Linguistische Berichte Heft 185

Surface dysgraphia is a (symptom of a) neurological language disorder with a loss of lexicosemantic writing routines and preserved segmental Phoneme-to-Grapheme conversion (PGC). Surface dysgraphic subjects write words and non-words phonologically plausible but with orthographic errors concerning words with ambigue PGC. Such an isolated use of the PGC routine allows us to focus analysis on segmental processing in writing. H.S., a German patient suffering from primary progressive aphasia, showed typical surface dysgraphic errors in writing words to dictation. The analysis of his responses revealed the implicit knowledge of syllabic principles and their application in writing. He could phonologically differenciate native and non-native sounding items and showed context-sensitivity in PGC (as for initial /f/). Furthermore H.S. succesfully applied the syllabic rules for inserting the silent '·syllable initial h''. Moreover the patient payed attention to the graphemic "syllable weight'' in realising the lengthening-h - that is he "stretched" graphematically short syllables by inserting it and "compressed" long syllables by leaving it out. Doing so he paralleled the probability for the existence of the lengthening-h in German. But he even took notice of the graphemic syllable weight in realizing lt/ as - which the German writing system does not use analogously. Thus the segmental ''surface•· routine works in relative autonomy of lexical routines as it is supposed in the logogen-model of word processing (cf. Patterson, 1988). Phonemegrapheme-conversion applies various syllabic rules and principles. The model will have to be specified in this point.

Montagovian semantics treats word meanings in terms of intensions and extensions which are meant to capture speakers' ability to decide which words are to be used for what objects, properties, events, etc. The aim of the present paper is to refine this view in order to model facts about language change, acquisition, and knowledge of language. I will reconstruct intensions in terms of (a) given core referents, and (b) contextually detennined ways to generalize these to füll sets of (real or counterfactual) extensions. This reconstruction of word meanings is clarifying in several ways. lt can explain how speakers can install the meanings of new words in a language on the basis of a limited set of intended referents, it makes clear how the limited knowledge of single speakers allows them to grasp the meaning of words in their language, and it allows to model certain types of historical meaning change and creative language use. The approach will finally be related to conceptual semantics, prototype theory, and classical truth value based semantics.

This paper is a state of the art report of forensic linguistic anonymous authorship analysis (FLA) of authentic extortion and threatening letters. lt summarizes briefly some theoretical, methodological and empircal aspects of forensic linguistic expert testimony given by the author in the last thirty years. The focus is ( l) on the range of variation of requests of linguistic expert opinion in real life; (2) the nature and methodological implications of data of verbal behavior relevant for FLA; (3) a perspective on a contrastive analysis oftext types in FLA; (4) an exemplaric illustration of feature configuration analysis of speaker specific behavior and its impact for FLA.