Mind as motion: Explorations in the dynamics of cognition Ed. by Robert F. Port and Timothy van Gelder (review)

BOOK NOTICES 219 words described. Unfortunately, to the best of my knowledge, there is no bilingual Lithuanian-English or English-Lithuanian dictionary containing information on the stress patterns of the Lithuanian words described. For this kind of information one must turn to an explanatory Lithuanian dictionary such as the multivolume, Lithuanian Academy of Sciences Dictionary , Lietuviu kalbos zodynas (Vilnius: Mokslo ir enciklopedijy leidykla, 1956) or the Dictionary of contemporary Lithuanian, Dabartinès lietuviu kalbos zodynas (Vilnius: Mokslo ir enciklopediju leidykla, 1993). It has been explained to me that such information is unnecessary for a Lithuanian, and most Lithuanian -English dictionaries are prepared for Lithuanians, not for English speakers. This latter assumption is probably true, but no matter how good the supplement by P, it does not alter the fact that for English translation and the Lithuanian stress pattern one must look in two different dictionaries. It is frequently difficult to get East European publications since the print run is usually rapidly exhausted ; the major advantage then of this new dictionary is that it is published by a reputable western press (Routledge), and we hope it will be in stock for a good long time. [William R. Schmalstieg, The Pennsylvania State University.] The development of morphological systematicity : A cross-linguistic perspective . Ed. by Hanna Pishwa and Karl Maroldt. (Tübinger Beiträge zur Linguistik 399.) Tübingen: Gunter Narr Verlag, 1995. Pp. 321. 'Development' here refers to Ll or L2 acquisition, historical change, and creolization. Edith Bavin, Dorit Ravid and Suzanne Schlyter describe Ll acquisition involving respectively Warlpiri, Hebrew, and French-Swedish bilingual children. Peter Mühlhäusler, Frederick C. Jones, and Eduardo D. Faingold discuss Creoles, Jones focussing on Krio and Faingold on definiteness and specificity. L2 acquisition is explored by Hanna Pishwa (the English progressive and German accusative), Richard Robison (English verb inflection), Carol Aisha Blackshire-Belay (German past tense forms), Heide Wegener (German noun plurals), Stephan Schmid (Spanish-Italian interlanguage), and Stephan Schneider (Italian). Finally, Romuald Skiba and Ullrich Steinmüller report on compounding in German technical terminology. There is plenty here for readers with general theoretical interests. Maroldt's introductory article, 'Morphology : A functional interface', where he discusses the roles of the 'morphological mode' and the 'syntactic mode' in grammar, indirectly raises the neglected question of why in all languages there are distinct grammatical categories word and sentence. His answer is that morphology is concerned with what is permanent and generic rather than what is transient: thus, tomato-knife means not any knife you cut tomatoes with but rather an instrument specifically designed for the purpose. I do not think this will work because of the existence of both transient morphological constructs (productively formed and semantically transparent wordforms) and permanent, or lexicalized, syntactic constructs (idioms). In M's use of the term lexeme it is correspondingly unclear whether he means 'word in abstraction from its forms' (Matthews' lexeme) or 'idiosyncratic item' (DiSciullo and Williams' lísteme). But although I am sceptical about M's answer, the question he raises is eminently worthwhile. For me, the other contributions which stand out are those of Wegener and Bavin. Plural forms of German nouns have been a recent battleground between rule-oriented and connectionist linguists. Wegener provides food for thought for both camps. Among learners of German as L2 in German primary schools, one of the first plural suffixes to be used is -er (as in Männer 'men'), even though its typefrequency is low and by anyone's criterion it is relatively marked. Yet the children do not overgeneralize -er, unlike -e, -en, and zero. Wegener suggests that -er is especially distinctive because it signals only plurality, never case, and because it triggers umlaut in umlautable stems consistently, unlike the more type-frequent -e, which triggers umlaut only sometimes. In childhood L2 acquisition, therefore, transparency and redundancy seem to be especially valued even though the realizations concerned violate constructional iconicity and may not be generalized Bavin shows that in Warlpiri a kind of inflection with a clear syntactic function (nominal case marking ) is acquired more slowly and erratically than the syntactically functionless allomorphy of tense and mood marking associated with the five inflection classes of Warlpiri verbs. Why are...