Go to Cyber Semiotic Institute Home Page
Go to Course Outline
Go to Lecture 2

A Course in Pictorial and Multimodal Metaphor

Instructor: Charles Forceville

Lecture 1: Preliminary Concepts and Terminology


A number of basic concepts and terms employed in discussions of metaphor will first be explained. The framework used is Max Black's (1979/1993) interaction theory, which will be com­ple­mented by other rele­vant studies. After the introduction of the pertinent concepts, they will be applied to some de-con­tex­tualized pictures in order to demonstrate how Black's theory can be made productive for non-verbal metaphors.

The interaction theory of metaphor is primarily associated with the work of I.A. Richards (1935/1965), Max Black (1962, 1979), and Paul Ricoeur (1977). A variant of the interaction theory underlies much current work on metaphor, including Lakoff and Johnson's influential Metaphors we live by (1980) and the Cognitivist Linguistics approach to metaphor that was anticipated by this study. Since Black's formulation of the interaction theory is the most lucid and applicable, his approach will provide the starting point for this course. A more detailed analysis of Black's model can be found in Forceville (1996: chapter 2; see also Forceville 1995); here I will omit the niceties and con­centrate on the essentials. Where pertinent I will replace Black's (1979) terms by ter­minology which nowadays has gained greater currency. I should em­phasize that while I see the following elaborations as being in the spirit of Black's ideas, they necessarily go beyond what Black could envisage 25 years ago. Hence the full respon­sibility for my interpretation of Black’s approach resides with me.


The interaction theory: target and source, target domain and source domain

A metaphor consists of two elements. One of these terms is that about which something else is said. It is the “topic” or “tenor” (Richards 1936/1965), the “primary subject” (Black 1962, 1979), or the “target” (Lakoff and Johnson 1980, 1999) of the metaphor. Since the term “target” is now commonly used, that will be the word generally employed here as well.

          The other term is the “something else” used to convey some­thing about the target. It is the “ve­hicle” (Richards 1936/1965), the “sub­sidiary subject” (Black, 1962), the “secondary subject” (Black, 1979) or the “source” (Lakoff and John­son 1980). I will use the term “source” here. Thus, in

(1.1) Love is a battle­field

(the title of an eighties' Pat Benatar song), "love" is the target and "battlefield" the source.

          Both target and source, however, are part of a whole network of related mean­ings, meanings that can usually be conveyed by words. Black labeled this network the “system of associated commonplaces” (Black 1962) and the "im­plicative complex" (Black 1979/1993); Kittay called it a "seman­tic field" (Kittay 1987). "Love" is part of a network that in­cludes concepts such as "lovers," "pas­sion," "sex," "ma­rriage," "respect," "sacrifice," “roses,” etc. Such a network, of course, can be infinitely extended. Some of the extensions are connected to "love" in a fairly accepted, conventional way. The conventionality of the connections can be gauged from the likelihood that these words will or could occur in the same textual passage or context in which "love" appears. Connections to other concepts are less self-evident and hence less conventional. Most people probably judge the link between "love" and "hap­piness" to be closer than that between "love" and "pain." But recall Catullus' famous love poem that begins "Odi et amo" ("I hate and I love") or, for that matter, the Nazareth classic “Love hurts,” and you remember it’s not all roses there.

          The source, "battle­field," is similarly part of a network of concepts. This net­work includes "so­ldiers," "vic­tims," "wounds," "pain," "sor­row," "victo­ries," "defeats," etc. But the networks con­sist of more than mere words or concepts and their denota­tions (roughly: objective meanings, as they are found in a dictionary); they also cover these words' connotations (roughly: personal or conventional overtones and emotions associated with the word) and attitudes toward them. Love can be a source of inspiration to a lover, but also a reason for re­venge, in­security, depres­sion, creativity; it can cause a divorce, a war (the Trojan war, for instance, was a result of Paris’ infatuation with Helena), or the conception of a child. Similar reasoning holds for battle­field. The net­work of which target and source of a metaphor are part, then, cannot be adequately described ex­clusively in terms of denotations; it also requires taking into account connotations and pragmatic considerations. The network is thus a category, and as Lakoff 1987 (elaborating the work of Eleanor Rosch) shows, categories are anything but stable, closed, and objec­tive units. A network embodies a wealth of related con­cepts, attitudes, cultural values, be­liefs, poten­tial actions, etc. The two metaphorical terms, target and source, then, are part of networks that will be referred to as the "target do­main" and the "source domain" respectively. Since conceptual domains are conventionally indicated by small capitals to distinguish them from their verbal instantiations (Lakoff and Johnson [1980] initiated this usage), we can say: in the metaphor "Love is a battlefield," "love" is the target, which is part of the target domain love; "battlefield" is the source, which is part of the source domain war. We should not forget, however, that there is no natural or exclusive link between a target and a target domain, or between a source and a source domain. Although it may sound sensible and uncontroversial to say that "love" is part of the domain love, one could also say that it belongs to the domain passion. Similarly, "battlefield" could also be said to belong to the domain battle. There is no objectivity here, and it is all the more important to realize this since the label chosen for the domain to which a target or source is seen to belong may have consequences for the further interpretation of the metaphor. The formulation love is a battlefield focuses primarily on source domain elements such as casualties, fighting strategies, trenches etc., whereas the more encompassing love is war is likely to cue more abstract source domain elements such as politics, powerplay, ally-seeking etc.

Interpreting a metaphor

What happens in a metaphor is that at least one feature typically as­so­ciated with the source (and therefore coming from the source domain) is projected (Black) or mapped (Lakoff and Johnson) onto the target. It is important to note that in order to make a metaphor possible in the first place, there must be some sort of resemblance between the target and the source; it is this similarity that is the basis on which the difference between the two can become productive. For instance, in “Love is a battlefield,” a minimal resemblance between the two domains is that in both two parties are engaged in a type of relationship with one another.

          In “Love is a battlefield” we could postulate that "the goal to hurt and kill the enemy" is a feature mapped from "battle­field" to "love." Other features that qualify for mapping are "making victims," "being costly in terms of lives and material," “being a cause of misery and bitterness.” In this fairly simple meta­phor, only a few features are mapped. It is to be noted that these features are not isolated, but linked to one another. Just as on the bat­tlefield soldiers fight the enemy, often at con­siderable personal and material cost, in order to kill or hurt him, thereby causing victims; just so a person quarrels with her lover, often causing grief to herself, intending to hurt him, thereby causing unhappiness. There is thus usually a structural relationship between a number of elements in the source domain and corresponding elements in the target domain, which enable the mapping of features from source to target. It is often the structural character of the mapping which makes metaphors interesting, insightful and persuasive. But unless the context in which a metaphor appears provides us with details about the features that are to be mapped (which it usually does), the interpreter of the metaphor must decide for herself which of the features are to be mapped. For instance, if I say about somebody “He’s a real George Bush,” it is not at all clear what feature(s) I intend you to map from Bush to the person I am referring to – partly because it is not clear in which context I utter this metaphor, and partly because you do not know how I feel about Mr. Bush.

          We can draw a few important conclusions from these observations:


(1)  By definition, not all features or characteristics of a source are map­pable. In fact, usually only a few are, while the vast major­ity are not. For instance, the fact that modern battles are fought with guns and tanks has no immediate coun­terpart in the target domain of love. Which features are to be mapped depends on many circumstances. Often the surrounding text of a meta­phor gives clues which features should be mapped. In addition, an audience that has specialized knowledge of a metaphorical source domain, or a certain attitude toward it, may come up with mappings unavailable to a general audience. To give an example: Guy Cook, discussing various interpretations of Hopkins’ complex poem “The windhover” (a dialect word meaning “kestrel”), points out that one interpretation of the poem, among various others, depends on the metaphor Christ is windhover. After a more general interpretation, Cook adds a number of relevant aspects of the domain (in Cook’s terminology: “schemata”), concluding, “the availability of this is, however, far less widespread than the other schemata [= “relevant features in the source domain” ChF] I have suggested, as it will only be present in those observers with ornithological knowledge” (Cook 1994: 246).


(2)  Neither the decision which features from the source are to be map­ped, nor how these features are to be la­bel­ed, nor the features in the target with which they are to be matched, are necessarily self-evident. In this sense, the interpretation of each metaphor requires work from its interpreter. Sometimes the context in which a metaphor appears gives the interpreter clues as to which features are to be mapped, but often this is not the case, and it will depend on the interpreter (and her personal experiences, her knowledge, her culture) how far she will go in deciding on the mappable features. The textual genre in which a metaphor occurs, too, plays an important role. A metaphor in a poem will probably motivate a reader to a larger degree to search for possible mappings than a metaphor in a journalist’s news report (for an empirical investigation of a similar hypothesis, see Steen 1994: chapter 6). But since mappable features are not always made explicit, different interpreters may infer (partly) different features, while the interpretation process is open-ended (see Sperber and Wilson 1995/1986; Forceville 1996: chapter 5). This is what makes metaphors both suggestive and risky ways of communication.


(3)   Sometimes a metaphor suggests the mapping of a single feature. When you call your boss an ass, it is very likely that the only feature you intend your interlocutor to map from the source domain ass to the target domain boss is “stupidity.” (If the boss happens to have protruding teeth, often wears grey suits, and/or has a hee-haw laugh, all these aspects may echo as part and parcel of the metaphor – and make the metaphor, perhaps, extra appropriate.)

          But in more interesting metaphors, it is not isolated fea­tures or properties that are mapped, but a number of features, along with the structural relationships that exists between them in the source domain. In As You Like It (II: 7) the character Jacques famously says:

 (1.2) All the world's a stage

Shakespeare himself provides a (partial) explanation of the metaphor by elaborating on the metaphor: "And all the men and women merely players;/ They have their exits and their en­trances;/ And one man in his time plays many parts." The source stage has in its domain such explicitly given ele­ments as “players,” “exits,” “entra­nces,” “many parts” (= roles). In interpreting the metaphor we begin by matching these elements with their corresponding elements in the target domain world: "players" matches with "men and women"; "exits" matches with "moments during which people are not in the centre of attention" or "people's deaths”; "entrances" matches with "people's births" or, perhaps, "moments in their lives during which people attract much attention"; "many parts" matches with "behaving (very) dif­ferently according to the situation in which people find themselves at a particular moment". In addition, what we match are the relations that connect these elements. That is: the metaphor gains in richness depending on the extent to which the structural rela­tions that exist between a stage, a play, players, parts, entrances and exists are retained in the mapping onto the target domain world. Notice that to retain the structural relationships in the mapping from source to target domain once we decide that “exits” matches with “deaths,” we will have to match “entrances” with “births”; whereas if we match “exits” with “moments during which people are not in the centre of attention,” we will have to match “entrances” with “moments in their lives during which people attract much attention.” We cannot “use” exits in two different senses simultaneously.

(4) Metaphors can provide a new or alter­native view of a given target domain by linking it with an “un­­expected” source domain, or by mapping unexplored features from a familiar source domain to the target. This happens because the chosen source domain, which has a certain structure, high­lights a similar, often latent structure, in the target do­main. Different source domains highlight different struc­tures in the target domain. The target domain structure resulting from "The world is a stage" is a very different one from that which emanates from "The world is a battlefield.” Each source domain “highlights” certain features in the target domain and “hides” others (see also Lakoff and Johnson 1980: chapter 3).

    Thus, the Dutch poet Jan van Nijlen wrote a wonderful sonnet called “De cactus,” from which the metaphor the poet is a cactus can be construed. Whereas we probably readily map the feature “prickliness” from cactus to poet (with due adaptation of a literal prickliness to a figurative one) van Nijlen builds up an entire source domain around cactus in the course of his poem, including the notion of the cactus having been transplanted from its natural habitat, the sunny prairies, to the drabness of Dutch window sills and, most pointedly, the fact that a cactus flowers only once every so many years. The correspondent features built up in the target domain are, presumably, that the poet, too, feels “alien” in his everyday surroundings, and derives his sense of joy (or the justification of his existence?) from the rare poem that he manages to produce.

    Inasmuch as rich source domains can provide elaborate structure to target domains, the resulting metaphors may amount to ideological frameworks for individuals or for communities. Gareth Morgan (1986) wrote an insightful book on how companies often (unconsciously) operate according to a specific metaphor. In his book he traces several of such metaphors, like an organization is a machine, an organization is an organism, and amply discusses how these various source domains structure the target domain organization, and how the resulting metaphors enforce or encourage certain actions while discouraging or forbidding other actions. Here each metaphor functions as a model of a company culture, and the different models partly conflict. Such an incompatibility often becomes visible in the case of a merger between two companies. One can imagine how behaviours that are consistent with a company is a machine cannot be simply translated into behaviours that are consistent with a company is an organism. Machines consist of parts, and parts are replaceable. In an organism one cannot so easily isolate a single part and substitute another part for it, for organisms are living tissue.

    New insights or angles on a target domain need not only arise from an unexpected coupling of domains; they may also be a consequence of  mapping ignored features of a familiar domain. Lakoff and Johnson (1980) give some examples in the course of the discussion of the structural metaphor theories are buildings. Some elements from the building domain are so frequently used as to have become clichés – indeed, the very fact of their having become clichés proves how deeply entrenched the metaphor has become in our folk theory of theories. Examples are: “the theory needs more support”; “so far we have put together only the framework of the theory”; “we need to buttress the theory with solid arguments.” But the fact that we have the theories are building metaphor integrated in our conceptualization of the world, also allows us, Lakoff and Johnson point out, to interpret novel extensions of this metaphor, such as “his theory has thousands of little rooms and long, winding corridors” and “he prefers massive Gothic theories covered with gargoyles” (1980: 52)


(5) An interesting phenomenon, and in fact a source of creativity, is that a well-chosen metaphor results in the highlighting of features that are not salient in either the target or the source domain. Gineste, Indurkhya, and Scart (2000) showed that participants in the experiment, when confronted with, for example, “her gaze, a flash of diamond” (exemplifying the metaphor gaze is diamond’s reflection) suggested mapped features that each of the two terms separately did not elicit. When, in this example, the target was presented in isolation, it evoked such features as “view” and “deep”; when the source was presented in isolation, it evoked such features as “wealth” and “brilliant”; but when the target and source were metaphorically coupled, the resulting metaphor yielded new features, such as “seduction” and “sharp.” Meta­phors thus can elicit what Gineste et al. call "emergent proper­ties." Even more exciting is that if the target domain has little or no structure, a meta­phor­ical coupling of this target with a richly-structured source domain can even import or impose structure on the target (Indurkhya 1991; see also Gick and Holyoak, 1980, Schön 1979).


Black’s (1979) “Stars of David”

Black elucidated his view of verbal metaphor (with hints toward cognitive metaphor) by some bril­liantly simple observations on visual representations of the "Star of David." Have a look at the “neutral” version in figure 1. One can think of this star as a single, monolithic “gestalt,” but it is also possible to “see” the star in different ways. Probably the alternative conceptualization coming to mind most easily is that of two superimposed triangles (figure 1a). Try to think of yet other conceptualizations. (Here are some examples, but the series is by no means exhaustive: figure 1b, figure 1c, figure 1d, figure 1e, after Black 1979: 32-34). The surface structure is the same in each case but the way we "see" the star is influenced by how we conceive it. Black’s illustration of “seeing-as” (e.g. seeing a star as two superimposed triangles) is particularly appropriate for present purposes, since it nicely demonstrates that perception (like reading) is concept-driven. Ana­logical to the different perceptions-cum-conceptualizations of the star, Black argued, a target domain "changes" in various ways depending on the source domain with which it is metaphorically coupled.

Denotations and connotations of depicted objects

A concept or object can be represented via a word, but in many cases also via a picture. Think of the concept sword. Now let us inventory some of the meanings that attach to it; that is, let us build up its semantic-pragmatic domain. A sword is a potentially lethal stabbing weapon associated with pre-modern times, in Western Europe primarily with medieval knights. Because of this, it has “noble” connotations. It has a certain (crucifixal) form, weight, and texture; it is worn on the body, in some sort of sheath; it is used in man-to-man combat. The connotations do not stop here: in some situations it is an indispensable element in a ceremonial outfit; in a Freudian context, it has phallic overtones, etc.

          In principle all these connotations could be evoked not only by the word “sword” but also by the picture of a sword – although the connotations will probably not be exactly the same for any two people. However, a picture cannot but depict a specific sword, and that specific sword, even if it is decontextualized, does not only share connotations with other decontextualized swords but also has connotations differing from them.

          Look at figure 2a. This, I propose, comes very close to what, in the Western world (but not necessarily elsewhere), is considered a prototypical sword. Note that there is only a minimal context in the form of an unspecified dark blue background, but already this minimal context begins to bestow, or strengthen some latent connotations on the sword – of value, of grandeur, of visual pleasure, perhaps of something museum-collection-ish. (Notice that street-wise fleamarket vendors often display their wares in a similar fashion for precisely this reason!) Compare this to figure 2b. Again, this looks like a rather prototypical sword. But the very fact that this is not a photograph but a drawn picture of a sword means that the connotations that, I proposed, cling to the sword in 2a do not adhere to the sword in 2b. A more prominent connotation here is arguably that it is an item in a list or catalogue. Certainly any aesthetic qualities a sword may have are downplayed here. But in comparison with the sword in 2a, the sword in 2b also looks less lethal, probably because in 2a the metal, enhanced by the reflection, evokes this quality more strongly than the sketched sword in 2b. While we should be careful not to attribute the presence or absence of these connotations to the difference between a photograph or a picture per se, the difference in medium affects us in a manner that should alert us to its potential effect on the activation of connotations – or its failure to do so.

          The next representation of a sword, figure 2c, is similar to that in 2a in being once more a photographic representation, but different in the sense that it is a different kind of sword. Indeed, such a curved sword is usually called a sabre. Although it is as beautiful and lethal as the sword in 2a, it has a more oriental look. Moreover, while we expect the sword in 2a in Crusaders’ hands, we tend to associate this sword more strongly with pirates. While these latter connotations are by and large conveyed by the sabre in figure 2d, too, the connotation “lethalness” is here reduced because we recognize that this is a child’s toy, made of foam rather than metal. So “childishness,” “cuteness,” as well as “colourfulness” are among the potential associations activated here.

          Finally, take a look at figure 2e. Here we see two objects that only Star Wars watchers recognize as swords, namely laser swords. One of them is green, and the aficionados know that it must be the sword of one of the heroes, Luke Skywalker. Hence connotations here comprise, thanks to the metonymic link, Luke himself, but by extension the whole Star War universe. Moreover, the representation of the swords is neither a photograph nor a plain drawing: it is a computer-generated image. This mode of representation itself potentially adduces connotations to the represented swords – for instance of artificiality or futuristicness.

          Here is another example: a series of visual representations of feathers. While feathers prototypically connote birds (that is what they metonymically refer to) and softness, individual feathers may elicit specific associations. The feather in figure 3a strikes the viewer as very ordinary – but note that “ordinariness” is in itself a potential connotation. In addition, it is a drawn feather and, more than the other feathers, it can be said to connote writing, since it seems to be the kind of feather one could sharpen and dip in ink. Figure 3b shows a feather that is less straight, and connotes fluffiness and possibly whiteness; most certainly the “writing” dimension here is non-existent. The most eye-catching characteristics of the feather in figure 3c is that it is coloured, with yellow and pink the dominant colours. Possibly “artificiality” is a connotation that adheres to it for many people. The feather in figure 3d, finally, is familiar to most viewers as belonging to a peacock. Conventionally, the feature “beauty” is attributed to (male) peacocks’ feathers. Note, however that although it is far less salient, the feature of “incompleteness” can also be detected in this representation. In a specific context this feature can be made relevant and hence salient.


A first approximation of “pictorial metaphor”

In the previous section I have briefly discussed two concepts, sword and feather, and some pictorial manifestations of each of them. While the denotation in the various sword-pictures and feather-pictures remains fairly stable (the denotation being the meaning that dictionaries give, say, “a weapon with a long blade, and a handle at one end” and “one of the very light things that form the covering on a bird’s body; a feather consists of lots of sort or smooth hairs on each side of a thin stiff centre,” respectively – both from the Collins Cobuild English Language Dictionary, 1987), the connotations, as we have seen above, to some extent vary from one representation to another. Since it is often connotations  rather than denotations that are mapped from a source to a target domain in a metaphor, it can matter a lot what representation of a sword or feather one chooses as source domain in a pictorial or multimodal metaphor. Thus “ornateness” and “lethalness” are strong connotations of the sword in figure 2a, but not of the one in figure 2d; while “beauty” conventionally attaches to the feather in figure 3d, it is not a connotation of the one in figure 3a.

          Now with five swords and four feathers we could start making metaphors although, because of the absence of a context, they would necessarily be highly unnatural, artificial ones. Even without a context, it is to be noted, there is a similarity between swords and feathers that serves as basis for a metaphor – we have seen, after all, that metaphorical tension between target and source can arise only thanks to some degree of similarity between the two. The similarity here is the more or less elongated form of both swords and feathers.

          Let us say we put the sword in figure 2a in the “target” slot, and the feather in figure 3b in the “source” slot so as to construe the metaphor sword 2a is feather 3b. The fluffiness of feather 3b being a salient feature, the metaphor could be interpreted to mean something like “the sword is ineffective/blunt/cute … as a feather.” If we were to construe the metaphor sword 2a is feather 3d, by contrast, the interpretation would more likely be “the sword is beautiful (in the manner that a peacock’s feather is beautiful).”

          We can also construe a metaphor with a feather in target position and a sword in source position. We could reverse the metaphor discussed above into feather 3d is sword 2a. Now the “lethalness” is mapped onto the (beautiful) feather, for instance if the feather-as-sword appeared on the hat of a lady wearing the hat to intimidate her rivals at a party – where of course the notion of literal “lethalness”  has to be somewhat adapted for the target domain, since the feather is not literally but figuratively meant to “kill” (think of the expression “dressed to kill”). In the case of feather 3d is sword 2d, on the other hand, the childishness, playfulness or perhaps ridiculousness is mapped from source to target.

          Of course what we have been doing in the two preceding paragraphs is like practicing swimming movements out of the water. We will not get a sense of the real thing until we perform the movements in the proper context. We will very seldom come across the kind of decontextualized examples discussed above  in “real life” (an exception may be scientific test environments: e.g., Kogan et al. [1980] used pairs of pictures to test children’s abilities to see literal and metaphorical resemblance). As a consequence, while examples such as the above are interesting in situations where creativity is an issue – such as in games of the kind “try to think of a context in which the metaphor X IS Y makes sense” – in real-life situations metaphorical source domains occur in a specific, given context, and the metaphors are chosen by their makers to serve a specific goal arising in this context (Forceville 1995).

          As discussed with reference to Gineste et al (2000) above, there is no “natural” way in which features from both domains are to be matched. Which features can be matched will depend on the context in which the metaphor occurs. Thus, the feature “ridiculousness” of the sword in figure 2d may remain latent until the sword is used in a feather 3d is sword 2d metaphor in a context where a lady has a hat with a feather which instead of intimidating her rivals at the party appears rather as ridiculous to them. This example also shows that, depending on the source domain (the sword in figure 2a or the sword in figure 2d) with which a given target domain (the feather in figure 3d) is metaphorically coupled, different features in that target domain are activated – and hence a different perception-cum-cognition of the target domain arises. Conversely, the same source domain (e.g., the sword in figure 2d) can activate different features in different target domains: for instance “cuteness” in the feather of figure 3b and “ridiculousness” in the feather of figure 3d. The point is that different combinations will lead to different “emergent properties” – and here lies an important clue to describing creativity.



Metaphors, in whatever medium, consist of two elements, a target (the topic or subject of the metaphor) and a source (the concept that is used to predicate something about the target). Both the target and the source are part of, in principle, infinitely expanding networks of (true or untrue) facts, connoted meanings, metonymic extensions, attitudes, emotions, etc. For this reason, it is common usage to talk about target and source domains. In each metaphor at least one,  but often more than one, feature from the source domain is mapped onto a corresponding feature in the target domain. In structural metaphors such as time is money (Lakoff and Johnson 1980: 7-9) the relationship has become conventionalized, and the metaphorical nature of the coupling is often no longer realized. In more creative  metaphors, the metaphorical coupling of domains (temporarily) transforms the perspective on the target domain. We are invited (or forced) to see, and think of, a target A in terms of a source B.

          Construing a metaphor requires, firstly, identifying the two pertinent domains involved and, secondly, their slotting as target and source domain, respectively. Thirdly, interpretation of the metaphor entails the identification of one or more features in the source domain that can be mapped onto the target domain. Since metaphors are deliberately conceived as parts of goal-directed (argumentative, persuasive, instructional) representations, the identification of relevant mappable features is usually guided by the local context in which the metaphor occurs as well as by the genre of the representation and, more broadly, the intentions of the author. However, if mappable features are not explicitly mentioned, various factors can influence their selection by members of the audience, such as an individual’s personal knowledge of and attitudes towards a source domain. Moreover, source domains may have very different salient connotations from one (sub)culture to another. This means both that different individuals may interpret the same metaphor slightly or vastly differently, and that metaphors can be understood slightly or vastly differently from how they were intended by their makers.

          Metaphors are central instruments in cognition, and do not only manifest themselves in language but also in pictures and sounds. What constitutes a metaphor, however, is partly affected by the medium in which it occurs. In the case of metaphors involving one or two domains that are pictorially represented, such (a) domain(s) is/are inevitably rendered in highly concrete ways, involving specific forms, textures, and colours, all of which may play a role in the mapping. In addition, the manner of representation and the material used (see Kress & Van Leeuwen 1996: chapter 7; Forceville 1999 is a critical discussion of this important book) to render it may influence the construal of the metaphor as well as its interpretation.


For questions, criticisms, and suggestions, e-mail: c.j.forceville@uva.nl



Black, Max (1962). Metaphor. In: M. Black, Models and metaphors. Ithaca, NY: Cornell University Press, 25-47.

Black, Max (1979). More about metaphor. In: Andrew Ortony (ed.), Metaphor and thought, Cambridge: Cambridge University Press,19-43 (second edition: 1993).

Cook, Guy (1994). Discourse and literature. Oxford: Oxford University Press.

Forceville, Charles (1995). (A)symmetry in metaphor: the importance of extended context. Poetics Today, 16:4,  677-708.

Forceville Charles (1996). Pictorial Metaphor in Advertising. London/New York: Routledge. [Paperback published in 1998].

Forceville, Charles (1999). Educating the eye? Kress & Van Leeuwen's Reading images: the grammar of visual design, (1996). Language & Literature, 8:2, 163-78.

Kress, Gunther, and Theo van Leeuwen (1996). Reading images: the grammar of visual design. London/New York: Routledge.

Gick, Mary L. and Keith J. Holyoak (1980). Analogical problem solving. Cognitive Psychology 12, 306-355.

Gineste, Marie-Dominique, Bipin Indurkhya, and Véronique Scart (2000). Emergence of features in metaphor comprehension. Metaphor and Symbol 15: 117-135.

Indurkhya, Bipin (1991). Modes of metaphor. Metaphor and Symbolic Activity 6:1, 1-27.

Kittay, Eva F. (1987). Metaphor: its cognitive force and lin­guis­tic structure. Oxford: Clarendon Press.

Kogan, Nathan, Kathleen Connor, A. Gross and D. Fava (1980). Under­stand­ing visual metaphor: developmental and individual differ­ences. Monographs of the society for research in child devel­opment. Serial no. 183, 45:1.

Lakoff, George (1987). Women, fire and dangerous things: what cat­e­gories reveal about the mind. Chicago: University of Chicago Press.

Lakoff, George, and Mark Johnson (1980). Metaphors we live by. Chicago: University of Chicago Press.

Lakoff, George, and Mark Johnson (1999). Philosophy in the flesh:the embodied mind and its challenge to Western thought. New York: Basic Books.

Morgan, Gareth (1986). Images of organization. Beverly Hills: Sage.

Richards, I.A. (1965/1936). The philosophy of rhetoric. New York: Oxford University Press.

Ricoeur, Paul (1977). The rule of metaphor: multi-disciplinary studies of the creation of meaning in language. Trans. by R. Czerny, Kathleen McLaughlin and John Costello. Toronto: University of Toronto Press.

Schön, Donald (1979). Generative metaphor: a perspective on problem-setting in social policy. In: Andrew Ortony (ed.), Metaphor and thought, Cambridge: Cambridge University Press, 254-283 (second edition: 1993).

Sperber, Dan, and Deirdre Wilson (1995). Relevance theory: communication and cognition. Oxford: Blackwell. [First edition published 1986.]

Steen, Gerard J. (1994). Understanding metaphor in litera­ture: an empirical approach. London: Longman.


Send comments or questions to Charles Forceville: c.j.forceville@uva.nl

copyright 2004, Charles Forceville.

Go to Lecture 2

Go to Course Outline
Go to Cyber Semiotic Institute Home Page
Go to Semiotic Review of Books Home Page