Refine
Year of publication
- 1991 (4) (remove)
Document Type
- Working Paper (4) (remove)
Language
- English (4) (remove)
Has Fulltext
- yes (4)
Is part of the Bibliography
- no (4) (remove)
Keywords
- Antonym (1)
- Grammatische Relation (1)
- Kognitive Semantik (1)
- Opposition <Linguistik> (1)
- Tagalog (1)
- Universalgrammatik (1)
Institute
- Extern (1)
- Informatik (1)
- Physik (1)
Oppositeness, i.e. the relation between opposites or contraries or contradictories, has a fundamental role in human cognition. In the various domains of intellectual and psychological activity we find ordering schemas that are based, in one way or another, on the cognitive figure of oppositeness. It is therefore not surprising that the figure and its corresponding ordering schemas show their reflexes in the languages of the world. [...] We shall be dealing with oppositeness in the sense that a linguistically untrained native speaker, when asked what would be the opposite of 'long' can come up with some such answer as 'short', and likewise intuitively grasp the relation between 'man' and 'woman', 'corne' and 'go', 'up' and 'down', etc. Thinking that much of the vocabulary of a language is organized in such opposite pairs we must recognize that this is an important faculty, and we are curious to know how this is done, what are the underlying conceptual-cognitive structures and processes, and how they are encoded in the languages of the world. We shall leave out of consideration such oppositions as singular vs. plural. present vs. past, voiced vs. unvoiced, oppositions that the linguist states by means of a metalanguage which is itself derived from a concept of oppositeness as manifested by the examples which I gave earlier. Our approach will connect with earlier versions of the UNITYP framework. However, as a novel feature, and, hopefully, as an improvement, we shall apply some sort of a division of labor. We shall first try to reconstruct the conceptual-cognitive content of oppositeness and to keep it separate from the discussion of its reflexes in the individual languages. We shall find that a dimensional ordering of content in PARAMETERS and a continuum of TECHNIQUES is possible already on the conceptual-cognitive level. In order to keep it distinct from the level of linguistic encoding we shall use a separate terminology, graphically marked by capital 1etters.
Grammatical relations – in particular the relation 'subject of' – and voice are of central concern to any theory of universal grammar. With respect to these phenomena the analysis of Tagalog (and the Philippine languages in general) has turned out to be particularly difficult and continues to be a matter of debate. What traditionally has been called passive voice in these languages […] appears to be so different from voice phenomena in the more familiar Indo-European languages that the term 'focus' was introduced in the late 1950s to underscore its 'exceptional' nature [...]. Furthermore, […] an inflationary use has been made of the term 'ergative' in the last decade; it can thus no longer be assumed that it has an unequivocal and specific meaning in typologizing languages, apart from the technical definition it might be given within a particular framework. But if the Philippine 'focus' constructions are neither passive nor ergative, how else can they be analysed? [...] In this paper a ease will be made for the claim that 'focus' marking should be analysed in terms of orientation, a concept used […] for capturing the difference between English (and, more generally, Indo-European) orientated nominalisations such as 'employ-er' or 'employ-ee', and unorientated nominalisations such as 'employ-ing'. This approach implies that 'focus' marking is derivational rather than inflectional as often presumed in the literature. This is to say that what is typologically conspicuous in Tagalog is not the 'focus' phenomenon per se, since this is very similar to orientated nominalisations in many other languages, but rather the very prominent use of orientated formations (i.e., derivational morphology) in basic clause structure.
It is well known that artificial neural nets can be used as approximators of any continous functions to any desired degree. Nevertheless, for a given application and a given network architecture the non-trivial task rests to determine the necessary number of neurons and the necessary accuracy (number of bits) per weight for a satisfactory operation. In this paper the problem is treated by an information theoretic approach. The values for the weights and thresholds in the approximator network are determined analytically. Furthermore, the accuracy of the weights and the number of neurons are seen as general system parameters which determine the the maximal output information (i.e. the approximation error) by the absolute amount and the relative distribution of information contained in the network. A new principle of optimal information distribution is proposed and the conditions for the optimal system parameters are derived. For the simple, instructive example of a linear approximation of a non-linear, quadratic function, the principle of optimal information distribution gives the the optimal system parameters, i.e. the number of neurons and the different resolutions of the variables.