Conklin Folk Taxonomies

7/21/2019 Conklin Folk Taxonomies

http://slidepdf.com/reader/full/conklin-folk-taxonomies 1/24

PROBLEMS IN

LEXICOGR PHY

Edi ed

y

FRE W HOUSEHOLDER

and SOL S PORT

ublished by

INDIANA UNIVERSITY, BLOUMINGTON

MOUTON Go. T ~ EAOWEHENETHERL NDS



yHarold C Canklin

a l e

University

0 t r oducti

on

Many lexical,

problems

are of

conrsidesable

importance to 1331

guirsts and

ethnographers.

i t h

the

interaetar

of

bath groups in

mind, would

like

to

discuss

certa in a~ p ec ts

f

folk

claersificatfan

which

I feel

deserve

more rigorous ~exfcogxapM c ttention than

they have typically

rsc ived,

n

dequ te ethnographic

descriptian,

of the c u b r e

[Goodenough

19571 of a

particular

society preeuppaaets a detailed

analysis

of the

communications system

and

of the

culturally defined eftuatfontu in

which

all

relevant

distinctfans in that system

occur. his

re-

gard accurate

knowledge

af both

the

grammar

and

Iexfcun

of the

local spoken language con~titutas minimum requbxament. When

the

ethnographer

works In an

area for

which dequ te statementar

about the

local

language are unavailable in publf

shed suurcee,

his

f irst and often continuing task fa

the?

canertructfon of set of valid

rules fur the

hterpretation o

the local language.

n his

phona-

Logical and grammatical

analysis

of

new

speech form s, he may

find

many helpful madeke

n

the

dstrcriptive linguistic literature.

attempting, however,

to account

o x

the

obligatory semantic

relations inherent fn his lexical carpus he

may

not be

61

fortu-

nate

While extant

dic

anarise a.nd vocabularies

do

provide

glasses

and

definitional

information,

many of the

nontrivial,

and

often

essential* memantic

and

contextual relationships

obtaining

among lexical item8

are often either

neglected or handled in an

i m pr e c i ~ e

nd

unsystematic

manner [cf Mewman 954:86

1

For formal linguistic analysis it f a necessary that utterances

be acceptable and interpretable grammatically,

For ethnographic

(including lexicographic)

analyeis

u t t e r a x e

r

must also be ac-

captable and

interpretable

semantically.

White an

appeal

to

meaning

does not h p r a v a grammatical ana lysis,

neither

does

an



intuitive

appeal

to morphos p t a c

i

farm yf eld

the

most

appropriate

analy~is f

meaning

and reference

see

1.5 below).

n fact, an

adequate grammar may

generate ~sernantically nacceptable prop-

ainitiuxla [Chamsky 1955:149* 1957:103-41 cf Landar 1960:352;

Frake 196f:f.13], Results ~f same recent atternpt~ o develop

non

intuitive pracadurss

fur

the evaluation

of

the

grammaticakxese

and

meaningfulness of

sentences

[e.g., Maclay and Sleator

19603

cf

Jaos 19581 indicate

that

this Wference

is

of considerable impax-

tancs. The

distinction

between these two aspects of the

analysis

of speech

is

apparent

even in the

treatment

of isolated forme.

the

course

of several years of linguisti and ethnographic

fiel d work

among the Hanun60 in the Philippines, it became

abundantly

evident

that providing such

segments as

ah, tabikuq,

isarnparzmsiarknr-

qaUstun,

and

lgda, balaynutl. tagraZxlarn, qf

run-

piidiq

each

with the

same

gloss (d is thct ) kind of plant

was--while

adequate for certain syntactic

purposes--most unsatisfactory

for

the taak of semantic analysis.

Had Inat modified

this

procedure,

would

have ended

up

with

more

than

2 lexical

iteme

[including

~ a v e ~ a l

undred

referential synonyms)

each

labeled identfcaliy

While

employing glosses Xfke

teas and tobacco in

the

f irst

two

cases

above) proved useful

in

labeling

familiar objects, the

m a -

jority

of these culturally

~dgnificzant

Hanuncia designations

referred

to entities

which to m e

were

quite

unfamiliar,

In t h i s type of

ethnographic context ans inds

many

instances where the problems

faced

traditionally by the cumpflsra

of

bilingual dictionariesr

are

caneiderably m a g d i e d [ N i b

19581

For

the s t h a g r a p h ~ r , he

eemtsntic structure

of

euch folk classification

is

of paramount

significance

Upon hi^ analysis

of it

depend^

the

accuracy

of

many crucial

sttatements about

the culture being

described,

Prob-

lems af amlyzing and presenting such structures

fn

a succinct

faahfon may

be

of interest even

to

lexicographers who

work

only

rehtfvfaly familiar

cultural

surroundings.

1 .

Folk classltffcatfon

In the lexicographic treztbent af

folk

classification, we are

concerned primarily with 1 ) the identification of relevant eryn-

bc t f c oagments, 2 ) the identification

o

fundamental

semantic

d t a r

fn

~pec i f i c ontexts, 3 )

the dslhaatian of

aigrzificant

a s t e of

semantic

d t s

in

particular domains,

and

(4) the

translation

(and

markrfng) of

these

un f ta

so

that important semantic

relationships

wdl1 not be obecured, discusering different system8 of clasei-

f y h g

xssgmsntrs of the

natural and social

environment,

the

neutral



Lexicographical

Treatment

of

Folk Taxonomies 121

term segregate [Canlclin 19541

serves as a label

for any terrni-

nolagfcally-distjslgui ~skbed i a F conventionally-named)

grouping of

objects.

1 . 1 Linguistic

structure. The

shape and

cambinataria1

structure

of the

linguf

etic forms which designate folk

segregates

are

irrelevant,

n

a

atrict

sense, to

the analysis of the

system of

classrification

itself I ., to the aernantic Btructure [CanMh

19573,

Labels

and categoriaa

can change

independently

and there-

fore must be analyzed separately,

O n he other hand* a knowledge

of the linguistic structure invalved is e s ~ e n t f a lor understanding

the

principles of

folk nannsnclatur

e; and In

working

out this atruc-

ture, clues for isolating folk

segregate

labels and far

eliciting

information

about such

segregatebit

may be found,

1,2. Lexical uni ts and

coartexta,

A full lexical statement (f .e ,,

n

adequate

dictionary) should

provide

a

emantic

explanation,

as wel l

as

phonological and grammatical identffication* for every

meaningful

farm

whom signification

cannot be inferred

from a

knowledge of anything

elee

n the language, t f s convenient to

refer

to these elementary lexical

u d t a

as

lexemes [cf.

Swadesh

19461 Newman 1954;

Jorden

955; Goodenough 19563

although

other terms

have

been suggested e, g,, idiom [Mockett

1956 ;

cf.

Householder 1959:508-24; Weinreich 1960:337]). o far

as

Iexernfc

~ h t ~

s

concerned,

the

marphusyntactic or aserurned

etymological

relations

of

a

particular

linguisti form ar e incidental; what

is

essential

is

that

i t

meaning cannot

be

deduced

from

its

gram-

matical structure. Single

morphemes

are necessarily lexemes,

but f o r

palymorphemf canetructiuns

the dsciaion depends on

meaning

and

use (Implying

an analysis of

the constraint8

imposed

by the semantic

structure, and the

specuication

of

relevant

h

mediate

contexts),

Formal segments

such

as black

bird VB. blackbird) or

in the

old

house

vs. in

the do~house) an be excluded from the lexical

statement because they

re

predictable,

meaningfully,

In that

they can be considered

semantically

endocentric [Nida 1951:12-3,

1 9 5 8 ~ 2 8 6 ;

cf.

Chaa 1 9 5 3 : 3 8 5 ] .

Put another

way,

those

construe

tiaras which are never

semantically

exocantric

may

be

classed as

nonlexemic

forms

e. g., sunburned

face,

long pink

strand),

Prob-

ems

do arise*

however, in

degrees of lexemalc exocentricit).

[Nida f958: 86] and,

again, i caution is not

exercised in

distin-

guishing

clearly between

grammatical, and

semantic

criteria.

The

compounds firewater

and

silverff

ah,

for

example,

are endoc

entric

morphosyntactical3y

either

on an attribute-plus- head

basis

or on



A Problemer in

Lexicography

the

perhaps

stranger grounds

of

formal selection

rules

[Lees

1960:128, 1S8J),

but ssmaxxUcaUy they

are

as exocentric s vodka

and moth*

he

study

of

bsegrsgate

Iabels n folk

classification, and

despite some of the difficultfes of technical definition noted, I

find

it

ttusefd to distinguish by explkft

semantic

criteria

h o

inds

of

lexsmic unitst unit ry

laxemas

no segmenta of

which

may desig-

nate categorietn which

are identical with, or superordfnats

to,

those designated by the

farms

in queetfon) and composite

lexemes

(one

ox mare aegmenta of which,

under

spec3fisd conditions,

may

[a] esignate

the ieama

categorfesr

a those designated

by

th forms

in quastfun [abbreviation], or [b designate catsgaries

supesordfnate

to

thaae

designated

by the farme in

question [generalimatfan],

see

2. 2 .2 .

)

Unitary Iexemes may

be either

s imp le

(unsegmentablef

or

coxnplm

(segmentable),

These distfnc tions

are

exemplified

below:

Unitary

e3imple Uxrita

xy

c

omplex

Composite

oak

paiaun oak

white

oak

pine

pfxlsapple

pitch pine

B O ~

grandson

son-

n-

aw

dart

(an

artifact)

darts [a game) Baldwin

apple

Jack

jack-in-the-pulpit

Port-orford cedar

dandelf

on

black- eyed Susan black-crwvned

night

heron

caterpillar (larva) cat*a- eye

caterpillar tractor

For contrast, consider a few similar but

nanlexemic

forms: cheap

pine, ine and oak, black-eyed Joe, dart (plural of dart [Wockett

7

956t229X . For a native speaker,

such

distinctions cause

little

concern,

but

in

new linguistic and

cuJturaX environs

difficultiess

may arise.

F o r =ampler on

fir s impsction,

the fallowing partially- iden-

tical ~an u n 6aorms [ConMin 19541might appear

to

belong to a

simple

paradigm

(they could all be recorded durislg a

conversation

about

ric

a

cultivation

nd weeding problems):

paray:parap

atta tail

paray. miyah

immature

wild pZdaq

(plant)

3 pZray qigkantufi

kind

of

wild

sedge



Lexfcagraphical

Treatment

of Folk Taxonarnias

kind of rice

* P ~ O ~ Qne ( e b e ) ~ice

that

rice

The gloaaas,

however,

indicate that asvsxal type@of

lexical

units

may be hvolved.

Are

there

any formal linguistic clues?

Each

of

the six

farms

is

easily segmented into two rnorphs,

as

h v e

indicated

by the use

o

do t s

Loo~e-joining, phonemically,

is

repreeented

by

a

single

raised

period.

E3xcept

for the closely-

joined

doubling

in

item number

-,

the f a m a

in this

~ e trovide

no

obligatory intonational ar

junctural

contra sts J?urthermore,

each

fo rm

occurs in many identical frames such a e filmy ti

--

8

is fare)

certainly

dffferant, Thurs*

or

most

of

he

seman-

tically

di dnc t

types

of

joining

eruggeatsd

by

the

g los~ les , here

are

no phaaalagical

c l u s ~

nd

few,

if

any, immediate, formal indica-

tions, ( A f u l l syntactic tittatemant

covering

the structure

of

corn-

pounds

would

separate

out aome of these forms n gram-

matical

grounds

[cf. Lees I96OJ

Given

the

necesaaxy semantic

Infarmation,

however,

these

distinctions can be noted carvily or

lexicagraphical

purpoBes by

rewriting

the

former ars

follows:

paramray

pgray-rn5yalr

p i

ray-qi~kantuh.

~ i x a v .

fhud

Thie procedure clearly mark^ 1,

2

and

3

semantically

exocentxic

- - -

unitary lexernes;

4 as

a

composite lexerne; and and as

non

lexemic

rsemanticolly andacentric

canstructltan-rr the Wtial Zexerns

of

which is

rsupsrordhatsly

related to 4

Minimally, forms - -

3 , 4, Sa, and 6a could be labeled kinhof plant, but by not at-

- -

tending

to

essGtial semantic distinctions this

type

of short cut

would obscure such important contrastive relations as the mutual

exclusion

of coordinate categories

1 : [ p a 9

mplied--but not

covered--by the specific growth sta;e term number

2 : :

5a or

-

- -

ba ,

and the poeaible total

inclusion

o eubordinate categories 4

by

a/6a; but not 1, 2, or

3

by Salba . Statements about

such

Fe

- - -

lations, hinted at in soma i l o sse and definitions,

may

be dam-

anstrated only

by

systematic pairing

in

minimal,

and

relatively



controlledJ linguistic and semantic contatcs,

1 3.

Lexical

sats and

domains.

n many ways

it can

be said

h t

the

more

discrete the phenomena referred to the simpler the

task of

treating the

aed~ocfated erminology

in a XezcicograpMcaUy

adequate

maaner [cf

Wallace

and

Atkinsn 19603

3f this is true fur

particular lexical items it

is

equally true far t h

eemantically

structured

sats

which such

items may cornprirre [Frake 19611

M b h a U y r

a

lexical

set

consist^

of

aU

rssmaatically

contrastive

lexemss

which bn

a

given, culturally

relevant context @hats e c l u -

sively

at Leamat one defining feature [Loxtaebury 1956:61- 2]

The

semantic

range

of

all

~ u c h

examea

defines

the domain

of

the la

ical set The initial e s t a b f i ~ h e n tf domain boundaries

while

wfdsiy

recogniz;ed aa

an

ideal goal, is often a very difficult

task

[cf Voegelin and o e g e l h 19571.

Effective

eliciting

frames

and

procedural t e s t s unad to deteranhe

such

boundaries

and canvhcing

demunstratioaz~ f

their

Intracdtural real t y

axr

e

subject

not often

d i ~ c u ~ s e d

n the Linguistic or

ethnological literature.

Some of the

essentf k

factors involved in

t h i ~

ype

of analysis are treated

briefly

below

under levels 2 . )

and

dimensions

3 . )

of

con-

trast.

In general, the

number and complexity of

boundary problems

increaaea

as one moves f rom the investigation

of lexfcal

domains

within

a

particular language

to an

attempt to umatchn

he

domains

of different languages [ ~ h m a n 9 5 3 ; cf. uine 1960~26 -791 . This

doe not.

however,

pr ecluda rigorous contrastive analysis.

1-4. Translation and semantic

structure.

W i t h

sw excep

fons, the lexical

items employed

3n

ay~tema

f folk classification

alwaya

comprise

a

segment of the

everyday vocabulary

of the

particular language [Gonklin

9573

The rules governing the

ab-

Ugatoxy

semantic

relations

among

the

categories

in

such

lexical

sets are t h u s

to

be determined, evaluated,

and described

far each

language, Such

rules

camat

be

prescribed

merely

on

the baisis

of familiarity S another system

with the

concrete dsnotata of

he

sets

fnvulved,

In the case of folk botany, f o x example, this

meaas that a

local system

of plant cfasslffcation

cannot

be

de-

scribed

accurately by

attempting to obtain

only vernacular

* agufvalsnta for

botanically

r

cognf

zed s p e d s

Translati~n

labela (glasere~) re frequently necessary, but they ghould be con-

sidered neither as defbitfons nor as

exact

equivalanta [Launsbury

1956:163;

POT

an attempt to

use

acsonyms

s

a partial mnemonic

solution

to

such translatian problem s,

see

Landar

et

al

1960:

7 This well- es tablished

and

perhaps obvious semantic

prfn-

ciple

is

s u m e t h s s forgotten

where

the aasumed absolute

nature



Lexic ographical T eatsrnent of

Fofk

Taxonomies

125

f cross- lhguist ic

s e n ~ e )

f

scientific

names or

of

other

long-

established traditional dlstbctionsl

n

certain

Western languages

is hvo ived

16brxlaxz

953;

cf, Simpsan l961:11].

1.5.

Syntactic

vs,

semantic

structure.

Implicit

he

pre-

ceding remarks is theasaumptian that the relation between formal

linguistic

(syntactic,

in

the

general, semiotic sense [Morris 19461

structure

and 'semantic structure need not be

ialarnarpkfc

fltouns-

bury

1956:

1891.

f

this assumption

is

taken

seriouslyr a

full

dic-

tionary

should state explicitly

the

necessarl and

sufficient con-

ditions far

the unambiguous structural interpretation o each

in-

cluded

lexerne

in the

context of the

total

'lexicon

as

well as

in

that

of

the

grammar. While

su h

coverage h r rarely

been achieved,

even for relatively

small lexical domains, J feel that recognition

of this goal ha8 considerable

relevance

for

this

f

scusaian.

A

brief illustration

may

help to indicate the kind

~f

crucial lexical

data

that

are often ignored, sspscia'bly where meanings

are

either

assumed

on

the

basis

af

semantic

patterning

in more

familiar

language, or where

they

are treated only partially as in the

der-

ivation of

definitianaX

statements from tranpilatioml lab els ).

C ~ a s i d e

the

following situation

(which,

with minor differences,

I

have

encountered on a numb er

of

occasions): a, woman, whose

brother x)

and

husband {y) are

both

named

Juan, has sont also

named Juan a) and a

daughter who

in turn

has

a,

son named Pedra

P). The genealogical situation is

diagrammed

in Ftg.

I we

can

ignore

the broken

lines for a

moment).

Two fluent

speakers

of

English, F a

Filipino whose

first language

was Tagalog, and

A,

a

native speaker of

a dialect of American English, bath

b o w

Pedro and.the specified

members of

hits family. The

fact

that

one

of

he

Juans x, y or 2; h s died i~

known

only

to

A

or

F)

who in

turn wishes to relate t h i a circumstance to his

friend

F

far

A).

A

~ctraightforward tatement cornpfe t h g

the

sentence

P'

Juan

died

would

seem

to do the

trick;

and, depending

on

the circum-

tances,

one

of

two

unitary

lexemas

(grandfather,

uncle)

might

be

used

to

fill the blank:

Pedro's Grandfather Juan

died.

Pedro's

Uncle Juan died.

However,

if A uses Grandfather F

may

ask Which grandfather ?

i F

uses Uncle,

A

may ask Which

uncle? indicating a kind of two-

ay ambiguity which

can

only be resolved by recognizing

that

despite

their unque

t ionabf

e grammatic alfty and rnorpho

syntactic identity,



Figure 1 A genealogical illustration o

contrasting

s y ~ t e m ~

o

kinsfiip classfficatfon



Lexicagraphical Treatment sfFolk

Taxonomies

f

7

A's

sentences

1 and

2

and

F s

sentences 1 and differ asman-

ticallyz

Sentemc e Ki nterm used

K Z n ~ p a f a ~ , ~ c l m d a d ( P x = p a r t n t s )

A1

Grandfather Y

(Pr

a)

Fl

H

x and y (PrFa, PrPrBx)

A2 Uncle

x

and

e

(PrBr,

PrPrBr)

F2 i t

;

(Pr r)

This, of course,

reflectls

only a small part of a very fundamental

strut

tural

f rencs in

Central

Philippine

and

North

American

systems

o k l n ~ h i p

lassification: univers l termhologfcal

rec-

ognition of

generation

in

the

f

oxmer

vs

.

uxlliva

r

aal

terminological

recognition of degree

of

collaterality in the latter

(these

wo ulimits

to

the

lexical extension

of kin class

msmbershfp

are

indicated on

the kinship diagram in Fig.

1

by the horizontal and diagonal broken

lines, re psctive3y , Although any careful investigator might

learn

this

systematic distinction after a few days of field work, the

principle

goes

unaccounted far n

the

relevant and extant bilingual

dictionaries,

The

restrictions Involved

in

this

il1ustration axe

just

as obligatory and ins scapable within the re

epective

semantic sy-

stems represented as

is

the distinction of singular VEP plural

in

English grammar,

2

Levels of

coxltraet

Folk tategoriee

within

the

same

domain may be related in

two

fundamentally different

ways:

by inclusion,

which implies

separate

levels of contrast, and

by

excluaiuni which here applies only with-

in single-level contrastive sets. There may also be ~lubcategosic,

or componential, intersection (see 3 below).

In studying semantic

relationships, as among folk

categories, it

h s often been

dem-

onstrated

th t

likenes s

gically

and edgnlffcantly

h p l i s

dff-

farence [Kelly l955:303-51,

It

Pe also pertinent,

however to

note

that

total.

antrast complete complementary erxcLusion)--whfc h

logically

relates such segregates

aa ant

nd ship or cough and

pebble-

is

less important than re trictad contrast within the

range of a

particular semantic subset

(compare the

r latioan with-

in

and

b a t w e ~ nhe partial

sets robin

wren sparrow; spaniel

errier

poodle:

and bird 3 .

henwe

speakofthe

category

dime eing included in

the

category coin e mply that every dims

l a

also a (kind of) coin--but

not

necessarily the reverse,

Further-

more

when

we

state

a )

that

the category

dime ontrasts

with

that of auartar and (bl

that

the catetrorv coin

contrasts

with

that of



Problems

in

Lexicography

bi l l we are

speaking of

two

instances

of

relevant

mutual exclusion

t

two

different levels of

c e n t r a ~ tCanklin 1 9 5 5 ,

1951 ;

Frake

19611.

Such

alignments

of folk categories

are common to

all

Tan-

guages,

though

systematic

indicatione

of

these

relationships

are

rare even iss the more

detailed

monolingual dictionaries.

2,1,

Hierarchic

structure.

Where the

articulation

between

succes~ive

evela,

each consisting of

a

s e t of contrastive lexical

u n i t s ,

i ordered vertically

by

inclusion such that each mono-

lexemic category at one level is totally included in only

one

category

at the

next

higher level, we can

speak of

a lexical

hierarchy*

The

two

xes of

such

structure

involve

the horizontal differentiation

of contrastive

but

coordinate

categories and

the vertical increase

of generafization OX spe~ i f i ca t ion esul thg from

ascent

to super-

ordhate (including) or descent to subordhate (Included)

levels,

raspectiveiy

[Gregg

1954;

Gonkl in

1957 ;

Beckner

1959 ;

5 5 - 8 0 ;

Frake 1961?117],

These axes

ss

fixed and camat be merged or

interchanged,

not

can the succession of levels be modified. Dime

irs

not contrasted

with

coin but

t the

same level

wit?i

nickel,

uarter, penny,

etc.

Subhierarchies

of

varying depthan are often

diacernfbPe within

larger hierarchic structures.

The

depth

in

levels) of

the subhierarchy

including

the categories

hawk,

pigeon,

nd

starling is less than that of the

subhierarchy

including hawk,

horse, and crocodile; i e,, the

fir

s t three

segregates

are included

in a

superordinate

category at a lower level than that of

the

Beg-

rcgate

ultimately bcludPng hawk,

or s and crocodile. The

embedding of shallow

subhierarchies

within increasingly deeper

ones

i~

characterfatic of many

systems of folk classification.

2 2 Folk taxonomy,

A

system

of

monolexemica~ly-labeled

olk

segregates related-byhierarchic inclusion is a folk taxonomy;,

egregates included

fn

tsuch

a class$ficat%on

re known as folk x

CaaMh 957; cf, Lawrence

95 ~53;

Sfrnpsan

1961:X9],

Same

of

the additional requirements of modal or regular taxonomic

aystems

[Woodger

1952:201ff,; Gregg 1954; Backner 1 9 5 9 : 5 5 - 8 ;

Simpson 196ll are:

I ) at the highest level,

there

1s only one

maximal (largest, imique)

taxon

which

includes aU

other taxa in

the

system; 2 )

the number of levels is

finite and

uniform through-

out

the sys

ern; 3 ) each taxon belongs to only one level;

4) there

is

no

overlap

ti.

e,, taxa at the same level are always

mutually

exclusive),

Fa& systems vary

widely wi t respect to these more

specific

*requirements,*

but the presence of hierarchically ard

ranged though less

@regularw

olk taxonomies is probably universal.

Mast af the examples

given

h r

are taken

f rom

folk botany, but

similar

iUustratiuras could

be taken

from

other

domainr~

Thomas



Lexicagraphical Treatment of F a k Taxonomies

12 9

1957; Frake

19611.

Several important differences distinguish

folk

taxa

from

t he

taxonomic groups of biological

syrstematics

[Canklfn 19571 Simp-

son

196

1

The farmer

usually

relate

only

to

Xacakly

relevant or

directly

obse~vable

hanomeata,

They a r e defined by criteria

which may

differ

greatly f rom culture

to culture.

The

number

and position

of

levels

of

contrast may change f r om one

sector

of

a folk system to another. There are no formal

rules

for the

aarnancfatural

recognition ax rejection of tgxa [cf Lawrence

1951

213-51,

though new groupings may be

added

productively

with

con-

siderable ease, 1In respect to any

particular

loca l biota, there is

no reason to expect the folk taxa to match those

of

systematic

biologyw-either

fn

n u m b e r

or in

range.

The

~a r run6o lassify their

local

plant

world, at

the

lowest

terminal)

Xeval

of

contrast,

into

more than

1800

mutually exclusrfva folk

taxa,

while botanists

divide the same flora--in terms

of

sgacfee--into less

than

1300

scientkfic taxit

2 . 3 . Special

problems. Although

they cannot

be discussed

here at length, a

number of

lexicographically

important problem s

encountered

in the analysis of

folk

taxonomies

include:

1)

Multiple and

interl c

king

Mar

archies. Unlike

scientific

taxa folk

segregates may belong shultaneoutilly to sevexal

di -

tin t hierarchic structures. The same segregates may be cXasrsed

as

terminal

categories

In

a

taxonomy baaed

an

fo rm

and

appear-

ance

and

also as terminal or nonterrninaf categories

in

another

taxonomy based on cultural treament e , g , , marphologica1ly dia

t inpi shed a d s of f loral segregates vs

functianak

categories

of

plants

as

food cultlgens, medicines,

ornamentals, etc.

[Conklh

19541. Subhierarchies

may

be interar tfculated in numerous ways

[ e ,

g. Soas 1956:296 71 and

there is always

the potentiality

of

partial inclusfan a x domain overlap.

2) Fxtrahferarchic relations. Not al l folk categories are

directly

related by

class inclusion

ax

contrast

within the range of

a particular superordinate

category.

For example,

numerous

difficulities may arise i lexemes designating separate ontogenetic

stages

or

parts of

member~ l

f particular segregate8 (see

1.2.

above) are not distinguished f rom hierarchically arranged

folk

taxa

[Chao 1 9 5 3 : 3 8 7 9 ;

Conklin

1954, 1957; Frake 1961 Part-of

part-whole)

relations are

often complicated

by ambiguities

[Magel

1961:381 3J

nut

encountered

in the analysis of khd-of class in

clusion)

relations (e. g. the eegregates plant,

stemt s p are not

elated taxonomically

like

plant, t ree,

elm).



particular folk

taxonomy, a rsfngle

taxon may be labeled by

pho-

aernically distinct forms, a s fn the case of

minor

dialect variants

r abbreviated terme [see 1,2.], we may apeak.uf referential syn-

o n p s or synonymouer lexemes); e. g r fb M , ive* fiver,

ive-spot, five-dollar bill,

n

many such cases,

it

may be dif-

ficult

to

demonstrate

taxonomic

identity and the

absence

of

catsgar ic overlap.

Alternative

s u b ructurfng

of the subhierarchy

may be involved.

Phonemically

identical

(homonymous)

Xexemes

may dersignata

separate

taxa of different ranges of genera1ia;atian

at su essive levels. Such

situatione

e.g , ,

animal

and man

in

he following

partial

contrastive set : animal.

vs. plant a n h a p

vs.

mad, m a d

s .

woman [cf.

~rake 1961:11?-91) re

not

un-

common

but they require careful.

contrastive pairing

and testing

far

inclusion

at

each

level involved.

Sdmilar

steps

must

also

be

taken

in working out problems concerned with diertinguishhg

polysemy from homonymy [ e l l s 1958:662- 3; cf,

Chamsky 1957:

9 5 ; Garvfn

1960:14?].

4)

Types

of contrast,

Paired

folk taxa

of some lexical aub-

setsl

are

related

by

simple, bh ar y, segre gate opposition. Many

larger sets and same dyadic ones jnvalve

important

t ypes a£ ae-

marttic contxast other than antanymy fcf. Lyons

1960:622

3 Struc-

turally,

bar

example, taxa may be

contra~tedn

aeria l , cam-

pletmentary,

or

d i sc

antinuous arrays. (For aubcategoric

attribute

relations,

eee

3.

below.)

2 4 Folk vs

botanical taxonomy.

Ideally* a

th

study of

nterrelated

l a i c a l

setla

in

folk

taxonomiao, priaxity

and

pref-

erence should be given t o

uxlaahously-agreed-upan, obligatory

dlstinctians

in

epecffied contexts. When

terated

by meaaa of

what

are

es~ential ly

ruchl

experiments--by

pairing and

contrasting

nsgatfvely and

positively--one

should be able

to

construct a model

(i,

e.,

a theoretical

statement)

of

the hierarchic

sltructure each

that aa~ertioner

of

membership and inclusion

in any of the implied

b x a are uaanimously and unambiguously

denied

whenever such

assertions are incongruent i ,e., meaningless w i t h b the s y ~ t e m )

[cf.

Joos X958:65],

The

assertion Poodles,

dogs,

and animals

are

kinds

of snails would

thus

be rejected by speaksra

of

m y

dialect of English-and an

vary

easily specified semantic groundis.

i t h i n a particular

universe

of discourse (a

taxonomic domain)

how can axle construct a nontrivial

modal by

means

af

which only

semantically acceptable, congruent propositions may

be

generated ?

An example

from Hanun6

folk

botany may

serve as a

partial

amwer.



Lexicographical Treatment of Folk

Taxonomf

as

131

In a

fjtituation

where

one

aaunoa

armer

wishes

to draw an-

other's

attention to a

particular fndividual pepper bush

Q, he

may,

of courme, attempt

to

describe some of Q's unique attributes with-

out naming

the

plant. Much

mare

often,

however,

even

n

the

caurae of

a unique

descrfptian, he wffl resort t o the

use of one

ox more

of at

least sight Lexical uni t8 each of which might complete

the frame rnSluq* g h d a p g

'Hey, take

a

look at this

I

ut at

different levels of

contrast

(allowing

for

different degrees

of deb~ired r required specificity):

kuwaq

'entity'

i.

e., something

that

can

be

named)

Ur kgyuh

'plant'

not a

rock,

etc.)

IV gilamnun 'herbacsou~

lant' (not

a woody

plant,

etc,

V ladaq 'pepper (plant)' (not

a

rice plant,

etc

VI

Xgda-

balapun ous yard pepper

(plant)'

not

a

wild

pepper

p l a t )

V U

lZda.baiayaun, 'hauseyardchilipepper

maGrat (plant)

not

a

hou~eyard reen

pepper

plant)

VU3

lzda.

ba3apun- Ucat-penis houseyard

mahGrat.

chfli pepper (plant)'

qiitfn-

kutiq

(not

a

member of any

of five

oth r

terminal

houseyard

chili

pepper

taxa

such as fgda.

alaynun,

ma rat,

tghud-manuk.

the

U ~ ~ ~ k '

-spurn variety).

i t h i n

th damafn of

EXanm6o plant taxonomyl f rom ldvel W

down, and specifically within the range of izdaq, from level V down,

conversations recorded

during

many taixnfJar sdtuatfone

would



ultimately

provide the

lexicographer

with

fifteen

unitary and cam-

posfta

lexsrnss (including a terminal s e t of eleven pepper plant

names) arranged

at

four levels in the fo rm of

a

discrete subhier-

archy

Fig, 2 .

Specification

below the level

of

the termbal

taxa

noted in the diagram Fig,

:l-11),

and

hence outside

tMs

system

of ciaes3licwtiun, may be pr

avldsd

only

by

semantically

endoc

antric

cons

ructiona de c ibing individual plant variations, on which u-

m n h a u s accord i e rare

and

unpredictable.

n this particular case

folk

taxa 1 5 14,

and

I 1

happen to carraspond

rather

clarsely

with

the scientific taxa

Capsicum,

C. amuum L., and C

frutescens

L.,

respectively;

but

the

twelve remaining folk

taxa involve distinctions

not recognized as significant botanical

subspeclet

s

by

taxonomic

botanists who have classified

the

same flora

Structurally speaking,

however, same

of

the

most important patterns of

semantic contrast

fnvtllve ot only

the

hierarchic separation

of

these varied, Xower-

level , folk axat

but also

a

large

number

of

nonhferarchic relations

governed

by

aublexemic class Intessection (see 3-21.

Although such

relatiana cannot

be diagrammed

with

the taxonomic implications of

Fig,

2*

nor

can

they

be

treated

effectively

at

al l

in

terms

of

our

hierarchic model,

they

should

aevertheles~

be

of

cansiderabZe

interest to

linguists

and others

concerned with sylsteme of

folk

claesificatian,

3. Dimenaionhn

of contrast

At any given level within a well-defined folk-taxanarnfc

aub-

hierarchy, the relations obtaining among

three

or

more coordinate

t ae may

invalve

varying

dirnensioax~,

or kinds of subcategory con

trast.

The

conjunction o these dimension^ x more

precisely of

the

values

(or

spsclfic

attributes

fcf.

Bruner

at

al,

l956:26-30

along t ie

several

dimensions, define the categories fIlv~lvedwithin

an ss

entially

paradigmatic i. e., n ~ n h i e r a x ~ h i ~ )ubsystem

[Lounsbuxy 3956 1960~27-8; for a

discussion of

the

structurally

similar

though

more typologicalXy-oriented procedures

of attribute

gpaca substructiaa and

reduction

see Greenberg I957 and Laaslrsfeld

19611.

3.1. Nunhie

rarchic

structure, Such rnultidimens ional

con-

trasts

do

not imply,

and

M e e d

do

ot allow, the

ordering of the

resultant categories

by hierarchic

fnclurion. These features sf

nunhierarchic

semantic

structures,

while

not

always

sharply

a s -

thguished

f rom

the

principles

U s r e n t

in

hierarchic aystems,

have

been

recognized and carefully

analyzed

in

a n u m b e r of

domains,

notably

in

kinehip [Guudea~ugh

951:BZ-

11O

1956 ;

Lounsbury 1956;



Figure 2

A

segment

of

Hanun6

plant

taxonomy.

All folk t x included in the taxon lidaq are

indicated.



5kake

1960 ; Wallace and Atkfns 1960 ] , color [CanMin 1955 ;

cf.

Lemaberg and Roberts 1956; Landar t

al,

19601,

orientation [ m u

gen 19571, disearaa [Frake 19611 and, beglarnhg with Jakobsan'a

pioneering

effffarta,

Sn

sruch

partly

rnodulatianal

[Jaoe

958

t QX

para-

digms

a8 case

and

pronoun

arystems

[Jakobson

1936;

Sabeok

1946;

Harris

1948;

Latz

1949; Wandesly

19521Auaterlitz 19591, The

folkawing -mpie of multidimensional contraat

fn a

regular para-

digmatic ~tructure

will

illustrate

some

of these

pointer.

3,2,

Significant claasificatian

vs,

catafagufng,

Xf,

omitting

t he high- level, wide- ranging h w a g F e e 2.4) w e st all t he

nun60 personal name subatitutest

occurring

fn variaust

frames

such as m5luqr qhda pag bhwat ni

'Hey,

take a look

at

what

did

(here),' we will inFariably end up

with

an exhaustive

and mutually

qcl ur ~ive exical

s i t cannrieting

o

juat

efght

d t s (in

each

cage xepreaenting

a single

morpheme),

Arranged in the

l e a ~ t

eaningful type of

catalogue,

an alphabetical index (as in a

ictionary), these

lexical un ts are:

dah

uh

ih

uh

ah

-

tam

yah

YE ?

'they'

'I'

*we8

'youY

'we WQ

'we all'

'he,

shef

'you all*

The @hapee provide little that i a ~trructurally u g g e a t i v a ~ut

the gloestes do indicate th t n orderbng in terms of efght

Utradi-

tfonal distinctions along three

q u a i -

semantf dheaeaiuns

1)

first person

:

nscund

psraon

:

third person

2 ) singular : dual : plural

3 ) axclu~live : inclusive

might be attempted,

But the re~ul t ing

pplied structure

is

hardly

elegant, aconamfcal, x convincing:

ku 1s

tab

ld

mfh

lpe

-

-

tarn Ipi

uh

63 - - -

yah

3 s

w

yuh 2p

dah 3p

U a

close examination

is

made of

the disttinctive contrasts in

volved, nut in

terms o

labels but in terms of

actual,

minimal,

obligatory differences, a m re satisfactory, economical, and



Lexicographical Treatment of Folk Taxonomies

135

semantically

verifiable solution is reached.

The

neces ary a d

sufficient conditions for defining

each

of the

eight categories de

pend

on

the

regular

intersection

of

sfx

components

which comprise

three

simple

oppositions

minimal

membership

aonmfnimal membership Ma)

inclusion

of speaker

e x c ~ u ~ f u a

f speaker S 1

inclusion

of hearer

excluaian of

hearer H .r

doh

y u h

m i h

f urn

y o

muh

k u h

f a h

m i

i

SH

MS

SH

M

SA

M S H

MSR

M S H

Figure

3

Paradigmatic structure

of

a ~ a n u x 1 6 o p r o n o m ~ ~

es .

These

relations can be represented in l i d t

or

diagrammatic form

Fig.

3 . Even

witbout

hrther

elaboration,

the basic semantic

atsucture of

t is

lexical set should now

be

clear,

h

assing, it

may

be

noted that

pronoun

systems

Tagalog,

nocano

[Thomas

19551,

Maranao [Mclcmughan

1959 ]

and B o r n e

other

Philippine lan

guagea exhibit very airnilar,

if not

identical, obligatory aemantic

r

lationahips

This

example

also

flluatratesl

a

vary

important, though perhaps

less

obvious, character1stic of paradigmatic relation8

t one

level

in

taxonomic subhierarchy

n

contrast

to

the nancommutat;Ive

re

lations o class inclusion governing the larger taxonomic sysltem

i t h i n

such

a contrastive

lexical set { st

in Fig,

31, ordered by



class inter section*

the constituent

categories cannot be arranged

in

a

taxaaamic hierarchy. Any

arrangement

e. g.

a circular, block,

or

branching diagram) superficially appearing ta contradict this

statement w3ll prove on

closer

inspection either a) to constitute

what the

biolagi~ts

all a

ke y

[Mayr et a1, 1953: 162-8; Simpson

961313-61 es~entfaf ly

nother

k h d o

cataXogue

or

finding l i t

ordered by

successiva-

-but not nec essar ily taxonomically

signHi-

cant--dfcHatamous exclusion* or b)t be baaed on same

other ar-

tific

ially

imposed, and hence semantically nonsignfficant, clas ai-

ficatfan.

4, Lexicagraphic

treatment

The

ways in which

th problem~l

mentioned n

t h h

paper may

be

treated

in

bilingual dictioaarie8 especidly ethnographic

dic-

tionaries, a r e

practically

unlimited.

That very few of the possi-

bflitiera

have

bean

explored to date f jl

df

sappofnting, but not dis-

couraging*

There have been a

number

of new attempts t expanding

the analytf

c

procedures

of

deexcriptive

linguistic

a

to

include more

rigorous, thorough, and theoretically

rewardfng

analysist o

B e

mantic stru ture

[e,

g.,

Goodenough 1956; Launabury 1956; Nida

1958; Frake

196

J

Despite th s more encouraging

signs,

:

realize th t

mast

dictionaries will continue

to

be

organized

prisna-

riiy as

alphabetical

indices.

Suggestions

regarding

the

ways

in

which

structural.

s rnantic information especially

with

reference

to folk taxonomies) might be mare adequately covered in

such

dic-

tionaries would include, wherever

pasaible:

1) consis nt marking

of each

entry

as

t

its status as

a

lexical unit

and

taxon,

its

im

mediately subordinate

taxa

and

oupsrordhate taxon, and all

caar-

dfpate

taxa included with t in this next

higher taxon

simple dia-

riti s

nd abbreviations can be devised for syeternatic

use fn

compilation

and

checking);

2 ) differential marking of translation

labels

and

of

definitions; 3 ) concise

indication of

distinctive at-

tributes which defhe categories belonging

to analyzed

lexical sets;

4) systematic crass-referencing to maximal

taxa

in a l l

major

subhierarchiea,

to referential

synonyms, and

to

al l

units involved

in

categaric

overlap;

and 5 ) frequent

use of

structural charts and

diagrams,

Where only l imited

oppo~tunities

re available

far

accomplishing such

tasks, priority might

be given

to

those

pasts

of

the lexicon which

on

the

basis of

nunintuitive and

fntracultural

criteria*

appear

t

involve semantic relations of an everyday*

obligatory

nature.

n number

of

segregates, paradigmatic com-

plexity, and hierarchic depth,

certain

lexical

domains are likely



LexicographicaX Treatment of Folk Taxanomf

est

1 3 7

to

be mare

highly structured than others [Brown 1956:307; Mi&

1958:283-4; Worth

1960:277; Frake 1961:

121- 21.

F o r the

student

of

olk

taxonomy,

focusing attentfan

on these domains

should

f

ad

not

only

to

more interssting

analytic problems

but

alao

to

reaulta

of greater

laxicographical and general cultural r levance

.

5

References cited

~ u s t e r 1 i t z ,Robert,

1959.

Usemanticcomponents of

pronoun

ays -

terns: Gilyak,*

Word

15 1:102-f09.

New Yark.

e c h e r , Marton, 1959 . The Biological Way of Thought. vifl*

200

pp.

New York: Columbia Udverroity Preas.

Brawn,

Roger

W. 1956. Language and categories, pp.

247-

31

2

(Appendix)

in A

Study

of

Thinking,

by

Jerome;

S

Bruner

et l. New

Yorkr

John

wiley and

Sons.

Brunerl Jerome S

Jacquaber 3. Goadnow,

and

George

A,

Auatin.

1956 .

A Study

f

hinking

(with

an

Appendix

on Language

by

Roger

W

Brown). xii

330

pp.

New Yarkt

John W h y

and

Sans.

Chao,

Yuan Ren. 1 9 5 3 Popular Chhsae plant

wards,

a

descrip-

tive lexico-gsarnmaticzaX study, Language

29,3:379-414.

Baltfmor

a

Chomsky, A. Noam,

1955

Semantic cone3ideratians in grammar,

Georgetown U l J

onograph

Serierr an Languages and

inguistics No. 8, pp. 141-150.

wasfig=, D.C

Chomsky,

No-.

1957 . Syntactic Structures, 16 pp, 's-

Gsavenhage:

Mouton,

Corzklirt Harold C

1954.

The

relattan

of Hanu60 culture to the

plant world.* Unpublished Ph.

D.

dig sertatfon

in

anthropaXogy,

Ya l e University, New Haven.

.

1955

Hanun60

color

categories

.

Southwestern

Journal

of

Anthxopaloq f 1 3 9 - 344,

Albuquerque.

.

1957

~Ethnobotan ica l

roblem s in

the campar

ative

study af folk axonomy. Paper read at the Ninth Pacific Science

Congress, 1957.

Bangkok.

f ~ u b l i s h e d i n P r o c e e d i n ~ s ,inth

Pacific Science Congress,

957,

4:299 301, Bangkok, 1962.1

FraEce Charles

0

1960. The Eastern Subanun

of

Mindanaa.

pp.

5

64

in Social

Structure

n Southeast Asia , ed. George

Peter

M u r d ~ c k . Chicago:

Quadrangle

Books.



. 1961. The dfagno~iaof

disease

among the

Subanuxl

o

b b n t t a ,

American. h t h r o p o l o g i ~ t

3 1:113-132.

Garvh, P a d L. 1960. On structurallat Georgetown

Univer~ i ty

Monograph

Series an.Language

and

inguistice

No.

11,

pp.

145-148.

washin&, D.C.

Gaadenaugh*

Ward

L 1951, Property,

Kin,

and Community

on

ruk.

ale

Univar

sity

Publications

in

Anthropology.

~ 0 . 4 6

90

pp,

New

Haven,

1956

CemponentJsrL

analysis and

the

study

of

meaning.

Language 32.2:

195-

216. 13altimars.

1957, YCultural

ant&opalogy and

bgufarticr~..

Georgetown

University

Monograph Serieti an Language8 and

inguistics No.

9

p p ~

167-173. washing&,

D.

C.

Greenberg, JasrephH,

1957

*The nature anduet8 of

l i~guir~t ic

typalogf

e

e,

International

Journal

of

American Lin gui~ti c

2 3 . 2 ~ 6 8 - 1 7 .

BaLtfmora,

Gregg, John

R

1954.

The

anguage

rf

Taxonomy:

an

pplication

of

symbolic logic

t o he

study of cl&sificatory syateme. xi.,

pp.

New

Yorkt Columbia

~Tiversity

ress

Harria, Zellig, 1948, Componential amiysia of a Hebrew para-

digm, Language 24.X:87-9i. BaLthars. <

Haugatn, E h r ,

9 5 1

The a

emantit

s

of

Icslandic orianta tion,

Ward 13,

?:a?-459,

New Yark,

o c k h t t ,Charlea

F.

i956

Idiom formation,

p

2 2 2 - 2 2 9 frz

For Roman Jakobsan, The Haguer

Mouton,

-

Muusshulder,

Fred

W

r,

1 9 5 9 . Review of:

A

course in modern

linguistics, by Charles

F

Hockatt, Language 3 5 3 : 5 0 3 - 5 2 7

Brtlthora.

Jakobr~an,

Roman,

1936, Beitrag zur allgemeinen

KaauaLehre,

Travawx

du

Cercls Linguistiqut dt

Prague

6~240- 88

Prague.

- -

Joos, Martin

1956

Review of: Machine

translation

of languages:

Fourteen

essaye.

Ed,

William N.

Lacke

and

A.

Donald

Booth.

Language 3 2 2 : 2 9 3 - 2 9 8 .

Baltimore

958.

USemolagy:

a

lhguf stfc theory ofmeanhg.

Studies in

L3nguiatics

13,3-433-70,

Buffalo.



LexicographicaP Treatment of

Folk

Taxonomies

139

Jorden Eleanor Wars,

1 9 5 5

YThea p t a x

of Modern

Colloquial

~ a p n e e e ,

Language

3 f , l (part 3 : i-vf, 1-135

Baltimore.

Kellyr

George

A.

1955

The

sychology

of

Personal

Canrstructe,

Val.

l. A

Theory of Personality. xzii

566 pp.

New

York:

-

W. W ~oGton.

Landar, Herbert 3.

1 9 6 0

A note on accepted and rejected ar-

rangements of

Navaho wards. XnternatianaX Journal

o

American

Linguistics 26.4:

351 354

Baltimore.

Landart Herbert J., Susan M. Ervin, and Arnold E Horawftz,

1960.

Navaho color categorieba. Language

3 6 . 3 : 368-

382

Baltfnnore.

Lawrence, George

H . M .

1951. Taxonomy

of

Vascular Plants,

xiv,

8 2 3 pp.

New

Yarkt

MacrnilXan,

Zaaarsfeld,

Paul

F.

1961, uThe

algebra of dichatomous

systems,

pp.

11-

15

7

in Shxdia in Item

Analysi~ nd Prediction,

d.

Herbert

Solomon.

Stanford:

Stanford

Univer aity

Lees*

Robert

B. 1960

The

grammar

o

English naminaXfzatione,

hternatiunal Journal of American Lhguinrtic a 26 3

(part

XI):

i-xxvi,

1

205

(~ublicat ion

2

Indiana University Research

Center

in Anthropology,

Folklore,

and Limguistics).

Baltimar

e

Lameberg, Eric

H

nd

J a h n M .

Roberts. 1956 @Theanguage

of experisnce a study in methodology,

International

Journal

o American Lingui~ticeMemoir

No,

13 xi, 33

pp.

~a l t lmare ,

Lota

John. 1949.

The h~emantlc

analysia

of the nominal base8

fn

Travaux du e x c l e Linguistique de Copenhagua

5 : 185 - 191

Capenhague,

Lousbury, FloydG,

1956 .

'A

semantic analysis of

he

Pawnee

kinahip usage.

Language 32:158-

194. Baltimore.

1960.

Similarity and contiguity relations in

language

and culture.

Georgetown

University Monograph

Series

an

Languages

and inguistics

No.

12, pp. l23ci28.

~ a s h t ~ o n ,

Lyons, J, 1960,

Review of:

Language change and linguistic re-

construction,

by

Henry M

Haanigswald.

Bulletin of-

he

School

o

Oriental

and

African

Studies

27,

J:62X-622,

London.

-



Maclay* Howard, and Mary

D

Sleator. 1 9 6 0 . Responses to lan

guage: judgments of grammaticalnes

a .

International

Journal

of American

Linguistics

26 4: 275-

282

Baltimore.

Mayr, Ernet,

E.

Gorton

Lfnslsy,

and

Robert

It . Ubsingsr.

1953.

Methodo and

rinciples

of

Systematic Zoology, xl

3 2 8 pp.

New Pork: ~ c ~ r a w -

ill:

McKaughan Howard,

1 9 5 9 . Semantic

components

of pronoun

ayetems:

Maranaa.

Word

15.1:101-

102. Mew York.

orris,

Charles W. 1946. Signs, Language,

and

ehavior.

365

pp. New Yark: Prentice-Hall.

Nagel, Ernest.

1961

The

tructure of Science:-

roblem8

in he

Xu*

of c i m t i f i c xplanation. x v - 6 8 pp.

New

York

and

Burlingame: Harcourt, Brace and World.

Newman, Stanley, 1954. Semantic problems

In

grammatical

ayatems

and

Xexernes: a

search

far method.

pp.

8 2 9 3 in

Language

fn

culture,

ed,

Harry

Hoijer.

Chicago: University

of

~ h i c a ~ o ~ r e s s .

Nidar Eugene A. 1951. 'A

system

for

the

description af semantic

elernenta. Ward 7 1:1-14. New Yark.

1958.

Analysis

of

meaning

and

dictionary making.

hternational

Journal

of American Linguistics 24.4:279- 2 9 2 .

Baltimore.

ohman, Suzanne. 1953. YTheoriesof the 'Linguistic Field. '

Word

2: 123-

134 New York,

Quine,

Willard

Van Qrman. 1960 Word

and

Object,

xvi,

294

pp.

Cambridge Maes.,

and

New

York

and London:

The Technology

Pre as of M.

T.

and

John Wdley $t

Son8 Inc,

Sebeak, Thomas A, 1946. Finnish and Hungarian

case

systems:

th ir farm

and function.

Acta nstituti Hungarici

Univeraitatia

Holmiensis Series B Linguistics 3. Stockholm,

Simpon, George Gaylord. 1961.

Principles

of Animal Taxonomy.

x i v , 247

pp.

New Yark:

Columbia University

Pres~s .

Swadeah, Morris. 1946

Chftimacha,

pp. 3 1

2

336 in Linguistic

Structures

of

Native North America,

by Harry Haijer,

e t

al.

New

York:

Tiking Fund.



Lsxic ographical Treatment of Folk Taxonomic

a

141

Thomas,

David.

1955. Three

analyses of the nocano

pronoun

aystexn.

Wo r d

11.2:204- 208. e w York,

1957.

An

introduction

to

Manaaka

lexicography.

6

typssc

ipt pp.

Na uli, Philippines: S3L.

Voagelin, Charles

F.

and Fiarexace

M.

1957. &pi domains, a

lexical

approach to the problem

of

selection.

International

Journal

of

American

Linguistics

Memoir No.

14,

8

pp.

~altimor;.

W a l l a c e

Anthony

F.C. ndJohn Atkine. 1960.

&Themeaning

of

kinship American Anthropologist

62.1:58-80.

Men a s h

Wsfnreich,

Urie l .

1960.

'Mid-century

linguistics

a t t a h m a n t ~

andfrustxations. Romance Phikolagy 8 3r32Q-341. Berkeley,

We l l s Rulon,

1 9 5 8 .

a srtructural treatment

of meaning

possible? Proceedings

of

the Eighth biternational

Congress

f inguist~~ slo, 1 9 5 1 pp. 654-666. Oda ,

Wonderly William

L.

1952

YSemantic

components in Kechua

persanmorphemsa. Language 28 3 :366-376 . Baltimore.

Woadger, 3 .

M.

1 9 5 2 .

Biology and

an~uage: n intraduction

to

the methodology

of

-

he biological

ac

isnc

es

Gluding

medicge.

x?

364

pp.

Cambridge:

The

niversity Press .

Worth, D.

S. 1960,

Review of: Leksikologbja

axxglijskogo jazyka

by A, I

Smirnickfj.

Word 6 .2 :

277-

284, New Y

rk

NOTES

1.

The

work an which

ost

of this paper s based has been

supported

by

the

National

Science

Foundation,

A

number

of

students and other friends have offered cons ractive

csitfciam

of

an earlier draft

of

this statement.

For

especially helpful and

more detailed comments 1 a m particularly indebted

to

Y.R. Chaa,

David Crabb, Arthur Danto,

C .

0

Frake

Paul

Friedrich,

W,

,

Gaodenough,

3.

H.

Greenberg,

E b r

Haugen P, , Lazarsfeld,

F.G. Launsbury

and

Valney Steifire,

Documents

Conklin Folk Taxonomies