Notes On Adding To The Láadan Vocabulary
by Suzette Haden Elgin
Now that the Láadan Working Group has begun putting substantial
amounts of material for the language on the Internet, more and more
proposed new words and morphemes are arriving. It seems, therefore,
that it might be useful to have a brief overview available about adding
new items to the vocabulary. I'll do my best to be brief without being
obscure, and to define my terms as I go along. I'll try not to slip
into LinguistSpeak (LS), but will provide the more common technical
terms [in square brackets] so that they'll be familiar for those who
may want to read further on the topic.
General Information
All living human languages have methods for adding new items [LS:
neologisms] to their vocabularies [LS: lexicons]. Some of these methods
are systematic and can be easily described; however, there is no rule
forbidding native speakers of languages to simply make up new items
from scratch. That is, no rule would prevent me from adding the new
word "kappid" to English to mean "upper surface of the entire left
thumb." Making that word succeed -- so that people would say it and
write it, and it would be added to the dictionary -- is an entirely
different matter. We don't know much about why a given newly-coined
item does or doesn't "make it" in a language. A few things we do know
are.....
1. It helps if the item gets introduced in a movie or novel
that's a smash hit and has a huge marketing budget behind it, or is
introduced by an organization of scientists, or some such thing.
2. Items that violate the grammar rules of the language -- for
any part of the grammar -- are unlikely to succeed. Consider my earlier
example, "kappid." If I tried to make that "mkappid" instead, native
speakers of English would immediately reject it, because the part of
English grammar that determines how the language sounds [LS: English
phonology] doesn't allow words to start with an M sound followed
by a K sound. M and K are both meaningful English sounds [LS: phonemes]
and either one of them could be first in a word, but "mk" at the
beginning of a word is unacceptable.
The Láadan sound system has a rule that says there can never be
a sequence of two identical vowel sounds inside a morpheme. (Note:
"Working" is just one word, but it contains two morphemes: "work,"
which is a morpheme that can stand alone, and "-ing," which is a
morpheme that can't stand alone.) This means that if there were native
speakers of Láadan, and someone tried to introduce "maath" as a
new morpheme, it would be rejected for violating that rule. Any time
two identical vowels would occur together, one of them has to have tone
(in Láadan, a higher pitch, indicated by an accent marker). So,
either "máath" or "maáth"would be acceptable, but just
plain "maath" is unacceptable in Láadan just as "mkappid" is
unacceptable in English.
(Note: If you are e-mailing a proposed new Láadan item and don't
have convenient access to the accent mark, just substitute a capital
letter for the vowel-plus-tone. In such a situation,
Láadan would be written as "LAadan.")
3. It's much harder for a proposed new morpheme to succeed if
it's part of a "closed class" of morphemes in a language. A proposed
new English pronoun, for example, or a proposed new English tense
ending, will inevitably face an uphill fight.
Because Láadan is a new language with no native speakers its
grammar isn't firmly set in the fashion that the grammar of English is,
and many of the new items that have been proposed for it have been part
of word-sets like the pronouns, the Speech Act morphemes, and the like.
That's still possible for Láadan in a way that it's not possible
for English or Cherokee or Japanese or Sign -- the boundaries of
"closed classes" haven't been established by centuries of speech and
writing and signing --but it will nevertheless be harder to do than
just adding a new noun or verb, and should be done with care. Once a
class of words becomes closed, the language is essentially stuck with
it; if that turns out to be awkward -- the way not having a
gender-neutral third person pronoun for English is awkward -- it's not
easy to fix.
How the Original Láadan Dictionary was Constructed
My goal when I constructed Láadan was that it should be as easy
to pronounce as possible -- and its pronunciation as easy to understand
as possible -- no matter what the native language of the learner might
be. Roughly speaking, the easiest linguistic structure for achieving
that goal is sequences in which consonant sounds alternate with vowel
sounds. How long or short the words and morphemes are isn't
particularly important; some languages have very short words, some have
words that are as long as a sizable English sentence. But the consonant
sound/vowel sound alternation really matters. I therefore used a number
of different strategies to preserve that structure as I put the
language together.
[
Note: The reason I keep saying "consonant sound" and "vowel sound" is
because English sometimes uses more than one consonant to write a
single consonant sound and more than one vowel to write a single vowel
sound. Linguists identify a phoneme -- a single meaningful sound -- by
writing it between slashes -- like /m/ and /sh/ -- so that it won't be
mixed up with the letters of the English alphabet.]
Another goal I had for Láadan was that it should be as easy as
possible to figure out what a particular word or morpheme means just by
looking at it [LS: that the language should have transparent
morphology]. I wanted the language to work like a Tinkertoy® set works,
so that people could take the pieces and fit them together easily to
make larger forms. For example: the Láadan word for "bee" is
"zhomid"; that word is made from "zho" -- the Láadan word for
"sound," and "mid" -- the Láadan word for "creature." The
meaning is transparent from the word's parts.
First, however, I had to construct the most basic elements of the
language -- the words/morphemes that are called "roots" and can't be
taken apart into smaller meaningful pieces. When linguists begin
working with a language for which no grammar or dictionary is
available, they ordinarily start with a set of roughly 100 very basic
words [LS: Swadesh list] made up of items like "eat" and "sleep" and
"food." I followed that practice, and began by constructing a core
vocabulary of those basic words; when I had those done I began adding
additional roots that I felt were needed. Sometimes I can explain to
some extent how I chose a particular shape for one of those words; much
of the time I can't.
For example... I can explain that I chose "oódóo" for
"bridge" because when pronounced its tune makes the shape of a humpback
bridge. I can explain that I chose "rul" for "cat" because the purring
of a cat sounds to me like "rulrulrulrul..." But the choice of "ana"
for "food" and "ina" for "sleep" was arbitrary; I have no explanation
for those choices other than that I tried to give them a shape that
could easily be combined with other morphemes. For any constructed
language that isn't based on some existing language, the hardest part
will always be putting together the inventory of roots.
The "Tinkertoy®" strategy leads naturally to sets of words that people
can easily recognize, even if they can't be certain of the exact
meaning. When you see a Láadan word that includes "mid" as one
of its morphemes, it should be the case that you can assume that the
word is the name of an animal, even if you can't be sure precisely
which one. (The word for "cat," because it's from the core vocabulary
and is a root word -- and all other animal names in the core vocabulary
-- will be an exception; the assumption is that people will learn the
core vocabulary first and then move on from there.) So:
"dithemid" is "cow"-- "mid" plus "dith," which means "voice"; "lanemid"
is "dog" -- "mid" plus "lan," which means "friend."
Just as Láadan doesn't allow two identical vowel sounds in a row
inside a single morpheme, it forbids any two consonant sounds or two
vowel sounds in a row both inside a morpheme and when you put morphemes
together. (This is consistent with the strategy of trying to maintain
consonant sound/vowel sound alternation at all times.) But obviously,
when you start combining morphemes you're going to run into situations
where you do have two consonant or vowel sounds in a row. Láadan
has three rules that take care of this problem:
1. When adding one morpheme to another would give you a sequence of two
vowel sounds, insert an H sound [LS: /h/] to prevent that.
2. When adding one morpheme to another would give you a sequence of two
consonant sounds, insert an E sound (the vowel sound in English "bed")
to prevent that.
3. When adding one morpheme to another would give you a sequence of two
_identical_ consonant sounds, you have two choices: (a) Follow Rule 2
and insert the E; or (b) drop one of the two identical consonant
sounds.
[1]
For example, "dom"/"remember" combined with "mid"/"creature" to mean
"elephant" would yield "dommid," which is not allowed. One choice, by
Rule 3(a), is "domemid"; that choice was rejected. The other choice, by
Rule 3(b), is to drop one of the two Ms, which yields "domid"; that is
the choice that was made. Either choice would have been acceptable.
The sound system of Láadan won't allow either "dithmid" for
"cow" or "lanmid" for "dog"; instead, they become "dithemid" and
"lanemid." Similarly, when I added the morpheme "-á" (which
means "one who does") to "bedi"(which means "learn") to make the words
"student" and "learner," I couldn't just make that word "bediá";
it had to become "bedihá." If I wanted to put the morpheme "du-"
(which means "try to") at the beginning of a verb that starts with a
vowel sound, I had to insert an H sound; so, "try to sell" (from the
verb "eb") had to become "duheb."
This word-building process can result in some very long words. If my
goal for Láadan had been that it should be composed as much as
possible of short words, that would be a problem. If Láadan were
a natural language, its speakers might decide over time that they
didn't like long words, and they might start shortening them. That sort
of thing happens in languages. However, there's a reasonable
probability that the speakers of the language would try to preserve the
two basic goals -- ease of pronunciation based on consonant sound/vowel
sound alternation, and ease of understanding based on the construction
of words from already known morphemes. I don't claim that that's
probable because it's "better" or "more logical" or "more esthetically
pleasing" or anything of that kind. It's probable because when I
constructed the language I built in many rules, at many levels of the
grammar, that all work together to maintain those two goals.
Those rules could all be changed by native speakers, certainly, over
generations; that happens with languages too. The end result would be a
perfectly fine language, but it would be a very
different language.
Additional Miscellaneous Word-formation Rules and/or Constraints
1. There are four Láadan sounds that can't be used as the final
consonant in a word or morpheme: Y, W, H, and R [LS: /y/, /w/, /h/,
/r/]. This isn't an arbitrary rule; it's part of the set of rules that
work to maintain the consonant/vowel alternation in the language
mentioned above. Although it's traditional to put those four sounds
into the class of consonants for the English alphabet, none of them is
-- in linguistics terms -- a true consonant.
2. I've had many queries about the set of morphemes that contain the
sequence BR, which includes "bre... ébre" ("if...then"), "bre"
("layer"), "bremeda" ("onion"), "bróo" ("because"), and the
three Repetition Morphemes "brada, bradan, bradá." Since
Láadan is said to have a rule forbidding consonant clusters,
this set should not exist -- but there it is. The explanation is very
simple: I made a mistake during the construction of the language and
added these items, which most certainly do violate that rule, and I
apologize.
There are several possible ways of handling this. [Deleting the
anomalous items isn't one of them, because languages don't work that
way; anomalies may die out over time because native speakers fail to
use them, but "legislative" attempts to stamp out particular morphemes,
even for very good reasons, are always a waste of time.] The
possibilities are:
a. Say that these forms are a historical accident -- which is true,
that they should be treated as exceptions to the consonant cluster
rule, and that no more items containing BR should be added.
b. Rewrite the rule about consonant clusters to say that the only
consonant cluster Láadan allows is BR. This would be simple, but
it would open the vocabulary to many more forms containing BR.
c. Rewrite the rule about consonant clusters to say that Láadan
allows no clusters of
true
consonants, rather than just no clusters of consonants. R (along with H
and W and Y) is not a true consonant, which would mean that BR is not a
violation of the rule. However, doing this would suddenly open up the
vocabulary to a large number of other sequences that would seriously
complicate the sound system. It would be better not to do this.
My personal preference is for alternative (a), because it is simplest
and introduces the fewest complications.
3. The consonant LH is a consonant like any other consonant, and it's
entirely legitimate to use it to add new words. However, it's also a
morpheme that has as its specific purpose the semantic information that
the item it occurs in has a negative [LS: pejorative] meaning. To use
it in any other way would create misleading and ambiguous items.
For example, the word for "apple" is "doyu"; one way to refer to a
rotten apple, or an apple with a worm inside it, or an apple that is in
some other way repugnant, is to call it "doyulh" instead of just
"doyu." Doing that doesn't create a new word of Láadan meaning
something like "nasty apple," it just uses LH the way English would use
an adjective. On the other hand, there is the word for "rape," which is
"ralh"; in that case, the negative meaning carried by LH is an inherent
part of the word.
For either of these uses of LH, the only rule is that LH must be added
in such a way that the alternating consonant/vowel pattern is preserved
and no other rule of the sound system is broken. For example, the word
for food is "ana." If you wanted to say of some particular food that it
was spoiled, or tasted awful, or anything of that kind, you could say
either "analh" or "lhana"; both are acceptable and both would be
understood to have that meaning. If you are constructing a new word of
Láadan that has an inherently negative meaning you would follow
the same rule, but your choice would be more final. It would be
unlikely, and uneconomical, for the dictionary of Láadan
to include both "analh" and "lhana."
===================================================
1. Rule 3(b) was discovered by Jackie Powers.
Contact:
Jackie Powers