Natural signs from Plato to Thomas A. Sebeok

2.1 Greek Antiquity

The distinction between the signs uttered by humans and those that we encounter elsewhere in nature, including the symptoms of human diseases, was sharp in Greek Antiquity, since the Greek language distinguished these two kinds of signs in its vocabulary. A word, verbal utterance, or proposition was a semaînon, whereas a sign of nature was a semeîon (Meier-Oeser 1997: 2–6) or sêma, which is a ‘mark,’ ‘characteristic,’ ‘indication,’ an ‘omen,’ or ‘portent’ (cf. Castañares 2014: 16). From the Ancient Greeks to the Stoics, natural and conventional signs remained lexically distinct in this way, although the difference was lost or ignored in Latin translations in which both semaînon and semeîon were translated as signum or nota (cf. Eco 1990: 114). What a semaînon had in common with a semeîon is that both mean (semaínei), but while a (natural) semeîon ‘gives evidence of,’ or ‘reveals,’ some fact or state of affairs, the meaning of a (verbal) semaînon was an immaterial content (Meier-Oeser 1997: 2–6).

Plato did not yet have a theory of natural signs, but in his dialogue Cratylus, the question whether words, in relation to their meaning, are originally natural or conventional was the topic of debate between Cratylus, Hermogenes, and Socrates. The dialogue was not about natural signs in general but about vestiges of naturalness in verbal signs. Of special interest in this dialogue is that the vestiges of nature that Cratylus, Hermogenes, and Socrates discussed were not indexical but iconic sign. Words (called “names” in this dialogue) are discussed as being associated with their objects of reference in a natural way in the sense that the former evince an iconic relation with the latter. While Hermogenes is convinced that words and their objects of reference are associated by convention and agreement, Cratylus defends the view that they are originally associated by “natural likeness.” For example, the glottal movements in the articulation of the phonetic expression [r] imitate the movement of something that “moves,” whereas the sound of [o:] (whose letter name, in Greek, is omega, ‘big o’) is an “icon” of something “really” big (Cratylus 432d). The study of names cannot only reveal their “natural correctness” (383a); since “the names of the things depict their essence” (424a), they can also disclose the essence of the things. The study of names hence enables him or her who knows the names to know the things (435d). Mediating between both parties in this debate, Socrates concludes in support of Cratylus, “representation by likeness is infinitely better than representation by any chance sign” (434a), but he also agrees with Hermogenes in concluding that names are given by a “legislator” who is the “name giver.” Since so many words are no longer icons of what they mean presently, Socrates then advises both Cratylus and Hermogenes to search for the essence of the things in the things themselves and not their names.

Aristotle includes reflections on the nature of the speakers’ and the hearers’ minds in his search for vestiges of nature in language. “Spoken sounds (phôné) are symbols (sýmbola) of affections (páthêmata) in the soul,” he declares in On interpretation (16a, 3–8), before he goes on to distinguish between natural and conventional signs: “A name is a spoken sound, significant by convention. […] I say ‘by convention’ because no name is a name naturally but only when it has become a symbol. Even inarticulate noises (of beasts, for instance) do indeed reveal something, yet none of them is a name” (16a, 26–28). Aristotle hence postulates an overlap between the signs of human culture and the ones of animal nature. His arguments have been much discussed by his commentators (Castañares 2014: 70; Eco 1990: 113–114; Pépin 1985). Not only humans but also some animals produce “vocal sounds” (phoné) that have or “reveal” meaning. The difference is that the sounds produced by humans are “articulate,” i.e. structured, whereas the ones of nonhuman animals are “inarticulate” (On interpretation, 16a, 28). The distinction between the acoustic utterances of human and nonhuman animals became the topic of countless treatises on signs since, of which Eco (1990) summarized the ones of Augustine, Boethius, Abelard, Aquinas, and Roger Bacon as follows.

Augustine of Hippo (354–430) did not distinguish between natural and conventional signs, but between natural and given signs (signa data), which he defined as “signs created expressly with a view to communication” (Manetti 1987: 166–167; cf. Gramigna 2010). As Augustine put it, “Given signs are those which living things give to each other to show, as best they can, their emotions, or something perceived or understood […] to express and transmit to the mind of another that which is in the mind of the one who gives the sign” (Doctr. Chr. II.3). About the senders of these signs, Augustine wrote, “Some animals too have signs among themselves. A cockerel, finding food, gives a vocal sign to its hen to hurry along” (Doctr. Chr. II.3). (Today we know that the rooster’s crow is not actually a signum vocis since birds produce sounds through an organ called syrinx, not by means of a voice.) Given signs are always intentionally produced by human or nonhuman animals. In contrast to given signs, which have an animate sender, natural signs that are not “given” may have either animate or inanimate sources. Augustine pays much attention to natural signs in animate as well as in inanimate nature:

When we see a footprint, we think that the animal whose footprint it is has passed by. When we see smoke, we realize that there is fire beneath it. When we hear the voice of an animate being we note its feeling; and when the trumpet sounds soldiers know they must advance or retreat or do whatever else the state of the battle demands. […] Natural signs are those that without a wish or any urge to signify cause something else besides themselves to be known from them, like smoke, which signifies fire. It does not signify fire because it wishes to do so; but because of our observation and attention to things that we have experienced it is realized that there is fire beneath it, even if nothing but smoke appears. The footprint of a passing animal also belongs to this category. The expression of an angry or depressed person signifies an emotional state even if there is no such wish on the part of the person who is angry or depressed, and likewise any other emotion is revealed by the evidence of the face even if we are not seeking to reveal it. (Doctr. Christ. II.1–2)

With intentionality as the distinctive feature of given signs, Augustine introduced a new semiotic paradigm, which broke with the Aristotelian tradition of separating the signs of humans from those of the rest of nature by the criterion of conventionality. Thomas Short contextualizes the historical importance of this redefinition as follows:

Augustine thus redrew the fundamental division of signs: instead of being between the natural and the conventional, it was between the natural and the purposeful, with conventional signs being a subset of those signs that occur for a purpose. Peirce’s idea of ‘legisigns’ and their replicas is exactly Augustine’s of signa data, and it applies to the same range of examples. Peirce, however, did not conceive of nature as that to which the purposeful is opposed. (Short 2007: 26)

2.2 Scholastic semiotics

Boethius (ca. 480–525) subdivided signifying sounds (voces significativae) into natural ones and those “by convention” (ad placitum). Peter Abelard (ca. 1079–1142) classified animal voices further by distinguishing between those that signify either through institution or without (ex vs. sine institutione). The former are “significative signs” (voces significativae) due to a “natural intention” of the animal. The latter are only “signifying signs” (voces significantes) in the sense that they indicate the presence of a sign emitting animal. Thomas Aquinas (1225–1274), while drawing the same distinction, calling the latter ex institutione, extended this subdivision to include also nonvocal sounds, such as the ones of musical instruments.

Roger Bacon (ca. 1219–1292), especially in his treatise De signis, has the most elaborate classification of signs among the Scholastics (Bacon 1267; Howell 1987; Maloney 1983). For Bacon, too, natural signs are not simply signs from a sphere of nature that would exclude human beings. Instead, they are signs “by their own essence” (ex essentia sua et non ex intentione animae), emitted without a sender’s intention in contrast to intentional signs “ordered by a mind” (signa ordinata). However, intentional signs also include, as a subclass, signs emitted naturally, although in an irrational way (sine deliberation rationis) through a natural impulse. Such signs may be human or nonhuman ones (e.g. sighs, groans, or laughter vs. the vocal signs of animals). The most important class of intentional signs are the conventional ones (ad placitum), produced by human beings only. This class includes verbal signs as well as nonverbal ones, such as the gestures used by monks or the kind of signs by which merchants advertise their goods.

Furthermore, Bacon subdivides natural signs proper into three classes. The first two consist of indexical signs in modern semiotic terminology, (1) signs by necessary or probable inference, such as the dawn from which we infer that the sun is rising, and (2) signs related to their object by a natural cause, such as the stag’s track in the sand or smoke as a sign of fire. As Maloney (1983: 137) points out, Bacon later abandoned the distinction between these two, reducing them to only one class of natural signs of the indexical kind.

Bacon’s third class of natural signs contains those that modern semiotics defines as icons, i.e. natural signs by likeness. These signify through “conformity and configuration” between the sign and its significatum (signum propter conformitatem unius rei ad aliud). Examples are “images, pictures, and the species of colors, tastes, and sounds” (De signis §5). To find paintings and pictures among the natural signs may surprise modern semioticians, who have become accustomed to definitions of images that emphasize their conventionality and cultural relativity (Eco 1976: 191–217). However, even though an artist must have produced the portrait of an emperor intentionally, such a portrait is still a natural sign since it “derives the ability to signify and to represent” its object through its configuration and hence by its own essence independently of the artist’s intention. “Naturalia autem dicuntur, quia ex essentia sua et non ex intentione animae signi rationem recipient” (De signis §3; cf. Maloney 1983: 135; Meier-Oeser 1997: 58; Tabarroni 1989: 198).

Although words and other verbal signs differ from natural signs insofar as they are intentional, conventional, ordered by a (human) mind, and uttered with a rational purpose, Bacon postulates several secondary ways in which verbal signs are also natural signs. One is because there is a relation of iconicity among the members of the class of objects to which a concept refers. If a concept serves the purpose of referring to a class of objects, these objects must evince a natural similarity among themselves before the given concept can be applied to them as a class:

The concept of horse has a different content from the concept of pig because the former is more similar to horses while the latter is more similar to pigs. This points to two important kinds of similarity that should be kept distinct: (1) the similarity that holds between a concept and what it is a concept of, and by virtue of which a concept is called a similitude or likeness; (2) the similarity that obtains between two individuals “of themselves,” i.e., two human beings or two stars. Ockham clearly believes that reality is carved up into kinds of individuals that just are similar to varying degrees. That reality is so-structured is an accepted metaphysical fact. (Pelletier 2013: 89)

Bacon distinguishes two more secondary respects in which concepts are natural signs. The first is related to what Ferdinand de Saussure (1916: 104), centuries later, called the “immutability” of the verbal sign: although a sign is arbitrary and conventional in relation to its meaning, no individual can change the relation between its signifier and its signified, once it has been established (or “imposed”). In this respect, Bacon postulates a causal and hence naturally indexical relation between the language of mind (oratio mentalis) and the things outside it (res extra animam) (Lambertini 1989: 126). Tabarroni (1989: 200) gives two reasons why this relation was considered natural. The first is associated with Aristotle’s claim, in De interpretation (16a, 3), that the concepts or “affections in the soul” associated with spoken sounds uttered in whatever languages “are the same for all.” The second is because concepts are signs “in virtue of the objective relations that characterize them as independent entities even before characterizing them as signs” (Tabarroni 1989: 206). Hence, concepts are caused by the object to which they refer and are therefore natural indices of them.

In addition, a concept was considered a natural sign because the mental image associated with it was considered an icon (similitudo) of the object it represents. In his treatise De multiplicatione specierum, Bacon interprets the way in which a mental representation is an iconic sign as follows: “Objects produce or generate a force (virtus) which is a likeness or image of the substance or proper sensible that generates it. […] Substances and accidents generate their own distinct species, and these in turn transform the sensory and intellectual faculties in which they are received” (Maloney 1983: 129).

The theory of natural signs was not an issue that divided the Realists (such as Bacon) from the Nominalists among the Scholastics. The Nominalist William of Ockham (ca. 1280–1349) interpreted concepts as natural signs, too. In his seventh Question on the Physics of Aristotle, he argues: “When the intellect apprehends a singular thing by intuition, it forms in itself an intuitive cognition […]. And just as a spoken word conventionally supposits for its significate, similarly this [intuitive] intellection naturally supposits for the thing it is an intellection” (Panaccio 2004: 12).

Peirce was familiar with the Scholastic theory of concepts as natural signs in contrast to words as conventional signs. In 1871, he quotes Ockham as one who thought

of a mental conception as a logical term, which, instead of existing on paper or in the voice, is in the mind, but is of the same general nature, namely, a sign. The conception and the word differ in two respects: first, a word is arbitrarily imposed, while a conception is a natural sign; second, a word signifies whatever it signifies only indirectly, through the conception, which signifies the same thing directly. (CP 8.20)

The history of the notion of the natural sign has many other chapters, which deserve more detailed research than can be presented here. Some of them have been outlined by Meier-Oeser (1997). Ayers (1986) has shown that the issue of the iconicity of concepts had a continuity until John Locke’s representational theory of ideas. For the period from Paracelsus to Novalis, Gaier (2001) has presented valuable insights. A comprehensive update on the issue can also be found in Lyons (2019). Further historical considerations must be left aside here. Instead, our focus is now on writings on natural signs by authors who were Sebeok’s contemporaries. We will first consider authors who approached the topic from various philosophical perspectives and then writers in the field of semiotics proper. Here, our discussion is restricted to papers not within the scope of Sebeok’s biosemiotics (including Jakob von Uexküll’s Umweltlehre), phytosemiotics, mycosemiotics, endosemiotics, or ecosemiotics (cf. Nöth 2001, 2008a) since contributions to these field of research directly influenced by Sebeok have been examined elsewhere in this issue as well as elsewhere (Cobley et al. 2011; Danesi 2001; Petrilli and Ponzio 2001).

3.1 Natural versus conventional

In 20th century analytical and cognitive philosophy, studies of natural signs in the tradition of the natural–conventional dichotomy were relatively rare. One of the few monographs with a historical perspective was published on Sebeok’s initiative in his legendary book series Approaches to semiotics in 1976. Its author, Bernard E. Rollin, then supervised by none less than Arthur C. Danto, has since become known for remarkable contributions to issues such as animal consciousness and animal rights. Despite its title, Natural and conventional meaning: An examination of the distinction, Rollin’s book is really on natural in contrast to conventional signs. Unfortunately, the misnomer “meanings” in the title indicates a lack of semiotic orientation. Nevertheless, Rollin offers a rather informative survey of philosophical views on the dichotomy between natural and conventional signs since the Presocratics with special attention to the Medievals and to John Dewey, before Rollin concludes by reaffirming the usefulness of the class of natural signs.

In the course of the next decade, Eike von Savigny (1977) and Jerzy Pelc (1988) offered two entirely ahistorical contributions to the issue with the aim at clarifying the meaning of the concepts of natural and conventional in ordinary language. Savigny surprises by including chance constellations, such as Rorschach blotches, preferred lottery numbers, or playing cards drawn by clients of fortunetellers in the class of natural signs besides giving better examples, such as weather signs. Pelc’s paper reveals that “natural” can mean ‘without any human intervention,’ as in the case of the lightning that is a sign of thunder, but also ‘with some human intervention,’ as when honey is called a natural product, although a human beekeeper was needed to produce it. The adjective “natural” does not exclude human participation entirely, also because humans show natural signs of aging, have unconditioned reflexes, and so on. Other questions concerning the meaning of “natural”: Is a technical defect that makes a watch stop natural or artificial? Is unintentional behavior natural or not? Apparently, there is no clear-cut division between nature and human culture. Instead, a set of well-defined criteria is needed.

3.2 Natsigns

In the tradition of Analytical Philosophy, David S. Clarke proposed a theory of natural signs in the third chapter of his book Principles of semiotic of 1987. Although the author dedicated this book to his teacher Charles Hartshorne, one of the editors of Peirce’s Collected papers, it shows almost no influence of Peirce’s semiotic. Clarke is not concerned with signs of, or in, nature, but with how humans interpret signs of natural phenomena. Based on a distinction between “long-range” and “short-range” interpretations of natural events, the author distinguishes two types of natural signs: first, signs interpreted in consequence of internalized linguistic generalizations of formerly experienced natural events; second, signs interpreted “directly” and without the mediation of generalized knowledge. By this distinction, Clarke is convinced of having found an explanation for the difference between black clouds interpreted as signs of rain (that comes much later) and a flash of lightning experienced as the sign of the following thunderclap (Clarke 1987: 44–45). Clarke defines the latter type of natural sign, the one that is interpreted without the mediation of verbal reasoning, as the most logically primitive sign of all and as the prototype of a natural sign. The author concludes that “only those events or objects whose interpretation does not require linguistic mediation are to be included in this restricted range [of natsigns], that is, those which directly signify on the basis of past correlations in experience between tokens of the sign and what it signifies” (1987: 50). Introducing the neologism of natsign for this class of natural signs, Clarke arrives at the following definition: “A natsign […is] an event having significance for an interpreter, which is not produced for the purposes of communication and whose interpretation does not require an inference from a linguistic generalization. Natsigns are thus contrasted with communicated signs, signs used and interpreted in the process of communication” (1987: 50).

Clarke’s theory of the two kinds of natural signs, distinguished with respect to the length of the time lapse between the interpreters’ perception and their interpretation of the sign, can be better explained in terms of Peirce’s semiotics of indexicality. Clarke’s natsign is simply a genuine index, whose characteristic, according to Peirce, is a “direct dual relation of the sign with its object” in “the mind using the sign” (CP 3.361, 1885). The other sign that Clarke defines as natural, the sign with a “long-range” relation between perception and interpretation, is not an indexical sign at all. It is a symbol, more precisely an argument, since its characteristic is that the interpreter, at least in the way Clarke describes it, interprets it in a process of reasoning using the method of deduction to arrive at his conclusion about what has happened. The fact that this conclusion is, so-to-speak, imposed by nature onto the interpreters’ mind is an element of Secondness in the Thirdness of this process of reasoning.

The interpretation of concepts as natural signs has also found some support in 20th century analytical metaphysics. With explicit reference to Ockham, Laird Addis (1983, 1989 proposes a theory of Natural Signs in which he argues that any thought “is, or contains as a constituent, an entity that is a natural sign” (1983: 345), an entity that “unlike linguistic and other conversational representing entities, intrinsically or by their very nature intend whatever they intend” (Addis 2010: 17). Natural signs, according to Addis, are signs of properties or relations, which are natural because we simply cannot conceive of them otherwise. For example:

The simple property of ‘being-scarlet’ cannot be conceived as having the property of ‘being-a-shape’ or ‘being-a-pitch’: its intrinsic character as being just the first-level property it is prevents it from exemplifying these particular second-level properties. The same goes for relations. Thus, it would appear that some properties or relations, by being just the properties or relations that they are and not some other properties or relations, might be natural signs, and indeed that only a property or a relation could, at ontological base, be a natural sign. (Addis 1983: 549)

3.3 Teleosemiotics

The study of natural signs has enjoyed an unexpected revival in a contemporary trend in the philosophy of mind that Ruth Garret Millikan and others have been pursuing under the designation of teleosemantics. Here, natural signs are not opposed to conventional ones but to intentional signs. However, instead of opposing natural and intentional signs, teleosemantics emphasizes what both have in common. Millikan’s teleosemantics is rooted in the theory of biological evolution. Ultimately, all signs, intentional verbal signs included, evolve from natural signs already used by nonhuman animals, but while the signs of animals are genetically replicated, verbal signs are replicated in the form of tokens reproduced from “some earlier token or tokens of its type that the speaker has heard” (1984: 22). Words are replicated from precedents and conventions “thought to have little tendency to emerge or reemerge in the absence of precedent” (2005: 7). Language conventions are “lineages of behavioral patterns” repeating precedents of successful communicative interactions (2005: 86). In the long process of their replication, cultural signs “never lose their character of being also natural signs” (2004: 105). A major purpose of all signs is to replicate in their usage. While the natural signs of animals survive genetically, words and syntactic patterns survive as semiotic types reproduced in the form of tokens (2005: 3). All verbal forms have their own purpose, which is to survive (1984: 28). Not restricted to a theory of verbal signs, Millikan’s teleosemantics is a general theory of signs (1984: 85), hence a semiotics, which explains the replication of genetically transmitted sign behavior of animals as well as the proliferation of artifacts in culture.

4.1 Greimas’s natural semiotics

A semiotician who conceived of a theory of natural signs in Saussure’s logocentric tradition is Greimas, first in his much-quoted paper “Toward a semiotics of the natural world” of 1968 (Greimas 1970), later in his articles on natural signs and the semiotics of nature in Greimas and Courtés (1982: 211, 374–375). “Natural signs refer back to things other than themselves, but this referential relation […] has different articulations, or variables, according to the cultural communities envisaged,” argued Greimas (1970: 20). The natural world, according to these premises, is merely “a place for the elaboration and practice of multiple semiotic systems” (Greimas and Courtés 1979: 375). Semiotics of nature, thus conceived, is a theory of how human culture interprets nature. Natural signs are thus nothing but cultural signs. They can never be truly “natural” since nature is always already “culturalized” whenever we take notice of it, because any individual “is inscribed from birth within a signifying world made up of both ‘nature’ and ‘culture.’ […] Nature, therefore, is not a neutral referent; it is strongly culture-bound” (1979: 374).

4.2 Barthes’s semiotization of nature through culture

Equally founded on the Saussurean dogma of the arbitrariness and conventionality of signs was the method of excluding nature from culture adopted by Roland Barthes since his Mythologies of 1957. Although Barthes, in these essays, did not develop a definition, let alone a theory, of natural signs, the nature of the natural is nevertheless their central issue. Barthes is not concerned with signs of nature proper, but with cultural discourse about nature, especially in the mass media of his time. His thesis is that in today’s media as well as in material culture, nature has disappeared or it has become a mere pretext for attributing a surplus value to commercial goods. Thus, whenever nature is referred to, it is already affected and transformed by culture. Not without a touch of nostalgia, Barthes deplores, for example, the substitution of natural wood for artificial plastic in children’s toys: “Current toys are made of graceless material, the product of chemistry, not of nature. […] A sign which fills one with consternation is the gradual disappearance of wood, in spite of its being an ideal material because of its firmness and its softness and the natural warmth of its touch” (Barthes 1957: 54). While nature has disappeared in culture, the media seek to reintroduce it in the form of everyday myths that aim at “naturalizing” culturally created objects. Under such premises, Barthes can write about nature only in inverted commas. After all, whatever is declared a sign of nature is only a disguised sign of culture, a myth. “What causes mythical speech to be uttered […] is immediately frozen into something natural” (Barthes 1957: 128).

4.3 Eco on natural as conventional signs

In his early writings, still strongly influenced by the structuralists, Umberto Eco was equally skeptical against admitting nature into the semiotic field of research, but his argument for excluding nature was different. His method was to admit nature, but only, so-to speak, very briefly. Consider how Eco (1976: 10), in his Theory of semiotics, excludes the class of symptoms from that of natural signs. According to his account, symptoms cease to be natural signs once they have been studied in medical treatises. By means of such treatises, which are conventions, the same signs change from the class of natural to that of conventional ones:

The first doctor who discovered a sort of constant relationship between an array of red spots on the patient’s face and a given disease (measles) made an inference: but insofar as this relationship has been made conventional and has been registered as such in medical treatises a semiotic convention has been established. There is a sign every time a human group decides to use and to recognize something as the vehicle of something else. In this sense, events coming from a natural source must also be listed as signs: for there is a convention positing a coded correlation between an expression (the perceived event) and a content (its cause or its possible effect). (Eco 1976: 17)

As signs that have become conventions through the doctrines of medical schools, symptoms lose their naturalness because culture, here the medical academy, has semiotized nature as soon as it interprets it. Even when interpreters infer fire from smoke, rain from a puddle, or the passing by of an animal from the trail that it has left, they do not interpret signs (Eco 1976: 17). Only when a cultural rule intervenes between the natural cause and its effect may a phenomenon be interpreted as a sign. “Our everyday life is filled with a lot of these inferential acts. It is incorrect to say that every act of inference is a ‘semiosic’ act — even though Peirce did so” (1976: 17). It is precisely on this subject that Eco, under Peirce’s influence, revised his semiotic theory in the 1980s. In 2017, inference even advances from a nonsemiotic mode of reasoning to the distinctive feature of sign interpretation in general:

It was in speculating on the Stoics’ notion of semeîon that I understood (through the Peircean theory of interpretation) that a sign is never equivalent to its interpretants. This relation is a relation of inference. The Stoics applied the notion of inference to natural signs (if smoke, then fire) but I tried to demonstrate that there is inference even when we interpret conventional signs, such as words, for instance. This becomes more comprehensible when I deal with the concept of encyclopedia. In any case, the way to represent the meaning of a sign-vehicle x is “if x then, according to the following circumstances and contexts, the following interpretants can be activated.” (Eco 2017: section x)

5.1 Indices

The index is a class of signs that Peirce often exemplifies by natural signs. In a genuine index, there is “a direct dual relation of the sign to its object, independent of the mind using the sign.” Specifying that an index “signifies its object solely by virtue of being really connected with it,” Peirce concludes: “Of this nature are all natural signs and physical symptoms” (CP 3.361, 1885). Peirce also considers an index without an animate source (such as a weathercock indicating the direction of the wind) as a “natural sign.” In contrast to a sign uttered by an animate being (CP 8.348, 1908), these natural signs “have no author” in their “direct physical connection” with their objects (CP 1.372, ca. 1885). Other examples of such natural sings are a low barometer, which makes us “suppose that the forces of nature establish a probable connection between it with moist air and coming rain” or a weathercock veering with the wind, which forces us “by the law of mind to think that direction is connected with the wind” (CP 2.286, ca. 1893). While many indices are natural signs, not all indices are since there are also indices among the verbal signs, for example, demonstrative pronouns.

Peirce’s list of the ten main classes of signs of 1903 (CP 2.254–264) distinguishes four indexical ones, among which three are natural signs. Most typical ones are the indexical rhematic sinsign (e.g., a spontaneous cry or a flash of light that blinds out eyes) and the dicent indexical sinsign (e.g., the weathercock or a specific medical symptom that conveys information about direction of the wind or about a disease, respectively). The third kind of natural index, the indexical dicentic legisign, is a hybrid class since this class also includes artificial sings, such as the red lamp of the oil pressure gauge, and signs emitted by humans, such as the voice that identifies a specific speaker, besides the natural ones, such as a symptom of a disease. Medical symptoms are only indexical dicentic legisigns insofar as they indicate a disease as a type, for example, when doctors recognize the set of symptoms A, B, C, and D as signs of the coronavirus disease in general. Peirce’s distinction between the indexical dicentic legisign and the indexical dicentic sinsign offers a semiotically more differentiated way of taking account of natural signs than that proposed by David Clarke, Roland Barthes, and Umberto Eco, as discussed above. While these authors defined natural signs interpreted according to culturally codified rules and conventions as symbols, Peirce can call them natural signs and still differentiate between them by distinguishing between natural indexical sinsigns and natural indexical legisigns (instead of “symbols”). In 1903, Peirce denounced the attempt as erroneous to classify a symptom of a disease both as an index and as a symbol (as Barthes and Eco later did). At the same time, he admitted that he himself had once committed this error of classification: “I had classed natural symptoms both among indices and among symbols, I restricted symbols to conventional signs, which was another error. The truth is that my paper of 1867 was perhaps the least unsatisfactory” (CP 2.340).

Symbols are often taken to be the class of conventional signs par excellence. However, this is mistaken if symbols are defined as Peirce defined them. Not the criterion of conventionality but those of habit and generality are Peirce’s main criteria of symbolicity (Nöth 2010). However, habits are not only human behavioral tendencies. They can be found everywhere, also in nonhuman nature (Nöth 2016a). In a definition of 1902, Peirce states that a symbol is “a sign merely or mainly by the fact that it is used and understood as such, whether the habit is natural or conventional, and without regard to the motives which originally governed its selection” (CP 2.307). In 1904, natural signs are even more clearly subsumed under the class of symbols, when Peirce states that the symbol “depends either upon a convention, a habit, or a natural disposition of its interpretant” (CP 8.335). Thus defined, the species-specific signs by which animals communicate among themselves are symbols since they are habits of sign behavior, irrespective of whether the habit is genetically determined or acquired by learning. Peirce even suggests that symbols may be found in inanimate nature, when he describes the evidence found by a paleontologist in the study of a fossil fish as a symbolic argument contained in the very data, even before the information it contains is discovered by the scientist. If the evidence found in the course of this research

will one day furnish that paleontologist the keystone of an argumentative arch upon which he will securely erect a solid proof of a conclusion of great importance, then, in my view, in the true logical sense, that thought has already all the reality it ever will have […]. For the fish is there, and the actual composition of the stone already in fact determines what the chemist and the paleontologists will one day read in them: and they will not read into them anything that is not there already recorded, although nobody has yet been in condition to translate it. It is, therefore, true, in the logician’s sense of the words, although not in that of the psychologists, that the thought is already expressed there. (EP2: 455, 1909)

The triad of icon, index, and symbol classifies signs with respect to the relation between the sign and its dynamical or “real” object, i.e. “the Reality which by some means contrives to determine the Sign to its Representation” (CP 4.536, 1905). From this object, says Peirce, “an influence emanates upon the Sign and produces, and is capable of producing and partly at least, in a mental way, an effect that may be called an Interpretant” (MS 634: 23, 1909). Now, insofar as this reality, which determines the sign to represent it, is of the sphere of nature, we have evidence of some influence of nature on signs. Furthermore, there is a “real” and, in this sense, natural disposition in any sign to create an interpreting action, which is evidence of the reality of Thirdness (cf. Nöth 2016b). In the case of an icon and an index, the role of nature in determining the sign is more prominent than in the case of the symbol, since an icon “partakes in the characters of the object” and an index is “really and in its individual existence connected with the individual object” (CP 4.531, 1905). In the case of the symbol, the determination of the sign is due to a “more or less approximate certainty that it will be interpreted as denoting the object, in consequence of a habit (which term I use as including a natural disposition)” (CP 4.531, 1905). The influence of nature on the sign is less in the case of the symbol, but it is not altogether absent, not only because the habit may also be a “natural disposition.” Even when the habit is due to a convention the sign user has adopted, there is an element of naturalness in it. Peirce describes this element when he defines habit as “some general principle working in a man’s nature to determine how he will act” (CP 2.170, ca. 1897).

5.2 Natural concepts

For Peirce, a concept is a symbol and as such also a legisign: “When a philosopher speaks of the ‘concept’ of matter, or the ‘concept’ of cause, or any other ‘concept,’ what he means by a ‘concept’ is a word or other legisign” (MS 1476: 6, ca. 1904). Of course, for Peirce, concepts are not generally “natural signs,” as the Scholastics would have it, but for Peirce, they are formed in a way that involves some kind of iconicity, in the sense of a natural correspondence with their object domain. Peirce expresses this idea most clearly in his Grand logic of 1893:

A concept is […] the living influence upon us of a diagram, or icon, with whose several parts are connected in thought an equal number of feelings or ideas. The law of mind is that feelings and ideas attach themselves in thought so as to form systems. But the icon is not always clearly apprehended. We may not know at all what it is; or we may have learned it by the observation of nature. (CP 7.467)

In ca. 1907, Peirce reiterates this notion when he defines a concept as a mental sign by means of which the object it represents “is moulded into some sort of correspondence with it” (MS 322: 9, ca. 1907). With such ideas, Peirce has been acknowledged as a precursor of the theory of “natural kinds” (Bird and Tobin 2018), which Peirce himself called “natural classes” (CP 1.203, 1902; Hulswit 2002: 97–132).