Relational knowledge is a hallmark of human cognition and the subject of a vast body of research. In this paper we argue that existing accounts of relations are inadequate because they have little to say abouthowrelations arise in the rst place and because they tend to be limited to particular sorts of relational tasks. We present a new approach to the learning and representation of relations, an approach that makes use of what we call micro-relation units (MRUs). Each MRU represents a relation between features of di erent objects rather than between objects themselves. We show howthis approach o ers an account of the grounding of relations, and we describe a neural-network implementation of the MRU framework and show howit enables a variety of relational tasks to be performed by the same system.

A series of three experiments investigated semantic and syntactic effects in the production of Adjective+Noun phrases in Dutch. Bilinguals (Dutch native speakers) were presented with English nouns and were asked to produce an Adjective+Noun phrase in Dutch which included the translation of the noun. In two experiments, we blocked items by either semantic category or grammatical gender.We found that participants performed the task slower when the target nouns were of the same semantic category than when they were from different categories; and faster when they were of the same grammatical gender than when they were of different gender. In a final experiment, both manipulations were crossed in order to both replicate the previous experiments and to test for interactions between the two effects. The results of the first two experiments were replicated, and crucially no interaction was found. These findings are compatible with models of lexical retrieval in production in which, first lexico-semantic and lexico-syntactic information are separable; second the flow of activation between the two is feedforward.

This report has been prepared in the form of an HTML document, which is available on the World-Wide Web at ftp:##ftp.cs.indiana.edu#pub#gasser#Playpen#TR1#tr#tr.html. It makes extensive use of color #gures, animated applets, and hyperlinks. There is a compressed Postscript version of the report available by anonymous ftp to ftp.cs.indiana.edu at #pub#gasser#Playpen#TR1.ps.Z. This includes the color #gures but not the applets and hyperlinks. Therefore we highly recommend the version on the Web

& Here we address the contentious issue of how nouns and verbs are represented in the brain. The co-occurrence of noun and verb deficits with damage to different neural regions has led to the view that they are differentially represented in the brain. Recent neuroimaging evidence and inconsistent lesion– behavior associations challenge this view. We have suggested that nouns and verbs are not differentially represented in the brain, but that different patterns of neural activity are triggered by the different linguistic functions carried by nouns and verbs. We test these claims in a functional magnetic resonance imaging study using homophones—words which function grammatically as nouns or verbs but have the same form and meaning—ensuring that any neural differences reflect differences in grammatical function. Words were presented as single stems and in phrases in which each homophone was preceded by an article to create a noun phrase (NP) or a pronoun to create a verb phrase (VP), thus establishing the word’s functional linguistic role. Activity for single-word homophones was not modulated by their frequency of usage as a noun or verb. In contrast, homophones marked as verbs by appearing in VPs elicited greater activity in the left posterior middle temporal gyrus (LpMTG) compared to homophones marked as nouns by occurring in NPs. Neuropsychological patients with grammatical deficits had lesions which overlapped with the greater LpMTG activity found for VPs. These results suggest that nouns and verbs do not invariably activate different neural regions; rather, differential cortical activity depends on the extent to which their different grammatical functions are engaged. &

Abstract not found

This paper computes the semantic representation of while as the pragmatically most relevant one which speakers select from a variety of grammatical constructions in which while may occur in current English. The semantic representation of while provides the condition for translating it into the adequate German equivalent. This computation is implemented in a unification--based formalism and may thus be applied in a machine translation system.

Human linguistic communication differs from the communication of other animal species in three main ways. First, and most importantly, human linguistic communication is symbolic. Linguistic symbols are social conventions by means of which one individual attempts to share attention with other individuals by directing their attentional or mental states to something in the outside world. Other animal species do not communicate with one another using linguistic symbols, most likely because they do not understand that conspecifics have attentional or mental states that they could attempt to direct or share (Tomasello, 1998a, 1999). This mental dimension of linguistic symbols gives them unparalleled communicative power, enabling their users to refer to and to predicate all kinds of diverse perspectives on objects, events, and situations in the world. The second main difference is that human linguistic communication is grammatical. Human beings use their linguistic symbols together in patterned ways, and these patterns, known as linguistic constructions, come to take on