Oligosynthetic languages

["T. Peter Park" <tpeterpark@erols.com>]

Dear Lobypli and IALlistsibs,

	In my "Beyond Whorf" post, I mentioned a "thought experiment" of
inventing a language containing only quality-words and no thing-words,
raising a child taught only to speak that language, and seeing whether
or not the child develops the concept of "things," "objects," or
"substances" on his own.

         Actually, a few dozen artificial languages of this type have
been invented in the last few centuries. They start in the 17th centrury
with the erudite attempts at a "characteristica universalis" by John
Wilkins, George Dalgarno, Francis Lodwick, and Gottfried Wilhelm von
Leibniz. They come down from the Baroque Era to bizarre 20th century
creations like the Austrian-American psychiatrist John Weilgart's *aUI*
(pronounnced "a-OO-ee"), the "Language of Space," allegedly taught to
Weilgart as a young boy by a little green elf-like space-man, as well as
semi-moronic crackpot efforts like dero-and-tero mythographer Richard
Shaver's "Mantong Alphabet," the supposed primordial tongue of ancient
Atlantis and Lemuria still preserved in the letters of the English
alphabet, published in his 1940's *Amazing Stories* tales about the
sinister doings of the deformed, dwarfish subterranean "dero" in their
cavern-cities. Umberto Eco described many of these "logical,"
"rational," and "philosophical" languages in *The Search for the Perfect
Language* (Blackwell, 1994). 

	Benjamin Lee Whorf, early in his linguistic career, toyed with a
hypothesis of "oligosynthesis" ("synthesis from a few elements"), the
suggestion that some natural languages (like Hebrew, Mayan, and Nahuatl)
might (1) have vocabularies very largely built up from strings of 2, 3,
or 4 short one- or two-phoneme basic roots and (2) have a relatively
small total repertory--only something like 40 to 60--of these roots. In
such "oligosynthetic" languages, Whorf believed, the words for such
common concepts as "mother", "dog," "tree", "eye," "sun," and "moon",for
instance, might be analyzable as compounds of two or three short
one-syllable (or even one-phoneme) roots, meaning something like, say,
"female-caring-person," "people-friendly-animal," "tall-erect- growing,"
"see-body-part,""big-up-light," and "little-up-light," respectively. A
language composed mostly of such "oligosynthetic" compounds would not
TOTALLY dispense with nouns, to be sure--but would still keep their
number to an absolute minimum, with most objects in the world denoted
largely by chains of "adjectives."

         All these "rational," "logical," "philosophical," or "cosmic"
languages, from Wilkins and Dalgarno to Shaver and Weilgart, break down
the thousands (or tens or hundreds of thousands) of concepts of ordinary
natural human languages into some 25 or 30 to 50 or 60 basic simple
concepts, each designated by a single letter, sound, syllable, or
symbol, with complex concepts expressed by stringing together the
letters, syllables, or symbols for simple concepts. Thus, for example,
they all express "mother" by something like
"human-female-progenitor,"or"reproductive-human-female-living-creature,"or
again like "not-active-round-person." These languages likewise represent
"apple" as something like "plant-reproductive-red-round" or like
"red-round-seed-holder-vegetable-living," "gold" as something like
"material-hard-shiny-yellow" or like "yellow-color-beautiful-matter,"
"dog" as something like "powerful-with-people-living-thing" or
"animal-domestic-four-legs-barking," and "I, me" as something like
"this-person," "person-here," "person-speaking," "this-human," or
"this-human-spirit."

         All these languages, in fact, have a few "nouns" or
"thing-words." However, by and large, they do strive very hard to denote
as many objects as possible as the sum total of their specific qualities
or characteristics. Thus, if anybody wanted to perform this experiment,
it would not be necessary to invent a "logical language" from scratch.
The experimenter could use the creations of Wilkins, Dalgarno, Lodwick,
or Leibniz, or more modern counterparts like John Weilgart's *aUI.*

	Linguists sometimes talk about the "poverty" of language as what makes
it useful as a medium of communication. The "poverty" of languages is
the fact that they do not have a separate word or name for each and
every particular individual object, event, action, or occurrence that we
could conceivably ever encounter, but only a few hundreds or thousands
of words for large broad general classes of objects, events, actions, or
processes. This is precisely what makes human languages useful as tools
for communications. If we had, for instance, a completely different and
unrelated word for every single horse but no general word for "horse,"
farmers could never communicate their anxious concerns to veterinaries,
and racing enthusiasts would have a very hard time collecting their
winnings. Instead of giving separate names to every single individual
object or occurrence, human languages divide the world into a large but
finite collection of classes or categories. Speaking of those classes
and categories, we make ourselves understood, and get the practical
business of our lives carried out. We rely on our friends, neighbors,
and colleagues understanding what we mean when we talk about "horses,"
"dogs," "trees," or "cars." Of course, our languages and cultural
linguistic conventions also do make it possible for us to talk about
individual members of a broad class if we really need or want to do so:
if we need to, we can talk about "Socrates," "Plato," or "John Smith" as
well as about "man," about "Dobbin," "Boukephalos," "Traveler,' or
"Secretariat" as well as about "horse," about "Fido," "Spot," "Rover,"
and "Towser" as well as about "dog," about "Mississippi," "Hudson,"
"Amazon," and "Nile" as well as about "river," about "Sirius,"
"Procyon," "Polaris," "Deneb," "Capella," "Alpha Centauri," "Tau Ceti,"
"Epsilon Eridani," and "Zeta Reticuli" as well as about "star," and so
on. Still, our common nouns are there, too--and, as the linguists remind
us, it is the presence and availability of common nouns and general
terms that makes languages useful as mediums of communication for
getting the everyday practical business of our lives done. We can
communicate with each other, and get all sorts of practical business
done, precisely because we can use general words like "dog" and "horse"
that leave out a great many idiosyncratic features of each individual
dog or horse.

        Still, despite what linguists say about the "poverty" of
ordinary natural human languages, natural languages can be seen as
extravagantly exuberant, redundant, and wasteful--and they have been
seen as wastefully redundant by many theorists in the last few
centuries. They find that it is in fact quite possible to simplify and
rationalize language by eliminating most of the hundreds or thousands of
words of natural languages for specific classes of objects, events,
activities, or processes. What they consider the wastefully large
vocabularies of natural languages could be easily simplified by
replacing most natural words with logical designations describing every
object, occurrence, or activity as the sum total of its component parts,
its significant qualities, or its separate individual steps or
sub-routines. Thus, we could easily dispense with a word like "mother"
if we could express it as "female-parent" or "person-female-progenitor."
Likewise, we would not need a separate word for "apple" if we could
instead just say or write something like "red-round-fruit" or
"plant-part-reproductive-red-round." Similarly, we could replace "I, me"
by "this-person" or "person-speaking-here," and "dog" with
"animal-quadruped-loyal-barking." 

	Instead of the several thousands or tens of thousands of common
familiar basic words of natural languages like English, French, German,
Latin, Greek, Russian, Sanskrit, Hindi, Estonian, Arabic, Hebrew,
Swahili, Chinese, Japanese, Nootka, Kwakiutl, Hopi, Aztec, Navaho, and
Quechua, we could now instead get by with a total vocabulary of just
something like 30 to 80 one- or two-letter roots for simple basic
elementary concepts like "material," "object," "person," "living,"
"quality," "social," etc., strung together into 3-, 4-, or 5-letter
strings expressing logically clear compound meanings like
"this-person-here" for "I, me" or "person-female-reproducing" for
"mother." This was the dream of 17th century theorists like Dalgarno,
Wilkins, and Leibniz, of 20th century linguists like Benjamin Lee Whorf
in his speculations about "oligosynthesis," and of occult-fringe
science-fictional "crackpots" like Richard Shaver with his "Mantong
Alphabet" and John Weilgart with *aUI*, the "Language of Space."
        
                     Regards,
                     T. Peter <tpeterpark@erols.com>
		     Garden City South, L.I., N.Y., U.S.A.