It. Metafora visiva; Fr. Métaphore visuelle; Germ. Visuelle Metapher; Span. Metáfora visual. A visual metaphor is a type of metaphor the component elements of which (generally designated topic and vehicle, or target and source) are rendered in visual terms. No unified theory of visual metaphors exists, both because scholars from various disciplines (art historians, philosophers, semiologists, linguists, psychologists, cognitive scientists, etc.) are involved in analyzing them, and because their definition varies based on the more general concept of metaphors of which it is a part. There are, however, several recurring questions: how is a visual metaphor identified? Can we retrace the semiotic-rhetorical strategies or semantic-cognitive processes that are common to both verbal and visual metaphors? Are visual metaphors a specific phenomenon or can they be reduced to linguistic metaphors? Such questions have already received some answers from literature now considered consistent, that – even before pictorial or iconic turns – investigated the rules of the construction and interpretation of visual metaphors in a variety of areas, including art and advertising.
Art history has always posed questions about the symbolic meaning of artistic images, that is the means through which they convey often complex and abstract concepts in non-literal, indirect, and allusive ways. Ernst Gombrich (1963) was one of the first to explicitly address the question of visual metaphors, analyzing the transfer processes based on which the forms of art metaphorically represented or evoked moral values. In the subsequent years, following the increasing popularity of a new concept of metaphor that highlighted their the more cognitive than ornamental value, studies of the metaphorical devices present in works of art multiplied. Virgil Aldrich (1968) applied Max Black’s interaction view (1962) to pictorial art, observing that visual metaphors were the result of the “interanimation” of two visual images (A and B), whose colors, forms, or positions cause us to link them visually into a single metaphorical unit (C); in this way, visual metaphors invite interpreters to see one thing as another thing with which the first is not ordinarily associated (see also Hausman 1989). During the same era, although from a different perspective, Nelson Goodman (1968) insisted on the metaphorical value of artistic expression: if I say that a gray painting is sad, I apply a label usual reserved for sentient beings to an inanimate object, creating a cognitively relevant categorial transgression.
The interaction view also influenced the attempts to respond to a basic question: given the non-propositional nature of images, how can we identify in them the difference between the literal and the metaphorical? Noël Carroll (1994) defines the recognition of visual metaphors in terms of homospatiality (the simultaneous presence of two or more elements in the same image) and noncompossibility (the physical impossibility that would, in real life, prevent those objects from sharing the same space). According to Carroll, where verbal metaphors use the verb “to be” to demonstrate identity (“A is B”), visual metaphors use homospatiality; where verbal metaphors exhibit obvious falsity, visual metaphors reveal noncompossibility. Analogously, John M. Kennedy (1982, 2008) sustains that the interpreter can identify a visual metaphor only by retracing the image of an incongruence that transgresses the usual visual code. However, according to Kennedy it is necessary to escape the temptation to consider all semantic transfers to be metaphors and to undertake an analysis and classification that makes it possible to map the wide spectrum of rhetorical figures that, especially in artistic images, coexist or intertwine with visual metaphors.
Initially, visual metaphors were studied in the context of the rhetoric of images, beginning with Roland Barthes’ semiological analysis of mass media products (comics, cinema, advertising). In his famous analysis of a poster advertising Panzani pasta, Barthes (1964) identifies both an iconic, non-codified and purely denotative message in the literal meaning of the image (where packages of pasta, tomatoes, peppers, and onions, etc. are visible) and an iconic, codified, connotative message which constitutes a second-level meaning that depends on cultural conventions (for example, in French culture of the time, tomatoes and peppers evoked the idea of “Italian-ness”). The image therefore has its own rhetoric, possessing both specificity and generality at the same time: specificity to the extent that it is subject to the physical limitations of the image; generality to the extent that its rhetorical figures coincide with those (such as metaphor and metonymy) already identified by the ancient rhetoric of verbal communication. In continuity with this program, Umberto Eco’s interpretative semiotics states that we can speak of visual metaphors because metaphors are defined by a type of semantic device: the exchange of two “sememes” (all of the properties that circumscribe the content of an expression) that share certain “semes” (the content traits that we associate with a given expression). It goes without saying that the semes shared by the metaphorizing term (that is, the vehicle) and the term being metaphorized (the topic) can vary in nature, and therefore can also be visual (Eco 1979).
However, perhaps the most systematic attempt to process a visual rhetoric is attributed to Group μ (an interdisciplinary group of scholars which formed in 1967 at the University of Liege and signed its works with this name). According to Group μ (1992) the same fundamental operations of general rhetoric can be found in the visual equivalents of metaphors: operations of addition, suppression, permutation. It is essential, however, to grasp the particular rhetorical effects of visual language, in which we can distinguish iconic signs (which represent things or people) and plastic signs (which instead represent shapes, colors, or spatial relationships). It then becomes possible to identify visual metaphors in the iconic, plastic, and iconic-plastic fields. The famous heads of the 16th-century painter Giuseppe Arcimboldo, in which various parts of the face are substituted by fruit and vegetables but are still recognizable as heads, imply a typically metaphorical operation of suppression-addition: the nose is replaced by a pear based on the properties common to both. The image of the cat-coffeepot, which appears on the Chat Noir poster created by Julian Key in 1966, also reflects the same operation of suppression-addition; the difference, compared to the previous example, is that it involves a reversible figure (this has no equivalent in verbal metaphors, which always have a precise directionality), so it is difficult to determine whether we are looking at a cat’s tail or the spout of a coffeepot. In an abstract painting by Victor Vasarely, Betelgeuse (1963), on the other hand, we can see an example of metaphorical process in the plastic domain. Here, the rhetorical transformations are geometrical-topological in nature and concern the family of shapes: in it, there are lines of perfect circles, both horizontally and vertically; at the intersection of 14 and 22 circles, where we would expect there to be another circle, there is a square. Finally, in a print by the Japanese painter and engraver Hokusai – The Great Wave, part of the collection Thirty-six views of Mount Fuji, which the artist created from 1826 to 1833 – the relationship between iconic signs and plastic signs show us a visual metaphor with a reversible structure: Fuji, the motionless wave; The wave, a mountain in motion. After all, even in iconic signs, visual metaphors are often signaled by plastic elements (formal similarities or chromatic affinities; typological organization: for example, the proximity or superimposition of two objects in a single space). While in verbal metaphors the characteristics in common between the topic and the vehicle only concern the content, in visual metaphors they also concern expression (Sonesson 2015; Trifonas 2020).
The most recent semiotic research has further investigated the expressive mechanisms used to build visual metaphors in advertising as well as in illustrated books, comics, or cartoons (for example, see Tsakona 2009; Platonova 2019). A social semiotic approach based on the visual grammar of Kress & Van Leeuwen (2006) has developed, analyzing the correlations between the compositional meaning structures of visual metaphors and their situational and cultural context (Feng & O’Halloran 2013).
Starting with Lakoff and Johnson’s work (1980), a cognitivist approach was established and discerned a conceptual phenomenon in metaphors: a projection from a source domain to a target domain, through which we understand one kind of thing in terms of another more familiar and/or concrete thing. This approach, in the past few decades, has dominated theoretical and empirical research on metaphors, and pushed us to further investigate the question of visual metaphors. As Charles Forceville (2016) notes, if metaphors are before all else a matter of thought, it is reasonable to hypothesize that the same conceptual structures are expressed in verbal language as well as in static and dynamic images (and in music, gestures, etc.). However, the comprehension of visual metaphors involves problems specific to them. In verbal metaphors that follow the structure “A is B” the existence of grammatical and syntactic rules makes it possible to distinguish the entity acting as the subject (the topic) from the entity acting as the predicate (the vehicle). In visual metaphors, on the other hand, the situation seems much less linear and cross-domain mappings are identified in other ways. Forceville, who, like the other exponents of this line of thought primarily studied advertising images, identified four types of visual metaphors: contextualized (an object is metaphorized based on the context in which it is located); hybrid (two objects, which normally constitute two distinct entities, become combined into a single Gestalt); based on a relationship of similarity (two objects are represented in such a way that they appear similar); and integrated (an object or shape is globally represented, even in the absence of context clues, as if it resembles another object or shape). Multimodal metaphors also exist, in which the target and source are expressed through at least two different systems of signs (for example, an image and a verbal message). In all of these cases, the steps that lead us to identify something as a metaphor do not change: (1) an identity relation is created between two phenomena that, in the given context, belong to different categories; (2) the phenomena are to be understood as target and source respectively; they are not, in the context, reversible; (3) at least one characteristic/connotation associated with the source domain is to be mapped onto the target domain; often an aligned structure of connotations is to be so mapped (see Forceville 1996).
Gerard Steen, integrating Conceptual Metaphor Theory with his own “Deliberate Metaphor Theory” attributes greater relevance to the communicative aspects of visual metaphors and proposes a partially different model of identification, called VisMIP (Visual Metaphor Identification Procedure). Steen and his collaborators object to the fact that Forceville’s three-step method serves to analyze but not to identify visual metaphors, as the identification is already presupposed in the first step. Their model, on the other hand, starts from the perception of an incongruence, from which follows its resolution through operations of conceptual mapping between source and target, and therefore the recognition of the metaphor and contextual processing (Šorm & Steen 2013, 2018). For example, if an anti-tobacco campaign shows an open packet of cigarettes with bullets inside in place of the (expected) cigarettes, the viewer must mentally replace the bullets with cigarettes in order to restore the expected scenario in an act of cross-domain mapping that attributes the characteristics of the bullets to the cigarettes (cf. Bolognesi 2016). We can also note that VisMIP implies an accentuation of the role played by the viewer, which contributes to building a meaning that is always potentially open and whose results cannot be predicted. We have in fact a visual metaphor when the “reader” of an image is stimulated to project at least one characteristic belonging to or evoked by the source domain onto the target domain. In this way, the spectator must build one (or more) non-reversible “A is B” identity relationships in which at least one of the two domains is expressed or cited through visual means that show perceptive incongruence.
V.C. Aldrich, Visual Metaphor, “Journal of Aesthetic Education”, 2/1 (1968): 73-86.
R. Barthes, Rhétorique de l’image, “Communication”, 4 (1964): 40-51.
M. Black, Models and Metaphors: Studies in Languages and Philosophy, Ithaca (N.Y.), Cornell University Press, 1962.
M. Bolognesi, Similarity is Closeness: Using Distributional Semantic Spaces to Model Similarity in Visual and Linguistic Metaphors, “Corpus Linguistics and Linguistic Theory”, 15/1 (2016): 101-137.
N. Carroll, Visual Metaphor, in N. Carroll, Beyond Aesthetics: Philosophical Essays, Cambridge, Cambridge University Press, 1994: 347-368.
U. Eco, Prospettive di una semiotica delle arti visive, in E. Mucci, P.L. Tazzi (eds.), Teoria e pratiche della critica d’arte, Milano, Feltrinelli, 1979: 69-87.
W. Feng, K. O’Halloran, The Visual Representation of Metaphor. A Social Semiotic Approach, “Review of Cognitive Linguistics”, 11/2 (2013): 320-335.
C. Forceville, Pictorial Metaphor in Advertising, London-New York, Routledge, 1996.
— Pictorial and Multimodal Metaphor, in N.-M. Klug, H. Stöckl (eds.), Handbuch Sprache im multimodalen Kontext, Berlin, Mouton de Gruyter, 2016: 241-260.
E.H. Gombrich, Visual Metaphors of Value in Art, in E.H. Gombrich, Meditations on a Hobby Horse and Other Essays on the Theory of Art, London, Phaidon, 1963: 12-29.
N. Goodman, Languages of Art. An Approach to a Theory of Symbols, Indianapolis, Bobbs-Merrill, 1968.
Groupe μ, Traité du signe visuel. Pour une rhétorique de l’image, Paris, Éditions du Seuil, 1992.
C.R. Hausman, Metaphor and Art: Interactionism and Reference in the Verbal and Nonverbal Arts, Cambridge, Cambridge University Press, 1989.
J.M. Kennedy, Metaphor in Pictures, “Perception”, 11, 1982: 589-605.
— Metaphor and Art, in W.R. Gibbs (ed.), The Cambridge Handbook of Metaphor and Thought, Cambridge, Cambridge University Press, 2008: 447-461.
G. Kress, T. Van Leeuwen, Reading Images: The Grammar of Visual Design, London-New York, Routledge, 2006.
G. Lakoff, M. Johnson, Metaphors We Live By, Chicago, University of Chicago Press, 1980.
M. Platonova, Poster, Poster on the Wall, Do you Really Mean It All?: Decoding Visual Metaphor ‘Global Warming’ in Public Awareness Campaigns, “Research in Language”, 17/2 (2019): 147-166.
G. Sonesson, Bats Out of the Belfry: The Nature of Metaphor, with Special Attention to Pictorial Metaphors, “Signs and Media”, 11 (2015): 74-104.
E. Šorm, G.J. Steen, Processing Visual Metaphor: A Study in Thinking out Loud, “Metaphor and the Social World”, 3/1 (2013): 1-34.
— VisMIP. Towards a Method for Visual Metaphor Identification, in G.J. Steen (ed.), Visual Metaphor: Structure and Process, Amsterdam, John Benjamins, 2018: 47-88.
P. Trifonas, The Semiotics of Visual Perception and the Autonomy of Pictorial Text: Toward a Semiotic Pedagogy of the Image, “Educational Philosophy and Theory”, 2020, DOI: 10.1080/00131857.2020.1761329.
V. Tsakona, Language and Image Interaction in Cartoons: Towards a Multimodal Theory of Humour, “Journal of Pragmatics”, 41/6 (2009): 1171-1188.