A Deterministic Parser With Broad Coverage

This paper is a progress report on a scries of three significant extensions to the original parsing design of (Marcus J980).* The extensions are: Ihe range of syntactic phenomena handled has been enlarged, encompassing sentences with Verb Phrase deletion, gapping, and rightward movement, and an additional output representation of anaphor-antcccdcnt relationships has been added (including pronoun and quantifier interpretation). A complete analysis of the parsing design has been carried out, clarifying the parser's relationship to the extended I R(k,t) parsing method as originally defined by (Knuth 1965) and explored by (Szymanski and Williams 1976). The formal model has led directly to the design of a "stripped down" parser that uses standard LR(k) technology and to results about the class of languages that can be handled by Marcus-style parsers (briefly, the class of languages is defined by those that can be handled by a deterministic, two-stack push-down automaton with severe restrictions on the transfer of material between the two sucks, and includes some strictly context-sensitive languages). 1 EXTENDING THE MARCUS PARSER While the Marcus parser handled a wide range of everyday syntactic constructions, there are many common English sentences that it could not analyze. One gap in its abilities arises because it did not have a way to represent the possibility of rightward movement that is, cases where a constituent is displaced to the right: A book [about nuclear disarmament] appeared yesterday. --> A book appeared yesterday [about nuclear disarmament]. Further, the only way that the Marcus parser could handle leftward movement was via the device of linking a "dummy variable" (a trace) to an antecedent occurring somewhere earlier in the sentence. For instance, the sentence, "Who did Mary kiss?" is parsed as, Who did Mary kiss trace!, where trace is a variable bound to its "value" of who, indicating the intuitive meaning of the sentence, "For which X, did Mary kiss X" . Jn the original parser design, a trace was of the category NP, so that only Noun Phrases could be linked to traces. But this meant that sentences where other than NPs are displaced or deleted cannot be analyzed. This includes the following kinds of sentences, where deleted material is indicated in square brackets.