasitic upon it" (Articulating Reasons 14, italics in original).

Brandom's metaphorical opposition between a downtown core of cognitive and linguistic practice and a parasitic suburban dependence on that core, articulates a strong judgment that is both philosophical and sociological in scope. Blended together are the judgment of a committed urban dweller against the uncontrolled growth of suburbs that suck the life out of inner cities (a pervasive, present problem generally across the United States, but specifically in Pittsburgh where Brandom teaches) and the judgment of a committed pragmatist against linguistic and philosophical practices that withdraw semantics into a suburb independent of discourse even while tacitly using discourse and depending on it for their existence. Just as in Steiner's cartoon, "Life without Parole," there is a judgment already in play, which globally organizes the various details of the cartoon into its comic figure, so here Brandom's commitment as a philosophical pragmatist and (probably) also as an urban citizen, provides a context of discur sive commitment that his metaphor responds (and is responsible) to. Implicit in all this is a grounding judgment that things could (and should) be otherwise, that arrangements--in both words and thoughts--are contingent, provisional, subject to rewording, rethinking, rearrangement.

(3) A theory of blending requires a theory of un-blending; the use of metaphor requires an awareness, a judgment that a particular metaphor both works and does not work in the context where it used.

As we have seen in Gagne and Murphy's research, the emergent properties of a blend (e.g., peeled apples) are more salient than the properties of their constituent elements alone (peeled and apple). This is not just a discourse context effect. It is also the organizing effect of a blend on all its constituent details.

Consider how you would respond to a request to pronounce the word whose first five letters are: "f-l-a-n-g." Solving this can be frustrating if you fixate on "a-n-g" as an apparent blend rhyming with familiar words like bang or gang or sang. It is a trap, and once you adopt this blend, you are likely to have a hard time un-adopting it. Focusing on the blend means that its separate elements are not immediately available, or they are available only through a deliberate analytic effort. Consequently, a cognitive capacity for blending requires a matching capacity for un-blending. Without it, there is no accounting for the ordinary freedom necessary in order to interpret or to speak at all. Moreover, we are never outside of a linguistic context or language game, even (or especially) when theorizing about language, and language is never a purely raw material: we know it always as language in use--what we hear or overhear, what we read, what we find ourselves saying at any given moment. So, writing and speaking are really rewriting and re-speaking, and a full account of blending has to include a concept of un-blending and re-blending. Or, more exactly, in Brandom's terms, a theory of discursive commitment implies (and requires) a theory of discursive dis-commitment and re-commitment. For we do not actually undo blends--Gagne and Murphy show they are too tough for that. Instead, like Hamlet seeing a cloud first as camel, then as a weasel, then as a whale, we just abandon one blend for another in the process of refraining our judgment of a situation. (The word in the puzzle is flange.)

Thinking Metaphor Discrepantly

Leonard Talmy, in his seminal paper on "Fictive Motion," argues that any maker or user of metaphor is, in principle, aware that a particular metaphor both fits and does not fit what it describes--as in Hamlet's invitation to Polonius to see the cloud as, not actually, but "almost" shaped like a camel. This double, contradictory awareness is evidence, again, that some form of judgment plays a leading role in the use of metaphor and the processing of blends. It is the kind of judgment playfully foregrounded in people's responses to "bad" puns (only really good puns are "bad"). What is distinctive about Talmy's approach is the posture of methodological indifference that he adopts toward questions of fiction and reality, metaphorical truth and literal fact. That is, he brackets out the usual preoccupation of analytic philosophy with the truth conditions of statements, while retaining a neutral perspective on questions of veridical difference. To this end, he uses the words fictive and factive as purely descriptiv e terms in analyzing sentences like, "The mountain range goes from Canada to Mexico." Of course mountains don't "go" anywhere. But the sentence is a model for the way cognition implements "veridically unequal discrepant representations of the same object" (100), without committing to one representation being objectively real and the other not. Mountains are massively immobile, and at the same time they also go from one place to another. That is how we think and how we see--discrepantly--and with a cognitive bias toward dynamism" that arises from the embodied nature of cognition (171-72).

Talmy comments directly on other cognitive accounts of metaphor, specifically on Lakoff and Johnson's typical schema--love is a journey, argument is war, seeing is touching--in which features from a source domain (journey, war, touching) are projected onto a target domain (love, argument, seeing). Talmy's comments apply equally well to Turner and Fauconnier:

[F]actively, love is not a journey, while in some fictive expressions, love is a journey. The very characteristic that renders an expression metaphoric--what metaphoricity depends on--is the fact that the speaker or hearer has somewhere within his cognition a belief about the target domain contrary to his cognitive representation of what is being stated about it, and has somewhere in his cognition an understanding of the discrepancy between these two representations.

("Fictive Motion" 168, italics in original)

I want to suggest that such an understanding of metaphoric discrepancy, wherever it may be located in cognition, is a deep form of what we commonly call irony, a concept of critical importance to the understanding of metaphor. By a "deep form" of irony I mean a function more general than speaker's irony, as when a speaker intends the opposite of what he literally says ("that pun is really bad"). The deep form of irony and its relation to metaphor is comparable to Kenneth Burke's account of the two tropes. Burke treats metaphor as perspective (thinking of the Rockies as "going" from Canada to Mexico is thinking about mountains from the perspective of embodied motion), and Burke treats irony as dialectic, or the perspective of perspectives, the position from which any metaphor is understood to be approximate, contingent, and revisable ("Four Master Tropes" 503-05, 512).

Irony is a powerfully important trope, and the model of conceptual integration requires it, because integration as a function is unthinkable without a prior, or at least attendant, function of disintegration. We cannot put things together unless we already know how they come apart. A model of figural thought based upon such a close, dialectical relation between metaphor and its ironic negation is, I believe, a model that might begin to describe the practice of a poet like Robert Creeley. For Creeley, metaphors arc figures of a contingently place-bound and time-bound ontology, a matter of being 'just in time," according to the title of his recent collection of poems. But to say this of Creeley's practice is to say something that is generally true of ordinary language use as well. Not only does this assume that all language use--in poetry, in science, in ordinary conversation--is metaphorical from the ground up, as has been forcefully demonstrated by Turner and Fauconnier, and by Lakoff and Johnson. But it also assumes that language is language-as-use before it is language-as-system, that parole is both conceptually and ontologically prior to langue.

The relation between a system and actions performed in accord with it, between langue and parole, has been a relation preoccupying linguistic and philosophical thought from at least the start of the twentieth century. A system is a set of rules inferred from actions, and while rules provide a ground for interpretation, they cannot predict actions or explain their causes. For rules do not apply themselves, and any attempt to devise a second rule for applying the first one opens up an infinite regress. Rather than a problem, this is a profoundly liberating circumstance; for it clears a space where it is possible to imagine a rhetoric, a politics, a poetics, an ethics. Above all, metaphor is an act, a commitment (and revisable re-commitment) of intuitive judgment. For without judgment, there can be no metaphor that we could recognize as something that someone might actually say in some situation or other.

Notes

(1.) A preliminary version of this essay was presented at the 1999 conference of the Society for Literature and Science (SLS) in Norman, Oklahoma. Thanks to James J. Bono, Vicky Kirby, Kenneth Knoespel, Jim Paxson, Martin Rosenberg, and Susan Squier for their helpful comments.

(2.) This was Charles Bernstein's Poetics seminar, November 4, 1999.

(3.) Thanks to Rosa Alcala, Peter Ramos, Kenneth Sherwood, and Jessica Smith for identifying these lines.

Works Cited

Brandom, Robert B. Articulating Reasons: An Introduction to Inferentialism. Cambridge, MA: Harvard UP, 2000.

_____. Making It Explicit: Reasoning, Representing, and Discursive Commitment. Cambridge, MA: Harvard UP, 1994.

Burke, Kenneth. "Four Master Tropes." Kenyon Review 3(1941). Rpt. A Grammar of Motives. New York: George Brazilier, 1945. 503-17.

Coupland, Douglas. Generation X. Tales for an Accelerated Culture. New York: St. Martin's, 1991.

Creeley, Robert. "Edges." Echoes. New York: New Directions, 1993. 87. Rpt. Just in Time Poems 1984-1994. Used by permission of New Directions Publishing Corporation.

Fauconnier, Gilles, and Mark Turner. "Conceptual Integration Networks." Cognitive Science 22 (1998): 134-87.

_____. "Conceptual Projection and Middle Spaces." UCSD Cognitive Science Technical Report 9401. University of California, San Diego (April 1994). 18 April 2002 <http://cogsci.ucsd.edu/cogsci/publications/9401.pdf>.

Gagne, Christina L., and Gregory L. Murphy. "Influence of Discourse Context on Feature Availability in Conceptual Combination." Discourse Processes 22 (1996): 79-101.

Lakoff, George. Women, Fire, and Dangerous Things: What Categories Tell Us about the Mind. Chicago: U of Chicago P, 1987.

Lakoff, George, and Mark Johnson. Metaphors We Live By. Chicago: U of Chicago P, 1980.

_____. Philosophy in the Flesh: The Embodied Mind and its Challenge to Western Thought. New York: Basic Books, 1999.

Lakoff, George, and Mark Turner. More than Cool Reason: A Field Guide to Poetic Metaphor. Chicago: U of Chicago P, 1989.

Neruda, Pablo, Spain In the Heart: Hymn to the Glories of the People at War, (1936-1937). Trans. Richard Schaaf. Washington, DC: Azul Edition, 1993.

Steiner, Peter. "Life without Parole" [cartoon]. The New Yorker. August 17, 1998. 59. Reprinted by permission of The Cartoon Bank.

Talmy, Leonard. "Fictive Motion in Language and 'Ception." Toward a Cognitive Semantics. Vol. 1. Cambridge, MA: MIT P, 2000. 99-174.

Turner, Mark. "Figure." In Katz, Albert N., Cristina Cacciari, Raymond W. Gibbs Jr., and Mark Turner. Figurative Language and Thought. New York: Oxford UP, 1999. 44-87.

Turner, Mark, and Giles Fauconnier. "Conceptual Integration and Formal Expression." Metaphor and Symbolic Activity 10 (1995): 183-204.

Margaret H. Freeman (freemamh@lavc.edu) is Emeritus Professor of English, Los Angeles Valley College. She defines her research within the emerging field of cognitive poetics, and has published several articles on the integration of cognitive linguistics and literature. She is currently working on a book-length cognitive poetics study of Emily Dickinson's writings.

COPYRIGHT 2002 Northern Illinois University
COPYRIGHT 2002 Gale Group