Descriptivism and anti-descriptivism

Listen to audio

In the Frege-inspired " descriptivist " conception, in order to be able to refer to an object, the subject must possess an identifying description of that object, or a globally identifying set of information (a mental file). However, it seems that we can refer to objects even if we don't possess an identifying description that allows us to single them out. The descriptivist conception also implies that the information in our possession must be correct for us to be able to refer to an object. But might we not discover that what we believe about a given individual to whom we refer by a proper name is largely wrong ? Could we not discover, for example, that Moses did nothing of the kind attributed to him in the Bible ?

Another type of objection concerns cases where reference is made not by means of proper nouns, but by means of pronouns or demonstratives. We have seen that the modes of presentation are different when the subject thinks J'ai le pantalon qui brûle (I've got burning pants) and when he thinks Il (or : this guy) a le pantalon qui brûle ( He ' s got burning pants). Now, Castañeda, Perry and the theorists of " the essential indexical" have established that the mode of presentation corresponding to an indexical like I cannot be a description, insofar as for any objective description Le F, the subject is not forced to realize that he himself is Le F.

In the alternative conception put forward by critics of descriptivism, what fixes the reference of a proper name is the fact that this name has been transmitted along a historical chain of communication that begins with the assignment of the name to the object itself (the " initial baptism ", as Kripke puts it). The subject is thus linked to the object through the name it uses and the chain of communication to which the name belongs. In the case of indexicals, too, a relation is at play, and it's this that fixes the reference. Here refers to the place where the subject is, not to the place that satisfies an objective description of the place in question in the subject's mind. I refers to the person the subject actually is, not to the person he thinks he is. In both cases, we can retain the idea that the mode of presentation is a "mental file", provided we abandon the idea that the reference of a mental file is the object that possesses the properties listed in the file. These properties correspond to what the subject believes about the referent, but the subject can be wrong, and so it's essential that it's not the properties in question that determine the reference. What determines the reference are the relations (known as " epistemically gratifying ") that the files exploit, and which serve as an informational channel in the sense that the information obtained via these relations feeds the file.

Abstracts from lectures 6, 7 and 8 (cancelled due to Covid-19)

A mental file is not defined as a collection or set of information about the referent. A set is individualized by its elements, so that if you change one of the elements, you change the whole. But a folder can remain the same folder even if you delete or replace any of the information in the folder. Like a cardboard folder containing documents, a mental file has its own individuality, a digital identity independent of what it contains. A file is therefore a thing like any other, i.e., in the terminology of philosophers, a particular - a " mental particular ", say Crimmins and Perry. What makes it possible to individualize a file, in this perspective, is not the information it contains, nor the type of file it is (based on this or that epistemically gratifying relationship), nor its reference, nor even a combination of all these things. We can well imagine situations where the subject, in the grip of an illusion, would open two distinct files of the same type, containing the same elements and referring to the same object, because he believes he is in the presence of two objects when in fact there is only one.

In the three lectures that remained to be given, but had to be cancelled due to the health crisis, this conception of files as mental particulars was invoked to defend the idea that mental files are, in Millikan's terms, " markers of identity " (lecture 7 - Identity, identification and coreference), to reanalyze the cognitive effect of discovering an identity (e.g., the identity of Emile Ajar and Romain Gary) in the light of the debate between the proponents of file merging and the proponents of linking (lecture 8 - Strawson's constraint and indexed files), and to return to the debate between Frege and Russell by asking whether what plays the role of mode of presentation is indeed an aspect of the content of representations (as Frege argues), namely meaning distinguished from reference, or whether it is the mental representations themselves, as Fodor argues (lecture 6 - Mental files as " particulars ").

Mental files considered as particular are vehicles, just as words and sentences are, in language, forms endowed with meaning. They are vehicles with content - both in the sense that they have a referent, and in the sense that they " contain " information about the referent -, but which should not be confused with the content they convey. This distinction opens up the possibility that it is these vehicles, these words of the " language of thought ", and not this or that aspect of their content, that play the role of mode of presentation. Indeed, if a subject apprehends the same object through two distinct representations, he may, even if he is rational, fail to realize that both times he is dealing with the same object. If we accept that a mental file is that which, in the subject's mind, represents an entity of the external world, then the existence of two distinct mental files suggests the existence of two distinct entities, even if these files are not distinguished by their content. Content, from Fodor's perspective, may well conform to the "Fido"-Fido theory, i.e. purely referential, since what plays the role assigned to the mode of presentation is in fact a " syntactic " entity, a vehicle, rather than a specific level of content. Lecture 6 aimed to defend the Fregéan position by showing that one aspect of the content of mental folders does indeed conform to the Fregéan notion of mode of presentation. A mental file is, through the epistemically gratifying relations it functions to exploit, associated with certain presuppositions that characterize it intrinsically and are independent of the environment. These presuppositions associated with the file are elements of content transparently accessible to the subject (unlike reference, which, because it depends on the external environment, is opaque to the subject), and they play a crucial role in the explanation of behavior and in psychological generalizations. They are part of what philosophers of mind call the " narrow content " of thoughts.