Handling Scope Ambiguities in English

HANDLING SCOPE AMBIGUITIES IN ENGLISH Sven Hurum Department of Computing Science 615 General Services Building University of Alberta Edmonton, Canada T6G 2H1 ABSTRACT This paper describes a program for handling "scope ambiguities" in individual English sentences. The program operates on initial logical translations, generated by a parser/translator, in which "unscoped elements" such as quantifiers, coordinators and negation are left in place to be extracted and positioned by the scoping program. The program produces the set of valid scoped readings, omitting logically redundant readings, and places the readings in an approximate order of preference using a set of domain-independent heuristics. The heuristics are based on information about the lexical type of each operator and on "structural relations" between pairs of operators. The need for such domain-independent heuristics is emphasized; in some cases they can be decisive and in general they will serve as a guide to the use of further heuristics based on domain-specific knowledge and on the context of discourse. The emphasis of this paper is on discussing several of the more problematic aspects of the scoping protocol which wcre encountered during the design of the scoping program. INTRODUCTION Natural languages contain a variety of "logical operators" which interact with each other to give rise to different types of ambiguity. The logical operators recognized by the scoping program include quantifiers, coordinators and negation, which are initially "unscoped" and must therefore be moved into position by the program, and adverbs, predicates and connectives (such as if-then). At the moment, other operators such as tense, aspect and modals are left in place and therefore assume innermost scope. There is some evidence that the handling of the scoping of quantifiers relative to such operators may require special treatment (eg. Fodor 1970; Enc 1981; S aarinen 1983). Three simple examples will illustrate some different types of scope ambiguity and their representation in an informal first order predicate logic, using restrictions on quantifiers and an infix notation for sentential formulas. The meanings of the different interpretations should be clear. For example, (4) may mean that John didn't meet either Jane or Mary (5) or that he didn't meet at least one of them (6). Further examples are given in Hurum & Schubert (1986) and Hurum (1987). Some alternative proposals for representing scope ambiguities are also discussed in the latter. (1) Someone loves everyone (2) (3x:person (Vy:person [x loves y])) (3) (Vy:person (3x:person [x loves y])) (4) John didn't meet Jane or Mary (5) --,[[John met Jane] v [John met Mary]] (6) [--,[John met Jane] v ~[John met Mary]] (7) Someone always comes late (8) (3x:person (always [x comes late])) (9) (always (3x:person [x comes late])) Until quite recently, designers of natural language understanding systems have given little attention to the problem of dealing with scope ambiguities. Two of the earliest attempts to incorporate quantifier scoping into natural language understanding systems in ,an integral way are described in Woods (1978) and Dahl (1979). Some more recent scoping algorithms are presented in McCord (1981), Warren & Pereira …

Content truncated to 500 words