Abstract
In this paper, I will be concerned with some of the distributional and interpretive properties of certain superficially subjectless non-finite VPs. Within the principles and parameters approach to syntax of Chomsky (1981; 1986) assumed here, these VPs actually do have subjects (phonetically null ‘PRO’) hence, are predicates of clauses. How this PRO selects an antecedent in a given structure (the theory of Control) is a major research question. And, of course, once PRO is admitted as a possible NP, an explanation of its distribution (only subject, and only of non-finite clauses) becomes another major research question.
I would like to acknowledge the helpful comments of Noam Chomsky, Mamoru Saito, and an anonymous reviewer, who commented on earlier drafts.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
Notes
For present purposes, the governing category of an item α can be taken as the minimal NP or S containing α and a governor of α. See Chomsky (1981) and Lasnik and Uriagereka (1988) for discussion of necessary complications in the notion.
The anaphoric nature of PRO is related to its inability to freely refer. The treatment of PRO as pronominal is justified, in part, by the observation that for both PRO and pronominals, if there is a binding antecedent, it must be ‘remote’, outside of its clause. Interestingly, though, the distributions of the two items are not captured in precisely the same way. For a pronoun, the remoteness of the antecedent is demanded by the fact that there will always be a governing category (since a pronoun will always need Case, hence, will always be governed), and any binder of the pronoun will have to be outside that governing category, by Condition B. For PRO, on the other hand, there will be no governing category, hence Condition B will be trivially satisfied. But the only even potentially ungoverned position in a clause is the subject position. Thus, any potential binder of PRO will necessarily be outside of the minimal clause of which PRO is the subject.
Note that the consequence that Condition A cannot determine antecedence for PRO is actually independent of the analysis of PRO as a pronominal anaphor subject to Condition B as well as Condition A. The requirement that PRO be ungoverned, whether deduced from interaction of Binding Conditions A and B or not, will automatically lead to this conclusion. This is so since an ungoverned NP has no governing category, hence, trivially satisfies Condition A no matter what index is assigned to it. Thus, (assuming usual versions of ‘governing category’) even if PRO were analyzed as a pure anaphor, its obligatorily ungoverned character would entail that Condition A is irrelevant to its indexing. Of these two accounts, obligatorily ungoverned anaphor or pronominal anaphor, then, the indexing of PRO doesn’t follow from Binding Conditions on either. The difference between the two accounts is that on the former, not even the distribution of PRO follows from Binding Theory.
For particularly interesting discussions along these lines, see, for example, Bouchard (1984) and Manzini (1983).
These and other Polish data in this paper are from Willim (1982).
An anonymous reviewer observes that the “controllability” phenomenon does not always reduce to theta role assignment in any straightforward way. For example, adverbial modification often suffices to create a complement of the appropriate type, as seen in the contrast between the two cases in (i) or (ii): (i) I tried to receive the good news *(with good grace) (ii) I tried to suffer the insults *(good humoredly) The reviewer speculates, based on these and related observations, that the constraint at work here and in the examples in the text is one on the event structure associated with the complements to particular classes of verbs, a property only indirectly associated with theta role assignment. Given this, the relevant difference between PRO and anaphors would perhaps not be one based directly on Control vs. Binding, as suggested in the text, but rather one stemming from crucial properties of configurations of Control.
Here, too, the reviewer notes certain exceptions. For example, intend does not demand PRO in its infinitival complement, but the semantic requirement does hold when (and only when) PRO is present: (i) *I intend to receive the letter (ii) I intend for her to receive the letter.
Neither Chomsky and Lasnik (1977) nor Lasnik and Fiengo (1974) examined the serve class.
See Chierchia (1984) for extensive discussion of ‘Control’ as involving VP complementation.
Chomsky (1981) observes that under one special circumstance, a pleonastic element can control PRO. He gives the following example: (i) It sometimes rains after PRO snowing Chomsky speculates that weather it is not a true expletive, but rather is a sort of argument, a ‘quasi-argument’. There is some evidence for this speculation. As discussed above, the subject of try is a strongly argumentai position. Yet (ii) is not all that bad for many speakers: (ii) It’s trying to rain (ii) contrasts sharply with clear instances of pleonastics in the same position: (iii) *It’s trying to seem that Mary is crazy (iv) There tried to be an investigation Therefore, following Chomsky, I assume that (i) is not a genuine counterexample to the generalization that “bound PRO always has an argument as its binder.”.
Or, as Noam Chomsky pointed out to me, licensing might only be checked at LF but must be assigned by S-structure. The choice between these alternatives is irrelevant to the arguments of this paper.
Chomsky’s further motivations for this will not be of concern here. I note, in passing, that this analysis of expletives is presented initially as an attempt to capture the observation of Burzio (1986) that expletive-argument pairs evidently conform to the constraints on NP movement chains. On Chomsky’s account, this is immediate, since such pairs are NP movement chains at LF.
Or assignment of arb in appropriate configurations.
See Safir (1985), though, for a possible way out.
A possible third reason involves Case transmission from the expletive to the argument, under the analysis of Safir (1985). But examples such as (56)–(57) above strongly suggest that there is, in fact, no Case transmission, and that an argument must be directly assigned Case even when it is associated with an expletive in a Case-marked position. See Lasnik (1992) for further discussion.
Chomsky (1991) offers an argument against this possibility, which I quote in full: “Note that we cannot assume the expletive to be unindexed — thus it might have raised, leaving an indexed trace.” Presumably what Chomsky is concerned with is distinguishing (i) from (ii): (i) There is likely [t to be a mani here] (ii) There is likely [t is a mani here] Suppose the distinguishing factor is antecedent government. The problem is that without an indexed antecedent, both examples, not just (ii), will be in violation of the Empty Category Principle of Chomsky (1981). This is true enough, at S-structure. But recall that under Chomsky’s proposal, by LF, all expletive-argument pairs have been converted into theta-chains. Then, (i) and (ii) would have become (i′) and (ii′) respectively. (i ′) A mani is likely [ti to be t′i here] (ii′) *A mani is likely [ti is t′i here] Now, if a finite clause is a barrier to antecedent government, (ii′), hence (ii), will be correctly excluded, while (i′), hence (i), will be correctly allowed. Note, by the way, that lack of an index cannot be an option for arguments. Neither Condition B nor Condition C could have the desired filtering effect for an unindexed argument. Further, the index must already be present at S-structure, given the dis-cussion of (60)–(63) above. For present purposes, I simply stipulate that an argument must have an index.
Note that there must have agreement features. The optionality will only be with respect to the specific values. If this were not so, (i) could presumably be generated with arrive in the unmarked, rather than plural, form. (i) *There arrive a man I assume that this will be excluded on essentially morphological grounds. In a finite clause, AGR must appear, and once it does, it must be fully specified for appropriate features via agreement.
Whether this line of reasoning carries over to expletive it is an open question. Chomsky (1986, 1991) resists extending expletive replacement to constructions with it, but the conceptual basis for there replacement — Full Interpretation — would seem to obtain. Chomsky (1981) does suggest that PRO controlled by expletive it is impossible. The following examples from Safir (1985) support this suggestion: (i) *(It) being obvious that I’m a hopeless coward, no one suspects me (ii) is, perhaps, a clearer example, with expletive it futilely attempting to Control expletive PRO. (ii) *(It) being obvious that I’m a hopeless coward, it is obvious that I’m a hopeless coward Here, even the associated clausal arguments are identical. With it, (ii) seems merely badly redundant. Without it, the example seems ungrammatical. (iii) is, possibly, a parallel instance with there. (iii) *(There) being a man in the room, there is a man in the room.
Bibliography
Berman, A.: 1970, ‘Agent, Experiencer, and Controllability’, Report No. NSF-24, Mathematical Linguistics and Automatic Translation, The Computation Laboratory of Harvard University.
Bouchard, D.: 1984, On the Content of Empty Categories, Foris, Dordrecht.
Burzio, L.: 1986, Italian Syntax, Kluwer Acad. Publishers, Dordrecht.
Chierchia, G.: 1984, ‘Topics in the Syntax and Semantics of Infinitives and Gerunds’, University of Massachusetts Ph.D. dissertation.
Chomsky, N.: 1965, Aspects of the Theory of Syntax, MIT Press, Cambridge.
Chomsky, N.: 1981, Lectures on Government and Binding, Foris, Dordrecht.
Chomsky, N.: 1986, Knowledge of Language, Praeger, New York.
Chomsky, N.: 1991, ’some Notes on Economy of Representation and Derivation’, in R. Freidin, ed., Principles and Parameters in Comparative Grammar, MIT Press, Cambridge.
Chomsky, N. and H. Lasnik: 1977, ‘Filters and Control’, Linguistic Inquiry 8, 425–504.
Epstein, S. D.: 1984, ‘Quantifier-pro and the LF Representation of PROARB’, Linguistic Inquiry 15,499–505.
Higgins, R.: 1973, ‘The Pseudo-Cleft Construction in English’, MIT Ph.D. dissertation.
Kajita, M.: 1967, ‘A Generative-Transformational Study of Semi-Auxiliaries in Present-Day American English’, Princeton Ph.D. dissertation.
Lasnik, H.: 1988, ’subjects and the θ-Criterion’, Natural Language and Linguistic Theory 6, 1–18.
Lasnik, H.: 1992, ‘Case and Expletives’, Linguistic Inquiry 23, 3.
Lasnik, H. and R. Fiengo: 1974, ‘Complement Object Deletion’, Linguistic Inquiry 5, 535–571.
Lasnik, H. and J. Uriagereka: 1988, A Course in GB Syntax, MIT Press, Cambridge.
Manzini, R.: 1983, ‘On Control and Control Theory’, Linguistic Inquiry 14, 421–446.
Safir, K.: 1985, Syntactic Chains, Cambridge University Press, Cambridge.
Stowell, T.: 1981, ‘Origins of Phrase Structure’, MIT Ph.D. dissertation.
Willim, E.: 1982, ‘Anaphor-Binding and Pronominal Disjoint Reference in English and Polish’, University of Connecticut MA thesis.
Editor information
Rights and permissions
Copyright information
© 1992 Springer Science+Business Media New York
About this chapter
Cite this chapter
Lasnik, H. (1992). Two Notes on Control and Binding. In: Larson, R.K., Iatridou, S., Lahiri, U., Higginbotham, J. (eds) Control and Grammar. Studies in Linguistics and Philosophy, vol 48. Springer, Dordrecht. https://doi.org/10.1007/978-94-015-7959-9_7
Download citation
DOI: https://doi.org/10.1007/978-94-015-7959-9_7
Publisher Name: Springer, Dordrecht
Print ISBN: 978-90-481-4149-4
Online ISBN: 978-94-015-7959-9
eBook Packages: Springer Book Archive