Noun and Adjective Stress

There seem to be sufficient commonalities between the stress patterns of nouns and adjectives that they warrant a single grouping. In disyllabics, the default stress is on the penult. In a 20,000-monomorphemic-word sample reported by Hammond (1999: 194), both disyllabic nouns and adjectives reveal penult stress over 80 percent of the time. More precisely, 81.7 percent (2,986 out of 3,652) of nouns and 81 percent (1,047 out of 1,294) of disyllabic adjectives followed this pattern. Below are some examples from both categories:

The exceptions to the penult rule fall into two groups. The first contains examples with weightless (unstressable) penults, because they have [ə] nuclei, and thus are stressed on the final syllable (ult) by default; for this reason, they might be considered exceptions:

The second group constitutes the real exceptions because they are stressed on the final syllable (ult) despite the fact that they have stressable penults with branching rhymes:

In trisyllabic and longer nouns, we formulate the following: stress penult if stressable (heavy/branching rhyme); if not stressable, then stress the next left heavy syllable. We show this with the following examples:

The words in the leftmost group are stressed on the penult because their penults are stressable (the first five qualify for their long vowel or diphthong nuclei, and the last four because of the closed rhyme). The words in the second trisyllabic group receive their stresses on the antepenult because their penults are not stressable (all with [ə] nuclei). The rightmost group consists of words that have more than three syllables, but the stress rule remains the same. The first word, barracuda, is stressed on the penult, as it contains a stressable penult, [u]. The remaining words (seven with four syllables, and the last one with five syllables) all have unstressable penults ([ə] nuclei) and thus are stressed on the antepenult. As for the frequency of such patterns, Hammond reports that this regularity accounts for over 90 percent of nouns (42 percent or 859 out of 2,074 trisyllabics have penult stress, and 49.5 percent or 1,027 out of 2,074 are antepenult-stressed because of unstressable penult). The exceptions, exemplified by the following, are below 10 percent:

These examples, mostly borrowings from French, retain the original final stresses. Most of these exceptional words have unstressable penults, and thus the rule predicts that the stress would go on the antepenult instead of the ult, a tendency revealed by several native speakers for the last 4–5 words on the list.

Some other trisyllabic exceptions, on the other hand, receive antepenult stress despite the fact that they have stressable penults, as shown in:

There is another group of words, trisyllabic or longer, that deserves attention. The words in this group, overwhelmingly coming from place-names, are stressed on the penult, despite the fact that it is not heavy, as exemplified below:

The stressed penults of these words (a)–(d) do not have branching rhymes (the rhymes are /ɑ/, /ε/, /Λ/, and /æ/ respectively).

There are some other words, again mostly place-names, which also carry the stress on their non-heavy penult:

The penults in words (e) and (f), /ɑ/ and /æ/ respectively, do not constitute heavy syllables, but are stressed, nevertheless. However, these words are somewhat different than the violations observed in words (a)–(d), because words in (e) and (f) do not have any other heavy syllable to the left of the penult (they all have [ə] nuclei). In other words, the stress is on the penult by default.

Trisyllabic adjectives, of which there are far fewer, also show a similar pattern to that of nouns. Hammond reports that over 90 percent of trisyllabic adjectives follow the expected path: 75.3 percent (502 out of 666) with penult stress, and 15 percent (100 out of 666) with antepenult stress because of unstressable penult.