A key debate in loanword adaptation is whether the process is primarily phonetic or phonological. Is it possible that researchers on each side are viewing equally plausible, but different, scenarios? Perhaps, in some language situations, adaptation is carried out mainly by those without access to L2 phonology and is, perforce, perceptually driven. In other situations, adaptation may be done by bilinguals who actively draw upon their knowledge of L2 phonology in adapting loanwords. The phonetic strategy would most likely be favored in situations where the vast majority of the population did not know the L2, thus having no possible access to the L2 phonological system. The phonological strategy, on the other hand, is most likely to be favored in situations where there is a high proportion of speakers who are bilingual in the L1 and L2. This possibility is tested by comparing the adaptations of English loanwords in 19th- and early 20th-century Quebec French, when bilinguals were few, to those of contemporary Quebec French, in which the rate of bilingualism is far higher. The results show that even when the proportion of bilinguals in a society is relatively small, they determine how loanwords are pronounced in the borrowing language. Bilinguals adapt loanwords on the basis of phonology, not of faulty perception of foreign sounds and structures. However, in a society where bilinguals are few, there is a slight increase in non-phonological influences in loanword adaptation. We address the small role played by non-phonological factors, including phonetic approximation, orthography, and analogy (true or false), showing that false analogy, in particular, may give the impression that phonetic approximation is more widespread in a loanword corpus than is actually the case.

In 1971–72 a lengthy discussion took place in Language concerning sentence stress (alias nuclear stress, alias intonation nucleus placement) and its relationship to syntactic structure (Bresnan, 1971, 1972; Lakoff, 1972; Berman & Szamosi, 1972; Bolinger, 1972). Bolinger had the last word in stating that placement of the nuclear accent ‘is not explainable by syntax or morphology’ but ‘is a matter of information, not of structure’ (Bolinger, 1972: 633). According to this view words with low informational content (i.e. highly predictable) in a particular context or situation are unlikely to receive a nuclear stress, whereas words with high information content (i.e. highly unpredictable) are likely to receive a nuclear stress. In addition, ‘the speaker will put the main accent as far to the right as he dares, when assertive pressure is high’ (Bolinger, 1972: 644).

Culicover & Jackendoff (1997) argue that ‘left-subordinating’ and-constructions (e.g. You drink one more can of beer and I'm leaving) should be differently represented in the dimensions of syntax and semantics, being coordinate in the former, and subordinate in the latter. Here we expand on their point by showing that their case is not an isolated one, but that there are many other instances of coordination-subordination mismatches. We will show that these facts make sense within a theory of grammar such as Autolexical Grammar (Sadock 1991) in which the autonomy of different components of grammar is assumed. Given such a view it is possible to postulate primitive notions of coordination and subordination that apply equally well to various components of grammar and thus predict the possibility of coordination-subordination mismatches.

1. Efforts to limit the power of transformations, and thereby reduce the class of possible grammars, have been concentrated in two areas: the conditions which restrict the application of the transformational rules, and the examination of the consequences of placing the burden of description on the non-transformational components of the grammar. Recent work on the non-transformational components has led to some interesting conclusions about how this power can be reduced. An article by Bresnan (1978) is a good example of this line of research. She cites a number of cases where transferring the burden of description from the transformational component to the lexicon and semantic interpretive components not only leads to a simplification in the from of the transformational rules, but also captures generalizations not adequately explained by transformations.

The classification of that is, or (in other words), and or rather as reformulation markers would seem to suggest that the utterances they introduce achieve relevance in the same way. However, the examination of a range of discrepancies between reformulations introduced by or, on the one hand, and reformulations introduced by that is, on the other, suggests that any account of the pragmatics of reformulation must be a non-unitary one. In this paper, I build on Burton-Roberts' (1993) suggestion that the reformulations introduced by or are meta-linguistic in character, and show how these can be distinguished from the reformulations introduced by that is, which must be analysed at the level of conceptual representation. I also show how this distinction corresponds to a distinction between the different ways in which a parenthetical construction may be pragmatically integrated with its host. As Potts (2005) would predict, parenthetical that is-reformulations are not themselves part of the truth-conditional content of their hosts. However, in contrast with or-reformulations, they communicate information about the propositional content of their hosts, and in this way can contribute to the identification of their truth-conditional content at the level of pragmatic interpretation.

To avoid the triviality of mere taxonomy in which languages are classified by such arbitrary criteria as ‘is/is not tonal’, ‘does/does not have unbounded movement rules’, ‘has prefixes/suffixes’, etc., language typology needs to be able to make implicational statements of the sort usually associated with the name of Greenberg.1 For instance, one wants to be able to say of a language not only that it is VSO, but also that it THEREFORE will have prepositions, will place modifying adjectives after rather than before the noun, will put titles before proper names, and so on. Likewise, if a language has the word-order SOV, then an interesting typology should allow one to predict that it will ipso facto be postpositional, put attributive adjectives before the noun, etc.

In this paper, we argue in favour of the NO IMPAIRMENT HYPOTHESIS, whereby L2 functional categories, features and feature values are attainable, and against the NO PARAMETER RESETTING HYPOTHESIS, according to which L2 learners are restricted to L1 categories and features, as well as against the LOCAL IMPAIRMENT HYPOTHESIS, which claims that the interlanguage grammar is characterized by inert feature values. An online experiment was conducted, investigating adult learners' knowledge of properties relating to clitic projections. Advanced learners of French (L1s English and Spanish), together with a native speaker control group, were tested on a variety of constructions involving clitics by means of the SENTENCE MATCHING procedure (Freedman & Forster 1985). L2 learners distinguished in their response times between certain kinds of grammatical and ungrammatical clitic placement, as did the native-speaker controls, suggesting the attainability of L2 properties distinct from the L1.

Variation in language on the basis of formality (register variation) is often neglected both in grammatical descriptions and in sociolinguistic analyses. I demonstrate here that in Sinhala, and perhaps in other diglossic languages, register variation in syntax cannot be ignored. In a Head-Driven Phrase Structure Grammar (HPSG) analysis based on a corpus of naturally occurring Sinhala texts, I propose an analysis of register variation in which the syntax of all observed registers is accounted for within a single grammar. I further explain how the approach to register variation developed here can be extended to other types of sociolinguistic variation.

The Principle of Phonology-Free Syntax (PPFS) is a proposed universal principle of grammar that prohibits reference to phonological information in syntactic rules or constraints. Although many linguists have noted phenomena that appear to them to be in conflict with it, the appearances are misleading in all cases we have examined. This paper scrutinizes four instructive cases in French that appear to falsify the PPFS. Section 1 deals with the alleged relevance of syllable count to the description of attributive adjective placement; section 2 addresses the validity of a rule mentioning consonantality in stating the agreement rule for adverbial tout; section 3 turns to the issue of preposition choice (e.g. en vs. au) with geographical proper names; and section 4 takes a look at a purported case of phonological reference in stating the rule for ellipsis of a clitic pronoun and an auxiliary in a coordinate structure. In each case we bring independent evidence to bear on the problem in order to show that the analyses employing phonology-sensitive syntactic statements are in error and the prediction of the PPFS is confirmed.

Part of the meaning of a verb phrase containing a form of one of the English modal verbs CAN and WILL1 is that the actions, events, etc., indicated by the following full verb took place, takes place and will take place. I shall use the term MODALITY to refer to the meaning of the modal verb and EVENT to refer to the meaning of the full verb and say that, in certain circumstances, there is ACTUALITY of the event or that the event is ACTUALIZED. I wish to discuss the circumstances in which this is so and attempt to provide explanations in terms of some of the characteristics of modality in general.

Gussmann (1980) has presented an analysis of Polish morphophonemics within the framework of SPE (Chomsky & Halle, 1968). His treatment appeals extensively to underlying representations containing segments which never appear in surface representations. He justifies this use of absolute neutralization with the claim that other rules of Polish phonology appeal to such underlying forms. Rubach (1984) has reanalysed much of Gussmann's description within the framework of Kiparsky's (1982) Lexical Phonology, retaining the analysis of the neutralized segments and even adding to the inventory of abstract elements. The arguments presented by both scholars are a paradigm of elegance and ingenuity, representing generative phonological analysis at its best. However, the considerable reliance they put on abstract underlying representations, morphologically conditioned rules, extrinsic rule ordering and other devices which give rise to a high degree of opacity arouses suspicion. For this reason I have sought to re-examine some of their material within a more ‘concrete’ interpretation of Lexical Phonology, making use of recent ideas from theories of non-linear phonology and theories of allomorphy.

The idea that certain morphological and phonological irregularities are due to speakers' desire to avoid homophony is widely invoked, yet has also come under strong criticism as an explanation which is neither necessary nor sufficient. In most cases there is no way to resolve the question, since the assumption that something is being avoided is itself a theoretical construct. In this article I attempt to address this last difficulty by looking at gaps in inflectional paradigms – where it is clear that something is being avoided – that plausibly correlate with potential homophony. These fall into two types: (i) lexical, where portions of the paradigms of two lexeme would be homophonous, and (ii) paradigmatic (i.e. syncretism), where forms within the paradigm of a single lexeme would be homophonous. Case studies of Tuvaluan, Russian, Mazatec, Tamashek and Icelandic confirm the effects of homophony avoidance as a genuine, if non-deterministic, principle.

Welsh has generally been analyzed as allowing two types of relative clauses and other A´-constructions, one involving movement leaving a wh-trace, the other involving a resumptive pronoun in situ. In this paper, I argue that, despite the appearance of agreement, which seems to license a null resumptive pronoun, relative clauses formed on a number of syntactic positions (object of periphrastic verb, object of preposition, embedded subject) may involve movement. Both movement and non-movement strategies are argued to be available for some syntactic positions (object of preposition, embedded subject), and separate constraints must therefore be established for the distribution of each. Resumptive pronouns are argued to be subject to a variant of the A´-Disjointness Requirement. For wh-trace, the Welsh evidence is compatible only with an account involving multiple cyclic movement via a VP-external position (SpecAgrOP) as well as SpecCP.

The present article discusses the notion of variation in discourse as an essential characteristic of language and the linguistic parameters that can be used for its study. The article describes the different traditions in the study of variation and places special emphasis on the role of prosodic analysis in the study of spoken discourse. The study explores the use of discourse markers in the London-Lund Corpus and describes their linguistic variation by introducing the notion of APPROPRIATENESS. This notion, which combines the quantitative and qualitative presence of elements in discourse, is based on a mathematical index that can describe discourse variation with a sound systematic criterion.

Besides the canonical Subject–I–VP structure, English has several inversion constructions in which the subject follows the inflected verb. The most familiar is Subject Auxiliary Inversion (SAI) which is analyzed as an instance of Head Movement (I–to–C-movement across the subject) in the generative tradition. In this paper we investigate Comparative Inversion (CI), which appears to be a special case of SAI in which ellipsis is required (Merchant 2003). Contrary to this analysis, we show that the subject can stay low in a noncanonical position, violating the Extended Projection Principle (EPP) in exactly those instances where it is under comparison and therefore heavily accented and contrastively focused. Our analysis shows that the non-application of the EPP is tied to regular interactions of syntax with phonology and syntax with semantics. We extend this in depth analysis to other English focus inversions and provide evidence that phonological highlighting and focus on the low subject can suspend the EPP. Thus, our analysis supports research programs which assume minimal syntactic structure and operations in interaction with interface constraints that are independently required for explanation.

In recent literature on transformational grammar there has been a good deal of discussion of tag sentences (Klima, 1964; Katz and Postal, 1964; Bolinger, 1967; Arbini, 1969). The type of construction under consideration is exemplified in the following (incomplete) paradigm: (1) John has gone, hasn't he? (2) John hasn't gone, has he? (3) John has gone, has he (?) (4) John hasn't gone, hasn't he (?) (5) Come early, won't you (?) (6) Come early, will you (?)

1. It may be claimed that current views regarding the nature of sound change fall into two broad categories: the more traditional attitude would treat an individual sound change as a complex trend or process taking perhaps several generations to establish itself, and then retaining its activity over a long period of time; whence the characteristic concern of classical historical linguistics with the establishment of absolute and relative termini post and ante quern, i More recently adherents of the generative–transformational school have interpreted sound changes as readjustments in the system of phonological rules; thus Postal (1968: 270) claims:

‘What really changes is not sounds but grammars. And grammars are abstract objects – sets of rules represented in human organisms.’

Recent developments in syntactic theory suggest that phrase structure is cross-linguistically more uniform than assumed so far, and that the order spec-head-complement may be the only permissible one. The present article takes issue with this view, showing that the derivation of final complementizers from initial ones by means of IP-raising faces serious difficulties. The discussion focuses on Bengali and similar languages which may be called ‘hybrid’ because both orders, IP-C as well as C-IP, are attested. Five arguments are raised which indicate that these orders are not derivationally connected. The discussion bears results which may also be of interest for linguistic typology, language change and acquisition.

The term morphological reversal describes the situation where the members of a morphological opposition switch their functions in some context (as with Hebrew gender marking, where -Ø~-a marks masculine~feminine with adjectives but feminine~masculine with numerals). There is a long tradition of polemic against the notion that morphology can encode systematic reversals, and an equally long tradition of reintroducing them under different names (e.g. polarity, exchange rules or morphosyntactic toggles). An examination of some unjustly neglected examples (number in Nehan, aspect in Tübatulabal, tense in Trique and argument marking in Neo-Aramaic) confirms the existence of morphological reversal, particularly as a mechanism of language change. This is strong evidence for the separateness of morphological paradigms from the features that they encode.