The binding variables scopes over the entire rule

between verbs, such as abhor (62), and corresponding deverbal adjectives (63), such as abhor-rent. Lexical rules came into fashion in computational lexical semantics in the early 1990s (see Section 4.1 above). Ontological semantics uses the facility of lexical rules for actual massive lex-ical acquisition, always paying attention to the relative effort expended in formulating the rule versus that needed for specifying lexical entries for a class of words manually (see Viegas et al. 1996b, Raskin and Nirenburg 1999). As a result, fewer lexical rules are proposed and those that are, generate numerous entries.

(62)

The lexical entry for abhorrent is generated from that for abhor using the following lexical rule:

Lexical rules overtly put in correspondence two types of lexical entry: that for the source entry and that for the target one. The binding of variables scopes over the entire rule, both its left-hand side (lhs) and the right-hand side (rhs). The above rule establishes that the semantics of abhor and abhorrent is identical (this is not always the case; see the example of criticize/critical below) but that the syntactic dependency changes from the verb to the adjective, as the direct object of the former becomes the head that the adjective modifies. The expression adj($var0) stands for the adjective whose entry is generated by the rule. In the lexicon entry for the verb, an additional zone, LR, will be created, in which each lexical rule applicable to this verb is listed with the string that is the lexeme of the target entry. A practical consideration for the economy of acquisition effort is whether it is preferable to populate the LR zone of a lexical entry or immediately create the target entry or entries.

Page 278