Linguistics in the Netherlands - Volume 31, Issue 1, 2014

Dutch perfect tenses are constructed with the auxiliaries zijn ‘be’ and hebben ‘have’. Over the years, several semantic generalizations have been proposed to account for this auxiliary choice. This paper evaluates these generalizations on the basis of examples with verbs of directed motion (draaien ‘turn’, keren ‘turn’, and stijgen ‘ascend, rise’), collected from the Internet. The paper concludes that these data are problematic for analyses of auxiliary choice in terms of telicity and ‘Inferable Eventual Position or State’, yet provide support for analyses of auxiliary choice that rely on the notion of ‘construal’ of the motion event. In particular, the paper argues that zijn is used when the event is construed as a ‘change of state’, while hebben is used when it is construed as a ‘type of act’.

This paper is concerned with the correct characterization of the licensing condition on clausal ellipsis and how it relates to the distribution of ellipsis. I argue, essentially following López (2000), that ellipsis is licensed when the ellipsis clause bears a relation to an antecedent in the discourse component. A relation between two discourse units can be established in two ways: (1) Either there holds a direct relation between the two discourse units or (2) there holds an anaphoric relation mediated by a discourse anaphor. In this paper, I show how this two-way distinction in setting up discourse relations accounts for the two-way split we find in the distribution of ellipsis.

In this article I investigate the internal syntax of Dutch nominal expressions like twee slangen van elk twee meter (two snakes of each two meter). The preposition van separates the quantifier elk from the expression twee slangen to which it is connected interpretatively. From a purely observational view, the quantifier could be said to be floating in the nominal domain. In this article I argue that this floating elk is an instance of so-called ‘binominal each’. So far, discussions of the binominal-each-phenomenon have concentrated on dependencies involving material in the clausal domain. In this article I show that dependencies involving binominal elk are also attested DP-internally. Syntactic issues that will be discussed in the context of the DP-internal binominal-each-phenomenon include: Constituency of the nominal expression, the nature of the DP-internal anaphoric dependency, the various manifestations of the binominal pattern, and word order rearrangements by DP-internal displacement.

Parentheses do not affect the semantic truth conditions of the host clause, but they do affect the discourse structure. We propose a maximally simple update system for the conversational context. Presuppositions are treated as past requests for the interlocutor’s consent. Parentheticals act like overt presuppositions unless they are linearly last in the utterance, in which case they can be taken as a current update request. This has consequences for the interlocutor’s ability to target a parenthetical message. We predict that sentence-final parentheses, and in particular attributive appositives, can be generically addressed, but medial ones only by a specific response. We also discuss why certain non-clausal parentheses, including identifying appositions, behave differently.

Except for finite verbs in main clauses, verbs in Standard Dutch cluster together in a clause-final position. In certain Dutch dialects, non-verbal material can occur within this verb cluster (Verhasselt 1961; Koelmans 1965, among many others). These dialects vary with respect to which types of elements can interrupt the verb cluster, varying from particles to various types of arguments and adverbs (Barbiers, van der Auwera, Bennis, Boef, de Vogelaer & van der Ham 2008). A study amongst forty Dutch dialect speakers reveals an ordered ranking of grammatical types, reflecting their acceptability in a verb cluster. I argue that this ranking directly follows from syntactic principles: The syntactic size and position of the intervening element affect its acceptance in a verb cluster. Potentially, these principles interact with a preference of performance dubbed ‘minimize domains’ (Hawkins 1994, 2003, 2004), which requires both the higher verb and the intervening element to be adjacent to the main verb, leading to two conflicting structures.

This article discusses a special type of equative comparison constructions: Potentiality equatives, i.e. equatives with a standard clause containing a predicate of physical or cognitive ability, or epistemic possibility. The focus in this article is on potentiality equatives involving a (negative) comparative or a superlative in the standard clause, as these seem to give a particularly clear insight into the properties of this construction type. It is argued that the broad syntactic variation in potentiality equatives is caused by their specific semantics, by their expressive force and by their idiomatic nature.

Domains where languages have two or more competing syntactic constructions expressing the same meaning may be problematic for bilingual heritage speakers. One such variable domain is the resultative constructions in heritage Ambon Malay, a variety spoken in the Netherlands by Dutch-Ambon Malay bilinguals. In Ambon Malay, resultatives are expressed mostly by means of verb serialization (SVC), although resultative prepositional phrases (PP) and adjectival phrases (AP) also occur. In Dutch, resultative constructions usually involve verb particles, PPs and APs. This overlap of structures poses the conditions for transfer effects between the two languages. The frequency distribution of SVCs, PPs and APs is investigated in semi-spontaneous speech from heritage speakers of Ambon Malay and compared to that of baseline speakers. Heritage speakers show an increase in the frequency of constructions shared by both languages (PPs and APs), while they underuse the constructions attested only in the heritage language (SVC).

In this paper we present an analysis of Spanish heritage speakers’ oral production of gender agreement outside the DP as an innovative source of support for the Interface Hypothesis (Sorace & Filiaci 2006). We demonstrate that, besides commonly known factors such as the gender, animacy and morphology of the antecedent, the interface domain in which gender agreement takes place also seems to play a role in how accurately heritage speakers apply gender agreement. Pronominal reference, located at the external syntax-discourse interface, turns out to be more problematic than adjectival predication, which pertains to the internal morpho-syntax interface. Furthermore, we discuss the possibility that, besides the amount of input heritage speakers receive, the quality of this input may also play a role in their gender agreement accuracy, given that the heritage speakers’ error pattern with respect to linguistic factors is very similar to that of first generation immigrants.

This study reports on the choice between a definite and an indefinite article by children with High Functioning Autism (HFA) and children with Specific Language Impairment (SLI). We carried out an elicited production task with 16 Dutch-speaking non-grammatically impaired children with HFA aged 6–13, 16 age-matched Dutch-speaking children with SLI, and 16 typically developing (TD) age controls. The results in the indefinite conditions reveal virtually no errors across groups. However, in the definite condition the HFA group, but NOT the SLI group, incorrectly produces indefinite articles significantly more often than the TD group. A more detailed analysis shows that 38% (6/16) of the children with HFA vs. 13% (2/16) of the children with SLI regularly produce indefinite articles in definite contexts. We propose that these children do not always calculate the additional (pragmatic) meaning of indefinites derived by scalar implicature (Horn 2006). Furthermore, development by age in the SLI group, but NOT in the HFA group, suggests that the failure to draw a scalar implicature is more persistent in children with HFA than in children with SLI. Concluding, our results show that non-grammatically impaired children with HFA are more prone to pragmatic impairments than children with SLI, suggesting a dissociation between grammar and pragmatics.

While Dutch welke ‘which’-questions are structurally ambiguous, number agreement cues can disambiguate them. Despite such agreement cues, children misinterpret object questions as subject questions (Metz et al. 2010, 2012; Schouwenaars et al. 2014). We investigated if adding another cue, specifically, topicality in a discourse context, helps the interpretation of which-questions in two groups of Dutch children (5;5, n = 15 and 8;5, n = 21). Using a referent-selection task, we manipulated number on the verb and postverbal NP to create unambiguous wh-questions. Moreover, the questions were preceded by a discourse which established a topic, relating either to the wh-referent or the postverbal NP referent. Nevertheless, both 5- and 8-year-olds misinterpreted object questions as subject questions, ignoring the number and topicality cues to resolve the (local) ambiguity of which-questions. Our results confirm the effect of a subject-first bias in children’s interpretation of wh-questions. We conclude that topicality, in combination with number agreement, is not strong enough to overrule this subject-first bias.

Speakers of Standard Dutch increasingly realise coda /r/ (e.g. paar) as a bunched or retroflex approximant. This coda variant is categorically distinct from onset /r/ (e.g. reizen), which is typically pronounced as a uvular trill or fricative, or an alveolar trill or tap. In this study, we investigate whether coda and onset /r/ in Dutch behave as distinct phonological categories. We present new articulatory and acoustic data on the realisation of /r/ in sandhi contexts, including the fake geminate context (e.g. paar reizen). Ultrasound data show that the presence of an onset /r/ conditions the deletion of a bunching gesture (typical of the coda /r/ realisation) from /r#r/ sequences. Nevertheless, the presence of a coda /r/ in such sequences is still acoustically traceable. We interpret these findings as a result of shared phonological identity between onset and coda /r/, and discuss them in the context of the place of allophony within models of grammar.

This paper analyses Dutch sentences from a commercial campaign which make use of the ‘X is zo NP/VP’-construction. I consider this to be a construction (cf. Goldberg 1995; Verhagen 2005), not only because of its template-like character in which the adverb zo is followed by an unusual phrase type, but also because of its unexpected semantic properties. The analysis of this construction reveals a part of the Dutch ‘constructicon’.

In this paper we discuss the four tones of Standard (Beijing) Mandarin Chinese. First, we will suggest a proposal for their phonological representation. Then, we discuss the order in which they are acquired in first language acquisition, relating this both to the representations we propose and to the relative frequency with which these tones appear in the most frequently used Chinese words. It turns out that the former predicts the order of acquisition more closely than the latter: We provide an explanation for this based on the early stage at which the tones are acquired.