Cross-linguistic perspectives on language processing (Q2778935)

The cross-linguistic studies in theoretical linguistics appeared at the end of the seventies, became nowadays a fruitful and hardly-worked research subfield, and involved both investigating a single phenomenon available in several languages (e.g. passivization), as well as investigating a single phenomenon available in a single language or a group of related languages (e.g. expletives in non pro-drop languages). Thus research can be reformulated in recent cross-linguistic studies of language processing as bearing on the issue of the universality of the human language processor, comprising (but not limiting to) the following components: (a) the structural relations (describing the syntactic representation of a sentence, accessed by the processor); (b) the operational primitives (building the syntactic structure): (c) the principles and preferences (guiding the processing strategies); (d) the general architecture of the language processor (serial vs. parallel models, autonomy of syntax vs. interactive models, etc.)NEWLINENEWLINENEWLINEThe main purpose of the highest quality papers enclosed in this book is a cross-linguistic (and psycho-linguistic) study on the human sentence processor, investigating a computational model of the language processor within a multilingual approach. The contents of the nine selected, outstanding contributions in the book is briefly described in what follows: (1) \textit{M. De Vincenzi}, \textit{V. Lombardo} (Introduction) represent the authorized and gentle introduction to the research field, and an extensive description of the papers enclosed.NEWLINENEWLINENEWLINEContributions (2) and (3) below present general theories of parsing and re-analysis, focusing on the parsing primitives, in terms of basic operations and structural relations. In (2) \textit{J. D. Fodor}, \textit{A. Inoue} (Garden Path Re-Analysis: Attach (Anyway) and Revision as Last Resort), the main point is that it is not the case when no more legal attachments exist. The immediate consequence of this point is the abandonment of the RALR (Revision As Last Resort) principle (Fodor and Frazier, 1980) and the coalescence of the Attach and Attach Anyway operations. The empirical data to justify the proposed model come from the analysis of the Japanese left branching structure and (recently) from English. (3) \textit{R. Frank}, \textit{K. Vijay-Shanker} (Lowering Across Languages) address the type of structural relations which the linguistic processor accesses when it updates the linguistic structures. Their main idea is that \(c\)-command relation on entities of the syntactic structure is more directly related to parsing difficulty than the dominance relation. The consequence is that parametric variations in linguistic primitives and parsing strategies are no longer necessary. This radically new perspective on parsing models will probably elicit considerable experimental research.NEWLINENEWLINENEWLINEThe next three papers investigate the parsing of modifiers (such as relative clause, prepositional phrases, adjectival phrases). (4) \textit{T. Baccino}, \textit{M. De Vincenzi}, \textit{R. Job} (Cross-Linguistic Studies of the Late Closure Strategy: French and Italian) make the first steps towards an experimental paradigm that allows the evaluation of parsing preferences across languages. (5) \textit{C. Frenck-Mestre}, \textit{J. Pynte} (Resolving Syntactic Ambiguities: Cross-Linguistic Differences) investigate the parsing principle influencing the attachment of a relative clause to a complex noun phrase, with experimental data from French. (6) \textit{M. J. Traxler} et al. (Architectures and Mechanisms for Sentence Processing) experimentally evaluate the attachment of prepositional phrase and relative clauses to complex noun phrases. If confirmed, the proposed approach poses real challenges to either serial or constraint-based language processors. (7) \textit{M. Meng}, \textit{M. Bader} (The Role of Case and Number Features in Syntactic Ambiguity Resolution) represents a cross-linguistic study with a monolingual approach, specifically German. The authors introduce the Case Preference principle, predicting that, in case of ambiguity, (a) a structural case (nominative or accusative) is preferred to a lexical case, and (b) the nominative case is preferred to the accusative case.NEWLINENEWLINENEWLINEThe next two papers (8) and (9) address the area of language production, in a multilingual approach. (8) \textit{H. P. Branigan}, \textit{M. Prat-Sala} (A Cross-Linguistic Perspective on Discourse Context and Syntactic Processing in Language Production) describe six experiments conducted in three languages (English, Spanish and Catalan), and confirm the universality of influence of discourse structure on the syntactic structure. Cross-linguistic variation in the syntactic structure production is explained. (9) \textit{M. Heydel}, \textit{W. S. Murray} (Conceptual Eflects in Sentence Prinming: A Cross-Linguistic Perspective) show that topicalization (in German) primes the production of a passive form (in English), thus giving a clear indication of the necessary abstraction towards activation of the conceptual form in sentence priming.NEWLINENEWLINENEWLINEThe volume provides a cross-linguistic perspective on the current research in sentence processing, and points out methodologies for the more accurate definition of theoretical primitives of parsing. The book is addressed to a large category of people for which natural language (re)present a major interest.