Lingvisticæ Investigationes - Volume 37, Issue 2, 2014

This paper addresses the status and formation of [non-Adj] sequences in French, such as non occupé ‘non-occupied’, non effaçable ‘non-erasable’. These sequences can be compared to two other types of negative constructions which are built in morphology, namely [non-N]N (e.g. non occupation, non violence) and [in-Adj] (e.g. inoccupé ‘unoccupied’, ineffaçable ‘unerasable’). Taking as a starting point the observation made by Gaatone (1971, 1987) that [non-Adj] have syntactic properties which differ from that of [in-Adj], I use authentic data drawn from different corpora to determine whether [non-Adj] are formed in morphology or in syntax, and I attempt to provide a constructional account (Goldberg, 1995; Croft, 2001; Booij, 2010) of these sequences.

Determining the morphosyntactic status of verbal constructions is an enduring issue in African linguistics. In Wolof, most verbal constructions are split predicate constructions involving a predicative marker, which encodes the greater part of the grammatical content, and a verb, which contributes to the lexical content. The aim of the current paper is to investigate the morphosyntactic status of these complex predicates. Based on several kinds of criteria (phonology, morphology and semantics), I show that the predicative markers must be analysed as phonologically dependent words (clitics). Thus, Wolof complex predicates clearly display syntactic construction features and cannot therefore be considered as morphological units.

The presence of semantically empty morphs (linking elements) in compounds poses a challenge for morpheme-based morphology. The purpose of this paper is to solve the problem using Distributed Morphology, which is a highly articulated version of morpheme-based morphology. Specifically, exploiting the empirical similarities between linking elements and expletives, I argue that linking elements are expletives in word domains. Moreover, I demonstrate that the Single Engine Hypothesis of Distributed Morphology theoretically supports the view of linking elements as expletives. In my analysis, linking elements function as markers of wordhood, just as phrasal expletives are markers of phrasehood. I confirm this view by showing that linking elements/expletives occur in all types of compounds.

Adverbs functioning as degree modifiers are not the only devices used for positive scaling; cross-linguistically, another very common means of intensification is the prefixation. This paper is concerned with data of nine intensifying prefixes of Modern Greek ( arxi‑ ‘archi’, θeο‑ ‘god’, kata‑ ‘over’, pan‑ ‘all’, olo‑ ‘whole’, para‑ ‘over’, pend(a)‑ ‘five’, xilio‑ ‘thousand’, iper‑ ‘hyper’) that constitute a representative sample of Greek intensifying prefixes. The purpose of this paper is to account for prefix-base combinations, taking into consideration the semantic typology of gradable predicates of Kennedy & McNally (2005) and more precisely focusing on the question of how the open or closed scale on which the bases map their arguments relates to intensifying prefix-base combinations. Our aim is to arrive at a general conceptual framework that can analyze all cases of intensifying prefixes. The paper also aims to account for the polysemy found in intensifying prefixes.

In this paper, we present a contrastive survey of a morpheme originally meaning ‘giant’ in German and Swedish. In both languages, this morpheme has developed into a prefixoid with simile or intensifying meaning. More recently, these prefixoids have been shown to occur as free morphemes as well, and it is the purpose of this paper to explore whether a quantitative analysis of synchronic corpus data can be used to determine whether the free forms are spelling variants, or whether they are truly new constructions that are the result of debonding. Drawing data from the COW corpus of contemporary web text, we compare bound and free forms on the levels of R1 collocations, semantic bleaching, and productivity. Our analysis suggests that the German prefixoid has undergone debonding, whereas the Swedish free forms are mere spelling variants.

This paper investigates the role of syntactic, semantic, and lexical factors in the diachronic development of German nominalization patterns. Drawing on an extensive corpus analysis of Early New High German and New High German texts, it is shown that (a) deverbal nominals in the suffix -ung tend to develop more reified meaning variants, which is reflected in the syntactic patterns in which the word-formation products preferentially occur, and (b) infinitival nominalization becomes more productive and is established as the new default word-formation pattern deriving nouns from verbs. These considerations fit in neatly with a cognitively-oriented theory of word-formation change situated in the framework of Construction Grammar.

The morphosyntactic property set associated with the syntactic node occupied by a word form is not invariably identical to the property set determining that word form’s inflection, as evidence from Bhojpuri, Turkish, Sanskrit and Hua shows. The difference between syntactic property sets and their corresponding morphological property sets may be represented as a property mapping relating two different kinds of paradigm: a lexeme L’s content paradigm specifies the range of property sets with which L may be associated in syntax ; its form paradigm specifies the (sometimes distinct) property sets that determine L’s inflectional realization. Thus, a language’s inflectional morphology doesn’t merely specify the realization of paradigm cells: it also specifies the sometimes nontrivial linkage of content with form at the interface of syntax and semantics with morphology.

Agreement minimally involves interaction between morphology and syntax, as a target’s features vary according to the morphological form of a controller in a given syntactic context. However, semantics can also play a role, and the term “semantic agreement” has been used to describe various constructions where morphosyntactic feature values of the agreement target do not match the formal features of the controller, reflecting instead meaning-based properties of the noun. In this paper, we deconstruct instances of “semantic agreement,” as there is good evidence to believe that more than just the semantics is involved in the agreement process. In some cases, e.g. Russian hybrid nouns like vrač ‘doctor’, the local context provides the agreement features, giving a type of “pragmatic agreement”. In other cases, socio-cultural information plays a role, showing a broader type of pragmatic agreement. In light of these observations, we offer a deconstruction of semantic agreement phenomena in order to show the complex ways morphology interacts with syntax, semantics, and pragmatics. Finally, we argue that the distinction between syntactic, semantic, and pragmatic agreement is paralleled by (and benefits from) earlier discussions of syntactic versus pragmatic control.

The present paper explores the processing of morphologically complex words in L2 Italian by means of as series of masked priming experiments associated with a LDT. We manipulated deadjectival nominalizations in -ità (e.g. velocità < veloce) and in -ezza (e.g. bellezza < bello), that differ in terms of numerosity, productivity (Rainer, 2004) and on surface frequency. Morphological priming effects were evaluated relative to both orthographic and identity conditions and the data revealed significant morphological priming effects emerging for words ending with the most productive suffix (-ità) and having a high surface frequency in Italian. Our data on derivation suggest that similarly to native speakers, L2 learners are sensitive to morphological information, but they integrate it progressively through L2 learning process.

Two masked priming experiments with Greek advanced ESL speakers were run in order to reproduce the experiments reported by Silva & Clahsen (2008): our data yielded similar derivational priming but divergent results for inflectional priming. After comparing the two sets of results and examining some discrepancies between the two studies, we provide an interpretation outside the decompositional framework: morphological priming effects are not viewed as low level perceptual saliency effects but rather as the result of the form-meaning systematic relations.