Introduction

This paper discusses several interlocking problems in intonational phonology, all having to do in one way or another with the scaling of tonal targets in utterance fundamental frequency (Fo) contours. Evidence is presented for two broad claims:

1. that downstep and prominence are intrinsically related phenomena of pitch register;

2. that the phonological control of register makes reference to hierarchical metrical structure of the sort that can be independently motivated by facts about rhythm, focus, and so on.

More generally, the paper attempts to clarify the essential differences between the now virtually standard theory of intonational phonology elaborated by Pierrehumbert and her colleagues (Pierrehumbert 1980, 1981; Liberman and Pierrehumbert 1984; Beckman and Pierrehumbert 1986) and the variant of this standard approach developed in my own work since the appearance of Pierrehumbert's thesis (especially Ladd 1983, 1986, 1987). Because the strictly experimental evidence discussed in the second half of the paper is quite limited – work to support some of the other interrelated claims is still in progress or being planned – the discussion of my overall approach is intended to enable the reader to evaluate the limited evidence presented here in its larger context.

I presuppose without comment Pierrehumbert's basic autosegmental premise that F0 contours in English and many other languages can be analyzed as strings of pitch accents, and that pitch accents are composed of one or more tones. (In addition to pitch accents, there are other elements occurring at the ends of various prosodic domains – Pierrehumbert's “phrase accent” and “boundary tone” – but these will not concern us here.)