An account of grammatical acquisition is developed within the parameter setting framework applied to a generalized categorial grammar (GCG). The GCG is embedded in a default inheritance network yielding a natural partial ordering (reflecting generality) of parameters that determines a partial order for parameter setting. Computational simulation shows that several resulting acquisition procedures are effective on a parameter set expressing major typological distinctions based on constituent order, and defining 70 distinct full languages and over 200 subset languages. The effects on acquisition of inductive bias, that is, of differing initial parameter settings, are explored via computational simulation.

Computational simulation of populations of language learners and users instantiating the acquisition model shows that: (1) variant acquisition procedures, with differing inductive biases, exert differing selective pressures on the evolution of language(s); and (2) acquisition procedures will evolve towards more efficient variants in the environment of adaptation. The reciprocal evolution of language acquisition procedures and of languages creates a genuinely coevolutionary dynamic, despite the relative speed of linguistic selection for language variants compared to natural selection for variant language acquisition procedures.

According to a standard view, the reciprocal pronoun has a fixed semantic value that defines a relation of weak reciprocity, and any stronger readings it may appear to have are pragmatic or lexical interactive effects (Fiengo & Lasnik 1973, Langendoen 1978). Dalrymple et al. 1995 counterproposes that the reciprocal pronoun has a flexible semantic value defining a range of readings of varying logical strength and that a semantic principle determines the reading required for a given reciprocal sentence on the basis of the meaning of its predicate. This article presents psycholinguistic evidence from adult speakers of English, Norwegian, and Dutch, and from child speakers of Dutch and Norwegian, which supports Dalrymple et al.'s analysis of the lexical content of the reciprocal pronoun but which also strongly suggests that the interpretive principle they posit is pragmatic rather than semantic in nature.

It is a familiar observation that phonological processes frequently fail to apply to geminates. A number of previous proposals attempted to account for these geminate inalterability effects in terms of a distinction between singly and multiply linked autosegments. Subsequent research, however, has observed that geminate inalterability is inviolable only in the domain of lenition processes. In this article, an account of this generalization is couched within a general optimality theoretic treatment of lenition, in which a scalar effort minimization constraint interacts with a set of lenition-blocking constraints. The geminate inalterability generalization follows from this effort-based approach to lenition, coupled with certain assumptions concerning the effort involved in geminates and their lenited counterparts.

Syntactic features like case, person, and gender are often assumed to have simple atomic values that are checked for consistency by the standard predicate of equality. The case feature has values such as nom or acc, and values like masc and fem are assumed for the feature gender. But such a view does not square with some of the complex behavior these features exhibit. It allows no obvious account of feature indeterminacy (how a particular form can satisfy conflicting requirements on a feature like case), nor does it give an obvious account of feature resolution (how person and gender features of a coordinate noun phrase are determined on the basis of the conjuncts). We present a theory of feature representation and feature checking that solves these two problems, providing a straightforward characterization of feature indeterminacy and feature resolution while sticking to structures and standard interpretations that have independent motivation. Our theory of features is formulated within the LFG framework, but we believe that similar solutions can be developed within other syntactic approaches.

This article tests earlier claims about the universality of patterns of polysemy and semantic extension in the domain of perception verbs. Utilizing data from a broad range (approx. 60) of Australian languages, we address two hypothesized universals. The first is Viberg's (1984) proposed unidirectional pattern of extension from higher to lower sensory modalities (i.e. intrafield extensions, like ‘see’ > 'hear'). The second hypothesized universal is that put forward by Sweetser (1990) regarding the extension of perception verbs to cognition readings (i.e. transfield extensions, like 'see' > 'know'). She suggests that vision has primacy as the modality from which verbs of higher intellection, such as 'knowing' and ‘thinking’, are recruited, and proposes that verbs meaning ‘hear’ would not take on these readings, although they often extend to mean ‘understand’ or ‘obey’. Though both hypotheses assign primacy to vision among the senses, the results of our Australian study show that Viberg's proposal remains intact, while Sweetser's is proved false. Australian languages recruit verbs of cognition like ‘think’ and ‘know’ from ‘hear’, but not from ‘see’. It appears that, at least as far as perception verbs are concerned, transfield semantic changes are subject to greater cultural variability than intrafield semantic changes. We argue that the same semantic domain can have its universal and its relativistic side, a foot in nature and a foot in culture, and conclude by demonstrating that there are good social and cultural reasons driving the extension of ‘hearing’, but not ‘seeing’, to ‘know’ and ‘think’ in Australian Aboriginal societies.

Milyilyi-lu kulirninpa, langa kulirninpa-lu brain-erg hear/think, ear hear-at him/her ‘Our brain thinks/hears, our ears think/hear’ Kukatja, from Peile 1997

In most standard dialects of English, there is a gap in the paradigm of the verb be where we expect to find amn't. But how do we know that this gap exists, since learners have no positive evidence that amn't is ungrammatical? It is even more puzzling since there is no gap when the subject is inverted (aren't I...?). Familiar explanations for this gap fail; in particular, it cannot be the result of conservative acquisition strategies. The explanation offered here is based crucially on a combination of multiple-default inheritance and function-based morphology, as embodied in word grammar. The gap is due to a Nixon-diamond conflict between two competing values for the same morphological function required by the categories negative and first-person. The inverted form is supplied by stipulation because of the functional pressure to supply a ‘casual’ form. Various dialect alternatives to the Standard English pattern are also considered. The success of this explanation shows that language must use default inheritance, multiple nonorthogonal inheritance, and morphological functions.

Four lexical features of a noun are relevant to agreement: (i) semantic conditions on reference, (ii) person, number, and gender features of the referential index, (iii) concord features, and (iv) declension class. These four features are correlated by a chain of binary constraints. When individual constraints are violated, the chain is broken, resulting in intricate patterns of mixed agreement. Three main types of mixed agreement are predicted, all of them attested in Serbo-Croatian. This theory helps explain Corbett's (1983) crosslinguistic agreement hierarchy.

Variations in postverbal constituent ordering have been attributed to both grammatical complexity (heaviness) and discourse status (newness), although few studies compare the two factors explicitly. Through corpus analysis and experimentation, we demonstrate that both factors simultaneously and independently influence word order in two English constructions. While past investigations of these factors have focused on their effects in language comprehension, we argue that postponing heavy and new constituents facilitates processes of planning and production.

We attempt to demonstrate that the substitution of a foreign segment in the borrowings of our database, which includes 14,350 segmental malformations from French and English loanwords in eight distinct languages, involves its replacement by a single native segment. This tendency is so strong in our database as to be virtually exceptionless, except where nasal vowels are concerned. These vowels are systematically adapted as an oral vowel followed by a nasal consonant (VN), a process we call unpacking. We document this process and suggest that it results from the fact that contrastive nasal vowels are fundamentally biphonemic, that is they have two root nodes. The influence of orthography is refuted, and a number of apparent counterexamples where segments other than nasal vowels seem to unpack are reanalyzed in terms of independent native processes.

We propose an account of linking patterns that does away with intermediary mechanisms such as thematic or actor/undergoer hierarchies. Instead, constraints on word classes, defined by both syntactic and semantic criteria, encode generalizations between semantic roles and syntactic arguments. We show that the generalizations a linking theory needs to capture can be modeled via the same mechanisms as other lexical generalizations, using conditions specified within the hierarchy of word classes. Each condition provides a partial specification of the mapping between semantic roles and syntactic arguments. We argue that this constraint-based, verb-class-based view of linking offers several empirical advantages: partial regularities and exceptions are easily accommodated, fine-grained semantic distinctions relevant to linking are countenanced, and cross-cutting similarities between semantic and syntactic verb classes are economically captured.

The lexical semantic structures of change-of-state verbs are explored via linguistic theory, corpus analysis, and psycholinguistic experimentation. The data support the idea that these verbs can be divided into two classes, those for which the change of state is internally caused and those for which it is externally caused (Levin & Rappaport Hovav 1995, cf. Smith 1970). External causation change-of-state verbs have been hypothesized to denote two subevents, internal causation change-of-state verbs only one event. Consistent with this difference, the psycholinguistic data indicate that, in both transitive and intransitive constructions, sentences with external causation verbs take longer to comprehend than sentences with internal causation verbs.

This article shows that endoclitics do exist in Udi, a language of the North East Caucasian family, and this fact poses a challenge to the lexicalist hypothesis. Clitics may be positioned between the morphemes of complex verb stems and immediately before the final segments of monomorphemic verb stems. The author argues, on the basis of accepted tests for wordhood, that complex verb stems are single words, not phrases. On the basis of criteria developed by Zwicky and Pullum (1983), it is argued that the clitics of Udi are true clitics. An analysis of the placement of clitics in various positions inside verb stems is proposed in optimality theory. The author shows that phonological phenomena do not provide an alternative basis for positioning these clitics and concludes that clitics in Udi are a counterexample to the lexical integrity hypothesis.

The advent of nonlinear phonology has resulted in an explosion of studies relating to Spanish syllable structure and stress placement, but most of these studies claim to represent linguistic competence and language structure, not actual mechanisms used by speakers in speech production and comprehension.

The present study is couched within Skousen's ANALOGICAL MODELING OF LANGUAGE (AML) (1989, 1992, 1995). AML attempts to reflect how speakers determine linguistic behaviors such as stress placement. According to AML, when an unfamiliar word needs to be stressed, speakers access their mental lexicon, search for words similar to the word in question, then apply the stress of the word(s) found to the word in question.

The 4,970 most common Spanish words served as the database for the study. AML correctly assigned stress to about 94% of these words. The errors it made closely reflect the pattern of errors made by Spanish-speaking children in a study by Hochberg (1988). Moreover, Aske's nonce word probe (1990) showed that native speakers are sensitive to a certain subpattern in Spanish stress assignment—a subpattern which does not receive representation in rule models. The analogical model of Spanish stress mirrors Aske's findings.

The primary purpose of this study is to present evidence, based on experimental data from Western European languages, that there is orderly variation in the choice of perfective auxiliary with intransitive verbs. Specifically, auxiliary selection is sensitive to a hierarchy of aspectual/thematic verb types: some verbs require a given auxiliary categorically, whereas others allow both auxiliaries to a greater or lesser extent depending on their position on the hierarchy. It is argued that this gradience has potentially important implications for the unaccusative hypothesis, and more generally for theories of the lexicon-syntax interface.

Previous studies have been largely unsuccessful in investigating Middle English quantity changes in terms of moraic or foot structure. The present study approaches quantity changes as the consequence of the phonologization of a syllable cut prosody (Trubetzkoy's (1939) 'Silbenschnittkorrelation') and provides new evidence for the relevance of syllable cut to the diachronic phonology of English. The evidence is from the author's partial reconstruction of the phonological system of the early Middle English dialect represented in the Ormulum (Northeast Midlands, circa 1180), based solely on the orthographic and metrical evidence provided by its author, Orm. Although Orm's innovative orthographic system previously defied analysis, it is fully interpretable once the role of syllable cut is recognized. The analysis provides crucial support for the relevance of syllable cut to the diachronic phonology of English, and ultimately to the explanation of the pan-Germanic similarities in the 'breakdown' of phonological quantity.

The German Perfekt has two quite different temporal readings, as illustrated by the two possible continuations of the sentence Peter hat gearbeitet in i, ii, respectively:

(i) Peter hat gearbeitet und ist müde.

Peter has worked and is tired.

(ii) Peter hat gearbeitet und wollte nicht gestört werden.

Peter has worked and wanted not to be disturbed.

The first reading essentially corresponds to the English present perfect; the second can take a temporal adverbial with past time reference (‘yesterday at five’, ‘when the phone rang’, and so on), and an English translation would require a past tense (‘Peter worked/was working‘). This article shows that the Perfekt has a uniform temporal meaning that results systematically from the interaction of its three components—finiteness marking, auxiliary and past participle—and that the two readings are the consequence of a structural ambiguity. This analysis also predicts the properties of other participle constructions, in particular the passive in German.

Der Alte folgte der Leiche und die Söhne, Albert vermocht's nicht. Man fürchtete für Lottens Leben. Handwerker trugen ihn. Kein Geistlicher hat ihn begleitet.

Goethe, Werthers Leiden—

The introductory linguistics course is the primary antidote that academic linguists can offer to commonly held, yet basically wrong-headed, views about language. It is essential, therefore, that this course be meaningful to the nonmajor student. Through a series of five vignettes, each by a different author and representing a variety of institutional types, we explore options other than ‘baby’ theoretical linguistics to the introductory linguistics course. A fundamental conclusion to these vignettes is that success in reaching the nonmajor turns on taking account of the institutional context and the student population and tailoring courses to be sensitive to these variables. This conclusion is driven home with an administrative view as to why every academic linguist and linguistics program should find the search for a successful introductory course compelling.