Beyond Whorf: "things," "qualities," and the origin of nouns and adjectives

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

ju'i lobyppli!

	  Coi! i mi du la tipitr park!  i mi du lo cnino ke lojban@egroups.com
cmima (for non-Lojbanists: "Attention, Lojban users! Hello! I'm T. Peter
Park, and I'm a new lojban@egroups.com list member"). 

              One aim of the development of Lojban and its predecessor
Loglan was to test the validity of the Sapir-Whorf Hypothesis, or
doctrine of linguistic relativity, developed by American linguists
Benjamin Lee Whorf (1897-1941) and Edward Sapir (1884-1939). The
Sapir-Whorf hypothesis claims that the structure of a language
conditions the ways in which a speaker of that language thinks, so that
the structures of different languages lead their speakers to view the
world in different ways. 

	Some of us are also probably familiar with the similar or related claim
of writers like Whorf, philosophers Bertrand Russell and A.J. Ayer,
semanticists Alfred Korzybski, C.K. Ogden, and S.I. Hayakawa, and
anthropologist Weston LaBarre that the concepts and categories of
traditional Western philosophy, from Plato and Aristotle to Descartes,
Locke, Kant, and Hegel, are simply the projection of the grammatical
patterns of Indo-European languages like Greek, Latin, English, French,
and German on the Universe. Thus, universals or Platonic Ideas like
Truth, Beauty, Goodness, Justice, and Man are said to simply reflect the
fact that Indo-European languages possess common and abstract nouns. The
bifurcation of the world into Things or Substances versus Qualities or
Attributes is explained as simply reflecting the grammatical distinction
of nouns and adjectives. Whorf devoted many of his essays to describing
in detail the differences between the basic English and Hopi
metaphysical world pictures by a detailed comparison of the respective
grammatical structures of English and Hopi. Thus, Whorf suggested on the
basis of comparative English and Hopi grammar, the way a people view
time, punctuality, continuation, and completion may be influenced by the
types of verbal tenses in their language. Whorf's theory has been
subjected to skeptical criticism by writers who contend that culture
shapes language rather than language shaping culture, or that no
correlation such as the one suggested by Whorf can in fact be found
between language and world-view.

          I myself suspect that language may influence thought in an
even more profound and fundamental way than the one posited by
Whorf--though in a way that is uniform for all known human languages and
cultures, essentially the same in English, Greek, Latin, Russian,
Estonian, Latvian, Arabic, Swahili, Chinese, Japanese, Tagalog, Hopi,
Eskimo, Cherokee, Shawnee, Nootka, and Kwakiutl. It's an insight I
arrived at while pondering the Whorfian claim that the familiar division
of the world into Things or Substances, Qualities or Attributes, and
Events or Actions simply reflects the Indo-European grammatical
trichotomy of nouns, adjectives, and verbs.

              All natural human languages that I know of, from all
linguistic families, all parts of the world, all cultures, all racial or
ethnic groups, and all levels of cultural development have (1) lots and
lots of words of the type "stick, stone, table, chair, knife, fork,
house, boat, man, woman, tree, dog, apple, head, eye, hand, foot, neck,
tongue, tooth, sun, moon, star", (2) lots and lots of words of the type
"big, small, long, short, good, bad, red, yellow, blue, green, black,
white, sweet, sour, rough, soft, hard, smooth, wet, dry," and (3) and
lots and lots of words of the type "eat, drink, bite, swallow, give,
take, walk, run, sit, stand, cut, hit, tear, pull, push, say, speak,
see, hear." This is true, as an empirical fact, of English, German,
French, Latin, Greek, Russian, Sanskrit, Estonian, Arabic, Hebrew,
Swahili, Xhosa, Zuilu, Yoruba, Chinese, Japanese, Korean, Indonesian,
Tagalog, Eskimo, Navaho, Tlingit, Nootka, Kwakiutl, Cherokee, Lakota,
Shawnee, Hopi, Zuñi, Aztec, Mayan, Quechua, Samoan, Hawaiian, Fijian,
and every known Native American, Sub-Saharan African and Australian
Aboriginal language, as well as of the vast majority of constructed
artificial languages like Esperanto, Interlingua, Volapük, Loglan, and
Lojban. It is also true whether or not any given language distinguishes
words of classes (1), (2), and (3) as morphologically distinct "parts of
speech" like Indo-European nouns, verbs, and adjectives with the
complicated conjugations and declensions we find in Greek, Latin,
Sanskrit, and Russian, or follows the example of Chinese, Loglan, and
Lojban in failing to make any such grammatical distinction (yhough such
languages still do make a grammatical distinction between words denoting
extra-linguistic "real world" objects, qualities, events and actions
versus grammatical operators--e.g., Chinese "full words" and "empty
words," Lojban gismu" and "cmavo"). Whether or not a formal distinction
in grammatical form is made between nouns, adjectives, and verbs, all
languages in practice do have lots and lots of common familiar words
belonging to each of these three broad semantic classes.

          I call them semantic classes. Exactly what is the basic
semantic distinction?

          Words like "red, green, blue, yellow, black, white, round,
square, wet, dry, sweet, sour, salty, bitter, rough, smooth, hard, soft,
good, bad, ugly, beautiful" refer to one particular specific sensory,
emotive, or aesthetic quality, which occurs over and over again in many,
many different things in the world. Words like "stick, stone, table,
chair, knife, fork, book, pencil, house, man, woman, tree, cat, dog,
head, apple, banana, leaf, head, eye, hand, foot, sun, moon, car, boat,
bicycle" refer to complex persisting or recurring bundles or aggregates
of such qualities that we notice as maintaining their character for
fairly long periods of time, or as intermittently disappearing but then
frequently recurring or reappearing in pretty much the same form as
before. Words like "eat, drink, bite, hold, grasp, give, take, send,
push, pull, cut, break, walk, run, sit, stand, go, come" denote certain
ways in which some such persistent bundles of qualities affect (or are
affected by) other such persistent bundles of qualities, or move (or
remain at rest) with respect to other such persistent bundles of
qualities.

             Now, if we take traditional Lockean-Humean British
Empiricism seriously, it might seem that the words like "yellow, green,
blue, sweet, sour,  round, rough, smooth, beautiful, ugly" are all that
a truly rational language ought to contain, since after all we are not
supposed to ever actually perceive anything except sensory, emotional,
or aesthetic qualities in all sorts of complex combinations and
sequences, combinations and sequences which can in principle be reduced
to a listing of their individual component qualities ("this is something
red, round, soft, and sweet accompanied by an emotion of pleasure").
However, as I just said, every single known human language on this
planet also in fact contains a vast number of words that really ought to
be superfluous according to a strict thoroughgoing Lockean-Humean
empiricism. Human beings just never ever seem to be content with having
a few hundred or few thousand words in their language for "black, white,
red, green, sweet, sour, rough, smooth, round, square, good, bad, ugly,
beautiful, scary, happy." They always invariably insist also on having
thousands of words for complex persisting or recurring bundles of
qualities, like "stick, stone, table, chair, book, pencil, tree, cat,
dog, apple, head, eye, hand, sun, moon, mountain, house, boat," and
thousands more for complex ways in which these persisting bundles of
qualities push, pull, cut, stab, tear, crush, incorporate, ingest, or
move around other persisting bundles of qualities, like "eat, drink,
bite, swallow, give, take, grab, grasp, cut, hit, chop, tear, come, go,
walk, run, jump, fly, swim, say, speak, love, hate, remember, sleep,
lie," etc.

             We learn very early in life that qualities do very often
come in persisting or recurring bundles, and that these persisting
bundles in turn have fairly recurrent ways of relating to the other
persisting bundles in their vicinity. All these persisting bundles, and
all these recurring ways for persisting bundles to act upon each other,
strike us as needing and deserving special names just as much as their
individual component qualities do. We feel from early toddlerhood on
that the world is so much easier, simpler, and more convenient to talk
about if we have words like "mommy, daddy, kitty, doggy,  milk, apple,
drink, eat, give" as well as words like "sweet, round, soft, hard, red,
green, black, white, hurts, sleepy." This, I feel, is an instinctive
basic biological and neurological need or impulse, hard-wired into our
brains and nervous systems, and probably a result of natural selection.
Our australopithecine, *Homo habilis*, and *Homo erectus* ancestors for
instance, had to react quickly, almost instantaneously, to a unified
holistic gestalt "SNAKE!!" rather than take their leisurely sweet time
nominalistically adding up one by one the individual qualities "long and
slender and flexible and scaly and legless and forked-tongued and
hissing"! However, while I believe it to be hard-wired into our brains,
I also think it is reinforced by language--by just about every human
language that there is.

            Now, a Whorfian thought experiment suggests itself. What
would happen if somebody invented an artificial language composed
entirely of words like "round, square, big, small, red, green, black,
white, sweet, sour, rough, smooth, dry, wet, happy, scary, itchy, hurts"
with no words for "stick, stone, man, woman, dog, apple" and no words
for "eat, drink, go, come, walk, push, pull", and then brought up a
child taught only to speak that artificial qualities-only language?
Would the child grow up truly lacking any concept of objects, actions,
or anything in the world except separate individual qualities? Or would
the child start modifying that language from almost the very start, by
inventing words for things (e.g., coining something like "miggert" for
"apple" and "loopty" for "chair"?) and/or by developing stereotyped
compound words for things out of compounded quality words (e.g.
"redround" for "apple")? And if the child did develop thing-words by
either or both processes, would he be very pragmatic and instrumental in
using them simply as conveniences of quick easy familiar speech, or
would he start developing a "things versus qualities" metaphysic around
them? Or would the fact that he had to personally invent nouns from
scratch as home-made jerry-built makeshifts cobbled together from
quality-words give him a kind of subconscious skepticism about their
reality, inhibiting him from developing a "things versus qualities"
metaphysic? 

	I recently found these speculations of mine echoed and confirmed in a
1985 book on the origin of language by Mount Holyoke College biology
professor Curtis G. Smith, *Ancestral voices: Language and the Evolution
of Human Consciousness* (Englewood Cliffs, N.J. : Prentice-Hall, 1985).
Curtis Smith attributed the origin of language to a sort of
"synaesthesia" between the senses, what Smith himself called intermodal
or cross-modal sensory processing. He used this to account for the
origin of nouns and verbs as well as of adjectives. On pp. 143-144 of
*Ancestral Voices*, in his chapter on "Language and Neural Codes," Prof.
Smith wrote:

	<<What is the poor brain to do to bring these different, mutually
incompatible classes of information [visual, auditory, olfactory,
tactile, etc.] together for some common processing step? There is no way
that these disparate inputs can be fed into any common processor without
being translated into a code that is capable of handling all of the
modalities at once. What might that code be? It cannot talk about sights
or sounds or smells. Such information would be meaningless to all but
the specialized portion of the sensory brain that had always been
committed to each of these senses. It cannot, in short, be a code that
deals with sensory signals emitted by some outside agent. It must be a
code that refers to the thing *itself*, not the stimuli it emits. The
new code symbol would not be "small, black and white, furry," nor
"pitter, patter, snuffle, stomp," nor yet "awful, acrid smell!" The code
would have to be a symbol that stood simply for *skunk*--a symbol for
the external reality itself, rather than a set of partial sensory
reports *about* the outside world. Sensory codes consisted entirely of
adjectives, and this universal cross-modal code introduced *nouns*. By
the same cross-modal process the nervous system developed a code that
integrates individual messages from muscles, stretch receptors, and
again the eye, to move beyond the body with a symbolic code that refers
to space and [pp.143/144] movement in the world outside of the skin,
rather angles of joints and stretch of muscles. Thus verbs were born.>>

	This, I think, helps beautifully to account for my own observation that
all known human languages without exception possess nouns and verbs as
well as adjectives, words for objects and actions as well as words for
qualities or individual discrete sense-data. If Curtis Smith and his
theories about cross-modal sensory processing are correct, the very
existence of language requires the existence of words for objects as
whole *Gestalts* and not just stringings-together of their various
qualities. To use Curtis Smith's own example, language from the very
beginning necessarily included words like "skunk" and never ever used
stringings-together of quality-words like
"black-white-furry-pitter-patter-stinky" more than perhaps to a very
limited extent! A language composed of adjective-chains like
"black-white-furry-stinky," if it had ever existed, would have defeated
the whole purpose of language--and could not perhaps have even existed
in the first place, as I see Curtis Smith's argument! Curtis Smith's
theory of linguistic origins, by the way, also suggests that, in talking
about the psychology of human sensory perception and the origin of our
mental concepts and complex ideas, the Gestalt psychologists may well
have gotten it more nearly right than John Locke and David Hume!
 
	---T. Peter <tpeterpark@erols.com>