Forthcoming in Mark Johnson (ed.), Journal of Metaphor and Symbolic Activity, volume 10, number 3.
Mark Turner
Department of English
Center for Neural and Cognitive Sciences
University of Maryland 20742
markt@umd5.umd.edu
Gilles Fauconnier
Department of Cognitive Science
University of California
San Diego CA 92093-0515
faucon@cogsci.ucsd.edu
This document is also available as a postscript file.
Abstract: We pursue here our exploration of conceptual blending and of the "many-space" model, which replaces the standard "two-domain" model. 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. New examples are presented. We show that meaning is not compositional in the usual sense, and that blending operates to produce understandings of composite forms. Formal expression in language is a way of prompting hearer and reader to assemble and develop conceptual constructions, including blends; there is no encoding of concepts into words or decoding of words into concepts. Blending is at work in many areas of cognition and action, including metaphor, counterfactuals, and conceptual change. We point out two fundamental aspects of this general process: cross-space mapping of counterparts, and integration of events.
In a customary view of conceptual metaphor--one we use in our own previous work--metaphor carries structure from one conceptual domain (a "source") to another (a "target") directly. (Lakoff & Johnson, 1980; Johnson, 1987; Turner, 1987; Lakoff & Turner, 1989; Sweetser, 1990; Turner 1991) In this approach, "conceptual domain" refers to a vast organization of knowledge, such as our knowledge of journey or dreaming or education. A conceptual domain has a basic structure of entities and relations at a high level of generality--for example, the conceptual domain for journey has roles for traveler, path, origin, destination, and so on. A conceptual metaphor consists of a (partial) mapping of the basic structure of one conceptual domain (the source) onto another (the target).
This "two-domain" model is highly parsimonious, and it is useful and effective for a number of purposes in cognitive studies--such as the ongoing hunt for conventional conceptual metaphors. But in Fauconnier and Turner (1994) we argue that the two-domain model is actually part of a larger and more general model of conceptual projection. We call this new and competing model the "many-space" model. The many-space model explains a range of phenomena invisible or untreatable under the two-domain model and reveals previously unrecognized aspects of even the most familiar basic metaphors.
We use the term "mental space" in contrast to the term "conceptual domain." A mental space is a (relatively small) conceptual packet built up for purposes of local understanding and action. Mental spaces are constructed whenever we think and talk. They are interconnected, and they can be modified as discourse unfolds (Fauconnier, 1994 [1985]). For example, we may build up a mental space to understand the phrase "my hike along the Appalachian trail in Carlisle, Pennsylvania in 1990." This mental space will include the hike, the hiker, the date, the location, and so on. It will recruit partial structure from the conceptual domain of journey, but only a small amount of the knowledge associated with journey will be explicitly recruited in building up the mental space. Additional structure becomes available through default and pragmatic procedures. A given mental space often recruits structure from more than one conceptual domain.
We propose a model of conceptual projection across four or more ("many") mental spaces rather than two domains. This introduces a higher degree of variability and a loss of parsimony, but with a corresponding increase in sensitivity and generality. The two-domain model is in fact a special case of the many-space model.
The many-space model assigns roles to the two input spaces ("source" and "target" in a metaphor or analogy) but also to two middle spaces--a generic space--which contains skeletal structure that applies to both input spaces--and a blended space--which is a rich space integrating in a partial fashion specific structure from both of the input spaces. The blended space often includes structure not projected to it from either input space.
The relationships among these four spaces, and the mechanisms of conceptual projection among them, are intricate, going beyond any of the mechanisms that have been proposed in philosophical traditions under the name of "compositionality." Moreover, this non-compositional conceptual complexity frequently plays an indispensable role in reasoning and inference. The many-space model brings to light essential and powerful work done by metaphor, blending, and conceptual projection in human affect, invention, and discovery.
We examine the relation of conceptual blending to formal linguistic expression. It turns out that there is a process of formal blending at the level of expression with the same general characteristics as conceptual blending at the level of thought. But interestingly, and rather unexpectedly, conceptual blends in thought are seldom mirrored by formal blends in language. Language has other means for prompting us to perform conceptual integration. We discuss some of them, which we think provide a novel perspective for the long-standing philosophical investigation into the relation between form and meaning.
The blend does much more than just make the positions easy to compare. We find that the blend is structured by a frame that was absent from the two input spaces, and is not a logical consequence of their composition: a match race between contestants, over a single course, with a winner and a loser. We showed in some detail that the richer structure of this blend is used cognitively in essential ways. With respect to truth-conditionality, it provides non-trivial information about the positions and speeds of the two boats, the likelihood of breaking the record, the value of making the attempt, and so on. With respect to intentionality, the race frame puts the crew of the catamaran into the position of trying to maintain a lead over the clipper, or in general coming up with a strategy that takes them first over the finish line. With respect to emotions, it provides the possibility of determination to win, anxiety over the prospect of defeat, or exultation at the actual victory. Emotions and intentions influence the outcome of an activity. In this case, they might contribute to success: conceiving the activity as a race might give the crew of the catamaran the psychological disposition and adrenalin they need to end up in one state rather than another.
This kind of process begins with two input spaces --1853 and 1993--and produces a third space in which cognitive and linguistic work is accomplished. This third space--the blend--is both less and more than the two input spaces. It is less in taking only partial structure from each of the two input spaces. For example, it does not take both dates from both input spaces. It does not take conditions of weather, differences in actual course run by the two ships, different principal purposes of the two ships (commerce versus adventure), and so on. The third space is also much more than the two input spaces: in particular, it has two boats instead of one , and a frame for racing that is absent from both input spaces. The fourth space of the many-space model is the generic middle space, a skeletal space that contains structure that is taken to apply to both of the input spaces.
In the previous study, we showed how the construction of blended spaces is involved in reasoning, imagination, action, emotion, and expression. Blending is a general cognitive operation, operating over categorization, the making of hypotheses, inference, and the origin and combining of grammatical constructions. Blending can be detected in everyday language, idioms, creative thought in mathematics, evolution of socio-cultural models, jokes, advertising, and other aspects of linguistic and nonlinguistic behavior.
Let us consider blending in an example of literary metaphor. In Shakespeare's King John, a messenger enters, looking fearful, and the King, reading the disturbance in his face, says:
So foul a sky clears not without a storm.
Pour down thy weather.[2]
We could read this simply as projecting metaphorically from the source to the target in the standard fashion: the appearance of the sky projects to the appearance of the face; the bad weather projects to the bad message; the event of precipitation projects to the act of delivering the message; and so on. This projection is all the more familiar to us because it is an instance of The Conduit Metaphor: a message is an object in a container; communication is transferring an object spatially from the speaker to the hearer.
So far, so good. But there is a subtler reading of this scene, according to which the rule of King John is being portrayed as fundamentally unstable. At this moment, King John, a usurper, has apparently succeeded in having the rightful heir to the throne murdered, causing a great defection of powerful nobles. King John's command is at once royal and illegitimate, effective and impossible. He is a king who is not a king, commanding in the role of king what he personally cannot effectively command. He is not naive or mad, and knows that his rule at this moment is profoundly conflicted and inconsistent, contradictory and insecure.
This is a subtle reading. These inferences, we suggest, involve a blended space, in which the messenger, the prime example of something absolutely under the king's command, is also nature, the prime example of something that is absolutely above the king's command. This is a blend of contraries. King John is commanding what he can command, but what he can command also turns out to be simultaneously what he cannot command. The tension in King John's command is symbolized in the identity of what he can command (the messenger) and what he cannot command (nature). It is a paradox, and a powerful one.
The tension of the blended space is reinforced by another blending of impossibilities. King John is metaphorically above the messenger, in the sense of having power over him. He is also probably physically above him: the messenger may be kneeling. By contrast, a human being on earth is spatially below a raining sky, and metaphorically below it as well, in the sense of being subject to its power. In the blended space, King John is literally and metaphorically above the messenger but literally and metaphorically below the sky. He is simultaneously above and below the messenger-sky. This paradox is compacted into a single expression: "Pour down!" This must appear a very odd thing to say to a bowing or kneeling messenger. In the source space of people and skies, John is unproblematically below; in the target space of kings and their dominions, he is unproblematically above. Only in the blended space is he both. This is fully possible in the blended space, but such a disagreement of usually concordant image-schemas is not possible in the target or the source.
This vivid presentation of John as inhabiting an impossibly unstable position has been taken as constituting the power and memorability of this passage, and as distinguishing it from an insipid expression such as "he has a stormy countenance." But these central inferences are unavailable except in the blended space. The reading that something is seriously wrong in the structure of John's kingship is not available in the source or in the target. It comes only from the blended space.
The scene is profoundly ironic, exactly because of the "ironic tension" in the blended space between the discordant image-schemas. Without the blend, there would be no tension, and without the tension, there would be no irony.
Just as in the boat race example, the new space cannot be obtained by standard compositionality from the input spaces. Neither the input spaces nor their union or intersection contains the discordant image schemas, the ensuing feeling of irony, and the implications relative to King John's fragile rule. The blend in this case recruits conventional metaphors but does new work with them. We explore a range of such interactions between blending and metaphor in Turner and Fauconnier (in preparation).
Now take a superficially very different example, offered by Dan Gruen, which involves the performance of a specific activity. Human-computer interfaces are often structured by the concept of a desktop, on which objects rest and can be manipulated and used to perform actions. The appearance of the computer screen carries icons corresponding to objects on a desktop. They can be opened and closed, put away, and so on. When working with the icons, we think of them and act upon them in some ways as we would on actual desktop material, and in some ways as when dealing with general computer commands. Clearly, the entire activity is coherent and integrated, once learned. It is not hampered by its obvious literal falsities: there is no actual desk, no folders, no putting of objects into folders, no shuffling of objects from one folder to another, no putting of objects into the trash, and so on.
Again we find that the created blended space has considerable structure that cannot be the product of simple compositionality of the inputs. For instance, dragging icons with the mouse belongs to neither moving objects on a desktop nor giving standard symbolic commands, or a fortiori using the machine language. The user is not manipulating this computer interface by means of an elaborate conscious analogy, but as an integrated form with its own coherent structure and properties. From an "objective" point of view, this activity is totally novel--it shares no physical characteristics with moving real folders, and it is novel even for the traditional user of a computer who has issued commands exclusively from a keyboard rather than from a mouse. Yet the whole point of the desktop interface is that the integrated activity is immediately accessible and congenial. The reason, of course, is that a felicitous blend has been achieved which naturally inherits, in partial fashion, the right conceptual structure from both input domains, and then cultivates it into a fuller activity under pressure and constraints from reality and background knowledge.
The desktop example also nicely illustrates the non-arbitrary nature of blending: not just any discordant combination can be projected to the blend. In the King John example, discordant image-schemas were deliberately built into the blend for the purpose of a central inference and of irony. In the desktop example, some discordant structure is irrelevant because it has no bad consequences--e.g, the trash can and the folders both sit on the desktop--but other discordant structure is objectionable--dragging the icon for a floppy disk to the trash as a command to eject the disk from the drive is notoriously disturbing to users. The inference from the domain of working at a desk that everything going into the trash is lost, and from the domain of computer use that everything deleted is irrecoverable, interfere with the intended inference that the trash can is a one-way chute between two worlds--the desktop interface and your actual desk.
Another point illustrated by the example is that input spaces are themselves often blends, often with an elaborate conceptual history. The domain of computer use has as input spaces, among possible others, the domain of computer operation and the domain of interpersonal command and performance. It is common to conceive of the deletion of files as an operation of complete destruction performed by the system at the command of the user. In fact, in the domain of actual computer operation, the files are not erased by that command, and can often be recovered. The user's sense of "deletion" is already a blend of computer operation and human activity. More generally, it is the fact that, by means of blending, keyboard manipulation is already conceived as simultaneously typing and high-level action and interaction that provides the appropriate partial structure to later blends like desktops with icons. The existence of a good blend can make possible the development of a better blend.
The construction of blended spaces is a key aspect of conceptual change in everyday life and in science. We have argued that complex numbers in mathematics achieved mathematical conceptual status only after the successful construction of an elaborate blend, even though the formal correspondences required between space and "imaginary" numbers had been known for centuries.
The expression "dolphin-safe" in all these cases provides motivation but not compositional prediction for the much richer blend that we need to understand the expression. The understander in all these cases must unpack minimal linguistic cues to reach large conceptual arrays over which to perform blending. Just as the race frame structured the boat race blend even though it could not be applied to either of the two input spaces, so the target scenarios here--genetic engineering or tuna fishing--are indispensable to the blend, regardless of their degree of association with the input domain of dolphins and the input frame of safety.
The illusion of central compositionality makes possible the mistaken view that such examples are marginal and exotic, constructed by exceptional processes that we do not have to consider if we are doing "core semantics." On this illusion, "dolphin-safe" and "football pool" work by different principles than "red pencil" and "green house," which serve as (wordplay-safe) canonical examples. But, following Charles Travis (1981), we observe that noncompositional conceptual integration is just as necessary in these "core" cases. "Red pencil" can be taken to mean a pencil whose wood has been painted red on the outside, a pencil that leaves a red mark (the lead is red, or the chemical in the pencil reacts with the paper to produce red, or . . .), a pencil used to record the activities of the team dressed in red, a pencil smeared with lipstick, not to mention pencils used only for recording deficits. For a set of houses that differ only in location and in the color of the kitchen linoleum, "green house" can mean the house with green linoleum, where "green linoleum" means the one with spots that are green, where "green spots" means spots created by using a green pencil, where "green pencil" means . . . .
The scenarios needed for these integrated meanings are no simpler than those needed for supposedly exotic examples like "dolphin-safe" and "fool-proof." The cognitive processes needed to construct these integrated meanings are the same as those needed to interpret the supposedly exotic examples, and these cognitive processes apply as well to the supposedly central examples like "green house" used to mean a house whose relevant exterior surface is green. But the situation is not symmetric: compositional models designed to explain these "core" examples do not suffice for the rest of the data, and so simply defer answering the general question. But for us, the fact that some interpretations stand out more than others--especially when the phrases are taken in isolation--stems from the existence of strong defaults. This difference has to do with the conceptual and linguistic defaults most likely to be activated in any given situation, not with the mechanisms of integration. In any event, an explanation is necessary for the invention and construal of the so-called exotic examples. If that explanation is general over the "core" cases as well, this shifts the burden onto those who need to partition the data in order to maintain compositionality. In fact, the valuing of cases like "brown cow" as central seems to derive less from their role in everyday language than from the fact that they have such strong defaults that they make the compositional explanation seem obvious.
In these circumstances, a theory of compositionality evidently depends covertly upon default principles. If we look across the central cases, we find that some of the relevant defaults are provided by cultural frames with rich structure, others by generic roles that run across many frames, others by local situation at the moment of utterance. This last case includes elicitation by linguists and philosophers: the subject is ostensibly asked to judge an expression in a way that is supposedly context-free, but in fact must construct a minimal context in which to interpret it. Such minimal contexts typically use the strongest defaults.
Restricted to defaults, compositionality would address only a small set of data, the default compositions. But since a theory of defaults is needed in any event, and since blending is independently motivated, we take the view that a general, and uniform explanation is based on conceptual integration through blending.
Langacker (1987) also makes a strong case for integration. He still allows for cases of full, as opposed to partial, compositionality, giving examples like "black bird," contrasted with "blackbird." In the present account, even these prototypical forms are themselves blends constructed along entrenched lines, using strong defaults. Of course, a blend constructed repeatedly across similar situations can be stored in its integrated form and used accordingly. But this is a difference of conventionality and familiarity, not in the mechanisms of achieving that integration. Just as "blackbird" is presumably stored as a unit, "black bird" with the default reading of "bird with black plumage" may also be stored as a unit. Understanding "black bird" as "bird assigned to wear the black falconry hood" for the first time, as in training for an animal show, requires on-line integration. With routine experience, it too could be stored as the default.
Blending is a general cognitive process that runs across cognitive operations. It is not limited to conceptual structures expressed by two-word juxtapositions any more than it is to conceptual structures expressed in metaphors like the King John example or in desktop interfaces or in non-metaphoric examples like the boat race. In all these cases, where there is conceptual blending, there is pressure to reflect it in formal organization. In the lines from King John, the blend is expressed compendiously in language that directs us to the blended space. This is because it contains forms appropriate for one space ( the imperative, the word thy ), and other forms appropriate for the other space ( storm, weather, sky ). Only in the blend do these forms pick out a coherent conceptual scene. Once the blended conceptual space exists, it can be referred to directly. In the boat race, a single expression, "barely maintaining a lead," expresses an integrated structure in the blend that corresponds to various structures in the input spaces.
This pressure to achieve formal organization to express conceptual blending is general. We see it at work across a range of constructions, from morphemes to sentences. Consider single word integrations like "Chunnel," referring to the tunnel that runs under the English Channel. Clearly, there is a conceptual construction that integrates structure from both the generic frame of a tunnel and the knowledge frame of the body of water between England and France. This integrated unit can serve as the site for integrating a great range of knowledge, from the relevant geology to problems of engineering, from the history of relations between England and France to issues of quarantine, disease, and ecology. This integration of content has as a corresponding grammatical form "the tunnel under the English Channel." This is already a tightly integrated form, relying upon generic conceptions of "under," and the unit status of a grammatical complex noun phrase. English makes available an even more compact compound noun construction, "the Channel tunnel." By fortuitous accident, a further integration of form is possible, given the phonemes in "Channel" and "tunnel." This integration is a formal blend, triggered by a partial phonological mapping between the two words channel and tunnel. Pressure to integrate produces in the case of English "Chunnel"; the corresponding accidents are lacking in French, leaving as the most integrated form "tunnel sous La Manche." This shows another important aspect of integration--that it is opportunistic. That this opportunism depends in any specific case upon apparently peripheral accidents can lead to the mistaken view that the operation is peripheral. Actually, the most central events and structures can arise exactly by opportunistic exploitation of accidents. Evolution teaches us that this is not paradoxical.
Consider an example of integration reflected at the morphological level: Suzanne Kemmer reports the following attested example:
"The Whiners' most common complaint is that they've been relegated to what Mr. Coupland calls 'McJobs.'"
"Mc" evokes a space of fast food and employment in that industry. "Jobs" evokes the more general frame of seeking employment. They have roles in common--workers, employers, wages, benefits, possibilities for advancement, and so on--providing a generic foundation for the blend. Two input spaces are set up, one for aspects of McWorld and another for seeking employment. A straightforward mapping links common roles. But that mapping 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. So this simple blend brings in analogical mapping, frame extraction, and generic role identification. The purpose of the blend is to bring inferences from it to the conception of crucial realities--such as the plight of young people in the modern economy--and to influence central decisions--such as political legislation and the development of policy. Its power and efficiency seem to derive from its conceptual status as an integrated blend with homogeneous internal structure and its corresponding formal integration into a single word. The striking thing about this unit is that it creates new conceptual structure while all the conceptual engineering that went into building it can be retrieved by a member of the relevant linguistic and cultural community from the single word, "McJobs." The general lack of difficulty that we experience in "unpacking" a conceptual blend is remarkable, presenting us with the general problem of explaining why it should be so easy, despite the obvious complexity of the structure and processes involved. This is true not only of "McJobs" but also of all the examples we cite--the boat race, King John, the desktop, "Chunnel," and so on.
In general, the formal unit names two elements in two different spaces, and directs the understander to find the rest. We will call these conceptual elements the named elements. Consider "land yacht" as a reference to a large, luxurious automobile. Clearly, "land" and "yacht" come from different domains: yachts are associated with water as opposed to land. "Land yacht" gives us land from one space and yacht from another, and asks us to perform a mapping between these spaces. In this mapping, yacht corresponds to luxury car, land corresponds to water, driver corresponds to skipper, the road for the car corresponds to the course for the boat, and so on.
Figure 1: Land yacht
Figure 1 shows how the conceptual blend depends on building an analogical mapping, and how the corresponding integrated syntactic form "land yacht" is built with "land" and "yacht" that name elements that are not counterparts in the mapping.
Consider "Language is fossil poetry." "Fossil poetry" works just like "land yacht": fossil comes from the domain of paleontology and poetry from the domain of expression. In the mapping, poetry corresponds to the living organism, while language corresponds to the fossil of that organism. The conceptual elements named in the integrated syntactic form "fossil poetry" are not counterparts in the conceptual mapping.
"Jail bait" refers to an under-age girl whom an of-age man finds sexually attractive. "Jail" comes from the domain of human criminality, while "bait" comes from the domain of fishing or trapping. In the mapping between them, attraction to the girl corresponds to attraction to the bait, initiating sex corresponds to swallowing the bait, and ending up in jail (for sex with a minor) corresponds to ending up caught. The conceptual elements named in the integrated syntactic form "jail bait" are not counterparts in the conceptual mapping.
One common aspect of these examples is that someone attempting to "unpack" the linguistic form does not begin from the assumption that the named elements are conceptual counterparts. When presented with such a linguistic form, we cannot predict on a priori principles the relationship between the named elements. In particular, notice that the generic roles of these elements are different in "land yacht," "fossil poetry," and "jail bait." "Land" is a locative, "fossil" is product of a process, and "jail" is a result. "Yacht" is a means, "poetry" an activity and its product, and "bait" an instrument.
Now consider "boat house." Again the same operations are involved. As in "land yacht," we have a connection between the two different spaces of land and water--houses are associated with land, and boats with water. In the mapping between them this time, the residents of the house correspond to the boats, the house itself corresponds to a protective shelter for storing the boats, leaving the house corresponds to being launched, and so on. "Boat" and "house" name elements that are not counterparts in this mapping.
Of course, there is no restriction that prevents the named elements from being counterparts. Consider "house boat," which again evokes two different spaces of land and water. In the space of land, the resident lives in the house; in the space of water, the sailor is aboard the boat. In the formal integration "boat house," "boat" and "house" are not conceptual counterparts; but in "house boat," the boat and the house are in fact conceptual counterparts, and they map onto a single element in the blend. Similarly, "jail house" evokes a domain of domestic residence and a domain of criminal punishment. In the mapping between them, the jail and the house are conceptual counterparts, and they map onto a single element in the blend.
In all of these cases, including those in which the syntactic form names elements that are blended conceptually--"house boat" and "jail house"--the blend is both less and more than the composition of the input spaces. In "land yacht," we ignore that yachts have a full crew and require no manufactured course, and so on. On the other hand, the blend contains more than the inputs: for example, the inputs may supply the knowledge that we are dealing with a vehicle, but not that it is a car as opposed to something else, or that many specific features that we link with luxury cars belong to the land yacht: passenger-side air bags, electric windows, leather upholstery, opera windows, and suspension built for comfort rather than speed.
In "fossil poetry," we ignore that fossils and their original organisms never exist simultaneously while language and poetry certainly do, or generally that poetry is not physical or biological. On the other hand, the conception of language as a derivative of poetry is the central inference of the blend but absent from the input space. In "jail bait," we ignore that someone intends to lure the fish while perhaps no one intends to lure the man. We ignore that the man is neither a fish nor (in the case where he merely admires and refrains) a criminal. In the blend, we make use of the particular social frame according to which the world is full of pitfalls and traps for the man, teasing him with what it forbids. In the space with the fish, fish are not capable of such a perspective, while in the space of criminal action, the world does not necessarily tempt people to commit crimes.
The situation is no different when the named elements happen to be conceptual counterparts. In "house boat," we ignore that houses have yards and are stationary or that boats are designed principally for travel. We also know many things from background knowledge about house boats that are not derivable from the inputs. We know that a house boat cannot be simply a regular boat put upon on land that happens to have people living in it, or a regular boat at mooring that someone has been living in for twenty years.
We see in these examples the falsity of the the general view that conceptual structure is "encoded" by the speaker into a linguistic structure, and the linguistic structure is "decoded" by the hearer back into a conceptual structure. An expression provides only sparse and efficient prompts for constructing a conceptual structure. The visible formal organization of these prompts is not a direct reflection of the conceptual structure that they prompt us to build or of the conceptual structure that called that formal organization into being.
The problem then is to find the relations between, on the one hand, formally integrated linguistic structure and, on the other hand, conceptually integrated structures built by the speaker or retrieved by the hearer. In general, we will find that the conceptual integration is detailed and intricate, while the formal integration gives only the briefest indication of a starting point from which the hearer is to depart in constructing this conceptual integration.
Consider "X is the Y of Z" constructions like "Vanity is the quicksand of reason." The conceptual mapping of counterparts in this case is elaborate and open-ended: reason corresponds to traveling animals, vanity to quicksand, mental activity to motion over a surface, mental focus to visual focus, and so on through a great list. The formal xyz construction gives only a starting point for constructing the conceptual integration. The formal xyz construction has the simple syntactic form NP be NP of NP. Its core semantic information is only that X and Z are to be gathered into a single space, and that there exists some entirely unspecified cross-domain mapping in which Y is the counterpart of X, and the X-Z relation is to be understood by projection from a Y-W relation in the y-domain for some unnamed W, where W is some element in a domain to which Y can belong. All the rest--what the domains are, their internal organization, the other unmentioned counterparts, the nature of the relevant relations, and so on--have to be constructed with no further formal prompting.
The point is that the formal integration gives only a small combination of elements from the two input spaces that are to be integrated: "Vanity" and "reason" come from one space, "quicksand" from the other. The understander performs the conceptual integration from this minimal prompt. We have seen the identical phenomenon in all of our two-word examples, like "jail bait" and "dolphin-safe," which at the formal level give us something minimal from each of the input spaces, and which leaves the conceptual integration to be worked out. Usually, the formally mentioned elements are not counterparts, as in in "boat house" or "door knob," but they can be, as in "house boat."
The case is similar for "the quicksand of reason," which is an NP of NP construction. "Quicksand" and "reason" point to elements in different spaces which are not counterparts. "Father of Sally" or "President of Mexico" or "distillate of barley" operate in the same way. In "father of Sally," the two relevant conceptual spaces are, on the one hand, a generic kinship frame that includes family relations, and, on the other hand, a specific space that includes a particular individual, Sally. We understand the formal expression as asking us to put the integrated conceptual packet father-daughter in the first frame into correspondence with X-Sally in the second space, where X is some unmentioned particular individual. We are to blend father-daughter and X-Sally. The formal level does not express this conceptual complexity or structure: it names only one element from each space, and these named elements are not conceptual counterparts.
Alternatively, just as "house boat" can name counterparts, NP of NP can name counterparts: as Christine Brooke-Rose surveys in great detail, NP of NP can name metaphoric counterparts, such as "fire of love." (Brooke-Rose, 1958) Charles Fillmore gives the example "One needn't throw out the baby of personal morality with the bathwater of traditional religion." These counterparts need not be metaphoric: "the nation of England," "the island of Kopipi," "the stigma of cowardice," "the feature of decompositionality," "the condition of despair."
We have found that single words ("McJobs," "Chunnel"), two words ("red pencil," "jail bait"), and two nouns connected by "of" ("father of Sally," "quicksand of reason," "fire of love") can all prompt the construction and linkage of two spaces and their blending into a third. This blending involves integration of both conceptual counterparts in two spaces and of events in one space. At the formal level in all these examples, only two elements are named. However, the two elements--one from each space--may or may not be counterparts.
In the boat race, a set of generic correspondences can be drawn between the 1853 run and the 1993 run. The space of 1853 and the space of 1993 share a common generic schema--sailing from San Francisco to Boston. The mapping is structural but not analogical or metaphoric--the spaces are instances of the same conceptual supercategory. "McJobs" is similar in that the connections depend upon existing categorization--jobs at McDonald's and jobs in general belong to the same conceptual domain. In addition, one of the spaces--jobs at McDonald's--is a subcategory of the other space--jobs in general--but a subcategory that brings in its own special scenarios and social considerations.
In the King John example, events of weather can be mapped onto events of communication. This is a creative local analogy that depends upon the conventional metaphoric mapping connecting ideas to objects and expressions to conduits. In the desktop, the mapping is an analogy, in this case a consciously elaborated analogy, such that the space to be understood--computer manipulation--carries explicit indications of the analogy. "Jail bait" and "fossil poetry" are also analogical, setting up counterparts in domains that do not belong to the same supercategory. "Land yacht" is analogical, but the two domains (travel on land, travel on water) belong to the supercategory of travel.
Integration of events. A fundamental motivating factor of blending is the integration of several events into a single unit. For example, although the boat race depends upon extensive connection of counterparts across different spaces, it also has integration of events: the sailing from one space is integrated with the sailing from the other space into a single event of racing, and this is the central point of the blend. In the desktop case, an action performed by the user of the computer is a single event that conceptually integrates the computer command and the manipulation of office items. It thus integrates both event components and conceptual counterparts.
Even metaphoric mappings that ostensibly look most as if they depend entirely on the construction of metaphoric counterparts can have integration of events as a principal motivation and product. "He digested the book" of course has metaphoric counterparts, such as food and book, but it also projects an integration of events. In the source, digesting already constitutes an integration of a number of different events. But its counterpart in the target is, independent of the metaphor, a series of discrete events--taking up the book, reading it, parsing its individual sentences, finishing it, thinking about it, understanding it as a whole, and so on. The integrity in the source is projected to the blend so that this array of events in the target acquires a conceptual integration of its events into a unit. On one hand, the metaphor blends conceptual counterparts in the two spaces--eating and reading. On the other hand, the metaphor helps us to integrate some distinct event sequences in the space of reading. The blend exploits the integrity of events already present in the space of eating, and exports that integrity of events to the target space of reading. In the "digesting" metaphor, we export the integrity in the blend to induce an integrity of events in the target (picking up book, reading lines, finishing book, thinking about it, etc.). In the boat race, we export the integrity of events in the blend to induce an integrity of events in 1993 (preparing the boat, raising money, waving goodbye to well-wishers at the dock, trimming the sails, keeping the log, arriving at Boston, parting afterward, etc. etc. etc.) In both cases, there is a great range of events in one space (reading, 1993) that comes to acquire the integrity of an event structure in the blend (digesting, race).
In summary, we often have a cross-domain mapping that connects elements, which are counterparts. Sometimes, there will be an integrity of events that either arises fresh in the blend or is projected to the blend from one of the input spaces, and this integrity in the blend can be projected to the other input space to integrate an array of events there.
Now we observe that there exists a process of formal blending that has parallel properties. Formal blending can occur independently of whether there is any background conceptual blending. Consider an ad on the back of a bus for the Del Mar racetrack, whose post time is 2 pm. The ad reads: "Hunch hour. 2pm." To get the punning effect of the formal expression, we must access both "hunch" and "lunch hour" simultaneously. This means doing pattern completion from "hunch hour" to "lunch hour." It means partial mapping of "hunch" to "lunch hour": "hunch" and "lunch" are noun phrases that differ only in their initial phoneme. It means projecting the "unch" of "hunch" and the "unch" of "lunch" both onto the "unch" of "hunch hour"; it means projecting the nominal compound structure (N1 N2) of "lunch hour" onto the corresponding nominal compound structure of "hunch hour." It means projecting the common noun status of the counterparts ("hunch" and "lunch") onto the noun status of "hunch" in the blend.
In some cases there is a formal blend but no conceptual blend. Suppose one speaker in a group wonders aloud whether the new mall down the road is still open at 9pm, and another speaker responds, "Well, they should be, since everyone knows about Amahl and the Night Visitors." Here, "Amahl" in the title of the opera is blended with the noun phrase "a mall," but there is no conceptual blend. Or consider a caption in Latitude 38 above pictures of the 1994 Vallejo boat race, which had two legs, with the upwind leg particularly sunny: "Vallejo 94 - Two legs, sunny side up."[3] This caption requires formal blending, partial projection, mapping between forms, pattern completion, and so on, but no conceptual blending: the race is not blended with a particular breakfast dish.
There are cases, potentially misleading, where the formal blend parallels the conceptual blend very closely. Page one of The Atlanta Constitution of 17 February 1994 carried a header reading, "Out on a Limbaugh," followed by a summary of the story on the inside pages: "Critics put the squeeze on Florida's citrus industry for its $1 million deal with broadcaster Rush Limbaugh." To get the punning effect of "Out on a Limbaugh" requires accessing "out on a limb" and "Limbaugh" simultaneously. Behind this formal blend is a conceptual blend with two input spaces, one with an agent who climbs out on a limb of a tree, another with the deal between the Florida citrus industry and conservative radio talk show host Rush Limbaugh. Just as limb and Limbaugh are blended conceptually, so "limb" and "Limbaugh" are blended formally. Conceptual counterparts that are conceptually blended have formal expressions that are formally blended. Even more, the formally blended element in fact refers to the conceptually blended element. The formal blend, in prototypical fashion for a blend, contains formal structure that is not calculable from the formal inputs: "out on a limb" has an indefinite article with a common noun; "Limbaugh" is a proper surname; although a proper surname in English can become a common noun indicating one of a group of people with that surname ("She's a Kennedy," "She's the poorest Kennedy") or one of a group of people analogically equivalent to a particular person of that surname ("He's an Einstein"), this is not what is going on in this example. In the example, "Limbaugh" has not become a common noun, referring to namesakes or analogs of Limbaugh. But it follows an indefinite article; this is a property of its counterpart "limb" in the other input to the blend. As a result the formal blend has new, previously unavailable syntactic structure - indefinite article + proper name.
The Limbaugh example might be taken as suggesting that formal blending and conceptual blending are always in close parallel, but this mirroring seems in fact to be quite rare. We have already seen cases of formal blending that have no corresponding blend at the conceptual level. It is equally easy to find conceptual blending with no corresponding formal blend. A basic metaphor always has a cross-domain conceptual mapping and a potential blend but provides many formal expressions whose vocabulary comes entirely from the source input space and that require no formal blending: "Look before you leap," said of the possible signing of a business contract, for example. There are even cases where conceptual blending and formal blending are both at work, but the formal blend runs contrary to the conceptual blend. For example, contestants on the BBC game show "My Word" were challenged to come up with an intelligible expression identical to the title of any popular show song except for a single letter in the last word. A typical response was "When you wore a pink carnation / and I wore a big red nose." The interesting response for our purposes was "Why can't a woman / be more like a mat?" (The original, from My Fair Lady, is "Why can't a woman / be more like a man?") At the formal level, "man" and "mat" are blended, but at the conceptual level, man is to be blended not with mat but with its opposing element: person who walks on the mat.
It is important, for the sake of elucidating the nature of blending, to account for the strong but mistaken intuition that blending should work in parallel at the conceptual and formal levels. Prototypical blending, as we have been using the concept, refers to a cognitive operation with many features and motivations. Blending involves partial projection from two input structures to create a third structure; often, a single element in the blend corresponds to elements in each of the inputs; the blend is not constructed by union or intersection of the inputs; the blend is not a skeletal or fixed mock-up of a few elements from the inputs, but has a life of its own, in the sense that it contains structure that is not calculable from the inputs and that it can be developed, once it is set up, on its own terms; the blend counts as a unit that can be manipulated efficiently as a unit; the blend provides full access to the input structures, without requiring continual recourse to them, and so on.
Formal organization has many mechanisms, and they often share features or motivations with blending. For example, combining a noun taken from one input space with a noun taken from the other input space to produce a nominal compound that counts as a noun phrase ("land yacht" or "door knob") does indeed produce a formally integrated unit that can then be manipulated as a unit and that permits us to exploit the opportunities for noun phrases that we get free in the language. This unit has integrity and efficiency; it subsumes otherwise unorganized structure; it provides a focus that is related to a great network of related formal structures. The unit is rich enough to work for us but not so rich as to be unmanageable; it is sufficiently compact to be efficient but not so compact as to be vanishingly thin. In Fauconnier and Turner (1995), we study more elaborate grammatical constructions of causation and caused motion, showing how they too depend centrally on blending.
In this paper, we have provided further evidence for the cognitive role of blended spaces in the operation of metaphor--as in the King John example--and in the invention of new actions and categories--as in the case of the desktop metaphor for the computer.
We have shown that formal expression in language is a way of prompting hearers and readers to assemble and develop the appropriate conceptual constructions, including blends. There is no encoding of concepts into words, or decoding of words into concepts. Meaning is not compositional in the usual sense.
We considered expressions that consist of names of elements from different input spaces. The named elements allow the retrieval of a partial mapping, and the construction of a blend based on that mapping. In a majority of cases, the named elements are not counterparts in the mapping.
Conceptual blending involves cross-space mapping of counterparts, and integration of events. Linguistic forms can also be blended, on the basis of cross-form mappings. However, typically, conceptual blends are not expressed by formal blends, nor do formal blends have to express conceptual blends.
Brooke-Rose, C. (1958) A Grammar of Metaphor. London: Secker & Warburg.
Fauconnier, G. (1994) Mental Spaces. New York: Cambridge University Press. [Originally published (1985) Cambridge: MIT Press.]
Fauconnier, G. & Turner, M. (1994) Conceptual projection and middle spaces. UCSD: Department of Cognitive Science Technical Report 9401. [A Microsoft Word for the Macintosh version is available over world wide web from infopath.ucsd.edu.] Compressed (UNIX) postscript version: 9401.ps.Z
Fauconnier, G. & Turner, M. (1995). Blending as a central process of grammar. Paper presented at the Conference on Conceptual Structure, Discourse, and Language, San Diego, November 1994. To appear in Conceptual Structure, Discourse, and Language, Ed. Adele Goldberg. Stanford: Center for the Study of Language and Information [distributed by Cambridge University Press].
Johnson, M. (1987) The Body in the Mind. Chicago: University of Chicago Press.
Lakoff, G. & Johnson, M. (1980) Metaphors We Live By. Chicago: University of Chicago Press.
Lakoff, G. & Turner, M. (1989) More than Cool Reason: A Field Guide to Poetic Metaphor. Chicago: University of Chicago Press.
Langacker, Ronald W. (1987) Foundations of Cognitive Grammar. Volume 1. Stanford: Stanford University Press.
Sweetser, E. (1990) From Etymology to Pragmatics: Metaphorical and Cultural Aspects of Semantic Structure. Cambridge: Cambridge University Press.
Travis, C. (1981) The True and the False: The Domain of the Pragmatic. Amsterdam: John Benjamins.
Turner, M. & Fauconnier, G. (in preparation). Blending and Metaphor. Solicited but not yet formally accepted for Cognitive Aspects of Metaphor. Ed., Shen, Y. & Kasher, A. New York: Routledge.
Turner, M. (1987) Death is the Mother of Beauty: Mind, Metaphor, Criticism Chicago: University of Chicago Press.
Turner, M. (1996) The Literary Mind. New York: Oxford University Press.
Turner, M. (1991) Reading Minds: The Study of English in the Age of Cognitive Science. Princeton: Princeton University Press.