Vulnerability in Heritage Speakers of Spanish in the Netherlands (original) (raw)
A Cognitive Account of Mood in Complements of Causative Predicates in Spanish
Hispania, 2014
Complements of causative predicates such as hacer in Spanish present a problem for analyses of mood that are based on semantic or pragmatic notions of assertion. The problem results from the fact that information expressed by these complements is presented both as true and new, and yet the complement verb appears in the subjunctive mood. This makes these clauses a counterexample to analyses that claim that asserted propositions appear in the indicative. This study proposes an account of the use of mood in these complements that combines the Cognitive Grammar notion of clausal grounding ) with Cristofaro's study of subordination (2003). It will be argued that the complement of causative predicates in Spanish is not independently grounded; that is, its temporal relation to the speech situation is established via the matrix predicate. Not being independently grounded, it does not have an autonomous profile ). Following Langacker (1991) and Cristofaro (2003, the current study claims that for a clause to be asserted, it must have an autonomous profile. This necessary condition is not met in complements of causative predicates and consequently they cannot be asserted.
Ser and estar: The syntax of stage level and individual level predicates in Spanish
… Wöllstein-Leisten & Claudia Maienborn (Hgg …, 2005
Ser and estar: The syntax of stage level and individual level predicates in Spanish * Studying the relevant works concerning the distinction between stage level and individual level predicates (SLP/ILP) in linguistics, one often encounters references to the Spanish copular verbs ser and estar (e.g. . In these copular verbs, the SLP/ILP-distinction seems to find its overt realization. The use of the verb ser is usually connected to ILP-characteristics, the use of estar to the SLP-phenomenon. We propose a minimalist account for the differences in semantic and syntactic behaviour of ser/estar (following Chomsky 1995). Contrary to Kratzer (1995), we assume an implicitly realized event argument for both SLPs and ILPs, which characterizes the spatiotemporal reference of the situation or eventuality expressed by the predicate (cf. Davidson 1967). This event position is localized in the predication phrase PrP as proposed by . The PrP represents an extension of the VP-shell analysis (s. Larson 1988) to non-verbal predication, as found in copulative constructions. We assume that complex interactions between the features of the Pr-head and the features of the minimalist T° (I°) (and probably also the C°) will result in either the SLP or the ILP interpretation. The Spanish data concerning ser and estar allow us to analyse the syntactical conditions which lead to the SLP/ILP-distinction, making the correlation between syntactic and semantic behaviour evident. We propose that both ser and estar are syntactical default strategies (last resort). If the predicate is a SLP and no verb is available in the numeration, then estar will be introduced into the derivation under Pr°. If the predicate is an ILP and no verb is available in the numeration, then ser will be merged under T°. Quantificational approaches to times, especially reference time are also taken into consideration. Where the SLP/ILP-distinction is not expressed syntactically (by ser or estar) we assume that the chosen interpretation results from spatiotemporal knowledge of the world, i.e. it is conditioned by pragmatics (s.
On the Internal Structure of Spanish Verbless Clauses
In this dissertation I discuss several aspects of the syntax, semantics and discourse properties of what I call Spanish verbless clauses -i.e. non-finite utterances with clausal properties: Spanish PredNP muy listo este tío 'very intelligent this guy', Spanish PP complement clause me sorprende lo caro del piso 'it amazes me how expensive this apartment is', and Spanish Qualitative Binominal Noun Phases (QBNPs): comparative QBNP el tonto del alcalde 'the idiot of the mayor' and attributive QBNP un tonto de alcalde 'an idiot of a mayor'. In each of these clauses there is a subject-predicate relationship. The analysis advanced here is that Spanish PredNP, PP complement clauses and comparative QBNP can be analyzed in terms of predicate inversion. Movement of the predicate in each case is due to a strong semantic feature [+X] that needs to be discharged in the course of the syntactic derivation. It is argued that in Spanish PredNP the strong feature evaluativity [+E] is responsible for predicate inversion, while in PP complement clauses and comparative QBNP the strong feature gradability [+G] triggers movement of the predicate over its subject.
Lingua, 2017
This work introduces a subset of informational features (termed core intentional features), different from standard pragmatic features such as topic and focus. Adopting the basic tenets of the Minimalist program, core intentional features are defined as edge features which sit in the relevant phases and are subject to parametric variation. They are assumed to drive the derivation of the sentence so that it constitutes an intentionally-adequate object (i.e. a categorical or a thetic statement) even in the absence of a particular communicative situation. The paper specifically focuses on one of these features, [DI] (discourse intention), and on how it determines the eventual position of the subject in a discourse-prominent language such as Spanish. A preliminary distinction is made between sentences that inaugurate the discourse (d-sentences) and sentences which are integrated in a particular context (context-dependent sentences). It is argued that the SV/VS order in Spanish follows from the conditions of valuation of [DI] in each case; in particular, valuation of [DI] in d-sentences will be a matter of structural and semantic prominence whereas in context-dependent sentences it will depend on pragmatic conditions. The paper also addresses a number of significant contrasts in the much-debated issue of the placement of the subject in Spanish, which receive a principled explanation under the theory of core intentional features proposed here.
Pragmatic Features at the L2 Syntax-Discourse Interface
BUCLD 35 …, 2011
A recent version of the Interface Hypothesis (e.g., Sorace & Serratrice, 2009) proposes a principled distinction between internal interfaces, those between narrow syntax and other linguistic modules (phonology, morphology, semantics), and external interfaces, those between syntax and other cognitive modules. A primary example of an external interface is the syntax/discourse interface: a major source of difficulty, causing delays as well as indeterminacy of judgments and residual optionality even at near-native levels of acquisition (Belleti, Bennati & Sorace, 2007). However, L2 acquisition findings to date are far from conclusive (Rothman, 2009; Valenzuela, 2005, 2006; Ivanov, 2009). This experimental study looks further into processes at this interface, teasing apart acquisition of syntactic, semantic and discourse knowledge. Lopez (2009) offers a new theoretical model of syntax-information structure interaction, proposing a pragmatic computation module that assembles sentences into Discourse Representation Structures in parallel to the syntactic computation. Topic and Focus are not primitives of the grammar, but descriptive labels. The crucial information structure notion are discourse anaphor and contrast, encoded by the features [±a(naphor)] and [±c(ontrast)] which in combination account for the conditions and effects of dislocation and fronting. In Clitic Left Dislocation (CLLD, ex. 1) or Clitic Right Dislocation (CRLD, ex. 2) the feature [+a] is operative: clitics double an overt argument with an obligatory link to an antecedent. A constituent which moves to the left periphery is marked [+c]: it presents a contrast (ex. 3 vs. 4). Table 1 presents the constructions and their features. Furthermore, CLLD and CRLD are felicitous in slightly different contexts due to differing constraints on the anaphor-antecedent relationship (Villalba, 2000). CLRD requires a relationship of identity between it and the discourse antecedent (ex. 5), while the anaphor-antecedent relationship in CLLD is much freer: it can be subset, superset, part/whole (ex. 6). Keeping in mind all these constraints, the learning tasks for English natives acquiring L2 Spanish involve: 1) syntactic knowledge of clitics; 2) discourse appropriateness of the clitic-doubled dislocations; 3) semantic constraints on the antecedent-dislocate relationship. We created 40 context-test sentences combinations: 10 CLLD as in (5) and 10 CLRD with and without identity between antecedent and dislocate, 5 Fronted Focus and 5 Rheme constructions as in (3) and (4), plus 10 fillers. Each context was followed by a sentence with a clitic and one without a clitic. Context stories and test sentences were presented both aurally and visually. Participants were asked to judge sentences as felicitous or infelicitous in the context of the story, by selecting on a scale of 1 to 4 or “I don’t know.” Syntactic knowledge of clitics was independently ascertained as a condition for inclusion in the overall study via a forced-choice task. Mean acceptance ratings of 20 near-natives and 20 advanced participants indicate that all three types of properties can be successfully acquired. However, there are differences in the acceptability of CLLD and CLRD, since the latter is much less frequent in the input, and not readily accepted by our 21 native speakers (SD>.9). However, the near-native and advanced learners confirmed the research hypothesis that even the semantic difference can be acquired. Implications of these results for the Interface Hypothesis are discussed.
A Systematic Investigation of the Spanish Subjunctive: Mood Variation in Subjunctive Clauses
A Systematic Investigation of the Spanish Subjunctive: Mood Variation in Subjunctive Clauses, 2021
The link to the full text of this Ph.D. thesis is available at Georgetown University's Institutional Repository: http://hdl.handle.net/10822/1062622 Standard Spanish grammar states that desideratives (querer que), directives (aconsejar que ), purpose clauses (para que), causatives (hacer que), emotive-factives (alegrarse de que), negated epistemics (no creer que), dubitatives (dudar que), and modals (ser posible que) embed subjunctive complement clauses. However, in spite of these classifications, some predicates exhibit a certain degree of mood variation. For instance, emotive-factives can take indicative complements (Crespo del Río, 2014; Faulkner, 2021). Similar variability between the moods may also come about in negated epistemic (Bolinger, 1991), dubitative (Blake, 1981), and modal clauses (Deshors and Waltermire, 2019). When both moods are acceptable, the choice between the two conveys a difference in meaning (Kanwit and Geeslin, 2014, 2017). Corroborating this hypothesis are findings from Faulkner (2021), which demonstrated that the mood variation occurring in emotive-factive clauses relates to the INFORMATIVENESS of the embedded proposition; i.e., propositions that are addressee-new versus addressee-old. The objective of the current dissertation is to investigate the standardly subjunctive environments in which the use of the indicative may become acceptable. Two hundred and twenty-three native Spanish-speakers completed a 128-item Acceptability Judgment Task (AJT), pertaining to the use of each mood in eight traditionally subjunctive complements (desiderative, purpose, directive, causative, modal, dubitative, negated epistemic, and emotive-factive clauses). Participants rated separate instances of subjunctive and indicative based on the (un)informativeness of the complement proposition. Statistical analyses of the ratings revealed that subjunctive is always preferred over indicative in traditionally subjunctive environments. However, certain predicates were more receptive than others to taking indicative clauses. Whereas preference-based expressions (desideratives, directives, purpose clauses, causatives) require subjunctive complements, emotive-factives and verbs of uncertainty (negated epistemics, dubitatives, modals) may accept indicative if the speaker intends to add the affirmative or negated proposition to the common ground. The likelihood that this occurs increases when the proposition (affirmative or negated) is informative. Since this was the case regardless of the participants’ countries of origin, said findings suggest that the Spanish mood system involves a split between two types of subjunctives, one that is required in preference-based contexts, and another that is default and can be replaced by the indicative. INDEX WORDS: Spanish, Verbal Mood, Mood, Subjunctive, Indicative, Mood Variation
How Much Room for Discourse in Imperative? The Lens of Interface on English, Italian and Spanish *
Studia Linguistica, 2021
This paper discusses root phenomena in imperative clauses, assuming as diagnostics conversational dynamics and the type of discourse categories that are admitted in their Cdomain, through a systematic comparative interface investigation in three languages (English, Italian and Spanish) based on an original experimental work. This novel perspective sheds new light on the syntax-semantic mapping and the interface (syntax-prosody) properties of imperative clauses, embedding the relevant proposal in a cartographic framework of analysis. Based on a twofold distinction of root phenomena-those which are widely allowed in Common Ground-active (Type I) contexts and those which can occur in non-Common Ground-active contexts (Type II)-it is proposed that imperatives are non-Common Groundactive propositions with no update potential, thus allowing only Type II root phenomena. Syntactically, imperative clauses are dominated by a super-ordinate Speech Act Phrase, including the SPEAKER and the ADDRESSEE as co-arguments, which explains the blocking effects identified in imperatives. * Earlier versions of this paper have been presented at the "Cambridge Comparative Syntax Meeting" (UK), the "26 th Generative Linguistics in the Old World" in Gottingen (Germany), the 24 th "Colloquium on Generative Grammar" in Madrid (Spain) and the 1 st International Workshop on the "Interface of Information Structure and Argument Structure" in Seville (Spain). We are very grateful to the audiences there. We also thank two anonymous STUL reviewers for their precious comments and suggestions. The research here has been partially funded by research project PGC2018-093774-B-I00 of Spain's Ministry of Science, Innovation, and Universities (MICINN). The order of authors is strictly alphabetical, so both of us are first authors. However, the empirical puzzle emerges when we are confronted with imperatives which do permit topics in their left periphery, as in (2), also from Haegeman (2012: 64): (2) a. The tie give to Bob, the aftershave give to Don. (Van der Wurff 2007) b. Anything you don't eat put back in the fridge. (Huddleston & Pullum 2002) This raises the question as to what properties imperatives must satisfy in order to license or not certain types of discourse categories. A plausible solution to this problem can be offered if we assume a distinction between different types of Topics. In this respect, Frascarelli & Hinterhölzl (2007) propose a typology of topics based on the systematic correlation between their formal properties and their function in discourse, which is encoded in dedicated functional projections in the left periphery of the sentence. The authors thus produce substantial evidence for the existence of (at least) three types of Topics, namely Aboutness-Shift (A-)Topics, Contrastive (C-)Topics and Familiar/Given (G-)Topics (cf. §3 below for details; see Cardinaletti 2009 and Menza 2006 for the distinction between A-and G-Topics in different constructions). Based on this distinction, a preliminary observation can immediately lead to the conclusion that shifting (i.e. A-)Topics can hardly be associated with the imperative mood. Consider for instance (3a-b) below, from Italian: 1 1 The list of abbreviations and symbols used in the article is the following: ALL = allocutive, AT = Aboutness-Shift Topic, CL = clitic (pronoun), CLLD = clitic left dislocation, CF = Corrective Focus, CG = Common Ground, CGr = Control Group, CT = Contrastive Topic, DIR = directive feature, DP = determiner phrase, F = feminine, FINP = Finiteness phrase, F 0 = Fundamental frequency, FORCEP = Force phrase, GT = Given Topic, IMP = imperative, IMPERS = impersonal, IND = indicative, IP = Inflectional phrase, IRR = interface root restriction, JUSS = jussive, LD = left dislocation, MF = Mirative Focus, NEG = negative, PL = plural, PST = past tense, REFL = reflexive, SAP = Speech act phrase, SG = singular, SUB = subordinate clause, SUBJ = subjunctive, TBU = tone bearing unit, TOP = topicalization, TP = Tense phrase, vP = Verb phrase. The present tense is assumed as deafult and, as such, it is not indicated in the glosses. Notice that throughout the paper, the same gloss will be provided for Italian and Spanish sentences when these are identical; otherwise, a different gloss will be given for each language. However, as translations show, English appears to provide some cross-linguistic differences. Cormany (2013) argues that non-contrastive topics are not allowed in English 2 and, in general, left-peripheral arguments are often unacceptable (from Jensen 2007), as illustrated in (1) above. Nevertheless, this is not absolute. The cases considered in (2a-b) and (the translations in) (4) show that a contrastive interpretation for the topic constituents, obtain acceptable results; hence, C-Topics can be realized in English imperative clauses. As for sentence (5), the translation shows that 'the ball' would be left in situ by speakers in this context. As a matter of fact, la palla/la pelota cannot be considered an A-Topic, because its mention is not used to introduce (or shift) the sentence Topic. Indeed, 'the ball' can be considered a background/given element in the context of a soccer game. Hence, it is a G-Topic in the framework assumed in this work and G-Topics are not dislocated in English, but rather realized through in situ destressing (as is argued in Bianchi & Frascarelli 2010). Similar to Topics , different types of Foci have been argued to exist in the literature, with different formal and discourse properties (cf., among others, Kiss 1998, Aboh, Hartmann & Zimmermann 2007, Frascarelli 2010a, Cruschina 2011, Bianchi, Bocci & Cruschina 2015). In particular, Bianchi & Bocci 2012 argue for a syntactic and semantic distinction between Mirative and Corrective Foci and, interestingly, the realization of these two Focus types in imperative clauses also obtains different results at a first observation. As for Mirative Focus (MF), the contrast between the declarative sentence (6a) and sentence (6b) immediately show that MF 'clashes' with the imperative mood (for reasons that 2 Cormany (2013:100-101) specifically argues that in English only contrastive topics are allowed in imperative clauses, in clear contrast with declaratives (in which both contrastive and non-contrastive topics are licensed), as is illustrated in (i)-(ii): (i) a. The book, John bought ___. b. *The book, buy ___! (ii) a. These stocks, the broker bought immediately. b. These stocks, buy immediately! (Those avoid at all costs!) This contrast is adduced to a relatively poor left periphery in imperative clauses (similar to Haegeman's 2004 notion of 'truncation'). As we will see in Section 5 and 6, however, a truncationbased analysis cannot account for cross-linguistic data. In addition, truncation cannot account for the grammaticality difference of examples in (1)-(2) showing that, under certain circumstances, topics can occur in the left periphery of imperatives. 3 The nature and properties of the relevant 'root operator' is not specified in Bianchi & Frascarelli (2010). Nevertheless, considering the formal and discourse properties of an A-Topic (cf. Section 2 below), this can be identified with Krifka's (1995) Assert Operator insofar as it has the function of updating the common ground by asserting a proposition that is informative, non-contradictory and implies alternatives. As this Operator is encoded in a functional projection in the C-domain (cf. also Meinunger 2004), it is perfectly in line with the present syntax-prosody-semantic interface approach 'Speak!' 'Speak (2PL)!' (35) a. No hables! b. No habléis! not speak.SUBJ.2SG not speak.SUBJ.2PL 'Don't speak!' 'Don't speak (2PL)!'
ENTAILMENTS, PRAGMATIC ASERTION AND MOOD IN SPANISH COMPLEMENTS
This paper examines a frequently overlooked class of expressions in Spanish that license the subjunctive mood in a complement clause. This class contains expressions such as poco/a/s "few", menos de "less than" and solo "only". The goal of the paper is to offer an account of the use of mood with these expressions that incorporates the data under discussion into previous pragmatic accounts of mood based on notions of assertion and informative value. The paper first offers a semantic characterization of this class of expressions that is based on their monotonic properties (Ladusaw 1980, Ladusaw 1983). Next, it explores the pragmatic effects of their semantic properties. Following Stalnaker (1978), I assume that the effect of a pragmatic assertion is to reduce the set of possible worlds that represents the presuppositions held by a speaker and their audience (referred to as the context set). It is argued that propositions under the scope of an upward entailing expression are more informative, and they are thus more relevant and higher in a scale of assertability (following Lunn 1989), in that they allow for inferences that further reduce the context set. Propositions under the scope of expressions that are not upward entailing lose some of their informative value, and thus they are lower in a scale of assertability, which correlates with the possibility of using subjunctive mood in a complement under their scope.
Cuza & Frank (2011)
This study examines the role of transfer from English in the acquisition of double-que questions in Spanish among 17 heritage speakers in the US. Results from an elicited production task, an acceptability judgment task and a preference task revealed significant difficulties in the production and acceptability of double-que questions. In contrast with interface vulnerability approaches suggesting no difficulties at the syntax-semantics interface, the participants showed a decreased level of use of double-que structures and no distinction in their acceptability of statements versus questions. However, results from the preference task showed sensitivity to double-que questions among 10 of 17 heritage speakers. It appears that only when the two structures are presented together were the heritage speakers able to perceive the semantic shift introduced by the double-que. The results suggest that transfer from the other language prevents the complete acquisition of these properties even at high levels of bilingual proficiency.