"Irregular" regular paradigms
Posted: Sun Feb 11, 2018 1:14 am
While working on Early Middle Laqar, I have been coming across paradigms that, at first glance, look bizarrely irregular, except that in fact they are derived purely through regular sound change. Two examples of partial paradigms of this sort I have come across are:
"philosopher" (from Proto-Laqar *ɬæwnæmi): sǫme (3sm. sonąmə, 3sf. sonąri, pl. sonąlə, pl. 3sm. sǫmilə, pl. 3sf. sǫmelri, def. sonąde, def. 3sm. sǫmide, def. 3sf. sonąride, def. pl. sǫmeldę, def. pl. 3sm. sonąlodę, def. pl. 3sf. sǫmelridę)
"number" (from Proto-Laqar *kʷʼiːnæ): k'inə (3sm. k'ęnə, 3sf. k'ęri, pl. k'ęlə, pl. 3sm. k'inalə, pl. 3sf. k'inəlri, def. k'ęde, def. 3sm. k'inòde, def. 3sf. k'ęride, def. pl. k'inəldę, def. pl. 3sm. k'ęlodę, def. pl. 3sf. k'inəlridę)
Furthermore, there is no overaching pattern amongst words with regard to what form they take for each of the possible forms, but rather many differing patterns depending all upon the exact shape of the original word in Proto-Laqar. For instance, all imperfective agent nouns derived from perfective verbs whose stems end in long vowels or weak consonants pattern closely with "philosopher". Likewise, all feminine nouns with the shape CV:NV of the original Proto-Laqar stem pattern closely with "number". And as you can see above, there are some parallels between the two paradigms of sǫme and k'inə, in part because they are both feminine nouns.
Has anyone encountered this in any of their conlangs, where they managed to create very significant apparent "irregularity" in what is, strictly speaking, regular sound change?
"philosopher" (from Proto-Laqar *ɬæwnæmi): sǫme (3sm. sonąmə, 3sf. sonąri, pl. sonąlə, pl. 3sm. sǫmilə, pl. 3sf. sǫmelri, def. sonąde, def. 3sm. sǫmide, def. 3sf. sonąride, def. pl. sǫmeldę, def. pl. 3sm. sonąlodę, def. pl. 3sf. sǫmelridę)
"number" (from Proto-Laqar *kʷʼiːnæ): k'inə (3sm. k'ęnə, 3sf. k'ęri, pl. k'ęlə, pl. 3sm. k'inalə, pl. 3sf. k'inəlri, def. k'ęde, def. 3sm. k'inòde, def. 3sf. k'ęride, def. pl. k'inəldę, def. pl. 3sm. k'ęlodę, def. pl. 3sf. k'inəlridę)
Furthermore, there is no overaching pattern amongst words with regard to what form they take for each of the possible forms, but rather many differing patterns depending all upon the exact shape of the original word in Proto-Laqar. For instance, all imperfective agent nouns derived from perfective verbs whose stems end in long vowels or weak consonants pattern closely with "philosopher". Likewise, all feminine nouns with the shape CV:NV of the original Proto-Laqar stem pattern closely with "number". And as you can see above, there are some parallels between the two paradigms of sǫme and k'inə, in part because they are both feminine nouns.
Has anyone encountered this in any of their conlangs, where they managed to create very significant apparent "irregularity" in what is, strictly speaking, regular sound change?