Skip Navigation


Logic Journal of IGPL Advance Access originally published online on October 16, 2007
Logic Journal of IGPL 2007 15(5-6):653-687; doi:10.1093/jigpal/jzm051
This Article
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
15/5-6/653    most recent
jzm051v1
Right arrow References
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Krynicki, M.
Right arrow Articles by Torres, J. M. T.
Right arrow Search for Related Content
Social Bookmarking
 Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author, 2007. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Games on Trees and Syntactical Complexity of Formulas

Michal Krynicki

Faculty of Mathematics and Natural Sciences, Cardinal Stefan Wyszynski University, Dewajtis 5, 01-815 Warsaw, Poland. E-mail: m.krynicki{at}uksw.edu.pl

Jose Maria Turull Torres

Information Science Research Centre, Dept. of Information Systems, Massey University, Private Box 756, Wellington, New Zealand. E-mail: J.M.Turull{at}massey.ac.nz


   Abstract

Ehrenfeucht-Fraïssé games have been introduced as a means of characterizing the relation of elementary equivalence between structures, or relational database instances in first order logic (FO), or equivalently Relational Calculus. In~the usual Ehrenfeucht-Fraïssé games the rules are determined by a linear ordering of a fixed lenght or, equivalently, by a special kind of tree – a chain of a fixed length –, where each point of that ordering or node of that tree corresponds to a quantification operation. Here we consider Ehrenfeucht-Fraïssé games whose rules are determined by arbitrary trees such that their nodes correspond either to quantification operations ("q–nodes") or to connective operations ("c–nodes"). By playing games on trees, we can refine the class of sentences which are considered in a given game, since a tree represents a particular class of sentences. We define and study several variations of tree games, for first and second order logic (SO). We give a sufficient condition for FO and SO equivalence restricted to formulae with up to n connectives, and hence also a sufficient condition for the non expressibility of a given query in those logics with formulae whose number of logical connectives is less than a given integer. We also give a sufficient condition to prove simultaneous lower bounds in both the number of connectives and in the quantifier types needed to express a given property in FO. If we consider only quantifier types, we get a characterization of the relation of preservation of sentences in the fragment of FO with the given set of quantifier types. We also study tree games for {Sigma}n and {Pi}n formulae. To illustrate the use of our games we use them to prove lower bounds in the connective size for several FO queries, like size of a database, size of a clique in a graph, size of a unary relation, transitive property in a graph, and degree of a node in a graph. Regarding SO, we prove lower bounds for quantifier rank for the parity query.

Finally, we give a precise characterization of the logic whose elementary equivalence is characterized by a given tree game, as well as several equivalent characterizations of the existence of a winning strategy for Duplicator in the classical Ehrenfeucht-Fraïssé game.

Key Words: Ehrenfeucht-Fraisse games • games on trees • definability • syntactical complexity.

Received for publication 15 September 2006.
Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?




Disclaimer:
Please note that abstracts for content published before 1996 were created through digital scanning and may therefore not exactly replicate the text of the original print issues. All efforts have been made to ensure accuracy, but the Publisher will not be held responsible for any remaining inaccuracies. If you require any further clarification, please contact our Customer Services Department.