Reference no: EM13796749

Exercise 1 Reflexives and ø-features

This exercise is intended to begin to develop your skills in syntactic argumentation. Follow the instructions exactly. Do not try to do more than is asked for each section.

Part A

There is a class of words in English called reflexive pronouns. These words are formed from a pronominal plus the word self in its singular or plural form.

Examples are:

Myself, ourselves

Yourself, yourselves

Himself, themselves

Herself

Itself.

In a simple sentence, these words are restricted in their distribution:

(1) *1 kicked yourself

(2) *Fie kicked yourself

(3) You kicked yourself.

Notice that examples like (3) have a special semantic property. The person doing the action described by the verb, and the person affected by this action, are one and the same. In (3), you are the kicker and the kickee. The words you and yourself are said to be coreferential, since they both refer to the same person. Other examples of expressions that can be co-referential are Anson and he in the following sentences:

(4) Anson thought that he had bought the paint stripper.

(5) I asked Anson if he was happy.

In (4) and (5), the pronoun he appears to take on its meaning via the expression Anson. In these particular sentences, another reading is also possible. He can refer to someone else entirely, just as she does in (6):

(6) The boy thought she was happy.

In (6), she is not coreferential with the boy, since coreferentiality appears to require matching ø-features.

Let us state this idea as an explicit hypothesis, the Coreferentiality

Hypothesis:

(7) For two expressions to be coreferential, they must bear the same ø -features.

We can see the coreferentiality Hypothesis as a kind of general interface rule which relates syntactic features to semantic interpretation.

In the examples with reflexives we have a case of obligatory coreference, in (3), you and yourself are required to be interpreted as coreferential. The same can be said for the following cases:

(8) He kicked himself,

(9) We kicked ourselves,

(10) They kicked themselves,

Using this notion of coreferentiality, we can state a hypothesis that will differentiate between the good and the bad examples above, We will call this hypothesis the Reflexive Generalization, for ease of reference. Parts B-F of this exercise will be devoted to revising this hypothesis, so what we have here is just a first attempt

(11) The Reflexive Generalization (First attempt):

A reflexive pronoun must be coreferential with another expression in the sentence.

The Reflexive Generalization has the consequence that a reflexive pronoun will have to be coreferential with another expression, and hence, have the same ø -features as that expression.

Task 1. Provide some further examples that support this result for person features.

Part B

Now look at the following data:

(12) You kicked yourselves.

(13) *We kicked myself

(14) *They kicked himself

Task 2. Explain how our generalizations account for these examples.

Task 3. Provide further examples that show the hypothesis working for number features.

Part C

Now look at the following data:

(15) *He kicked herself

(16) *She kicked itself

Task 4. Explain how these examples are captured by the hypotheses, and provide more examples which show that gender features are relevant.

Part D

The following examples show that this hypothesis is not enough to explain the distribution of reflexive pronouns:

(17) *Myself saw me

(18) *Himself saw him

Task 5. Explain why the Reflexive Generalization is not enough to rule out these examples.

Task 6. How might you alter the hypothesis so that it covers these examples? (Hint: there are two possibilities here: one involves the order of the words, the other involves the case of the pronoun inside the reflexive.)

Part E

Look at the following examples:

(19) *1 thought he liked myself

(20) • You said she liked yourself

Task 7. Whichever hypothesis you came up with for Task 6, explain whether these data are problematic for it or not.

Part F Summary

These data show that the distribution of reflexives is not just conditioned by their ø-features, but that word order, case, and other aspects of syntactic structure may enter into their analysis too. This is a general fact about syntactic problems. The rarely confine their solutions to just one part of the grammar.