Hi all,

This is the second installment of a preliminary sketch of the conlang 
that I'm currently most actively working on, tentatively named 
Nenyuvai.  The first installment, which I posted on 23 September, 
covered the general background, overall typology, phonology and 
orthography.  I'm not sure how many more installments there will be, 
because I'm not sure how much detail I'll end up going into.  (I'm 
trying to keep it brief, but that's always hard, especially for me.)

I'd like to start by apologizing for the long delay between 
installments.  My original plan was to post one every week or so, but 
it's turning out to be a bigger job than I thought.  (There's nothing 
like trying to explain something in writing, as succinctly as possible, 
to make you aware of parts of it you don't know yourself, but didn't 
know that you didn't know.)  I also keep getting sidetracked with other 
arcane projects, like trying to understand what's currently known about 
how the ancient Greek musical modes worked, and to use Scala (a freeware 
tuning software package) to hear what they actually sounded like.

IV.  Morphology

IV.A.  Word classes

Nenyuvai has the following word classes and subclasses.  In this list, 
an asterisk indicates an open class.  Some of these subclasses can be 
further subdivided, but these are mostly cross-cutting subdivisions that 
do not lend themselves to this kind of hierarchical listing; therefore I 
am not including them in this listing, but will discuss them in the 
section on the appropriate class or subclass.

- nouns*
--- proper names*
--- common nouns*
- pronouns
- determiners
- quantifiers

- verbs*
--- active verbs*
--- stative verbs*
--- mixed verbs*
- auxiliaries (a.k.a. light verbs)
- copulas

- adverbs
- prepositions
- conjunctions
- interjections (a.k.a. illocutionary particles)
- miscellaneous particles

Nominals and verbals undergo various kinds of inflection, while 
particles are invariant in form.  Note that there is no separate class 
(open or closed) of adjectives, although determiners and quantifiers can 
behave either like pronouns or like attributive adjectives; in general, 
this language uses stative verbs in lieu of predicative adjectives, and 
relative clauses in lieu of attributive adjectives.  Note also that 
adverbs are a closed class; the functions of open-class adverbs (that 
is, adverbs derived from adjectives, in languages that have adjectives) 
are performed either by prepositional phrases, with nominalized stative 
verbs as their objects, or by various kinds of subordinate clauses.

IV.B.  Nominals

IV.B.1.  Nouns

Nouns are inflected for number and state.  There are two numbers 
(singular and plural) and three states (absolute, construct, and 
pronominal).  (I have borrowed the term "state" and the terms for the 
various states from the terminology used in traditional grammars of 
Semitic languages.)  There is a distinction between animate and 
inanimate nouns, as mentioned below, but this distinction is overtly 
marked only on determiners, not on nouns.  There are no case 
distinctions, since the argument alignment is hierarchical; there is a 
distinction between proximate and obviative, but this distinction is 
marked only by word order and context, not morphologically.

The class of nouns can be subclassified in several ways.  These 
subclasses do not form a clear hierarchy, but are (for the most part) 
orthogonal; they cross-cut each other, so that one subclass cannot be 
treated as a subset of another.  These subdivisions are not marked on 
the noun, and so are not considered inflectional categories.  They 
include, but are not necessarily limited to, the following:
- proper names vs. common nouns;
- count nouns vs. mass nouns (uncountables);
- animate vs. inanimate nouns (note: some nouns may be treated as 
animate in some contexts and inanimate in others; this may or may not 
reflect a difference in meaning);
- possessibility classes, based on the distinction between alienable and 
inalienable possession: some nouns can only be possessed alienably; some 
only inalienably; some either way (with a difference in meaning); and 
some cannot be possessed at all;
- number classes: whether, and if so, how, a noun is marked as singular 
or plural.

IV.B.1.a.  Number

There are two subclasses of nouns that do not take number marking: 
proper names, which are inherently singular (in Nenyuvai, you can't say 
"two Anns"; you can only say "two women named Ann"), and "mass nouns" or 
"uncountables" (e.g., _kulya_ "water"), which are semantically neither 
singular nor plural, but which are grammatically treated as singular, in 
that any determiners that refer to them or modify them are singular.

But the great majority of nouns are "count nouns", which are marked for 
number, and which can be grouped into several subclasses according to 
how they form their plurals (or, for a very small but important group of 
count nouns for which the plural is the unmarked form, how they form 
their singulars).

IV.B.1.a.(1)  Count nouns with unmarked singular and marked plural

The majority of nouns mark the plural by total or partial reduplication 
of the first syllable: for example, _gèz_ "book", _gegèz_ "books"; 
_veldimah_ "teacher", _veveldimah_ "teachers".  If the first syllable 
starts with a cluster rather than a single consonant, the second 
consonant of the cluster is dropped in reduplication: _tyèl_ "daughter", 
_tetyèl_ "daughters" (not _*tyetyèl_).  (As you may remember from the 
section on phonology in the previous installment, the only permitted 
syllable-initial clusters consist of an obstruent followed by a liquid 
or glide.)  If the initial consonants of the first two syllables are 
such that they could be combined (in that order) to form a permitted 
syllable-initial cluster, the vowel of the second occurrence of the 
reduplicated first syllable is dropped so that the two consonants do, in 
fact, form such a cluster: _taràh_ "man", _tatràh_ "men" (not _*tataràh_).

Although this pattern is by far the most common form of plural marking, 
there are several other patterns that occur unpredictably; thus the 
plural of a noun must be memorized along with the singular.  Among these 
patterns are the following:
- infixation of _-ha-_  or _-ya-_ before the stressed syllable: _nalín_ 
"woman", _nahalín_ "women"; _atán_ "god", _ayatán_ "gods"; _etín_ 
"goddess", _eyatín_ "goddesses";
- a suffix _-ya_, with a shift of stress to the syllable before the 
suffix (this tends to occur with nouns whose singular already contains 
reduplication): _máma_ "mother", _mamáya_ "mothers";
- this suffix in combination with raising of the vowel in the formerly 
stressed syllable: _tsóma_ "priest", _tsumáya_ "priests"; _tséma_ 
"priestess", _tsiméya_ "priestesses";
- replacement of a long vowel by a diphthong in a stressed final 
syllable: _kadán_ "chief", _kadain_ "chiefs".

[Brief digression into derivational morphology:  As the above examples 
show, Nenyuvai has some pairs of nouns whose meanings differ only in the 
sex of the referent, and whose forms differ only (with a few exceptions) 
in the vowels, with back vowels being associated with the male member of 
the pair and front vowels with the female.  (This is _not_ a "gender" 
distinction in the linguistic sense of the term, because: (a) it applies 
only to a very small subset of nouns; (b) it does not trigger any kind 
of agreement.) The relevance of this to number marking will become clear 

IV.B.1.a.(2)  Count nouns with unmarked plural and marked singular

There are a few nouns for which the plural is the unmarked default, with 
the singular marked by a suffix _-(a)t_.  Most of these are 
gender-neutral kin terms formed by combining two gender-specific nouns 
of the type mentioned in the brief digression above (sometimes with a 
plural suffix, but more often not).  For example:
_máma_ "mother" + _bába_ "father" > _mambabáya_ "parents" (singular 
_mambabáyat_ "parent");
_saráf_ "brother" + _seríf_ "sister" > _safseríf_ "siblings" (sing. 
_safserívat_ "sibling");
_tál_ "son" + _tyèl_ "daughter" > _taltyèl_ "children, offspring" (sing. 
_taltyellat_ "child") (note: this means "child" only in the sense of 
"son or daughter"; there is a totally different noun _tíka_ meaning 
"child" as opposed to "adult").

IV.B.1.a.(3)  Mass nouns and singulatives

The same suffix that is used to form singulars of default-plural nouns 
(see above) can also be used to form singulatives from mass nouns.  For 
example, from _kulya_ "water" we can derive _kulyat_, which, depending 
on the context, might mean "glass of water", "drop of water", or some 
other countable unit of water.  Singulatives are count nouns, with 
plurals formed in the usual way; the plural of _kulyat_ is _kukulyat_.  
I'm not sure yet how productive this process is, i.e., whether 
singulatives can be derived from all mass nouns or only from some.  I do 
know that the meaning of a singulative is not totally predictable from 
the meaning of the corresponding mass noun; for example, the mass noun 
_marén_ means "light" (regarded figuratively as a substance), but its 
derived singulative _marénat_ (plural _mamarénat_) does not mean "ray of 
light, beam of light" as one might expect, but rather "lamp, light 
source, object emitting light".

IV.B.1.b.  State

State is a grammatical category which indicates whether or not a noun is 
inalienably possessed by an immediately following constituent.  
(Alienable possession is shown by a different construction; both kinds 
of possession will be discussed more fully in the section on syntax.)  
If a noun is unpossessed, it is in the absolute state.  If it is 
possessed by another noun, or by a determiner or a free pronoun, or if 
it is modified by a relative clause, it is in the construct state.  
(Relative clauses in Nenyuvai are treated grammatically as inalienable 
possessors of their head nouns.  As I mentioned earlier, they are the 
functional equivalent of attributive adjectives -- instead of "the old 
man", one says literally "the man who is old" -- and therefore occur 
much more frequently than in languages that have adjectives.)  If it is 
possessed by a bound pronoun, it is in the pronominal state.

The absolute state is the unmarked default.  The construct state is 
marked by a suffix _-u_ (with some phonotactic complications that I 
won't go into here).  The pronominal state is marked by the appropriate 
pronominal suffix.  For example:

Absolute state: _karás_ "friend" (depending on context, this can mean "a 
friend" or "the friend")

Construct state: _karázu Yán_ "John's friend"

Pronominal state:
           _karasku_ "my friend"
           _karassi_ "your(sg.) friend"
           _karasta_ "his/her/its(proximate) friend"
           _karazna_ "his/her/its(obviative) friend"
and so on for all the other possible pronominal suffixes.

Because of the phonotactic complications in the way certain nouns form 
their construct state, and because of the unpredictability of the 
plural, the glossary entry for an "ordinary" noun (that is, a 
default-singular count noun) gives three forms: the absolute singular, 
the construct singular, and the absolute plural.  For example:

karás, karázu, kakrás -- friend.

Well, that's it for nouns.  I was going to go on to the morphology of 
other kinds of nominals (pronouns, free and bound; determiners; 
quantifiers), but this is getting rather long even by my standards, so I 
think I'll put those off to the next installment (which hopefully won't 
be as long in coming as this one was).  So, until next time...

-- Tim