An Explanation for Minimal Attachment and Right Association

This paper describes the principles of RightAssociation and Minimal Attachment and explains how the Sausage Machine and ATN decribe these principles. It is then shown that these two models cannot explain these principles. It is then shown that a production system grammar can both describe these principles as well as suggest why they must be true. A. The Sausage _Machine Lyn Frazier and Janet Fodor (FF) in [II have proposed a two stage model of the Human Sentence Parsing Mechanism (HSPM), called the Sausage Machine (SM). FF proposed that the syntactic analysis of sentences by hearers or readers is performed in two stages. The first stage combines words in phrasal nodes as they are received. They call this the "Preliminary Phrase Packager" (PPP) or the ItSausage Machine". The second stage combines these phrases into sentences. This stage is called the "Sentence Structure Supervisortl (SSS). The main principles of Frazier and Fodor are these: 1) Right Association... an ambiguous constituent should be "attached into the phrase marker as a right sister to existing constituents and as low in the tree as possible" 16 p.21111 Cl p. 2941 2) Minimal Attachment,..an ambiguous item "is to be attached into the phrase marker with the fewest possible number of non-terminal nodes linking it with the nodes already present" Cl p. 3201 *This paper describes work done in the Department of Artificial Intelligence and School of Epistemics, Edinburgh University, Scotland Right Association suggests that "terminal symbols optimally associate to the lowest non-terminal node." This predicts the preferred interpretation of: [ll Tom said that Bill had taken the cleaning out yesterday. [21 Joe called the friend who smashed his new car up. [3l John read the note, the memo and the letter to Mary. [41 The girl took the job that was attractive. (from Cl p. 2971) In each of these sentences, the preference is to attach the final modifier to the lowest right node, just as their principle predicts. This principle also predicts the difficulty in the following