Elements of Peircean phenomenology: From categories to signs by way of grounds

1.1 The categories of triadic structuralism

With this aim, the three Peircean categories are treated as one possible, but not exclusive, result of the Husserlean free variation in the imagination, also known as ideation. From this point of view, there is no reason to think that the Peircean categories are universally applicable, as Peirce himself was led to think; but they seem to be at least partially adequate for the understanding of one particular ontological domain, which is singularly important, at least from the standpoint of human beings, or any other kind of living creatures. This domain, I submit, concerns the relation of the acting and perceiving subject to the world at large (“experiences” in the sense of CP 7.524–7.538). In other words, the three categories describe intentionality in the sense of Brentano and Husserl, that is, the directedness of the mind to the things of which it takes cognizance. As is well-known, Peirce himself did not recognize the distinction between mind and matter, supposing the former to shade gradually into the other. Thus, he posited a “quasi-mind” at one end of the relations that he acknowledged. This may be a metaphysical truth, but here I am only interested in the experience given to phenomenology, in which mind and matter are very different things. Indeed, it is precisely because the mind and the body are experienced as in some sense different, that it makes sense to talk about the mind as embodied – and, correlatively, of the body as minded.

Structuralism is the idea that all meaning is produced by the opposition of terms, or, at least, that meaning is always perceived by means of an opposition of terms. Let’s call the former “strong structuralism” and the latter “weak structuralism” (cf. Sonesson 1989: 81, Sonesson 2009, and Sonesson 2012). Beginning with the work of Lévi-Strauss and Jakobson, we tend to take for granted that this opposition is basically an opposition between two terms at a time. Structuralism, however, does not have to be dyadic. Its important feature is the stipulation that the categories there are, no matter how many, are defined in relation to each other. And this is certainly the case with Peirce’s categories.

What Peirce proposes must be characterized as a strict triadic structuralism. Indeed, to take the general case, everything must pertain to Firstness, Secondness, or Thirdness. Particulars, but not generals, it is true, may partake of them all. This is in fact not very different from dyadic structuralism, even as applied to language. A phoneme, on a structuralist reading, necessarily has a specific feature or the opposite one. This is not the case with concrete sounds. But, rather than have both that specific term (and its opposite), the sound is thought to realize some intermediate case.

As I have argued elsewhere (Sonesson 2007), the question whether something has two or three parts has no meaning before determining the domain for which the model is valid, as well as the criteria determining the relevant properties according to which the division is made. Since the domain of the Saussurean sign is that which is internal to the sign system, its content being all the time opposed to the “real world” it interprets, it would be triadic – to the extent that reality outside the sign system is included in the domain to be analyzed. As for the Peircean sign, it really comprises six instances, if all criteria of division are included, since there are two kinds of objects, and three kinds of interpretants, but only one kind of representamen. As soon as we abandon the idea of our subject matter beings signs, as Peirce himself suggested late in life, it may be easier to make sense of these divisions.

But there is something more to Peirce’s triadic structuralism. Jakobson, Lévi-Strauss, and their followers seem to be content to affirm that everything comes by twos, but they impose no limits on the content of the units those opposed, if it is not that one unity must, in one sense or another, have properties which are opposed to the properties of the other. More exactly, the unities must have properties that are identical, without which the opposition does make sense. It will be noted that Prague structuralism, as represented by Trubetzkoy, does not impose any such specific requirement. In any case, the triadic structuralism of Peirce is different, because it supposedly requires the units themselves that are triadically opposed to be somehow intrinsically instances of Firstness, Secondness, and Thirdness. And this is where Peirce’s conception goes beyond classical structuralism.

Quite apart from the necessity of always making threefold divisions, there is the question of the content of each of the three categories. Firstness, Secondness, and Thirdness mean so much more than just being the first, the second, and the third category of an obligatory segmentation of the world into triads. Peirce’s phenomenology is in fact very short, as Ransdell (1989) rightly observes, because he rapidly proceeds to tasks that he takes to be beyond phenomenology. There is every reason to deplore this, in particular if we follow Peirce in identifying phenomenology with the study of the categories. In fact, Peirce has a lot to say about the categories, but always in passing, on the way to more serious work, and never entering into any detail. This is where one may start regretting that Husserl, with his sense for detail, his meticulous way of proceeding, and his habit of returning over and over again to the same task, never really happened upon Peirce’s categories. But we must start from what we have got.

Often, Peirce simply claims that Firstness is something that exists in itself, Secondness must be related to something else, and Thirdness requires a more complex relationship, either a relation between three things, or a relation between relations, or perhaps both at the same time. One of the more formal definitions of the three categories reads as follows:

Firstness is the mode of being of that which is such as it is, positively and without any reference to anything else. Secondness is the mode of being of that which is such as it is, with respect to a second but regardless of any third. Thirdness is the mode of being of that which is such as it is, in bringing a second and third into relation to each other. (CP 8.328)

Firstness and Secondness could here almost be understood as somewhat distorted equivalents of Husserl’s (1913: 1, 225) distinctions between independent and dependent parts, with the exception that there is no proviso for the difference between mutual and one-sided dependence.^[1] This then raises the question what the business of Thirdness is. If it involves a relation between two terms, instead of only one term and a relation, as Secondness could perhaps be understood to be, or a relation between relations, why then should we not go on defining Fourthness, and so on? Of course, Peirce himself claimed that all relations beyond Thirdness could be dissolved into several relations, but Thirdness itself could not be so resolved. It is not clear whether this is indeed a phenomenological fact. Actually, this must, among other things, depend on what exactly is to be understood by Firstness, Secondness, and Thirdness. Thus, for instance, is there really no relationship in Firstness? When it is used to define a kind of sign, the icon, it must already be supposed to be part of a relationship, even before it is seen as a sign namely, the relation of similarity.^[2] Indeed, Peirce himself repeatedly says that Firstness cannot be grasped as such. And what about Secondness? Is Secondness second, because it is made up of two things – in which case it would already be made up of three items, two things and a relation? Or should the second thing be conceived as a relation hooked up to an element, as I suggested some time ago (Sonesson 2012). Thirdness, in a similar way, then would have to contain three hooks, one of which is already filled up with an element describing the nature of the relationship.

These numerical values of the categories seem to be borne out by Peirce’s application of them to signs. An icon, it will be remembered, is a sign, in which the “thing” serving as expression is similar in one respect or another to (or has properties in common with) the “thing” which serves as its content – but the sign is an icon only if the similarity between these relata (the elements which are related) obtains independently of the sign relation and independently of possible relations between the relata as such. Similarly, an index is a sign in which the “thing” serving as expression is connected in one respect or another to the “thing” which serves as its content – but the sign is an index only if the connection between these relata obtains independently of the sign relation. In a symbol, in contrast, there is nothing, apart from the sign relation, which relates expression to content. Clearly, the numerical values have their part to play.

However, I do not think it sufficient to say that Firstness, Secondness, and Thirdness correspond to a one-place predicate, a two-place predicate, and a three-place predicate, respectively, as Ransdell (1989) maintains. Peirce probably thought so, for instance when he claimed that an

act of attention has no connotation at all, but is the pure denotative power of the mind, that is to say, the power which directs the mind to an object, in contradistinction to the power of thinking any predicate of that object. (CP 1.547)

But this cannot explain the workings of the categories. Rather, Firstness must be a one-place predicate with one term in the slot, Secondness a second-place predicate furnished with two terms, and Thirdness a three-place predicate featuring three terms. According to Peirce, “A fork in the road is a third, it supposes three ways: a straight road, considered merely as a connection between two places is second, but so far as it implies passing through intermediate places it is third” (CP 1.337). In this sense, the fork is not only the place where the road splits, but from where it goes to different places.

There are some passages where Peirce seems to be much clearer about the numerical import of the categories:

[E]xperience is composed of: 1st, monadic experiences, or simples, being elements each of such a nature that it might without inconsistency be what it is though there were nothing else in all experience; 2nd, dyadic experiences, or recurrences, each a direct experience of an opposing pair of objects; 3rd, triadic experiences, or comprehensions, each a direct experience which connects other possible experiences. (CP 7.524–538)

But, as this quotation already suggests, there clearly is much more to Firstness, Secondness, and Thirdness than numerical values. However, first we should note that the categories are all “experiences,” that is, they are “phenomena” or intentional acts. But then Firstness cannot be so simple, because it must already involve an experiencer – or at least a quasi-experiencer.

1.2 From quantities to qualities in the categories

There are many places, nevertheless, where Peirce imputes a much more concrete content to each of the categories. Since it is impossible to look at all the (only partly overlapping) descriptions of these categories offered all through Peirce’s writings, a few instances pertaining to each category will have to do here, most of them taken over from the discussion in Sonesson (2009; cf. Table 1). Reasoning in terms of sufficient and necessary properties, there does not seem to be much hope of finding any more general term able to subsume this welter of divergent properties. And yet, in spite of what is, on the face of it, the range and diverseness of the contents attributed to the categories, they certainly are much more specific than what is contained in the purely numerical definitions.

Perhaps it could be argued that the three categories are, formally, quite apart from their content, themselves of the order of Firstness. Indeed, given these descriptions, Firstness, Secondness, and Thirdness certainly sound very much like what Vygotsky (1962) would have called “chain-concepts,” characteristic of small children and what at the time were known as “savages.” Since Wittgenstein presented them as “family concepts,” spread all over ordinary language, these terms have been somewhat rehabilitated. Eleanor Rosch conceived the idea of the prototype, according to which a category is defined by a central example that seems to embody what is important to the category, with other members being ranked at different distances from the prototype. In a number of experiments, Rosch showed this explanation model to make sense beyond phenomenology. One of the most interesting experiments involved placing objects on a spatial layout in relation to some object that was taken to be the prototype of the category. Rosch and Mervis (1975) reflected on the relations between the prototype and Wittgenstein’s family concept, arguing that the difference consists in the former being related to a central example, while the second lacks any such instance.^[3]

At first, one may tend to see in the Peircean categories some kind of “chain-concepts” or “family concepts,” but I think a few of the members of the “chains” can really be considered to make up the prototype of the categories. This could be seen as a generalization of the claim, made over and over again by Peirce, that some instances of his categories are “degenerate.”^[4] The others, then, would be the prototypes or ideal types. According to Ransdell (1989), all instances of signs repertoried by Peirce that are not signs in the proper sense are degenerate. If degeneracy should here be taken in the sense of mathematics, degenerate items are objects that change their nature so as to belong to another, usually simpler, class. Thus, for instance, a point is a degenerate circle, namely, one with radius 0. This actually seems to go even further than the prototype concept, to the point of appearing less useful.

In the case of Firstness, this central idea is difficult to grasp, but it certainly has something to do with fleetingness or streaminess. Secondness is dominated by the idea of reaction/resistance. And law or regularity tends to be the most prominent element of Thirdness. However, I think the following quotation from Peirce goes a long way in showing that (double-sided) resistance is the ideal type of Secondness:

A door is slightly ajar. You try to open it. Something prevents. You put your shoulder against it, and experience a sense of effort and a sense of resistance. These are not two forms of consciousness; they are two aspects of one two-sided consciousness. It is inconceivable that there should be any effort without resistance, or any without a contrary effort. This double-sided consciousness is Secondness. (EP 1: 268)

Secondness is perhaps the easiest category to grasp: it is about effort and resistance. Or we could say: resistance to the world (“putting your shoulder against”) something, as well as the world resisting back (“a sense of resistance”).^[5] Firstness can then only be understood as opposed to Secondness: something appearing, bringing about an event, catching the attention which starts of the chain of Secondness, in which we live. Thus, Thirdness may stand for reflection, meta-consciousness, the observation of the reaction, which, as products, may give rise to rules and regularities.^[6]

Peirce, it will be remembered, always refers to the difficulty of talking about – and even conceiving – Firstness on its own: it needs the presence of Secondness. This shows a decidedly structuralist bend, which we will be exploring in the following.

1.3 The golden bough of Peircean phenomenology

The Peircean sign is a sign only in a very Pickwickean sense of the term. It is one of three specifications of Firstness, Secondness, and Thirdness. Taken literally, it is a combination of a fleeting moment with something which resists and something which is a rule. Even if we suppose this characterization to say something about the properties of expression, content, and the relation between them, respectively (which is not at all obvious), it is certainly a description that applies to numerous phenomena apart from signs or representations, which means, as Peirce (CP 4.3) was to point out in later life, that to use the latter terms for something in fact much more general than the sense these terms habitually carry, is “injurious.” Moreover, this formula doesn’t tell us anything about the specificity of the sign. No doubt this idea is contained in the idea of degeneracy, but this has the curious effect of extending the name of a more specific instance to a lot of widely divergent phenomena, without however defined that specific phenomenon, but only the general class of classes. It is like saying that the point is a degenerate circle, but defining the circle as if it were a point.

Nevertheless, the Peircean triad may have something to say about meaning in a much more general sense, for which we should perhaps reserve the Peircean term “semiosis.” Maybe this is what Peirce was thinking about when, at a later stage, he complained that his notions were too narrow, and that, instead of referring to signs, he should really be talking about mediation or “branching” (see MS 339, quoted in Parmentier 1985: 23).

It was suggested above that the prototypical meaning of Secondness is resistance including the resistance to resistance, and so on. In the theatre of our experience, there must be something to initiate this chain of resistances or reactions. It is a thing no matter which that first grasps our attention – that is, Firstness. In the primary sense, Thirdness is simply the observation of something occurring and the reaction to this occurrence. In accordance with this conception,

a sign [or, as I would say, semiosis] is whatever there may be whose intent is to mediate between an utterer of it and interpreter of it, both being repositories of thought, or quasi-minds, by conveying a meaning from the former to the latter. (MS 318, quoted in Jappy 2000)

In many passages of Peirce’s works the object is not described as that which the sign is about, that is, to which it refers, in the sense in which this term is used in linguistic philosophy; instead, it is that which incites somebody to produce a sign which may or may not coincide with the referent. It is in this sense that the object is Secondness: it concerns the relation between the reality perceived and the expression produced. Similarly, the interpretant must be seen as the result of the receiver taking in the whole event of the utterer’s creating an expression starting out from some feature of experience. Because it refers to the relation between the utterer and that which he reacts to, it is not only an elementary relation, it is Thirdness. Indeed, this idea is very well illustrated by the notion of “branching,” which Peirce used to characterize his later concept of mediation, or even by his example of “a fork in the road” (CP 1.337) Conceived in this way, Peirce’s theory appears to be about the situation of communication, but much closer to what we now would describe as a hermeneutical model than to the model known from the theory of information. Indeed, Peirce’s characterization of semiosis, quoted above (from MS 318) is reminiscent of my formulation of the import of the Prague school model of communication, as an offer of an artefact from one subject to another in the form of a task of interpretation (see Sonesson 2014).

Even describing that which Peirce is concerned about as an act of communication may be too specific a notion. Instead, it could be characterized as an observation being observed. Summarizing all of Peirce’s different attempts at pinning down the nature of Firstness, we could probably say that it is something that appears without connection to anything else. It is thus prior to all relationships. Secondness is not only the second term that comes into play, but also it is made up of two parts, one of which is a property, and the other a relation. It is something the function of which is to hook up with something already given. In this sense, it is a reaction, in the most general sense, to Firstness, where the first part is the connection to the property independently appearing and the second part describes the nature of this relationship. Thirdness is not only the third term which is ushered in, but it consists of three parts, two of which are relational; one which is hooked up to the term of Firstness and another which is connected to the relation of Secondness, together with which we find a third term describing the relationship between these two terms. It is thus an observation of the reaction. Appearance is monadic, reaction is dyadic, and observation is triadic (Figure 1).

Figure 1:

An attempt at the reconstruction of the Peircean triad.

Such a characterization has rather little to tell us about something as specific as the sign. It is really about something much more general and elementary: something first appearing to consciousness, the reaction of a mind (or, if you like, a quasi-mind) to this occurrence, and then the mind taking account of its own act. In its first stage, this clearly has something to do with what Husserl calls intentionality, the directness of the mind to that which is beyond the mind. More specifically, it all seems to be a story told about attention. Forgetting for the moment about the sign, we will go on to suggest that “the ground,” which is intermediate between the categories and the signs, is best understood as co-attention: the bringing of two items together into focus (See Gurwitsch 1957).

2.1 Ground and relevance

As applied to signs, iconicity is one of the three relationships in which a representamen (expression) may stand to its object (content or referent) and which can be taken as the “ground” for their forming a sign: more precisely, it is the first kind of these relationships, termed Firstness, “the idea of that which is such as it is regardless of anything else” (CP 5.66), as it applies to the relation in question. According to Greenlee (1973: 64), the ground is that aspect of the referent that is referred to by the expression, for instance, the direction of the wind, which is the only property of the referential object “the wind” of which the weathercock informs us.^[8] On the other hand, Savan (1976: 10) considers the ground to consist of the features picked out from the thing serving as expression, which, to extend Greenlee’s example, would include those properties of the weathercock permitting it to react to the wind, not, for instance, its having the characteristic shape of a cock made out of iron and placed on a church steeple. Since Peirce talks about “the ground of representamen,” it seems that Savan must be right as far as Peirce’s intentions go. And yet, it seems to me that, in order to make sense of the notion of iconic signs, we must admit that both Greenlee and Savan are right: the ground involves both expression and content. Rather than being simply a “potential sign-vehicle” (Bruss 1978: 87), the ground would then be a potential sign. Indeed, if we take seriously Peirce’s claim that the concept of “ground” is indispensable, “because we cannot comprehend an agreement of two things, except as an agreement in some respect.” (CP 1.551), then it must be taken to operate a modification on both the things involved. In other words: the ground involves an abstraction applied to the content with respect to the expression, in the same time as an abstraction applied to the expression with respect to the content.

The operation in question, I submit, must be abstraction or, as I would prefer to say, typification. In one passage, Peirce himself identifies “ground” with “abstraction” exemplifying it with “the blackness of two black things” (CP 1.293).^[9] It therefore seems that the term ground could stand for those properties of the two things entering into the sign function by means of which they get connected. i.e. both some properties of the thing serving as expression and some properties of the thing serving as content. In case of the weathercock, for instance, which serves to indicate the direction of the wind, the content ground merely consists in this direction, to the exclusion of all other properties of the wind, and its expression ground is only those properties which makes it turn in the direction of the wind, not, for instance, the fact of its being made of iron and resembling a cock (the latter is a property by means of which it enters an iconic ground, different from the indexical ground making it signify the wind). If so, the ground is really a principle of relevance, or, as a Saussurean would say, the “form” connecting expression and content: that which must necessarily be present in the expression for it to be related to a particular content rather than another, and vice-versa. This phenomenon in well-known from linguistics, where often conventional rules serve to pick out some properties of the physical continuum, differently in different languages, which have the property of separating meanings, i.e. of isolating features of the expression on the basis of the content, and vice-verse. The difference is, of course, that in the iconic and indexical grounds, the relation that determines one object from the point of view of the other, is basically non-conventional (cf. Sonesson 1989: Ch. 3.1).

Figure 2:

Bühler’s Organon model (with “abstractive relevance” and “abstractive supplementation”: see Bühler 1982 [1934]: 28).

If the ground is a form of abstraction, as Peirce explicitly says, then it is a procedure for engendering types, at least in the general sense of ignoring some properties of things and emphasising others, for the purpose of placing them into the same class of things. And if it serves to relate two things (“two black things” for example, or “the agreement of two things” in general), it is a relation, and it is thus of the order of Secondness, i.e. “the conception of being relative to, the conception of reaction with, something else” (CP 6.32). All this serves to underline the parallel with the principle of relevance, or pertinence, which is at the basis of structural linguistics, and much of semiotics inspired by it (the work of Louis Hjelmslev and Luis Prieto, notably). But we could take this idea further, adding to the notion of ground a more explicitly constructive aspect. To many structuralists (those of the Prague school notably), relevance is a double movement, which both serves to downplay non-essential elements and to enhance others which were anticipated but not perceived: thus, it depends on the twin principles of “abstractive relevance” and “apperceptive supplementation” embodied in Bühler’s Organon model (see Figure 1 and Bühler 1982 [1934]; Sonesson 1989: Ch. 2.4.2), as well as on the Piagetian dialectic between accommodation and assimilation (see Sonesson 1988: Ch. 1.3.1).

2.2 Grounding iconicity

Conceived in Peircean terms, iconicity is one of the three relationships in which a representamen (expression) may stand to its object (content or referent) and which may be taken as the “ground” for their forming a sign: more precisely, it is the first of these relationships, Thus, the ground seems to be a part of the sign having the function to pick out the relevant elements of expression and content, similar, in that respect, to the “form” (the principle of relevance) of the Saussure/Hjelmslev tradition (see Sonesson 1989: 202). Contrary to the indexical ground, which is a relation, the iconic ground consists of a set of two classes of properties ascribed to two different “things,” which are taken to possess the properties in question independently, not only of the sign relation, but of each other. Indexicality as such involves two “things,” and may therefore be conceived independently of the sign function, but iconicity should be possible to conceive independently even of the second “thing” involved.

The blackness of a blackbird, or the fact of Franklin being an American, to use some of Peirce’s own examples, can be considered iconicities; when we compare two black things or Franklin and Rumford from the point of view of their being Americans, we establish an iconic ground; but it is only when one of the black things is taken to stand for the other, or when Rumford is made to represent Franklin, that they become iconic signs. Just as indexicality is conceivable, but is no sign, until it enters the sign relation, iconicity has some kind of being, but does not exist, until a comparison takes place. In this sense, if indexicality is a potential sign, iconicity is only a potential ground (see Sonesson 1994, Sonesson 1997, and Sonesson 1998; also see Table 2).

Many semioticians, in particular those who deny the existence of iconic signs, apparently believe pictures to be typical instances of this category. There are several reasons to think that this was not Peirce’s view. Pure icons, he states (CP 1.157), only appear in thinking, if ever. According to Peirce’s conception, a painting is in fact largely conventional, or “symbolic.” Indeed, it is only for a floating instant, “when we lose the consciousness that it is not the thing, the distinction of the real and the copy,” that a painting may appear to be a pure icon (CP 3.362; see also Sonesson 1989: Ch. 3.1). It will be noted then that a pure icon is thus not a sign, as the latter term is commonly understood (although Peirce will sometimes state the contrary). Peirce specifically refers to the case in which the sign loses its sign character, when it is not seen as a sign but is confused with reality itself (which could actually happen when looking at a picture through a key-hole with a single eye), when, as Piaget would have said, there is no differentiation between expression and content (See Sonesson 1989, Sonesson 1992, Sonesson 2007, and Sonesson 2011).

In point of fact, if would seem that, at least sometimes, the pure icon is taken to be something even less substantial: an impression of reality, which does not necessarily correspond to anything in the real world, for “it affords no assurance that there is any such thing in nature” (CP 4.447). Thus, it seems to be very close to the “phaneron,” the unit of Peircean phenomenology (itself close to the Husserlean “noema”), which is anything appearing to the mind, irrespective of its reality status (see above Section 1 and Johansen 1993: 94). In this sense, the Peircean icon is somewhat similar to that of cognitive psychology, for it involves “sensible objects” (CP 4.447), not signs in any precise sense: however, it still comprises all sense modalities.

In most cases, when reference is made to icons in semiotics, what is actually meant is what Peirce termed hypo-icons, that is, signs which involve iconicity but also, to a great extent, indexical and/or “symbolic” (that is, conventional, or perhaps more generally, rule-like) properties. There are supposed to be three kinds of hypo-icons: images, in which case the similarity between expression and content is one of “simple qualities”; diagrams, where the similarity is one of “analogous relations in their parts”; and metaphors, in which the relations of similarity are brought to an even further degree of mediation. Diagrams in the sense of ordinary language are also diagrams in the Peircean sense, e.g. the population curve which rises to the extent that the population does so. The Peircean concept is however much broader, as is the notion of metaphor, which would, for instance, also include the thermometer. Moreover, no matter how we choose to understand the simplicity of “simple qualities,” the Peircean category of images will not include ordinary pictures (which would be metaphors of metaphors; cf. picture perception), although Peirce sometimes seems to say so: if anything, an Peircean image might be a color sample used when picking out the paint to employ in repainting the kitchen wall.

Contrary to the way in which icons have been conceived in the later semiotic tradition, diagrams, rather than pictures, are at the core of Peircean iconicity: at least, they are of most interest to Peirce himself. Indeed, mathematical formulae and deductive schemes, which are based on conventional signs, are those most often discussed in his work.

There is still another sense in which pictures are far from being central instances of icons. As was noted above, the fact that an object serving as the expression of an icon, and another object serving as its content, possess, in some respects, the same properties, should not be a result of one of them having an influence on the other. In the case of an icon (contrary to the case of an index), “it simply happens that its qualities resemble those of that object, and excite analogous sensations in the mind for which it is a likeness” (CP 2.299). Since both Franklin and Rumford are Americans, Peirce claims, one of them may serve as a sign of the other; but the fact that Franklin is an American is quite unrelated to Rumford’s being one. But there is at least one sense in which this is not true, not only of a photograph (which Peirce often pronounces to be an index), but also in the case of a painting or the image on a computer screen: in each case, the “thing” serving as the expression is expressly constructed in order to resemble the “thing” serving as the content, although a direct physical connection only exists in the first instance. Leonardo painted the canvas known as Mona Lisa in order to create a resemblance to the wife of Francesco del Giocondo, and, although the resemblance is of a much more abstract kind, the same is true of Picasso painting Gertrude Stein or Kahnweiler. And it is as true of a synthetic computer picture showing a lamp as of a photograph with the same subject.

Peirce’s claim that the properties of expression and content pertain to them independently seems more relevant to identity signs (like Franklin representing Rumford) than to pictures. In another sense, on the other hand, pictures are far more iconic than, for instance, objects representing themselves: they can do with far less indexicality and convention. From this point of view, and contrary to what has been suggested by Morris (1971 [1946]: 98), and which often is repeated in theatre semiotics, an object is not its own best icon.

When used to stand for themselves, objects are clearly iconical: they are signs consisting of an expression which stands for a content because of properties which each of them possess intrinsically. And yet, without having access to a set of conventions and/or an array of stock situations, we have no possibility of knowing, neither that something is a sign, nor what it as sign of: of itself as an individual object, of a particular category (among several possible ones) of which it is a member, or of one or other of its properties. A car, which is not a sign on the street, becomes one at a car exhibition, as does Man Ray’s iron in the museum. We have to know the show-case convention to understand that the tin can in the shop-window stands for many other objects of the same category; we need to be familiar with the art exhibition convention to realize that each object merely signifies itself; and we are able to understand that the tailor’s swatch is a sign of its pattern and color, but not of its shape, only if we have learnt the convention associated with the swatch (cf. Sonesson 1989: Ch. 2.2.2, 1994).

Convention is thus needed, not only to establish the sign character, but also the very iconicity of these icons. Since iconicity can be perceived only once the sign function, and a particular variety of it, is known to obtain, the resulting icons may be termed secondary (Sonesson 1994). This also applies to “droodles,” a kind of limiting-case of a picture exemplified by Carraci’s key, in which a triangle above a horizontal line is discovered to represent a mason behind a stone wall, once we are told so; as well as the manual signs of the North American Indians, which, according to Mallery (1972 [1881]: 94–95), seem reasonable when we are informed about their meaning.

In these cases, knowledge about the sign function already obtaining between the two “things” involved is clearly a prerequisite to the discovery of their iconicity. The opposite case, in which it is the perception of iconicity which functions as one of the reasons for postulating a sign relation, would seem to be more germane to Peirce’s conception of the icon. Such a primary icon is actually realised by the picture sign. Indeed, we know from child psychology and anthropology that no particular training is needed for a human being to perceive a surface as a picture (cf. picture perception). The possibility of this feat remains a mystery: they properties possessed in common by the picture and that which it represents are extremely abstract. It has been suggested that picture perception is only possible because there is a taken-for-granted hierarchy of things in the world of everyday life which makes certain objects and materials more probable sign-vehicles than others (Sonesson 1989 and Sonesson 1994).

It seems clear, then, that the iconic ground is a kind of Second Firstness, and that the iconic sign is a Third Firstness. Still, there remains a problem with pure iconicity, even apart from the fact that is may be “injurious” to apply the term to something which is not even, according to Peirce, a relation, and thus can have nothing to do with similarity, apart from being its requisite. This is that, since even pure iconicity is a kind of “experience,” as Peirce clearly maintains, it cannot really be monadic, since it must exist for a subject (or at least a quasi-subject). But since the experiencing subject must be added to all the categories, we may perhaps choose to ignore it for the time being.

2.3 Grounding indexicality

From a strictly Peircean point of view, indexicality is simply that property which makes something which is a sign into an index. However, by a slight shift of emphasis, which has at least some justification in Peirce’s work, it could be conceived as a property which, when added to the sign function, creates an index, but which, in addition, may have other parts to play in the constitution of meaning. That might account for the ambiguities of the Peircean notion, as well as for some of the uses to which it has been put subsequently.

Given the long period through which Peirce’s thinking evolved, and the state in which it came down to the public, it is not surprising that indexicality, like so many Peircean notions, should be so variously, and probably inconsistently, defined, and that many of the examples given hardly fit in with the definitions (also see Goudge 1965). Indexicality, in any case, pertains to the general category of Secondness, which means it concerns two items and/or the relation between them. The sign being a Third, there is every reason to think that it cannot be constituted by indexicality alone. Perhaps Peirce is really considering “potential sign-vehicles” in order to investigate their “capacity to serve as signs” (Bruss 1978: 87). More substantial arguments can be derived from a consideration of the Peircean concept of “ground,” as was suggested above (in Section 2.1.).

Generally put, an indexical ground, or indexicality, would then involve two “things” that are apt to enter, in the capacity of being its expression and content (“representamen” and “object” in Peircean parlance), into a semiotic relation forming an indexical sign, due to a set of properties which are intrinsic to the relationship between them, such as it is independently of the sign relation. This kind of ground, which is a relation, is best conceived in opposition to an iconic ground, which consists of a set of two classes of properties ascribed to two different “things,” which are taken to possess the properties in question independently, not only of the sign relation, but of each other, although, when considered from a particular point of view, these two sets of properties will appear to be identical or similar to each other. This is the sense in which indexicality is Secondness, and iconicity Firstness.

Such a view of indexicality as the one reconstructed above best fits in with the most general formulations given by Peirce, according to which it depends on there being a “real connection,” an “existential relation,” a “dynamical (including spatial) connection” and even, in one of its many conceivable senses, a “physical connection” between the items involved (CP 1.558, 1.196, 2:305, 3.361, 8.335). From this point, it seems natural to go on to argue that indexicality is involved with “spatiotemporal location” (Burks 1949: 683), which underlies the “indices” of such logicians as Bar-Hillel and Montague, the “egocentric particulars” of Russell and the “shifters” of Jespersen and Jakobson. In fact, however, as Savan (1976: 25) observes, location in time and space will only result, to the extent that some system of co-ordinates has been conveyed by other types of signs – or, as we would add, can be presupposed by the ongoing practice of the ordinary world of our experience.^[10]

More generally, many of the examples adduced by Peirce would justify us in going along with Jakobson (1979), when he claims that indexicality is based on “real contiguity,” and is connected with the syntagmatic axis of language, and the rhetorical figures of metonymy. To Jakobson, however, metonymy actually involves, not only the relation of contiguity of traditional rhetoric, but also that of part to whole, known in rhetoric as synecdoche. This distinction may be re-established inside the category of indexicality (see Nöth 1975: 20–21), and could be described more generally in terms of contiguity and factorality (cf. Sonesson 1989: 40).

There is, however, another series of definitions which suggest that indexicality is, is some way, dependant on there being a relation of causality between the expression and content of the potential sign: that is, the index supposedly “denotes by virtue of being really affected by that object” (CP 2.248). Apart from this, Peirce also makes a number of other claims about indices, many of which are repeated by Dubois (1983: 48–49, 60) when trying to demonstrate that photographs are indices: that they refer to unique, singular objects (CP 2.283); that they testify to the existence of its object (CP 2.316); and that they show up the object without asserting anything about it (CP 3.361); and that they point, by “blind compulsion” to the object of reference (CP 2.306). All these generalizations do not seem to hold true for many of the examples suggested by Peirce himself, as I have shown elsewhere (see Sonesson 1996).

Although the definition by causality is probably the most commonly quoted of all the definitions Peirce offers of indexicality, it has come in for serious criticism. Some commentators would reject the relation between causality and indexicality altogether, while others would see it as merely coincidental. Burks (1949: 649) takes Peirce to task for confusing the semiotic relation with mere causality, when treating, for instance, the weathercock, which is causally affected by the wind, as an instance of indexical signs: it is not clear, however, why causality should preclude indexicality, since the fact of the wind causing the weathercock to turn must be seen by the observer to be a contiguity in order for it to receive an interpretation.

More to the point, Goudge (1965: 55) claims that not all examples of indexical signs given be Peirce are susceptible of receiving a causal explanation: The Pole Star, for instance, may be an index of the north celestial pole, but it is in no way caused by that astronomical location. Nor is a personal pronoun, or even a pointing finger, actually caused by the person or thing for which it stands; and if they may be said to motivate it, then this is also true of all other signs. Moreover, if could be added that even some cases which are often taken to confirm the causal explanation are actually doubtful: the causal agent may not be that which is signified, or may not signify in the same respect in which it is the cause. Of all the innumerable causes that have to concur in order for a rap on the door to occur at a particular moment, the door and the material of which it is made, and a particular person and his moving hand may seem to be the most important. However, if, at this moment, no person in particular is expected, the sign will only carry some very general meaning such as “there is somebody (probably a human being) outside the door who wants me to open it and let him in.” Nor the particular person, nor his hand or the door, which are the causal agencies, are here parts of the meaning of the sign (Sonesson 1989: 39).

The idea that indices must point to their object by “blind compulsion” could be taken as a special case of causality, this time applied to the interpreter, and thus more properly described as motivation. Greenlee (1973: 86) believes this to constitute a contradiction on the part of Peirce, since the interpreting mind is on the level of Thirdness, and thus lies outside the definition of indices, which derives from Secondness. It seems, however, that the contradiction, if there is one, should be located at another point, for already the “immediate object” must (perhaps contrary to the “dynamical object”) be a mental unit. There is certainly an extremely Pickwickean sense in which all indices force us to attend to their objects, but in that sense the observation applies to all signs, and even to other kinds of meaning.

Nevertheless, causality seems to be the definition which fits best with the idea of indexicality being Secondness, that is, reaction, resistance, and even resistance to resistance. When the door you try to open resists, and you put your shoulder against it (EP 1, 268), it seems that a least the latter part of the action cannot be understood in purely causal term, if we, as is customary, restrict the latter notion to the outright physical domain. Of course, such a restriction may not apply to Peirce’s notion of causality, since his world is not clearly divided into physical and mental domains. If causality also includes what is usually termed “purpose,” “motive” or “intention” (in the everyday sense, not that of Husserlean phenomenological), it could no doubt serve to account for the shoulder put against the door, but hardly for the Pole star.

If we return to Peirce’s (CP 7.524–7.538) definition of Secondness as consisting in “dyadic experiences, or recurrences, each a direct experience of an opposing pair of objects,” and if the ignore “recurrence,” which seems to start us out on some quite different path, there also seems to be a case in between reaction and mere contiguity, something similar to the classical structuralist opposition. We can think of the indexical ground (and also the iconical ground, but not, of course, iconicity as such) as being equivalent to what Husserl terms couplings or pairings. In Husserl’s (1939: 174, 1950: 238) parlance, items which are co-attended form a paired association, or a coupling, when both items are directly present; they are an appresented pairing, or simply an appresentation, when one of the items is present and the other is not; and an appresentation becomes a sign when it is the absent item which is the theme. As I have pointed out elsewhere (see Sonesson 2012), these couplings may coordinate two items bound together be similarity or opposition, as well as by contiguity (and then we recover the relations underlying classical rhetoric which have been spelled out in the structural rhetoric built up by Groupe µ 1992). To account for this in Peircean terms, we might want to say that the opposition is a degenerate version of the reaction, and the contiguity in turn is a degenerate version of the opposition. Still, such a negative formulation leaves us wondering what is really making up the difference.

As discussed above, indexicality emerges as a potential sign, or, better, as a particular kind of ground characterising indexical signs, but which may also be found outside signs. Perception would seem to be profused with indexicality. Indeed, proximity is a basic factor of perception according to Gestalt psychology, and is also one of the relationships included in topological space perception. The relation of part to whole is fundamental to Gestalt relations themselves. All indexical relations involve either contiguity or factorality. Those indexicalities which are not as yet signs, being based on items which are not situated on different levels of directness or thematization, or not clearly differentiated, may be described as contexts (or ‘pairings’, in Husserl’s sense). Any experience of two elements being related by proximity, conceived as a primordial perceptual fact, may be considered an actual perceptual context involving contiguity. An actual perceptual context involving factorality is any experience of something as being a part of a whole, or as being a whole having parts (cf. Sonesson 1989: Ch. 1.2.5).

When only one of the items is directly given, and the other precedes it in time, or follows it, we may speak of an abductive context (protention and retention, respectively). The term abduction is employed here in Peirce’s sense, to signify a general rule or regularity which is taken for granted and which links one singular fact with another. All experience taking place in time is of this kind, for instance our expectancy, when seeing the wood-cutter with the axe raised over his head, that on the following moment, he is going to hit the piece of wood (contiguity protention), and on the moment just preceding, he lifted the axe to its present position (contiguity retention). Abductive contexts involving factorality would be, using some Peircean examples, the gait of the sailor, the symptom as part of the disease, part and whole in a picture, the partly destroyed Minoan fresco, a jig-saw puzzle, a piece of torn paper (the last three examples combine factorality and contiguity). We may use the term proto-index for an indexicality which is only momentarily a sign, as would be the “tableau vivant” of the wood-cutter, the photographic pose (which is a limitation in time), that what is seen in the view-finder (with spatial limits), and indeed many of the examples given above, to the extent that the flow of indexicalities is momentarily halted. The archaeologist’s art, from this point of view, would consist in transforming indexicalities of decayed cultures into proto-indices accessible to us.

2.4 Grounding symbolicity

Contrary to the icon and the index, on Peirce’s view the symbolic sign would seem to be literally groundless. Or perhaps it would be more correct to say that its ground coincides with the sign function. In the case of the index, Peirce stipulates that the connection between its relata must obtain independently of the sign relation, and in the case of the icon, he adds that iconicity must also be independent of possible relations between the relata as such. In the case of the symbol, however, there is no comparable stipulation. Nevertheless, if we follow up on Peirce’s late insight that what he had hitherto called the sign is really something much vaster, which we could perhaps call mediation, as Peirce does, or simply semiosis, the sign function is simply one of many possible kinds of Thirdness.

Elsewhere, taking my inspiration from Husserl and Piaget, I have suggested that we can minimally define the sign by the following properties (Sonesson 1989, Sonesson 1992, Sonesson 2011, and Sonesson 2013):

1. it contains (a least) two parts (expression and content) and is as a whole relatively independent of that for which it stands the referent;

2. these parts are differentiated, from the point of view of the subjects involved in the semiotic process, even though they may not be so objectively, i.e. in the common sense Lifew, (except as signs forming part of that Lifeworld);

3. there is a double asymmetry between the two parts, because one part, expression, is more directly experienced than the other;

4. and because the other part, content, is more in focus than the other; and

5. the sign itself is subjectively differentiated from the referent, and the referent is more indirectly known than any part of the sign.

Perhaps this definition is not sufficient, but it will at least separate out a smaller class of phenomena within the big category of Thirdness. And it will allow for the fact that, as iconic and indexical signs are based on pre-existing iconic and indexical grounds, some symbolic signs may rely on some kind of Thirdness (rules or regularities) instituted prior to the sign. It should not be difficult to find examples of Thirdness which is not of the type of the sign: traffic rules, rules of polite behaviour, the rules of chess, etc. Indeed, some iconic signs may be icons of symbolic properties which themselves do not form signs. Thus, for instance, there is an example from the Sign language of the North American Indians, mentioned by Garrick Mallery (1972 [1881]): the sign meaning “woman” is made up of gestures showing the braids on both sides of the head. Here the sign simply depicts the convention in North American Indian society for how woman should fashion their hair.

This may be the place to admit that there is something unsatisfactory with the account given of Thirdness summarized in Figure 1. It amounts to construing Thirdness as being some kind of meta-position, and while this fits well with half the quotations from Peirce (in Table 1), its relation to the other half is, to say the least, unclear. One is reminded of the notion of categorical (or abstract) attitude characterized by Gelb and Goldstein in opposition to the concrete attitude (see Gurwitsch 1966). The former is a perspective based on generalities and essentials rather than specifics. It parses a situation into properties, isolates them, and makes meaningful projections. The concrete attitude is certainly also necessarily for leading a full life, but in healthy people, both attitudes can work together and adapt to circumstances. According to Gelb and Goldstein, many psychiatric disorders are characterized by an inability to adopt an abstract attitude or shift readily from the concrete to the abstract and back as required by circumstances. If we identify Thirdness with the categorical attitude, it seems that both Firstness and Secondness would have to pertain to the concrete attitude. The problem with this explication is that Gelb-Goldstein’s attitudes are themselves rather fuzzy notions.