On the nonuniformity of weight-to-stress and

stress preservation effects in English



Joe Pater

McGill University



 

Secondary stress placement in English is affected by syllable weight,
 by the placement of stress in the stem of a derived word, and by
 lexical idiosyncrasy. However, the effects of all of these factors is
 'nonuniform', in the sense of Prince (1993): in some contexts they
 unequivocally determine stress placement, while in others they have
 little, or no effect. To the extent that it has been dealt with, this
 nonuniformity has tremendously complicated prior analyses of English
 stress. At the root of these complications is the tenet of
 inviolability, or full satisfaction. Once full satisfaction is
 replaced by Optimality Theoretic minimal violation, both descriptive
 and explanatory advances are made possible.

 

In section 1 the usual, productive patterns of weight-to-secondary
 stress are accounted for in terms of a small set of ranked
 constraints. Section 2 contains a brief discussion of primary stress
 placement, demonstrating that quantity sensitive primary stress
 placement is compatible with the relatively low rank of the
 Weight-to-Stress constraint that the facts of secondary stress
 placement require. Section 3 examines lexical exceptions, and apparent
 cases of cyclic stress preservation, and argues for an account based
 both on stress preservation, formalized as prosodic faithfulness, and
 on lexically specific constraint ranking. By interspersing prosodic
 faithfulness constraints, and lexically specific constraints into the
 hierarchy established for regular stress, the same principles that
 determine regular prosodification can be used to straightforwardly and
 precisely capture the distribution of lexical, and stem based stress.
 Residual data, and issues, are discussed in appendices.
  

Comments are welcome, and I would be happy to send a hard copy
 of the manuscript if downloading is impossible.