The metaphor and the iconic attitude

1.1 The narrow concept of sign

For the discussion of metaphor in this paper it is important to take note of the concept of sign that will be used in the following. It will be referred to as the narrow concept of sign (Lenninger 2012). The narrow concept implies that signs are meaning processes but also that not all meaning processes form the kind of meaning called signs (Piaget 1945). In people’s lives (for simplicity the discussion is here restricted to human beings) many kinds of meaning are processed all the time. Perception of the physical and social environment and proprioception of one’s own body and feelings are examples of fundamental meaning processes that may not necessarily at a given moment be part of an actual sign from the point of view of the perceiving subject. These meanings can nevertheless be complex and undergo changes in a lifetime (such as when one refines one’s ability to predict, plan, and reflect on situations in real life). The narrow concept of sign focused on (human) thought and communication must therefore comply with these circumstances (Piaget 1945; Vygotsky 1934/2001). This is what Piaget has called (subjective) differentiation.

In short, the sign is a compound unit of meaning processes that involves at least an expression, a content, and, potentially, a referent. Two of the constituent parts, the expression and the content, join together to form a relational unite that is relatively independent of the third part of the sign complex, that for which it stands in the communicative situation and which can be called the referent (Sonesson and Lenninger 2015: 193). This means, for instance, that the “sign expression” can be changed in order to affect its “content” without necessarily implying that the reference must be changed in all respects. In the elaboration of sign relations, meanings can be tested and different views on a reference concept can be emphasized or constructed. Because of the relations between expression and content, and its reference, respectively, the sign process opens up for possible interpretations of a reference meaning.

One more criterium for the sign relation must be emphasized here: the meaning relations in a sign process are asymmetric from point of view of the perceiver — the levels of meanings are not mixed up and cannot be reversed so that the expression becomes the content, or vice-versa, without a change in the understanding of the whole situation. If the perceiver mistakes a picture of a dog for a real dog, then there is not, in that instance, any sign from point of view of the perceiver (Sonesson and Lenninger 2015). This is not, however, the same as requiring that the perceiver actually reflects on the specific sign relations in every communicative situation. In ordinary communications with pictures and language, the differentiation between the sign (as an expression with its content) and its possible references is part of the world taken for granted. Authors, teachers, and artists work with these relations all the time — and for their efforts to be fruitful also their receivers should be receptive for something being a sign. Nevertheless, analytic awareness of the meaning relations and their impact on meaning can differ among the addressed receivers and still afford valid interpretations.

Thus, perceptions of pictures have referential meaning; hence they are signs. As Gibson (1980: xiii) formulates it, the picture affords “a sort of partial second-hand perception for stay-at-home observers,” which is different from the direct affordances of the picture as a canvas, a board, or an oddly shaped stone. But not only can a picture present existing dogs, and thus “show a dog” although no real dog is present at the moment, they can also visualise fantasy dogs that have not yet been seen. However simple this may sound, the differentiation in the conception of sign, from the point of view of the perceiver, is nevertheless that which accounts for a wide range of possibilities in the communication and sharing of ideas.

2.1 Creativity in connection with similarity

In an article from 1954, Max Black asks a simple but interesting question: how does one recognize a case of metaphor? He rejects the detection of a similarity relation as being relevant or sufficient to the recognition of a metaphor (a metaphorical expression). The metaphor, Black (1954: 285) concludes, creates similarity rather than “formulates some similarity antecedently existing.” Later in the same text he continuous thus: “It is easy enough to mutter ‘analogy,’ but closer examination soon shows all kinds of ‘grounds’ for shifts of meaning with context - and even no ground at all, sometimes” (ibid.: 298–290).

According to Black (ibid.), metaphorical constructions are creative and have the potentiality to overcome conceptual gaps in language. New meanings can take form in creative couplings of concepts (ibid.; Hausman 1989). However, according to the interaction view on metaphor suggested by Black (1954: 286), “connexion,” not similarity, is key. The interpreter, he maintains, is forced to connect the two ideas presented in order to obtain a metaphorical expression. Moreover, the apprehension of the connexion has relevance in the sense that metaphor as a creative resource in language use goes beyond the mere extension of meaning (ibid.: 289). While the rules of language determine some linguistic expressions as metaphors, other metaphors are due to variables in context, or result from perceiving implied intentions. Black’s own example of the latter is when Churchill called Mussolini “that utensil:” this statement made sense as a metaphor due to Churchill’s tone of voice and the background knowledge of the political and historical situation possessed by the audience:

… we must also recognize that the established rules of language leave wide latitude for individual variation, initiative, and creation. There are indefinitely many contexts (including nearly all the interesting ones) where the meaning of a metaphorical expression has to be reconstructed from the speaker’s intentions (and other clues) because the broad rules of standard usage are too general to supply the information needed. (Ibid.: 277)

In this passage, Black rejects the possibility that something like an independent language system can be enough to explain all metaphorical expressions. The rules of language are open to new initiatives and to creativity by meaning-making across contexts. In this text, Black challenges what he calls the comparison view of metaphor, that is, the “view that a metaphor consists in the presentation of the underlying analogy or similarity” (ibid.: 283). The metaphor, he suggests, rather creates than reveals similarities. His main objection to the comparison view is that the idea of a similarity in the metaphoric gap suffers from a vagueness that borders upon vacuity, and he sums up:

… likeness always admits of degrees, so that a truly “objective” question would need to take some such form as “ Is A more like B than like C in respect of P?” — or, perhaps, “Is A closer to B than to C on such and such scale of degrees of P?” (Ibid.: 283)

Black’s appraisal of the openness of language rules as having a creative role in language, but his denial to the potentiality (or openness) of meaning-making by similarity is intriguing — especially for our present purpose, which is to probe the weight of the notion of metaphor as an instance of iconic signs. It seems that Black treats non-specificity as a quality in supporting language creativity while also holding it to be a deficiency for understanding the roles and potentialities of similarity relations in metaphor.

2.2 The metaphor sign and Peirce’s division of hypoicons

According to Kirby (1997: 534; see also Isenberg 1963), Aristotle predominantly conceive metaphor as a highly valued skill in “seeing likeness in things.” The Aristotelian metaphor, in Kirby’s view, basically involves an insight into a relation of similarity between two items. Kirby (1997: 535) also points out that this insight is also apparent in Charles Sanders Peirce’s formulation of the sign constituents of the metaphor. Indeed, Kirby (ibid.: 538) suggests that applying a semiotic model on metaphor

will in fact make it possible to incorporate the features and strengths of other models, such as those based on comparison, substitution, or interaction, and will be suitable for the assessment of verbal texts in either prose or verse, as well as of visual and other nonverbal texts.

In a famous passage on sign meanings, Peirce defines metaphor as a subtype of iconic signs (EP2: 273–274, cited in Section 3.4). Peirce divides iconic meanings into image, diagram and metaphor. Thus, from this point of view a metaphor is an iconic sign — but this of course implies that not all iconic meaning relations are characterized as metaphors. The division presents a framework for approaching the concept of metaphor based on the observation that (1) in being an iconic sign, the metaphor derives its meaning from relations between qualia (here understood as the perception of qualities and attributes) — that is, more specifically, on the axis of similarity, (2) iconic relations have the potentiality to differ.

Despite its abstruseness, Peirce’s definition of hypoicons (subtypes of the icon) can be helpful. Peirce’s subdivision of the icon elaborates mainly one relational parameter (that of sharing qualities) into three variable conditions of similarity: as image, diagram, or metaphor. The proposal assumes that the three hypoicons constitute only one instance of Peirce’s more general trilogy of signs (cf. signs in the narrow sense) in iconic, indexical, and symbolic or conventional signs. The proposal also takes for granted that practically all real sign meanings are mixtures of such relations (Colapietro 1989). Crucial for the further analyses of sign meanings, however, is not only that different sign types (such as signs dominated by indexical, iconic, or conventional relations) are combined, but also that their internal organization (such as being dominated by image, diagram or metaphor meanings) is dynamic — even while, at any given time, one of the relations may dominate the understanding of the sign as a whole (Jakobson 1971).

A conclusion that can be drawn from this way of reading Peirce is that the potentiality for metaphoricity can be found in almost any sign relation. In addition, from this point of view, a potentiality for metaphor can be found in almost any sign use while at the same time not being the dominant relation (feature) of its meaning as a whole. It is in this sense that one can understand the metaphor as an instance of an iconic sign relation specified by its exploitation of a potentiality for iconicity, rather than as belonging to a specific means of expression, such as verbal language or pictures. This, it can be noted, brings back Black’s dilemma of defining metaphor expressions by way of tracing similarity relations alone.

3.1 Synesthetic relations and similarity

Before continuing the discussion of different types of similarity relations, let me briefly comment on synesthetic relations. Theories on synesthetic relations in the brain suggest explanatory models that interconnect meanings derived from processing different senses in the brain by means of associations, or lawful translations, from one modality to another (Day 1996; Ullmann 1964). The point is that brain-centred theories cannot replace or rule out the impact of a semiotic sign-based theory. It has already been said that any actual sign meaning discussed here only can occur thanks to its interpreters. Theories of inter-linkages in the human brain concern mental predispositions or possibilities for human cognition. But signs are not only meaning relations bound to subjects’ brains; signs belong to conscious interlocutors who develop the power to use them, share them and explore them. Signs are creative (Donald 1991). Predisposed linkages based on brain anatomy may explain why some cross-modal combinations are easier to share, or need less conventional support, than others in communication — but this is not enough to explain the voluntary creativity in sign-using cultures that seems to be independent of synesthetic mapping.

3.2 Similarity in the visual metaphor

At a general level, there is an organisation of the perceived world as we know it. In perception, we tend to recognize a world in which things seem to fit in as “real” or natural parts. This perception, in a world taken for granted, occurs, according to Husserl, within our natural attitude. Sometimes it happens in this privileged world that things appear as strange, “unreal,” or even as illusions (Gregory 1997), magic (Gibson 1980), or fantasy (Husserl 1980). As a material thing and a surface, the picture fits with our expectations about objects within the natural attitude. Also, as a perceptual experience, depiction is meaningful (showing fantasy or realism) in the sense of being related by similarity to our expectations about the visual world. As a communicative device, however, the picture also belongs with a common world of working (cf. Schütz 1967). This is a social aspect of the common world that differs from the natural attitude and that will be taken up again in discussing the iconic attitude in Section 4.

Sonesson (2008) has suggested a division of iconicity into a primary and secondary kind. Primary iconic signs are iconic signs that in themselves provide sufficient perceptual similarity to specify their meaning in some respect. A realistic picture, considered as prototypical, provides such perceptual information about its object. Consider for instance the felines at the “panel of lions” in the Chauvet Caves; after spending almost 40,000 years in darkness hidden from human eyes, they nevertheless provide enough perceptual information for people today to specify them as “profiles of female lion heads and the front parts of their bodies.” In this respect, meaning is not unclear for us; however, other levels of meanings may be obscure. The visual information afforded does not tell us anything about the meanings of things depicted, nor about the purposes of making these drawings. In that respect, the gaps in time and culture occlude some but not all aspects of meaning.

A criterion for primary iconic signs is that the sign expression is, from the point of view of the perceiver, self-evidently manifested as such, and is not confused with its reference. In contrast, secondary iconic signs derive from situations where the sign relation or similarity is not clear to the user if not pointed out by convention or an otherwise enhanced context. In cases of secondary iconic signs, more information must be added in the sign context in order for its similarity or sign relation to become clear. Secondary iconic relation can be the case if there is either too little or too much similarity for the sign relation to be apparent for its perceivers. Cases of providing too little similarity is discussed through the examples of droodles and diagrams in Section 3.3. Elsewhere, Sonesson (2008) has exemplified cases of providing too much similarity with the custom of using samples in retail and exhibitions. The samples are not the objects for sale; they are only the token used for the costumer to be able to make the right choice in ordering this type of thing. The customer who is not familiar with this standard in retail may fail to notice the sign relation and expect to bring the exhibition piece home.

Note, however, that the partition in iconic signs into the primary and secondary iconic kinds does not conflict with Peirce’s subdivision into hypoicons — but rather offers a complementary distinction that was neglected by Peirce. The prototypical picture is a sign with primary iconic ground, and so are also realistic figurines and film signs. From this point of view, a prototypical picture is what often is called a realistic picture. Perceptual similarity dominates its character as a sign.

3.3 Primary iconic ground and the prototypical picture

As already pointed out, similarity relations differ not only in quality (which properties that manifest them and how clear the similarity is rendered from the point of view of the perceiver) but also in type. Pictures involve visual meaning arising from visual experiences of the world around us while also fundamentally diverging from these experiences. Consider a prototypical diagram (i. e., diagram expression), for instance, a pie chart. For a pie chart demonstrate its similarity with an extrinsic phenomena or experience that it measures, supplementary information is needed. No matter how many measure points one specifies, the pie chart will never afford perceptual similarity with, for instance, the distribution of populations it measures. While a portrait imitates visual affordances of its object, the pie chart measures only proportional qualities of its content. To know which facts the prototypical diagram measures, additional information is necessary. The relevance of similarity in a pie chart differs from the relevance of similarity in a picture.

Notice that the secondary iconic ground covers all iconic signs that are not conceived to be primary iconic. Given this discrimination, secondary iconic signs do not form a homogeneous group. Moreover, the division in signs having primary and secondary ground is made from the point of view of the perceiver. That is, in case of a primary iconic ground some experiences are taken for granted and are thus expected to be clear also for one’s fellow perceivers. The boundaries or thresholds in these taken-for-granted meanings differ, which is something which also can be discovered in the creativity of iconic sign use. To make these observations less abstract the droodle (visual riddle) is discussed below (see further Price 2000).

Because of the scanty visual information in Figure 1, it can be considered as illustrating a case of secondary iconic ground. A clue like “it shows a bear climbing up a tree — from behind the tree” gives additional information about how to possibly interpret the display. The information indicates in the first place that it is a picture and thus affects the way one looks on it. (Look for tree trunk and paws!) Thus, the droodle in Figure 1 relies on habits from perceiving pictures with sufficient perceptual information to render depicted objects on their own. Discovering similarity in blurry pictures or in pictures showing only scarce details are examples of similar situations where the habit of seeing pictures is linked to the interpretation of perceptual similarity.

Figure 1:

One suggested meaning for this droodle, or pictorial riddle, is that it shows a bear climbing up a tree — seen from behind a tree trunk (in Price 2000).

A diagram and a droodle have in common that the interpreter needs contextual information in order to fill their similarity relations with reference meanings. However, the relevance of similarity differs between them. A diagram relies on relations of proportional similarity. Droodles, on the other hand, depend on the possibility to detect perceptual similarity in a depicted object in spite of the limited information. Note that other instances may rely on combinations of both these alternatives. Figure 2 shows the famous maps of “the motor and sensory homunculi” from medical literature that was published by Penfield and Rasmussen (1950). The illustrations show “relative sizes and locations of discrete functional regions within the sensorimotor cortex” (Snyder and Whitaker 2013: 277). Thus, in the illustrations pictorial and diagrammatic iconicity are combined into iconographic maps. Different cases of perceptual and proportional similarity relations can be distinguished in these images.

Figure 2:

Visuographic maps combining pictorial and diagrammatical similarity, first published in Penfield and Rasmussen (1950: 44, 57; see also Snyder and Whitaker 2013). The drawing in section (a) aims to show a sensory homunculus and that in (b) the corresponding motor homunculus.

Pictorial elements that rely on perceptual similarity (showing prominent body parts such as hands and face) are combined with a cross-section image of the human brain. The face in (b) is perhaps closer to primary iconicity than the face in (a). The profile in (b) depicts a “full” face seen from this angle, while the frontal view of the face in (a) is incomplete. Moreover, the pictorial elements (e. g., face, hand, foot, etc.) are presented in an unfamiliar way according to how one generally expects a human body to be built. Drawings of body parts are lined up along the contour lines of the cross-section image of the human brain, exhibiting unexpected size relationships between them. Moreover, the identification of the cross-section image of the human brain requires familiarity with either seeing the human brain in this “strange” condition or with similar pictures of it and is thus a case of secondary iconicity. In addition, the interpretation of these images relies also on verbal information and the convention of using lines pointing out names and locations on illustration images. A more thorough discussion of the iconic relations that can be drawn from these images should be presented elsewhere (see also Snyder and Whitaker 2013 for a critical discussion) but, taken together, these images presuppose to the ability to interpret a combination of several different meaning systems.

3.4 Metaphors and hypoicons

Pictures are not among Peirce’s principal interests. Yet, perception is crucial in his theory of meaning (Jappy 2013). Peirce notices the links between ordinary perception and pictures, but he never explores the consequences for the iconic ground to qualify as a sign in the narrow sense. Indeed, according to Peirce, the metaphorical relation is characterized by representing the sign relation prior to the thing represented. Metaphors are thus as a metarepresentations and, as a consequence, stress the meaning of something being a sign. In addition, metaphors are a metarepresentations conditioned by the way of processing the potentialities in meaning by similarities:

A possibility alone is an Icon purely by virtue of its quality; and its object can only be a Firstness. But a sign may be iconic, that is, may represent its object mainly by its similarity, no matter what its mode of being. If a substantive be wanted, an iconic Representamen may be termed a hypoicon. Any material image, as a painting, is largely conventional in its mode of representation; but in itself, without legend or label it may be called a hypoicon. Hypoicons may roughly [be] divided according to the mode of Firstness which they partake. Those which partake of simple qualities, or First Firstnesses, are images; those which represent the relations, mainly dyadic, or so regarded, of the parts of one thing by analogous relations in their own parts, are diagrams; those which represent the representative character of a representamen by representing a parallelism in something else, are metaphors. (EP 2:273–274)

The quote is from the short passage where Peirce (in 1903) suggests the division of the iconic signs. It says that the character of any iconic sign is constituted by its mode of partaking of “qualities.” Moreover, when Peirce asks us to accept the bracketed meaning as a hypoicon, he does so in terms of a Representamen. The division in hypoicons involves the structural conditions for how an iconic Representamen relates qualities to the Object and Interpretant.

Note that, just as the division into primary and secondary iconic ground, the subdivision in hypoicons suggests that there is more than one way of deriving meaning on the bases of qualia. Perceptual similarity, however, is not singled out as a particular resource or means for establishing a sign relation in the narrow sense.

The division in hypoicons can be considered a useful attempt to specify operational structures in a sign relation given by the constraints of qualities, irrespective of perceptual similarities between qualia. That is (1) by partaking of simple, unrelated qualia in the representation; (2) by partaking of related qualia in the representation; or (3) by partaking of the representative character of a representamen into a representation of something else. Note, that although “a painting” can illustrate the possibility of exercising image hypoicons, Peirce does not say that the picture per se is the manifestation of such representative relations. More probably, a picture (or parts of it) can manifest all three of the hypo-iconical operations (see discussion in Lenninger 2012).

Note also that when Peirce (EP 2:273–274) detaches the picture from its legend and label, he suggests that we should lay aside our cultural heritage with reference to what is known about the picture (pictorial narrative, provenience, genre, technology, and so on). Note also, however, that this is not the same as asking us to suspend our awareness of the sign relation! Based on its primary iconic ground, the picture remains a surface still looking like something else, perhaps a face (or possibly only a mixture of its single features such as redness, saturation, grain, etc.).

4.1 The iconic attitude

So far, the focus has been on showing that similarity is not a single kind of relationship. The relevance of similarities differs. It has also been stressed that some similarities are tightly coupled to a sign relation, but that this is not necessarily the case for all similarity relations. An important point however is that via the sign relation one may gain insight into structurally different organizations of similarity relations.

Image, diagram and metaphor hypoicons could be described as different instances of iconic attitudes directed to an iconic ground. Iconic grounds are essentially provided for by different conditions of similarity relations between expression and content. Note again that metaphorical relations are not confined to sign resources that predominantly rely on primary iconic ground, such as in pictures. In fact, metaphors are often expressed verbally and, indeed, metaphorical concepts have been argued to be at the foundation of human meaning-making (Lakoff and Johnson 1980). Remember also that sign meanings are rarely — if ever — strictly iconic, indexical, or conventional, but are rather mixtures of these. Relations of contiguity and conventionality partake in iconic meanings. It is the pertinence of the whole that makes something into an iconic, indexical, or conventional sign.

It is flexibility and not static hierarchic arrangements that organizes meanings into iconic signs (cf. Jakobson 1971). Relations of contiguity and conventionality participate as aspects in the character of iconic meanings. The nature and relevance of their partaking affect the categorisation of iconic sign.

I borrow the formulation of Alfred Schütz to describe an attitude. An attitude is:

… briefly, the subjective meaning man bestows upon certain experiences of his own spontaneous life. What appears to the observer to be objectively the same behaviour may have for the behaving subject very different meanings or no meaning at all. (Schütz 1967: 210)

The attitude is a first-person position on meaning; moreover, it is situated and has an actor-action relation with respect to an actual object of meaning. In the natural attitude, one’s own meanings are not reflected upon — but are taken for granted as being general. In the natural attitude, individual experiences are unified with the expectation of a common ground, and thus can be expected to be true also for others (if reflected upon). However, people have the cognitive flexibility to be surprised! Moreover, one can expect others to be surprised as well.

The iconic attitude responds to flexibility in thinking and the potentiality of an iconic ground. Simultaneously the iconic attitude provides the necessary stabilization to predict the range of meaning relations in a specific sign relation. Vagueness, it was said above, is a quality in icons, and the iconic attitude accords with the range of operations making use of it — which do not only consist in precisions, but also in the acceptance of vagueness in the iconic ground (cf. Jappy 2013).

Independently of whether metaphor relations are implicit in pictures in general, metaphor conception does not always partake of the predominant iconic relation in picture perception. When metaphor relations are salient, meaning relations produced by conjunctions or the colliding of experiences are enhanced, or tested, via the sign relation. Operating on the generalities of vagueness, the metaphor construction as a whole refers to a new focal condition of constraints that can be understood only by means of perceiving the metaphor itself (cf. Hausman 1989).

4.2 A lion person figurine in human culture

Within the narrow conception suggested here, signs are not just any meaning relation. Signs are taken to be special: signs have reference and signs can indicate or present meanings that are in that moment only present as sign meaning. Sign meaning can be very precise in its description of a state or an experience. In elaborating sign meanings, the communicator can also test, or give proposals, with respect to different versions of its references.

Near Ulm, in Germany, archaeologists have made unique findings. In 1939, the first fragments of a 30 cm high animal-shaped ivory figurine were found. The figurine is estimated to be 35,000–40,000 years old. The figurine, which has been reconstructed, has been named “the lion person” (Löwenmensch) or “the lion-headed figurine” (Figure 3). Together, the two names mirror the accuracy and vagueness in meaning produced by the primary iconic ground. Owing to similarity relations, we perceive a figurine with a human body posture but with a head of a lion. The figurine itself, however, does not reveal how to conceive the somewhat unexpected combination of intelligible parts. In addition, because of its realistic traits (in parts) we can expect that the figurine was produced on purpose, at some level of intentionality (cf. Lewis-Williams 2004). However, we cannot know for certain what was its original meaning or meanings. Did the figurine originate as a realistic replica of an animal-shaped human (perhaps shaman) or was it conceived as an animal shaped as a human (perhaps a God), or was it simply a lion figurine? Its prehistoric target domain, or content, is obscure to us. It is possible that the erect body posture originates from an error of realism, based on the priority of the human body. It is also plausible that it was primarily a consequence of the shape of the piece of ivory used. In the end these questions boil down to the question whether the figurine was conceived as a “literal” visual statement or can be traced back to a manifestation of metaphorical thinking. Note that we do not know when figurines like this one were first produced or spread between groups of people. In the same area (Blaubeuren area), other human or animal-shaped figures have also been found, and these figures show different levels of abstraction (e. g., the famous figurine called Venus of Hohle Fels, a flying bird, a horse, and a coil-shaped human body). Among found objects, so far, the lion person stands out in size (the other figurines are smaller) but also in shape.

Figure 3:

A line drawing of the lion person figurine (Der Löwenmench). The figurine presents the possibility of a metaphor construction in visual meaning. The figurine is carved in ivory, estimated to be 35,000–40,000 years old, and was found in pieces in 1939 at Mount Hohlenstein in Germany. The Figurine can be seen at the Ulm Museum, Germany.

Let me be clear: the proposal here is not that the lion person is the first case of metaphor expression in human culture — there would be many reasons to object to the validity of such a claim. The main reflection I derive from this is that irrespective of whether the figurine was intended as a metaphorical expression or not the mere existence of a thing like the lion person triggers the potentiality of purposive metaphor thinking. In perceiving an object like this, the flexibility and freedom in visual meaning construction is brought up to the surface and can be shared with co-perceivers. That is, the vision based on the accuracy, and materiality, of meanings produced by primary iconic grounds opens the potentiality of seeing a lion, or another animal, and this vision is easily picked-up and thus shared as meaningful in a communicative and creative culture.