Accéder au contenu principal

Communal Norms of Representation are Conceptual

Community wp20

The thrust of the previous post is that there are two main levels of analysis when discussing representation in science: relevance and accuracy, or what a representation is about and what it says.

Let us focus first on conditions of relevance. Consider any model M, and any object O. The question I want to ask is: in what sense might we say that it is possible that M represents O, or that it must be so (independently of whether it represents it well)?

We should first dispell an ambiguity in this question. Are we talking in general, or in a particular context? I have argued in my past research (Ruyant 2021) that there is an important difference between two senses of “represent”: either it refers to norms at play in the epistemic community (such as: the Lotka-Volterra model represents a prey-predator system), or it refers to a specific use in context (the model of the pendulum that I’m using represents the oscillation of my clock). However, the first sense is generally more prominent. In the contextual case, we are more likely to say “being used to represent”, or “represents for its user” than “represent” simpliciter.

In the generic, normative sense, models usually represent kinds of objects, or generic. In the example above, “a prey-predator system” does not refer to one system in particular, but to a type of system, as when we say “a triangle has three sides”: we are not referring to one specific triangle, but talking about triangles in general. The same goes when we say that a model represents the hydrogen atom, or oscillators. These are generic locutions, as in “the dodo is extinct” or “Tigers are predators”.

There are exceptions to this: when a model represents one particular object, for example, the solar system or the Sun. However, these are rare in science, and even the model of the Sun is often thought to inform us about many similar stars. So, I will neglect these cases and consider that scientific models represent kinds.

Now, let us reformulate our initial question. It becomes: consider any model M, and any kind object K. The question I want to ask is: in what sense might we say that it is possible that M represents Ks (or the K or a K), or that it must be so? And I would say that the sense of possible or must in which we might say that is conceptual. It is a matter of applying norms to our representations, of understanding these norms, and both the source and target of necessity are internal to our representations.

Consider potential objections:

  • Aren’t there external sources of necessity? After all, we can refer to a kind without having a full description of its nature. Must it not be the case that M correspond to characteristics of Ks that are only known a posteriori in order to represent it? I don’t think so. If misrepresentation is possible, then there need not be such constraints. Whether M is accurate depends on the characteristics of Ks, of course, but this is a matter of accuracy, not relevance.

  • Can’t the model be normative in some cases? Can’t the aim be to tell how we want Ks to be, not how they are? Perhaps, but again, this is not captured at the level of relevance. A model in electronics that represents how transistors must be, for example, does this by means of its content, which reflects the aims of modellers, but the fact that it represents transistors to start with is independent from what these aims are exactly. It is a conceptual matter.

Note that I don’t consider that a model of K can also represent a sub-kind or a super-kind of K. I understand “represent” in a very specific way. Otherwise, we could claim that a question such as “does a model of a whale represent a fish?” or “can a model of a mammal represent a whale?” is an empirical matter (because it depends on the fact that whales are mammals or fish). But it seems to me that these questions trade on an ambiguity between the two senses of representation mentioned above (the communal norm / contextual use distinction). In this respect, the locution “The K” is more appropriate, if sometimes not very natural: it is clear that a model of the whale does not represent “the” mammal, and conversely.

My conclusion is that whether a model represents a kind (communal norms of relevance) is a purely conceptual matter. If there are apparent exceptions to this, this is to the extent that we focus on particular uses on concrete objects instead of communal norms that concern kinds. There can be epistemic uncertainty about whether a model should be used to represent a particular object, or our aims can matter when selecting this or that object as a target of representation when using a model, but we are no more considering the level of kinds here.

Having said that, the use-level is also interesting, so we will consider it in the next article.

Commentaires

Posts les plus consultés de ce blog

Modalities in Scientific Representation

The aim of this blog is to present the result of my Marie Skłodowska-Curie research project "Modalities in Scientific Representation", realised in Madrid between September 2021 and December 2023 (which was in direct continutation with my project realised at UNAM between 2019-2023). A large spectrum of discourse in science is typically analysed in modal terms, that is, in terms of what is possible or necessary. This includes scientific explanations, physical laws and constraints, causation, counterfactual reasoning and probabilities. There is a large amount of literature on the metaphysics of natural necessity in science at a general level, with proposed analyses in terms of governing laws of nature, or dispositions, etc. However, necessity comes in many varieties and other kinds of modalities than natural necessity are involved in scientific discourse. This is certainly the case for epistemic modality, involved, for example, in assessing the credibility of a hypothesis, bu...

Bonus: A Puzzle about Conceptual Changes

Let us say a bit more about a puzzle that came about in previous posts. Let p be "the sum of the angles of a triangle is lower than 180 degrees" and C a modal operator "it is conceivable that". Consider the situation before anyone even conceived of non-euclidean geometries as a consistent alternative to euclidean geometry. Then, intuitively speaking, p was not conceivable (¬Cp). But we now know that p is true of some triangles, because the geometry of the universe is non-euclidean (p). On the other hand, it is quite natural to think that conceptual modality is factive, that is, that if something stems from conceptual necessity, then it is also true (¬C¬p→p), and this implies the converse, that if something is true, then it is conceivable (p→Cp). But these three premises are mutually inconsistent, hence our trilemma: one of the following statements must be false. ¬Cp p p→Cp So, we have exactly three options: (A) either we can claim that p always was...

Epistemic Representation: The Basics

Now that we have reached a rough taxonomy of modalities, let us talk about representation. I think it will be enough to have this rough picture in mind, which, I reckon, should be uncontroversial aspects of representation: Inferentialism: An epistemic representation (such as a scientific model, a map, a realistic drawing) can be used to make inferences about target systems of a relevant kind in concrete situations. Misrepresentation: It is possible, for a representation, to represent its target inaccurately. I take inferentialism to be a minimalist theory of representation. This does not tell against theories that attempt to give more substance to representation, but only defines a common ground. As for the possibility of misrepresentation, it has been challenged notably on the ground that representation would be a “success term” or a thick term ( Poznic 2018 ; Chakravartty 2010, pp. 209–10 ), but I’m utterly unconvinced that this is the case. “Cruel” is a thick term, and...