ComSyn Lectures

Speakers: Irina Morozova & Sjef Barbiers (LUCL)
Title: Why does water have a blue colour? Modification of Nouns and the Possession of Properties
When: Thu 17 April, 16:15 – 17:30
Where: Lipsius 2.23
Zoom: link / Meeting ID: 633 9946 0316 / Passcode: AgNKj$2d

There are at least three ways in Dutch to say that a car is green:

(1)  

Shape and size adjectives also have the paradigm in (1), material adjectives do not (*een houten materiaal tafel ‘a wooden material table’). Exactly the same pattern holds for Russian. We argue against treating elements like kleur ‘colour’ in (1b) as the counterpart of a silent noun in (1a) (Kayne 2005) and propose an analysis in which (1b) is not derivationally related to (1a), but is derived from (1c) instead. We argue that in this case the relation between an adjective and a noun is mediated by a dimension classifier (DimCL), realised as kleur ‘colour’ in (1b,c). The nominal nature of DimCl leads to establishing a possessive relation between the noun and the DimCl, which is realised as the preposition van ‘of’ in Dutch (1c) or as genitive case in the Russian equivalent of (1c). As opposed to this, material adjectives (e.g., hout-en ‘wood-en’) do not involve possession, hence the patterns in (1b) and (1c) are impossible. Surprisingly, however, the pattern in (1b) also does not hold when the modifier targets a mass noun as in blauwe (*kleur) water ‘blue water’. We take this as a key observation and build our analysis on this.

Speaker: Melle Groen (University of Tübingen)
Title: Hypothetical comparison clauses in Dutch dialects
When: Thu 20 March, 16:15 – 17:30
Where: Lipsius 2.23
Zoom: link / Meeting ID: 651 1518 0819 / Passcode: &YaUi^H9

Hypothetical comparison clauses (HCCs; e.g. ‘she acted as if she were ill’) show considerable morphosyntactic variation, both intra- and cross-linguistically. In some languages, they are transparently composed of an equative particle and a conditional clause (e.g. English as if, French comme si, etc.). In Dutch however, HCCs are not compositional in this way: they are most commonly introduced by alsof, which is cognate to English ‘as if’, but of ‘if, whether’ is not used as a conditional marker in contemporary Dutch. Furthermore, Dutch dialects exhibit a wide range of variation with regard to the complementizer domain of HCCs (see De Rooij 1965): this variation includes various combinations of functional elements, like als ‘as’, of ‘if’, dat ‘that’, and wen(t) ‘when’; verb-first order in the embedded clause, as well as a lexical element (ge)lijk ‘equal’.
 
In this talk, I present an overview of the morphosyntactic variation in HCCs among the Dutch dialects, and argue that the restrictions on this variation point to one common underlying structure. I will also argue that the data on HCCs have implications for the analysis of the Dutch complementizer domain. Specifically, they imply that the Dutch C-domain consists of two CP layers, rather than three (as proposed by Hoekstra 1993, Zwart 2000), and that the complementizer of ‘if, whether’ should not be uniformly regarded as an interrogative head above the general C/Fin head dat ‘that’, but should instead receive a different analysis in the northern and southern dialects, respectively.

Speaker: Miriam Schiele
Title: Preposition Omission and Focus in German Fragments: A Case for a Q-Based Approach
When: Thu 6 March, 16:15 – 17:30
Where: Lipsius 1.23
Zoom: link / Meeting ID: 640 7326 7318 / Passcode: Mq.cK08f

Preposition omission such as in Peter was talking to someone but I don’t know (with) who is allowed in English but not German. Merchant (2001) attributes this to syntactic factors, arguing that English allows preposition stranding under regular syntactic movement but German does not. In contrast, Nykiel & Hawkins (2020) propose a processing-based explanation, suggesting German’s complex case-marking system necessitates the presence of the case-assigning preposition in elliptical clauses to ensure proper processing of the case-marked DP. This talk presents new data on German fragments, examining preposition omission, case marking and focus through an acceptability judgment task. Not only do the findings challenge both Merchant (2001) and Nykiel & Hawkins (2020), but they provide strong support for the Q-based approach to clausal ellipsis.

We are happy to announce the ComSyn programme for the spring semester. All talks will take place on Thursdays at usual time, 16:15-17:30 CET. We will continue to stream offline talks via Zoom. For online talks you are also welcome to join us in person. Please note that the talks on 6th Match and 17th April have been switched around. The talk on 6th March will be an online talk.

We are looking forward to many interesting discussions, so save the dates!

Speaker: Thomas Grano (Indiana University)
Title: Size/flavor correlations in the grammar of attitudes and modals
When: Thu 5 Dec, 16:15–17:30
Where: Lipsius 2.08
Zoom: Link / Meeting ID: 685 9579 1979 / Passcode: S*ZuXg0^

Cross-linguistically, complements to ‘think’/‘believe’ verbs tend to resemble declarative main clauses whereas complements to ‘want’ tend to be structurally smaller. Mandarin Chinese illustrates this trend in a particularly striking way because it has a mental attitude verb 想 xiǎng that can be used to report both beliefs and desires, with a syntactic restriction: when the verb combines with a full CP, it reports a belief, whereas when it combines with a vP, it reports a desire. In this talk, I propose to relate this phenomenon to an independent but similar cross-linguistic fact about modality: namely, epistemic modals combine with tensed complements whereas root modals combine with structurally smaller complements that do not include tense. Toward a unified account of the attitude facts and the modality facts, I propose that epistemic modals and belief predicates operate on world descriptions (i.e., propositions), which are encoded by full CPs, whereas root modals and desire predicates operate on temporally truncated, future-oriented situation descriptions, encoded by smaller pieces of structure. The upshot is that at least some of the grammar of attitude reports falls out from a more general grammar of modality, broadly construed.

Speaker: Richard S. Kayne (NYU)
Title: The DP-Internal Origin of Datives
When: Thu 21 Nov, 15.45-17.00 (please note the time change!)
Where: Lipsius 1.33
Zoom: Link / Meeting ID: 629 5474 2232 / Passcode: u^7gBg9!

Bringing together Szabolcsi’s work on Hungarian possessive sentences, the possessor-raising tradition having to do with inalienables, locative datives, temporal datives, datives with modals, datives with psych-verbs and dative subjects in French causatives, I will suggest that dative arguments invariably originate DP-internally.

Speaker: Jenneke van der Wal (LUCL)
Title: Preverbal focus in Kîîtharaka revisited
When: Thu 31 Oct, 16.15-17.30
Where: Lipsius 0.01
Zoom: Link / Meeting ID: 651 1843 4869 / Passcode: 3wF.zpnn

Term focus in Kîîtharaka (Bantu E54, Kenya) can be expressed by marking the focused constituent by ni and placing it in clause-initial position, as illustrated in (1) for the question word ûû ‘who’.

(1) N’ûû agûpéere rûûtha rwa gûtaa rûûyî?
ni           ûû       a-kû-pa-îre             rûûtha              rû-a        kû-ta-a   rû-ûyî
foc/cop  1.who  1sm-prs-give-pfv  11.permission  11-conn  15-fetch  11-water?
‘Who gave you permission to fetch water?’

This construction has been analysed in two ways: Harford (1997) proposes an analysis as a biclausal cleft, while Muriungi (2005) and Abels and Muriungi (2008) argue that the structure is a monoclausal focus construction and propose a Focus projection in the left periphery.

In this talk, I systematically check the properties of the copula/focus marker, the relative marking (which turns out to also be tonal!), the scope interpretations, and movement diagnostics, showing that these new data force us to reconsider the monoclausal analysis. The alternative analysis as a reduced cleft needs discussion in this seminar.

This is joint work with Patrick Kanampiu.

Speaker: Fábio Bonfim Duarte (Federal University of Minas Gerais)
Title: Head-finality, Predicate Fronting and Spell-out Domains
When: Thu 17 Oct, 16:15–17:30
Where: Lipsius 0.01
Zoom: Link / Meeting ID: 666 5595 3390 / Passcode: q@X8&R2Y

Tenetehára root clauses exhibit a complex subset of sentence-final particles that encode aspect, evidentiality, agreement and tense. The purpose of this talk is to argue that they are syntactic heads sitting in different positions in the inflectional spine of the sentence. As such, we will posit that the head I^o is not expressed by means of a syncretic head, but by an inflectional complex comprised of several head-final functional categories. The main evidence in favor of this hypothesis concerns the fact that these various functional particles do not occupy a uniform syntactic slot. These final particles systematically occur after the predicate, thereby producing the head-final structure [[SVO]-Infl]-Infl]-Infl]-Infl]. Based on these empirical facts, the central hypothesis I will develop in this talk is that this head-final order is the result of the application of successive roll-up movements of the complement of the relevant Infl head through its specifier position. This syntactic derivation corresponds to what Travis (2000, 2005) refers to as an intraposition operation, in which several applications of complement raising proceed in a bottom-up fashion. This hypothesis conforms to our analysis that vP is not a Spell-Out domain, since vP must participate in further syntactic operations in C/T/IP phase. According to this theory, the Spell-Out of the vP is delayed until the C/T/IP phase is merged in the derivation, making the predicate fronting possible. Pursuing this line of reasoning, we will also assume that it is the presence of the unvalued [uPRED] feature on the Infl heads that force the iterative movements of the complement of the Infl heads through their specifier positions. I will then consider this intraposition operation as a typical case of externally-driven movement forced by some unvalued features on some higher Probe in the sense that it is not motivated by needs of the complement of the Infl heads. Viewed in this way, I will claim that the heads in the Infl complex carry an unvalued [uPRED] feature, which is valued by the interpretable [iPRED]-feature of the vP projection. This feature is then deleted under the AGREE operation between the Infl head and the complement that moves through its specifier. Thus, the complement of the Infl head always carries the [iPRED]-feature needed to erase and value the [uPRED]-feature of the relevant Infl head in the course of the syntactic operation. This syntactic operation explains why adjunction of the verb to a functional head in the C/IP region is entirely disallowed in Tenetehára.

Please note the following change in the ComSyn program.

It has come to our attention that the talk by Thomas Grano on Oct. 10 coincides with a lecture in the LUCL Colloqium series. This talk is therefore rescheduled to Thu Dec. 5, in the usual timeslot, Lipsius 2.08 (where we’ll broadcast the Zoom talk).

Here’s the updated program with the four remaining talks:

Speaker: Gert-Jan Schoenmakers (Utrecht University)
Title: An experimental investigation of syntactic and discourse-processing claims about filler-gap dependencies: Adjunct islands and parasitic gaps
When: Thu 19 Sep, 16:15–17:30
Where: Lipsius 0.01
Zoom: Link / Meeting ID: 682 3942 3068 / Passcode: ttW+g7aq

Explanations for linguistic phenomena are often sought either in the syntax or in non-syntactic components of human cognition, e.g. through demands of discourse processing. The unacceptability of filler-gap dependencies in island and parasitic gap constructions is a notable example: while syntactic theory claims that grammatical constraints impede long-distance dependency formation, other researchers claim that (discourse-inaccessible) referring arguments contribute to processing complexity to the extent that acceptability is significantly reduced. This talk presents new data to empirically evaluate such claims in Dutch adjunct island and parasitic gap constructions, evaluating Culicover and Winkler’s (2022) recent Uninvited Guest Hypothesis in particular. Furthermore, we reflect on our findings from the perspective of syntactic theory as well as other discourse-processing accounts of filler-gap dependencies.