The goal of this paper is to investigate the structural distinctions that exist between prefixoids and prefixes in modern Dutch, with an empirical focus on inseparable complex verbs (ICVs). Adopting a derivational, separationist approach to the syntax-morphology interface (Distributed Morphology), we maintain that the fundamental difference between prefixoids and prefixes is that the former are clitics while the latter are initially merged with √(roots) lower in the first phase (vP) prior to when these complex predicates are labeled (by v). As clitics, prefixoids are adjoined to the left-edge of the first phase (with the phase head v). Our theoretical claims are further substantiated by data extracted from the Dutch Celex Lexicon, providing a detailed assessment of the categorical status of the stem that prefixoids and prefixes attach to, their ability to contribute a resultative interpretation, and their potential effects on the argument structure of the ICV.
In recent decades, several Germanic languages have seen the rise of new social dialects in multilingual urban areas. Syntactically, these contemporary urban vernaculars (CUVs) are characterised by allowing a new set of exceptions to the verb second (V2) pattern, as a clause initial constituent may be followed by subject-verb and not verb-subject word order. Here we present an empirical investigation, based on ethnographic material, of word order variation of V2 and V3 in CUV Swedish, showing that V3, compared to non-subject initial V2, is prevalent and clearly the preferred word order in spontaneous speech. We also provide a new formal account of the C-domain of CUV Swedish with obvious relevance for other Germanic varieties, CUVs and non-CUVs. The structural difference between V2 and V3 does not depend on different positions of the finite verb, nor on a different number of projections in the C-domain. Instead, V3 varies with V2 if not just one but two constituents may move across the Fin head from within TP. This proposal is supported by the finding that arguments, and not just adjuncts, may occupy the initial position of V3 sentences.
Drawing on novel data collected from 216 speakers (108 speakers of Icelandic and Faroese, respectively), we provide a micro comparative examination of the distribution and interpretation of double object (DO) and prepositional (PP) ditransitives in Insular Scandinavian. Overall, PPs are more acceptable in both Icelandic and Faroese than has been previously reported, with a substantial amount of variation in Icelandic. Even though verbal semantics interacts with PPs more so in Icelandic than in Faroese, different verbs in the same semantic class pattern differently in both languages, complexifying claims that have been made about the semantic restrictions on PPs (Þráinsson, Höskuldur. Syntax of Icelandic. New York: Cambridge University Press, 2007 and Þráinsson, Höskuldur et al. Faroese: An Overview and Reference Grammar. Føroya Frόðskaparfelag, Tόrshavn, 2004). Furthermore, our findings suggest that ditransitive verbs in Insular Scandinavian do not entail successful transfer of possession, irrespective of verb class. This is counter to claims that have been made for English, for which ‘give’-type verbs are argued to entail successful transfer (e.g., Rappaport Hovav and Levin in J Linguist 44(1):129–167, 2008. https://doi.org/10.1017/S0022226707004975 ; Beavers in J Semant 28(1):1–54, 2011. https://doi.org/10.1093/jos/ffq014 ; Bruening in Nat Lang Linguist Theory 39:1023–1085, 2021).
In this paper we experimentally investigate (i) how native speakers of English interpret negative quantifiers (e.g. nobody, nothing) when used as fragment answers to negative wh-questions in the absence of a biasing context, and (ii) whether their preferences correlate to their interpretation of full sentences with a negative quantifier co-occurring with a negative marker. Despite an overall preference for double negation interpretation, in which each syntactic negation contributes independently to the semantics, our results show that speakers also allow single negation to different extents both for fragment answers and for full clauses. Results also revealed a correlation between participants’ interpretation. If they interpreted fragment answers as single negation, then they also tended to interpret full clauses as single negation, and the same was true for double negation interpretations. We account for these findings by postulating the existence of two different lexical variants for English negative indefinites (a negative quantifier one, ¬∃, and a Negative Concord Item one, ∃[neg]) that can explain why both a double negation and a single negation reading are possible for our participants when interpreting full sentences. For fragments, we show that both double negation and single negation readings can obtain with either of the two lexical variants for nothing, nobody and the like in approaches to ellipsis with different degrees of strictness on syntactic identity of the fragment with the antecedent.
This paper presents a theoretical approach to adjunct control in German and English participial clauses. Based on a corpus study of 1600 non-finite adverbial clauses with a participial head in German and English, we argue that there is a clear correlation between the syntactico-semantic scope of a participial adjunct and its control status: while adjuncts modifying the event generally display obligatory control (OC), adjuncts modifying the matrix proposition or utterance occur with non-obligatory control (NOC). To capture this theoretically, we analyse OC as an upward Agree relation between a matrix argument and PRO, as also suggested in other recent approaches to control. Since this Agree relation can only be established without further ado when the adjunct is in the c-command domain of a matrix argument, OC is restricted to adjuncts in the verbal domain. For adjuncts above T, an NOC relation is established, which is licensed by a salient perceiver or referent that is syntactically encoded in the C-domain. Thus, we argue that the syntactic height of an adjunct is crucial for its control status and that OC and NOC are generally in complementary distribution in adjuncts. However, in contrast to previous work, we suggest that the control relation is established phase-wise, which leaves room for a certain amount of flexibility if the first potential goal is not a suitable controller.
A growing body of research on phonological features claims that they are not universal elements of language. Rather, it argues that they are emergent, meaning that they are encoded based on phonetic cues as well as sociohistorical factors. In emergent feature theory, features are probabilistically encoded. The clearest phonetic cues result in the clearest featural encoding. But phonetic cues can be blurrier for certain segments and in certain contexts. In these situations, multiple feature encodings are possible. Little research has explored the range of emergent specifications that are possible. This paper argues that coda nasals in early Germanic languages were marked with the feature [low], while onsets were not. Due to the lowness specification of coda nasals, the blocking of Northwest Germanic a-umlaut, Proto-Germanic pre-nasal raising, and high vowel lowering conditioned by coda nasals and post-velar fricatives in Old Norse can be accounted for in a novel and straightforward manner.
In this paper, we draw on corpus data to show that in Icelandic, verbs that assign distinct cases can be coordinated and share a single object: the verb on the right determines the case that the object bears. However, it turns out that fine-grained details of how case is realized on the object morphologically have an effect on which verb is more likely to come first: the one that assigns accusative or the one that assigns dative. If the object is syncretic for accusative and dative, then there is no preference, and both word orders are equally frequent. If the object is not syncretic, then there is a preference to put the dative-assigning verb last. But this preference is not equal for all objects: when the accusative is realized by a zero affix, the preference is weaker than when the accusative is realized by a non-zero affix. We present an analysis of these facts that is grounded in formal spellout mechanisms and the following two guiding principles: choose the order that expresses the most case features, and choose the order that uses the fewest mechanisms. Non-syncretic objects vary in the strength of the preference in a way that can be connected to these two principles, as long as we make certain specific assumptions about how spellout works, most importantly that zero affixes result from the absence of Vocabulary Insertion, rather than the insertion of a phonologically empty symbol, and Impoverishment exists as a spellout mechanism, one that is distinct from Vocabulary Insertion.
In determiner sharing, a quantifier may be omitted from a coordination in the context of another ellipsis. This paper proposes a novel analysis on the basis of new German data: determiner sharing arises from the interaction of clausal ellipsis and split topicalization. I show that the apparent parasitism of determiner sharing can be derived without any further assumptions. The success of this analysis supports Move-and-Delete approaches to ellipsis.
Heterofunctional Coordination (HC), in which conjuncts bear different grammatical functions (as in English What and when to eat to stay healthy), is assumed to be solely multiclausal in Germanic languages, i.e., to be underlyingly a coordination of clauses. This is supposed to distinguish Germanic from Slavic, where monoclausal HC is also possible, in which the surface conjuncts are coordinated directly. In the case of German, this assumption has not been supported by any empirical studies. This paper offers two such studies—based on corpora and on acceptability judgement experiments—which, however, do not confirm the assumption that German HC is strictly multiclausal. In particular, numerous examples of monoclausal HC constructions may be found in German corpora, while judgement experiments show a great variability of acceptance rates of monoclausal HC in German and demonstrate that the acceptability of such constructions depends on various factors. As this variability does not seem to reflect processing effects, we conclude that a gradient (non-binary) grammaticality approach is needed to model German HC. While the focus of this paper is on deriving the right empirical generalizations, we also include an appendix containing a proof-of-concept sketch of such a gradient grammaticality analysis, which builds on Minimalist Gradient Harmonic Grammar and on ideas from Linear Optimality Theory and the Decathlon Model.
DP-conjunctions with a mismatch in person features call for additional resolution rules to determine the values the agreement target has to copy. Across languages, resolution for person features typically follows a hierarchy of the form 1 ≻ 2 ≻ 3 —with some well-known exceptions, namely German and Dutch coordinations conjoining 2nd and 3rd person which allow for both agreement options. This paper takes a closer look at resolution agreement in German, Dutch, Norwegian, Swedish, and Icelandic. The German(ic) anomaly provides evidence for the presence of an underlying binary feature system and the need for set union as a resolution mechanism. The pattern is derived within the framework of Distributed Morphology where vocabulary insertion happens late and is thereby sensitive to decomposed and unified feature sets. Crucial for the account is an independently motivated impoverishment rule that tracks the absence of 1st person inclusive exponents in Germanic.
A generalization that has emerged in the literature on verb clusters in West Germanic languages is that so-called 2-1-3 orders (where verb n selects verb n+1) are absent with core clustering verbs. In this paper we show that Dutch permits 2-1-3 order as an optional variant of the more widely acknowledged 1-2-3 order. The 2-1-3 order is subject to a range of restrictions involving the lexical items in the cluster, the morphological form of these items, whether other elements intervene, and where focus is placed. We argue that these restrictions are best understood if the 2-1-3 order is derived from an underlying 1-2-3 cluster through a post-syntactic inversion rule. This rule shares various properties with other inversion rules but cannot be reduced entirely to a familiar rule type.
This paper argues in favor of the hypothesis that there are two definite articles in natural language, a weak article which expresses uniqueness and a strong article which expresses anaphoricity. The study is based on the distribution of definite articles in Icelandic, and they are found to alternate along the same empirical dimension as weak and strong articles in German. Furthermore, the Icelandic pattern manifests a structural interaction which is similar to English do-support, and we refer to it as the-support. We argue that the similarities between do-support and the-support suggest that studies of the two phenomena can benefit from considering them to be related at a deep abstract level. Thus, consequences ensue for the theory of suffixation under adjecency.
Long extraction, that is, the displacement of a constituent across a clause boundary, is considered a process of broad applicability. This view is challenged by the claim that extracting a phrase from a that-clause into a relative clause is ungrammatical in German. Since the evidence for this claim is extremely limited, we ran three acceptability experiments investigating long extraction in German. As expected from a large range of studies on long extraction in German, long extraction was judged as less acceptable than corresponding sentences without long extraction. Importantly, long extraction was equally acceptable across the three contexts that were tested—relative clauses, embedded questions and main clause questions. Our experiments, thus, show that long extraction applies across different structural contexts in German, as expected if long extraction is a general syntactic process. In addition, this paper presents new evidence concerning the sources of individual variation with regard to the acceptability of long extraction. First, we confirm that long extraction gets less acceptable when going from the South to the North of the German speaking area. Second, we tested whether individual participants differ with regard to how easily they accept non-standard constructions. To this end, we ran an additional experiment on verb-cluster formation, including sentences that are ungrammatical according to prescriptive grammar but that are, nevertheless, accepted by many speakers of German. The acceptability of long extraction correlates with the acceptability of non-standard verb clusters even when regional background is controlled for.
This paper presents a cross-dialectal study of grammatical gender in Norwegian nominal phrases. Specifically, we investigate the decline of the feminine gender in three age groups across seven different dialects. The dialects vary in their morphological richness of gender marking: some dialects traditionally have more distinctive marking of the feminine gender. With an elicited production experiment, we investigate gender marking on the indefinite determiner and the definite suffix. We find that feminine gender is in decline in all dialects, but there are clear differences between the locations and between age groups. The feminine indefinite determiner ei is replaced by the masculine en at different rates and to a different degree in the various dialects. We furthermore find that the feminine definite suffix -a is retained in all locations except for Stavanger. We argue that the decline of the feminine gender can be explained by an interplay between the morphological richness of the given dialect and dialect contact. The former helps to retain the feminine as a separate category, while the latter accelerates the loss of the feminine.
This paper investigates the properties of nominal phrases and demonstratives used as verbal anaphora in Norwegian, Danish, English, and Scots-English, e.g. English Can John make good curry? – That he can; Norwegian Anja ligger godt an, det samme gjør Madelène lit. ‘Anja is in a good position, Madelène does the same [thing]’. Following Lødrup (Proceedings of NELS 24, 1994), Houser et al. (Proceedings of WECOL 34, 2007), Bentzen et al. (J Comp Ger Linguist 16:91–125, 2013), these anaphoric expressions are argued to be surface anaphora and to conceal elided vPs. Contrary to previous analyses, the nominal phrases are argued to themselves be contributing meaning beyond the vPs they conceal; they are argued to be overt background arguments for an ellipsis-licensing head with semantics similar to Rooth’s ∼ operator (Nat Lang Semant 1(1):75–116, 1992). The paper also explores cross-linguistic variation in the discourse/antecedence conditions on such anaphora, and their fronting behavior. In Danish and (general) English, such anaphora must generally topicalize, whereas in Norwegian and Scots-English, they can more freely appear in situ (in post-auxiliary position). Developing Mikkelsen’s (J Linguist 51(3):595–643, 2015) analysis of Danish det, this behavior is encoded as a feature [uTop] which must be checked; Norwegian is argued to have more possibilities to check this feature in situ than Danish, while in Scots-English, that is argued to be a propositional anaphor, lacking the relevant feature.
In Alemannic dialects of German, [n] is particularly vulnerable to assimilation, deletion, and epenthesis. Although these changes are not necessarily uniform across all Alemannic varieties, the Alemannic dialect areas all exhibit some, if not all, of these processes. In this article, we present data from a diverse array of Alemannic dialects and show that [n] behaves similarly throughout Alemannic, assimilating to the place of following stops, deleting word-finally, and repairing hiatus through epenthesis. We contend that coronal [n] is interacting with so many processes because it is unmarked in terms of place and manner. This paper contributes to the phonological literature on dialectology and Markedness Theory. First, by considering similar processes which occur across multiple Alemannic dialects, we show how Alemannic prefers eliminating or modifying word-final [n]. Second, this analysis gives insight into theories of segment (un)markedness; thus, the data presented in this paper support descriptions of unmarked segments as undergoing assimilation, deletion, and epenthesis, while they challenge markedness accounts by scholars who bar [n] as an epenthetic segment. Third, we provide data for a language family in which one segment undergoes all three processes of assimilation, deletion, and epenthesis; this is unprecedented in the literature on unmarked segments, which typically focuses on languages which possess only one of these three processes.
Gothic preverb compounds illustrate several interesting characteristics, including multiple preverb stacking, idiomatisation, tmesis (i.e., separation by clitics), and P-copying (i.e., multiple pronunciation of the preverb). This paper is a close examination of the morphosyntax of these compounds, highlighting novel empirical generalisations about the Gothic language with key theoretical implications for our understanding of Germanic complex verbs and the alternations they participate in. In particular, this paper proposes a structural distinction between preverb compounds which are obligatorily semantically transparent and those which are optionally idiomatic. In arguing that transparent compounds involve the mechanism of preposition incorporation and m-merger, paralleling recent accounts of clitic doubling, while idiomatic compounds involve a thematic high applicative projection, this paper captures nuanced differences in these compounds’ case assignment and argument licensing behaviour. These structural differences will be shown to derive these two compound types’ constrained interaction with the aforementioned phenomena of stacking, tmesis, and copying. In addition, this paper compares Gothic complex verbs to their cross-linguistic correlates within and beyond Germanic, whilst also providing a diachronic pathway for the development of (multiple) preverb compounds.
The present paper investigates participial root configurations, i.e. participial clauses that are grammatically independent of a host clause. Unlike previous work, which has focussed on either directive or (non-directive) performative uses of so-called past participles (i.e. participles that have passive and/or perfect(ive) interpretations), the present paper establishes a typology of ‘root participles’ in Germanic and contrasts the properties of four main types: (1) directive (RPdir), (2) expressive (RPexp), (3) commissive (RPcom), (4) representative root participles (RPrep). The main claim with respect to the properties of these distinct types is that they differ in terms of whether they include a verbal or an adjectival (passive) participle. In fact, arguments based on argument structure, orientation, aspect, and adverbial modification are presented to substantiate the claim that types (1) and (2) are formed with verbal and types (3) and (4) with adjectival participles. Additionally, the distinct types will be shown to differ in their status of either being non-sentential (i.e. structurally different from potential clausal counterparts) or merely elliptical (just phonologically reduced): types (1) and (3) can be shown to be non-sentential and hence receive a dedicated syntactic analysis, where special attention is paid to the contribution of the (imperative vs. declarative) left periphery.
Possessive alternates (prenominal and postnominal) have mirrored properties in Italian and Norwegian when taking into consideration frequency, derivation, and markedness; i.e., the variant that is base-generated in one language is considered the derived one in the other language. Thus, in both languages there is a variant used for unmarked contexts (i.e., topic) and for marked contexts (i.e., contrast). Previous studies have shown that Italian children acquire the use of the variants with ease, whereas Norwegian children were found to overuse the marked variant, even in unmarked contexts. Here, we reanalyse the co-occurrences of the possessive and the noun in the monolingual corpora for the two languages available on CHILDES, by focusing more attentively on the contextual use of the variants, to reveal whether the same principles underly the acquisition process. Our findings contradict the previous claims on the acquisition of Italian but are in line with the previous findings for Norwegian. Both groups of children overuse the marked but base-generated variant, indicating the relevance of syntactic economy in language acquisition.
The present paper proposes an alternative analysis of so-called expletive es in German. It is argued that es has semantic content that serves to anchor the utterance in the context. In particular, I argue that es constitutes a weak demonstrative element binding a situation argument. The account gets rid of the assumption that the relevant head in the clause is endowed with an EPP-feature and restores the original principle underlying it, namely the requirement that every predicate needs a contentful subject argument it can be predicated of. The account also explains in more depth the obligatory and optional occurrences of es and proposes that there are essentially two occurrences of es to distinguish in terms of their syntactic properties.
