SWETWOL is implemented in the framework of Koskenniemi's (1983) two-level model. It contains a 48,000 item lexicon and a full inflectional description. Special attention was paid to the design of a computational analysis of productive Swedish compounds. Recall (coverage) and precision of SWETWOL meet high standards. SWETWOL has been extensively tested on various types of texts.

Grammatical information in ASL can systematically be marked on the face. Such nonmanual marking extends over the c-command domain of the trigger, and therefore provides information about the hierarchical organization of the language. Consistent with evidence available from the distribution of non-manual markings—as illustrated with respect to wh-marking and negation—a basic clausal structure for ASL is proposed. Furthermore, we suggest, contrary to generally accepted claims about ASL, that both Tense and Agreement are structurally present in all ASL main clauses. This analysis allows for a uniform account of the licensing of null subjects in ASL. Evidence in favor of this analysis, and against a dual licensing mechanism (as proposed in Kegl, 1985, and Lillo-Martin, 1986, 1991b), is presented.

For both of the two Danish construction han sidder og sover “he sits and sleeps” and han kom gående “he came walking (along)”, it has been suggested that the second verb is governed by the first in the same way as a main verb is governed by an auxiliary or modal verb. Specifically, this would imply that the second verb is located in the non-finite verb slot in the topological sentence pattern. It is shown that these suggestions are not justified and that the Danish verbal constructions only allow infinitives and past participles as verbal complements of another verb.

In this article it is demonstrated how verbal and non-verbal expressions, i.e. speech and gestures, interact to convey social and cognitive information in an excerpt taken from a conversational discussion. The theoretical framework for the article is that of information processing. Within this framework, verbal and non-verbal expressions in the conversation are characterized as discourse operations which develop and refine information in a generative search in analogy with an AI production system. The discourse operations function as underspecified instructions that in interaction with given information states generate new information states in search of a desired information state.

This paper examines the role of physiology of the hand relevant to particular handshapechange data from Taiwan Sign Language. The physiology of the hand is shown to be completely responsible for one aspect of handshape-change (that fingers must become unspread when closed). In addition, the physiology provides the explanation for two tendencies in handshape-changes (i) that when fingers extend, they tend to spread, and (ii) that when less than all five fingers perform a handshape change, the fingers that do so are the thumb, index, and middle or a subset.

This article discusses different transcription systems and models of sign language phonology that fall into two major categories: The innovative work of Stokoe, with later adaptations from different European countries on one hand, and on the other, several autosegmental models. The focus in the discussion is how these different models deal with hand orientation and specification of movement directionality. On the basis of this discussion, features and concepts for a new model are proposed.

American Sign Language uses spatial locations as an integral part of its systems of pronominal reference. Referents mentioned in an ASL discourse may be associated with points in the signing space (“referential loci”). Pronominal signs are directed to those loci, unambiguously indicating their referents. Referential loci are sometimes analyzed as overt phonological representations of the semantic indices associated with referents in the discourse. In this paper I show that referential loci do not function merely to index or identify referents, but rather that they carry a great deal of additional information pertaining to the conceptual location of referents in the discourse space. I examine contexts in which one referent is associated with two distinct loci in the signing space, and offer a preliminary formulation of principles governing locus selection in such contexts.

One way of expressing a particular point of view depends on deixis; the sender ‘lends’ his means of expressing his own point of view to another referent. In Danish Sign Language there are three such basically deictic means of expressing a particular point of view, namely shifted reference, shifted attribution of expressive elements, and shifted locus. Eveathough especially the last two phenomena may seem special to sign languages, they have clear functional parallels in spoken languages.