Preface
Acknowledgments
Pretext: Professing English in the Age of Cognitive Science
1. User's Manual
2. Floor Plan
3. Poetry and Invention
4. The Body of Our Thought
5. The Poetry of Argument
6. Conceptual Connections
7. The Poetry of Connections, I
8. The Poetry of Connections, II
9. The Poetry of Connections, III
10. Cultural Literacy and Poetic Thought
11. Envoi: Cognitive Rhetoric and Literary Criticism
Appendix
Index

Preface

The coming age will be known and remembered, I believe, as the age in which the human mind was discovered. I can think of no equal intellectual achievement. The purpose of this book is to propose a reframing of the study of English so that it comes to be seen as inseparable from the discovery of mind, participating and even leading the way in that discovery, gaining new analytic instruments for its traditional work and developing new concepts of its role.

I speak from a known frame about a new frame, or more accurately about a new activity of framing. I offer a provisional version of that new frame. I conduct illustrative case studies within it. Of course, I am invested in the intricacies of these particular case studies, having lived with them. But my final investment is not in the details of these particular studies or even in their chosen themes and problems; it is rather in the activity of reconsidering and reframing the study of language, literature, and mind.

This book is offered as a means to that activity. I present a series of incitements. I do not map the borders and divisions of future inquiry but ask the reader to begin a revision that will affect the path of our research in ways we cannot know. The crucial result of this book is to be found not within this book but rather within the reader who has been persuaded to begin that reconsideration.

The human mind is linguistic and literary; language and literature are products of the everyday human mind. The future for which this book is written is one in which traditional humanistic studies will be centered once again upon the study of the human mind, and the study of the human mind will be centered upon human acts that are the subject matter of the humanities. To give an adequate suggestion of this future, I am forced to work in greater detail than the current state of research permits. This book is something like a detailed report about a future that cannot be known in such detail.

The constructive project of the book--to shape an inquiry into language, literature, and mind as inseparable objects of analysis--begins in chapter 1, the "User's Manual." Readers who feel by temperament or prior interest attracted to this project but who see nothing at stake in a discussion of the profession of English can begin without loss at chapter 1. It is a guide to the use of this book as an instrument for revisiting and revising basic concepts so automatic and powerful in our thought as to constitute a kind of floor plan of our thinking. I discuss the entrenched nature of these concepts, their ubiquitous unconscious use, and the sources of our resistance to their loss. In "Floor Plan," I examine particular basic concepts that we must change if we are to reconstitute the profession along the lines I propose. I consider concepts of the human person, the body, the brain, thinking, consciousness, concepts, language, literature, and the humanities. My purpose in these chapters is finally not to survey the existing floor plan of our thought but rather to offer a replacement.

These chapters are the opening frame. Collectively, they offer a design for research that might be called "cognitive rhetoric." The second part of the book offers topical introductions to that design through four case studies. Each provides a specific point of access to the kind of research that would be necessary at the project's start. I think of them as early laboratory experiments in the mode of investigation I propose.

The third part of the book--titled "The Poetry of Connections"--presents an overview of larger issues in cognitive rhetoric. I close the book with a practical application and a brief envoi. The practical application brings the instruments of cognitive rhetoric to bear upon questions of cultural literacy, to indicate how cognitive rhetoric might serve as a basis for our pedagogy. The envoi sketches possible lines of research and instruction under a new frame of inquiry whose purpose would be to rejoin the study of language with the study of literature, in concert with the study of the mind and the brain.

What follows is therefore not a humble book, but as it is my attempt to reconstitute the profession, I hope the arrogance needed to write it at all will be excused. It is a proposal to change our conception of the humanities--or at least that branch of the humanities that concerns language and literature--by grounding it in the study of human cognition. It is my attempt to do what I can to restore the profession of English to its once-central position. I hope especially that my colleagues in the profession of English will take it as the offering of someone dedicated to our profession and to its future at the center of humanistic and scientific activity.

Language is the Mirror of the Mind

--ask yourself whether our language is complete; whether it was so before the symbolism of chemistry and the notation of the infinitesimal calculus were incorporated into it; for these are, so to speak, the suburbs of our language. (Wittgenstein, Philosophical Investigations)

--Sex is the poor man's opera

--Because of its many canals, Venice is sometimes called the Venice of Italy. (James McCawley)

"Money is the root of all evil," "Brevity is the soul of wit," "The wages of sin is death," "Politics is the art of the possible," "Religion is the opiate of the masses," and "Language is the mirror of the mind" share a form first noted by Aristotle:

As old age (D) is to life (C), so is evening (B) to day (A). One will accordingly describe evening (B) as the "old age of the day" (D + A)--or by the Empedoclean equivalent; and old age (D) as the "evening" or "sunset of life" (B + C). (Poetics, 1457b)

The genitive phrase in each case specifies a concept in a way incompatible with our default version of that concept.[1] Roots belong to plants, not to evil. A soul is possible for a human being, but not for wit. We get wages for work, not for sin. A mirror reflects a face, not a mind, because a face is visible but a mind is not.

How do we understand these sayings? We might be tempted to assume that we understand them because we have heard them used, and meaning is use. We might say that we just know metaphoric meanings of "root," "wages," "soul," and so on, and that we just use these metaphoric meanings to make sense of these sayings. But that would not explain how we understand parallel sayings that are altogether novel and even strange, such as "Causation is the cement of the universe," "Custom is the tyrant of mankind," "Art is the sex of the imagination," "Brevity is the soul of lingerie," "Today is yesterday's pupil," "Vanity is the quicksand of reason," "Cunning is the dwarf of wisdom," and "Sex is the ancilla of art." It is amazing that we understand and invent such phrases so easily.

The engines of metaphor do not run on their own. Principles of metaphor operate on familiar knowledge, allowing us to understand what we do not know in terms of what we know very well. In the preceding sections of "The Poetry of Connections," we considered general principles of great scope. Here, we will examine specific knowledge of limited scope. Ultimately, we will consider ways in which these general principles and this specific knowledge interact in understanding.

Conceptual domains that we know very well serve as basic source domains for metaphoric understanding. Eating, dress, learning, buildings, travel, combat, and plants are basic source domains, grounded in our forms of life.

Consider wealth, labor, and commerce. Our knowledge of this basic source domain is rich and familiar, enabling us to activate detailed ranges of it at will; a mere word like "money," "price," "treasurer," or "wages" activates for us this feltwork of knowledge and relations.

Metaphors such as "The wages of sin is death" direct us to use a basic source domain--in this case labor--to understand the relation between two things--in this case death and sin. Such constructions are instances of a special form that I will call "xyz metaphor." In this construction, "x," "y," and "z" are noun phrases, and "y" and "z" are connected by some linguistic form, such as "of" or "to," that typically indicates some relationship between the concepts y and z. The specific knowledge we are to use in constructing a meaning for the xyz metaphor comes from a source domain with which y is associated. This source domain is usually a basic source domain.

Consider, as examples of xyz metaphors, the following expressions, all of which have wealth and commerce as their basic source domain:

Children are the riches of poor men.

The wages of sin is death.

Vigilance is the price of liberty.

Trouble is the price of progress.

Labor is the capital of the workingman.

Memory is the treasurer of the mind.

A leader is a dealer in hope. (Napoleon I)

Gratefulness is the poor man's payment.

Evermore thanks, the exchequer of the poor. (Richard II, 2.3.65)

We understand these metaphors instantly, but not because they are expressions of conventional basic metaphors or generic-level metaphors. We understand them as directing us to consider the relationship between x and z metaphorically in terms of something unspecified from the y-domain. For example, "Children are the riches of poor men" directs us to consider the relationship between children and poor men metaphorically in terms of a domain containing y, namely, riches. To do this, we must be very familiar with the conceptual domain containing y, because the domain we are to use and the relevant useful knowledge within that domain are not mentioned in the xyz metaphor; we are given only the single word, "y."

For example, we take "Children are the riches of poor men" as directing us to understand the relation between children and poor men metaphorically in terms of some unmentioned knowledge from some unmentioned domain containing riches. An example of such a domain would be wealth and commerce. We know the possible relationships in this domain between y (riches) and the other parts of the domain. When we hear "riches of," we know the possible words that can follow. "Riches of kings," "riches of Rockefeller," or in general "riches of rich people" are possible phrases, referring in conventional ways to conventional knowledge we possess.

But in an xyz metaphor, "z" is not one of the things that can follow "y of," because we believe that y and z do not bear a relationship in some conceptual domain that we conventionally express by saying "y of z." For example, the phrase "poor men" is not something that can follow "riches of." "Riches of poor men" expresses a relationship between riches and poor men that is counter to our conventional knowledge: poor men do not have riches. We take an "x is the y of z" expression of this sort as prompting us to construct a metaphor.

An xyz metaphor prompts us to understand the conjunction between x and z metaphorically in terms of a conceptual domain containing y. In particular, we are to find some w in our conceptual knowledge that stands in a relation to y which we can refer to conventionally by the expression "y of w," and we are to map the relation between y and w onto the conjunction of x and z.

For example, "Children are the riches of poor men" prompts us to find a w, rich people, related to y, riches, and to map the relationship between y and w, that is, between riches and rich people, onto the conjunction of children and poor men. We are to understand the relationship between children and poor men metaphorically in terms of the relationship between riches and rich men. We do this according to the general principles discussed in the previous two chapters, with the additional help that the xyz metaphor cues us to map a relation onto a relation. In mapping the y-w relationship onto the x-z conjunction, we are guided by the invariance principle: we map onto the x-z conjunction the image-schematic information in the y-w relation.

Xyz metaphors are more varied in their appearance than we have so far considered. Not only "x is the y of z" but also "x is the y to z," "x is the y toward z," "x is the y for z," "the y for z is x," and many other constructions can prompt us to perform just the same sort of metaphoric mapping. In such cases, the xyz metaphor has the form of an equation between two noun phrases, where we can take one noun phrase ("x") as referring to some conventional concept x, but where the other noun phrase (namely, "y of z" or "y to z" or "y in z" or, in general, "yRz") is referentially weird because, if we use our conventional ways of understanding it, then it refers to something that runs counter to our conventional knowledge. We take any such equation as prompting us to construct a metaphor: we are to try to bring x and z together into a conceptual domain, and understand that domain metaphorically in terms of some conceptual domain containing y. In particular, we are to try to understand the relationship between x and z in terms of some relationship that does exist in our conceptual knowledge between y and some unmentioned w, such that the noun phrase "yRw" is not referentially weird, but expresses a conventionally known relationship between y and w.

We must distinguish between the skeletal form of the xyz metaphor and the lexical content that fills it. The skeletal form is general. The lexical content is specific, and differs from case to case. The lexical content guides us more or less specifically in each case we encounter. The skeletal form, which we know as a general construction, guides us generally, for all cases. Here, I characterize the semantics of that skeletal form, as it interacts with and is independent of the semantics of the lexical material that fills it. We all know the skeletal form, and use it to make sense of specific cases. In specific cases, we get more or less extra guidance from the lexical content.

For example, in "Money is the root of all evil," "root" is a relational noun, and strongly suggests plant as the unmentioned w we are to use, so that the root-plant relation is to be mapped onto the money-evil conjunction. Here, the lexical content of the xyz skeletal form gives us some guidance. But in "Vanity is the quicksand of reason," the lexical content gives us much less guidance. However much guidance the specific lexical content gives us, what we always have is the guidance given by the skeletal form itself, which is the subject here. In certain cases, virtually the only guidance we have in making sense of the xyz metaphor is our knowledge of the semantics of that skeletal form. In these cases, the particular xyz metaphor--such as "Vanity is the quicksand of reason"--contains lexical content that does not point to the metaphoric mapping. The metaphor is not in the words. The metaphor is rather in what we do mentally to determine that a mapping is in order and to construct that mapping.

An xyz metaphor gives us in general almost none of the information we need to understand the metaphor that it suggests. In general, it merely points us in a direction. We are equipped to understand it by bringing to bear a powerful commonplace conceptual apparatus that includes knowledge of basic source domains. We know different possible relationships between riches and rich people: riches can be spent by rich people, riches are valuable things to rich people, and riches bring pleasure to rich people. We seek such relationships in the y-domain that can plausibly be mapped onto the x-z conjunction. For example, we cannot map onto the conjunction between children and poor men the relationship that, as rich people literally spend riches, so poor people literally spend children, because it runs counter to our knowledge: poor men do not routinely spend their children. But we can map onto the conjunction between children and poor men a different relationship between riches and rich people: a poor man's children can be valuable to him, bringing him pleasure and joy.

In achieving this understanding, we use no prefabricated basic metaphor or even a generic-level metaphor. We arrive at this understanding imaginatively through taking a basic source domain we understand very well, and--constrained by the invariance principle and our conventional knowledge of the target--mapping from that basic source domain onto the target domain. In doing this, we are constrained not to violate the image-schematic structure of the target domain. For example, we do not take the metaphor as meaning that poor men can put children in the bank or buy stocks with them or offer them as collateral in a leveraged buy-out or pay income taxes with them or bribe foreign corporations with them or any of the many other things that rich people can do with their riches. We know not to take the metaphor as indicating any of these things because they all violate either our conventional knowledge or the event structures associated with the target.

Let us consider the basic source domain of travel, or more accurately, of physical progress toward physical locations. Consider the following metaphoric expressions:

Adversity is the first path to truth. (Byron)

Death is the end of woes. (Spenser)

[Of a hansom] The gondola of London. (Disraeli)

Self-respect is the gate of heaven.

Anarchy is the stepping stone to absolute power. (Napoleon)

Custom is the guide of the ignorant.

It is death that is the guide of our life.

The cross is the ladder of heaven.

Logical consequences are the beacons of wise men. (T. H. Huxley)

Commonplaces are the tramways of intellectual transportation. (Ortega y Gasset)

Cheese--milk's leap toward immortality.

The bar in America is the road to honor. (William Wirt)

We understand these xyz metaphors as prompting us to bring x (for example, The bar in America) and z (for example, honor) into a conceptual domain that is to be understood in terms of another conceptual domain (for example, travel) containing y (for example, road). This requires that the domain containing y, and the particular knowledge we are to use from that domain, be extremely familiar to us, because again that domain and that knowledge are not mentioned; we are only given the single word, "y."

Consider "The bar in America is the road to honor." We know that "road to" needs some destination to follow the "to." Our specific knowledge of the relationship between a road and its destination is that the road is something we can choose to engage with in a conventional way, namely, by traveling it, and that engaging with it in the conventional way will result in a change in the relationship between ourselves and the end of the road, namely, we will be at that end. The image-schematic structure of this relationship is to be mapped onto the conjunction of The bar in America and honor: that is, the bar is something we can choose to engage with in a conventional way, and doing so will result in a change in the relationship between ourselves and honor, namely, we will be honored. In order to conduct this mapping, we must understand the target domain image-schematically in terms of a journey, and the metaphor explicitly asks us to do that. It explicitly asks us to understand the bar as a path, and honor as the end of that path. We are therefore guided to understand conventional engagement with the bar (that is, becoming a member of the bar and working as a member of the bar) as traversing that path. Understanding the bar in America as a path that we can traverse, and that leads to honor, is the result of image-schematic mapping.

In principle, we can arrive at this understanding without using a prefabricated basic metaphor or even a particular generic-level metaphor; in practice, in this case, the general principles of conceptual connection have resulted in relevant basic metaphors and generic-level metaphors that already contain the image-schematic mapping.

How do such conventional conceptual metaphors come to exist? The xyz form has something to teach us on this score. Suppose we have many specific target domains, each containing some means. Suppose we understand each of these means in each of these specific target domains in terms of paths. Then the cognitive pattern inhering in all these metaphoric understandings would be the conventional conceptual metaphor MEANS TO GOALS ARE PATHS TO DESTINATIONS, and could be used to structure yet other target domains.

Let us drop down a level in specificity. Consider two conceptual domains, JOURNEYING as the source and SOLVING A PROBLEM as the target. Mapping the image-schema of a directed path onto this target domain in such a way as to preserve topological structure (such as ordering) results in mapping the destination onto the solution, the path onto the means of solving the problem (that is, onto the mental work that one must complete in order to reach the solution), the traveler of the path onto the mind doing the thinking, distance traveled along the path onto the degree of success in solving the problem, and so on. The result is a specific-level metaphor, MENTALLY SOLVING A PROBLEM IS JOURNEYING ALONG A PATH TO A DESTINATION. Repeatedly understanding mental problem-solving in this way entrenches the specific-level metaphor, and makes it conventional, until it becomes a basic metaphor. MENTALLY SOLVING A PROBLEM IS JOURNEYING ALONG A PATH TO A DESTINATION is in fact a very basic metaphor that we use frequently.

This combining of the invariance principle and the extremely specific and textured knowledge we have of basic source domains underlies the entire range of xyz metaphors. The range of such metaphors is infinite: we can and do construct them at will, often with seeming effortlessness.

We can understand the target relationship in terms of more than one source domain, as in Walter Lippman's "Social movements are at once the symptoms and the instruments of progress." We can understand more than one target conjunction in terms of a single source relation, as in "As poetry is the harmony of words, so conversation is the harmony of minds." We can understand x as a component of more than one x-z conjunction, as in "The idea is the old age of the spirit and the disease of the mind." Other complex xyz metaphors appear in the appendix.

The cognitive process behind the "xyz metaphor" is not restricted to the linguistic construction "x is the yRz" where R is some linguistic form, like "to" or "of," that we take as expressing a relationship between y and z. The cognitive xyz structure also underlies compound noun forms like "disc jockey," "road hog," and "budget ceiling." These are abbreviated expressions of the xyz metaphor. Consider "disc jockey." In this case, a person is x. y is jockey. We expect jockey to be associated with race horses. But instead, this jockey is associated with a z--namely, discs or records--that does not belong to the domain of race horses. The person (x) is a jockey (y) not with respect to horses (w) but rather with respect to discs or records (z). We understand this metaphor just as we understand all xyz metaphors: the relationship between x (the person) and z (discs, or records) is to be understood metaphorically in terms of the well-known relationship between a jockey (y) and horses (w). Consider another example. A pair of dogs runs to the front door when they hear a noise outside. They push aside the curtains that cover the glass panes in the door, and sit looking out the glass panes at anything that goes by. They bark their approbation, but in general stare out transfixed by more or less nothing. The dog owner refers to this door and what the dogs see through it as "dog TV," which is to say, television for dogs. We understand this compound noun by understanding metaphorically that this glass-paned front door (x) is TV (y) not for human beings (w) but rather for dogs (z). Again, y belongs to a very familiar basic source domain (television entertainment); we know that television is for human beings (w); we are given a relationship not between y and the expected w, but rather between y and a z that is not part of the source domain containing y. English is rich in compound nouns formed along these lines: "tax bite," "computer jock," "dog heaven," "card shark," "land shark," "record jacket," "record sleeve," "fire storm," "wave train," "brain spasm," and "price war." I have heard a large, expensive car referred to as a "land yacht." A children's program on TV is referred to as "baby crack," where crack is an addictive drug. A vendor of aloe vera plants in a flea market says, "You have your aloe vera, honey? That's your kitchen medicine cabinet." In certain cases, the xyz form can become one word, like "brainwashing," as in the phrase "Brainwashing is evil." The activity known as "brainwashing" stands to minds as washing stands to things washed.

Backlash on the Unmentioned Concept

Occasionally, the source and target in an xyz metaphor will overlap in their components, leading to the possibility of a "backlash" comment on the source.[2] Consider "Children (x) are the riches (y) of poor men (z)." The missing w is rich men: The children-poor men (x-z) conjunction is the target, to be understood in terms of the source relation between riches and rich men (y-w). But children is a possible component not only of the target but also of the source: there is a natural connection not only between poor men and children but also between rich men and children; this allows us to read the expression as implying that rich men have sadly violated this natural connection by allowing riches to supplant children. This is a backlash reading; it is a comment upon the source, not the target. In particular, it is a comment on the unmentioned concept, rich men, in the source. "Sex (x) is the opera (y) of poor men (z)" also has a backlash reading: sex, which is a possible component not only for the target but also for the source, has been supplanted for rich men by opera.

Implications of XYZ Metaphors

An xyz metaphor, such as "Fame is the beauty-parlor of the dead" or "Humanity is the sin of God" or even "disc jockey," gives us almost none of the information we need to understand it; it merely points us on our way toward a meaning. It mentions y but does not mention the domain containing y or the possible w's to which y can bear a relationship in that domain, or the relationship that actually obtains between y and such a w. It does not tell us to perform a mapping. It does not tell us how to perform that mapping. In order to understand an xyz metaphor, we must bring to bear elaborate and detailed conceptual knowledge not referred to in the expression. This is the common situation of all language: expressions do not mean; they are prompts for us to construct meanings by working with processes we already know. In no sense is the meaning of an xyz metaphor or of any utterance "right there in the words." When we understand an utterance, we in no sense are understanding "just what the words say"; the words themselves say nothing independent of the richly detailed knowledge and powerful cognitive processes we bring to bear.

The construction of such meaning is, as we have seen with xyz metaphors, open-ended. Our understanding of an xyz metaphor can be skeletal, rich, or in between, depending upon our interpretive energies. It is misguided to claim that the words of the expression tell us just exactly so much, and that anything else we find in the expression is the result of our "reading meanings into" the utterance. All reading is reading in. If one stopped at "just what the language says," then one would never begin understanding at all.

Consider, as an example of this open-endedness, "Vanity is the quicksand of reason." We understand "quicksand of" as referring to some conceptual domain containing quicksand, in which quicksand bears some relation to some w that we can express linguistically as "quicksand of w." For example, stereotypically, quicksand is part of the desert, so "quicksand of w" might suggest to us "quicksand of the desert." In this case, we would see the metaphor as asking us to see vanity as part of reason in the way quicksand is part of the desert. While this reading is possible, and could be pushed, let us instead pursue the alternative approach in which we assume that stereotypically vanity conflicts with and is not part of reason. Then we would more plausibly take "quicksand of w" as evoking things that travel across the desert, such as human beings, animals, and vehicles. The implication would be that as quicksand acts on things traveling in the desert, so vanity acts on reason.

In that case, "Vanity is the quicksand of reason" would be a special instance of the more general metaphor that MENTAL PROGRESS IS JOURNEYING, expressed in "Let's walk through the problem," "P = NP is the Mt. Everest of complexity theory," and "We have to work our way around that problem." How might we lock together MENTAL PROGRESS IS JOURNEYING, or the general principles of metaphoric understanding that result in this metaphor, with our knowledge of quicksand, vanity, deserts, and reason to arrive at a metaphoric understanding of the vanity-reason conjunction in terms of the quicksand-travelers relationship?

We have stereotypical knowledge that quicksand acts on travelers crossing the desert. We can use this knowledge of the source domain in a variety of finely grained ways to understand that just as quicksand catches travelers unaware, so vanity catches reason unaware; that vanity, like quicksand, is a hidden danger that can be skirted if one uses proper caution; that the danger of vanity, like the danger of quicksand, is ever-present, requiring constant precaution; that the domains of discovery over which reason wanders contain hidden traps; and so on, and on, and on, bringing to bear a great deal of very finely grained knowledge about the source domain.

We have choices in what finely grained knowledge we bring to bear. If we bring to bear the conventional but erroneous stereotypical knowledge of quicksand that it occurs where there is no water and that it swallows up victims whole, then we conceive of reason as a person quickly swallowed whole by vanity. If we bring to bear experiential knowledge that quicksand almost surely will not swallow a traveler, we might understand that reason can encounter but pass through vanity if reason simply keeps its wits and moves quickly and fluidly without becoming distracted by momentary encounters with vanity. But if reason dallies with vanity, or dallies unaware that it is in vanity, then reason might become bogged in the midst of vanity, and its progress halted. At that point, reason might call on one of its companions--love, anguish, self-loathing, nihilism--to pull it out of vanity. If none are there or can help, reason might try to extricate itself. If shrewd or properly trained, reason might succeed. If not, its motions to escape will only sink it deeper. And while vanity will almost certainly never swallow reason, reason might be caught half-submerged in vanity, unable to escape, slowly dying, constantly aware of the increasing probability of its near death. And so on and on and on.

Consider "London is the cesspool of the Empire." This xyz metaphor has been extracted from Arthur Conan Doyle's more elaborate version, "London, that great cesspool into which all the loungers of the Empire are irresistibly drained." This metaphor evokes two vast domains and prompts us to investigate the fine details of our knowledge of these domains. It places one foot in each of two domains: the first is London, the explicitly mentioned x. Connected with the domain of London is the phrase "loungers of the Empire." But this phrase is also used to modify the y domain, cesspool. This y domain is also explicitly mentioned, and the phrase "drained" is connected with that y domain. What are the possible mappings here? The loungers of the Empire, we are told, get "drained" into London. So we must map cesspool onto London, and whatever gets drained into a cesspool onto loungers. What gets drained into a cesspool? We know that the answer is liquid waste, which is not mentioned. So we may try to understand loungers of the Empire metaphorically in terms of liquid waste. How would we do that? There is generic-level information in our concept of liquid waste: it is undesirable and fluid, and the waste product of something. None of this knowledge is referred to in the expression. We understand loungers metaphorically in terms of liquid waste by understanding the loungers, at the generic level, as undesirable, fluid, and the waste products of something. Of what might they be the waste products? We might take the expression as indicating that Empire is the source of these waste products. This leads to another set of metaphoric mappings. The Empire must have waste products, so it must be a dynamic process, or a collection of processes (like a body or an industry, or a town full of industries and bodies, or a market-place full of industries and bodies and ecological systems), all of which have waste products. Such processes have crafted products, births, deaths, and so on, but certainly they have waste products. Therefore, while the Empire may have many components, it also, like all dynamic systems associated with "cesspool," must have metaphoric liquid waste products. We know that liquid waste is drained. Drainage takes place in physical environments with either natural or artificial structure that impels, by gravity or other means, the waste products toward the drain. We may take this as suggesting that, as gravity and physical structure impel liquid sewage toward the cesspool, so some component of the structure of Empire impels loungers toward London.

Let us pause to notice how much knowledge we have to bring to bear to arrive at such an understanding, even though that understanding seems automatic and effortless. The possibilities of physical structure must be mapped to the possibilities of political, societal, economic, transportational, or, in brief, Imperial structure. The mapping of physical structure onto Imperial structure entails a mapping of causality between the two domains: something in the physical structure causes liquid sewage to be impelled to the cesspool; something in the Imperial structure causes the loungers to be impelled to London.

Why should this impelling structure of Empire work on loungers but not on others? Asking this question (unconsciously) may prompt us to yet a further metaphoric understanding: as waste has no life, no power, no will to resist the flow, so too are loungers impotent to resist--much less employ--this structure of Empire. We may take ourselves as nudged toward this interpretation by the cue "irresistibly." Presumably capitalists and artists are drawn to London, but drawn, not drained. How else might loungers in an Empire be like waste products in a house, town, market, and so on? Is it that, as waste products are those things the industries and bodies have no use for, so loungers are those people that the Empire has no use for? Is it that loungers, like waste products, have nothing to contribute?

Let us stop there with this metaphoric understanding. It is clear that with the appropriate interest and energy, it could be extended to the length of a novel. The lessons we draw from the way we go about understanding such an xyz metaphor are:

--The invariance principle guides us. In this metaphoric mapping, paths are mapped to paths, an image-schema of causation is mapped to the same image-schema of causation, the general shape of the target event is preserved, and so on.

--We bring to bear a tremendous range of detailed knowledge, including such knowledge as the association of London with the British Empire, and its association with the work ethic and Victorian values. Vast ranges of conceptual knowledge not in the least explicit in the expression are indispensable for understanding the metaphor.

--Understanding a metaphor is often actually understanding whole systems of metaphoric mappings, involving many components in the source and the target.

--There is no natural terminus to understanding a metaphor. It is nonsense to say that the reader should stop when he has determined "just what the linguistic expression says," because the linguistic expression itself does not mean.

--The power of language lies not in words, but in the mind.

Language is Fossil Poetry

Earlier, we considered cases like "Art is a jealous mistress," "This world is an unweeded garden," and "The Queen, that bright Occidental star," where the source domain was specified in some unobjectionable way. But in certain illuminating cases the specification runs counter to our conventional knowledge. Consider "Language is fossil poetry." Many things can be fossil, but we do not think of poetry as one of them. Consider "Daniel Webster struck me much like a steam-engine in trousers." Many things can wear trousers, but we think a steam engine is not one of them. Such a specification runs counter to our concept of what it specifies. I will call such specifications "weird." How do we understand metaphors containing weirdly specified source concepts?

Specifying the Source with Something That Belongs to the Target

Consider "Daniel Webster is a steam-engine." This is a bare metaphor, which we understand by means of THE GREAT CHAIN METAPHOR. We know that Daniel Webster is one form of being, a human being, and that a steam engine is a different form of being, a machine, and that our understanding of Daniel Webster metaphorically in terms of a steam-engine by means of THE GREAT CHAIN METAPHOR must bridge over exactly this distinction.

Many metaphoric expressions simply take the principal information separating the target from the source and pin it weirdly on the source, as in "Daniel Webster is a human steam-engine." This weird specification then emphasizes exactly that distinction which we must bridge over when we use THE GREAT CHAIN METAPHOR.

The information pinned on the source can also be something metonymic of the principal distinction separating the target from the source. For example, wearing trousers is metonymic of being human. Hence, "Daniel Webster is a steam-engine in trousers."

We take such weird specifications as emphasizing just that distinction we must bridge over in using general principles to construct a metaphoric understanding. I label such a weird specification merely weird, to distinguish it from really weird specifications that we will encounter in a moment. A merely weird specification pins something from the target onto the source. Metaphors that involve source domains that have been specified in a merely weird fashion are extremely common, and occur in phrases such as "Macaulay is a walking encyclopedia," "My big brother is a living garbage disposal," "He's a living dream," "This airplane is a flying garbage can," "He's a walking time-bomb," "Old men are walking hospitals," "She's a walking beauty parlor," "She's a computer with legs," "This car is a rolling disaster area," and "He's a human vegetable." The obese beggar Iros in the Odyssey is referred to as a "walking famine." Less common examples are "A minister of finance is a legally authorized pickpocket," "Hope is a waking dream," and "A father is a banker provided by nature."

The principal distinction on the Great Chain between Helen and Venus is mortality versus immortality. Hence "Helen is the mortal Venus." The principal distinction between a mixed metaphor and a sore thumb is that a mixed metaphor results from cognitive behavior while a sore thumb results from physical behavior. Hence "A mixed metaphor is a cognitive sore thumb." Understanding Helen in terms of Venus while emphasizing the mortal versus immortal contrast between them is to understand that Helen in the mortal world is the metaphoric equivalent of Venus in the immortal world. Understanding mixed metaphor in terms of a sore thumb while emphasizing the contrast of cognitive versus bodily between them is to understand that a mixed metaphor is the metaphoric equivalent in the realm of cognitive behavior of a sore thumb in the realm of bodily behavior. The case is similar for "Cynicism is intellectual dandyism."

Commonplace Transformations

We know how to transform things mentally from one state into another. Certain transformations are commonplaces. Specifications often ask us to transform a concept according to a commonplace transformation.

We have concepts not only of objects like dog and table, but also of actions like theft and departure and of events like healing and death. Many such actions and events can be thought of as transformations--operations that transform something from one state into another. Enlarging an area, coloring an object, rearranging flowers, coining metal, breaking a piece of pottery, freezing water, educating a mind, and getting a body into shape are all commonplace transformations.

Our knowledge of a commonplace transformation includes what it works on. Education operates on minds, not water. When we think of freezing water into ice, or of educating a mind by taking it from ignorance to knowledge, we are thinking of a commonplace transformation. A commonplace transformation is a conventional transformation as applied to the sort of thing it conventionally applies to. For example, we think of a polished stone as the result of a commonplace transformation, namely, polishing, which transforms the stone from its natural state to a polished state.

A commonplace transformation of a concept results in a specification of that concept, as follows: a commonplace transformation of some concept x creates some final transformed state of x, and we think of that final state as a specification of x. Polishing is a commonplace transformation that we think of as applying to, among other things, stone. The result of applying polishing to stone is polished stone. We think of a polished stone as a specification of stone.

This raises a curious point. At the conceptual level, an unpolished stone and a polished stone are tightly connected. This should seem bizarre: an unpolished stone and a polished stone are completely incompatible. No unpolished stone can be a polished stone and conversely. Yet we do not think of unpolished stone and polished stone as far apart. Apparently, we feel that commonplace transformations preserve some part of the essence of a concept.

Many specifications can be thought of as caused by commonplace transformations. We think of a jealous mistress as a specific kind of mistress, namely, one who is jealous, but we can equivalently think of a jealous mistress as one who is jealous because she has been made jealous. Any such transformation of a person from one affective state to another by means of some action or perception is a commonplace transformation.

Commonplace Transformations That Are Specifications Transforming the Source into the Target

One way to understand what it means to bridge over such a distinction is to understand that some commonplace transformation transforms the source so that it is no longer separated from the target by that distinction. Consider "A rhombus is a square pushed over." Pushing over is a commonplace transformation that can be applied conceptually to shapes. The specification pushed over applies to the target, rhombus, but not to the source, square. This distinction between the target and the source is removed by applying to the source a commonplace transformation that removes the distinction.

For many of the metaphors involving sources specified with something fitting the target but not the source, it is possible to see the specification as suggesting the application of a commonplace transformation that would transform the source into the target. In "Las Vegas is the American Monte Carlo," the specification "American" fits the target, Las Vegas, but not the source, Monte Carlo; we understand that it emphasizes the principal distinction between Las Vegas and Monte Carlo: Las Vegas is American, but Monte Carlo is European. We must bridge over this distinction in constructing the metaphor. So far, so good. One way of bridging over this distinction is to take "American" as suggesting a commonplace transformation, namely, Americanize. We often talk of Americanizing non-American parts of the world. Americanizing them involves changing their economies, their cultural patterns, the stereotypical styles of their citizens, and so on, to bring them into line with American forms. We think of Monte Carlo as aristocratic, imbued with an atmosphere of European high culture and pretense. To Americanize it might mean to eliminate its aristocratic and cultural atmosphere, substituting instead the jumble of American demographic types coupled with vulgarity in architecture, speech, and behavior. Monte Carlo connotes old money and the psychology that goes with monetary security. To Americanize it might mean to substitute instead new money and the psychology of the lottery.

It should be clear that a commonplace transformation is not necessarily a realistic transformation. It is instead a transformation that we can apply conceptually. It is implausible that one could ever Americanize Monte Carlo (or that one would want to), but one can imagine it. One cannot actually push over a drawn square, but one can visualize it mentally as being pushed over. One cannot actually humanize a vegetable or a tree or a car, but conceptually, it is quite easy for us to think of humanizing such things, as we see in animated cartoons. When, in a metaphor like "He's a human vegetable," something fitting for the target but not the source is attached to the source, thereby emphasizing the distinction separating the target and the source, we know that we must bridge over that distinction to construct the metaphor. One way to understand what it means to bridge over that distinction is to understand that a commonplace transformation separates target and source.

Specifications That Are Commonplace Transformations Inappropriate for Both Source and Target

Consider "Language is fossil poetry." The specification, fossil, is really weird because it is inappropriate for not only the source but also the target. We think of a fossil as the result of a commonplace transformation, fossilization. But we think of this transformation as applying only to organisms and physical traces of organisms, such as bones, footprints, leaf imprints, and so on. It certainly does not apply to poetry. Therefore, we cannot think of fossil poetry through commonplace transformation of poetry in the way we can think, for example, of fossil bones or fossil footprints as commonplace transformations of the concepts bones and footprints. The commonplace transformation fossilization does not apply to poetry. Similarly, we think of various things as the results of fossilization, but language cannot be one of them. So, in this case, the commonplace transformation is inappropriate for both the source and the target: it cannot conventionally be applied to the source, and the target cannot be the result of applying it to anything.

Yet, in a different way, fossilization can apply to poetry. Although the specific-level information in the commonplace transformation fossilization cannot be applied to poetry, the generic-level information in fossilization can be applied to poetry. A commonplace transformation has a Nature, and consequently a behavior. Fossilization, for example, has a Nature: it behaves in a certain way, on certain sorts of things. This Nature has both specific-level information and generic-level information. When such a commonplace transformation is inappropriately applied to a concept, we take it as a cue to apply the generic level of The Nature of the Transformation to that concept. For example, the generic level of fossilization is: a transformation in which the passage of long time transforms or takes the impression of part of a changing, interactive world, resulting in a rigid fixity that has none of the original freshness or changeability or capacity of what was fossilized but that bears its marks. If we apply this generic-level Nature of Fossilization to poetry, then the result is: a transformation or impression-taking of poetry through the passage of long time, thereby creating a rigid fixity that has none of the original freshness or changeability or capacity of poetry, but that bears its traces. We are therefore to understand language as this fixity. Language is the final state that derives from the original state, poetry, and the mechanism of this derivation is the generic-level of The Nature of Fossilization. The passage of time has turned poetic inventions into a rigid language system that still bears the traces of poetry.

Really Weird Specifications and XYZ Metaphors

A metaphor with a source that is specified in a really weird way may have an equivalent xyz form. "Language is fossil poetry," for example, is equivalent to the xyz form "Language is the fossil of poetry." Let us inspect the details of this equivalence. We understand "Language is fossil poetry" by taking the generic level of the commonplace transformation of fossilization, applying it to poetry, and mapping the result onto language. We understand "Language is the fossil of poetry" in the same way, as follows: We know that we are to map the relationship between fossil (y) and that which has been fossilized (w) onto the conjunction of language (x) and poetry (z). What is that relationship? It is that the first is the result of applying the commonplace transformation fossilization to the second. What is entailed in mapping that relationship onto the target? Guided by the invariance principle, we map the generic level of that source relationship onto that target conjunction: as a fossil is the result of fossilizing something, so language is the result of the generic-level fossilization of poetry. In both cases, then, we map onto language the result of applying the generic level of fossilization to poetry. The xyz version and the version involving the really weird specification are equivalent.

Conclusion

In this study of the poetry of connections, we have seen a high-level coherence among the principles of conceptual connection. First, we discussed basic metaphors, then generic-level metaphors that both underlie basic metaphors and enable us to understand metaphoric expressions when no basic metaphor does the trick, and then higher-order principles of metaphoric understanding that underlie all metaphoric understanding, including basic and generic-level metaphors.

We observed how we employ these higher-order principles to understand bare metaphoric expressions that are not instances of basic metaphors and that contain no cues of the sort we find in xyz metaphors. We found that we use these higher-order principles in conjunction with our commonplace notion of The Nature of Things.

We examined our processes for understanding a particular construction, the xyz metaphor: we take it as directing us to understand the x-z conjunction in terms of some y-w relation drawn from a basic source domain like eating, dress, combat, travel, and so on. This led to a theory of the origin of basic metaphors: a basic metaphor is the entrenched result of applying higher-order principles of metaphoric understanding to a certain basic source domain in order to understand a certain target domain.

We have taken up cases where the source domain is specified in a way that is merely weird or really weird. We understand the merely weird case--where something from the target is pinned on the source--in just the way we understand any bare metaphoric expression, save that we take the merely weird specification as cueing us to the principal distinction between source and target that we must bridge over in applying higher-order principles to arrive at a metaphoric understanding. We considered commonplace transformations as conceptual connections between two concepts where one is viewed as the initial state and the other is viewed as the transformed state. We saw that some merely weird specifications can be understood as commonplace transformations that conceptually transform the source into the target. We then took up the case of really weird specification, where a commonplace transformation that is inappropriate for both the source and target is applied to the source. We saw that we can understand such metaphoric expressions as asking us to perform a mapping at the generic level: we are to apply to the source the generic level of the transformation and map the result onto the target. We then saw that source domains that have been specified in a really weird fashion have equivalent xyz metaphors, where the conceptual processes involved in understanding them are isomorphic.

This book, which is meant as a start-up program for the reader's own activities, has now concluded its illustrative guided tours intended to suggest lines or modes of research open in the study of English in the age of cognitive science. They are initial laboratory experiments, provisional and topical only, offered in the hope of leading to a reconsideration of what our research into language and literature might become if we began to think of language and literature as acts of a human mind in a human brain in a human body that must make sense of its environment if it is to survive.

We have one more tour to take. It concerns not the nature of our research in the age of cognitive science, but the nature of our teaching.