The language's name is <hBaṣvî> which translates to "the words".
The language's name derives from an irregular formation of the noun <baṣvâ>. While its plural is regular (as it is a 5. declension noun) in <baṣvî>, it has a definite prefix <h-> that is otherwise exclusive to second-declension nouns.
The language is very much inflected, and most of its words inflect for some category.
For all the time I've spent on it, I do not have an actual phonology for it; I just went for a roughly Avestan feel.
The tentative consonant inventory is:
<p b t d ṭ ḍ k g q>
<s z θ ð ṣ ẓ x h>
<m n ṇ ŋ>
<v r (l) j w>
<mn* pt* bd*>
The co-articulated consonants, marked with an asterisk, haven't shown up (with the exception of one noun with <mn>) in the notes so far but I did write them out really early on and haven't disproven them since.
The lateral <l> is extremely rare. It occurs only in words where there already is a rhotic <r> and is dissimilative; it isn't a phoneme.
The vowel inventory is more definite:
<a â å e ê ə i î o ô u û>
<ą ą̂ ę ǫ ǫ̂ ų ų̂>
I assume that the circumflex denotes a long vowel.
The vowels come in sets; each short vowel has an equivalent long vowel, as seen here:
Code: Select all
a > â
e > ê
i > î
o > ô
u > û
ą > ą̂
ę > ą̂
ų > ų̂
ǫ > ǫ̂
å > ǫ̂
The schwa <ə> has a special set of equivalents:
Code: Select all
ə > â
jə > jê
və > vû
wə > wû
nə > ną̂
mə > mų̂
You might have noticed I've used only the angled brackets; that is intentional. I do not actually have a consistent phoneme correspondence (or an inconsistent one, really, just none at all) between the actual phonemes and the graphemes. I decided that there are as many phonemes as there are graphemes and that there is a perfect one-to-one correspondence between them, just not what each grapheme corresponds to.
I don't have phonotactics for the language since I went by feeling, but general observations:
- There is no "voicing" assimilation (that is, <sb> is distinct from <zb>, but, surprisingly, there isn't any <zp> or its ilk)
- The sonority hierarchy doesn't play much of a role (that is, <ṣv> is as frequent as <jr>, <mx> and <xm>)
- Some clusters do not seem to appear very much, some such as homorganic plosive+fricative clusters do not appear at all
- Vowels aren't frequently found together, though some such as <əu> and <əą> are found on occasions. These seem to form diphthongs; a hiatus seems to be indicated with an apostrophe (I have only one instance of an apostrophe here, and it's in <θwi'â>, the vocative dual of the fourth person pronoun)
- Geminates are allowed in medial positions; they do not seem to occur in a cluster with other consonants often. I write them as doubled consonants for convenience.
- While initial clusters do not appear to have more than two consonants (except probably in second declension nouns that add a plural <x-> and a definite <h-> prefix), they do not seem to have a limit as to what goes into them (<θm> is just slightly less common than <ṣp>)
- Final clusters also seem to be limited to two consonants.
Huh. This'll be an endeavour.