Gender and Noun Class

Mon May 13 08:05:20 UTC 2013

Dear all,

With the mention of an argument of which gender and noun class are 
viewed as opposite ends of a continuum, we've now an exceedingly broad 
response to my question, suggesting that there was indeed merit in 
asking it in the first place.

Unfortunately, I don't think we're any further in answering my question. 
  Most respondents appeared to have an intuitive idea of what gender and 
noun class mean, but I'm afraid they didn't always match each other very 
well, and it's not a matter of knowing the subject poorly.. more than 
one respondent has written a book on the topic!

One thing I did notice that most respondents did somehow feel that 
semantics are at the core of gender systems in at least SOME way, 
whether it be animacy, biological sex, or some other means of 
classification, and even if it's only a small core of words in which 
they are semantically differentiated.

Noun class would appear, then, to be more neutral in general (I'm not 
sure people would feel that semantic correlations are as important in 
noun classes), but I'm afraid that from an African perspective this one 
remains more problematic than it might in general due to the Bantuist 
traditions.  And despite the fact that noun class has NOT been used to 
refer to Dutch or Scandinavian languages, I've often felt that the 
2-gender system where masculine and feminine have collapsed are good 
examples in which sex is no longer a distinguishing factor 
synchronically, but the term persists due to historical/comparative 
reasons.  So noun class would seem appropriate for these systems, 
despite its actual lack of use (unless you go by the criteria of number 
of categories as the difference between noun class and gender).

As I looked through some of the specific suggestions people made on what 
to read, I noticed that a lot of you use "gender/noun class" even in 
your own writings, again suggesting that while the two ideas go hand in 
hand, there isn't a neutral word or phrase in reference to a nominal 
categorization system.

(Grev, agreement/inflectional class work to a point and are indeed 
things I've thought of, but unfortunately both inflection and agreement 
are utilized in Uduk so picking one would be a bit arbitrary).

Martin, I noticed that you seem to have some skills in coining new 
terms, and this is NOT something that comes easy to all of us!  It's not 
simply a matter of courage; sometimes it's also very challenging, 
particular for the younger linguists who have not studied Latin. Perhaps 
you or someone else with more terminology-coining abilities could come 
up with a neutral term which simply refers to a system of nominal 
categorization? It doesn't have to refer to HOW the categories work in 
practice, how many there are, how you distinguish them, why they exist, 
or anything other than the fact that there ARE categories.  These other 
ideas are all factors which can be dealt with with more specific terms 
such as gender, agreement class, etc.

I don't think we can fix all of the issues regarding cross-linguistic 
comparison vs. individual descriptions, but I believe this particular 
issue has a relatively simple (and non politically-charged) solution 
compared to most. It's simply that we lack a neutral term, and we need 
to make one. And no matter what ends up happening with Uduk (even if we 
do solve this issue in another way), I think creating a term for this 
concept might still be a good idea.

All the best,

Don

PS: This is just an addendum for those of you whose curiosity was piqued 
regarding the specifics of the Uduk system, some of which I had 
deliberately left off earlier.  For further questions/comments about 
Uduk specifics it's probably best to contact me by email off-list.

There are two classes, for simplicity we can call the masculine and 
feminine despite a lack of semantic correlation.  As case marking and 
gender marking varies depending on word order, I discuss all of these 
together to some extent.  Uduk is a split system in case marking in 
which only post-verbal arguments can get marking; all pre-verbal 
arguments are neutralized. What the system actually IS has been 
problematic to figure out, and I've gone back and forth in my 
description between split marked nominative and split ergative.

Masculine nouns generally are not marked, although some irregular 
plurals of animate/humans do receive a marker ī-, which puts them into 
masculine (no matter what their original category was); this is not 
regular however, and masculine is mostly marked by the absence of 
anything else. Post-verbal masculine A/S are marked with ā (or sometimes 
à due to phonological reasons), which phonologically attaches to the end 
of the verb but morphologically is part of the noun.  It has been 
somewhat problematic to decide where to write it.

Feminine nouns generally are marked with a proclitic à- (or à= if you 
prefer), but a post-verbal feminine A/S is marked with mā (or mà for the 
same phonological reasons as with ā/à).  Feminine O trigger different 
verb conjugations compared to with masculine O, so despite an occasional 
overlap between post-verbal O and post-verbal S/A marking on the noun, 
there is no chance of ambiguity due to different verbal paradigms (verbs 
also conjugate differently depending on whether S/A is pre- or 
post-verbal). Feminine nouns in isolation do not require the à- marker 
(it's optionally used), but do require it when in clauses/phrases and 
there's not another marker replacing it.

Post-nominal adjectives (one of two categories of adjectives, the first 
of which is more nominal and is somewhat head-like, the second of which 
is the main category of adjectives and mostly comes from stative verbs) 
always agree with the head noun, and each adjective modifying a feminine 
noun has a proclitic à- attached to it.  Adjectival modifiers of 
post-verbal S/A also have a proclitic à-.  Adjectives on masculine nouns 
do not have any markers on them in any situation.

Conjunctions, prepositions, and many other grammatical words do have 
alternations depending on the following noun; if a feminine noun would 
follow, the à- proclitic replaces the final vowel of the grammatical 
word (dhàlì, 'and' +masculine; dhàlà, 'and' +feminine).  This may be 
phonological rather than grammatical, however, and requires further 
testing. Genitive constructions have different forms for masculine or 
feminine heads (gì vs. mā/mà) but not for dependents.

Areas where I can make generalizations on the semantics of the 
categories are limited.

Singular personal names are always feminine, and plural personal names 
(e.g. īDòn: Don and the person/people he is associated with) are always 
masculine. Borrow words from Arabic generally are put into the feminine 
category, suggesting it is more of a default than masculine (despite 
being more marked morphologically).  Size, animacy, biological sex, are 
not factors that I can discern in the system.

 From what I've seen, this system seems to be most useful in 
distinguishing arguments, particularly with all of the word order 
variations, but I can't say for sure yet.  Koman languages are a small 
family with very little descriptive workin general, so I canʼt do 
comparative or historical work at this point very easily.  Perhaps in 
the future I'll be able to see the diachronic connection to other gender 
systems in Nilo-Saharan languages, once more data appears.

On 05/11/2013 04:27 PM, David Beck wrote:
> Personally, I've always favoured using "noun class" as the general
> term and "gender" a for type of noun class, for many of the reasons
> given above. However, for those interested, Igor Mel’čuk has a
> detailed and very informative discussion of the issue in his Aspects
> of the Theory of Morphology where he makes the case for the two being
> at opposite ends of a continuum. It might help Don make up his mind.
>
> David
>
> ================================
>
> David Beck, Professor 4-63 Assiniboia Hall Department of Linguistics
> University of Alberta Edmonton, AB  T6G 2E7 Canada
>
> Phone: (780) 492-0807 FAX: (780) 492-0806
>
> http://www.ualberta.ca/~dbeck/
> http://www.arts.ualberta.ca/~totonaco/
>
>

-- 
Don Killian
Researcher in African Linguistics
Department of Modern Languages
PL 24 (Unioninkatu 40)
FI-00014 University of Helsinki
+358 (0)44 5016437