1 SYNTAX OF THE WORLD’S LANGUAGES 3 Freie Universität Berlin, 25-28 September, 2008 26 September, 2008 Section B Anton Zimmerling (Moscow) meinmat@yahoo.com Clitic-Second Languages and Verb-Second languages in a diachronic perspective 1 The paper discusses the relation of Germanic word order systems based on the Verb-Second Constraint (=V2 systems) and word order systems with clitic clusters in clausal 2 nd position (= systems with Wackernagel’s Law ~ W-systems ~ systems with C-oriented clitics). V2 languages are typologically rare outside Europe, while languages with 2 nd position clitics are attested both in Old and Modern Indo-European Languages (cf. Hittite, Avestan, Old Greek, Old Novgorod Russian, Serbo-Croat, Czech, Slovene, Pashto, Osset) and in a number of non-Indo-European languages, cf. Kabyle Berber (Afroasiatic), Lummi (Salish), Makah (Wakashan), Warlpiri and Djaru (Pama-Nyungan), Quiavini Zapotec (Otomangean), Mayo (Uto-Aztecan) etc. It is plausible that all languages with Wackernagel’s law, irrespective of their genetic origin, share a number of constraints on clitic placement (clusterization rules, movement patterns, orientation of clitics), cf. [1], [2], [12], [5]. 1. V2 languages are mostly attested in languages from the Germanic group 2 . W2 languages are widely attested in genetically unrelated languages from different families [13]. 2. There are, however, no *V-3, *V4 languages, and no *W-3, *W5 languages etc. Is there a magic of clausal 2 nd position or a general principle of constituent ordering? 3. Approaching Second: V2/V1 languages and W2/W1 languages. Topological constraint Germanic languages South and West Slavic languages V2 ~ W2 German, Dutch, Swedish, Norwegian, Faroese Serbian/Croatian/Bosnian, Slovene, Bulgarian, Czech, Slovak V2/V1 ~ W2 / W1 Icelandic, Yiddish Macedonian, Upper Sorbian 1 Research is a part of the project “The typology of free word order languages and models of inversion” funded by the Russian Foundation for Humanities (grant RGNF 06-04-00203a), whose support is gratefully acknowledged. I am grateful to the audience of the SWL 3 conference for the valuable feedback. All responsibility for the statements made here is my own. 2 A similar, though not identical mechanism (the so called Operator-Second Constraint) is attested in Kru languages (Niger-Kongo family). Verb-Second phenomena are also attested in Rhaeto-Romance languages and Kashmiri, in both cases under the influence of Germanic languages.