(Pseudo-)Inflected infinitives and Control as Agree. Anabela Gonçalves, Ana Lúcia Santos, Inês Duarte - PDF

Description
Gonçalves, A., A. L. Santos & I. Duarte (Pseudo-)Inflected infinitives and Control as Agree. In Karen Lahousse & Stefania Marzo (eds.) Selected papers from Going Romance Leuven Amsterdam: John

Please download to get full document.

View again

of 20
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Biography

Publish on:

Views: 18 | Pages: 20

Extension: PDF | Download: 0

Share
Transcript
Gonçalves, A., A. L. Santos & I. Duarte (Pseudo-)Inflected infinitives and Control as Agree. In Karen Lahousse & Stefania Marzo (eds.) Selected papers from Going Romance Leuven Amsterdam: John Benjamins. pp (Pseudo-)Inflected infinitives and Control as Agree Anabela Gonçalves, Ana Lúcia Santos, Inês Duarte Abstract This paper discusses the distribution of inflected infinitives in standard and non-standard European Portuguese. In the standard variety, inflected infinitives are generally available in non-obligatory control contexts (subject and adjunct clauses), but can only occur in obligatory control contexts when the temporal orientation of the complement is not specified by the matrix verb. An explanation for this fact is offered along the lines of an Agree theory of control. This analysis also accounts for the possibility of controlled inflected infinitives, which occur in non-standard varieties of European Portuguese. Controlled inflected infinitives bear morphological inflection but do not license nominative. We argue that these non-standard inflected infinitives make the Agree operation underlying control visible (this operation does not have a morphologically overt counterpart in the standard grammar of EP). We also argue that some speakers accept these pseudoinflected infinitives as a strategy to make partial control readings explicit in independent tense contexts. 1 (Pseudo-)Inflected infinitives and Control as Agree 1 Abstract This paper discusses the distribution of inflected infinitives in standard and non-standard European Portuguese. In the standard variety, inflected infinitives are generally available in non-obligatory control contexts, but can only occur in obligatory control contexts when the temporal orientation of the complement is not specified by the matrix verb. An explanation for this fact is offered along the lines of an Agree theory of control. This analysis also accounts for the possibility of controlled inflected infinitives, which bear morphological inflection but which cannot license nominative, occurring in non-standard varieties of European Portuguese: they are considered the result of the spell out of the Agree operation that takes place in control contexts. 1. Introduction Several studies on the distribution of inflected infinitives in E(uropean) P(ortuguese) (Raposo 1987, Ambar 2007, a.o.) establish a strong correlation between these infinitives and embedded CPs with independent tense on the basis of lexical properties of matrix verbs. Under the standard GB account of inflected infinitives presented in Raposo (1987), the availability of inflected infinitives is a consequence of the interaction of two parameters, the Null Subject Parameter and the Infl Parameter: the inflected infinitive results from the possibility of having an infinitive Infl node with overt agreement features (the Infl Parameter), combined with the option of a Agr specified for Case (the Null Subject Parameter). Raposo (1987) also claims that inflected infinitives only occur as complements to verbs selecting for tensed CPs (i.e. what has been renamed complements with independent tense ). If so, and if tense (in)dependence is a purely lexical property, we should not expect to find any correlation between tense (in)dependence and specific syntactic configurations, that is, a particular verb should select for either a tense independent or a tense dependent complement, regardless of the finiteness of the CP complement, contrary to fact (see (1)). Moreover, since PRO is not allowed to co-occur with overt phi-features in infinitival clauses, controlled inflected infinitives are not expected, a prediction borne out for standard EP but not for non-standard varieties of this language (see (2)). (1) a. O João quis que a Maria saísse amanhã. the João wanted that the Maria leave.sbjv tomorrow João wanted Maria to leave tomorrow. b. *O João quis sair amanhã. the Joao wanted leave.inf tomorrow (2) Decidiram irem todos os presentes à Lagoa 2 decided.3pl go.inf.3pl all the presents to.the Lagoa All those who were present decided to go to Lagoa. (https://sites.google.com/site/boletinsarquivo/) Thus, this study aims at (a) showing that tense (in)dependence of a CP complement is not only dependent on lexical properties of the main predicate but also on the syntactic configuration; (b) arguing that the lexical property of temporal orientation (and not tense independence) explains the distribution of inflected infinitives; (c) adding a new argument for an account of the inflected infinitive that takes the occurrence of overt inflection in the infinitive and the licensing of pro/overt subject as two independent facts, based on a theory of control as Agree and on the occurrence of controlled inflected infinitives (or pseudo-inflected infinitives) in complement clauses of non-standard EP, such as (2). The paper is structured as follows: in section 2, we briefly describe the distribution of inflected infinitives in standard EP both in obligatory control (OC) and non-obligatory control (NOC) contexts; section 3 presents the analysis of standard inflected infinitives in subject OC contexts and its consequences for the distribution of tense (in)dependence, as well as the role of temporal orientation in those contexts; in section 4, we suggest an approach to pseudo-inflected (or controlled inflected) infinitives in nonstandard varieties of EP as a corollary of the analysis sketched in section 3; finally, we make some concluding remarks in section 5. Throughout the paper, only subject control will be considered. 2. The distribution of inflected infinitive in obligatory control vs. non obligatory control contexts As it is well known, a long and lively debate on what counts as the best theory of control has been going on in the literature. Since this debate is not the focus of this paper, we will simply state that we adopt Landau s (2000, 2004) approach to control as Agree, and we believe that the facts presented here may ultimately be seen as an argument in favor of this theory. Landau argues for the distinction between raising and control and refines the distinction between OC and NOC contexts, showing that locality conditions determine the distribution of OC: it applies only in θ-marked infinitives internal to VP, that is, in complement clauses, whereas NOC applies in island contexts, that is, in preverbal subject infinitives and in adjunct clauses. This distribution is accounted for if OC, but not NOC, involves an Agree relation between the matrix functional head F [T, in the case of subject OC] that agrees with the controller and PRO/T-AGR. (Landau 2000, 14). The divide between OC and NOC is particularly interesting in standard EP, a language with inflected infinitive, which is generally available in 3 NOC contexts (see (3)-(4)), but severely restricted in OC contexts (see (5)- (7)). (3) É melhor [nós/pro irmos ao cinema]. is better we go.inf.1pl to.the cinema It is better that we go to the cinema. (4) Os miúdos telefonaram [para os pais os irem the kids called for the parents CL.3PL go.inf.3pl buscar]. pick up The kids called in order for the parents to pick them up. (5) a. O júri declarou irem dois candidatos à the jury declared go-inf.3pl two candidates to.the final. final The jury declared that two candidates go to the final. b. *O júri declarou dois candidatos irem the jury declared two candidates go.inf.3pl à final. to.the final (6) a. O júri lamentou [ CP irem poucos candidatos the jury regretted go-inf.3pl few candidates à final]. to.the final The jury regretted the fact that few candidates (would) go to the final. b. O júri lamentou [ IP poucos candidatos irem the jury regretted few candidates go.inf.3pl à final]. to.the final (7) a. Os pais quiseram ir ao cinema. the parents wanted go.inf to.the cinema The parents wanted to go to the cinema. b. *Os pais quiseram irem os meninos the parents wanted go.inf.3pl the children ao cinema. to.the cinema c. *Os pais quiseram os meninos irem the parents wanted the children go.inf.3plu ao cinema. to.the cinema. d. *Os pais quiseram irem ao cinema. the parents wanted go.inf.3pl to.the cinema e. *O pai quis irem ao cinema. the dad wanted go.inf.3pl to.the cinema 4 The examples in (3) and (4) illustrate the free distribution of inflected infinitives in NOC contexts: subject and adjunct clauses, respectively. On the contrary, in OC contexts, there is a sharp contrast depending on matrix verbs: as Raposo (1987) showed, complements to declarative, factive and epistemic verbs allow the occurrence of inflected infinitive (see (5), (6)), subject-verb inversion being obligatory with declarative verbs (see (5b) vs. (6b)); in contrast, complements to volitional verbs exclude it (see (7)). These general facts characterize the standard distribution of inflected infinitives. As it is well known, Raposo treated the contrast between (5)-(6) and (7) as a consequence of a purely lexical property of the matrix verbs: the fact that they did or did not select for an independent tensed complement clause. However, as we show in the next two sections, his definition of tense (in)dependence is not accurate enough and hence we will both provide a refinement of this concept and examine the role of another temporal relation between the embedded and the matrix clauses in the licensing of inflected infinitives: temporal orientation. 3. The role of tense (in)dependence and temporal orientation in the distribution of inflected infinitive in OC contexts Tense (in)dependence as a result of a particular syntactic configuration As mentioned in the introduction of this paper, Raposo s (1987) classical analysis of EP inflected infinitives associates the distribution of inflected infinitives with a lexical property of matrix verbs: the possibility of selecting for a tensed CP, that is, a type of complement that could be renamed as an independent tense complement. However, examples (8) and (9) show that independent tense in an infinitival complement is not a sufficient condition for the occurrence of inflected infinitives in standard EP, contrary to Raposo s prediction. Assuming that the co-occurrence of the embedded future-oriented adverbial amanhã tomorrow with a matrix verb in the past tense is a reliable test of tense independence, we conclude that both decidir to decide and prometer to promise select for independent tense complements. Yet, inflected infinitive is not allowed in these contexts, as the (b) examples show. (8) a. Eles i decidiram [-] i ir ao cinema amanhã. they i decided [-] i go-.inf to.the cinema tomorrow They decide to go to the cinema tomorrow. b. *O João i decidiu [-] j irmos ao cinema the João i decided [-] i go.inf.1pl to.the cinema (amanhã). (tomorrow) (9) a. Eles prometeram acabar o trabalho amanhã. they promised finish.inf the work tomorrow 5 They promised to finish the work tomorrow. b. *Eles prometeram acabarmos o trabalho amanhã. they promised finish.inf.1pl the work tomorrow The problem is not solved by more recent definitions of (in)dependent tense. Ambar (2007) also resorts to this concept in order to account for the distribution of indicative, subjunctive and inflected infinitive in EP complement clauses. As Raposo (1987), she argues that (in)dependent tense is a lexical property of matrix verbs, which may select for either [+T] (independent embedded tense) or [-T] (dependent embedded tense). Moreover, Ambar (2007) claims that tense (in)dependence is correlated to the possibility of having the t(ense)-features on C-V valued internally to the embedded CP (a tense independent context) or outside it, in the matrix domain (a tense dependent context): inflected infinitives are licensed inside complete CPs, i.e., CPs whose features are internally valued. Thus, tense (in)dependence is already accounted for both in lexical and in syntactic terms, although leaving the contrast in (1) unaccounted for. Within the theory of control as Agree, Landau (2000, 2004) suggests a three-way distinction of embedded tense: anaphoric selected tense; dependent selected tense and free (non-selected) tense. Dependent tense is defined as describing a situation where the tense of the embedded clause is constrained by (though, crucially, not necessarily identical to) the matrix tense. (Landau, 2004, 822). Landau (2004, 851) specifically suggests that EP inflected infinitives are incompatible with irrealis tense (although they are compatible with realis dependent tense, as in factives, or independent tense, as in declarative and epistemic complements or in subject and adjunct clauses). Nevertheless, the following sentence shows that inflected infinitive is compatible with irrealis 3 : (10) Ela acredita acabarem os primos o trabalho she believes finish.inf.3pl the cousins the work amanhã. tomorrow She believes that her cousins will finish the work tomorrow. Furthermore, Landau tests selected tense (anaphoric vs. dependent) with deictic temporal adverbials (such as yesterday / tomorrow) whose features are incompatible with the tense features of the matrix verb. He concludes that: (a) in dependent selected tense (or tensed) contexts, a future-oriented adverbial is allowed despite the morphological tense of the matrix verb; (ii) in anaphoric (or untensed) contexts, this kind of conflicting temporal information produces ungrammatical results (Landau 2004, ). Yet, this type of test is not accurate, since the deictic / non-deictic interpretation of adverbials may change the grammaticality judgments (see (11a) and (11b)). (11) a. *O João quis comprar o jogo amanhã. 6 the João wanted buy.inf the game tomorrow b. O João quis comprar o jogo no dia seguinte. the João wanted buy.inf the game in.the day after João wanted to buy the game the next day. Although both temporal expressions occurring in the infinitival domain have a future interpretation, in (11a) the adverb amanhã tomorrow forces the temporal perspective point (TPpt) of the embedded sentence to be the utterance time whereas in (11b) the temporal PP no dia seguinte the next day forces the TPpt of the embedded sentence to be the time of the matrix clause. This difference is crucial to the grammaticality of the sentences. Gonçalves, Cunha & Silvano (2010) provide a finer-grained definition of tense dependence, which instead involves the semantic concepts of temporal domain (Declerck, 1991) and TPpt (Kamp & Reyle, 1993). The temporal domain concerns the time interval taken up by a situation or by a number of situations that are temporally related to each other; TPpt is defined as the point from which a situation is viewed. Thus, according to Gonçalves, Cunha & Silvano (2010), tense dependence obtains when two situations share the same temporal domain; in this case, the TPpt of the embedded infinitive is the time of the matrix clause (Silvano 2002), which precludes deictic adverbials related to the Utterance Time. This is the case of (11b). If the situations do not share the same temporal domain and the TPpt of the embedded clause may be different from the time of the matrix clause, an independent tense context results (see (12)). (12) a. O João decidiu comprar o jogo amanhã. the João decided buy.inf the game tomorrow João decided to buy the game tomorrow. b. O João decidiu comprar o jogo no dia seguinte. the João decided buy.inf the game in.the day after João decided to buy the game the next day. Some generalizations become clear when we use this definition. First, there is a strong association between tense independence and NOC contexts: a structurally opaque domain for OC (that is, a NOC context) is always tense independent, as in sentential subjects (13) and adjuncts (14). (13) Ir ao cinema amanhã não agradou à Maria. go.inf to.the cinema tomorrow not pleased to.the Maria To go to the cinema tomorrow did not please Maria. (14) A Maria comprou cervejas para beber amanhã. the Maria bought beer to drink tomorrow Maria bought some beer to drink tomorrow. Second, inflected infinitival clauses always show independent tense, as illustrated in (15), with a sentential subject including a deictic adverb 7 incompatible with the future meaning of the auxiliary ir to go in the main clause, and also (16), with a complement clause: (15) Termos ido ao cinema ontem não lhe have.inf.1pl gone to.the cinema yesterday not him vai agradar. go.ind.3sg please.inf The fact that we went to the cinema yesterday will not please him. (16) Os pais lamentaram irmos ao cinema amanhã. the parents regretted go.inf.1pl to.the cinema tomorrow The parents regretted the fact that we will go to the cinema tomorrow. Finally, according to Gonçalves, Cunha & Silvano s (2010) definition, tense (in)dependence is not a purely lexical property of matrix verbs but rather a temporal relation which obtains in specific syntactic structures with specific verbs. The fact that an obligatory tense dependence reading is only generated in the infinitival complement clauses of a particular subset of verbs is the lexical component of this property. In other words, it is the semantic effect of a particular syntactic configuration which may only obtain when some OC verbs occur. This accounts for the fact that some OC verbs obligatorily occur with tense dependent infinitive complements (see (17)), whereas others do not (see (18)). (17) *Os pais quiseram ir ao cinema amanhã. the parents wanted go.inf to.the cinema tomorrow The parents wanted to go to the cinema tomorrow. (18) Os pais decidiram ir ao cinema amanhã. the parents decided go.inf to.the cinema tomorrow The parents decided to go to the cinema tomorrow. Given that obligatory tense dependence only occurs in OC contexts (it never occurs in non-inflected infinitive domains in NOC contexts nor in inflected infinitive complement clauses), we hypothesize that tense dependence may be seen as a possible side effect of subject control. Subject control involves the formation of an Agree chain (Landau 2000) having matrix T as a probe, as illustrated in the simplified representation of subject control structures in (19). In section 3.3, we explore the idea that an Agree chain involving [T] features in matrix T and embedded C-T may generate a tense dependent reading. (19) T matrix DP matrix subject [CP C [TP PRO T [VP PRO ]]] Agree 3 Agree 1 Agree 2 Agree 4 8 In sum, Raposo s (1987) correlation between independent tense and inflected infinitives should be taken in the opposite way: true inflected infinitives are not restricted to independent tense contexts; instead, true inflected infinitive clauses do not show up with dependent tense because the relevant Agree chain cannot be formed, as it happens in control structures Temporal orientation and the licensing of inflected infinitives In the previous section, we accounted for the relation between tense independence and inflected infinitives: an inflected infinitive clause is always an independent tense context. However, several questions remain, namely: which property underlies the distribution of inflected infinitives in OC contexts? What distinguishes (20b) from (21b)? (20) a. O João i afirmou [-] i ir ao cinema amanhã. the João said [-] i go to.the cinema tomorrow João said that he will go to the cinema tomorrow. b. O João i afirmou [-] j irmos ao cinema the João said go.inf.1pl to.the cinema amanhã. tomorrow João said that we will go to the cinema tomorrow. (21) a. O João i decidiu [-] i ir ao cinema amanhã. the João i decided [-] i go to.the cinema tomorrow João decided to go to the cinema tomorrow. b. *O João i decidiu [-] j irmos ao cinema the João i decided [-] i go.inf.1pl to.the cinema amanhã. tomorrow We suggest that the relevant difference is the temporal orientation of the embedded sentence. The matrix verb in (20b) does not specify the temporal orientation of the embedded sentence, that is, it is neuter (in the sense of Cunha & Silvano 2006) with respect to th
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks