Wed Mar 9, 2011
Dennes Room, 234 Moses Hall, 6–8 PM
|Working Group in the History and Philosophy of Logic, Mathematics, and Science
Peter Ludlow (Northwestern University)
Linguistic Phenomena, Data, and Theory Need to be Distinguished: I argue that much of linguistic practice turns on the role of what I will call linguistic phenomena. To a first approximation, linguistic phenomena are facts about the acceptability and interpretation of linguistic objects. I argue that these facts or phenomena are explained by the theory of grammar. Linguistic data, on the other hand, provide evidence for the linguistic phenomena. As we will see, there are many potential sources of data, some (especially linguistic intuitions) much more controversial than others.
Linguistic Intuitions are Linguistic Judgments: Among other sources of data, generative linguists often use so-called linguistic intuitions as evidence when constructing our theory of grammar. Many linguists and philosophers take these intuitions to be “Cartesian” in the sense that they are the inner voice of competence. Contrarily, taking a leaf from Williamson (2004) I ague that linguistic intuitions are not “the voice of competence,” but are merely defeasible judgments about linguistic phenomena or facts.
Linguistic Judgments are Reliable (enough): Thinking of linguistic intuitions as judgments can help us avoid being traumatized by the possibility of error in our linguistic data gathering (error is simply something that we have to live with). It can also help us to see that some informants may be better judges than others. I make the case that linguistic judgments are useful, economical, and on the whole reliable sources of data.
Linguistic Judgment is Scientific Experimentation: I argue that when linguists isolate particular judgments to make the case for their theories they are involved in a fairly standard form of scientific investigation. That is, they are setting up controlled, replicable experiments in which they and other experimenters are judging whether a sentence is acceptable and/or what its interpretation might be. This leads to the question of which judges are good judges and why. I point out that this is similar to the question of what makes for a good experimenter in any science and leads us as well into what is sometimes called “the experimenter’s regress.” I’ll offer that we can break the regress by identifying good experimenters using easy cases where judgments converge.