Exercise 3 Pronouns and coreference

Part A

This exercise flows on from the last two, but looks at how pronouns and reflexives are distributed. Consider the following data:

(1) *I kicked me

(2) I kicked myself.

(3) You kicked you

(4) You kicked yourself.

(5) *We kicked us

(6) We kicked ourselves.

Task 1 We can formulate a hypothesis to cover these data, using the same kinds of concepts. We will term this the Pronoun Generalization:

(7) A pronoun cannot be coreferential with another pronoun.

Task 2 Explain how this hypothesis extends to the following data:

(8) He kicked him.

(9) They kicked them.

(10) She kicked her.

Part B

Recall the two grammars that you chose between in the last exercise. Grammar A stated that imperatives had an unpronounced subject, while grammar B stated that imperatives had no subject at all, but that only second-person reflexives could appear in imperatives.

Now consider the following data:

(11) Kick me!

(12) Kick them!

(13) *Kick you!

Task 3 Assume Grammar A is the correct grammar, and that the Pronoun Generalization holds. Explain how the new data given above fall out in this grammar.

Task 4 Assume Grammar B is the correct grammar, and that the Pronoun Generalization holds. Explain how the new data don't immediately follow, and revise Grammar B by adding a new hypothesis to fix it.

Task 5 Now explain which of the two grammars is preferable, and why.

Further reading

Section 2.2

The notion of features in syntax derives from work on features in phonology. See Chomsky and Halle (1968) for discussion and further references. Matthews (1974) gives an overview of many of the kinds of morphological processes that are found in languages. For a more detailed overview of issues in morphology see the papers in Spencer and Zwicky (1998).

Section 2.3

See Dalrymple and Kaplan (2000) for a discussion of some of the issues raised by the text, and Harley and Ritter (2001) for an alternative using feature geometry. The Hopi data are taken from the latter source. An excellent descriptive discussion of constraints on how languages vary with respect to the features they use is Greenberg (1966). The idea that morphology and semantics are both interpretative components of grammar goes back to early work of Chomsky (see Chomsky, 1970).