Mapping the mind and the body: on W.H. Auden's personifications - Critical Essay

Craig A. Hamilton

"The will of one by being two At every moment is denied."

(W. H. Auden, "The Sea and The Mirror," Collected 413)

1. Introduction: Mind, Body, and Personification

Despite our traditional view of the body and mind as divided, one figurative way for representing the two in our thought and in our language is the unified method of metaphor. For example, because we have easier access to bodies than to minds, our everyday notion of body language means we "map" from the body to the mind to interpret the behavior of others. As Simon Baron-Cohen has argued, this normal psychological mapping provides us with a "theory of mind" that some autistic children appear to lack for successful social cognition. We may not be fully aware of such mappings, but they conceptually link the body to the mind, making the body indexical of the mind in a way that closely integrates them and nearly negates dualism. Apart from these non-verbal mappings, mappings from the body to the mind in language, according to Eve Sweetser, motivate how physical verbs like grasp take on mental meanings like "know" during a language's evolution. As we would imagine, these connections between mind and body, between the mental and physical, also appear in literature, particularly in bodily descriptions, the language of emotions, and the use of mind or body metaphors. W. H. Auden is one case in point since he was forever writing about the mind and the body in his poems. Usually, Auden depicts body and mind in general via metaphor and personification in particular. What makes Auden's mind and body personifications strange, however, is that they are unlike the imaginary abstractions we often associate with personification.

Personification was vital to Auden. According to Bernard Bergonzi, the four main features of Auden's poetic style were the "copious use of the definite article; unusual adjectives and adjectival phrases; surprising similes, which have a reductive or trivializing effect; and personified abstractions" (70). Indeed, Auden's personified abstractions bothered readers like Karl Shapiro, who wrote in his 1947 Essay on Rime, "An all-purpose abstraction is a form / Dear to the tired mind that must malinger / And precious to the talentless [...] History is but one / Of Auden's ill-starred words. Luck is another" (qtd. in Bergonzi 68). Irvin Ehrenpreis also found that "Auden at his best did not stop at personification; he embodied the abstractions in curious or supreme examples" such as southern Italy's limestone landscapes in the poem "In Praise of Limestone" (498-99). Shapiro, Bergonzi, and Ehrenpreis have all recognized Auden's tendency to personify and embody abstractions. However, Auden also tended to personify the mind and the body, entities which are perhaps less abstract than we recognize.

We personify when we metaphorically ascribe agency to normally inanimate objects, turning non-existent or imaginary entities into realistic actors or agents. As the cognitive linguist would describe it, to personify is to "map" information from a "source domain" onto a "target domain" (what I. A. Richards once called vehicles and tenors). Mapping occurs simultaneously at conceptual and linguistic levels. Novel metaphors in language often reflect conventional metaphors in thought. In eroding classical boundaries between figures of thought and figures of speech, personification is apt for study from the cognitive viewpoint because a metaphor in language normally reveals a related conceptual metaphor in thought (Gibbs 311). Simply put, one metaphor can hide another. Therefore, it is fruitful to consider personification as both a product of thought and a product of speech.

Personification is one of our most basic and frequently utilized metaphors. Its high frequency in children's literature suggests that we can understand it very early in life and that it is our "prototypical metaphor" built from "nonhuman topic--human vehicle" mappings (MacKay 87). Not surprisingly, the same holds for Auden, who presumably knew that the personification of abstractions in literature reached its apex in eighteenth-century English poetry. The figure seemed to fall into disuse after Wordsworth's radical break with the trend in "Yew Trees" (Knapp 128). Personified abstractions that allegedly functioned allegorically, popular in the seventeenth and eighteenth centuries, were therefore more or less abandoned by the Romantics. As Steven Knapp has suggested, personifications such as Sin or Death in Milton's Paradise Lost, for example, became problematic in the eyes of eighteenth century writers. The view that personification as a poetic device could no longer be easily defended followed from the divisi on between "arbitrary magic" and "rational Aristotelian mimesis" for critics like Addison and from Lord Kames's separation of the "natural" personification of "insensible objects" from the "unnatural" personification of "deities, angels, devils, or other supernatural powers" (Knapp 60-61). Since no Romantic would have wanted willingly to write seemingly "unnatural" poems, "natural" "mimesis" was apparently preferred. The issue, as Knapp puts it, was no minor one:

Some of the danger implicit in the transformation of rhetorical personifications into agents [...] lies in its apparent reversibility. "Imaginary" agents disrupt the realistic texture of epic partly because they represent an alien mode, but also because they call into question the status of ostensibly "real" or "historical" agents. If personifications are animated through the intensification of metaphor (or more precisely, through the intensification of a metaphoric vehicle at the expense of its supposed "tenor"), then mimetic agents may have a converse tendency to slide "back" into metaphor (that is, the agent may turn out to be the vehicle of a previously unsuspected or forgotten tenor). The reversibility of personifications thus makes the boundary between rhetoric and agency less secure than it might have seemed. As figurative language seems more violent and opaque, agents seem more transparent and abstract. (60)

Those familiar with research into metaphor by cognitive linguists will find the "reversibility of personification," its "alien mode," and differentiation between agents and personifications somewhat odd. George Lakoff, Mark Johnson, and Mark Turner have argued for two decades that personification cannot be described in such ways. These researchers would argue that Knapp's passing mention of tenors and vehicles seems to take for granted that which he could be accounting for, in the same way that they would find Katherine Hayles's recent remark about "simple anthropomorphic projection" (2) onto objects in art to be overlooking the fundamental cognitive mechanisms behind personification. Of course, Knapp and Hayles are not cognitive linguists, so taking them to task for ignoring the cognitive dimensions of personification may be unfair. However, because accounting for our mappings of the human onto the nonhuman inherently involves metaphor in a way that is far from simple, the mappings merit explaining. For exam ple, Knapp's remark about "personifications on vacation from allegory" (128) in Wordsworth's poetry is witty, but it only adds confusion where clarity is needed because of long misunderstandings in literary history over allegory (Crisp 6-8). Love, History, Luck, and so on, are probably better understood by personification than by any other source domain. Because we are hard pressed to imagine these target domains literally, it seems we cannot avoid metaphor to imagine them. If so, Auden's personifications are justifiable. That said, literary critics still need a story about personification that makes sense and accounts for examples that include the talking bodies and talking minds found in some of Auden's greatest poems. A story like this no doubt begins with the rhetorical tradition before cognitive linguistics arrives on the scene.

Perhaps the oldest view of personification that literary scholars are familiar with comes from the rhetorical tradition. For instance, in the sixteenth century, Erasmus (d. 1536) referred to two types of personification: "prosopographia" and "prosopopoeia" (51). He classified prosopographia with the personification and description of abstractions such as Justice or Calumny (52). By turning an abstract entity into an agent embodying a moral value, the value was understood through its personification. To combine the agent with the moral value represented creates a fictive character (e.g., Justice), which we understand as standing for the ideal of the same name. Erasmus also classified prosopopoeia as a figure that either described people or presented "a man far away, or long since dead, speaking" (53). For Erasmus, prosopopoeia could be used to "bring back to life the ancient noblemen" who would "rightly exhort us with [...] words" when rhetoricians repeated the words of those they envoiced in an argument or sp eech (53). In essence, Erasmus's distinction of prosopopoeia (i.e., envoicing) from prosopographia (i.e., personification) is one that persists. For example, Richard Lanham's definitions of prosopographia as that which "vividly describes the appearance of a person, imaginary or real, quick or dead" (123), and of prosopopoeia as first and foremost occurring when an "animal or an inanimate object is represented as having human attributes and addressed or made to speak as if it were human" (123), are problematic. As we will see below, "Our Weakness" (the persona of part IV of Auden's "Memorial for the City") is a talking body that is both an instance of prosopographia and an instance of prosopopoeia. That is, Our Weakness personifies an abstract body incapable of speaking although the body is staged by Auden as someone speaking. Even Rodney Edgecombe's recent efforts in this journal "to tidy" (4) the confusion caused by many terms that are synonymous with personification assumes categorical differences based on the targets personified while ignoring the similarity of the source domains involved. Other literary critics who simply equate personification with prosopopoeia alone (Paxson 1) also reveal the limitations of classical rhetoric to handle figures like Auden's.

The rhetorical tradition suggests that personification is merely a categorization issue. All we need to do to deal with it is simply label it either prosopographia or prosopopoeia and move on. But such a task overlooks how we personify, why we personify, or what metaphorical domains are involved. To simply call "Our Weakness" an instance of prosopopoeia is not just wrong: it does not explain how the personification occurs when we read Auden' s poem. If we are to study metaphor at all it must be studied as a joint venture of our thought and our language because the same domains hold for them both. Of course, it is too late for us to question Erasmus or his predecessors, but classical rhetorical categories are limiting when we find examples like Auden' s that do not fit nicely into them. In that case, the story of personification in cognitive linguistics, which begins in 1980 with Lakoff and Johnson, can be useful for clarifying the issue.

For Lakoff and Johnson, personification is an ontological metaphor that "allows us to comprehend a wide variety of experiences with non-human entities in terms of human motivations, characteristics, and activities" (Metaphors 33). Since "motivations, characteristics, and activities" are central to our basic notion of a human being, when we use these traits in source domains to construct a target domain via metaphor, the product of this cognitive process is a personification. As Lakoff and Johnson claim, in personification generally "we are imputing human qualities to things that are not human-theories, diseases, inflation, etc. In such cases, there are no actual human beings referred to. When we say 'Inflation robbed me of my savings,' we are not using the term 'inflation' tQ refer to a person" (Metaphors 35). The target of this expression is inflation, while the source is not a single specific person per se but human traits that become salient for the mapping. in this case, we have personal motivations and a ctions, such as the greedy act of robbing that we map onto "inflation," and the resulting state of affairs whereby money is inevitably or unexpectedly lost. Figure 1 represents this mapping.

"Inflation robbed me of my savings" is meant to convey that money was lost unexpectedly. That is, a certain state of affairs in the target domain results from a certain source domain activity (i.e., robbing). We map a personal trait like having a motive (greed) to carry out an action (rob) onto "inflation," which helps us to personify it. So when Edgecombe states that "Incarnation involves the assumption not only of a human form, but also of a personality" (6), the "assumption" of "personality" results from active cross-domain mapping from a source to a target rather than a passive encounter with the metaphor. On this view, Knapp's idea of the reversible nature of personification does not hold since cross-domain mapping principles, as Lakoff and Johnson would argue, mean that metaphor is neither arbitrary nor reversible. To clarify conceptual domains, and clearly recognize corresponding elements within those domains, is just one advantage of the cognitive approach to metaphor.

While personification may not be arbitrary, whether or not it is "a single unified general process" (Lakoff and Johnson, Metaphors 33) is open to question. Depending on the context, it is possible to read INFLATION IS A PERSON in general as INFLATION IS AN ADVERSARY in particular. However, a general theory of personification metaphor that posits the existence of an infinite number of different source domains is not helpful. There may be no limit to what we can personify (e.g., fear, time, trees, weather, etc.) but we cannot confuse targets with sources or source subcategories (an adversary is a person) with targets. Mapping a personified source domain onto a target domain of an idea or an object or an emotion will yield personification, so there must be something coherent about this source domain. It is the one common denominator for personification metaphors. Indeed, so pervasive is personification that what I call its inevitability arises because of what Daniel Dennett defines as our "intentional stance" in the world. For Dennett, the intentional stance is one we take on a regular basis in life and it works as follows:

[F]irst you decide to treat the object whose behavior is to be predicted as a rational agent; then you figure out what beliefs that agent ought to have, given its place in the world and its purpose. Then you figure Out what desires it ought to have, on the same considerations, and finally you predict that this rational agent will act to further its goals in the light of its beliefs. A little practical reasoning from the chosen set of beliefs and desires about what the agent ought to do; that is what you predict the agent will do. (17)

If Lakoff and Johnson tell us how we personify in terms of cross-domain mappings, Dennett's triangulation between observer, object, and observer's calculation about what the object will do, roughly explains why we personify. And if Dennett is right to argue that we are cognitively predisposed to project or map intentionality across domains, then Auden's personifications of mind and body are the results of this normal predisposition. Indeed, Baron-Cohen's research, which suggests that autism arises from a child's failure to map between body and mind when it comes to registering the behavior of others and displaying accurately his or her own emotional states, bolsters Dennett's claim about the everyday nature of projecting intentions. As Dennett explains, "Treating frogs, birds, monkeys, dolphins, lobsters, honeybees--and not just men, women, and children--from the intentional stance not only comes naturally, but also works extremely well within its narrow range. Try catching frogs without it" (108). If the cog nitive mechanism for projecting intention onto, say, a lobster is the same one for projecting ontological and volitional capacities onto a distinct mind or body in a given Auden poem, then a cognitive take on personification is useful.

After Lakoff and Johnson in 1980, Lakoff and Turner in 1989 told the EVENTS ARE ACTIONS (75-80) story about personification. Lakoff and Turner's model applied to many poetic personifications of Time, Death, and Love. They explain EVENTS ARE ACTIONS as follows:

First, the source domain of actions is a subcategory of the target category of events; that is, every action is an event, though the converse is not true. Indeed, it is exactly the events without agents that the EVENTS ARE ACTIONS metaphor applies to. [...] Second, because actions are events, the mapping from actions to events has a structure somewhat different from the other mappings. Each action consists of an event plus the agency which brings that event about. The mapping thus adds structure to the event domain making the event the result of an action and introducing the agent who brings that action about.

(75; my emphasis)

In this story, a personification's source domain is an ACTION and its target domain is an EVENT. Lakoff and Turner's best evidence for this mapping is Death as the Grim Reaper. Personification of Death the Grim Reaper is due to the fact that an EVENT (dying) is understood as an ACTION (reaping) done by an agent (reaper). As my italics above suggest, the issue of agency is vital because personification occurs when an action's agent becomes an event's causal agent. (1) On this view, personification is a way to describe events. It also suggests one way out of the confusion caused by contrasting personification to metaphor when we fail to see that agency is a human trait we map from source to target. That is, agency is not just a property of the source domain: it is mapped to the target to make causal agency a crucial property of the personified entity in the target domain. This suggests that Knapp's contrast between personifications and agents is misguided. People are always already agents because of our intenti onal stance towards them in this world.

When we personify we bring events like death down to human scale so as to understand them concretely as personified agents analogous to human beings. With the Grim Reaper, we transform death from an event into an action caused by an agent. On this view, agency, is inherent to personification and not exclusive of it. In sum, we personify to make the world make sense to us on human scale. (2)

2. Auden's Personifications of Mind and Body

Bergonzi and others noticed Auden personifying abstractions. "Reason's depravity that takes / The useful concepts that she makes I As universals" (Collected 231), as Auden wrote in "New Year Letter," is just one example. But there is a difference between personifying "Reason" and personifying the body or the mind. The difference is that a person entails both a body and a mind. To personify a non-human target like Reason is one thing, but to personify a human mind or body is another thing. Why? The nature of the mind and the body involves their special relation to a personified source domain. That relation in simple terms is our understanding of bodies and minds as comprising people. Lakoff and Johnson refer to "motivations, characteristics, and activities" (Metaphors 33) as inherent to personification's source domain. However, corporeality is just as central to the domain as is having ontological characteristics like feelings, desires, emotions, and the will to act. That is, the source domain of "person" take s as given essential elements like mind and body. However, when we map these elements into target domains, then we sense a difference between, say, personifying history and personifying the flesh. Of course, to consider a "person" as more than the sum of a mind and a body is part of our normal mental development. What the developmental psychologist Peter Hobson calls our "concept of a bodily-cum-mental person" (211) suggests that our default notion of "person" in childhood encapsulates something that has a mind and a body. Still, this does not prevent Auden from separating the two in dualistic fashion. Indeed, what are we to make of Auden's habit of splitting the mind from the body in poems that include, among others, "1929," "The Quest," "The Sea and The Mirror," "Horae Canonicae," "For the Time Being," "The Age of Anxiety," "Friday's Child," "You," "Ode to Terminus," and "Talking to Myself"? (3) Something odd appears to be happening when Auden turns from personifying abstractions to personifying minds and b odies.

In keeping in line with classical dualism, Auden repeatedly treats the body as other or as object, buying into the very dualism some readers may sense him as challenging. For instance, consider these lines from "Aubade," written in 1972:

               Few bodies
comprehend, though, an order,
few can obey or rebel,
so, when they must be managed,
love is no help: We must opt
to eye them as mere Others,
must count, weigh, measure, compel.
(Collected 881)

As we intuitively sense, dividing the body and the mind is linguistically feasible when either one is personified as an autonomous agent with the ability to act independently of the other. As readers of Auden's poetry will discover, he tends to personify the mind and the body in order to distinguish them. However, what makes these divisions peculiar is their relation to the same source domain of a "person." Here Auden argues that our bodies always seem alien to us, and the systematic conceptual patterns found in Auden's poetry reveal a division of mind and body again and again. Even so, the workings of these less-than-abstract personifications have yet to be fully explained. This is strange since Auden's preoccupations with dualism span his career. They are present in late poems like "Talking to Myself" from 1972 (where death is the divorce of the mind from the body), and they are present in some of his earliest poems from the 1920s. For example, at the age of twenty, he was already writing in 1927 of how the "mind lies to tarnish, untouched, undoing, / Though body stir to sweat, or, squat as idol, brood" (English Auden 21).

Perhaps two poems that go furthest in separating mind from body are "The Mind to Body Spoke," one of Auden's first poems, and "Memorial for the City," one of his masterpieces. Dualism runs throughout "The Mind to Body Spoke," a wonderful poem that Auden wrote when he was about twenty. Written around November 1927, the poem opens as follows with this stanza:

The mind to body spoke the whole night through:
'Often, equipped and early, you
Traced figures in the dust, eager
To start, but on the edge of snow
As often then refused me further;
Proffered a real object, fresh,
Constant to every loyal wish.
                                     (English 441)

The poem's four stanzas depict the mind divided from the body in what looks like a conversation. In the opening stanza above, the mind tells us that the body has "traced figures," "refused," and "proffered." These are actions that the body has performed on its own, to the chagrin of the mind. The opposition of the two is also suggested by the mind as a late riser who has chattered all night, while the body is an early riser. In the next stanza, the mind has regrets. The body would not take it to the "Dark Tower," although both body and mind found time "on the hill crest" marching on away from life. But since "applause" contrasts with the "disillusioned bell," the mind and the body perceive events differently in stanza two where the mind will stand in a storm even when the body's physiological reaction to the danger has made it "frantic." Stanza three, however, suggests there was once a moment where mind and body worked in concert, the voice disguising the "jabber of the blood." The body's poise keeps it calm while Auden implies that the mind behind the voice was troubled.

Finally, the three lines of the last stanza are spoken neither by the mind nor the body but by another persona: "Cocks crew, and sleeping men turned over. / Rain fell for miles; ghosts went away. / The jaw, long dropped, stopped a reply" (English 442). The blather of the mind has little effect on the external world although the dropped jaw appears surprised at what it has had to listen to. The poem's final line reveals that the body too is personified. In a rather impersonal tone, the jaw stops the body's "reply" to the mind. Whereas the body can only speak audibly, Auden suggests that the mind is able to speak inaudibly, is able to speak against the wishes of the body or without its help. Whereas the mind is directly quoted in the first three stanzas, the body's reaction is indirectly reported in stanza four. We never know what the body would have said: it wanted to answer the mind but the jaw did not comply. And yet, the ability of the body to act on its own is a sign of its autonomy, just as the will of th e mind to "speak" freely is a sign of its autonomy. Of course, if the mind rather than the body controls the jaw, Auden may be arguing that the body's attempt to reply is actually thwarted by the mind. In this case, the body can never speak for itself even if it exists apart from the mind here as the mind's interlocutor. Either way, my point is that regardless of where we stand on the issue of bodily control, this very dualism comes through to us in the poem via the personifications presented by Auden. In order to yield the active personifications in the poem, human traits like motivations are mapped by readers from a personified source onto the targets of body and mind. More than mere metonyms for a person, the body and the mind for Auden here are individual personified beings at odds with one another. Ironically, the two are realized by the exact same conceptual process of metaphorical personification, which complicates Auden's dualism with regard to poetics in ways hitherto unnoticed.

In contrast to "The Mind to Body Spoke," the body does get its chance to speak in the enigmatic poem "Memorial for the City." Finished in June 1949, this poem is a "direct result" or "record" of Auden's 1945 Germany trip and is dedicated to the memory of the theological writer Charles Williams, who died in 1945 (Carpenter 337). Hired by the Pentagon because he was fluent in German and knew the country well, Auden toured post-war Germany in summer 1945 to conduct civilian interviews. The Allies wanted to know if prolonged bombing had helped bring about Germany's surrender, and as Auden later said of his stint with the United States Strategic Bombing Survey's "Morale" Division, "We asked them if they minded being bombed. We went to a city which lay in ruins and asked if it had been hit. We got no answers that we did not expect" (qtd. in Pearsall 108). Auden' s long silence about the haunting trip was broken in 1949 with "Memorial for the City," a four part poem.

Part I depicts the world after the Holocaust as a "Post-Vergilian City / Where our past is a chaos of graves and the barbed wire stretches ahead / Into our future till it is lost to sight" (Collected 592). Part II recounts nearly nine hundred years of revolutionary history (from 1075 to 1917) involving cities like Rome, Paris, and Moscow. The chronology lends the part a teleological structure before the action "returns to the present" (Fuller 420) in part III. (4) By the time readers get to part III, the "abolished City" (Collected 594) is hopelessly divided by barbed wire. Auden writes that even "our Image," which is to say the body made in God's image, "has no image to admire, / No age, no sex, no memory, no creed, no name" (Collected 595). When readers get to part IV, this is what they find in the opening lines:

Without me Adam would have fallen irrecovably with Lucifer; he never
 would have been able to cry O felix culpa.
It was I who suggested his theft to Prometheus; my frailty cost Adonis
 his life.
I heard Orpheus sing; I was not quite as moved as they say.
I was not taken in by the sheep's-eyes of Narcissus; I was angry with
 Psyche when she struck a light.
                                                       (Collected 595)

These are the first four lines of an eighteen-line monologue spoken "in riddling terms" (Fuller 420) by a peculiar persona. Since part III ends with the command, "Let Our Weakness speak," Auden provides a name for the persona--the unnamed "image" mentioned earlier in part III--but then leaves the rest to us. Edward Mendelson, the great Auden critic, says all the lines in part IV are "spoken by the body in language it might use if it were autonomous" (322). And yet, how do we understand the monologue here? If the body is speaking in its own terms, then it must be construed as autonomous when we personify it. In other words, Mendelson's "if" is superfluous since autonomy is presupposed by the personification mapping. Still, the riddle is intriguing. On the one hand, since we know that answers to riddles cannot be obvious, we know that "Our Weakness" is not specific enough to answer the riddle's implicit question: who is speaking? Despite the definite article in the poem's title, "the City" that Auden memorializ es is never truly specified as one single city in the poem. On the other hand, Auden's catalog of at least twenty-three allusions in the monologue's eighteen lines gradually moves from the general to the specific and elaborates our mappings.

Throughout part IV, Auden alludes to Adam, Lucifer, Prometheus, Adonis, Orpheus, Narcissus, Psyche, Hector, Oedipus, Orestes, Diotima, St. Anthony, Tristan and Isolde, Galahad, Faustus and Helen, Hamlet, Don Quixote, Don Giovanni, Figaro, Prince Tamino, the Ancient Mariner, Captain Ahab, and a city he calls "Metropolis." These allusions fall into two categories: (1) Ancient and Christian, (2) Medieval and Modern. Ancient allusions in the first eight lines refer to ancient Greek or Roman literature, while three Christian allusions are made to Adam, Christ, and Saint Anthony. The first category of allusions occupies roughly the first half of part IV. The second category of allusions, from medieval and modern literature and opera, takes up the next ten lines of the poem. The allusions reinforce the personification of Our Weakness--the body as narrating persona--although they are so wide and inconclusive in range that the only unifying presence for each allusion is the omnipresent human body, the "I" of nearly ev ery line. The body is that which accompanies all the characters referred to in the course of their actions. In this manner, what some see as "allusion as a technique of exclusion in Auden's work (Smith 318) does not apply here. We need not know what all the allusions mean so as to personify Our Weakness. At some point readers will decide that "the city" that Auden memorializes is in fact the human body. As a body that seems to contradict the anonymous and a historical status suggested for it in part III of the poem, Our Weakness then becomes less generic and more specific with each passing allusion in part IV. It is a multifaceted agent present at a range of specific fictional and historical events, an agent taking part in those events alongside the personalities involved. And yet, this explanation does not answer the lingering question: how do readers cognitively build such an elaborate metaphor, a metaphor which posits a counterfactual body possessing agency (it speaks intelligently), ontology (it feels not hing by various speeches), and intentionality (it will return for the city's judgment)?

3. Personification as Conceptual Integration

Lakoff and Turner's EVENTS ARE ACTIONS story nicely explains personifications like Death the Grim Reaper, but the story is misleading when something other than an event is personified. Turner sensed as much in The Literary Mind, when claiming that the "target story in EVENTS ARE ACTIONS need not be an event performed by an actor" (45). If we agree that minds and bodies are not events, then EVENTS ARE ACTIONS may not account for our understanding of something like "Our Weakness." Perhaps another story then is needed for cases like these. The recently proposed model of "conceptual integration" (Fauconnier and Turner "Formal," "Networks," "Mechanism," Way) might just offer such a story.

Conceptual integration as a theory offers another description for literary personifications to the cognitively inclined literary critic. It also provides an answer to the how question posed above with regard to Our Weakness. Fauconnier and Turner's theory of blending (as conceptual integration is often called) is "complementary" to the Lakoffian conceptual metaphor theory (Grady, Oakley, and Coulson 101). Whereas conceptual metaphor theory assumes two essential domains (source and target) and mapping between them for the production of metaphor, blending theory assumes at least four mental spaces: two input spaces, a generic space which identifies what the two inputs abstractly share as common ground, and a blended space where emergent meaning arises from selective projections out of the input spaces. The typical model for this mental process, in visual terms, is seen in figure 2. The basic network is comprised of four circles, with arrows linking them all up, and with dotted lines indicated correspondences ac ross the input spaces.

For those unfamiliar with Fauconnier and Turner's model, examining blends is always a bottom-up process: one finds an artifact that results from blending (e.g., an analogy) and unpacks the psychological steps that produced the result. Analyzing blends is very rarely a top-down process where one places two random input spaces side by side and then predicts what the resulting blend will be. In other words, blending is a process, not a predictor of results, and its utility arises when items like analogies, counterfactuals, and metaphors are not always describable by the dual-domain model offered by conceptual metaphor theory. Fauconnier and Turner first developed the blending model to try to better describe phenomena like analogies and analogical metaphors where more than two terms or domains were activated. Since then, the framework has been applied widely in many fields for a variety of reasons and research programs. (5)

More recently, Turner has proposed that the blending model even explains the genesis of writing ("Successive"). In order to write words and read writing, human beings first needed to imagine this as an extension of a basic communication situation where one person speaks to another person. Then people had to imagine marks on a page as visually standing for words people understood audibly. People grasp the basic communication situation of verbal exchange very early on in life, but they only learn to read and write several years later, if at all, and only after making a tremendous effort (Sperber). Fauconnier and Turner would argue that learning how to read and how to write is a matter of gradually acquiring a conceptual integration network for extended communication situations. At its most subtle level, of course, storytelling requires conceptual integration. The story of story told by Turner in The Literary Mind, for example, is that stories are "complex dynamic integrations of objects, actors, and events" whi ch deeply depend on "abstract stories that apply to ranges of specific situations" (10). Since we cannot know or memorize all the stories we will ever hear, read, imagine, or produce, cognitive efficiency demands that we have a supple mechanism for understanding or producing the infinite number of stories we will encounter in our lifetimes. Conceptual integration is such a mechanism, linking general to specific elements in real or imaginary situations.

To return to Auden, Our Weakness is a personification constructed by conceptual integration in two ways. First, when we encounter any poem there is always already a blend involved. As Todd Oakley has argued in his recent analysis of blending in lyric poetry, to read a poem we must presuppose a storytelling input space where one person is talking to another, and another input space where a specific persona narrates the events depicted in the poem for us. In the case of Auden's poem, we imagine that Our Weakness narrates part IV of the poem to us because we blend together a STORYTELLING input space with an OUR WEAKNESS input space to yield a talking (and hence counterfactual) BODY, which becomes the focal participant and persona of part IV of "Memorial for the City." This talking body blend is also present for the talking mind in "The Mind to Body Spoke." However, it is more specific in "Memorial for the City" because the generic persona becomes highly specified owing to Auden's many allusions. This brings us t o a second explanation of the blend in the poem. The allusions related to Our Weakness prompt readers to blend the general and the specific even if the full contents of various allusions escape them. In general, the allusions ask us to solve the "who is talking here?" riddle by taking the STORYTELLING input space to be GENERIC and the OUR WEAKNESS input space to be SPECIFIC. The talking BODY in the blend that we specify through Auden's two dozen allusions seemingly redefines the first-person speaker step by step in the poem. On the one hand, Auden's "I" metaphorically coheres throughout the poem at one level of blending since we take it to be the same persona delivering each line to us. As Fauconnier might explain, an identity principle maintains (12) across mental spaces so that, for example, in the line "I heard Orpheus sing; I was not quite as moved as they say," a single pronoun indicates both a generic "I" (persona) and a specific "I" (physically present body at the scene with Orpheus).

To understand Auden's personification along these lines is to see "Our Weakness" as a result of two blends. First, it is a blend built by integrating a storyteller in an unspecified narrative with a persona in a certain poem. Second, it is a blend built by integrating the persona in the poem with figures in each highly specific allusion. The first blend, as Oakley might argue, is not specific to Auden since it arises from our default cognitive model for reading poetry. The second blend, however, directly results from Auden because his allusions here are specific and unique to this one poem. Also, a GENERIC IS SPECIFIC basic metaphor (Lakoff and Turner 162) relates to the input spaces since a specific OUR WEAKNESS input space contains the contents of the allusions. (6) In each allusion, the body is historical, an omnipresent companion from Adam's time until Melville's Moby Dick character, Captain Ahab. Thus, our construal of the body here as a "person," because of its corporeal presence in the poem, leads to w hat we might simply call Auden's personification of the body. However, making such a statement takes for granted that which needs explaining: how, in cognitive terms, such a personification is possible.

To embody and personify Our Weakness, we conceptually integrate the abstract and concrete to yield an emergent figure: a speaking body possessing those traits we attribute to people. In the blend, the BODY has a name just as a person would ("Our Weakness"); it has memories of events (historical presence) like the Fall; it has the agency required to burst out with an eighteen-line monologue (voice); and it has intentions (it will "hear" the city judged).

Figure 3 above is a cognitive model of the blend prompted by part IV of "Memorial of the City." The generic story space is shared by both inputs since the basic communication situation represented in Input I and the poetic situation represented in Input 2 have subject matter delivered by a speaker to a hearer (any story's basic ingredients). The information within Input 1 is general relative to our world knowledge, while the information within Input 2 is specific relative to Auden's poem. The blend itself is specific because of the allusions and it is counterfactual because of our world knowledge: bodies just by themselves cannot speak. Even if they could speak, they might not speak as elegantly as Our Weakness does. In essence, figure 3 is meant to represent the cognitive mechanism we possess to be able to personify an entity like Our Weakness in Auden's poem. (7) As Fauconnier and Turner put it, however, "such a diagram is always a snapshot of an imaginative and complicated development" (Way 85). That is, t he diagram is just a diagram--no more, no less--although it exists to model in visual terms how a reading mind processes and produces Auden's personification online.

At this point some critics might disagree with me and reject Fauconnier and Turner's findings as irrelevant. While we rarely argue about facts, we do argue about the explanations of those facts in literary studies. Of course, hostility to Turner's ideas (cf. Gross) is symptomatic of the view that cognitive science, or any science for that matter, has nothing to offer the literary critic involved with explications de texte. Skepticism is normal in literary studies as long as historically-oriented hermeneutics reigns supreme, but it becomes less so when we deal with poetics, with the cognitive mechanisms for signification that are already studied under various rubrics (cognitive approaches to literature, cognitive rhetoric, cognitive poetics, cognitive philology, or cognitive theory) on different campuses worldwide. These projects (1) aim to discover as many findings from cognitive science as shall be useful to the analysis of literature and (2) apply those findings that are relevant to the analysis of literatu re. For the moment, figurative language research in cognitive linguistics offers the most help to literary scholars, but the field will no doubt expand in future as the phenomena accounted for by cognitive science gradually increase. In short, if conceptual integration as "a theoretical framework for exploring human information integration" (Coulson and Oakley 176) is to be useful for literary critics, it will be useful in how the theory forces us to rethink or redefine some of the processes we take for granted when reading and understanding literature.

If conceptual integration does not account for Auden's personification, then we might ask what rival hypothesis can account for it. Perhaps one explanation would be to elide the personification problem altogether by simply calling Auden's persona a good old fashioned literary character. As Ralf Schneider successfully argued recently in this journal, a literary character is "a mental model that the reader construes in the reading process through a combination of information from textual and mental sources" (2). Essentially, as Schneider cogently argues, to take a view like this would be an exercise in categorization (5) since we would situate something like Our Weakness in the category of person based on similarities we find between people and personae like Auden's. According to Schneider, "categorization" is a top-down cognitive process for mentally construing a literary character, while "personalizing" (seeing the character as a specific entity) would be a bottom-up process (14). In this fashion, one rival h ypothesis might call Our Weakness nothing more than a character (categorization) narrating in the context of Auden's poem (personalization). As a character, it would have human qualities since we assume characters are simulations of humans. As a personalization, it would have the specific traits of those characters that Auden alludes to.

A rival hypothesis like this of characterization seems adequate, but it rehearses the problem mentioned earlier in reference to the rhetorical tradition. To categorize Our Weakness as a character or a characterization or a personalization does not explain the cognitive mechanisms involved in such procedures. Additionally, Lakoff and his colleagues insist contra Sam Glucksberg and Matthew McGlone that metaphor is not categorization, while Fauconnier and Turner might add that a character in Schneider's terms is a blend created by blending a mental source with a textual source from the bottom-up and from the top-down. Clearly, then, arguing against the conceptual integration hypothesis is a real challenge. As Fauconnier and Turner state, "In one sense, everybody knows everything about blending and is a complete master of the operation, in just the sense that each of us has complete unconscious 'knowledge' about vision but almost no conscious knowledge of our unconscious ability" (Way 54). In essence, the researc h program on blending is an effort to make explicit an operation that usually escapes our notice. As Turner has argued about talking animals found in children's books ("Blending"), personifications of such characters are blends that even toddlers produce rather effortlessly. However, personification's simplicity does not mean that the process itself is simple to explain or that it is unworthy of explanation. Of course, to revamp conceptual integration theory means finding ways to falsify it. That is no easy task. Like other prominent theories, arguments are always already available for countering rival hypotheses by showing that the countering claims themselves arise somehow from conceptual integration at grammatical, figurative, or literal levels.

4. Conclusion

Literary critics may no doubt find blending theory useful for a number of things, but they might face what sociologists call "unknown unknowns." You never know what you do not know, so you never know what else might account for the evidence under investigation. When you think you have explained something fully and to the best of your ability, another unknown factor may in fact offer a better answer. In this sense, everything I have said here should prompt a new investigation into literary personification if cognitive theories of this trope force us to redefine our concepts. Having said that, despite Shapiro's criticisms of Auden's personifications, the trope is justifiable and even inevitable to the argument in poems like those analyzed here. For certain, personification is not arbitrary, it is not ornamental, and it is certainly not a symptom of Auden's waning talent. Personification is as persistent in Auden's work as it is in everyday thought and language. Many verbs that we insert into the phrases "The bo dy ..." or "The mind will in some manner reveal our attribution of ontology, volition, and agency (in short, human traits) to them. For example, the body is, feels, has, desires, knows, hopes, seems, and reminds. The mind thanks, defeats, gets ready, embraces, resigns, reckons, and speaks. Auden supplies verbs like these in his poems when personifying the body and the mind, which means that the verbs' various "participant roles" often "profile" an agent relevant to each verb's semantics (Goldberg 46-53).

Auden's consistent use of personification challenges the commonplace notion that his changing views about the mind and the body, over the course of his career, worked themselves into his representations of the mind and the body in his poetry. This view, taken recently by Mendelson in Later Auden (277-416), is questionable when analyzing the evidence through a cognitive linguistic lens. As I see it, Auden personifies the mind and the body throughout his poetry because they can only be imagined metaphorically. And because anthropomorphizing from our intentional stance is straightforward personification, Edgecombe's claim that we can cross "from personification into anthropomorphism" (4) seems misleading. Whether we prefer the Greek term or-the Latin term, the conceptual processes that these terms represent are one in the same. The many labels Edgecombe provides for personification indicate different targets, not different sources. This is something we may only notice when taking cognitive views on personificati on. If we hold the metaphor to be primary to our conceptual life, personification's target domain can vary but not its source domain.

As for reading Auden, there are two consequences to bear in mind. First, the same metaphor of personification is used to represent two apparently different targets. This might suggest that in place of dualism we find unification in Auden' s figurative language. If so, this would be an advance in Auden scholarship made possible by cognitive theory. Second, Auden's mind and body metaphors pose a challenge to dualism. Our belief in dualism needs revising if "l'esprit n'est pas situe dans la tete" (Varela 10) and if "the mind is inherently embodied" (Lakoff and Johnson, Philosophy 3). If the mind is not in the head, what is it and where is it? As cognitive scientists Stephen Kosslyn and Olivier Koenig have put it, "The mind is what the brain does" (qtd. in Crane and Richardson 131), a statement which the cognitive linguist Bert Peeters has altered into "The mind is what the brain does for a living" (3). Remarks like these ground psychology in biology in a manner that may have surprised past scholars. But as the e mbodiment hypothesis from cognitive semantics purports, the body's role in this combination is fundamental. For cognition to happen, the mind must be embodied. No body, no cognition. Indeed, Johnson refuses to distinguish the mind from the body for he thinks that the mind is the body just as the body is the mind. Such ideas should fuel a new analysis of dualism in literary studies. I offer my research on Auden as an initial step in that interesting direction. (8)

Notes

(1.) Turner has explained EVENTS ARE ACTIONS in more detail elsewhere (Reading 123, 161-62).

(2.) The cognitive linguistic story of personification trails off with Lakoff and Turner since Lakoff and Johnson in 1999 only mention it in passing when they refer to MIND AS BODY (248) and MIND AS PERSON (339) metaphors for the mind.

(3.) These poems, and others, are studied in depth in Hamilton's Opaque Enigmas.

(4.) Auden's understanding of revolutions in the 1940s is thoroughly explicated by Mendelson (310-28).

(5.) For a full list of blending projects by scholars around the world, please see Turner's Web site at <http://www.wam.umd.edu/~mturn>.

(6.) For a more detailed discussion of GENERIC IS SPECIFIC, please see Grady ("Typology" 91-95).

(7.) As Oakley puts it, blending "is a complete model of cognitive framing [...] that gives a plausible account of the intervening cognitive operations that occur between the perception of signs and their interpretations" (14).

(8.) For useful comments on earlier versions of this article, I thank Mark Turner, Joseph Grady, Todd Oakley, and two anonymous reviewers at Style.

Works Cited

Auden, W. H. Collected Poems. Ed. Edward Mendelson. 1976. New York: Vintage, 1991.

_____. The English Auden: Poems, Essays, and Dramatic Writings, 1927-1939. Ed. Edward Mendelson. New York: Random, 1977.

Baron-Cohen, Simon. Mindblindness. Cambridge, MA: MIT P, 1995.

Bergonzi, Bernard. "Auden and the Audenesque." Encounter (Feb. 1975): 65-75.

Carpenter, Humphrey. W.H. Auden: A Biography. Boston: Houghton, 1981.

Coulson, Seana and Todd Oakley. "Blending Basics." Cognitive Linguistics 11.3-4 (2000):175-96.

Crane, Mary Thomas and Alan Richardson. "Literary Studies and Cognitive Science: Toward a New Interdisciplinarity." Mosaic 32.2 (1999):123-40.

Crisp, Peter. "Allegory: Conceptual Metaphor in History." Language and Literature 10 (2001):5-20.

Dennett, Daniel. The Intentional Stance. Cambridge, MA: MIT P, 1987.

Edgecombe, Rodney. "Ways of Personifying." Style 31.3 (1997):1-13.

Ehrenpreis, Irvin. "Inside Auden's Landscape." W. H. Auden: The Critical Heritage. Ed. John Haffenden. London: Routledge, 1983. 496-504.

Erasmus. On Copia of Words and Ideas. Trans. Donald King and H. David Rix. Milwaukee: Marquette UP, 1963.

Fauconnier, Gilles. Mental Spaces. Cambridge: Cambridge UP, 1994.

Fauconnier, Gilles, and Mark Turner. "Conceptual Integration and Formal Expression." Metaphor and Symbol 10.3 (1995):183-203.

_____. "Conceptual Integration Networks." Cognitive Science 22 (1998):133-87.

Fauconnier, Gilles. "A Mechanism for Creativity." Poetics Today 20 (1999):397-418.

_____. The Way We Think. New York: Basic, 2002.

Fuller, John. W.H. Auden. A Commentary. Princeton: Princeton UP, 1998.

Gibbs, Raymond W., Jr. "Why Many Concepts are Metaphorical." Cognition 61 (1996): 309-19.

Glucksberg, Sam and Matthew McGlone. "When Love is Not a Journey: What Metaphors Mean." Journal of Pragmatics 31 (1999): 1541-58.

Goldberg, Adele. Constructions: A Construction Grammar Approach to Argument Structure. Chicago: U of Chicago P, 1996.

Grady, Joseph. "A Typology of Motivation for Conceptual Metaphor: Correlation vs. Resemblance." Metaphor in Cognitive Linguistics. Eds. Raymond Gibbs and Gerard Steen. Amsterdam: John Benjamins, 1999. 79-100.

Grady, Joseph, Todd Oakley, and Seana Coulson. "Blending and Metaphor." Metaphor in Cognitive Linguistics. Eds. Raymond Gibbs and Gerard Steen. Amsterdam: John Benjamins, 1999. 101-24.

Gross, Sabine. "Cognitive Readings; or, The Disappearance of Literature in the Mind." Rev, of Reading Minds by Mark Turner. Poetics Today 18 (1997): 271-297.

Hayles, N. Katherine. "Simulating Narratives: What Virtual Creatures Can Teach Us." Critical Inquiry 26 (1999): 1-26.

Hamilton, Craig. Opaque Enigmas: Mind, Body, and Metaphor in W.H. Auden's Poetry. Diss., University of Maryland, 2001.

Hobson, Peter. "Understanding Persons: The Role of Affect." Understanding Other Minds: Perspectives from Autism. Eds. Simon Baron-Cohen et al. London: Oxford UP, 1993. 204-227.

Johnson, Mark. The Body in the Mind. Chicago: Chicago UP, 1987.

Knapp, Steven. Personification and the Sublime: From Milton to Coleridge. Cambridge, MA: Harvard UP, 1985.

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

___. Philosophy in the Flesh. New York: Basic, 1999.

Lakoff, George, and Mark Turner. More Than Cool Reason. Chicago: U of Chicago P, 1989.

Lanham, Richard. A Handbook of Rhetorical Terms. 2nd ed. Berkeley: U of California P, 1991.

MacKay, Donald. "Prototypicality Among Metaphors: On the Relative Frequency of Personification and Spatial Metaphors in Literature Written for Children Versus Adults." Metaphor and Symbol 1.2 (1986): 87-107.

Mendelson, Edward. Later Auden. New York: Farrar Straus Giroux, 1999.

Oakley, Todd. "Visceral Poetics." Unpublished essay, 1999.

Paxson, James. The Poetics of Personification. Cambridge: Cambridge UP, 1994.

Pearsall. Cornelia. "The Poet and the Postwar City." Raritan 17.2 (1997): 104-20.

Peeters, Burt. "Does Cognitive Linguistics Live Up to Its Name?" Unpublished essay, 1999. <http://www.tulane.edu/~howard/LangIdeo/Peeters/Peeters.html>

Schneider, Ralf. "Towards a Cognitive Theory of Literary Character: The Dynamics of Mental Model Construction." Style (forthcoming).

Shapiro, Karl. Essay on Rime. London: Secker and Warburg, 1947.

Smith, Stan. "Remembering Bryden's Bill: Modernism from Eliot to Auden." Critical Survey 6.3 (1994): 312-22.

Sperber, Dan. "Vers une lecture sans ecriture?" Text-e Colloque Virtuel. January 2002.<http://www.text-e.org/conf/index.cfm?fa=printable&ConfText_ID= 12>

Sweetser, Eve. From Etymology to Pragmatics. Cambridge: Cambridge UP, 1990.

Turner, Mark. "Blending in Children's Literature." Keynote Address. 10th Languaging Conference, University of North Texas. 28 January 1998.

_____. The Literary Mind. New York: Oxford UP, 1996.

_____. Reading Minds. Princeton: Princeton UP, 1991.

_____. "Successive Blends in Material Culture, Part II: Writing, Speech, Sign Language, Gesture." 7th Annual ICLC Conference, UCSB. 26 July 2001.

Varela, Francisco. "Quatre Phares Pour l'Avenir des Sciences Cognitives." Theorie Litterature Enseignement 17 (1999): 7-22.

Craig A. Hamilton (Craig.A.Hamilton@nottingham.ac.uk) teaches rhetoric and literary linguistics at the University of Nottingham, where is a lecturer in English. He has published on Joyce, Lawrence, Pound, HD, Shaw, and Strand. At present, he is writing one book on Auden and plans to edit two others on stylistics and rhetoric soon.

COPYRIGHT 2002 Northern Illinois University
COPYRIGHT 2003 Gale Group