Advances in Role and Reference Grammar

424 MARY L. NUNES

theory, or the semantically-(un)restricted GFs of LFG, can be applied to

the nominal to effectively articulate the complex, but principled syntactic

relations which exist within the verbally-derived English noun phrase.

Notes

* This paper is a revised and abbreviated version of my UC Davis M.A. thesis

(Nunes 1990). I would like to thank David Wilkins for his invaluable comments,

which greatly facilitated the development of this analysis. I am particularly in­

debted to Robert D. Van Valin, Jr., not only for providing the theoretical

framework with which to analyze the English nominal, but for so willingly and

repeatedly discussing, reading, and insightfully commenting on the analysis. Most

of all, I am grateful to my husband, Jim, whose boundless patience and unfailing

encouragement are the cornerstones of this work.

1. For reviews of analyses in which such assumptions are made, see Chapters 2 and
   3 of Nunes (1990).


2. See Andrews (1985) for a definitional discussion distinguishing among the notions
   "grammatical relations," "grammatical functions," and "semantic roles."


3. See Comrie (1976) for a discussion of how languages vary in the degree to which
   clausal-vs.-NP syntax reflects the "action nominal" status of constructions, where
   Comrie's "action nominal" includes both Ross's "action nominals" and what will
   be identified in the current work as derived "process nominals."


4. In terms of the RRG set of thematic relations, the "object being transferred"
   would be identified as the theme argument of the V/vN; the "recipient" would be
   identified as the locative-goal.


5. Levi's examples have been modified here to the extent that her complex nominal
   [CN] contexts have been deleted. This does not interfere with the intent of her
   examples, as she wishes to exemplify the types of nominalizations which produce
   vNs, "whether or not these forms surface as component parts of CN structures"
   (1978: 167). The examples included in the current discussion are based on exam­
   ples from pages 168, 169, and 173 of Levi's work.


6. A performance object is something that comes into being by virtue of the comple­
   tion of some performance or activity. See 2.3.4.2 for discussion.


7. Actually, in light of the RRG analysis presented below, wherein accumulated
   action vNs are distinguished from result vNs, the term "non-process" more accu­
   rately identifies the type of vN which Grimshaw identifies as "result." Because
   Grimshaw does not make the same RRG distinction, her "result nominal" sub­
   sumes both types of nominals, and the tests apply to both. It should also be noted
   that (6) and (7) do not reflect Grimshaw's presentational format, but are based on
   her discussion and examples from the pages cited before each figure.