Re: Ideas for deriving verbs from nouns

Amanda Babcock sikayal:

> Instead I seem to have started inventing a completely new language, with, er,
> none of the above as such.  Oops :)
Heh, I have that problem, too.  Oh, well.  I figure that all of my lang
sketches will sit around until I need them for something, since all of my
langs thus far are someplace incorporateable into my fictional world of
Aratasa.

> Anyway, I'm getting stuck on deriving verbs from nouns.  A little background:
> I've always been obsessed with the idea of a language where either all nouns
> are derived from verbs, or where all verbs are derived from nouns.  I tried
> the former in high school, quickly giving up because it seemed so unrealistic;
> now I'm trying the latter.
My true love conlang Yivríndil derives all verbs from nouns; my friend
Brett made a language where all sentences are composed entirely of
verb--*entirely* of verbs.  There were no nouns or adjectives, merely
noun-like and adjective-like clauses, all coordinated with an elaborate
system of subordination.  (Brett's a brilliant conlanger and got me
started on it, but I haven't been able to convince him to join this
group.)

> Ordinarily, this wouldn't be quite the problem that I'm finding it to be,
> but I also got hung up on making it a trigger system.  I liked Kristian
Erm, since know nothing about triggers I can't help you with that.

> Therein lies the rub.  I suspect that my problem is that I'm trying to
> verbify nouns in the lexicon, and nominalize verbs in the grammar, and
> tripping over myself coming and going.
That does seem to be quite a problem, but as I said, I can't help you
there.

> However, upon further reflection, I can't justify to myself why the noun
> should be "gift" and not "giver", "recipient" or one of the other roles.
> Or "giving", in which case I'd seriously have to consider whether I'm
> inventing nouns at all or just verbs whose infinitives can be used as nouns.
OK.  As I said, my main conlang Yivríndil works in exactly this way, and I
wrestled with exactly this problem.  Eventually I came to the conclusion
that I was trying to have a one-to-one correspondence between nouns and
verbs, and this simply isn't possible.  After much consternation I
realized that I would have to derive nouns with no direct verb correlates,
and derive verbs with no noun correlates.  Thus, when choosing a noun to
serve as the root for a particular verb, I just pick one according to my
current tastes.  The resulting system is messy and irregular and very
natlangish.  Here's an example with the word "giver, give" and related
derivatives.

The basic word is <rilda>, meaning "giver."  The choice of the agent as
the root instead of the object "gift" or target "recipient" is entirely
arbitrary, though I tend to prefer using the agent or the object.  The
basic verb is <rildaya> meaning "to give."

The object deverbal is formed with the suffix -m, attatched directly to
the noun.  Thus:

<rildam> = gift

In this system it makes perfect sense to form an object-focused deverbal
directly from the noun, since all verbs are derivatives of nouns.  But
now, why should <rildaya> have the basic meaning of "to give", and not the
verb derived from <rildam>, e.g. <rildamya>.  There is no systematic
reason--it just happens that <rilda> is the root and the verb comes from
the root.  With other nouns the root happens to be the object or an
abstract noun--there is no real reason, and there cannot be any real
justification for one over the other.  The best you can hope for is to
make the system regular--nouns always serve as the subjects of the verbs
they form, for example.  I have not chosen to do this completely
regularly, though in Yivríndil *most* nouns serve as the subjects of the
verbs they form.

As for the lack of correspondence between nouns and verbs: I can form
nouns <rildam> "gift" and <rildasí> "giving", but there are no verbs
derived from those nouns--they would either be nonsensical or synonymous
with <rildaya>, the verb derived from <rilda>.

And I can form the verb <tairildaya> meaning "to give back" w/ the
reciprocal prefix tai-, but that verb doesn't have any direct noun
correlate.  If you simply lop off the verbal ending you get *<tairilda>,
which isn't a word.  However, you *can* form an abstract noun by removing
the verbal ending and adding the abstract noun ending: <tairildasí>
"giving back".  And if you want you can add an agentive ending to form
<tairildasín> "one who gives back".

As I said, it's messy and natlangish.

> The trigger system invites me to consider all the arguments of the verb as
> equally valid, but I do plan to derive new nouns from the verb derived from
> the core noun, so I have to pick one to be the root and derive the others.
> Figuring out which one to use is the problem.
Restating what I said above, you'll have to be arbitrary.

> In English, it seems like in verbs with a focus role, the focus exemplifies
> the verb more than any of the other roles do.  Is this true of other
> languages?  Is it a reasonable way to go about things?
I think so.  You also will have problems with transitive/intransitive,
probably--what do you do with nouns like "beauty"?  Is the verb "to be
beautiful" or "to become beautiful" or "to make beautiful"?  In Yivríndil
I've decided that the verb means "to be beautiful," with a derivative that
means "to make beautiful."  But if the verb of "beauty" means "to
be beautiful", then why shouldn't the verb of "giver" mean "to be
a giver/ to be generous"?  In Yivríndil there actually *is* a
verbal derivative of "giver" which means "to be a giver", which
was my solution.  However, you'll have to figure those things out for
yourself.

The Yivríndil system actually has to distinguish three different classes
of nouns to handle this properly:

1) Nouns which denote agents or objects and which form transitive verbs
with the full range of possible derivatives.  In Yivríndil these
derivatives include simple, causative, and copular (copular = "to be X",
as in the "to be a giver" mentioned above.)

2) Nouns which indicate emotions and form transitive verbs.  These nouns
form only two verbal derivatives: "to be X-ish" and "to cause to be
X-ish."  For example, <thol> "fear" forms the verb <thoyya> "to be
fearful" = "to fear", and the verb <tholahya> "to cause to fear."

3) Nouns which indicate states and form intransitive verbs.  These nouns
form two verbal derivatives: "to be X-ish" and "to cause to be X-ish".
(The second one is obviously transitive, but the first, which is
considered more basic, is intransitive).  For example <hara> "beauty"
forms the verbs <haraya> "to be beautiful" and <harahya> "to make
beautiful."

> Finally, can anybody suggest pointers with examples (conlang or natlang)
> on various systems of deriving verbs from nouns?
Well, I showed you quite a bit of the Yivríndil system.  I hope it was
helpful!


Jesse S. Bangs jaspax@u.washington.edu

"If you look at a thing nine hundred and ninety-nine times, you are
perfectly safe; if you look at it the thousandth time, you are in
frightful danger of seeing it for the first time."
--G.K. Chesterton

Replies