Document Type


Publication Date

January 2009

Journal Title

Interfaces in Linguistics: New Research Perspectives


Partee (1991) challenged the significance of the observation that certain adverbs (e.g., only) reliably associate with phonologically prominent words to truth‐conditional effect, noting second occurrence (i.e., repeated or given) focus (SOF) appears to lack a phonological realization. Rooth (1996), Bartels (2004), Beaver et al. (2004), Jaeger (2004), and Fry and Ishihara (2005) argued that, while not intonationally prominent, an SOF word can be marked by increased duration and/or increased rms intensity. An acoustic study of verb‐noun homophone pairs is reported. Three sophisticated speakers uttered five repetitions of the targets, embedded in discourses, in first occurrence (FOF), SOF, and unfocused (NF) contexts, both pre‐ and postnuclearly. Syntagmatic comparison shows a durational effect for focus and ordering in the FOF/NF case, yet only an effect of ordering in SOF/NF; similarly, an effect of spectral balance (H1‐H3) was found for FOF/NF but none for SOF/NF. Results suggest, contrary to previous assumptions, that simple durational difference is not a straightforward cue of prominence. Sluijter and van Heuven (1996) and Campbell and Beckman (1997) have shown spectral balance calculated by various means is a cue to focus; its absence in these data raises doubts that prominence is reliably produced (let alone perceived) on SOF words.