The goal of this article is to investigate the syntax of subjects in Brazilian Portuguese (BP), developing a unified analysis for referential and non-referential subjects, on the one hand, and for the so-called topicsubject constructions, VS word order clauses, third person subjects with a generic interpretation without the indefinite pronoun ‘se’, existential and meteorological predicates, on the other hand. By hypothesis, in these contexts, the subject is filled by a (null or overt) DP with (spatial/ temporal) locative interpretation. It is claimed that the relevant facts can be accounted for if the pronominal/inflectional system of BP is analysed as displaying a split in two subsystems: one, which is constituted by the first and second person, being inherently definite/ referential, and the other, which is constituted by the third person, being underspecified for the definite/referential feature. Our proposal is that the split in the pronominal/inflectional system is the key to account for the distribution of third person (spatial/ temporal) locative subjects, either lexical or null in BP, both in matrix and embedded clauses. In the discussion, we present previous analyses on the syntax of subjects of BP, pointing out that the absence of consensus is due to their theoretical implementation (which is often complementary), although their contribution converge with respect to the properties of the third person, as well as to the role of discourse orientation, allowing for the presently proposed unification.

Foreword - Diadorim Especial 2017

This paper examines the properties of inclusive generic constructions, focusing on languages where the inclusive generic pronoun is a null category. We investigate empirical data from a set of languages with and without agreement to test Phimsawat’s (2011) hypothesis that the inclusive generic pronoun lacks all phi-features, and therefore has the least restricted reading, due to there being no restriction on the reference. We show that this hypothesis cannot hold true universally, as phi-features trigger agreement in inflecting languages. We show that there is a correlation between presence of agreement and restriction to human reference for null inclusive generic pronouns, based on comparison of a set of languages without agreement (Thai, Mandarin Chinese, Korean, Vietnamese, and Sinhala) with a set of languages with agreement (Finnish, Brazilian Portuguese, Hebrew, Basque, and Tamil). An explanation in terms of feature architecture is proposed for this correlation. A prediction for generic PRO is discussed and shown to be inconclusive or false.

This paper focuses the distinction between verbal and adjectival passive sentences. For that differentiation to be accounted for, a previous proposal considering a specific functional node for eventive passives - passiveP (LIMA JÚNIOR; AUGUSTO, 2015) is retrieved. It is arguably assumed that the approach based on passiveP, besides maintaining a uniform analysis to active and passive sentences (as VoiceP in Collins (2005)), deals well with intervention issues, and is prosperous in addressing parametric variation and language acquisition facts. As far as the tripartite distinction among the different types of passives is concerned (EMBICK, 2004; DUARTE; OLIVEIRA, 2010), it is proposed that a main bipartite distinction between eventive and adjectival passives may be retained, which is here attributed to the presence of passiveP. Concerning stative and resultative passives, an agreement operation between the auxiliary verbs and the participle (LUNGUINHO, 2011) is assumed to allow for different readings to be obtained. A fourth group of passive-like sentences, involving participles, which have lost their connection with their original verbs, is also syntactically distinguished and treated as actual copular constructions.

In standard Brazilian Portuguese (BP), as well as in other Romance languages, possessives have uninterpretable number features, which are valued via nominal agreement. However, dialects of BP, especially the one spoken in Minas Gerais, have shown that 2nd person possessives, in postnominal position, do not have number agreement with the noun. In order to account for these facts, I will argue that, in this grammar, number features on 2nd person possessives are reanalyzed as being: (i) associated with the person (rather than the noun) and (ii) valued. From the first postulation, ‘seu’ is expected to be the possessive for 2nd person singular, and ‘seus’ for 2nd person plural. From the second postulation, no number concord is expected to be triggered on the possessive. In addition, based on Danon (2011) and Norris (2014), I will argue that cardinals divide BP DPs into two domains in that phrases located above NumP are marked with the plural morpheme, while phrases below it are unmarked. In this sense, because prenominal possessives precede cardinals (NumP), they must be marked with the plural morpheme for nominal agreement; whereas postnominal possessives, which follow NumP, must be unmarked. Free from the plural marking associated with nominal agreement, postnominal 2nd person possessives favor the reanalysis of the morpheme ‘-s’ as indicating the number associated with person features.

In the present study, we analyze the acceptability of 3rd person accusative clitic pronouns (3P-ACC) in written modality as compared to 3P tonic pronouns in the same context. Contemporary Brazilian Portuguese has lost 3P-ACC clitics in oral modality. For this reason, some linguists have preferred not to use 3P-ACC clitics in experimental items, even in reading experiments. Against this trend, we conducted an acceptability judgment task to investigate whether the low frequency of 3P-ACC clitics and the high frequency of 3P tonic pronouns in oral modality drive the evaluation of these pronouns in written modality. Our results suggest that, for students ranging from technical high-school to post graduation, both pronouns are highly accepted. Thus this leads us to maintain that both 3P-ACC clitics and 3P tonic pronouns are adequate for experimental items of reading tasks.

Recent studies on the representation of pronominal subjects in Brazilian Portuguese (BP) show the preference for overt indeterminate referential subjects. A “side effect” of the change in the Null Subject Parameter in BP is the progressive decline in the use of the standard strategies - structures with verb in the third person plural with a null subject and sentences with the pronominal generic clitic se - and the appearance of alternative strategies – an increasing frequency of sentences with overt nominative pronouns, especially você (you) and a gente (the people =we). Our purpose is to refine Duarte’s (1995) data of indeterminate subjects, collected from a sample of the speech of graduates from Rio de Janeiro. Our theoretical framework associates the Theory of Language Variation and Change (WLH, 2006 [1968]) with the Principles and Parameters Theory (CHOMSKY, 1981, 1995), which guides our analysis, from the hypotheses raised to the interpretation of the empirical data (DUARTE, 2016). Our results show that the different strategies are not in direct competition: they can be distributed in three diferent groups, according to a set of features they share, with respect to arbitrary and generic reference. At one extreme we find [+3rd person/+plural] category, which excludes the speaker, represented by the dying arbitrary clitic se and the pronoun eles (they), preferably overt. At the other extreme, we have a [+3rd person/+singular] category, which may or may not include the speaker and the addressee, represented by the generic clitic se, the zero strategy (with a 3rd person singular verb form) and você (you), which is preferably overt. Finally, we have a [+1st person/+plural] category, which does include the speaker, represented by nós (we) and a gente (the people, the folks=we), with considerable advantage with respect to the former. The variation in each category disposed along our scale is not a stable phenomenon: each point has a strong competitor to represent each degree of indeterminate reference as the change progresses.

Based on the Principles and Parameters framework (CHOMSKY, 1995), this article presents a diachronic analysis of European Portuguese (EP) Wh-questions and a comparison with the same structures in Brazilian Portuguese (BP). I also present a brief analysis of Wh-question patterns in sociolinguistics interviews recorded in two periods (years 1970/80 and 2010). The initial hypothesis, inferred from recent theoretical descriptions, was that WhVS order would be more frequent in EP and that WhSV, whenever attested, would be constrained by the presence of the cleft-structure. Taking into account the fact that EP is considered a consistent Null Subject Language (NSL), 1st and 2nd person subjects as well as anaphoric subjects would be preferably null. Therefore, EP would exhibit a different behavior from BP, which has become a WhSV system; the rare cases of SV are attested with unacccusative verbs and a lexical DP subject; and this change, as shown by the analyses reported here, runs parallel to the re-setting of the value of the Null Subject Parameter in BP. The sample analyzed for the diachronic study, comprising theater plays written across the 19th and the 20th centuries, is comparable to the Brazilian sample. For the interviews, I used the samples NURC and Concordância for PB and Cordial-Sin and Concordância for EP. The methodology to codify and submit the data to statistical treatment follows the variationist approach (TAGLIAMONTE, 2006; GUY AND ZILLES, 2007). The results show that EP prefers the WhV pattern, confirming its status of a NSL; as for overt subjects, WhVS is the preferred pattern; however, we can observe a slow decrease of VS in the last quarter of the 20th century, suggesting the implementation of a competition with WhSV, triggered by the introduction of clefting in the second half of the 19th century. Once introduced in the system, the cleft structure expands to all Wh patterns, which is confirmed in contemporary speech data.

Ellipsis phenomena are studied according to two conflicting hypotheses: the semantic approach, which claims that the identification of the ellipsis-antecedent is semantic; and the syntactic approach, which argues in favor of a structural identification between ellipsis and its antecedent. Alternative analyses have argued that there are different types of ellipsis, some of them licensed by information structure, especially topic and focus. In this paper, in order to test the hypotheses presented above, I realized a prosodic study of elliptical sentences produced by 12 female native speakers of BP between 20 and 35 years old. The test consisted of five sentences to be completed by the participants plus 28 images containing two scenes; in each scene there was at least one different item (action, complement or subject). The results showed an interesting pattern. In the case of ellipsis, the given information was generally deaccentuated, while the contrasted information (action, complement or subject) were marked intonationally by High (H) or LowHigh (LH) contour. If we take into account that contrasted itens are moved to the left periphery, leaving a trace in their original position, than we can deduce that in all these cases there are deacentated syntactic structures

This paper aims to investigate possible prosodic differences in two word orders attested for object relative clauses in Karitiana (Tupi-Arikém). In an elicited production task, Vivanco (2014) observed that there are two possible word orders for object relative clauses in this language: OSV and SOV. Assuming the tenets of the Autosegmental Metrical Theory (Pierrehumbert, 1980, 2000, Ladd 2008), we revisit Vivanco’s (2014) data and demonstrate that there is a prosodic difference between these two orders, namely, an obligatory L*+H pitch accent on the stressed syllable of subjects in SOV relatives. On the other hand, OSV object relative clauses do not obligatorily show any specific intonational contour.

This study aims to provide a preliminary account of the morphosyntax of nominalization in Wayoro (Tupian family), an endangered language which is situated in the state of Rondônia (Brazil). In this paper I describe some of the morphosyntactic and distributional properties of nouns and verbs. An additional objective of this paper is to examine the nominal and/or verbal properties of the constructions involving the morpheme {-p} ‘nominalizer’. The data suggest that there are two kinds of constructions involving the morpheme {-p}: a lexical, or VP nominalization, and a clause nominalization.

After Everett’s (2005), the theoretical basis of Generative Grammar has been severely attacked. The mere possibility of existing a language with no self-embedding was taken by many people interested in linguistics be a conclusive proof that the notion of universal grammar is a spurious theoretical construct. However, the arguments for the unavailability of self-embedding in Pirahã are mostly based on the absence of evidence for self-embedding in the language. This calls for a serious discussion of the validity of scientific arguments within linguistics, as absence of evidence is not the same as evidence of absence. Unlike many of the language analyzed throughout this volume, Pirahã is a language with almost no morphological clues to its syntactic structure. Thus, its syntax may not be transparent, and relying on Pirahã superficial simplicity (E-language) as a way of measuring its underlying complexity (I-language) can be a serious faux pas. In this paper, we show that once one investigates the syntax-semantics interface, focusing on interpretative dependencies, evidence for self-embedding in the language are found. Thus, the beauty of the Pirahã grammar as a systematic recursive system is expressed in the cognitive abstract processes of its speakers.

This article aims to describe and examine the verbal agreement system in the Tenetehára language (of the Tupí-Guaraní linguistic family). We assume the hypothesis that the agreement displacement phenomena – which are sensitive to person hierarchies – come from the mechanism of Agree, that operates on articulated f-feature structures in cyclic syntax (Rezac, 2003; Béjar, 2000ab, 2003; Béjar; Rezac, 2009). We explore such agreement displacement in order to understand its syntactic and morphological character and its parameterization in Tenetehára. The analysis of the target language shows that cyclicity and locality derive a preference for agreement control by the internal argument, rather than by the external. Furthermore, the articulation of the probe derives when the agreement control displaces – in terms of cyclic syntax – to the external argument, which is sensitive to the following person hierarchy: 1>2>3[+foc]>3[-foc] (Duarte, 2007). In sum, when the resulting syntactic configurations are submitted to Transfer, properties of the morphological component further parameterize the outcome. Thus, the agreement displacement phenomenon in Tenetehára characterizes at least three classes of derivations corresponding to direct, inverse and direct-inverse contexts.

This paper aims to describe the syntax of the constructions that present the verbal extensions {-ik-} and {-uk-} and their allomorphs in Shimakonde, a Bantu language classified as P23 in the Guthrie classification (GUTHRIE 1967-71). This language is spoken in the northern regions of Mozambique and Tanzania. These verbal extensions are reported in literature under the labels of stative, impositive, pseudo-passive, neuter, and quasi-passive (DOKE, 1947; SATYO, 1985; MCHOMBO, 1993; DUBINSKY & SIMANGO, 1996; BENTLEY & KULEMEKA, 2001; LIPHOLA, 2001; NGUNGA, 2004; KHUMALO, 2009; LEACH, 2010; LANGA, 2013). The addition of the {-ik-} or {-uk-} morphemes to the verb structure usually demotes or suppresses the external argument, turning a basically transitive predicate into an intransitive one. This paper aims to investigate in Shimakonde if alternations from a dyadic to a monadic predicate, through the use of one of the aforementioned morphemes, are instances of the phenomenon known in literature as causative/anticausative alternation (HASPELMATH, 1987, 1993; LEVIN & RAPPAPORT HOVAV, 1992, 1995; NAVES, 1998, 2005; VAN HOUT, 2004; OLIVEIRA, 2011; KALLULLI, 2007). In order to do so, I analyze the grammatical role of this morpheme with two Shimakonde native consultants from different Mozambique districts (Mocimboa da Praia and Montepuez). The fieldwork activities consisted of translations of sentences from Portuguese to Shimakonde, testing the grammaticality of the proposed sentences. In order to examine the data that were collected, I adopted the Alexiadou, Anagnostopoulou and Schäfer (2006) refinement of the verbal categories by Levin & Rappaport Hovav (1992, 1995). One of the results obtained is that the verbal extensions display an atelic reading (giving rise to stative interpretation) or a telic reading (giving rise to anticausative or passive interpretation). To account for the different interpretations in these constructions, I propose distinct associations between Asp head and Voice head in accordance with Kratzer (1996), Pylkkänen (2002), van Hout (2004), and Oliveira (2010).