Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives

SRB Highlights

This review appeared in Volume 8 (3) of The Semiotic Review of Books.

Toward a Cognitive Rhetoric

by Jørgen Dines Johansen

Reading Minds. The Study of English in the Age of Cognitive Science. By Mark Turner. (1991) Princeton: Princeton University Press. x+298pp. ISBN 0-691-06897-6; ISBN 0-691-00107-3 (paper back).
The Literary Mind. By Mark Turner. (1996) New York/Oxford: Oxford University Press. ISBN 0-19-510411-0.

It may be argued that the sort of cognitive linguistics, semantics and philosophy practised by George Lakoff and his circle of colleagues, associates, and students at Berkeley has now reached maturity. Although Metaphors We Live By (Lakoff and Johnson 1980) attracted a lot of attention, it was a kind of "How to Do Metaphor Studies with a Sledge-Hammer", a very provocative book but one lacking nuances and extended argumentation. In 1987, however, both Lakoff and Johnson independently published brilliant and important scholarly books: Women, Fire And Dangerous Things and The Body in the Mind, respectively. In addition, much research has been done and a steady stream of publications by colleagues and associates along the lines laid down by this approach has followed influencing linguists, psychologists, philosophers, and others worldwide.

It is fascinating to follow the development of this approach to cognition and representation. Primarily, of course, in order to profit from the insights it offers, but also to see whether it has potential for development beyond what has already been achieved. The latter is not obvious. Glossematics, for instance, was never taken beyond the point to which Hjelmslev himself brought it. The value of glossematics lies rather in the influence it has exercized on other structural approaches of language. Like Russian Formalism and the Prague School it has been part of a broader trend. Also its successor, structural semantics as it was laid down by Greimas (inspired by Hjelmslev) - I think it is fair to say - remained within his understanding of it. I am no expert on generative grammar, but it is my impression that it has not advanced further than Chomsky himself has taken it; and in all three cases some might say that creativity and stamina have diminished.

There are two ways, however, in which a given theory, or approach, can develop, in breadth and in depths. And whereas glossematics was not very much developed in breadth, both structural semantics and generative grammar were applied outside their original field. Within literary studies, the Chomskyan approach very soon proved abortive, whereas the case of Greimas is different, since Greimas himself applied structural semantics to literature in the last chapter of Semantique structurale (1966), in his book-length study of a short story by Maupassant, and on many other occasions. The debt of literary studies to structuralism and especially to structural semantics is far from negligible.

The reason for which I mention the tradition from Russian Formalism to the Prague School and from Glossematics to Greimas's structuralism is that there are links and similarities with Lakoff's variety of cognitive studies. Lakoff himself studied with Roman Jakobson, and it is evident that Lakoff's approach offers a developed and up-dated version of many of the insights of structural linguistics in general and of Jakobson inparticular -- albeit with its own agenda.

However, for a literary scholar and semiotician, the advent of cognitive semantics (and of pragmatics) meant that the gap between literary studies and linguistics which was created by generative grammar has been bridged again.

Like Jakobson and Greimas, Lakoff has been directly involved in literary studies. He wrote a thesis on the grammar of the folktale, and he co-authored with Mark Turner More Than Cool Reason. A Field Guide To Poetic Metaphor (1989), probably the best introduction to the study of metaphors available to the student of literature. It is Mark Turner, however, who is the literary scholar, and who has developed the cognitive approach to literature in two books of his own: Reading Minds. The Study of English in the Age of Cognitive Science (1991) and The Literary Mind (1996). Consequently, I will concentrate on his contribution to the study of literature and on its relation to the work of cognitive linguistics, philosophy, and neuroscience.

In Reading Minds Turner states that it is his ambition to reconstitute the profession of teaching English. A profession that has been severely damaged by the contemporary practice of literary scholars and critics who, instead of analyzing literature for the benefit of students and public, are writing theory for each other. Thus they are cut off from the full human world and live within "an exclusive Disney world for literary critics" (Turner 1991: 4). His own understanding of the profession is conditioned by his understanding of what literature essentially is:

Language and literature are suffused by the full human world of the everyday. That is their basis. This full human world, the world that comprehends language and literature, exists for us independently of any academic theory. Whether the academic theory is invented or not, whether it attains an ascendency or vanishes, the full human world, its language and literature, abide, exactly because our grasp of this world and our operation within it depend not upon academic theories but upon a commonplace conceptual apparatus that is textured and powerful. Literature is the highest expression of our commonplace conceptual and linguistic capacities. Literary criticism touches home base, the full human world, to the extent that it sees literature as the expression of everyday capacities and helps us to understand them. (Turner 1991: 4)

Consequently,Turner suggests that the common ground and foundation of literary studies should be sought in the relationship between brain, body, and life-world, i.e., both in the analysis of the bodily presence in the world and in the study of the workings of the brain.

Let me try to situate this cognitive approach to literature in the critical landscape by a historical detour: In the Poetics of Aristotle we find two necessarily related, but opposed conceptions of literary aesthetics, an aesthetics of similarity and one of difference. The aesthetics of similarity is foregrounded in his idea of literature as mimesis, as the imitation of men in action. However, no sooner has he stated this fundamental principle of imitation, than he points out that literature is different from life, because (1) it depicts men either better than they are (i.e., of more active virtue: tragedy) or worse (i.e., more ridiculous or grotesque: comedy). He further states that(2) it may represent human action and interaction as it ought to be, not as it is. In addition, he states that (3) literature may better present what is probable but impossible, than possible but improbable. And finally, he states that (4) the language of literature should be different from everyday language. In the further history of poetics we constantly find this opposition of an aesthetics, or poetics, of similarity and one of difference. The realism of the 19th century, for example, was an aesthetics of similarity, whereas, in the 20th century, Formalism, in its first phase, propagated the aesthetics of difference. The aesthetics of literary theory from the perspective of cognitive science is definitely an aesthetics of similarity, at least as conceived by Turner. Not because the specificity of literature as an individual kind of discourse is denied, but because one of its main points is that literature uses, and refines, principles of cognition that are fundamental and operative in everyday life. Literature utilizes the everyday language and the conventional and stereotyped ways of thinking which coincide with it. Literary theory must develop methods of analyzing the common ways in which texts are structured, and, Turner points out, the analysis of that which is so common that it is almost invisible is more difficult to analyze than that which sticks out as being unfamiliar. Furthermore, the originality of literature does not consist in the use of different principles of patterning than everyday cognition, but in the more complex ways they are used. Thus, at this point at least, the mimetic nature of literature is conceived primarily as a mimesis and display of the basic ways in which the mind works.

The cognitive approach has, according to Turner, two advantages over most other critical theories: 1) It is committed to data, thus it should be possible to prove it wrong. 2) It is not monolithic, but is able to integrate insights from other theories, to take over where they left off. This also means that it has insured itself in regard to the future, because it closely follows the development of the other disciplines that constitute the broad field of cognitive science (linguistics, psychology, philosophy, anthropology, AI, neuroscience, etc.), and it will adjust to their findings and attempt to integrate elements from them in its own approach. In its self-understanding, then, it is not a dogmatic doctrine immune to alien influences, but it is a part of a common endeavor to inquire into the nature of human cognition.

In Reading Minds Turner wants to go back to basics, i.e., to analyzing the general and unconscious processes of cognition in the light of the different disciplines of cognitive science: Because "the mind is the brain and the human person is patterns within the brain"(33-4). Further, most mental activity is unconscious, and the body and the world are represented in the brain, and it is only through it (and the organs of sense) that we know and ascribe meaning to them.

One immediate consequence of this view is a changed understanding of the nature of concepts. They are not mental objects defined by critical features, but cognitive processes that activate different neural patterns depending upon the context and the aspects called forth by it. And concepts are changing over time, as regards both phylogenesis and ontogenesis. Further, our unconscious thought processes are impregnated with the representation of the body in the world through the image schemas (cf. Mark Johnson 1987) that are responsible for our moving, and surviving,in the physical world and on which basic concepts (not linguistic expressions) such as up and down, forward and back, right and left, continuity and discreteness, linearity, circularity, path, progress, boundaries, links, interiors and exteriors, centers, penetration, covering, etc. are based (38). Such basic structuring devices, that have been analyzed by Lakoff, Johnson, Turner, Sweetser, and others, do not only pattern our orientation in space (and time), by metaphorical extension, they are also projected onto, and thus conceptualize, non-spatial relationships. To illustrate their point, let us look at Mark Johnson's analysis of the image schema of LINKING and Lakoff's analysis of the CONTAINER schema. The analysis of concepts are concentrated on two levels. In Johnson's own words:

(a) The basic level, at which we distinguish elephants from giraffes [...] and at which we distinguish walking from running, and standing from sitting. This is the level of understanding that we have evolved to permit us to function passably well in our environment. [...] (b) The image-schematic level, which gives general form to our understanding in terms of structures such as CONTAINER, PATH, CYCLE, LINK, BALANCE, etc. This is the level that defines form itself, and allows us to make sense of the relation among diverse experiences. (Johnson 1987: 208)

First, links are physical and biological phenomena. We owe our very existence to being linked with our birth mother through the umbilical cord. This cord exemplifies the basic idea of the link as the linking of at least two entities: A--------------B. In addition to spatial and physical links we experience temporal links between events that we do not necessarily perceive as physically related at one instance in time. The experience of temporal linkage allows us to understand our life world in terms of causal connections, and Johnson adds, "without such causal conjunction we could never experience our world as a relatively comprehensible place" (118).

Linking and links need not be physical, as Johnson puts it: "The severing of the umbilical cord launches us into an ongoing process of linking, bonding, and connecting that gives us our identity (117). Johnson does not expand on human bonding, but it should be pointed out that physical connectedness is evidently prominent in the early stages of infant's bonding. The importance of the psychoanalytical concept of "holding", (e. g. the mother's holding the infant) has been stressed among others by Winnicott (1971). Although physical closeness and contiguity continues to play a major part in human bonding throughout life, bonding transcends mere physical contiguity. It also concerns caring, benevolence, emotional comfort, security, trust, respect, and love.

Furthermore, bonding may be extended to persons with whom the individual has never had any physical contact, e.g. charismatic leaders, and to non-human entities, gods and demons, abstract ideas (the nation, progress, revolution, justice, etc.). In human bonding the emotional dimension is foregrounded, viz. the metaphor "emotional ties" which testifies to its importance. This locution is at the same time an example of metaphorical extension of the image-schema of physical linking to mental states. Johnson points to another extension of the simple LINK schema by saying that:

[it] makes possible our perception of similarity. Two or more objects are similar because they share some feature or features. Those shared features are theircognitive links in our understanding. Here, obviously, we have a highly abstract notion of linkage, in which the "third thing" that binds or relates two objects is a perceptual or logical feature. The link schema must be metaphorically interpreted to apply to abstract objects or connections, since there is no actual physical bond of the required sort to relate the objects. (ibid. 118-119)
Obviously, there is an important difference between two individuals being linked through an umbilical cord and the relatedness of two abstract entities because of shared features, such as for instance the phonemes p and b, which are related by being both bilabials, but distinguished by the fact that the former is unvoiced the latter voiced, and between the relation of similarity between a source domain and a target domain in a metaphor such as "life is a journey" -- one of Lakoff's favorite examples, or the linking of two variables by implication -- if q, then p -- within logic.

Bonding implies a metaphorical extension of the LINK schema based on emotional responses, and the original emotional relationships lie in concrete, physical relations of contiguity between individuals. As to the second extension of this schema, Johnson mentions that logical predicates and propositions are related via "logical connectives" ("and," "but," "or," "if-then") in an act which he calls "intellectual linking" (ibid. 119). What, however, is the force, or evidence, for such intellectual links? Why, in cases of correct deductive reasoning, are we compelled to acknowledge the soundness of the argument? Here we find an interesting resemblance to the semiotics of C.S. Peirce, because linking, in Johnson's sense, crosses the boundary between indexical signs that are related to their objects by contiguity (the fingerprints physically links a person to the site of the crime) and iconic signs that link sign and object by virtue of their similarity. According to Peirce the compelling nature of intellectual linking lies in the fact that it is iconic and more specifically diagrammatic:

...necessary reasoning makes its conclusion Evident. What is this "Evidence"? It consists in the fact that the truth of the conclusion is perceived, in all its generality, and in generality the how and why of the truth is perceived. What sort of a sign can communicate this Evidence? [...] It is true that ordinary icons, -- the only class of signs that remains for necessary inference, -- merely suggests the possibility of that which they represent, being percepts minus the insistency and percussivity of percepts. [...] It is, therefore, a very extraordinary feature of Diagrams that they show, -- as literally show as a Percept shows the Perceptual Judgment to be true, -- that a consequence does follow, and more marvellous yet, that it would follow under all varieties of circumstances accompanying the premises. (Ms, 293, c. 1906: 13-14)
Lakoff has also stressed the possible role of image-schemas in reasoning. He points out that the traditional representation of classical logic by Venn diagrams is a way of using the CONTAINER schema in logic (see Lakoff 1987: 444-461). According to Peirce not only the CONTAINER schema as exemplified in the Venn diagrams, but also the other way of representing the syllogism (all M are P/all P are S//all M are S) are diagrammatic, because the position of the middle term in both premises is shown (see Peirce 3.363).

However, the representation by means of the CONTAINER schema seems to relate to basic ways of perceiving and interacting with the world: Our bodily and procreative functions tell us that, in a very fundamental way we are ourselves containers that need to be filled, and that receive, store, transform, secrete, and discharge what is put in them. This is why Isak Dinesen somewhat ironically lets one of her characters in "The Dreamers," from Seven Gothic Tales (1936), suggest that basically man is nothing but a "minutely set, ingenious machine for turning, with infinite artfulness, the red wine of Shiraz into urine." Furthermore, in addition to filling our own and each others containers, including our mental containers, our minds, we spend much time putting things, again including our own bodies and minds, into and taking them out of containers. We are born from and we live in containers, and when we die we are put into containers; some in well-wrought urns, others in an oak coffin within a leaden coffin within a marble sarcophagus in a mausoleum.

Peirce's icon, index, symbol distinction and Johnson's and Lakoff's work on basic level concepts and image-schemas are so important, because both offer partial explanations of how meaning is possible. Both in Peirce and in cognitive semantics a major objective is to mediate between the senses and perception on the one hand, and perception and cognition on the other hand. Johnson, following Kant, explains that the office of schemas is to accomplish a mediation between images or objects of sensation and concepts. Schemas are able to do this, because they can be "a rule-following or rule-like activity for creating figure or structure in a spatial and temporal representations" (Johnson 1987: 155). Consequently, neither to Peirce nor to Johnson and Lakoff meaning is, and must not be, confined to propositions. In Johnson's words:

it [meaning] permeates our embodied, spatial, temporal, cultural, and value-laden understanding. The structures of imagination are part of what is shared when we understand one another and are able to communicate within a community. (ibid. 172)
Such insights and analytic procedures are the backbone of Reading Minds, since it is Turner's conviction that some neural patterns are inherently meaningful, because they correspond to bodily situations and are related to episodic memory. Brain patterns, here concepts, not directly connected to such bodily situations, become meaningful by being indirectly related to them by virtue of projection (45-6). Turner, very appropriately, analyzes the bodily basis of symmetry (68-98), he shows the central role of symmetry in literature (specifically in verse and poetry). He further speculates that, in addition to being grounded in the bilateral symmetry of our bodies, symmetry may have another foundation in the symmetrical projection from the bundles of neurones from one hemisphere to the other, each hemisphere projecting to the other a copy of the perceived world. Although our default concept of rationality and rational argumentation is that of a conscious and disembodied activity, Turner's analysis of the understanding of argumentation within the general and juridical rhetoric of Cicero shows that it is fundamentally spatial and dynamic, since the central concept is stasis. For example, the ending of an argument by the prosecution is understood as the reaching of a provisional stasis. This stasis, however, is disturbed by the rejoinder of the defense, and the whole argument is described as forcing the other party out of his position by advancing counterargument as if it was a physical fight.

Turner goes on to treat the relationship between categories, on the one hand, and analogies and metaphors on the other hand. His position is that there is a difference of grade rather than a difference of kind between them, because our categorizations are dynamic and changeable, and established categories may change precisely because of the force of analogies (e. g. light is a wave). Nevertheless, he recognizes the existence of non-analogical equivalences such as identity, definition, class inclusion, inference, causal statements, legal status, etc., but he also shows that for each of these there is a analogical parallel (Achilles is a man vs. Achilles is a lion). The difference is that analogies bring forth the maximal difference between the terms compared and they would be false if taken literally. He suggests, however, to talk about degrees of generative entrenchment of concepts and connections, because this would allow for the fact that some concepts and connections do seem to be indispensable, while leaving us with a middle ground of well-established, but changeable ones.

From here he goes on to analyze metaphor according to the model established by Lakoff, Johnson, and himself, both on the basic level and on the level of image schemas. Since the analysis of metaphor is changed in The Literary Mind let us just give the basics of the traditional Lakoffean analysis: A metaphor is established by the projection of elements and structures from one conceptual domain, the source domain, onto another, the target domain, to the effect that the concept of the latter is structured by the former (very often the source domain will be physical or at least concrete and well known, while the target domain tend to be mental, abstract, or little understood). Let one of Turner's many examples suffice, namely LIFE is a PLAY:

man is actor
fellow man is co-actor
behavior is character
the others are audience
death is curtains
attention is spotlight
draw attention is stealing the show
wait to act is waiting in the wings
act spontaneously is improvising
The point is that because such a metaphor is established by the projection of a whole conceptual structure it is potentially eternally productive ; there is no end to, and no predictability concerning, the continuous addition of novel variants to the metaphor LIFE IS A PLAY. The rest of Turner's book is a meticulous and very rewarding inquiry into the different forms of metaphors.

Reading Minds is a brilliant exploration of what Jakobson has called the poetic principle, namely the principle of equivalence projected from the axis of selection onto the axis of combination. Broadly speaking metaphors and analogies are definitions by means of categorization (there are other possibilities, e. g. ostensive definition). Theyare, however, special kinds of definitions, because they do not state the identity of definiendum and definiens. Instead they state similarity and/or equivalence between concepts and terms while preserving the categorical difference between the two parts of the comparison.

This way of thinking about metaphor, and poetry in general, goes back at least to Aristotle, and Turner's handling of the so-called xyz constructions [e. g. children are the riches of the poor, i.e., x (children) : y (the poor) :: z (riches : (u (the rich))] is Aristotelian to the core, something that Turner recognizes by quoting the key passage from The Poetics (1457b, Turner 1991: 198). He further stresses the importance of rhetoric, indeed, he characterizes his approach as one of cognitive rhetoric. What Turner's approach offers is not discoveries of new properties of literature or differentiae specificae setting it apart from other kinds of discourse, what he offers is a new perspective, that of cognitive science, and a very close and convincing argument.

Whereas in Reading Minds Turner explores similarities from a paradigmatic point of view, in his latest book, The Literary Mind, he is now extending his research to stories (parables, fables), although he still focuses on the mechanisms of similarity and comparison.

However, before going into his analysis of stories, a brief account should be given of the much needed revision of the theory of metaphor presented in this book on the basis of his collaboration with Gilles Fauconnier, the author of Mental Spaces (1994). In the earlier studies of metaphor by Lakoff and his group, many outsiders, myself included, found the framework and perspective profoundly interesting and important, and the concrete analyses often brilliant, but, at the same time, had to reject their model of metaphor as unacceptable for two reasons: (1) its unidirectionality: the mapping, it was claimed, is only from source to target, and (2) the denial that similarity plays an important role in the discovery or invention of metaphors. In the revised model by Fauconnier and Turner these obstacles are removed. Instead of talking about a projection from source domain to target domain, they talk about projection from input spaces onto a common blended space, something that firstly allows for an input from more than two spaces (thus making it easier to handle complex metaphors), and secondly allows multidirectional influences in the blended space. Furthermore, in addition to input spaces and blended space, they talk about the generic space containing what is common to the input spaces (thus making room for the common features of the input spaces as well as making explicit the abstract structure on which the metaphor is built).

"When the cat's away, the mice will play" is an input space which, like most proverbs, is projected into a blended space in which the other input space will be the concrete situation to which the proverb is applied, for instance, as Turner points out, to "stories of the office, the classroom, infidelity, congressional oversight committees, computer antivirus utilities, and so on, over an unlimited range" (Turner 1996: 87). According to the analysis by Turner and Lakoff (in Lakoff and Turner 1989), this is an instance of the mechanism called "generic is specific", i.e., a concrete input contains an abstract structure, located in the generic space, that can be applied to an indefinite number of inputs in the blended space, because they share this abstract pattern, and consequently, the input spaces will have counterparts that allow their blending (the absent wife will be the cat, the unfaithful husband the mouse, or vice versa (see also LIFE IS A PLAY above). It may also be possible to express the content of the generic space differently from that of the input space, for example, "when the person in control is absent, the other(s) will overstep the bounds regulating status quo." However, it is difficult, to say the least, to give a one-sentence paraphrase of the fundamental ambiguity of the proverb: for instance, is the cat a ferocious predator or is it your friendly neighborhood police officer, and what about the mice? Turner expresses the relation between the original and the revised metaphor model as follows:

When we take our data exclusively from deeply entrenched projections such as LIFE IS A JOURNEY that have a deeply entrenched vocabulary, the generic and blended spaces are less easily noticed, and the projection looks as if it carries positive meaning from one input space (the source) to another (the target). This has lead to the customary model of projection of meaning as direct, one-way, and positive. This is a useful and parsimonious model, but it is adequate only in limiting cases (Turner 1996: 90).
This revision, and complexification, of the metaphor model is interesting. Among other things, it raises the question whether the new model renders the analytic work done according to the original model redundant. Personally, I am convinced that this is not the case. Although it may be advisable to adjust the analysis in certain cases. The revised model is an amelioration that leaves the insights of the previous model intact, but (1) makes it possible to handle more intricate cases, and (2) removes some serious objections that the original model provoked. The fact that the revised model belongs to the family of metaphor models resting on the so-called similarity-comparison view, prevailing from Aristotle to Max Black (see Black in Ortony 1993: 30) -- a view from which Lakoff and Johnson (1980) distanced themselves, indeed opposed -- is of course somewhat ironic, but of minor significance. Even if the work done by Lakoff, Johnson, Turner, Fauconnier, Sweetser and others turns out to lie within this tradition, as I am convinced that it does, this does not belittle its importance, because novelty of perspective and thoroughness of argumentation are the important ingredients in outstanding scholarship, and their research certainly fulfil these criteria.

Beginning a discussion of The Literary Mind with his account of the revised model of metaphor is also appropriate for understanding the extension of Turner's research to stories. The bridge between the study of similarities and that of stories can be shown in the following passage from Aristotle's Rhetoric:

Fables are suitable for public speaking, and they have this advantage that, while it is difficult to find similar things that have really happened in the past, it is easier to invent fables, for they must be invented, like comparisons, if a man is capable of seizing the analogy;...(Aristotle1926: 277, Rhet. II.xx. 7-8, 1394a).
Although Turner does not quote this passage, it seems to me that it expresses a substantial part of what he is doing in his new book, because projection still plays a major role in his argument. According to Turner stories -- which he also calls narrative imagining -- involves both understanding sequences as wholes and parsing their elements and categorizing them as agents, events, objects, actions, etc. It also involves predicting, evaluating and explaining. Stories are based on, and understood by means of a discourse involving intentions, motives, and agency:

When we watch someone sitting down into a chair, we see what physics cannot recognize: an animate agent performing an intentional act involving basic human-scale categories of events like sitting and objects like chair. But physics offers a representation of the world that leaves out agency, motive, intentionality, [...]. The basic elements of physics are not tied to the human scale; sitting and chair are elements of a story but not elements of physics (Turner 1996: 14).
Turner's general description of story structure is sensible, but it does not add to the thorough study of narratology from Propp to Marie-Laure Ryan (which he does not mention). The novelty of his approach consists in the combination of three perspectives on narrative: (1) the crucial role of what he calls minimal spatial stories, (2) his description of the projection of story onto story, and (3) his idea of the projection of story onto language, i.e., story as the basis for grammar.

In accordance with the general position of this school of cognitive science, Turner states that small spatial stories, e.g. the moving of physical objects, have a privileged position within story making. One of the reasons they are privileged is their close connection with neurobiology: "We have a neuro-biological pattern for reaching out and picking up. This pattern underlies an individual event of reaching out and picking something up and helps us create the category reaching out and picking up (ibid. 17). Although this pattern is activated in situations that are always slightly different a "skeletal complex image schema of dynamic interaction" is built up. Such image schemas of small spatial stories are then projected onto small non-spatial stories (you throw away an empty can, but you also throw away an opportunity). Although Turner finds the claim that all abstract reasoning is always grounded in spatial and bodily stories "too extreme for the available evidence," he contends that we may comfortably say that "for many abstract concepts, the spatial and bodily instances are the archetypes." Further, since models are beginning to be developed of how the brain "might develop both perceptual and conceptual categories of spatial and bodily stories," it is plausible, Turner argues, "that our understanding of social, mental, and abstract domains is formed on our understanding of spatial and bodily stories: But plausibility is the most we can assert on this evidence" (ibid. 51). It is seen how consistent the approach of this branch of cognitive science is: in some form or another this argument is repeated again and again from Metaphors We Live By (1980) onwards.

This is one basic form of projection, the other is the one Aristotle mentions when he recommends to use fables in public speaking. In Aristotle's division of rhetorical devices, both fable and comparison are species of the example. His classification can be sketched in this diagram:


Aristotle explicitly mentions comparison and fable as the two species of the inventedexample (or rhetorical induction). The reason is that they are both used as an argument to make probable that something was or will be the case. The point is to "find similar things," namely similar to the case under discussion, the case in which the orator and the parties of the dialogue are involved. As concern comparison and fable, there is no reference to a historical past but an invention of a parallel case. Comparisons are examples constructed by changing the variables but preserving the relationship, as when Socrates says that magistrates should not be chosen by lot, because this "would be the same as choosing representative athletes not those competent to contend, but those on whom the lot falls" (ibid., this parallelism can be constructed as a four-term analogy: magistrates : government :: athletes : sports). Comparisons indicate a parallelism between two realms of experience, but without creating an independent fictional universe.
The other possibility is the fable. Here a full-fledged alternative, i.e., a fictional universe is created (in Aristotle's case the reference is to animal fables). This is the possibility explored by Turner under the name of parable, i.e., the understanding of "[f]orms of social and psychological causation are understood by projection from bodily causation that involves physical forces. This is parable" (ibid. 18). The parabolic mechanism, then, is the projection of one kind of story onto another. Turner has high hopes for this mechanism, since he claims that it may explain the general story about human thinking, the story of how we know: "The general story is that human beings construct small spatial stories and project them parabolically" (ibid. 15). The first example, which is also the content of the first chapter of the book, is, by the way, an animal fable, by means of which the vizier, her father, attempts to persuade Sharazad not to marry the king, because of the eminent danger of her being decapitated the next morning. He further shows that very many stories of The Arabian Nights are parabolically related to the frame, the curing of the king of the nasty habit of executing wives.

The third way in which minimal spatial story is active is, according to Turner, in the development of language. He rejects the idea that human language depends on a special neurological module of grammar (Chomsky's position). Instead he suggests the following scenario for the development of language: (1) minimal stories, (2) parabolic projection, and (3) grammatical structure. Story structure, he points out, implies distinction, categorization, combination, and hierarchy, precisely operations that characterize grammar. He further speculates that story structure is projected onto sound, and that the articulation of sound is instrumental in creating grammar.

Once grammar is developed, it interacts with the conceptual structure of spatial story to create a generic space containing abstract structures, and abstract structures make non-spatial stories possible. Thus it is the grammatical identity of "Mary threw the stone out of the window" and "Mary threw the job out of the window" that makes the latter possible (by the way, Peirce analyzes the basic identity of such propositions similarly, namely as that of a trivalent rheme: -- throws -- out of --). Turner also argues for his position by analyzing the mental spaces created by speakers in relating stories (moment of speaking, viewpoint and focus) and by a comparison with the linguistic analysis of tenses.

In addition to the genealogical relationship between storytelling, parabolic projection, and communicating by language, Turner points out a very important functional feature: they are all used to transcend the singularity of our individual experiences and our singular lives. And thus we are back where we started: projection as the prime instrument of understanding not only in the individual, but among individuals.

The Literary Mind is an extremely ambitious book that from the vantage point of literature undertakes to explain the basic workings of the human mind. Obviously, the account of the origin of language it offers is speculative -- Turner is well aware of this -- but is nevertheless interesting and has a certain plausibility. Its analysis of the uses of projection in literature belongs to the venerable tradition of rhetoric, and is a further step towards the redemption of the idea of a cognitive rhetoric advanced in Reading Minds. In general it is a very rewarding and tightly argued book. There are, however, two points that invite discussion.

The first point concerns what, in my opinion, is an inadmissible clash of different discourses. On the outset, it should be stated, however, what is not the issue. I certainly do not object to the hypothesis that the mind is dependent on the brain in the sense that certain neural processes are necessary for both the conscious and the unconscious working of the mind. Neither do I object to the hypothesis that the individuality of a person depends upon the specificity of the processes in her or his brain. There seem to be overwhelming indications that this is so from changes in how the mind works and the make-up of the individual person concomitant with changes in the anatomy and/or function of the brain (e. g. the havoc brain damage is creating in the functioning of the mind). Further it seems that every brain is different from others, because its neural networks are influenced by the way it is stimulated. Nevertheless, it will be remembered that Turner himself is careful to distinguish discourse dealing which narrative structure from that of physics, because the first necessitates reference to human agency, intention, motive, etc., while such concepts of intentional agency are absent in the second one. And in most parts of the book he keeps them apart .

In one crucial place, however, it seems to me that Turner confuses the way processes are described in the discourse of phenomenology and in that of the natural sciences. Again, Turner is well aware of this difference in discourse. He says:

We expect our neurobiology to work at least loosely the way our perception seems to work, and we think (wrongly) that introspection reveals at least roughly how our perception works. We expect phenomenology to indicate the nature of neurobiology. But it does not. It appears that there may be no anatomical site in the brain where a perception or a concept horse resides, and, even more interestingly, no points where the parts of the perception or concept are anatomically brought together. The horse looks obviously one thing; yet our visual perception of it is entirely fragmented across the brain. (ibid. 110)
So far so good, but later Turner starts cashing in on the neurobiological analogue (as he calls it), when he opposes a traditional description of the mind which he characterizes as follows:

In its essential lines, it claims that there are certain basic, sober, and literal things themind does; that imaginary and literary acts are parasitic secondary , peripheral, exotic, or deviant; and that when neuroscience gets its act together, we will come to understand that the brain does the thing pretty much in the ways we always expected. [...]
It is possible that this story is wrong to the core. The brain does not seem to behave at all in the ways we expected it to, based on our notion of stable and unitary concepts. On the contrary, our notion of concepts as stable and unitary seems to be a false guide to neurobiology. Blending may seem exotic to us, but in fact it may have a fundamental neurobiological analogue. (ibid. 113)

What I object to as inadmissible is the argument that a certain understanding of the mind may be "wrong to the core," because of the distribution of the brain processes. On this line of reasoning we should also give up our understanding of intentional action as unitary, because attention and motor control are spread over different areas of the brain (see Posner and Raichle 1996). The point is that although it is often, but not always, possible for a neurologist to predict with almost practical certainty which phenomena will follow from certain kind of brain injuries, there is, nevertheless, no way of directly relating phenomenological and neurobiological discourse, because, as Turner makes explicit himself, certain crucial concepts within both discourses cannot be translated into one another. And a further point is that we cannot dispense with any of them, if we want to understand human intention and interaction.

In the final analysis, this issue has to do with questions of modelling and representation. The model of mental spaces used to describe blending is an interesting and ingenious way of representing the interaction of mental elements -- concepts or, as I suppose Turner would rather have it, conceptual structures-- and its relation to a model of brain processes is quite indeterminate. Consequently, if the understanding of the mind to which it gives rise is preferable to an idea of the mind resting on the distinction between the literal and the literary, it must be because it explains the mental data better than does the other model, not because of a putative similarity with models of neural activities. It may be that in the distant future we will have no need for two distinct discourses, one of mind and one of brain, but even this is doubtful at present; it seems that we have to keep them apart.

This brings me to my second point. Because, as a corollary of both the description of projection and blending as fundamental operations of the mind and of contemporary brain research, Turner reaches the conclusion that the everyday mind is literary. Indeed there are two ways in which the mind is literary:

Narrative imagining -- story -- is the fundamental instrument of thought. Rational capacities depend upon it. It is our chief means of looking into the future, of predicting, of planning, and of explaining. It is a literary capacity indispensable to human cognition generally. This is the first way in which the mind is essentially literary. (ibid. 5)

This projection of one story upon the other may seem exotic and literary, and it is--but it is also, like a story, a fundamental instrument of mind. Rational capacities depend upon it. It is a literary capacity indispensable to humancognition generally. This is the second way in which the human mind is essentially literary (ibid.).
Obviously, one may choose to call the mind literary for the reasons given above. I find, however, that in doing so one runs the risk of confusing some very important issues. And the following paragraph seems to testify to the fact that Turner is somewhat aware of this:

Although literary texts may be special, the instruments of thought used to invent them are basic to everyday thought. Written works called narratives or stories may be shelved in a special section of the bookstore, but the mental instrument I call narrative or story is basic to human thinking. Literary works known as parables may reside within fiction, but the mental instrument I call parable has the widest utility in the everyday mind (ibid. 7).
The reason why I object to stressing that the mind is literary is that narrative, or narrative imagining, is certainly not confined to literature, and neither is the mechanism of projecting story onto story. In writing what is supposed to be a true history of,say, the election campaign of an American president, narrative imagining and narrative structure are certainly used, but it will end up in another section in the bookstore. And the fascinating story of the blending of narrative spaces in the strange boat "race" between the Northern Light, actually sailing in 1853, and Great America II, sailing the same distance in 1993, of which Turner gives a fascinating account, is a projection of stories into a blended space, but it has nothing to do with literature.

In my opinion, Turner's attempt to turn topsy-turvy the distinction between literal and literary is adverse to a clarification of some very important points concerning literature in its historical and institutional context. Let us, however, first look at the issue from the point of view of ontogenesis. In most accounts of the psychological development of the child from psychoanalysis onwards it is hypothesized that the distinctions between perception and mental imagining, between outside and inside, between the personal body and other bodies and objects, and between self and others are developed during the early years of life (although different psychologists give different chronologies). An extremely important part of this process is the ability to test the reality of the phenomena appearing to the mind: are they perceptions or figments of imagination? are they just frightening night-mares or do they present an outside danger from which we should flee? This ability to test the ontological status of a phenomenon has of course a very high survival value, without it we would not be able to manage on our own: after all, people suffering from serious mental illnesses must be taken care of. However, if these distinctions are correlative and slowly built up in the individual, then, from an ontogenetic point of view, it makes little sense to claim that the mind is literary. The emerging mind is neither literary nor literal, because the literal/literary distinction is the outcome of a development. Narrative imagining and blending may be general operations of the mind instrumental in building up this distinction, but certainly not placed on the literary side of it.

Furthermore, the fact alone that story cannot be confined to the literary, means that we need more distinctions than the one between literary and literal. In fact, most often the distinction is not drawn between the literal and the literary, but between the literal andthe figurative, for example, the distinction between sensus litteralis and sensus allegoricus used in the exegesis of the Bible.

This distinction is different from the distinction between the fictional and the non-fictional, the "factional". Using one of Turner's examples, we can illustrate the difference as follows: "Is it really true that Mary threw the job out of the window?" (1) "No, I have just been teasing you, she started working there last week." (2) "No, she threw it in the waste bin." In the first case, the utterer pretends that something is the case. He has created a counterfactual world, and he is very well able to distinguish this world from the actual state of affairs. In the second case, he is improving on the figurative expression. And if Mary did start working last week, then both "Mary threw the job out of the window" and "Mary threw it in the waste bin" would be false; and the truth and falsity, respectively, of these statement is not effected by their metaphorical nature. One could of course also ask "Did she really threw the job in the waste bin?" Then one sensible answer could be, "Well, because she acted nervously, she let go of the opportunity." This answer implies that she might not intentionally have let the opportunity slip through her hands, but she had not been able to give the impression of a cool, career-minded person. "let go of the opportunity" is just as figurative as "throw something (non-trash) in the waste bin," but it has a different meaning, because it rather implies passivity than activity or agency.

Furthermore, although statements using figurative language and figurative conceptual structures may be just as univocal as literal statements, we nevertheless need to distinguish between them, because the different mechanisms of parsing and categorizing depend upon such a distinction. If somebody says "Albert is a rat" it is important to know whether it is actually meant that a specific creature is a rat and not a common hamster, or that some person is disliked by the speaker for his rat-like (whatever that means) appearance and/or behavior (most often this question is decided in advance by the context).

However, neither the case when "Mary threw the job out of the window" is said in spite of the fact that she got it, nor when the person Albert is called a rat are instances of literature. They are instances of the (speech) act of misleading, or just kidding, and of using figurative language; and we do these things all the time in our everyday life. Thus, they are not sufficient, and not even necessary preconditions for literature. They are not sufficient, because they neither distinguish literature from other forms of pretending, or from lying, nor from the everyday use of figurative expressions. They are not necessary, because a work of literature need not use figurative language, and because, for instance a poem commemorating the death of a person may establish no universe apart from the actual world. These are, of course, limiting cases; most often literature does use figurative language and figurative conceptual structures (it may even have a particularly high density of them), and/or most often it does create a fictional universe. This is why Turner's distinction between the literary and the literal, and his characterization of the mind as literary obscure the question of the status of literature. This question cannot be answered exclusively within the framework of cognitive rhetoric, because it is bound up with historical and institutional issues that are (not yet?) integrated in this approach. In a Goodmanean way, one could say that the question of "what is literature?" can hardly be asked separately from "when is literature?" (and apart from "why is literature?"). Here, a bare mentioning of two examples will suffice: The Homeric epics have certainly not originally been considered as literature in our sense of the concept. Indeed, the concept of literature itself is the outcome of a very complicated historical development, and it has been, and still is, a category with a (changing) radial structure and fuzzy boundaries. Especially the boundary between the telling of fictional stories and the relating of the historical past is fluctuating, among other things, because the rules of history writing have been changing (Thucydides thought it was fine to write the speeches of the historical characters himself). From the 6th century BC and onwards, there was a growing opposition to viewing the epics as relating reliable information neither about the historical past nor about the gods, and slowly, in a non-synchronic fashion, their status changed within the different parts of society. Furthermore, this borderline is still moved back and forth today in discussions of documentaries and such a curious concept as the "non fiction novel" coined by Truman Capote in connection with his book In Cold Blood.

Consequently, the way Turner distinguishes between the literary and the literal seems to be neither warranted from the point of view of ontogenesis, nor from that of the social and institutional history of literature. And, in my opinion, the analogic argument from neuroscience is spurious as well, because even if it is sensible to believe that neural processes are an important part of what causes the mind to work (together with input from the outside world and from the body), models of mind and models of brain processes do not yet converge, indeed they are not even yet isomorphic.

However, although I consider these objections as serious ones, they only affect his work on literature on a certain level. The level that is concerned with the admissibility or inadmissibility of making generalizing inferences from analyses of features of storytelling and projection. The analyses, however, are themselves ingenious and well-founded. Consequently, Turner's work must be highly recommended. His approach to literature is a very balanced and successful amalgamation of traditional literary and rhetorical scholarship and an inventive way of combining this kind of study with novel perspectives. Further, it is a welcome antidote to today's fad of theorizing for theory's sake. In Turner, on the contrary, we are always close to the mechanisms involved in actual text production, he is concerned about showing us how texts work.

It is common practice to tell what is not (yet) achieved by the approach under review, i.e., the sins of omission, and I have myself indicated that an idea of literature is missing which is sensitive to the fact that it is defined and defines itself differently at different times and places. Indeed, in my understanding, Turner is analyzing general cognitive operations just as much as literature. It should be added, however, that Turner may very well be right, indeed, that there are no structuring principles or critical features that are exclusively literary (this, by the way, was also Roman Jakobson's position). Literature is the result of mental operations that are generally used, but as a kind of discourse it is in modern times, partly for structural partly for functional reasons (that change over time), kept apart from other kinds of discourse.

Anyway, it will be curious to see in which direction Turner will develop his approach.But it may already be safely said that what he has achieved till now is important to the study of literature. Indeed, on the present North American scene it seems to be one of the most promising approaches. And for semioticians his books are particularly interesting, because cognitive semantics in general is so close to semiotics. In fact, I see it as a part of the study of language and literature that, in our century, began with the works of Saussure and Russian Formalism and that, in different forms, has continued to this day. And to the semiotic study of literature they are certainly of special interest as well, because they combine a structural and rhetorical approach with a cognitive one.


Aristotle (1926). Art of Rhetoric (trans. J.H. Reese). London: Heineman (Loeb).

--- (1927). The Poetics (trans. W. Hamilton Fife). London: Heineman (Loeb).

Black, Max (1962) Models and Metaphors. Ithaca: Cornell University Press.

--- 1993, orig 1977). "More about metaphor." I Ortony, A. (ed.) Metaphor and Thought.

Fauconnier, Gilles (1994). Mental Spaces. Aspects of Meaning Construction in Natural Languages. Cambridge/New York: Cambridge University Press.

Goodman, Nelson (1976). Ways of Worldmaking. Indianapolis: Hackett Publishing Co.

Greimas, A. J. (1966). Sémantique structurale. Paris: Larousse.

Johnson, Mark (1987). The Body in the Mind. The Bodily Basis of Meaning, Imagination, and Reason. Chicago: University of Chicago Press.

Lakoff, George (1987). Women, Fire, and Dangerous Things. What Categories Reveal about the Mind. Chicago: University of Chicago Press.

--- (1988) "Cognitive Semantics." In Meaning and Mental Representations (Eco, Santambrogio & Violi, eds.) Bloomington IN.

--- (1993). "The Contemporary Theory of Metaphor." In Metaphor and Thought (2. udg.1993, A. Ortony ed., cf. nedenfor) pp. 202- 251.

Lakoff, G. & Johnson, M. (1980). Metaphors We Live By. Chicago: University of Chicago Press.

Lakoff, G. & Turner, M. (1989). More than Cool Reason. A Field Guide to Poetic Metaphor. Chicago: Chicago University Press.

Master Metaphor List. (1991) Second Edition Compiled by George Lakoff, Jane Esperson, and Alan Schwartz. Cognitive Linguistics Group. University of California at Berkeley. October, 1991. (pp. 217). (database containing a very large systematic list of different kinds of metaphors)Metaphor and Thought. (2. udg. 1993, Ortony, A., ed.) Cambridge: Cambridge University Press.

Posner, M. J. & Raichle, M. E. (1996). Images of Mind. New York: Scientic American Library.

Sweetser, Eve E. (1990). From etymology to pragmatics. Cambridge Studies in Linguistics 54. Cambridge: Cambridge University Press.

Turner, Mark. (1987) Death Is the Mother of Beauty: Mind, Metaphor, Criticism. Chicago: University of Chicago Press.

--- (1991). Reading Minds. The Study of English in the Age of Cognitive Science. Princeton: Princeton University Press.

--- (1996). The Literary Mind. New York/Oxford: Oxford University Press.

Winnicott (1971). Playing and Reality. London: Tavistock Publications.

Jørgen Dines Johansen is professor of general and comparative literature at the Center for Literature and Semiotics, Odense University, Denmark. He is vice president of Sémiologie ddu spectacle; member of the bureau of Association Internationale de litterature comparée (1985-91), and of the Executive Committee of the International Association for Semiotic Studies (IAAS) (1986 ); he is also secretary of the Nordic Association for Semiotic Studies (NASS) (1987-93) and a former director of the North European Regional Center for Semiotics (1989-90). His fields of research are literary theory, semioitcs, and the history of European literature. He has published Novelleteori efter 1945 (Postwar Theories of the Short Story. A study in Literary Taxonomy] (1970), Dansk kortprosa I-II [Danish Short Stories], and Analyser af dansk kortprosa I-II [Analyses of Danish Short Stories] (1970-72), Om fortolkningssituation [On Interpretation] (1972), Novelle og kontekst [Short Story and Context] (with Søren Baggesen, 1972), Psykoanalyse, litterature, tekstteori I-II [Psychoanalysis, Literature, Text theory] (1977), Danish Semiotics (with Morten Nøjgaard, 1979), Communication et sujet (1981), Hvalerne venter [The Whales are Waiting. Studies in the Works of Klaus Rifbjerg] (1981), Litterer verdi og vurdering [Literary Value and Evaluation] (with Erik Nielsen, 1984), Lille psykoanalytisk leksikon [Concise Dictionary of Psychoanalysis] (1986),Dialogic Semiosis (1993), and Tegn i brug [Signs in act] (with Svend Erik Larsen, 1994).

Go to Semiotic Review of Books Home Page
Go to SRB Highlights
Go to SRB Archives