The Emergence of 2P Clitics in Slavic and the Order of Cliticization

This is the 2nd version of my SLS 2008 talk given in Columbus. The revisions are technical: I checked the spelling of the Old Russian examples and fixed minor bugs. I discuss the order of cliticization in the Slavic languages and compare the reconstructions proposed by Zaliznjak (1993, 2008) and Migdalsky (2008). I argue that Proto-Slavonic already had the class of 2P clitics. The Old Novgorod Russian (ONR) system of clitic placement cannot not be derived from the Old South Russian (OSR) / Old Church Slavonic (OCL) system, but the latter can be derived from a ONR-like system. Modern Slavic systems with 2P clitics originate from a ONR-like system. Andrej A. Zaliznjak‘s ( 2008) claim that 'weak clitics‘ (dat., acc. pronouns, BE-auxiliaries) behaved as non-clitic words in OSR/OCL is not tenable. In the OSR/OCR pronouns and auxiliaries behaved as verbal clitics. This fact implies that after the disintegration of Proto-Slavonic, clitics belonging to the latest cliticization layer did not have 2P-properties any longer. The BE-auxiliries could still cliticize and develop in the direction of 2P clitics in a group of conservative Slavic dialects (ONR, Slovene, Carpatian Ukrainian). Comments. The Old Novgorod Russian clitic template was first described by Zaliznjak (1993), who proved that ONR had Wackernagel's law and a class of 2P clitics incl. particles, clitic pronouns and 1-2 p. auxiliaries. Zaliznjak's book was well-known in Russia but remained virtually unknown to those scholars who pursued clitic studies in the western part of the world, cf. Franks, King (2000), Pancheva (2005) and Migdalsky (2007, 2009). In 2008, Zaliznjak published his book "Old Russian Enclitics". I learned it shortly before this SLS 3 talk in Columbus (2008) and made last-minute changes in my presentation posted here. Directly after my talk, the "Journal of Slavic Linguistics" invited me to review Zaliznjak's book for them. But parallel to that, I got another invitation to review Zaliznjak (2008), from the Russian journal "Russkij jazyk v nauchnom osveshchenii". My review Zimmerling (2009) appeared there already in January 2009, so I could not publish the same review twice even though in a different language. Instead, I published my own analysis of 2P cliticization in Slavic in my book Zimmerling (2013) and journal papers Zimmerling (2012a, 2012b; 2012c, 2021), Zimmerling, Kosta (2013) as well in the book chapter Kosta, Zimmerling (2013). See the 'Related works' section. The point that the clitic-internal ordering in the ONR template [Particles > Dative pronouns > Accusative pronouns > 1-2 BE-auxiliaries] follows the time of cliticization was first made by myself in 2002 in the book Zimmerling (2002: 70). I quote this page in my presentation. In 2008, Zaliznjak repeated that point iand provided the same arguments in his book on Russian enclitics (2009: 16). I quote it in my presentation too. A conсlusion from this analysis (not drawn by Zaliznjak, who never reverted to clitic syntax since that) is that unless the 'diachronical' principle of cliticization is given up, it is impossible to reconstruct one and the same clitic template for all varieties of Old Slavic languages, since the auxiliary slots are placed differently in different Slavic areas. However, one can postulate different clitic templates for Old East Slavic, Old Balcanic Slavic and Old West Slavic, which I did in Zimmerling (2013) and Zimmerling, Kosta (2013). I did not make any revisions in the second uploaded version on my SLS 2008 talk but checked the spelling of all exampled and fixed minor bugs. Today, I would like to fix the use of the term 'macroparameter'. In early 2000-s, I called Wackernagel's law 'macroparameter of syntax': what I meant it that W's law is not a single (prosodic or syntactic) principle but rather a bundle of syntactic parameters corresponding to class of the world's languages. However, the term 'macroparameter' in the formal typology, cf. Longobardi (2005), Baker (2008) is firmly associated with the architecture of UG, while all parameters defined for a specific language samples and datasets are identified as 'microparameters'. I realize this terminological clash is misleading since I never claimed that Wackernagel's law, V2 constraints and other 2P phenomena are based on any UG principles.