- Give three grammatical sentences that are generated by the current grammar (as it will stand by the end of class on Monday, April 15). Provide a phrase structure tree for one of these sentences.
- Give three sentences which are, as a matter of fact, grammatical — i.e., which are sentences of English — but which are not generated by the current grammar. Explain what prevents each sentence from being generated. [For purposes of the assignment, assume that the lexicon include all the English words which belong to our current word classes (lexical categories)].
- Give three ungrammatical sentences which are generated by the grammar. Provide a phrase structure tree for one of them.
- Consider the sentence in (1):
(1) They mentioned the secret village in the forest.
This sentence is ambiguous. First, explicate the ambiguity using paraphrases. Does our current theory provide a way to understand the fact that (1) is ambiguous? Explain your answer to this question as clearly as possible.
EXTEND THE CURRENT GRAMMAR
5. The data in (2)-(8) require changes to our current grammar. Explain why. Then revise the grammar to account for them. That is: add or change rules so that the grammar generates the grammatical ones but does not generate the ungrammatical ones. You should explain your revision, motivating everything you can. To illustrate how your analysis works, draw constituent structure trees for some (but not all) of the examples.
- *A very senator proposed the measure.
- A very disingenuous senator proposed the measure.
A disingenuous senator proposed the measure.
- *The rather professor danced in the forest.
- The rather fat professor danced in the forest.
The fat professor danced in the forest.
- *Yolanda investigates notoriously executives.
- Yolanda investigates notoriously corrupt executives.
Yolanda investigates corrupt executives.
–1 /3 —
- (5) a. Some students are grumpy.
b. *Some students are very.
c. Some students are very grumpy.
- (6) a. You sound frantic. b. *You sound rather. c. You sound rather frantic.
- (7) a. Every sad person seems inert.
b. *Every sad person seems very.
c. Every sad person seems very inert.
- (8) a. The puppy walked with the affable politician.
b. *The puppy walked with the rather politician.
c. The puppy walked with the rather affable politician.
Note: When doing your write-up, you should replicate the examples above in the text of your essay. That is, don’t merely refer to them by their numbers on this assignment sheet — but incorporate them into your own write-up.
If you analysis is properly constructed, it should also automatically account for the data in (9) below.
(9) a. A slightly seedy but rather attractive hotel stood at the edge of town. b. We vacationed in a very grand, extremely expensive and rather
Both of the examples in (10) contain a subtle ambiguity. Does the analysis you have constructed (in combination with the existing grammar) account for that ambiguity? In other words, does it assign two distinct trees to each sentence?
(10) a. We read an article about an extremely corrupt and disingenuous senator. b. The puppy licked the very stinky and playful child.
–2 /3 —
GUIDELINES TO KEEP IN MIND:
- Please read the guidelines for written work in the syllabus and follow them closely.
- When asked to make a revision to the grammar, do so explicitly. Write a new rule or change an existing one. When you do this, give a full and explicit statement of the new or revised rule.
- When you propose a revision, always discuss it. Say why you are making a particular proposal rather than some other conceivable proposal. Consider and weigh alternatives. Here are some questions you might ask yourself:
- * if I make a change to a rule in response to a specific sentence, what will the implications be beyond that sentence?
- * will my change lead to generation of ungrammatical (and therefore unwanted) sentences?
- * will it prevent the generation of other grammatical sentences (that were otherwise generated by the grammar, before you made the change) In general, if you are proposing a revision to account for a single example, you are likely on the wrong track. Successful solutions will account for a large body of data (i.e., several grammatical or ungrammatical sentence types). To put this another way, if you wish to account for five new types of examples, and you find yourself proposing five different adjustments or revision (one per example), then you are almost certainly on the wrong path. In other words: our search is for economical and elegant solutions, not hacks or kludges