Search

In the present essay some of what have been the more influential conceptions of phonology are examined in the light of what I see as 'intrusions' into its domain, starting with a strand of orthographic influence. I am concerned with the extent to which various 'phonological' proposals might be described as 'graphophonological' rather than strictly phonological. This examination also reveals the interaction of the graphophonological impulse with the impact of other considerations that are not proper to synchronic phonology. Most pertinent here is the more familiarly controversial formulation as synchronic generalizations of what are substitutes for the diachronic regularities reflected in morphophonological alternations: what I call 'anachronic' phonology, associated, for the most part, with the conflation of morphophonology and phonology; another instance of 'pseudo-phonology'. Scrutinized here are particularly proposals concerning sound structure associated with the classical littera, with the ('taxonomic') phoneme, and with the morphophoneme (or 'systematic phoneme'). Finally, after an evaluation of an overtly graphophonological proposal, the orthographic rather than phonological value of 'CVCV' phonology is explored in relation to the inadequacies of the Linear B syllabary. [ABSTRACT FROM AUTHOR]

The article examines the striking correlation of the directive infinitive with negation in Dutch. This correlation is interpreted in the light of recent findings on the form of the prohibitive in the languages of the world and is explained in terms of the negative-first principle and politeness. The validity of the two explanations is further investigated from a cross-linguistic perspective. The preference for preverbal negation is shown to be even stronger in prohibitive than in declarative sentences. The difference in politeness between positive imperative and prohibitive speech acts is argued to be reflected in a wide range of languages. On the whole, the article illustrates the fruitful, two-way interaction between the study of language usage and typology. [ABSTRACT FROM AUTHOR]

This article addresses the question of control in Polish in the context of the current Movement vs. Agree debate and the nature of Obligatory Control (OC). More specifically, it aims to show that many phenomena which are allegedly inconvenient for the movement-based theory – such as control into infinitives introduced by a lexical complementiser and the controller's case independence of the semi-predicate or predicative adjective – can be dealt with without postulating a case-marked OC PRO. The controller's escape from a CP-infinitive is made possible by a small modification of the mechanics of the Phase Impenetrability Condition (PIC) and the ‘derivational window’. The optionality of case transmission and case independence in certain contexts is ascribed to the optional phase-status of infinitives, the clitic-like properties of the bare complementiser and the application of the default case mechanism to semi-predicates and adjectival predicates. [ABSTRACT FROM AUTHOR]

German Glide Formation converts /i/ to [j] before vowels, e.g. ‘Spain’, but the rule is consistently blocked in neologisms containing the suffix – esk, e.g. hippiesk [hrpiℇsk] / *[hrpjℇsk]. It is argued below that the underapplication of Glide Formation in such examples follows from a requirement that the stem in a derived word must be identical to the unaffixed base. The base in such examples will be shown to be a free-standing morpheme as opposed to a bound root. The analysis proposed will be shown to be supported in additional examples in which Glide Formation is blocked from applying to the [i] preceding the suffix – aner, but only if the suffix follows a free-standing morpheme, e.g. ‘adherent of Schumi (Michael Schumacher)’. [ABSTRACT FROM AUTHOR]

Since Booij (1994, 1996) it has become increasingly clear that inflectional morphology can take part in lexeme formation and compounding. Booij (1994) recognized the need for substantive constraints on the ways inflection can feed derivation, and restricted its derivational use to deictic categories, including Number. Pursuing this search for constraints, I propose that Number is a single morphological category covering two abstract functions (cf Beard 1995), and that it can be inherent only when it expresses the more ‘lexical’ of those functions, and thus means more than the grammatical feature would. This ‘lexical’ Number expresses properties of the lexeme but stands halfway between the lexical core and the properly inflectional categories. It encodes mereological (part-whole) properties of the noun's interpretation, thus paralleling the role of Aspect in the verbal domain, and like Aspect it can be integrated to different degrees in the grammatical system of a language. In some languages, this type of information has a specific morphological expression (so-called collective affixes). In others, it appears only as non-canonical semantics (and sometimes form) for Number inflection. Inherent Number, both as a component of lexeme-formation and as fixed Number value on certain nouns, consists in the expression of Nominal Aspect through the morphology of Number. Morphology is not ‘split’, but its uses are. Inherent inflection, specifically Number, arises in certain languages as a by-product of the separation of (morphological) form and meaning. The article develops these views by presenting first a relatively detailed exemplification from several sources (section 1), followed by some critical reflections on the peculiarities of these constructions, to the effect that inherent Number must be qualitatively different from inflectional Number (section 2). Section 3 sets out in detail the hypothesis that inherent Number is the inflectional expression of Nominal Aspect, and section 4 concludes the argument by hypothesizing that Number not only can, but must have a distinct interpretation as a lexicalized property than as a regular inflectional one. [ABSTRACT FROM AUTHOR]

This paper is concerned with the relationship between (proper) names and word structure, and specifically with the classification of names and with the role of (classes of) names in lexical derivation. The major source of exemplification is English. §1 outlines the categorization of names proposed in a sister study devoted to syntax of names (Anderson in preparation), as well as other relevant parts of the syntactic description given there. In §2.1 different kinds of personal and place names are differentiated and their more salient morphosyntactic characteristics commented upon. This is followed in §2.1 by a consideration both of the historic sources of names and of some of the properties and functions of systems of naming; and there is noted the typical de-semanticisation of names compared with the common words that are their typical historical source, such that the synchronic role of common (descriptive) elements in name systems tends to be restricted. These discussions are relatively informal, but § offers a more (lexical) derivational processes that can form names and with the role of names in derivational processes forming other names or items of other categories, and the light these throw on the semantics of (classes of) names and naming. [ABSTRACT FROM AUTHOR]

The present article offers an Optimality-Theoretic analysis capturing the distribution of ‘superheavy syllables’ in Modern English, i.e, syllabics consisting of a long vowel + con- sonant(s), a diphthong + consonant(s), or a short vowel + two or more consonants, e.g. feel, farm. It will be shown that superheavy syllables are restricted to surfacing either (i) in word-final position, (ii) before a compound boundary, or (iii) before a consonant-initial suffix. By contrast, superheavy syllables are typically non-occurring in other environments. In the present article the following theoretical claims are made. First, superheavy syllables are analyzed as trimoraic. Second, environments (i)-(iii) above describe the right edge of a phonological word. The distribution of superheavy syllables is captured below with the constraint ALIGN-3μ, which ensures that the third mora in a syllable be situated at the right edge of a phonological word. Finally, ALIGN-3μ will be shown to be violated in certain (predictable) word-internal cases, e.g the syllable [aun] in mountain, such facts will be accounted for by various language-specific constraint rankings. The present surface based analysis will be argued to be superior to an earlier one proposed for Modern English in a derivational framework by Borowsky (1986, 1989). [ABSTRACT FROM AUTHOR]

This paper offers a speculative account of the acquisition of syntactic word class categories by children, based on a notionally founded theory of categories. Categories are differentiated by (non-)possession of simplex semantic features which may combine in varying degrees of complexity. While less complex categories are universal, more complex combinations, such as that characterising adjectives, are not necessarily so. Markedness, or relative accessibility, correlates with relative complexity of representation. With respect to word classes as such, the child's ‘initial state’ is assumed to be undifferentiated, with the first categorisation being based on referentiality. And evolving systems are constrained by a minimality requirement on change of representation which ensures that acquisition of marked categories is late. Initial syntactic structure and its development are built on the basis of the perceived subcategorisation requirements of the emerging word classes. [ABSTRACT FROM AUTHOR]

This paper analyzes the system of Modern Standard German (MSG) noun plural formation in the framework of Optimality Theory (Prince & Smolensky 1993, hereafter OT). We concentrate our analysis on the thorny issue of why vowel mutation (umlaut) in the stem is used in combination with other plural suffixes or alone in order to mark plurality. In this paper, we maintain that a historical process of umlaut ceased early on to be an active phonological process in German Umlaut became reinterpreted as a morphological marker of plurality which then spread analogically into other classes. We then capture the synchronic distribution of plural markers in Standard German using a set of independently motivated constraints and constraint rankings based both on phonological properties of the lexical input and on universal principles of morphology adapted from current work in Natural Morphology (Mayerthaler 1981, Wurzel 1984a; hereafter NM). We show that OT can provide a principled analysis of the umlaut+suffix plurals and predict more ‘natural’ patterns or plural formation in non-standard dialects and the responses of native speakers to nonce word experiments. In several respects, general principles of NM anticipate developments in OT, although this has not been generally acknowledged. In this paper, we make explicit the relationship between two theories that share much of the same conceptual ground. Specifically, we incorporate NM ideas of markedness, iconicity. and uniformity into an OT constraint framework. [ABSTRACT FROM AUTHOR]

There exists an interesting correlation between aspect realizations and typological ergativity in those languages that exhibit ‘split ergativity’: the ergative properties (morphological, sometimes and never totally, also syntactic) are displayed only in the past and the perfect tenses or perfective aspect (among which dominantly the periphrastic perfects), as opposed to the present tense where the nominative-accusative system is adhered to. In typological research, this phenomenon has been known to be restricted to what has been called ‘Type-B ergativity’, i.e. only in languages with an ergativity split (Trask 1979:388). Type-A languages of the ergative type, do not exhibit this strict correlation. The question why this should be of such overwhelming empirical evidence has not received an answer (Trask 1979:396: ‘[...] why should the perfective so often be constructed on an creative basis?’). In this article, such a link is drawn. It is claimed that the type of ‘syntactic ergativity’ (or ‘unaccusativity’) discussed in generative syntax in fact boils down to perfective predicates and constructions which are expressed in terms of object predicates (‘predicatives’ or ‘small clauses’). [ABSTRACT FROM AUTHOR]

This article talks about universal semantic primitives as a basis for lexical semantics. It notes that the semantic system of a language is like a set of Lego blocks, of different shapes and sizes. The meanings of words are like objects constructed out of various Lego blocks. The purpose of lexical semantics is to study such objects, to deconstruct them into their constitutive building blocks, and to seek generalizations about the different types of building blocks and different ways of putting them together. The main difficulty of lexical semantics is that while it needs a solid foundation in the form of well justified semantic primitives, no set of such primitives is given at the outset, rather, the primitives themselves must be found through large-scale lexicographic investigations, both monolingual and cross-linguistic. This double task of finding the primitives via lexicographic description and basing lexicographic description on the primitives may seem self-contradictory and thus impossible to accomplish.

This paper explores how adults acquire a new lexical item, and how they integrate it into their overall lexical knowledge. It suggests that the findings are relevant both to lexical semantics, and 10 studies of the mental lexicon. In particular, the paper examines how the word wimp and its derivatives (mainly wimpish, wimpishness, and wimp out) have become widely-known in British English in a time-span of around ten years. Wimp-usage predated entry into standard dictionaries, so dictionary consultation is unlikely to have played an important role. This paper therefore analyses a corpus of over 500 occurrences of wimp-words from British newspapers 1990–3, mainly from The Times and Sunday Times, on the assumption that word learning can occur from reading. and that newspaper usage overlaps with ‘normal’ usage. It shows that over 80% of wimp-word tokens contain information on their meaning in the immediate surrounding text. This involved one or more of the following: reference to the sex of the wimp (usually male), collocation with a word indicating feebleness (e.g. ‘paihetic wimp’), contrast with a ‘strong’ non-wimp (e.g. ‘From wimps to warriors’), overt negative evaluation (e.g. ‘reviles as a wimp’). covert negative evaluation (e.g. ‘Who needs an enclosed cockpit? Wimps’). Finally, it argues that ‘the wimp effect’ reinforces the idea that a desirable male is one who is a belligerent action-man, and so promotes and sustains cultural stereotypes. [ABSTRACT FROM AUTHOR]

This article argues that the concept of "markedness" is inapplicable to syntactic phenomena, and that "markedness reversal," frequently invoked to explain the local predominance of a Marked form in putatively "marked" environments, is unnecessary once it becomes clear what the qualitative motivation is for the quantitatively imbalanced distribution of variants. The argumentation relies on the difference between (open-ended) syntax and the inherently closed domains of phonology and morphology, and on the relevance of generally acknowledged communicative principles, such as "contextual support" and "vantage point".

This article seeks to explore lexical processing in an agglutinative language and the organization of the lexicon. From the point of view of lexical processing the major empirical question that arises in relation to agglutinative languages is whether the morphologically complex words are represented in a holistic or a morphologically decomposed form in the mental lexicon and whether lexical access requires some form of morphological parsing of the word. In this article several possible models are being differentiated. These include holistic word entries, separate morphemic entries, and serially specified morphemic entries. It also briefly reports some preliminary results from two pilot studies which bear on the validity of these models. Both experiments examined the lexical processing of morphologically simple and complex Hungarian words using different on-line techniques.

The enterprise in parsing research is to find out how the mind creates syntactic representations. The veritable theory of parsing will be both part of a well-motivated mental model of language and language processing and accommodate the experimental evidence. The goal of this paper is to compare two sorts of parsing models that are being discussed in literature, taking into account two criteria. The two types of models the author will discuss are the Licensing Parser and Immediate Attachment Parser. In section one, the author defines these models. In the rest of the paper, the two kinds of models are evaluated.

This work is part of a research project on the on-line parsing of Italian sentences. In particular I will discuss the processing of Italian sentences that are structurally ambiguous or that differ in syntactic complexity: (1) declarative sentences with post verbal subjects; (2) declarative sentences with unaccusative and unergative verbs; (3) single clause wh-questions with non-referential (who) and referential (which-N) wh-items; (4) wh-questions with ‘who’ and ‘which’ items, and extraction out of embedded clauses; (5) wh-questions with ‘who’ and ‘which’ items, and extraction out of an embedded clause with a lexically filled complementizer (‘that’). The hypothesis is that the parser is sensitive to syntactic complexity, in particular to the complexity of chains, a linguistic object that defines relations among positions in sentences (Chomsky, 1981; Rizzi, 1990). The processing hypothesis is that in case of chain length ambiguity, the parser will always choose the shortest one and that more complex chains will be harder to process than simpler ones. This preference is expressed in the Minimal Chain principle (MCP), a principle that applies at S-structure, determining the decisions made at ambiguous points and the complexity of unambiguous sentences. The results from reading time and comprehension tasks show that readers are sensitive to chain complexity in that, following MCP, they always choose the shortest chain, and that longer chains are processed more slowly than shorter ones. [ABSTRACT FROM AUTHOR]

A theory of human syntactic processing must be both grammatically responsible and psychologically plausible. This article seeks to explore the structural relations in the grammar and the parser. In the first section of this article the author outlines his grammatical assumptions. Section two outlines the proposed parsing model. The empirical justification here, for the most part, is based on English data. In section three, the author turns to an examination of syntactic processing in Japanese. It is argued that a wide range of processing phenomena in Japanese can be accounted for without altering or parameterizing the parsing model. Sections four and five contrast the proposed parsing model with alternative approaches such as the characterization of automatic structural reanalysis and the underspecification theory.

The paper discusses the functioning of Lower in three different frameworks: Laskowski's (1975) early cyclic framework, Gussmann's (1978, 1980) simultaneous rule application hypothesis and Rubach's (1981) strict cyclicity approach. The latter is related to the lexical phonology framework developed for Polish in Rubach (1984) and especially in Booij and Rubach (1987). The strict cyclicity framework is shown to be basically correct in its account of the Polish data. It is argued furthermore that an analysis of the prefixal part of Polish derivatives provides convincing evidence that Polish prefixes fall into two classes: cyclic and noncyclic. On the other hand, the analysis of the sufflxal section of derivatives supports the claim that the features triggering Lower must be autosegmentalized. A preliminary investigation of the Polish data strongly suggests that the autosegmentalized features can be successfully analysed on the same plane that Halle and Vergnaud (1987) postulate to account for stress phenomena in a number of diverse languages. [ABSTRACT FROM AUTHOR]

The reports on generative grammar in Europe. The extended standard theory of generative transformational grammar led a very diasporadic life in Europe in the mid seventies. There were a few pockets of linguists interested in this esoteric field, but most of them were young--students or assistants--and they were scattered and hardly aware of each other. The first two official conferences were again held in Amsterdam, the Netherlands, and the very first one was not only again very successful but also very important in the development of the generative theory. The topic chosen for this discussed the question as to whether all transformational processes could be reduced to rules that obey certain domain restrictions such as the subjacency condition, or whether some rules had to be considered truly unbounded as was being proposed by writer Joan Bresnan and others. The papers read at the conference brought to light an astounding wealth of arguments based on material from a! great variety of languages. Virtually all of the evidence pointed in the direction of the view that all of syntax is subject to some locality principles. Thus it may be claimed, in fact, that this particular debate, one of the most heated ones around that time was in effect settled at this conference.

This article discusses a note on government, binding, and case assignment in Polish. It is argued that case marking of Polish emphatics can be given a principled and theoretically interesting account by the Government and Binding Theory. Polish makes use of two types of emphatics. These are HB (head bound) emphatics, which are categorically adjectives and appear in all cases and end-of sentence (E)-emphatics which are categorically adverbs and are marked for two cases only, the nominative and the dative. The former are assigned case by a case agreement rule for adjectives, the latter, by a special case marking rule for E-emphatics which makes crucial use of the binding principles and the theory of government.

This article provides information on the sentential nature of prenominal adjectives in German. It notes that generative grammar has made a considerable move from abstract structures towards more surface oriented ones. One major exception to this lies in the enrichment of surface structures with empty elements like NP-trace or PRO. As the goal of generative linguistics is to explain what the underlying mental representations of grammar look like, how languages can be acquired by children the way they are, this move can be justified if a) theory postulating empty elements for certain positions helps to keep a number of subtheories of grammar simpler and if b) postulating empty elements will not lead to a proliferation of stipulations concerning the nature of and rules governing these elements.

This article demonstrates that only grammars of a particular type, generative-transformational ones, can meet the conditions of theoretical orientation of linguistics being directed towards explanatory adequacy. Since 1957, five models of such grammars had been proposed. All of them are axiomatized systems that generate infinite sets of sentences with associated structural descriptions but differ with regard to the significance attached to individual components of these descriptions. The distinction between surface and deep structure, came as a result of the tendency to avoid repetition of identical information with respect to the selectional requirements of related expressions.

This article is meant as a contribution to the theory of grammar. It attempts to clarify the notion of grammatical relation and applies relational analysis to a couple of illustrative examples from syntax and morphology. The role of sequential order in the expression of syntactic relations is assessed and considered minimal. The concept of relationality, of the inherence of a grammatical relation to a morpheme, is clarified. And two kinds of dependency relations, government and modification, is identified as having their source in different types of relationality.

It is held by many grammarians that the relation of syntax and semantics is characterized by the Fregean principle of compositionality. The principle can be stated in various ways. In this article, the following formulation is adopted: the literal meaning of an expression is uniquely determined by the literal meanings of its subexpressions and their mode of composition. And the author believes that the validity of the formulation is beyond doubt and thus any grammar, whether organized to reflect it directly or not, may ultimately be required to satisfy it. And one of the systems that are precisely designed to reflect the principle is Montague Grammar.

Scientific definitions of word all yield results which do not correspond to traditional word-boundaries. On the other hand traditional words cover all sizes of linguistic units, from the bare morph to the complete sentence. A word may belong to any "level" or to more than one. This means that the word might be left to everyday usage as a more or less stable but incidental agglutination of elements, but has no place in linguistic description. Consequently, word-disciplines like morphology and word-formation have to be abandoned. These fields are indeed said by some to be on the decline, yet the majority of structural linguists currently use word, word-formation, etc, as if nothing has happened. This inertia is due to the human tendency to synthetic thinking (mostly involving speaker's orientation in describing), a vice invalidating some Structuralist models (both Item and Arrangement and item and Process, Descriptive Order, φ-derivation) as well as, in a different way, the Generative model. My suggestions are: abolish all a priori divisions of the description including the one into "words" and "word-parts", hence give up word-formation with its traditional compounds, derivatives, etc. Use the same analytic, descending method with the Glossematic double definition for all kinds of syntagms, above and below the phrase level, to show constituents and their nucleus-satellite relation, as illustrated by a detailed analysis of a Greek sentence at the end of the paper, using FREI's graphic representation. [ABSTRACT FROM AUTHOR]

This article presents four arguments for a deep structure re-analysis taking place in the fifteenth century. Two arguments are for a deep structure for Modern English and the other two are plausibility arguments. This re-analysis is also argued for on the basis that it gives a natural account of the changes taking place. This "naturalness" is based on typological facts whereby under this analysis the changes are seen to be a consequence of the SOV-SVO change, rather than the result of foreign influence, etc.; and the general claim that root transformations provide the initial focus of syntactic change.

This article argues that a contrastive grammar should be viewed on the one hand in its relation to a grammatical basis describing the languages to be contrasted and on the other hand in its relation to the learning and teaching problems which it claims to explain. The first relation is hereby emphasized and it is shown what form a generative contrastive description, i.e. a contrastive description based on the generative grammar, might take. A generative contrastive description which contains a contrastive component for the explicit representation of specific contrastive relations between sentences, is shown. The requirement of comparability is fulfilled by a multilingual generative grammar which describes more than one language. The requirement of contrastability is taken into account by the contrastive component.