Scarica Sintesi dettagliata del libro Figurative Language e più Sintesi del corso in PDF di Lingua Inglese solo su Docsity!
FIGURATIVE LANGUAGE – BARBARA DANCYGIER e EVE SWEETSER
The basics of metaphor
The concept of a mapping
Thinking about figurative expressions requires that we develop hypothesis about how words can provide access to concepts which are not literally associated with them. Such vocabulary choices are not random but are often very systematic. A person likely to talk about her views being under attack is also likely to talk about defending her position, using strategies or weapons , planning campaigns, or even being a casualty in some ideological conflict. This entire range of vocabulary items can be used to talk about situations where the conflict is at the level of ideas. This kind of usage has been described by referring to two conceptual domains, and postulating a mapping between them, such that some of the conceptual structure of one is projected into the other. A conceptual metaphor is a unidirectional mapping projecting conceptual material from one structured domain, called the source domain, to another one, called the target domain. When we say that the mapping is unidirectional, we mean that the construal is asymmetric. Metaphoric mapping: a unidirectional relation between two conceptual domains (the source domain and the target domain ) which sets up links (mappings) between specific elements of the domains’ structures. A conceptual connection of this kind may be further reflected in metaphoric expressions, linguistic usages of source-domain forms to refer to corresponding aspects of the target domain. Mappings are typically represented by analysts using the sequence TARGET IS SOURCE written in small caps, while the names of domains are otherwise conventionally capitalized. The mapping discussed above was originally described in Lakoff and Johnson as ARGUMENT IS WAR. Some analysts find it more appropriate to label the mapping ARGUMENT IS COMBAT. There are ready mappings available between the oppositional stances of the two participants in the two scenarios. But these mappings are not alone in licensing the construal whereby the “weapons” in an argument are ideas expressed by the opponents. There is a much more general mapping which motivates this specific one. The IDEAS ARE OBJECTS metaphor gives object status to an abstraction and has therefore been called an ontological metaphor because it so radically reframes the ontological status of the abstraction. It is rare for a complex domain to be understood metaphorically via only one set of mappings. Since a debate or an argument is a communicative event, it is also partially structured by other mappings, specifically those that allow us to construe the concept of communication. One of the mappings that was discussed in detail even before the Theory of Conceptual Metaphor had been formulated was the COMMUNICATION IS EXCHANGE OF OBJECTS metaphor, originally known as the Conduit Metaphor. This metaphor represents acts of communication as acts of physical objects to each other, taking turns. The objects exchanged are linguistic meanings, while the containers in which they are exchanged are the linguistic forms; the “containers” filled with meanings are sent to the addressee, so he can unpack the meanings at the other end of the communicative channel. The Conduit Metaphor explains the cohesiveness across expressions such as I can’t put it into words! And Did you get what I said? – not to mention graders’ comments such as You’re packing too much into this paragraph, This sentence is empty , or I didn’t get anything out of this section. The metaphors discussed here share structure. The Conduit Metaphor includes an understanding of the relation between meaning and form, while the war/combat metaphor adds the adversarial nature of the exchange. But it is important to note that there are underlying conceptual patterns which can reappear in the context of different mappings; that is, mappings are not necessarily fully
independent of each other. Conceptual structure can be more or less detailed or specified. There is a very schematic conceptual structure that is shared between the ARGUMENT IS WAR and the COMMUNICATION IS OBJECT EXCHANGE metaphors: the schematically shared target-domain structure is that of communication, and the schematically shared source-domain structure is that of objects being moved from one participant to another. The choice of which label to use to represent a metaphoric domain thus needs to take many aspects of the mapping into account. On the target-domain side, one could say Argument, Dispute or Debate. And on the source-domain side, the choice of War or Combat discussed above reflects certain assumptions about the types of situations and roles available for projection. Metaphor analysts are making somewhat arbitrary choices between labels, but the issue is in fact different. Domains are labelled with available expressions, representing specific construals. The choice of construal licenses a range of linguistic expressions, but the structure of conceptual frames and categories is complex, and there is much potential overlap between them. The fact that categories often have not just a boundary, but a core and a periphery already means that a given category label may apply better or worse. The choice of the domain label is also a choice about the level of schematicity.
Frames and domains
The concept of a domain is basic to the formulation of metaphor as a unidirectional mapping. A domain is a chunk of conceptual matter which either contains structure to be projected into another domain or receives such a projection. Determining the content or limits of a domain without ambiguity may present problems: something as broad as Cognition could be thought of as a “domain”, as could something as narrow as Tests, or something intermediate such as Education. A number of theorists have preferred to talk about metaphors as mappings between frames. The term frame was introduced to linguistics by Fillmore to represent a “prefab” chunk of knowledge structure; a lexical frame is a frame paired with a lexical item or lexical items that represent it. The definition of a frame also involves gestalt structure: that is, an expression referring to some aspect of a frame structure gives conceptual access to the entire structure, so that evoking one aspect of a frame provides access to the entire frame, and individual frame components are understood in the context of the entire frame. For example, the word husband cannot be understood other than in the context of the frame of Marriage, which also includes a wife, a legal and spiritual bond between them, other family relationships. Mentioning husband, wife, divorce , or in-laws requires reference to the entire frame: there can be no marriage without spouses, no spouses or in-laws or divorce without marriage. More specific frames such as Wedding are parts of the Marriage frame; specifically, the Wedding frame is understood to constitute the normal initiatory step in the larger Marriage frame. The reader will have noted that names of frames are conventionally capitalized: thus “marriage” in our text refers to the meaning of the world marriage , while “Marriage” refers to the cognitive frame involved in that meaning. Frames are linked chunks of conceptual structure which get evoked together. Lexical frames are of course frames that are evoked often enough to be given names – which means they are general enough or schematic enough to be by language users. Complex frames have roles, and relations between those roles: in the Marriage frame, roles such as Husband or Wife or Mother-in-Law are filled by different individuals. Frames display various levels of specificity, complexity, and cultural embeddedness. The Marriage and Wedding frames are the focus of a cultural reframing concerning the partners eligible to enter into a married state. The dispute involves the relationship of Marriage to other frames. The concept of a frame stresses additionally that the very mention of any aspect of the transaction evokes the transaction frame. The concept of a frame stresses additionally that the very mention of any aspect of the transaction evokes the transaction frame.
which instantiate this schematic structure: being in a crib and trying to get out, being outside a room with a closed door and trying to get in or opening a box. We point out here that image schemas are abstractions from specific instances of experience. From these specific instances, they abstract more schematic categories and schemas. Other basic schemas we could list are Path, Force, Counterforce, Balance, Control, Cycle, In/Out, Centre/Periphery, Link. Image schemas constitute the very most schematic level of conceptualization of experience, but they do connect to frames. They are only the bare bones of frames – they are structural commonalities which can be noticed between richer, more-filled-out frame structures. They are abstracted from experience, which is inevitably richer and more detailed; we never experience a generic container, buy only specific ones. The schema for Path is an abstract line which delineates movement from one location to another. But it participates in a wide range of frames. It underlies the frame of a Path(way), understood as a visible trail or pathway in space, marked by a different surface, a lack of objects on it, suitability for human use, directionality, etc. But it also underlies the more complex frame of Travel, where the path may be imaginary (an itinerary drawn on a map) or involve means of moving along a path (means of transportation). In metaphor analyses, the distinction between different levels of specificity in mapped frames is crucial. The mapping we would propose would be COMPETITION IS A RACE, not COMPETITION IS A PATH, because the Path schema is a component of the Race frame, and it is the Race frame as a whole which explains the inferences transferred in this metaphor. Path schema is a necessary component of the Race frame. The issue is thus the appropriate level of schematicity and complexity – the central consideration in metaphor analysis. Lakoff and Johnson’s claim was that these very schematic structures correlate with other aspects of human experience in ways that result in pervasive metaphoric mappings.
Primary Metaphors, conflation, and MIND AS BODY mappings
Grady and Johnson developed the Experiential-Correlation Theory of metaphor into an understanding of two closely related concepts: primary scenes and Primary Metaphors. Primary scenes are early and pervasive correlations – often specifically between physical experiential input and subjective judgment or assessment. Primary Metaphors result from these primary scenes, as small children pull apart the paired experiences of height and power, or height and quantity. There is significant evidence that small children are not thinking in terms of separate domains in these cases. Height and Quantity is a clear example of a cross-domain correlation which is conflated at earlier stages of development. Primary Metaphors: metaphoric mappings emerging directly from correlations in experience, in primary scenes. Primary Metaphors do not rely on frame-rich domains, instead they build on cross-mapping of domains of experience. Another crosslinguistically common metaphor is KNOWLEDGE IS VISION or KNOWING IS SEEING. The correlation is obvious: a great deal of our cognitive assessment of our surroundings is based on visual experiential input. The visual cortex is very large, and thanks to links to other areas of the brain, it provides assessments of likely tactile, motor, and other affordances of objects. An important question is why the direction of these mappings is so frequently from a physical experiential domain as source to a more subjective one as target. One might here appeal to the needs of communication. Domains such as Physical Perception are understood to be relatively intersubjectively accessible, while internal judgments, cognitive states, and emotions are invisible, accessible only to the experiencer if not expressed.
But another point made by Lakoff and Johnson is that, for Primary Metaphors, the direction of cuing parallels that of metaphoric mapping. And this is true even in cases where both of the two domains are physical. Many pervasive metaphors involve understanding a more abstract domain in terms of a more concrete one. Given Grady’s and Johnson’s hypothesis about paired physical and subjective- experiential structures, we would expect Primary Metaphors specifically to have this asymmetry. And we might restate this as understanding a more subjective and less mutually accessible domain in terms of a more intersubjectively accessible domain. Lakoff and Johnson presented a wealth of examples of social and cognitive structure being understood in terms of physical experience; Sweetser gave the family label MIND IS BODY metaphors to some of these. However, it is not the case that MIND IS BODY could ever have been thought of as a single systematic set of mappings. To give an example of the problems this would pose, let’s look at KNOWING IS SEEING and UNDERSTANDING IS GRASPING side-by-side. The frames of Seeing, Grasping, Knowing and Understanding are more complex than skeletal image schemata; rather, they are generic event or situation structures, with familiar roles and relations. In a scenario more complex than the schemas examined by Johnson, a person might well hold an object specifically in order to visually inspect it, and thus to find out more about it; it is even possible that these three domains are partially conflated in some infant experience. But you don’t need a light source to grasp something firmly; nor do you need to use hands to see a well-lit object. So if you grasp something, it doesn’t necessarily follow that you see it, or vice versa. These metaphors are not parts of a single coherent mapping of mind to body; rather, they are two distinct mappings, where the frame of physical vision is mapped onto that of knowledge acquisition, and where the frame of grasping physical objects is mapped onto that of comprehension. This includes mapping of causal and aspectual structure: your causal control over holding onto an object corresponds to your control over a concept you understand, and coming to understand something is catching hold of an object when you didn’t previously have hold of it, while ceasing to understand something is losing hold of an object. Furthermore, even within our visual experience, there are a number of separate structured frames, and not all of these are mapped. It is not all of vision, but the visual frame of a light source shedding light on objects, which is mapped in KNOWLEDGE IS VISION – not, for example, the assessment of ambient light in a space. Emotions are another area where subjective experiences are described in terms of bodily experiences. Anger is perhaps the best-explored example in English. A pervasive metaphor in English is ANGER IS HEAT AND PRESSURE OF A CONTAINED LIQUID. An angry person can be said to be simmering, boiling, or streaming with rage ; and in the end this person might blow her top or explode. It is the case that body heat and blood pressure rise in angry people, which is the basic experiential correlation. Steam-producing levels which could physically harm adjacent people. It is the social expression of anger which is socially harmful to adjacent people. The role of experience may affect the mapping in more partial ways as well. Other metaphors of anger include ANGER/AN ANGRY PERSON IS A WILD BEAST; one can see that the inferences here are very different, since you won’t make a wild beast less dangerous by letting it growl, for example. You have to protect people from it. So the inference here might be that people should not show anger. Note that the pervasive SELF IS A CONTAINER mapping surfaces here too: you would need to keep your anger shut up inside if it were a beast. The Self is also a container of knowledge, of emotions at large and other aspects of cognitive/emotional experience. The example shows that correlations in experience are not found only in Primary Metaphors but may constitute a part of metaphoric mappings in which richer frames are used.
“Two-directional” metaphors
This is not of course to say that individual meaning changes within particular languages can move in the directions seen in these strong historical trends. There are multiple possible metaphoric construals of any important target domain, both within and across languages. Productive metaphor motivates expanded uses of forms; it does not force historical-change processes to spread a new meaning to the level of a broad convention. Older literal meanings very frequently coexist for long periods of time with their metaphoric counterparts. Even in the case of a very basic shared metaphoric mapping between domains, individual lexical items may have their own historical developments. A particular word may lose its source-domain meaning and thereby refer only to one domain: that is, it may lose the metaphoric motivations for its meaning, even while the general metaphor remains productive. Of course, reanalysis of a linguistic form may also cut off its original metaphoric motivation. Lakoff and Turner’s classic example of this is English pedigree , which comes from the French piede-grue “crane’s foot”, an image metaphor for the shape of a genealogical tree. Since English speakers mostly did not know that his French form meant “crane’s foot”, they only conventionalized the target-domain meaning “genealogy” and the metaphor was lost. In other cases, a culture can lose a metaphor which was once active and thus cut off word meanings from their original metaphoric motivations.
Conventional or creative?
Many people associate the term metaphor with a figurative, possibly literary, trope, in opposition to colloquial, literal, everyday use of language. However, the distinction between literal and figurative language is misleading. Like literal language, figurative language is grounded in experience and cognition; the difference is in the structure of the grounding relations. Metaphor involves two conceptual structures and a set of projections between them,, while metonymy, personification, irony, hyperbole, synesthesia, and oxymoron rely on different, but equally conceptually motivated, patterns. It is the presence of this inter-frame mapping pattern that determines whether a linguistic expression is metaphoric or not. The level of conventionality of an expression is a criterion totally independent of the nature of figurative thought and language. Some of the conceptual patterns we call metaphor are fully conventionalized, for example when calling independent-study educational material a study guide. The expression relies on the domain of Travel, where the presence of a guide simplifies the process of reaching the destination; in the domain of Studying, reaching a satisfactory level of knowledge and understanding can be facilitated by appropriate texts, so they can be called guides. But the same pattern can be used in a less predictable context, as when Romeo addresses the vial containing poison as bitter conduct and unsavoury guide. This usage would be recognized as a metaphoric expression. Romeo is committing suicide, and so the poison is metaphorically going to lead him to death, his Purpose and hence his metaphoric Destination. So the metaphor is appropriate, but this is not a typical construal of what poison does. In these two instances, the same pattern explains two expressions that differ widely in their level of conventionality. Idioms are another important area of interest for studying the relationship between conventionality and metaphoric structure. Many idioms are metaphoric. Although idioms may be processed as indivisible wholes, analysable idioms have been shown to evoke both source and target domains in processing. Conventionalized metaphors have often been termed “dead”, presumably because the patterns they represent no longer motivate new expressions. But all metaphor use relies on a certain level of activation in the context of use, and so such metaphors are often better described as “sleeping”, as they can easily awaken and become newly productive. This is true of ordinary conventional metaphors such as KNOWING IS SEEING, bet even in the case of idioms, which are presumed “dead”, we can imagine someone saying something like The cat is now out of the bag. In such
cases, the conceptual pattern is revived to be reactivated with a level of detail not typically present in the expression. This shows that, though dormant, the pattern is still alive.
Experimental support for Cognitive Metaphor Theory
The question of conventionality touches another important theoretical conundrum: for conventional cases where clear polysemy exists between original source- and target-domain senses, what is the cognitive reality of the polysemy relationship? From the beginning, experimental psychologists have been interested in testing the cognitive claims made by Cognitive Metaphor Theory. At the same time, many linguists have found it hard to believe that idiomatic or fixed metaphoric usages were live metaphors and processed as such. There are a number of serious problems with this noncognitive view of conventional metaphors. One is that if the everyday idiomatic metaphors are not “conceptual” but just linguistic, we need some other explanation for how they help us to understand related novel and poetic metaphors. Experimental psychologists and psycholinguists have begun to test the cognitive reality of metaphoric usage. Gibbs was among the first to test processing of metaphoric idioms, and he found quite solid evidence that subjects were processing source-domain images as quell as target- domain meanings. More recently, there have been even stronger las results suggesting that metaphoric source and target domains are cognitively linked to each other. For example, linguists have shown that an experience in the source domain will prime particular choices of mapping to a target domain; in this case, they found that actual physical-motion experiences affect subjects’ choice of time metaphors. Other investigations of the role of spatial cognition have also consistently demonstrated its role in processing abstract concepts. Reading about actual physical journeys with different characteristics affects the time that subjects take to process very conventional metaphoric motion usages. Looking at the cognitive connection between verticality and emotion, other linguists asked subjects to move marbles downwards from a higher rack to a lower one, or alternatively upwards from a lower rack to higher one. Bergen gives an overview of brain-imaging studies which appear to show mixed results. Literal meaning is simulated in the brain. In some contexts, subjects respond in the same way to hearing or reading motion expressions used metaphorically, but in other cases the motor areas suggested by the source-domain language do not “light up” in the brain. Possible explanations for the non- activation include a certain level of “bleaching” of idiomatic expressions and differences in how the stimuli are processed under different conditions. Source-domain simulation effects are more likely to occur in the brain when sentences are presented more incrementally, so the speed with which the linguistic expressions are processed appears to be a factor. But the kinds of simulation involved may be different. If we hear or read about grasping actual physical objects, the brain can simulate the situation in all its rich detail, including the concrete object being grasped. Bergen’s conclusion is that simulations of metaphoric expressions in the brain are weaker – because less richly simulated – than simulations of fully concrete situations. However, it is clear that there is some observable effect in the brain. This experimental work does clearly show that, in the relevant cognitive domains, the metaphoric connection is not just linguistic; it is cognitive. Patterns of mapping exist independently of linguistic evocation: warm or cold drinks, or heavy objects, evoke social like or dislike and cognitive importance, with no linguistic mention of the source domain. Many of the same observations have been made about the processing of linguistic forms: brain areas associated with one domain are activated by language from another.
Metaphoric structure: levels and relations
Inheritance and compositional relations between metaphors
JOURNEY or whether it is an example of PURPOSEFUL ACTION IS MOTION; it is both, since one is a more elaborated instance of the other. Such complex elaboration of Primary Metaphor are common, as are compositional combinations of mappings involving distinct but compatible metaphors. Being physically contained could also mean that the person restricted cannot fully experience the world either in terms of perception or interaction and motion toward goals. A complex expression may activate many mappings at once – and we will not accuse it of being a “mixed metaphor” as long as those mappings are compatible with each other. Rather, it is a compositionally complex metaphor, built up of compatible components. We refer to this as compositional structure. Just as grammatical constructions are composed of multiple mutually compatible structures which are instantiated; similarly, multiple compatible metaphors can be instantiated in more specific examples such as these. Compositionally complex metaphor: a metaphor which is a fully specified instance of more than one higher level metaphoric mapping. This may mean that multiple Primary Metaphors are involved. Additionally, all uses of the Location ESM at least imply the passage of time. And indeed time, as well as events and activities, is conceptualized in terms of motion through space. The primary experiential correlations are clear since all motion actions are activities in both space and time. In many Location ESM examples, time is profiled, as in That’s where we are now and it’s a long way to go , which construes the location both in terms of time and with respect to the current stage in the process and also suggests that the purpose is distant both temporally and in terms of the stages which still need to be accomplished. It is difficult to separate Event Structure both from Temporal Structure and Experiential Structure. The temporal structures on the two sides of a metaphoric mapping need to be compatible: it could not be the case, for instance, that the end of a life could be the start of the journey onto which the life is mapped, or that the start of the life could be the end of the journey. This is part of what has been proposed as the Invariance Principle. Certain structures are “invariant” across metaphoric mappings: in particular, temporal sequence and aspectual stages must be mapped coherently between the inputs in a metaphoric blend. There is another low-level schema that can be exploited as part of our understanding of Event Structure – that of Holding an Object or having it. This schema gives rise to the Object Event Structure Metaphor, in which attributes such as bodily states are described in terms of a participant “having” them. Some of these attributes can be transferred to others, which can be a matter of simple movement or intentional transfer – in the latter case, we can also talk about causation as transfer of possessions. The Object ESM is so named because it treats events and situations as objects rather than locations. In some ways it is the inverse of the Location ESM: the Object ESM treats attributes or situations as mobile objects which can be acquired, while the Location ESM treats them as fixed locations to and from which participants move. The Object ESM is common and is an interesting example of reliance on ontological metaphor, such that attributes are conceptualized as physical objects which can be possessed or given away.
Levels of complexity
The connections between the domain of mental constructs and processes and the domain of human nutrition (Food) exist at many levels of conceptualization, for there are multiple frames and subframes involved. The Food domain, being the source, brings the complexity of its frames into our understanding of mental life. The subframes range from those related to the food itself through the role of Nutrition in health and Well-Being, to the viewpointed processes of Absorbing and Digesting Food. The complexity of the subframes is also the basis on which this broad family of
mappings builds up a complex construal of mental processes. Some of the subframes can be organized hierarchically, in the sense that most of them inherit a schematic ontological metaphor IDEA ARE OBJECTS mapping, in the specific subinstance of IDEAS ARE EDIBLE OBJECTS. Edible objects are typically evaluated in terms of Nutritional Value, Attractiveness, and Stage of Preparation, and each of these subframes yields interesting expressions in the domain of Mental Activity. This broad range of metaphors depends both on the basic mapping allowing us to think of ideas as edible objects and also on another basic family of mappings that allows us to think about the functioning of the mind in terms of the functioning of the body and to present mental processes as bodily processes. MIND IS BODY is not a single metaphor, but a name for a whole range of metaphors which conceptualize humans’ mental and emotional selves in terms of their physical selves acting in the physical world. In this case, we are mapping a range of bodily functions, such as requiring nutrition and eating, onto a range of mental functions such as considering, believing, remembering, and rejecting ideas. What the above examples make clear is that the complexity of the domains involved in metaphor and the complexity of the resulting range of metaphoric usages depend in a direct way on the complexity of the frames involved, which is in turn proportional to the frames’ importance in human experience and culture. The complexity may be an effect of the range of subframes, and the result may be quite a varied array of meaning. Multiple subframes can be metaphorically exploited at the same time, and the ones which are more salient culturally depend in many ways on the more directly embodied ones. Any important area of abstract experience can be thought about in many different metaphoric ways. Ideas are not just edible objects, but malleable objects, objects that are building blocks for more complex structures, and so on. Any metaphor we consider is situated in a complex network of other mappings – mappings from the source domain to other target domains, more or less schematic mappings which underlie or extend the mapping we are considering, and mappings from other source domains onto that target domain. These connections between mappings are crucial to the role of metaphors in conceptualization. We should expect links and relations among metaphors to be the core feature of metaphoric thought, directly related to the conceptual power of metaphoric mappings. The connection across mappings can emerge on various levels. They can be based on the more schematic levels of embodiment and ontology, so that, for example, ideas can be viewed as either edible objects or as malleable objects; however, in both cases, the mind also has to be viewed as a body and thus capable of interacting with whichever types of objects in appropriate manners. At the other end of the spectrum, the rich cultural content of metaphoric domains may yield very specific mappings. However, the richness of the concepts of a feast relies on both more schematic and more specific levels of framing and conceptualization: the nature of feast or the variety involved in a feast are specific, but they rely on the basic understandings of food and eating, of material objects and the functions of the body. To sum up, each of these metaphors participates in a hierarchy of mappings, and their power is based in the human ability to connect the body and its material interactions to complex cultural scenarios. Metaphors are thus connected, not only vertically, but also horizontally (with many metaphors relying on treating mental constructs as different types of material objects and treating the mind as a body in different ways). Relationships among mappings affect metaphoric construal of specific domains in more than one way. Croft observer that there are metaphoric uses of food vocabulary which cannot be explained through a single unidirectional projection from the source to the target. As he argues, such examples display different argument structure because they rely on the framing from both the source and the target. This connects the IDEAS ARE FOOD metaphors to other frames in which are preposition into profiles the flow of material into a container. Croft’s conclusion is that metaphoric mappings, while asymmetric, should include the contribution of both the assumed source and the assumed target.
experiential correlations between physical experiences and attendant subjective evaluations such as judgements about ongoing existence, attributions of causal dependence, and so on. Even though it is not itself a Primary Metaphor, it is deeply motivated by Primary Metaphors – whereas there would be no motivation for the reverse mapping BUILDINGS ARE THEORIES, because there would not be Primary Metaphors on which to base that more specific one. THEORIES ARE BUILDINGS is also an example of the relationship between more skeletal, general metaphors and their more specified subcases. Primary Metaphors are by nature extremely general and motivate potentially huge numbers of subcases; the Location ESM permeates all of our understandings of human actions and states. Also, in THEORIES ARE BUILDINGS as in other metaphors, inferential structure is transferred. That is inevitable in Primary Metaphors, which are about “cuing” inferences in one domain from data in another. So also in these submappings, we know what it would mean to have your theory come down around you , to have to rebuild from the foundation , or to have even the foundations undermined. In the case of undermined foundations, more rebuilding is needed than in the case where a building’s foundation still survive; and we transfer that inference to an assessment of the extent to which theorizing needs to be redone. Causal and aspectual structure is also preserved across mappings. If you are building a building , it isn’t yet complete, and if you’re building a theory , it is still being developed. Causing the physical structure to be stronger or higher or broader corresponds to making the theory better reasoned or more ambitious or more extensive in coverage; damaging the physical structure corresponds to making the theory less convincing or less likely to continue to be believed. These mappings could not go in the opposite direction: physical damage to the building could not correspond to improvement of the theory. In the case of metaphors with primary experiential bases, this kind of causal and aspectual coherence is ensured by the correlation between the input domains. For MORE IS UP, the gradually increasing height of liquid in a container just does correlate with increasing quantity, not with decreasing quantity; the scale of heights corresponds to the scale of quantities, and adding more liquid causes the level to go up, while removing liquid causes it to go down. The power of the THEORIES ARE BUILDINGS metaphors is shown in scientific discourse. The Metaphor Analysis team led by Lakoff has coined the term cascades to talk about the kind of inheritance relationships that link levels of metaphoric structure. The primary characteristics of cascade relationships between metaphors are that higher, more schematic structure is fully inherited by lower more elaborated subcases and one specific metaphor can inherit fully the structures of multiple higher-level metaphors. In talking about the THEORIES ARE BUILDINGS metaphors, we have no need to separate them from the PERSISTENCE IS REMAINING ERECT or the DEPENDENCE IS PHYSICAL SUPPORT metaphors. Those cascade naturally downwards to participate in the more specific metaphors and are included in their structure. Some details of building structure don’t seem to be incorporated: they are not present to be inherited from the Primary Metaphor structures. One very interesting fall-out of this kind of approach to levels of metaphor is that we can see why it is so rare to encounter a truly novel metaphor. What we encounter instead are either elaborations and/or compositional combinations of known metaphors. The classic example the fast lane on the freeway of love is an elaboration of LOVE IS A SHARED JOURNEY. Even these subcases still have their own inferences, added to the inferences available from the more general mappings: the risk and emotional excitement which go with very fast speeds emerge from this specific case, and suggest that the love relationship is risky an exciting. More interesting are compositional cases such as glass ceiling. It refers to the situation where women and minority employees somehow never get promoted above a certain level - but where the reason is not obvious, since the companies hire women and minorities at lower levels and appear to be open in their employment and promotion policies. To understand this metaphor, we must combine A CAREER IS A JOURNEY with AUTHORITY IS UP, a combination that yields the
career ladder and climbing the corporate ladder, where PROMOTION IS UPWARDS MOTION. The ceiling is a barrier specifically to upwards motion, and DIFFICULTIES ARE OBSTACLES. But we need another piece too. The reason the ceiling is “glass” is that you don’t know it is there: the social exclusion is unnoticeable , like a physical barrier that is transparent and thus invisible. And this requires KNOWING IS SEEING. Innovative metaphoric usages are usually either elaborations of conventional higher-level metaphors, or compositional combinations of conventional higher-level metaphors. This makes them easy to learn, since they don’t need really new conceptual foundations. Further, since many of the Primary Metaphors are based in primary scenes, the higher up you go in the metaphor hierarchy, the more likely you are to find crosslinguistically shared structures.
Image metaphors
We have not yet talked about image metaphors , the kind of metaphors that involve mapping specifically of images from one domain onto another. Image metaphors appear to be differently motivated, and subject to different constraints, from conceptual metaphors. Human perceptual structure is constantly mapping inputs onto each other, without necessarily involving inferential structure or broader categorial generalizations. This metaphor doesn’t seem to have significant structure mapped beyond the image similarity. Nor does there seem to be any broader network of metaphoric mappings into which this one fits or from which it inherits structure, as LIFE IS A JOURNEY inherits from the Location ESM. Not a great deal is known about image metaphors as a class, but they are structurally distinct from conceptual metaphors. Another way in which image metaphors are different from conceptual metaphors is that they don’t tend to have a basis in experiential correlation. Primary Metaphors need primary scenes to motivate them. And that motivation is passed on downwards to the complex metaphors which inherit Primary Metaphor structures. There is no experiential correlation at the specific level between discriminatory promotion practices and transparent glass ceilings in buildings. But there are primary experiential correlations between vision and knowledge, between verticality and authority, and between motion/barriers and purposive action/difficulties. These primary mappings cascade downhill to motivate the specific case of glass ceiling , even though it has no direct correlational motivation of its own. Image metaphors don’t seem to work this way. There not only is no experiential correlation between hourglasses and waists, or sweet flavours and harmonious sounds, there doesn’t need to be. The brain’s structure connects these shapes and these modalities. And this brings us to synesthesia. In language, it is a frequent phenomenon for vocabulary from one perceptual domain to be used to describe phenomena in another perceptual domain, as in sweet sound. The connections across domains of perception are difficult to classify in terms of concrete or abstract meanings, but what such uses share with metaphor is the pattern whereby a domain basic to human experience yields a conceptualization applied to another such domain – although the target domain in this case is also an area of basic experience. Such mappings have been talked about as synesthesia , clearly in reference to a clinically recognized condition in which experiences in one area are received in another as well. Symptoms vary, but various forms of the condition link colour and numbers, colour and letters, touch and hearing; that is, subjects physically perceive, for example, the colour blue when processing a particular numeric value. There is a significant difference between having synesthesia in this dramatic way, and this kinds of associations which result in synesthetic linguistic usages. But all humans seem to have some crossmodal associations, and most languages seem to lexicalize at least some such associations.
Objectifications, in comparison, are most common in description of how we manipulate abstractions. In a sense, it is the flip side of personification since, instead of providing a construal of an effect something has on us, they offer a construal of our interaction with something. Adjectives used to describe these metaphoric objects indicate the properties of the abstract entities, in terms of the physical properties of physical entities. From the perspective of linguistic choices, objectification and personification play important roles in more complex construals of causation and mental activity and are frequently components of other metaphors.
Making the abstract concrete
Metaphor is a thinking of the abstract in terms of the more concrete. For Primary Metaphors, the asymmetry between source and target domains is not really just one of concreteness but is based in the cognitive asymmetry between the two domains. Vertical Height is not more concrete than Quantity, but it is more assessable than Quantity and serves as a cue for assessing Quantity, rather than the other way around. Of course, not all conceptual metaphors are primary: many of them are more complex and can be described as subcases or combinations of Primary Metaphors. However, they will inherit cognitive asymmetries, including target-source differences in degree of concreteness, from their components. Grady and Johnson both refer to the basis of Primary Metaphors as being a correlation between a more directly sensory domain and a more subjective-assessment domain, where often the more concrete sensory domain ends up being the source domain is the target. The concrete/abstract contrast in metaphor is really about more and less intersubjectively accessible domains. Not only do we experience height physically, but because we do so, we expect others to assess height using the same input we use, and to predictably share our assessment of relative heights, while assessments of quantity or volume are both more difficult and less likely to be shared. We have given examples where two concrete domains (Vertical Heights and Quantity or Volume) are linked by experiential correlation and become source and target domains in a Primary Metaphor: target domain certainly do not have to be abstract. Nor do source domains need to be concrete. Conceptual metaphors can rely on source domains which are not construable as concrete. So even for metaphors with a clear experiential basis, there are immediate difficulties in maintaining the idea that metaphors are about understanding the abstract in terms of the concrete. Consider ARGUMENT IS WAR – or ARGUMENT IS COMBAT. Of course it is true that War and Combat involve physical force and fighting, while Argument can be entirely a linguistic communicative activity. But this does not fully resolve the problem of abstraction. Certainly an argument is more linguistic and social than a physical fight. And we can even say this is related to a primary mapping: that is, disagreement can lead simultaneously to argumentative communication and to physical struggle, which are correlated in a primary scene. But there are added complexities to consider. Some fights have a great deal of abstract conceptual structure. And both arguments and wars typically involve people who have conflicting desires and beliefs about some situation, so there is real shared structure in the domain of participant intentions. An added crucial point is that not all these metaphors have equal degrees of experiential grounding. The Primary Metaphor is clearly something more like ARGUMENT IS COMBAT, not ARGUMENT IS WAR. So the Primary Metaphor here is a mapping from concrete to abstract.
Metaphor families
Morgan developed the concept of metaphor families to describe cases where a group of domains includes shared frame structure, which permits mapping back and forth among that group of domains. One such family is united by the Competition frame, which is part of the structure of such
varied domains as War, Combat, Argument, Business Competition, Elections, Sports Competitions, and Predation. We might rephrase this metaphor family relationship partly in terms of inheritance since the general Competition frame is inherited by all of the subdomains. But there is something more here. Within a metaphor family, we can see metaphors mapping in a number of directions between domains, with some of those mappings being more productive than others. Several point emerge from Morgan’s analysis. One is that concreteness and abstractness are important and are relative rather than absolute. Indeed, War and Predation may be more concrete even than Team Sports, and all of these seem more concrete than Elections or Business Competition or Arguments. And indeed this is probably part of the reason why we don’t understand Combat in terms of Elections, or as Business Competition, even though the reverse mappings are common. Another emergent point is the importance of the shared Competition frame which enables all these varied mappings; such shared inheritance of frame structure can enable mappings even in the absence of correlational motivation from a primary scene. A third emergent point is that, despite the shared Competition frame, the differences between these domains are just as important in determining mappings are the similarities. Although inheritance relationships are involved both in metaphor families and in metaphor subcases, they are involved in very different ways. The subcase examples involve more general metaphoric mappings being inherited by more specific ones. The frames involved are not necessarily similar: one would not say ideas are “like” objects or that they share basic frame structure with them, nor that one is a subcase of the other. In the case of metaphor families, a set of frames shares generic structure inherited from a more general frame, like Competition. Then, if metaphoric mappings are built up between these somewhat similar frames, it is sometimes possible to build metaphors in many different directions.
Mental spaces and blending
Blending is a type of multidomain mapping. Metaphor is a reconceptualization of one domain (the target) in terms of another (the source). But looking at a depiction of a metaphor as a mapping between two domains, one might wonder where in the representation to find that conceptualization or construal. There’s a place for source-domain structure and frames, and a place for target-domain structure and frames – and a specification of mappings between particular substructures and frames in those two domains. But the metaphoric cognitive structure which emerges from those mappings is different from either the source or the target considered alone. But it doesn’t seem to have a representation of its own. Nor is there any place in a two-domain table or diagram where we are representing the generic structure shared between source and target domain which guides those mappings. Things like aspectual, causal and scalar structure not only are shared by source and target domains but seem to constrain the mappings between the two domains. The Heat/Pressure model of Anger exemplifies all three of these kinds of pairings. We map the scale of increase or decrease in heat and pressure onto a scale of degree of anger. Just as we know that liquid which is steaming or simmering is not on the verge of violently blowing off the top of the container, so we know that a person who’s simmering is not as furious as someone ready to blow her top or explode. Speakers can identify the shared aspectual structure involved; a liquid has to first start being heated by some heat source and has to go through the whole intermediate scale of heats and pressures before it reaches the final result of explosion. This maps onto the idea that an angry person doesn’t typically instantly “blow her top”, but rather, her anger develops over time before that happens. And we could not map if the other way; it would be incoherent to have explode refer to a lower level of overt anger-manifestation than simmer.
specific participants, deictic features, etc. In other words, frames have the potential to be conceptually elaborated and become less skeletal spaces, but they may remain simple.
Spaces and frames: types and relations
It is important to distinguish differing levels of specificity or schematicity. We can outline the major types of input structures and some types of relations between input structures. The simplest type of structure is exemplified by Primary Metaphors, where the primary content of the input frames is image-schematic structure. A mapping like STATES ARE LOCATIONS, or TIME PERIOD ARE BOUNDED REGIONS, does not give a lot of detail to either side of the mapping: States, Time Periods, Bounded Regions, and Locations are all very schematic. Similarly, MORE IS UP, LESS IS DOWN evokes the highly schematic structures of Quantity and Verticality. Of course, Primary Metaphors emerge from an individual’s full sensory experiences of specific inputs such as heights and quantities, or of specific experiences states and locations. But their power like precisely in the fact that they are abstracted away from those specific cases to form the basis for conceptualizing the correlated general parameters. On the other hand, many complex metaphoric mappings are more fleshed out than MORE IS UP. In saying that someone missed the boat or reached a dead-end in her career , we evoke more complex and more specific imagery of boats or dead-end streets, even though the very schematic Location ESM and its subcase LIFE IS A JOURNEY are present as a major component of that more specific complex structure. So the metaphoric expressions miss the boat and reach a dead- end are specific subcases of both the Location ESM and its subcase LIFE IS A JOURNEY. The level of specificity of mappings depends of course on the levels of specificity of the frames mapped. We will use the term superordinate frame or schematic frame to talk about frames which do not have visual and motor imagery. The frames of Sinking and Soaring are specific cases of the superordinate Up and Down schemas. There is no motor imagery, and only minimal schematic visual imagery, in the abstract Up and Down frames, but there is a specific type of motion and much richer visual imagery in, say, the expression Prices sank and Prices soared, which combine the Up/Down schema with CHANGE IS MOTION, with specific manners of motion and moving objects. At the other end of the schematicity spectrum will include many different frames. Some of the them might include general knowledge about collegial relations in academia, friendship, coffee shop, and gossip. Others might include the specific appearances of the two participants, their voices, their likely styles of clothing, and so on. This is at the subordinate level of categorization and it is highly complex and multiframed as well. In an extended discourse, mental spaces form networks of spaces linked in terms of temporal sequences and/or causality chains. Thus temporal and causal relations among spaces constitute another dimension of analysis. This is a property of frames and spaces that holds whether or not linguistic expressions themselves refer to a sequence of events. We discussed above the metaphoric construal of an extreme expression of anger as an explosion and this now makes another important point, namely that frames are dynamic, with aspectual and causal structure. And the spaces that they structure are dynamic too: we can run them and see what the next stage or the result will be. We will need to treat causal and temporal sequences between spaces as a unique kind of relationship. Let us call this the succession relation : if one space is the successor or predecessor of another, they cannot be simultaneous, and they have to be fitted into slots of cause-result or result-cause, and their sequencing related to the broader aspectual structure of the events involved. Another important relation is that of paired networks of spaces, as in cases where we need to construe two causal chains – one real and the other one negated or counterfactual. Another crucial dimensions of spatial categorization is that of conventionality vs. novelty. Frames rely on shared cultural knowledge, though they may be shared by groups of various sizes – from
an entire community to a group as small as a family. They are part of our “prefab” knowledge about the world. We are able of ordering food in a restaurant, as such without conscious work. And frames help us structure our experience. These cognitive structures reside in long-term brain patterns and remain accessible to our abilities or memories without being constantly activated. On the other hand, the specific dynamic situation which we’re imagining when we’re told the story of yesterday’s coffee date is not being pulled from long-term memory. This we might call an online space – to refer to the original formulation in Mental Spaces Theory, where mental spaces were talked about as cognitive structures that emerge “online” as discourse progresses. Metaphoric mappings and other kinds of blending can involve these very different kinds of structures, all of which are frame-based in content. Generic or schematic structures may be mapped onto each other in primary or other higher-level mappings, but they also often guide mappings between more specific structures.
Blending processes and types of blends
The processes which transform two or more independent input spaces into one blended space are critical in producing the effects of blending.
How to build a blend
Let us consider several expressions used to describe the effects of longterm exposure to a certain view of the world and the need to reverse those effects: mental resuscitation, pollution of our minds, infotoxins, mental detox. All of these expressions rely on the same organizing frame of an organism affected by unhealthy influences and in need of healing. Together, they map the input space of a living organism onto the input space of the mind. To understand the structure of the blend, we need to look at various aspects of how it emerges. First of all, it is important to identify how various elements are connected across the inputs: here, we wan to analyse the way various substances which can negatively affect an organism are cross-mapped with the standard vision of the world presented by the media and other representatives of the establishment. The body of an organism can be poisoned by substances which bring about illness; in the same way, in the blend, human minds can be “poisoned” by information provided through official channels. The cross- mappings determine how the organizing frame of a poisoned organism will be used to restructure the understanding of the mind. The inputs cannot be projected into the blend as wholes; rather, the structure needed for the blend is selected and then projected into the blended space. In this case, most of the space topology comes from the organism input rather than from the mind input – this is why we would call the blend metaphoric , since it involves conceptualizing one space in terms of another. But the projection is selective since the blend does not focus on other possible aspects of organisms but primarily on biological health and interaction with toxic substances. The structure projected into the blend forms a coherent whole, the emergent structure. Having put that structure together, we can run the blend – that is, understand the inferences it yields in terms of various aspects of the setup: the kinds of infotoxins our minds are exposed to, how one would clear people’s minds of the effects of the poison, the sources of the “poisoning”. The impact and power of the blend is not in its structure alone, but in the inferences it yields. At the same time, the blend allows us to reason differently about the inputs. In this case, the blend yields a new, emotionally loaded view of the types of information we are exposed to and the effects it has on the clarity of our thoughts. Blends have very general structural properties. Fauconnier and Turner pointed out that mapping (or projection ) is selective. Not all of the biological organism frame was projected onto the information-absorbing mind frame. Aspects which will not map may simply be omitted: for example,
