Hebrew Stress: Can’t you hear those trochees?

This paper makes the claim that intonation is best understood as a part of a language’s prosodic system, specifically its stress feet.

Previous research (Bat El 1993) has established that Hebrew has final stress by default. Nouns can have lexically marked stress, but unmarked nouns are assigned final stress. Verbs roots invariably have final stress.

Analyzing Hebrew in terms of foot structure, however, has proved to be highly debatable. Suggestions include unbounded right-headed feet (Bat El 1993), iambs (Landau 1998), trochees (Graf 1999) and a grammar that has both iambs and trochees (Ussishkin 2000).

In this work, I supply previously unavailable data from the intonational patterns of the language. In Hebrew, the pronunciation of a stressed syllable involves a long vowel on the stressed syllable, and a High tone. The exact location of the High tone is determined by the following principles:

• High tones are shifted one syllable to the right of the stressed syllable
• High tones are not shifted onto a phrase-final syllable (extra-tonality)

When you take words in isolation, these two principles put a High tone on the stressed syllable with final or penult stress, and one syllable after the stress with antepenult or preantepenult stress. Thus: (underline marks stress)

H H H H

At the phrasal level, two more principles are observed:

• A High tone from one stressed syllable never shifts to a following adjacent stressed syllable
• While High tones always shift inside a word, High tones don’t always cross over to another word.

An Autosegmental analysis of the facts is shown to have limited descriptive power and little theoretical insight. An analysis in terms of tonal domains (Kisseberth 1994), however, is shown to be superior. All the above tonal phenomena are analyzed in terms of left-headed binary domains.

The proposed domains are then shown to be nothing else than the metrical feet of the language, i.e. trochees. Now the interaction of tone and word structure is understood as a part of the prosodic structure of the language.

This analysis further suggests that footing can be done at the phrasal level, or across words – a possibility normally overlooked in the literature.

References: