Theiling Online    Sitemap    Conlang Mailing List HQ   

Exocentric Derivation in Gweydr

From:David J. Peterson <dedalvs@...>
Date:Friday, February 3, 2006, 3:07
In our Word and Paradigm Morphology class, we just read part
of Bochner's thesis, in which he discusses exocentric compounds.
These are compounds in English whose meaning is not the sum
of its parts.  A quick example is "rocky road".  This is the name of
a kind of ice cream, and has nothing to do with either rocks or
roads.  This gave me an idea about a system of derivation for

Gweydr now has five prefixes which look something like what
you have below:


The vowels of the first four can change based on predictable
phonlogical principles, but other than that, the prefixes are invariant.
These prefixes have no meaning at all.  They combine with stems,
however, to form different words.  Here are some examples:

Root: m6ks "berry"
bwEm6ks "blackcurrant"
kj6m6ks "cranberry"
towm6ks "blueberry"
ruwm6ks "redcurrant"
Tijm6ks "gooseberry"

So here you have a set of five prefixes which simply derive different
berries from the root "berry"--not unlike the prefix "cran" in
(or "rasp" in "raspberry"), in fact.  These same prefixes can be
to a stem which doesn't actually occur in isolation: /sine/.  It takes a
suffix to form the word "gemstone", and with the five prefixes, it forms
a class of precious gemstones:

Root: sine (from sineby, "gemstone")
bwezine "ruby"
kj&zine "diamond"
t8wsine "emerald"
riwsine "sapphire"
Tijsine "amethyst"

Application of a given prefix is blocked by a root which begins with
the same letter.  For example, here's the paradigm for /tews/, "nut":

Root: tews "nut"
bwetews "almond"
kj&tews "walnut"
*tewtews (nonexistant)
rywtews "pecan"
Tijtews "chestnut"

And finally, the prefixes can be added to a root that doesn't correspond
to a word at all.  Here the root is /st/:

Root: st
bwEst "stool"
kjast "ladder (long)"
towst "ladder (short)"
ruwst "footrest"
Tijst "rug"

(Note: The word for "foot" is /dZAnu/.)

The point is that if you combine all the /bwe-/ words, or all the
/kja-/ words, and try to find a common meaning amongst them,
you'll come up empty.  Additionally, there are no systematic relations
between, say, the /bwe-/ word and the /kja-/ word in one paradigm
when compared to the /bwe-/ word and the /kja-/ word of a
different paradigm.  However, the words are clearly separable
into prefix and root, and can be grouped by prefix and root.

Finally, though the prefixes in each group can be grouped together
semantically, they aren't by any means the exclusive members of
each group.  For example, /mEsl/, "topaz", could just as easily have
gone in with the gem group, but as it has its own form already,
there was no need.  Additionally, even though other nuts exist
(e.g., the butternut, /pY4ts/), there will never be a */tewtews/.

So, that's it.  Just something to add.  This strategy is kind of halfway
between noun classes and exocentric compounding.  The difference
is that: (a) they're not compounds, and (b) they're nowhere near as
prevalent as noun classes.  It's kind of like the system of English
Latinate prefixes (ab-, ad-, pre-, trans-) and Latinate roots (cur, mit,
fer), except that there's more meaning associated with the Latinate
prefixes than with the Gweydr prefixes.

"A male love inevivi i'ala'i oku i ue pokulu'ume o heki a."
"No eternal reward will forgive us now for wasting the dawn."

-Jim Morrison


Roger Mills <rfmilly@...>