This is not, however, true of syntactically similar coordinate clauses that receive a coordinate interpretation, as shown in 13c. Another picture of himselfi appears in the newspaper, and Susan thinks Johni will definitely go out and get a lawyer.
- Handbook of Intensive Care.
- The Nature and Function of Syntactic Categories by Robert D. Borsley - casysdipoor.cf.
- Product description.
If another picture of himselfi appears in the newspaper, Susan thinks Johni will definitely go out and get a lawyer. As Culicover and Jackendoff argue at length, these facts suggest that the possibility for binding a reflexive pronoun depends on semantic properties that are independent of syntactic structure. They do not mention any need for the idea of prototypes or prototypicality. However, they do in passing discuss some interesting data that are not readily explained in terms of independent syntactic or semantic principles.
Consider the examples in 14a—b below: 14 a. Big Louie saw you with the loot and he put a contract out on you. Coordinate interpretation only b. If Big Louie saw you with the loot, he put a contract out on you.
However, there is no obvious syntactic or semantic reason why this should be so. Simply changing the verb form to past tense should not change the constituent structure of the sentence. Likewise, there appears to be no semantic anomaly involved, as indicated by the fact that the past tense if-clause in 14b is perfectly acceptable with a subordinate interpretation. A similar problem is posed by the examples in 15a—d below. The sentence in 15a contains left-subordinating and, and receives a subordinate interpretation similar to that of 15b.
However, the subordinate interpretation disappears if we add a third clause, as in 15c. Since syntactically coordinate structures normally allow three conjuncts, there is no obvious syntactic reason why constructions with left-subordinating and should not also allow three conjuncts. Furthermore, there is no obvious semantic reason for this either. As shown in 15d , a three- clause sentence in which the first two clauses are subordinated to the third is perfectly acceptable when the first two clauses are if-clauses.
Subordinate interpretation preferred b. Subordinate interpretation c. Coordinate interpretation only d. Subordinate interpretation As in the case of ANs, independent syntactic and semantic principles fail to predict the patterns of data described here. FRANCIS 23 language may play some role in these cases, we can again argue that the lack of a subordinate interpretation for sentences such as 14a and 15c is at least partly the result of a deviation from the prototype.
In sentences with left-subordinating and, a syntactically coordinate structure is used to express a subordinate relationship, thus violating the prototypical correspondence between coordinate structure and coordinate meaning.
- Natural Compounds: Flavonoids!
- Fitness Cross-Country Skiing;
- Top Authors.
Although in many contexts, left- subordinating and is completely natural and fully grammatical, we have seen that some syntactically coordinate sentences with and cannot receive a subordinate interpretation. In the case of 14a , past tense marking appears to preclude a subordinate interpretation.
This fact is puzzling, since past tense marking is semantically compatible with a subordination relation, as shown in 14b. I propose that because a sequence of coordinated past tense clauses is normally used perfectively to express a sequence of events in a narrative, and because sequences of events are semantically coordinate, past tense marking strongly implicates a coordinate interpretation of 14a. Why, then, is present tense marking in 11a fully compatible with left- subordinating and?
Since present tense is rarely used to describe sequences of events but normally used to express subordination relations involving hypothetical situations, the violation of the prototype exhibited by left-subordinating and is less salient in present tense than in past. Interestingly, a subordinate interpretation is in fact compatible with past tense marking in certain discourse contexts in which a habitual or generic pattern of behavior is being described 16 In those days, Big Louie saw you with his sister and he put a contract out on you.
Similar to the case of present tense, the mismatch involved with left-subordinating and is less salient in a generic past-tense context as in 16 than in a non-generic past-tense context as in 14a , because the generic context does not implicate a coordinate meaning and so does not highlight the coordinate properties of the sentence.
We can understand the patterns in 15 in a similar manner. Sequences of three clauses normally express sequences of three events rather than subordination relations, whereas subordination relations are more typically expressed by two-clause sequences. Similar to the examples in 14 , the violation of the prototype is less salient for two-clause sequences involving left-subordinating and than for three-clause sequences, contributing to the unacceptability of sentences like 15c in a subordinate interpretation. This is because the two-clause structure in 15a instantiates the expected form-meaning correspondence for expressing subordination as much as is possible within the context of a syntactically coordinate sentence , thus downplaying the syntactically coordinate nature of left-subordinating and.
FRANCIS 25 structure in 15c draws attention to the coordinate properties of the sentence, in effect excluding the possibility of a subordinating interpretation of and. These two case studies of syntactic mimicry provide some evidence in favor of a modular approach to grammar in which independent principles apply to distinct levels of linguistic structure.
The Nature and Function of Syntactic Categories
Such an approach is helpful, for example, in explaining the binding facts for constructions with left-subordinating and, as well as the presence of semantically empty words in constructions with ANs. However, independently-needed principles of syntax and semantics are apparently not enough to explain some of the distributional and interpretive properties of these items.
I have proposed, following McCawley , that these cases can best be handled with reference to prototypes. While relatively subtle violations of the prototype are allowed, more blatant violations are not. In this paper, I have reviewed some of the issues involved in evaluating prototype theories of syntactic categories and have argued in favor of a notion of prototype in grammar.
We have seen that such principles are helpful for understanding otherwise puzzling phenomena, such as the occurrence of meaningless morphemes in some cases of categorial mismatch e. McCawley , ; Sadock ; Francis The two cases of syntactic mimicry discussed in this paper support the existence of prototypes in grammar. In short, prototypes constitute expectations that appear to constrain the manner and degree to which syntax and semantics can operate independently of each other, thus rendering more blatant deviations from the expected form-meaning correspondences ungrammatical.
Ultimately, of course, syntactic categories are cognitive in nature, and additional psycholinguistic and neurolinguistic evidence will be needed to determine their structure. At the least, the linguistic evidence for prototypes discussed in this paper provides a plausible hypothesis to test in future studies of syntactic categorization.
Although constructions do not necessarily have a prototype structure, they are similar to the prototypes discussed in this paper in that they involve a conventional mapping between a form and a meaning. FRANCIS 27 2 Another possibility is that categorization decisions may be based on clusters of previously stored exemplars, as hypothesized in exemplar-based theories e. However, exemplar-based theories are beyond the scope of this paper and will not be considered here. This difference between prototypes and squishy categories will not concern us here.
However, it is semantically bleached in the sense that it does not serve its usual function of indicating indefinite reference. Mufwene p. For example, Newmeyer , advocates a classical approach to categories and assumes in addition a modular theory of grammar to accommodate prototype effects, as discussed in Section 3. Mufwene argues that only a small, semantically-defined subset of stative verbs is actually incompatible with the progressive. However, he agrees with McCawley that these nouns display an unusual combination of syntactic and semantic properties, contributing to their defective distributions.
Coordinate reading only ii If she has two children, then she's eligible for the tax credit. FRANCIS 29 The reviewer suggests that perhaps i lacks a subordinate interpretation because there is a separate prototype for the coordination of simultaneously valid states, which specifies present tense. This seems to me to be a plausible interpretation of the facts. Binominal noun phrases in English. Transactions of the Philological Society What some concepts might not be. Cognition Baker, Mark C. Lexical categories: Verbs nouns and adjectives.
Cambridge: Cambridge University Press. Bolinger, Dwight. Syntactic diffusion and the definite article.
Bloomington: Indiana University Linguistics Club. Bresnan, Joan. Category mismatches. In Theoretical approaches to African linguistics, ed. Chomsky, Noam. Aspects of the theory of syntax. Remarks on nominalization. In Readings in English transformational grammar, ed.
Rosenbaum, — Boston: Ginn. Croft, William. Syntactic categories and grammatical relations.