Sanskrit Syntax: Selected Papers Presented at the Seminar on Sanskrit Syntax and Discourse Structures.

• Author: Kobayashi, Masato
• Position: Book review

Sanskrit Syntax: Selected Papers Presented at the Seminar on Sanskrit Syntax and Discourse Structures, 13-15 June 2013, Universite Paris Diderot, with an Updated and Revised Bibliography by HANS HENRICH HOCK. Edited by PETER M. SCHARF. Providence, RI: THE SANSKRIT LIBRARY, 2015. Pp. xxx + 522. $60.

From the title, readers might imagine a comprehensive description of Sanskrit syntax like J. S. Speijer's Sanskrit Syntax (Leiden: Brill, 1886), or a collection of papers on syntax such as Studies in Sanskrit Syntax edited by Hans Henrich Hock (Delhi: Motilal Banarsidass, 1991). Unlike these works, however, the volume under review covers diverse topics, ranging from Paninian grammar to the computational processing of Sanskrit. It contains selected papers from a seminar held in Paris in 2013, of which the main interest was the computational formalization of Sanskrit grammar, the parsing and tagging of Sanskrit texts, and the creation of Treebank corpora. In that respect, this volume continues the line of Sanskrit Computational Linguistics, edited by Gerard Huet, Amba Kulkarni, and Peter Scharf (Berlin: Springer, 2009).

Hans Henrich Hock ("Some issues in Sanskrit syntax," pp. 1-52) surveys works on syntax since 1991 with a special focus on his own contributions. Together with the seventy-two-page bibliography at the end of the volume (pp. 399-470), which updates Deshpande and Hock's bibliography in Hock 1991 (pp. 219-44), it provides a selective overview of syntactic topics currently at issue. On the first topic of how free Sanskrit word order is, the works cited agree that Sanskrit is a configurational language, i.e., has a hierarchical phrase structure, but differ on whether there is an unmarked word order and whether the syntactic tree has linear ordering of its constituents or not. Hock supports Schaufele's approach, which assumes a head-final basic word order, because it conforms well to the overall SOV typology Sanskrit shows. On the second topic of the relative-correlative clause structure. Hock reaffirms his own claim that the Sanskrit relative clause is conjoined to the main clause, citing cases where there is no clear relationship between the relative and correlative pronouns, such as relative clauses serving as conditional clauses, and relative clauses containing interrogation and imperative modality. In the section on gender agreement. Hock takes up what he calls "upside-down" agreement, in which pronominal subjects adopt the agreement features of their predicates and not of their antecedents. Hock finds its parallel in the agreement of the locative absolute...