A statistical approach to African personal pronouns | ||
Guillaume Segerer & Martine Vanhove - CNRS LLACAN, France | ||
4b. N-m systems in AfricaIn both WALS and MPLA samples, N-m languages are scarce. Note that there is one mistake in WALS concerning Berber Tamazight where 2sg m is a feminine marker, not the personal marker. Which means that the proportion is 3/30 two of which are non-paradigmatic. Our database gives a proportion of 41/738 (6%), only 6 of which are non paradigmatic. 39% of the languages have an N (i.e. n or ɲ) for 1st person in at least one paradigm. 19% of the languages have an m for 2nd person in at least one paradigm. In the 738-lgs sample, if the presence of N in 1st p. and m in 2nd p. are independant from each other, there should be 738*0.39*0.19 = 56 languages showing the N-m pattern. There are actually only 42 lgs showing this pattern, which is 25% less than expected. |
||
< | ALT8 (8th Biennial Conference of the Association for Linguistic Typology), University of California, Berkeley, July 23-26, 2009 | > |