Abstract
In this paper we analyse anaphoric pronouns in control sentences and we investigate the implications of these kinds of sentences in relation to the Propositional Theory versus Property Theory question. For these purposes, we invoke the categorial calculus with limited contraction, a conservative extension of Lambek calculus that builds contraction into the logical rules for a customized slash type-constructor.
This is a preview of subscription content, log in via an institution to check access.
Notes
- 1.
This paper will deal only with what is called obligatory control (or exhaustive control in the nomenclature of [14]), cases where the controller must be, for each verb, a NP in a particular syntactic position in the sentence, not with types of arbitrary control (cf. [10, 24, 25]). An anonymous reviewer observes that our approach seems to provide resources to treat also cases of split control, where both the subject and the object matrix clause can jointly form the controlled embedded subject, as in the following Portuguese examples, where the inflection on the infinitive form (INFL) marks the plural predication:
There is not complete agreement in the literature as to whether split control is a type of obligatory control. In footnote 17 we will show how can we deal with split antecedents. As this reviewer notes such cases of control were discussed in the LFG Glue framework. Indeed, [2], following [5], observes that anaphoric control, but not functional control, allows split controlled antecedents. In future research we hope to compare our proposal for control with the one made in the related resource-sensitive formalism of LFG [2, 3], which we unfortunately do not have space to discuss here.
- 2.
BP stands for Brazilian Portuguese, EP for European Portuguese, SPA for Spanish and IT for Italian. The infinitival subjects are highlighted. As is well known, Portuguese has an inflected form of infinitive (INFL), that is, an infinitive form that carries ending marks of agreement with their subject in both person and number. It is generally assumed that inflection is obligatory if there is an overt subject within the infinitive clause. And it is also generally assumed that the inflection must be deleted in cases where the reference of the (null) subject coincides with the reference of a matrix constituent. Thus, both rules confront each other in cases of control sentences with overt subjects.
- 3.
An anonymous reviewer indicates that the so-called Richard constructions, as exemplified below, have a number of suggestive parallel features with the variety of control which we deal with. Despite some similarities, it is important to note that Richard constructions are cases of copy raising; raising verbs, unlike control verbs, do not select for a thematic subject in the predicative complement, and copy raising verbs, unlike our control examples, require a pronominal bound copy in their complement clause:
\(\begin{array}{ll} \text {i.} &{} \text {Richard seems like he is ill.}\\ \text {ii.} &{} \text {Richard seems like he is in trouble.} \end{array}\)
The phenomenon of copy raising is also attested in Portuguese. But two differences between typical Richard constructions and cases of copy raising in Portuguese must be pointed out: Firstly, the lexical copy is not obligatory in the Portuguese constructions; and secondly, the embedded copy in Portuguese can be not only a pronoun but also a lexical DP (cf. [4]):
For a treatment of copy raising in a resource-conscious framework or in a generative grammar we refer the interested reader to [1, 21], respectively.
- 4.
A anonymous reviewer objects to the use of Gentzen format as “about as unfriendly as possible”. Gentzen calculus, labelled and unlabelled natural deductions, proof nets, categorical calculus, etc. are all of repute, all have their respective advantages and disadvantages, and are all notations for the same theory. We think that it is better to try to understand each notation than to censure one. The Gentzen format is really not so hard to read.
- 5.
Both systems also admit the Cut rule, i.e. adding the Cut rule does not give rise to any new theorems.
- 6.
In this respect, Jaeger adopts Jacobson’s proposal [12].
- 7.
Jaeger’s proposal in itself does not capture either Binding Principle A (locality of anaphors) or Principle B (antilocality of personal pronouns). A categorial approach to locality of anaphors is given by modalities in [17] and a ‘negation as failure’ categorial approach including antilocality of personal pronouns is given in [20].
- 8.
In SPA and IT subjects of (adverbial or subject) infinitive constructions necessarily occupy the post-verbal position. By contrast, in BP such subjects normally occupy the pre-verbal position. In EP subjects within infinitive clauses normally occur in the post-verbal position, but the pre-verbal position can also be admitted. Hence, the nominal argument position in the embedded verb in control structures is justified:
. - 9.
If the controller were a pronoun, as in the example below, then the complex category s|n can be derived, but this is in virtue of the matrix subject pronoun.
. - 10.
As in other Romance languages, object pronouns in Portuguese take the clitic form: (l)o/(l)a. In Brazilian spoken language the third person (non-reflexive) clitics are not commonly used; instead, the (nominative) form ele(s)/a(s) is usually used for accusative object:
\(\begin{array}{ll} \text {i.} &{} \text {Visitei-o ontem. (EP)}\\ \text {ii.} &{} \text {Visitei ele ontem. (BP)}\\ &{} \text {`[I] visited him yesterday.'} \end{array}\)
- 11.
Jaeger’s pronominal type does not distinguish between pre- and post-verbal position, this last difference being incorporated in the infinitive verb type in our previous proposal. As we have said before, Jaeger’s proposal does not capture Principles A (locality) and B (antilocality) of the Binding Theory: that reflexive pronouns must be bound in their own clause and that accusative pronouns cannot take a c-commanding antecedent in their own clause. In order to take account of the pronominal position, we shall use lifted pronominal types, for example, \(\left( s|n\right) /\left( n {\backslash } s\right) \) for pre-verbal subjects.
- 12.
An anonymous reviewer observes that this difference could be made by using features instead of introducing a new connective. Although we could have chosen that option, we have preferred to extend Jaeger’s proposal because using both proforms we can obtain the corresponding lifted types, and so, we can also distinguish between a pre- and a post-verbal pronoun.
- 13.
Observe that the double free pronoun reading \(\lambda x. \lambda y. ((saw'\ y)\ x)\) for He saw him, which in Jaeger’s system gets the category \(\left( s|n\right) |n\), corresponds on our proposal to the type \(\left( s||n\right) |n\) or (s|n)||n.
- 14.
Note that the bound reading for the object n|n is obtained by using, not |R, but |L.
- 15.
There are two readings for Pedro quer ele ajudar ele: a reflexive bound reading \(\left( \left( wanted'\ \left( \left( help'\ p\right) \ p\right) \right) \ p\right) \) and a free reading \(\lambda x.\left( \left( wanted'\ \left( \left( help'\ x\right) \ p\right) \right) \ p\right) \). In the first case, the clause corresponds to the type s and in the second, to the type s|n. There is no derivation for the type s||n with the subject free reading \(\lambda x.\left( \left( wanted'\ \left( \left( help'\ p\right) \ x\right) \right) \ p\right) \). Considering that in BP the object pronoun can take the nominative form, it seems we have to take the sequence of types \(n,\left( n{\backslash } s\right) /\left( \left( s||n\right) \sqcup \left( n{\backslash } s\right) \right) ,n,\left( n {\backslash } s\right) / n,n||n\) into account for the sentence in (12). But, it must to be remembered that infinitive subjects are usually preverbal in BP.
- 16.
- 17.
To analyse split control as exemplified by the Portuguese sentence below, we suggest the following lexical entry for the prepositional control verb convencer a/de ‘convince’, where g groups individuals:
-
\(\mathbf{convenci }:\left( n {\backslash } s\right) /\left( n\bullet \left( n {\backslash } s\right) \right) : \lambda x.\lambda y.\left( \left( \left( convinced\ \left( \pi _{2}x\ \pi _{1}x\right) \right) \ g(y, \pi _{1}x\right) )\ y\right) \)
-
\(\mathbf{a/de }:\left( n {\backslash } s\right) /\left( \left( s||n\right) \sqcup \left( n {\backslash } s\right) \right) :\lambda x.x\)
-
- 18.
Since the pronoun ele is assigned the type n||n and the nominal phrase os amigos d’ele contains this pronoun it seems that we have to admit that the pronominal type n||n is also assigned to it, and consequently, the sentence in (ii) could be derived:
Nevertheless, in this case ele is the complement of the preposition, and it is used as the third person possessive pronoun in order to avoid the ambiguity between the second and the third reading for the possessive seu(s) (‘your’/‘his’/‘her’). Observe that the preposition de ‘of’ cannot take a first pronoun as its complement: *os amigos de mim/eu/nós.
- 19.
Although for reasons of space we do not do so here, we believe our eventual, lifted pronoun type, proposal can prohibit this non-locality when it is semantically modalised; cf. the way non-locality for reflexives is blocked in [17].
- 20.
Note that the sequents are not derivable in the reverse direction. Hence, assigning lifted types preserves some but not all of the distribution of unlifted types.
- 21.
In the previous proposal for control sentences this fact is captured by assigning different pronominal types for the control argument—s||n—and the object pronoun—n|n. Notwistanding, as in Jaeger’s proposal, there is no way to block the occurrence of an object pronoun in a subject position.
- 22.
.
.
References
Asudeh, A.: Richard III. In: Andronis, M., Debenport, E., Pycha, A., Yoshimura, K. (eds.) CLS 38: The Main Session, vol. 1, pp. 31–46. Chicago Linguistic Society, Chicago (2002)
Asudeh, A.: Control and Semantic Resource Sensitivity. J. Linguist. 41(3), 465–511 (2005)
Asudeh, A., Mortazavinia, M.: Obligatory control in Persian: implications for the syntax-semantics interface. In: Handout from ICIL 4, Uppsala University (2011)
Barbosa, P.: Overt subjects in raising and control complements and the null subject parameter. In: LSA Annual Meeting Extended Abstracts (2010)
Bresnan, J.: Control and complementation. Linguist. Inquiry 13, 343–434 (1982)
Chierchia, G.: Anaphoric properties of infinitives and gerunds. In: Cobler, M., MacKaye, S., Wescoat, M. (eds.) Proceedings of the Third West Coast Conference on Formal Linguistics, pp. 28–39. Stanford Linguistics Association, Stanford (1984)
Chomsky, N.: Lectures on Government and Binding. Kluwer, Dordrecht (1981)
Chomsky, N.: The Minimalist Program. MIT Press, Cambridge (1995)
Davies, W.D., Dubinsky, S.: The Grammar of Raising and Control. Blackwell Publishing, Oxford (2004)
Dowty, D.R.: On recent analyses of the semantics of control. Linguist. Philos. 8(3), 291–331 (1985)
Hornstein, N.: Movement and control. Linguist. Inquiry 30(1), 69–96 (1999)
Jacobson, P.: Towards a variable-free semantics. Linguist. Philos. 22(2), 117–184 (1999)
Jaeger, G.: Anaphora and Type Logical Grammar. Springer, Dordrecht (2005)
Landau, I.: Elements of Control: Structure and Meaning in Infinitival Constructions. Kluwer Academic Publishers, Dordrecht (2000)
Landau, I.: A Two-Tiered Theory of Control. MIT Press, Cambridge (2015)
Mensching, G.: Infinitive Constructions with Specified Subjects. Oxford University Press, Oxford (2000)
Morrill, G.: Intensionality and boundedness. Linguist. Philos. 13(6), 699–726 (1990)
Morrill, G.V.: Type Logical Grammar: Categorial Logic of Signs. Kluwer Academic Publishers, Dordrecht (1994)
Morrill, G., Valentín, O., Fadda, M.: The displacement calculus. J. Logic Lang. Inform. 20(1), 1–48 (2011)
Morrill, G., Valentín, O.: Displacement logic for anaphora. J. Comput. Syst. Sci. 80(2), 390–409 (2014)
Polinsky, M., Potsdam, E.: Expanding the scope of control and raising. Syntax 9(2), 171–192 (2006)
Raposo, E.: Case theory and Infl-to-Comp: the inflected infinitive in European Portuguese. Linguist. Inq. 18, 85–109 (1987)
Rosenbaum, P.: The Grammar of English Predicate Complement Constructions. MIT Press, Cambridge (1967)
Stiebels, B.: Towards a typology of complement control. In: ZAS Papers in Linguistics, vol. 47, pp. 1–80 (2007)
Wurmbrand, S.: Syntactic vs. semantical control. In: Zwar, C.J., Abraham, W. (eds.) Proceedings from the 15th Workshop on Comparative Germanic Syntax (2002)
Acknowledgements
The first author was supported by a Doctorate scholarship (BEPE) granted by FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo, process number 2015/09699-2). The second author was supported by an ICREA Académia 2012, SGR2014-890 (MACDA) of the Generalitat de Catalunya and MINECO project APCOM (TIN2014-57226-P). We thank Formal Grammar reviewers for comments and suggestions. All errors are our own.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendix
Appendix
(A) \(\mathbf{pedro}{+}{} \mathbf{quer}{+}{} \mathbf{ele}{+}{} \mathbf{chegar}: Sf\)
\(Nt(s(m)): { p}, {\forall }a((Na{\backslash } Sf)/((Si{||}Na){\sqcup }(Na{\backslash } Si))): \lambda A\lambda B(({ want}\ ({ A}\ { B}))\ { B}),\)
\({\forall }v((Sv{||}Nt(s(m)))/(Nt(s(m)){\backslash } Sv)): \lambda C{ C}, {\exists }aNa{\backslash } Si: { arrive}\ \Rightarrow \ Sf\) (See Fig. 14)
Derivation of (A)
\((({ want}\ ({ arrive}\ { p}))\ { p})\)
\((\mathrm {B})\;\mathbf{a}{+}{} \mathbf{policia}{+}{} \mathbf{for}\) çou \({+}{} \mathbf{os}{+}{} \mathbf{manifestantes}{+}{} \mathbf{a}{+}{} \mathbf{eles}{+}{} \mathbf{ajudar}{+}{} \mathbf{ele}: Sf{|}Nt(s(m))\)
Derivation of (B)
\(\lambda A((({ force}\ (({ help}\ { A})\ (\iota \ { protesters})))\ (\iota \ { protesters}))\ (\iota \ { police}))\)
Rights and permissions
Copyright information
© 2016 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Corbalán, M.I., Morrill, G. (2016). Overtly Anaphoric Control in Type Logical Grammar. In: Foret, A., Morrill, G., Muskens, R., Osswald, R., Pogodalla, S. (eds) Formal Grammar. FG FG 2015 2016. Lecture Notes in Computer Science(), vol 9804. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-53042-9_11
Download citation
DOI: https://doi.org/10.1007/978-3-662-53042-9_11
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-662-53041-2
Online ISBN: 978-3-662-53042-9
eBook Packages: Computer ScienceComputer Science (R0)