Peirce, Trichotomies, and Emergence

In this post, I’m going to amplify my comment in the previous post that Peirce’s trichotomies are a type of stratified emergence.

In “A Guess at the Riddle” and “The Categories Defended,” Peirce sets forth a number of arguments defending his use of three categories: (1) Firstness, Secondness, and Thirdness are distinctive, such that Secondness isn’t simply a matter of two monads, and Thirdness isn’t just a combination of dyads; (2) anything higher (Fourthness, Fifthness, etc.) can be reduced to a combination of Thirdness; (3) looking the other direction, that Firstness and Secondness can’t be collapsed into Thirdness; yet at the same time, (4) Secondness contains an element of Firstness, and Thirdness contains an element of both Secondness and Firstness, and in this relation of containment, the lower categories are not present as mere parts, but as necessary parts.

I will not review Peirce’s categories and the arguments above; in this post I’m focusing instead in this conjunction of qualitative distinctions, irreducibility, and necessary containment. It is remarkably similar to the modern concept of emergence, according to which a new entity (property, structure, etc) develops out of preexisting, more basic ones. The emergent entity has causal powers that could not be predicted from its constituents and cannot be reduced to them. It depends on its constituents for its existence, but it’s also capable of “acting back” on them, such as by setting boundary conditions or controlling their activity (downward causation).1 New powers mean new “levels” of entities: emergence leads to the stratification of reality. The most basic is the physical stratum, which has its own strata, e.g., sub-atomic particles, atoms, molecules, and four fundamental forces. From the physical stratum emerged the biological stratum. Biology depends on the workings of the physical stratum, but it also does new things, like reproduce itself and evolve. From the realm of biology emerged creatures able to interpret signs in their environment and produced among themselves; humans are the most sophisticated of these species, but it is scarcely alone in possessing some form of semiosic capacity. But the semiotic relationship—meaning per se—isn’t reducible to biology, it’s a different kind of power. (Obviously reductionists disagree.)

That example is diachronic, but once the entity has emerged, its stratification continues synchronically: semiosis still depends on and contains a biological substrate and biology still depends on and contains physical processes as long as that creature lives. Emergence doesn’t only apply to entities (ontically); as I’ve argued previously, it also characterizes ontology, as conceptualized in critical realism.2 The domain of the real is that of powers and generative mechanisms existing as substantive but (so to speak) static possibilities or potentialities. When they interact, concrete events and manifestations arise, constituting the domain of the actual. Of these, a subset are perceived, experienced or detected, which for Bhaskar constitute the empirical domain. He adjusted that later to the subjective domain, but I argue that it should be conceptualized more broadly (and less anthropocentrically) as the semiosic domain.

The three domains clearly parallel Peirce’s three categories (see e.g. CP 1.23-24, 1.325, 8.216). More, Peirce conceptualized the three categories as having a line of dependency that is fully consonant with emergence:

Thirdness as such . . . can have no concrete being without action, as a separate object on which to work its government, just as action cannot exist without the immediate being of feeling on which to act. (Essential Peirce, vol. 2, 345)

The categories, then, describe an emergent relationship.

However, Peirce’s categories more emphatically incorporate action and process than critical realism (and the latter’s is far from weak). I say this because Peirce constantly underscores the active, processual, evolutionary character of semiosis. This is undoubtedly best known through his argument that every sign becomes—or better, has the potential to become—the object of a new sign, which in principle means it could spawn an infinite process of semiosis, growing in complexity, specificity, vividness, generality, or in some other manner:

Symbols grow. They come into being by development out of other signs, particularly from likenesses or from mixed signs partaking of the nature of likenesses and symbols. We think only in signs. These mental signs are of mixed nature; the symbol-parts of them are called concepts. If a man makes a new symbol, it is by thoughts involving concepts. So it is only out of symbols that a new symbol can grow. (Essential Peirce, vol. 2, 10)

Another processual element resides in what Peirce called the logical interpretant: a habit-change, consisting of “a modification of a person’s tendencies toward action,” and also such things as a habitual association of ideas or a way of thinking, which we might describe as actions of thought (CP 5.476-78). This is one more instance of dynamism and transformative elements in his thought, one in which a rule itself may changes and semiosis can ultimately reenter the body. Nor is all of it necessarily conscious:

The action of thought is all the time going on, not merely in that part of consciousness which thrusts itself upon attention, but also in those parts that are deeply shaded and of which we are too little conscious to be much affected by what takes place there. (Essential Peirce, vol. 2, 23)

Today we’d reject the idea that what happens in the subconscious doesn’t affect us, but my point is the depth to which semiotic activity goes.

But Peirce’s signs (particularly symbols) don’t only grow through their activity: they also ramify. Saussure’s sign-structure, the simple signifier + signified combination, is static. The only differences among them that one might point to are as parts of speech (nouns, verbs, etc), but these involve no structural distinctions in the Sr/Sd relationship: the relationship was solely an arbitrary, conventional one (Peirce’s symbols). In contrast, from the very first Peirce had three distinct sign-types (icon, index, symbol), arising from different types of relationship between the sign and object, relationships that were themselves defined by his three categories. Over the years Peirce realized that this taxonomy, although satisfactory for many purposes, was in logical or scientific terms too unrefined, and he identified ten categories, which in his later years he expanded to 66.

As I explained in my previous post, the methodology behind these distinctions is governed by the trichotomies. They operate as an engine of ramification, and they do so because the relationship between them is emergent and stratified. In the two paragraphs following his argument that symbols grow, he writes:

we may liken the indices we use in reasoning to the hard parts of the body, and the likenesses we use to the blood: the one holds us stiffly up to the realities, the other with its swift changes supplies the nutriment for the main body of thought. . . . The reasoner makes some sort of mental diagram by which he sees that [a] conclusion must be true, if the premise is so; and this diagram is an icon or likeness. (Essential Peirce, vol. 2, 10)

Thus in the reasoning process, icons have a special weight or function, a point that melds well with cognitive science’s findings concerning image schemas. (See also CP 2.170.) The symbol produced at one point in thought changes the quality of its sign-structure (or the emphasis of its quality) from conventional to iconic—and, moreover, to a potential, as it is not inevitable that a new step in reasoning will occur.

In addition, when he developed the ten sign categories, Peirce realized that the icon-index-symbol distinction was already too concrete and actual, in fact a Second, and he would have to back up to identify consisting of a triad of semiotic Firstness, a set of thoroughly qualitative potentialities. Thus in his derivation of the ten categories, the icon-index-symbol triad came second, and the next was determined by Thirdness (consisting of rhemes, dicents and arguments). But note that because the three triads are distinguished by the trichotomy of Firstness, Secondness, and Thirdness, the structure of ten sign-types is itself stratified and emergent: they have a necessary order, a line of dependency, and because they are qualitatively different they cannot be reduced to any of the others.

For these reasons, I hold that Peirce’s system of trichotomy provides a method for analyzing emergent entities, and perhaps of emergence as such.

However, two additional points. First, the fact that Fourthness, Fifthness etc can be resolved into combinations of Thirdness implies that the growth of symbolic thought will remain at the level of symbols, and not produce a novel type of sign. This is particularly significant when considering reflexivity as recursion: purely discursive (especially linguistic) reflexivity loops within semiosis, although it may reach a final interpretant in the form of a habit-change. (There’s an ethical point about reflexivity there!) Of course, nevertheless within the level of Thirdness, there can be (and is) increasing degrees of complexity.

Second, Peirce didn’t explore the fact that because signs are triune—a form of Thirdness—it’s possible to back up further to find the forms of Firstness and Secondness preceding the semiotic relationship. This is what I am doing in analyzing agency, which is rooted in our material being. But agency itself is triune (embodied, causal, and intentional), and as agency, it is a Second. The Firstness preceding it, I’ll hypothesize, is the natural one I discussed above: physics, biology, semiosis (biosemiosis, including non-human semiosis). Perhaps there is a triad preceding even that, but that notion might lead us into “triads all the way down,” an infinite regression of who knows what sort.