This chapter considers the distribution of inflected infinitives in European Portuguese (EP), with a particular emphasis on (exhaustive and partial) obligatory control (OC) contexts. It is shown, by means of large-scale survey data, that all EP speakers accept both inflected and uninflected infinitives in OC contexts but with substantial micro-parametric variation. An analysis is proposed whereby there are two distinct kinds of OC in EP, one derived via movement (Hornstein, 1999) and one resulting from Agree plus failed movement. Points of micro-parametric variation concern: the kind of non-finite clause selected for by various matrix predicates, patterns of obviation, and the precise way in which the thematic feature [D: ] is valued.
In this chapter, we describe the distribution of inflected infinitives in Brazilian Portuguese (BP); discuss the possible interpretations of their subjects when null and address the debate between Modesto (2010) and proponents of the movement theory of Control (MTC, particularly Boeckx & Hornstein, 2006; Rodrigues & Hornstein, 2013) as to whether inflected infinitives constitute a serious argument against the MTC. Data were experimentally tested using a grammaticality judgment task, presented here. Results indicate that inflected infinitives are perfectly grammatical in obligatory Control contexts for BP speakers; but its grammaticality is uncertain in non-control contexts, supporting Modesto’s claims.
This paper traces the diachrony of three alternative configurations of infinitival complementation with causative and perception verbs in Portuguese, namely the faire-infinitive, the Exceptional Case Marking and the inflected infinitive constructions. It is shown that the faire-infinitive construction is the earlier pattern of infinitival complementation with causative and perception verbs. The ECM construction is the subsequent innovation and creates the conditions for the appearance of the inflected infinitive. Hence the diachronic development of the structures analyzed in the paper defines a clear path of functional enrichment of the infinitival complements of causative and perception verbs, which gradually acquire greater syntactic autonomy. The structurally ambiguous configurations that lie behind each step of the change are identified in the paper.
Szabolcsi (2009) shows that there are languages where control and raising infinitives have overt subjects, in compliance with (1):
The overt subjects of control complements can only be pronouns.
The overt subjects of raising complements can be pronouns or lexical DPs.
Drawing on data from European Portuguese (as well as Spanish and Italian) we show that the evidence underlying (1) constitutes a strong case in favor of a non-raising approach to obligatory control. Relying on the observation that many consistent Null Subject Languages (NSL) allow for explicit subjects in raising and control complements, we develop an account that aims to capture the association between this phenomenon and the null subject property.
Brazilian Portuguese and Spanish allow subject NP-raising from non-finite clauses, but both languages also allow referential subjects in existential clauses with finite complement clauses. The latter have been referred to as hyper-raising in BP (Martins & Nunes, 2009) and further-raising in Spanish (Fernández-Salgueiro, 2005, 2008). Both structures have been argued to have a matrix subject in an A-position resulting from A-movement raising from the embedded clause. Whereas BP has been argued to require matrix subject-verb agreement, differently from Spanish, we show that there is more variation in this respect. We compare these structures, and adopt a unified A-movement analysis for them, allowing variation only in the specification of agreement (phi-)features in the matrix clause.
Subject obviation refers to the impossibility of the subject of an embedded clause to be co-referent with the main clause’s subject. This is verified in complementation structures of some predicates that select the subjunctive, but not in those of other predicates. A semantic explanation is proposed, based on Giannakidou’s concept of veridicality, for why only some of the predicates that select the subjunctive mood trigger obviation.
We offer a new insight on the acquisition of control, by considering data from Portuguese. We explore two main issues: (i) the distinction between obligatory control and non-obligatory control and (ii) choice of controller. Our results show that children distinguish certain obligatory and non-obligatory control contexts: children accepted a sentence-external antecedent in infinitival subjects but not in infinitival complements. Additionally, our study confirms previous results suggesting that subject control in promise-type contexts is delayed. However, it equally shows that there is no absolute bias for object control at early stages. We evaluate predictions made by contemporary accounts of control, namely the movement theory of control (Hornstein, 1999) and the theory of control as Agree (Landau, 2000, and subsequent work).
This chapter aims at analyzing the acquisition of complements to causative verbs in Mozambican Portuguese (MozP) as L1 by pre-school children. By presenting the results of a completion task (Santos, Gonçalves, & Hyams, 2016), we show that Mozambican children, like Portuguese ones, take propositional complements as complete functional domains, which explains the high occurrence of inflected infinitives. However, MozP presents an innovative pattern in which the causative verb selects for a complement with the structure [DP XP]. We claim that this complement corresponds to a pseudo-relative, a structure which is severely constrained in Portuguese.
This study investigates the acquisition of morphosyntactic and interpretative properties of the Portuguese inflected infinitive by Chinese and Spanish speakers, using three experimental tasks. Assuming a Full Access approach (Schwartz & Sprouse, 1996) and the Interface Hypothesis (Sorace, 2011), we predict that full acquisition of the morphosyntactic properties is possible, albeit showing L1 effects, and that interpretative properties develop later than narrow syntactic properties. Both predictions are supported. Our findings show that the advanced learners have largely acquired the relevant morphosyntactic properties, although there is evidence of developmental delays, particularly in the Spanish group. Moreover, both advanced groups performed non-targetlike regarding the interpretation of overt subjects; however, the Spanish group showed an advantage over the Chinese group with null subjects.
This study aimed to verify if Portuguese children show an interpretative asymmetry between null and overt subject pronouns in indicative and subjunctive complement clauses. In the indicative, children overaccepted the dispreferred coreferential reading with overt pronouns (argued to be licensed post-syntactically); children performed more adult-like with null pronouns (considered to be licensed in syntax) when there was only one intrasentential antecedent (the matrix subject). However, when a matrix object antecedent was added between the preferred matrix subject antecedent and the null embedded subject pronoun, they often accepted the dispreferred reading of disjoint reference. In subjunctive clauses, children incorrectly assigned coreferential readings to both pronouns. We assume that subjunctive obviation is partly dependent on lexical-semantic knowledge, taking time to be acquired.
This chapter considers the distribution of inflected infinitives in European Portuguese (EP), with a particular emphasis on (exhaustive and partial) obligatory control (OC) contexts. It is shown, by means of large-scale survey data, that all EP speakers accept both inflected and uninflected infinitives in OC contexts but with substantial micro-parametric variation. An analysis is proposed whereby there are two distinct kinds of OC in EP, one derived via movement (Hornstein, 1999) and one resulting from Agree plus failed movement. Points of micro-parametric variation concern: the kind of non-finite clause selected for by various matrix predicates, patterns of obviation, and the precise way in which the thematic feature [D: ] is valued.
In this chapter, we describe the distribution of inflected infinitives in Brazilian Portuguese (BP); discuss the possible interpretations of their subjects when null and address the debate between Modesto (2010) and proponents of the movement theory of Control (MTC, particularly Boeckx & Hornstein, 2006; Rodrigues & Hornstein, 2013) as to whether inflected infinitives constitute a serious argument against the MTC. Data were experimentally tested using a grammaticality judgment task, presented here. Results indicate that inflected infinitives are perfectly grammatical in obligatory Control contexts for BP speakers; but its grammaticality is uncertain in non-control contexts, supporting Modesto’s claims.
This paper traces the diachrony of three alternative configurations of infinitival complementation with causative and perception verbs in Portuguese, namely the faire-infinitive, the Exceptional Case Marking and the inflected infinitive constructions. It is shown that the faire-infinitive construction is the earlier pattern of infinitival complementation with causative and perception verbs. The ECM construction is the subsequent innovation and creates the conditions for the appearance of the inflected infinitive. Hence the diachronic development of the structures analyzed in the paper defines a clear path of functional enrichment of the infinitival complements of causative and perception verbs, which gradually acquire greater syntactic autonomy. The structurally ambiguous configurations that lie behind each step of the change are identified in the paper.
Szabolcsi (2009) shows that there are languages where control and raising infinitives have overt subjects, in compliance with (1):
The overt subjects of control complements can only be pronouns.
The overt subjects of raising complements can be pronouns or lexical DPs.
Drawing on data from European Portuguese (as well as Spanish and Italian) we show that the evidence underlying (1) constitutes a strong case in favor of a non-raising approach to obligatory control. Relying on the observation that many consistent Null Subject Languages (NSL) allow for explicit subjects in raising and control complements, we develop an account that aims to capture the association between this phenomenon and the null subject property.
Brazilian Portuguese and Spanish allow subject NP-raising from non-finite clauses, but both languages also allow referential subjects in existential clauses with finite complement clauses. The latter have been referred to as hyper-raising in BP (Martins & Nunes, 2009) and further-raising in Spanish (Fernández-Salgueiro, 2005, 2008). Both structures have been argued to have a matrix subject in an A-position resulting from A-movement raising from the embedded clause. Whereas BP has been argued to require matrix subject-verb agreement, differently from Spanish, we show that there is more variation in this respect. We compare these structures, and adopt a unified A-movement analysis for them, allowing variation only in the specification of agreement (phi-)features in the matrix clause.
Subject obviation refers to the impossibility of the subject of an embedded clause to be co-referent with the main clause’s subject. This is verified in complementation structures of some predicates that select the subjunctive, but not in those of other predicates. A semantic explanation is proposed, based on Giannakidou’s concept of veridicality, for why only some of the predicates that select the subjunctive mood trigger obviation.
We offer a new insight on the acquisition of control, by considering data from Portuguese. We explore two main issues: (i) the distinction between obligatory control and non-obligatory control and (ii) choice of controller. Our results show that children distinguish certain obligatory and non-obligatory control contexts: children accepted a sentence-external antecedent in infinitival subjects but not in infinitival complements. Additionally, our study confirms previous results suggesting that subject control in promise-type contexts is delayed. However, it equally shows that there is no absolute bias for object control at early stages. We evaluate predictions made by contemporary accounts of control, namely the movement theory of control (Hornstein, 1999) and the theory of control as Agree (Landau, 2000, and subsequent work).
This chapter aims at analyzing the acquisition of complements to causative verbs in Mozambican Portuguese (MozP) as L1 by pre-school children. By presenting the results of a completion task (Santos, Gonçalves, & Hyams, 2016), we show that Mozambican children, like Portuguese ones, take propositional complements as complete functional domains, which explains the high occurrence of inflected infinitives. However, MozP presents an innovative pattern in which the causative verb selects for a complement with the structure [DP XP]. We claim that this complement corresponds to a pseudo-relative, a structure which is severely constrained in Portuguese.
This study investigates the acquisition of morphosyntactic and interpretative properties of the Portuguese inflected infinitive by Chinese and Spanish speakers, using three experimental tasks. Assuming a Full Access approach (Schwartz & Sprouse, 1996) and the Interface Hypothesis (Sorace, 2011), we predict that full acquisition of the morphosyntactic properties is possible, albeit showing L1 effects, and that interpretative properties develop later than narrow syntactic properties. Both predictions are supported. Our findings show that the advanced learners have largely acquired the relevant morphosyntactic properties, although there is evidence of developmental delays, particularly in the Spanish group. Moreover, both advanced groups performed non-targetlike regarding the interpretation of overt subjects; however, the Spanish group showed an advantage over the Chinese group with null subjects.
This study aimed to verify if Portuguese children show an interpretative asymmetry between null and overt subject pronouns in indicative and subjunctive complement clauses. In the indicative, children overaccepted the dispreferred coreferential reading with overt pronouns (argued to be licensed post-syntactically); children performed more adult-like with null pronouns (considered to be licensed in syntax) when there was only one intrasentential antecedent (the matrix subject). However, when a matrix object antecedent was added between the preferred matrix subject antecedent and the null embedded subject pronoun, they often accepted the dispreferred reading of disjoint reference. In subjunctive clauses, children incorrectly assigned coreferential readings to both pronouns. We assume that subjunctive obviation is partly dependent on lexical-semantic knowledge, taking time to be acquired.