Journal statistics
How to use the archive
When the cursor changes from an arrow to a pointing finger, it indicates a link that you may click on to select that item.
The blue buttons at the top can be selected to return to the overview or to run a search for specific information.
Most volumes have been split into two parts, each with separate articles. When you select a specific number, all of the articles will be listed with details of the author, date of publication, language and the size of the file that will be downloaded if you choose that option. Just click on the yellow "Download" button to transfer a copy of the article to your own computer or device.
- Language Family: None
- Topic #1: Phonology
- Topic #2: Vowels
Abstract
Vowel contrasts based on tongue root position are common in African languages. Many such languages (“/2IU/ languages”) have a tongue root contrast in high vowels, yielding two sets of phonemic high vowels, /i/, /u/ and /ɪ/, /ʊ/, as in the very common nine-vowel system /i ɪ e ɛ a ɔ o ʊ u/. Many other languages (“/1IU/ languages”) have an [ATR] contrast only in non-high vowels, as in the very common seven-vowel system /i e ɛ a ɔ o u/. The problem of determining, in the course of descriptive fieldwork, which type of vowel system is found in a language has sometimes proven surprisingly difficult in practice. The difficulty has revolved around the high [-ATR] vowels [ɪ], [ʊ], which have frequently been mistaken for other vowels, generally [i], [u] or [e], [o]. Confusion with mid [+ATR] vowels [e], [o] has been especially common, due to the auditory similarity of the two sets (Casali 1995a, 2008, Starwalt 2008). Largely in consequence, a good number of West African /2IU/ languages have been misanalysed as /1IU/ languages (Boyd 2015, Casali 1995a, Dakubu 1997, Rennison 1986, Schuh 1995). While it can sometimes be challenging to distinguish /2IU/ and /1IU/ systems solely on the basis of impressionistic phonetic data, however, the two systems show very striking differences in their typical phonological patterns (Casali 2003, 2008, 2016). These differences can furnish valuable clues to a language’s likely vowel-inventory type in early stages of phonological fieldwork. This paper outlines several specific diagnostic tests based on phonological patterning that can aid in vowel-inventory analysis. The proposed diagnostics rely on straightforward observations about the behavior and distribution of high vowels and relate to general tendencies for the high vowels [i], [u] to be more widely distributed in /1IU/ than in /2IU/ languages and for the same vowels to more readily trigger [+ATR] spreading in the latter.