Volume 48, Issue 3

This article contributes to the study of approximative adverbs almost and barely by providing psycholinguistic evidence for the asymmetry of their meaning components. The experiments reported are discussed against the background of a set of tests targeting the theoretical status of the meaning components. The first experiment addresses the role played by each meaning component in textual coherence, whereas the second experiment addresses the interpretation in isolation of a sentence containing an approximative adverb. The results argue for a pragmatic difference in the role of the meaning components, along the lines of Horn's (Assertoric inertia and NPI licensing: 55–82, 2002) proposal, pertaining to the way in which the implications of approximative adverbs contribute to context update.

Exaptation (Lass, Journal of Linguistics 26: 79–102, 1990) is the recycling of meaningless linguistic “junk”, which may be recycled to create entirely new words. Two cases are examined here: – There are two auxiliary verbs do in modern English. The first (Chomskyan do) has the distribution that is famously outlined in Syntactic structures. The second, a homophonous auxiliary, occurs in (negative) imperatives, and was completely indistinguishable from the other periphrastic auxiliary in Elizabethan English. It is subject to very different constraints. The result is that English now has a prohibitive, as well as a declarative negative. – Khmer has an enormous number of alliterating near-synonyms which differ only in one or two segments (the coda, or the entire rhyme) of the stressed final syllable. It is possible that many of these arose as (nearly?) meaningless “servant words” that accompanied an original root in symmetrical compounds. Examples of this sort may call into question two dogmas of functional linguistics: that grammaticalization is irreversible, and the widely shared view that erosion is the only productive process of linguistic change.

In the Germanic and Romance languages (among others) a universal quantifier can combine with a numeral and form a floating quantifier. I refer to these quantifiers as universal numeric quantifiers or simply ∀NumQ. The following examples from Dutch and Romanian demonstrate this phenomenon: Dutch (i) a. Alle drie de studenten hebben het boek gelezen. all three the students have the book read b. De studenten hebben alle drie het boek gelezen. the students have all three the book read Romanian (ii) a. Toţi trei studenţi-i au citit carte-a. all three students the have read book the b. Studenţi-i au citit toţi trei carte-a. students the have read all three book the The aim of this article is to show three things: that a ∀NumQ occupies the same position as a bare universal quantifier, namely, the head position of a Quantifier Phrase; that a ∀NumQ must be base-generated in Q° in its entirety and cannot be derived by moving a numeral from a position inside DP up to Q°; that the derivation or rather creation of a ∀NumQ can easily be understood if one combines aspects of the theory of word formation in Di Sciullo and Williams (On the definition of word, MIT Press, 1987) with the theory of the morphology of numerals in Booij (The construction of Dutch numerals, University of Leiden, 2008).

In this article, I discuss Finnish case alternation adpositions from a theoretical and corpus perspective. First, I argue that postpositional PP constructions with genitive case denote the standard spatial meaning of which an extension is marked with partitive case. Also, I show how word order interacts with case assignment. Both findings are formalized in a bidirectional Optimality Theoretic framework. Second, I show that case alternating behavior does not occur unrestrictedly in newspaper corpora. Adpositions in principle tend to assign the same case to the same object over and over again, and only a small subgroup of highly frequent nouns is assigned both genitive and partitive case by the same adposition(s). This suggests that (adpositional case) alternations are only allowed for highly frequent constructions.

In this article, on the basis of a set of new data that allow us to assess the position of subjects, it is shown that in Japanese, ordinary subjects (which are marked with either nominative or dative case) are moved to Spec of TP, while the subjects of oblique-subject constructions do not undergo subject raising. We argue that subject raising to TP is motivated if the clause is constrained by the nominative-case requirement, which dictates that a clause must have at least one nominative argument. Nevertheless, there are also cases among idioms where nominative subjects remain in vP-internal position; that is, idiom subjects, which are interpreted non-compositionally as part of clausal idioms, do not undergo subject raising. In Japanese, subject raising to TP motivated by the EPP requirement of T is most typically instantiated, owing to the fairly persistent nominative-case requirement, but still, it is not unitarily implemented, since subjects do not undergo raising to TP if they receive oblique marking, which brings out the effect of voiding the nominative-case requirement, or constitute part of idiomatic expressions.

Although contrastive stress in English facilitates context disambiguation, this paper shows that Taiwan Mandarin adult speakers were insensitive to contrastive stress in resolving ambiguity in zhi ‘only’ associated arguments in triadic constructions. The first experiment, which contained sixteen true/false responses to sentences containing zhi associated arguments inferred from corresponding narrated stories, was conducted to test the judgments of one hundred college students. The results showed that the direct object (DO) focus tended to be interpreted as the most prominent in both the dative and double object constructions, contra to the results attested in the dative construction by English-speaking adults in Gualmini et al. (Children’s insensitivity to contrastive stress in sentences with ONLY: University of Pennsylvania, 2003) and Dutch children in Szendrői (Acquisition evidence for an interface theory of focus: 457–468, LOT, 2003), and the prediction of the VP default focus by Reinhart's (Interface strategies: Optimal and costly computations, MIT Press, 2006) focus interface strategy. The second experiment with a multiple-choice questionnaire of the same sixteen contexts as in the first experiment was designed for a different group of fifty-one participants to identify which element (indirect object (IO), DO or VP focus) was interpreted most prominently. The results again showed the correction rates of the DO focus were higher than those of the IO, and those of the default VP focus was the lowest, contra to Reinhart's prediction. The conspicuous nature of the DO focus in triadic constructions in both experiments is accounted for by its syntactic basicness to the predicate and its preferred thematic prominence over the IO. The DO focus surfaces in Mandarin in which stress is not lexically distinctive; hence prosody is not served as the primary cue for disambiguation.

Much recent work has argued that the major lexical categories can be distinguished in terms of pragmatic functions (e.g., Baker, Lexical categories: Verbs, nouns and adjectives, CambridgeUniversity Press, 2003; Bhat, The adjectival category, John Benjamins, 1994; Croft, Radical construction grammar: Syntactic theory in typological perspective, Oxford University Press, 2001; Hengeveld, Non-verbal predication: Theory, typology, diachrony, Mouton de Gruyter, 1992). Typically, such pragmatic accounts argue that nouns distinguish themselves by referring, verbs distinguish themselves by predicating and adjectives distinguish themselves by modifying. The current article argues that such accounts are prone to two distinct sets of problems. The first set of problems arise from the definitions of the pragmatic functions that are employed in these accounts. Thus, the definitions of predication and modification that feature in such accounts are typically so similar they render attempts to distinguish verbs and adjectives in terms of them vacuous. Moreover, the definitions of all three pragmatic functions are often so vague and general that they apply with equal ease to words of all three major lexical categories. When more specific definitions are given, however, they typically exclude words from their intended category while continuing to include words from other categories. The second set of problems arise from the lack of direct evidence for the pragmatic functions. Such an absence of evidence gives rise to disputes over issues as basic as whether a given lexical category performs a given pragmatic function or not, whether there are two, three or more pragmatic functions and whether pragmatic functions are performed by words, phrases or different units altogether. It is argued that in the absence of direct evidence such basic disputes cannot be satisfactorily resolved. It is concluded that these problems are as serious as those which afflict semantic and morphosyntactic approaches to lexical categories and thus that the prospects for a coherent explanation of lexical categories remain as remote as ever.

While agreeing with Smith that the lexical categories cannot be defined pragmatically, I disagree with his implication that Baker 2003 is a pragmatic theory. Rather, I attempt to clarify my view that the definitional differences between nouns, verbs, and adjectives are primarily syntactic in nature. These syntactic differences stand in a natural correspondence to the pragmatic differences discussed in the literature in many cases, but when syntax and pragmatics go separate ways, category distinctions clearly follow the syntax. Given this, it stands to reason that the accuracy of the definitions can only be assessed if the syntactic structure of relevant examples has been accurately determined. I discuss a few suggestive examples of how this might be done in some “contentious” cases.

Smith offers a critique of the theory of parts of speech in Croft (Syntactic categories and grammatical relations: The cognitive organization of information, University of Chicago Press, 1991, Radical construction grammar: Syntactic theory in typological perspective, Oxford University Press, 2001) inter alia. Smith tries to make a functionally-based universal-typological theory of parts of speech provide an answer to the problem of defining word classes and giving those classes the same names across languages (“noun”; “adjective”); this is not possible and not what I intended. Smith conflates semantic properties with pragmatic properties, and he conflates different pragmatic properties that cannot be conflated. There are challenging issues in defining pragmatic functions and their linguistic reflexes, but Smith's critique only briefly touches on them.