LINGUIST List 11.1959

Sat Sep 16 2000

Disc: Linguistics & Nominalising Languages

Editor for this issue: Karen Milligan <>


  1. Dan Moonhawk Alford, Re: 11.1953, Disc: Linguistics & Nominalising Languages
  2. Zylogy, Re: 11.1953, Disc: Linguistics & Nominalising Languages

Message 1: Re: 11.1953, Disc: Linguistics & Nominalising Languages

Date: Sat, 16 Sep 2000 10:52:42 -0700 (PDT)
From: Dan Moonhawk Alford <>
Subject: Re: 11.1953, Disc: Linguistics & Nominalising Languages

On Sat, 16 Sep 2000, Richard S. Kaminski <> wrote:

> Dear Linguist:

> I have been following what might aptly be called the "nominality pros
> and cons" discussion with great interest from its outset. I do not
> intend to go into great detail in this posting; rather, I should like
> to insert some brief observations and a question.

> Allow me to preface my remarks by stating that my background
> predisposes me to think in terms of nominality, inasmuch as the
> languages that I have studied are indeed quite "nouny," so to speak. 
> That said, I must hasten to add that I have found Moonhawk's
> anti-nominality position, and his various expositions thereof, most
> fascinating, particularly in the context of the current discussion;
> the broadening of perspective that this has afforded me is, to my
> mind, most edifying. This is not to suggest that I am now abandoning
> my Western way of viewing the world on the strength of my having read,
> and digested, these postings; far from it. However, I think that I
> can now resolve a question regarding such "verby" languages as
> Moonhawk mentions which has heretofore stumped me.

> For a long time, whenever this anti-nominality issue arose, there was
> one point at which I would find myself completely stopped. (The fact
> of my not knowing any of these languages certainly didn't help,
> either!) I found myself asking, "How can a language *not* nominalize? 
> After all, doesn't *action* presuppose a thing that *acts*? Doesn't
> *being* presuppose a thing that *exists*?"

Richard, you have accurately characterized here how *I* originally reacted
to my friends' claims -- especially since I'd been studying an Algonkian
language myself! -- and probably that of most people following this
thread. Thanks for jumping in.

> But now I think I get the idea, namely that this is a question of
> *prioritizing* and not an "all-or-nothing" issue. In other words, in
> these "non-nouny" languages under discussion, the *name* of an entity
> is not considered as important as its *attributes,* e.g., what it
> does; how it looks, sounds, smells, tastes and/or feels; its
> relationship to the rest of nature; these are just a few possibilities
> that come readily to my mind.

I will add that animals are usually named for what, out of their database
of Natural knowledge, is unique or rare -- a behavior that others don't
have. So once you've said the behaving, any speaker is free to call forth
an associated noun. And we're talking more kinesthetic than visual as far
as sensory bias is concerned.

> I should like to direct my question to Moonhawk in particular, though
> not to the exclusion of any and every other individual who may have an
> answer. The question is: Is the idea expressed in the latter half of
> the preceding paragraph a reasonably accurate summary of the
> anti-nominality position being offered here?

Except for the fact that I myself have never seen or labelled it as
*anti-nominality* at all, and neither does your description, it is indeed
an accurate summary. So let's see if I can expand carefully on that.
Perhaps we could call it "noun-unnecessary".

"Prioritizing such that the attributes are more important than its name"
- yes. We can think about that prioritizing as something *people* are
doing, perhaps as a social/cultural/cognitive habit in how they use the
structures of their language. I mentioned at one point the Blackfoot
speaker who was unwittingly overlaying English dichotomous thinking onto
Blackfoot, a bleedover effect from being bilingual. We could see that as a
social/cultural/cognitive habit of non-dichotomizing ("just because I say
this one's sacred doesn't mean the others around it aren't sacred too").

Prioritizing doesn't "hate" any "opposite" (as I was accused of "hating
nouns" recently), but is about paying attention to something else a little
more (the dancing instead of the dancers).

Look at Navajo, with at one estimate over 350,000 "words" for GO,
depending on the how, the KIND of going, being described. I think it's
easy to see that the verbs are incorporating something, but it's not
objects or names -- it's manner, the dancing, the way of going, not an
object going per se, though that's how we'd say it, or how Natives might
say it for the linguists' benefit, taking audience into consideration.
(Which, by the way, is itself an issue worthy of discussion.)

Or my Cheyenne duck/rattlesnake example where *Se?Se* means "duck" but
*Se?Se-novotse* means "rattlesnake," with the second morpheme indicating
"goes down into hole." A rattlesnake is clearly not a duck that goes down
in a hole by any stretch of metaphorical or any other thinking. That means
we must go back and reanalyze what the first morpheme means, seeing it as
an attribute and not a name. As soon as we do that we notice that *Se?Se*
is a reduplicated form in Cheyenne, indicating a repeated motion/sound
such as we might call a zigzag rustling -- which is what both a duck and a
rattlesnake do when they are going away from you, and maybe you just catch
the last glimpse to report. You don't report an object, but a process.

And that "going away from" is highly important as well, given that I've
reported Ms. First Rider and Dr. Henderson's claim that there are no (what
we call) pro-nouns in their language at all -- no "me", only coming toward
(me) or going away from (me): more deictic than pronominal.

So if we find a sentence like *Se?se(novotse) navoomo* in Cheyenne, we
find the repeated motion/sound, the verb *-voo-* "see", the combined
prefix/suffix pair *na- -o* indicating "going away from [speaker]", and
finally the animate-gender marker *-m-*.

Since this was brought up by another respondent (how can there be animate
gender markers for nouns if there are no nouns?), this is a good place to
show what they do. Now we've seen how *Se?Se* is a reduplicated rustling
motion/sound -- manner, how -- and it is marked in the verb as an animate
process/relationship with *-m-*, rather than inanimate (dead or inert
body/process/relationship), *-ht-*. And that changes with the
circumstances rather than being more fixed as in English: a tree might be
animate, a branch that falls off inanimate, but a figuring carved from the
branch animate.

If we add now the Blackfoot claim that there's no "language" per se as in
linguists' conceptions going on at all, just 80 roots that get shuffled
around -- creating new words on the fly instead of relying on and teaching
vocabulary words -- it all starts making more sense, I hope. Each root
points to a kind of in-/animate process/relationship with nuances from
what it's combined with. Restated, these roots are a classificatory system
by which the phenomena of the world are noticed and reported.

Thanks to everyone, incuding the Algonkianists who weighed in. for pushing
me to explain myself further. (I know noting of Iroquois.) I would be
happy if any Algonkianists on the list would take this as a serious
alternative interpretation of existing data -- using an indigenous
linguistics lens -- and see if it helps them make more "sense" of anything
in their data. (I know when I thought about that incredible table I was
forced to set up for Cheyenne, with actors and goals, it all just melted
away in insight when I turned that into more deictic information while
writing this.)

warm regards, moonhawk

Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: Re: 11.1953, Disc: Linguistics & Nominalising Languages

Date: Sat, 16 Sep 2000 13:20:03 EDT
From: Zylogy <>
Subject: Re: 11.1953, Disc: Linguistics & Nominalising Languages

Something lacking in the discussion about nominalizing versus verbalizing 
languages thus far is any historical/typological perspective- why languages 
should be skewed one way or another in their discourse preferences for full 
versus grammaticalized nouns and/or verbs.

Note that pronominals and demonstratives are often transparently related to 
each other in form in many of the world's languages. This is usually true of 
non-1st/2nd person, and perhaps then 3rd (and "4th") are the least-marked 
equivalencies. There are, however, quite a few languages in which even 1st 
and 2nd person pronouns relate to distance demonstratives. If these are late 
or rarer developments then we should perhaps posit a hierarchical 
implicational scale. On the other hand perhaps there is a flip-flop based on 
morphosyntactic type. Without detailed survey its impossible to know at the 
present time (unless someone out there has done this already).

In any event such forms, regardless of origin, have higher "headedness" than 
normal lexical nouns. Thus affixational preadaptation, if you will. And I 
won't go into scope/anaphora, but its all there. Highly "verby" languages 
thus sport heavy use of these highly grammaticalized nominals in continuing 
discourse. And as mentioned in the previous posting, in languages such as 
Wakashan, Salishan, Chemakuan, etc., arguments have been put forward that at 
the lexical level, at least, all or most forms have predicative weight. As 
higher level nominals appear to exist at some levels of analysis, we must 
conclude here that "nominal" is a grammaticalized category, for the most 
part. Still there, but shifted upwards in the syntactic hierarchy.

In severely "nouny" languages, like Mongolian, Tungusic, Turkic, various 
Australian and a host of other relatively freshly agglutinative types with 
lots of case, etc., true verbs are the casualties of severe attrition. Ask 
any specialist. Practically all verbs in these languages are built through 
combination of nouns/adjectives (no real distinction here at the lexical 
level between these) or phonosemantically transparent particles with 
auxiliaries. Many of the auxiliaries can stand as verbs in their own right, 
but there aren't that many of them. Some are more grammaticalized than 
others. Still the point is that they are in fact grammaticalized. And therein 
lies the rub.

In both types of skew from neutrality with regard to "nouniness" or 
"verbiness" we see grammaticalized forms taking over the bulk of the usual 
tasks associated with nouns and verbs. Bound pronominals/demonstratives and 
auxiliaries have broad semantics compatible with grammaticalization trains. 
They stand higher in the syntactic hierarchy than their corresponding regular 
lexical equivalents.

And there seems to be an association with morphosyntactic type that cannot be 
simply dismissed as accident. Nichols has pointed to such a skew. Most of the 
languages I'm familiar with that are extremely "verby" to such an extent that 
all roots have predicative force are verb-initial. And most of the languages 
that are extremely "nouny" are verb final. There may also be phonological 
associations- witness the flourishing of vowel harmony in the latter type, 
and perhaps consonant harmony in the former. Have we some sort of symmetry 
game going on here? Something holistic typologists (what few might admit to 
such a conceit) might wish to examine more thoroughly?

And then there are nominal classifiers. Also broad semantics. Also the 
tendency to grammaticalize, perhaps eventually all the way to simple gender 
markers. And perhaps instrument/path terms should be considered here as well. 
Serial verbs? The skew from neutrality may be multidimensional, and real 
languages may thus exhibit a mixture of states synchronically as they 
traverse the state space diachronically.

So what's it all about? Maintainance of discourse continuity, of course, 
focussing on participants, or activities, or qualities, whatever. Ideophones, 
on the other hand, appear to exist to "break" the prevailing continuity. I 
was driving along the road enjoying the sunshine when CRASH a tree came out 
of nowhere and destroyed the passenger side. Interestingly, even ideophones 
may be "grammaticalized"- in Chinook, for instance, verb particles which 
modify relatively "nouny" predicates (perhaps a survival from a prior 
Penutoid "nouny" state) have higher pro-control values than one would expect, 
and are present in much smaller numbers, as compared to ideophone sets in 
other kinds of languages. They resemble, in fact, auxiliaries (even though 
they themselves are bound to a tiny set of performative auxiliaries). 
Examination of other languages with high pro-control ideophones or 
expressives leads me to suspect an association of some sort with ergativity. 
Auxiliaries usually have a definable control value re other verbs- 
"pro-control" in ideophones generally has the opposite polarization in 
construction with these auxiliaries- as if some sort of balance were 
required. I term it pro- as ideophones are not themselves verbs, though 
historically they can evolve into them by lexicalization of the ideophone on 
the one hand, and grammaticalization of the auxiliary in construction, on the 
other. This ain't rocket science.

Ergative languages often utilize antipassive constructions to maintain topic 
continuity, as if the unmarked state of the grammar was shifted over in the 
opposite direction. Which it is, given a variety of typological associations 
of ergativity. So the utilization of high pro-control, auxiliary-like 
ideophones? Perhaps they link otherwise disconnected discourse.

But it all gets back, ultimately, to morphosyntactic type. The entire 
nouny/verby debate is focussed on one axis. I'm not sure that we can really 
understand the polarization between these without first having the bigger 
picture in mind. If the big picture is based on principles of system closure, 
symmetries, and recombination, so much the better. 

Jess Tauber
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue