On the evolution of language, as exhibited in the specialization of the grammatic processes, the differentiation of the parts of speech, and the integration of the sentence; from a study of Indian languages

SMITHSONIAN INSTITUTION BUREAU OF ETHNOLOGY.

J. W.

POWELL,

DIRECTOR.

ON THE EVOLUTION OF LANGUAGE,

AS EXHIBITED IN

THESPECIALIZATION OFTHE

GBAMMATIC

PROCESSES,

THE

DIFFERENTIATION OF

THE

PARTS OF SPEECH,

AND THE

INTEGRATION OF THE

SENTENCE;

FROM

A STUDY OF INDIAN LANGUAGES.

J.

W. POWELL.

ON THE EVOLUTION OF LANGUAGE,

ASEXHIDITEDIN THESPECIALIZATION OFTHE GRAMMATICPROCESSES, THEDIFFER- ENTIATION OF THE PARTS OF SPEECH, ANDTHE INTEGRATIONOF THE SENTENCE

;

FROM A STUDY OF INDIAN LANGUAGES.

By

J. W. Powell.

Possible ideas

and

thoughtsare vastinnumber.

A

^distinct

word

for everydistinctidea

and

thought

would

requirea vast vocabulary.

The

problem inlanguage is toexpress

many

^ideas

and

thoughts with com- paratively fewwords.

Again,inthe evolutiou ofany language, progressis froma condition where fewideas areexpressed

by

afewwords to ahigher, where

many

ideas are expressed

by

the use of

many

words; but the

number

of all possible ideas or thoughts expressed is increased greatly out of pro- portionwith the increase ofthe

number

ofwords.

And

^still again, inall ofthoselanguageswhich have been most thor- oughly studied,aud

by

inference in all languages, ^it appears that the feworiginalwords usedinauy language remain as theelements forthe greater

number

finallyused. In the evolution of alanguagethe intro- duction of absolutely

new

material is a comparatively rare phenome- non.

The

oldmaterialiscombined

and

modified in

many ways

to form the new.

How

^has the small stock of words foundas thebasisof alanguage been thuscombined

and

modified?

The way

inwhich theoldmaterialshave been used givesrise to

what

will here be denominated

the

g-bamtvtatic processes.

I.—

THE PROCESS BY COMBINATION.

Two

or

more

words

may

be united to form a

new

one, or toperform the officeofa

new

one,

and

four

methods

or stages ofcombination

may

benoted.

a.

By

juxtaposition,

where

the twowords areplaced together

and

yet remain as distinctwords. This

method

is illustrated inChinese,where the wordsinthe combination

when

taken alone seldom give a clewto their

meaning when

placedtogether.

b.

By

compounding, where

two

words are

made

into one, in which case the original elements of the

new word

remain in an unmodified condition, asin housetop, rain-bow, tell-tale.

i

4

ON THE EVOLUTION OF LANGUAGE.

c.

By

agglutination, inwhich caseone or

more

of the elementsenter- ing into combinationtoform the

new word

is

somewhat

changed

—

^the

elements are fused together. Yet this modification is not so great as to essentiallyobscure theprimitivewords, asin truthful,where

we

easily recognize theoriginal words truth

and

full) and holiday, inwhich holy

and

dayare recognized.

d.

By

inflection.

Here

one or

more

of the elements enteringinto the

compound

has been so

changed

that it can scarcely be recognized.

Thereis a constanttendency to

economy

in speech

by

which wordsare gradually shortened asthey arespoken

by

generation aftergeneration.

In those words which are combinations of others thereare certain ele-

ments thatwear out

more

rapidlythanothers.

Where some

particular

word

is combined with

many

other different words the tendency to modify

by

wearthis oft-used elementis great. This ^is

more

especially the case wherethe combined

word

isused in certain categoriesof com- binations, aswhere particular words are used to denote tense in the verb; thus, did

may

be used in combination with a verbto denote past timeuntilit is

worn down

to the sound of d.

The same wear

occurs where particularwordsareused toform cases innouns,

and

a varietyof illustrationsmight begiven. These categories constituteconjugations and declensions, and for convenience such combinations

may

becalled paradigmatic.

Then

^theoft-repeatedelements of paradigmatic combina- tions areaptto

become

excessively

worn

and modified,sothattheprimi- tivewordsorthemestowhich they are attached seemtobebutslightly changed by the addition.

Under

these circumstances combination is calledinflection.

As

amorphologic process, no well-defined plane of demarkation be- tween these fourmethodsofcombinationcan be drawn,asone runsinto another^;but, ingeneral, words

may

besaid tobe juxtaposed

when

two wordsbeing placed together the combination performs thefunction of a

new

word, while inform thetwo words remain separate.

Words may

be said tobe

compound when

two or

more

wordsarecom- binedtoformone, nochangebeing

made

in either.

Words may

besaid tobeagglutinated

when

theelementary words arechangedbutslightly,

i.e.,only to the extent that their original forms are not greatly ob- scured;

and

words

may

besaid to beinflected

when

inthe combination theoft-repeated element or formative parthas beenso changedthatits origin is obscured. These inflections areusedchiefly in the paradig- matic combinations.

In the preceding statementithasbeen

assumed

that therecanberec- ognized, in these combinations of inflection, a

theme

or root, as it is

sometimescalled,

and

a formative element.

The

formativeelement ^is used with a great

many

^different wordsto define orquahfy

them

; that

is,to indicatemode,tense,number,person,gender,etc.,ofverbs, nouns,

and

other parts ofspeech.

When

ⁱⁿa languagejuxtaposition isthe chief

method

ofcombination,

poweli.]

COMBINATION VOCALIC MUTATION. 5

there

may

also be distinguished two kinds of elements,in

some

sense corresponding tothemes and formativeparts.

The theme

^isa

word

the meaning of which ^isdetermined

by

the formative

word

placed

by

it

;

that ^is, the theme^is a

word

having

many

radically different meanings

;

with which

meaning

it is to be understood ^is determined only

by

the formative word, whichthus serves as itslabel.

The ways

ⁱⁿ which the

theme

wordsare thus labeled

by

theformative

word

are very curious, but the subjectcannotbe enteredintohere.

When

words arecombined

by

compounding, the formative elements cannot so readilybedistinguishedfromthe

theme

^; norforthepurposes under immediateconsideration can

compounding

bewellseparatedfrom agglutination.

When

words are combined

by

agglutination,

theme

and formative part usually appear.

The

formative parts are affixes; and affixes

may

be dividedintothreeclasses,prefixes,suffixes,

and

infixes. Theseaffixes are often called incorporatedparticles.

In those Indian languageswhere combination ^is chieflybyagglutina- tion, that is,

by

the use of affixes, i. c, incorporated particles, certain parts of the conjugation of the verb, especially those which denote gender, number,

and

person,are effected bythe use ofarticlepronouns

;

but in thoselanguages wherearticlepronouns arenot found the verbs are inflected to accomplish the

same

part of their conjugation. Per- haps,

when we come more

fullyto studytheformativeelementsinthese more highly inflected languages,

we may

^{discover in} such elements greatly modified,i. e.,

worn

out,incorporated pronouns.

II.—

THE PROCESS BY VOCAEIC MUTATION.

Here,in orderto form a

new

word, one or

more

of thevowels of the old

word

arechanged, as in

man —

men, where an e is substituted fora;

ran

—

run,whereuissubstitutedfora; lead

—

^led,where^e,with ^itsproper sound, issubstituted foreawith itsproper sound. This

method

is used to a verylimitedextent in English.

When

the history of the words in

which itoccurs is studied itis discoveredto be but an instance of the wearingout ofthe differentelementsofcombined words^; butintheHe-

brew

this

method

prevails toa very largeextent,and scholars havenot yet been able todiscoverits origin incombination astheyhavein Eng-

lisli. It

may

or

may

nothave been an ^originalgrammatic process,but

because ofitsimportance incertain languages ithas been found neces- sary todeal with it asa distinct

and

original process.

ON THE EVOLUTION OF LANGUAGE.

III.—

THE PROCESS BY INTONATION.

In English,

new

wordsare notformed

by

this method, yetwords are iutonedforcertain purposes, chiefly rhetorical.

We

^{use the rising in-}

tonation(or inflection, asit isusually called) toindicatethataquestion

is asked, and variouseffects aregiven to speech

by

thevarious intona- tions of rhetoric.

But

this process is used in other languages toform

new

words with which toexpress

new

ideas. In Chineseeight distinct intonations are found, by the use of which one

word may

be

made

to express eight different ideas, orperhapsit is better to say that eight words

may

be

made

of one.

IV.— THE PROCESS BY PLACEMENT.

The

place orposition of a

word may

affect its significant use.

Thus

in English

we

sayJohn struck James.

By

theposition ofthosewords toeach other

we know

^that

John

is theactor,

and

that

James

receives theaction.

By

the grammatic processes language is organized. Organization postulates thedifferentiation of organs

and

their combination iutoin- tegers.

The

integers of language are sentences,and theirorgans are the parts of speech. Linguisticorganization, then, consistsin the dif- ferentiation of the parts of speech

and

the integration of thesentence.

Forexample, letus take thewords John,father,

and

love. John^is the

name

of an individual; love is the

name

of amentalaction,

and

father the

name

of a person.

We

put

them

together,

John

loves father,

aud

they express athought; John becomesa noun,

aud

is thesubjectofthe sentence; lovebecomes a verb,

and

is thepredicant_; father a noun,

and

isthe object; and

we now have

an organized sentence.

A

sentence ^re- quires parts of speech,

aud

parts of speech are suchbecause theyare used asthe organic elements of asentence.

The

criteriaof rankinlanguagesare, first,grade oforganization, ^i.e.,

the degreeto which the grammaticprocesses

and methods

are special- ized, and the parts of speech differentiated; second, sematologic con- tent, that is, the

body

of thoughtwhich the languageis competent to convey.

The

grammaticprocesses

may

be usedfor threepurposes:

First, forderivation,

where

a

new word

to express a

new

ideais

made

by combining two or

more

old words,or

by

changing the vowel^of one word, or

by

changingthe intonation of one word.

powell.]

INTONATION — ^PLACEMENT.

⁷

Second, for modification, a

word may

be qualified or defiued

by

the processes of combination, vocalicmutationor intonation.

It should bere benotedthat the plane betweenderivation and quali- fication is notabsolute.

Third,for relation.

When

words as signsof ideas are used together to express thought, therelationofthe words

must

beexpressed

by some

means. In English therelation of words ^is expressed both

by

place-

ment and

combination, ^{i. e.,}inflection foragreement.

Itshould here benoted that paradigmaticinflections areusedfor two distinctpurposes, qualification and relation.

A word

^isqualified

by

in- flection

when

the ideaexpressed

by

the inflection pertains to the idea expressed

by

the

word

inflected; thus a

noun

is qualified

by

inflection

when

its

number and

gender are expressed.

A ^word

^{is related}

^by

^in-

flection

when

the office of the

word

in the sentence ^is pointed out thereby; thus, nounsare related

by

caseinflections ; verbs are related

by

inflections for gender, number, and person. All inflection foragree-

ment

is inflectionfor relation.

In English, three of thegrammatic processes are highlyspecialized.

Combination is used chiefly for derivation, butto

some

slight extent forqualification

and

relation in the paradigmatic categories.

But

its

use^{in this}

manner

as

compared

with

many

^{otherlanguages} has almost disappeared.

Vocalicmutationisusedtoaverylimited extent

and

only

by

accident,

and

can scarcelybesaid to belong to the English language.

Intonationisused as agrammatic process onlyto a limited extent

—

simplyto assist in forming theinterrogativeandimperativemodes. Its use hereis almostrhetorical; in all other casesit ispurelyrhetorical.

Placementis largely used in the language, and ^is highly specialized, performing theofficeof exhibiting the relations of words to each other in the sentence; i. e., it is used chiefly forsyntacticrelation.

Thus

one of the four processes does not belongto the English lan-

guage; the others are highlyspecialized.

The

purposesfor whichthe processes are used are derivation, modifi- cation, and syntactic relation.

Derivation is accomplishedbycombination.

Modification ^isaccomplished

by

the differentiation of adjectives and adverbs, aswords, phrases, and clauses.

Syntactic relation is accomplished

by

placement. Syntactic relation

must

not be confounded with the relation expressed

by

prepositions.

Syntacticrelation is the relationofthe parts of speech toeach other as integralparts ofasentence. Prepositions expressrelationsof thought ofanother order.

They

relate words^toeach otheras words.

Placementrelates words toeach other asparts ofspeech.

In the Indian tongues combination isusedforall threepurposes, per- formingthe threedifferent functionsofderivation, modification,

and

re-

8

ON THE EVOLUTION OF

LANGXIAGE.

lation. Placement,also, is used for relation, andfor both kinds ofrela- tion, syntactic

and

prepositional.

With

regard, then, to the processes

and

purposes for whichthey are used,

we

find in the Indian languages a low degree of specialization _; processes areused fordiverse purposes, and purposesareaccomplished

by

diverseprocesses.

DIFFERENTIATION OF THE PARTS OF SPEECH.

Itis nextin orderto considerto

what

degree the parts of speech are differentiated in Indian languages,as

compared

with English.

Indian nouns are extremely conuotive, that is, the

name

does

more

than simply denote the thing towhich ^it belongs; in denoting theob- jectit also assigns to it

some

qualityor characteristic.

Every

object has

many

qualities

and

characteristics,

and by

describingbuta part of thesethetrueofficeofthe

noun

isbutimperfectly performed.

A

^strictly

denotive

name

expressesno one quality or character, but embraces all qualities and characters.

In TJte the

name

for bear is he seizes, orthehugger. In thiscase the verb isused forthe noun, and in so doing the Indian

names

the bear

by

predicatingoneof his characteristics.

Thus noun and

verbare un- differentiated. In Seneca the north isthesunnever goes there,

and

this sentence

may

be used as adjective or

noun

; in such, casesnoun, adject- ive,verb,

and

adverb are found as onevocable orword,

and

the four parts ofspeech are undifferentiated. In thePavdnt language aschool- houseis called po-hunt-in-in-yi-Mn.

The

^firstpart of theword,po-M,nt, signifies sorcery ispracticed,

and

is the

name

given

by

the Indians to any writing, from thefact that

when

theyfirst learned of writing they supposed it to bea

method

of practicing sorcery^; in-in-yi is the verb signifyingto count, and the

meaning

ofthe

word

has been extendedso as to signify to read:

Mm

^signifies

wigwam, and

^{is derived} from the verbkdri, tostay.

Thus

the

name

ofthe school-house literally signifies astayingplacewheresorceryis counted, or where papers areread.

The

Pavcmtin

naming

a school-house describes the purpose for whichitis

used. These examples illustrate the general characteristicsof Indian nouns; they are excessively conuotive; a simply denotive

name

^israrely found. In generaltheirname- wordspredicate

some

attribute oftheob- jectnamed,

and

thusnoun,adjective,andpredicantare undifferentiated.

In

many

Indian languages there is no separate

word

for eye, hand, arm,orother partsandorgans of the body, butthe

word

isfoundwithan incorporatedorattachedpronounsignifying

my

hand,

my

eye; yourhand, your eye^; his hand,Ids eye, etc., as the case

may

be. If theIndian, in

naming

theseparts, refers to his

own

body, he says

my

^; if herefers to

powbll.]

ARTICLE PRONOUNS ^GENDERS.

9 the

body

of the person to

whom

he ^{is speaking,} he says your, &c. If an Indian shouldfind a detachedfootthrown fromtheamputating-table of an

army

field hospital, he would say something like this: I have found

somebody

his foot.

The

^linguisticcharacteristiciswidely spread, thoughnotuniversal.

Thus

theIndian has no

command

of a fully differentiated

noun

ex- pressive ofeye, hand, arm, orother parts

aud

organsofthe body.

In the pronouns

we

often havethe mostdifficult part of an Indian language. Pronounsareonly toa limitedextent independentwords.

Among

^thefreepronounsthestudent

must

early learn todistinguishbe- tween the personalaudthedemonstrative.

The

demonstrativeprououus are

more commonly

used.

The

Indian ^is

more

accustomed to say this person or thing, that person or thing, than he, she, or it.

Among

the freepersonalprououusthe student

may

fiud an equivaleut of the pro-

noun

I,anothersignifying

I

andyou; perhaps anothersignifying

I

and

he,

and

one ^signifying we, more than two, including the speaker aud thosepresent;

and

another including the speaker and personsabsent.

He

will alsofind personal prououusin the second

and

thirdperson,per- haps with singular, dual, and plural forms.

To

a large extent theprououus are incorporated in theverbsas pre- fixes,infixes, or suffixes. In such cases

we

will call

them

article pro- nouns. These article pronouns ^pointout with great particularity the persou, number,

and

gender, both of subject

and

object,

and

sometimes oftheindirect object.

When

thearticlepronounsareusedthe persoual pronouns

may

or

may

^not be used; but itisbelieved that the personal prououus will always be found. Article pronouns

may

uotalways be found. In those languages which are characterized by

them

they are used alike

when

the subject

and

object nounsare expressed

and when

they areuot.

The

student

may

at firstfind

some

difficulty with these articlepronouns. Singular,dual,

aud

pluralformswillbefound. Some- times distinctincorporated particles will be usedfor subjectandobject, butoften thiswill not be the case. If the subject only^is expressed, one particle

may

be used; if the object ouly is expressed, another particle; but if subject

and

object are expressed an entirely different particle

may

standforboth.

But

^{it is} in the genders of these article pronouns that the greatest difficulty

may

be found.

The

student

must

entirely free his

mind

of the ideathatgender is simply adistinction of sex. InIndian tongues, genders are usuallymethodsofclassificationprimarilyintoanimate and inanimate.

The

animate

may

be again divided into male

and

female, but thisisrarely the case. Often

by

thesegendersall objectsare.classi- fied bycharacteristicsfound intheir attitudesor supposedconstitution.

Thus we may

have the animate

aud

inanimate,oneor both, divided into the standing, thesitting, and the lying; orthey

may

bedivided intothe watery, the mushy, the earthy, the stony,the icoody, and thefleshy.

The

gender of these article pronouns has rarely been

worked

out in any

10 ON THE EVOLUTION OF LANGUAGE

language.

The

extenttowhich theseclassifications enter intothearti-

cle-pronounsisnot wellknown.

The

subject requires

more

thorough study. These incorporated ^particles are here called articlepronouns.

In the conjugation of the verb they take an importantpart, and have

by some

writers been called transitions. Besides pointing out withpar- ticularity the person, number,

and

gender or the subject and object, they perform the

same

offices that are usuallyperformed

by

thosein- flectionsof the verb that occur to

make them

agreein gender, number,

and

person with the^subject. Iu those Indian languages wherethearti- cle pronouns are not found, and the personal pronouns onlyare used, theverb isusuallyinflected to agreewith the subject orobject, orboth, inthe

same

particulars.

The

articlepronouns as they point out person, number, gender,

and

case of the subject

and

object, are not simple particles, butare toa greater or lesser extent

compound;

their

component

elements

may

be brokenapart

and

placedin differentparts ofthe verb. Again, thearti- clepronounin

some

languages

may have

its elements combinedintoa distinct

word

in such a

manner

thatit will notbeincorporatedin the verb, but willbeplacedimmediatelybeforeit. Forthisreason theterm

articlepronoun has been chosenratherthanattachedpronoun.

The

older term, transition,

was

given to

them

because oftheiranalogyinfunction to verbalinflections.

Thus

the verb ofan Indian languagecontains within itself incorpor- atedarticle pronouns which pointout with greatparticularity the gen- der,number,

and

person of the subject

and

object. In this

manner

verb, pronoun,

and

adjective are combined,

and

to this extent these parts of speech are undifferentiated.

In

some

languages thearticlepronounconstitutesadistinctword,but whetherfree orincorporated ^itis a complex ^tissueofadjectives.

Again, nouns sometimes^{containparticles within themselvesto predi-} cate possession,

and

to thisextentnouns

and

verbs areundifferentiated.

The

verbisrelativelyof

much

^{greaterimportanceinanIndian tongue} than in a civilized language.

To

a large extent thepronoun ^is incor- porated in the verb as explained above, and thus constitutes a part of

its conjugation.

Again, adjectives are used as intransitive verbs, as in

most

Indian languagesthereisno verb tobeused asa predicant or copula. Where,

iu English

we

would say the

man

^is good the Indian would say that

man

good, using the adjective as an intransitive verb, ^i.e., asa predi- cant. If hedesired to affirmitin the past tense, theintransitive verb good would be inflected, or otherwise modified, to indicate the tense;

andso, in likemanner,alladjectives

when

usedtopredicatecanbemodi- fled to indicate mode, tense, number, person,

&c,

asother intransitive verbs.

Adverbs

are used as intransitive verbs. In English

we may

say he is there; the Indian would saythatperson there usually preferring

powei.ll

WORDS USED

AS INSTRANSITIVE VERBS. 11 the demonstrative to the personal pronoun.

The

adverb there would, therefore,be used asa predicant orintransitiveverb, and might becon- jugated todenotedifferentmodes,tenses,numbers,persons,etc. Verbs

will often receive adverbial qualifications

by

the use of incorporated particles, and, still further,verbs

may

^{contain within} themselves ad- verbial limitationswithout ourbeing ableto tracesuchmeaningstoany definite particles orparts ofthe verb.

Prepositions are intransitive verbs. In English

we may

saythehat is on the table; the Indian would say that hat on table; or he might change the order,

and

say that hat tabic on; bnt the preposition on would be used as an intransitive verb to predicate, and

may

be conju- gated. Prepositions

may

often be found as particles incorporated in verbs, and, still further, verbs

may

contain within themselves preposi- tional meaningswithout our being ableto tracesuch meanings to any definite particles within the verb.

But

the verb connotes such ideas thatsomethingisneededto completeitsmeaning,thatsomething being alimiting or qualifying word, phrase, or clause. Prepositions

may

be prefixed,infixed, or suffixed tonouns, ^{i. e.,}they

may

be particles incor- porated in nouns.

Nouns may

be used as intransitiveverbs under the circumstances

when

inEnglish

we

would use a

noun

as the

complement

of asentence aftertheverb to be.

The

verb, therefore, often includes withinitselfsubject,direct object, indirect object, qualifier,

and

relation-idea.

Thus

itis that the study ofanIndian language is, toa largeextent, thestudy ofitsverbs.

Thus

adjectives, adverbs, prepositions,

and

nouns are used as ^in- transitiveverbs; and, tosuch extent, adjectives, adverbs, prepositions, nouns, and verbs are undifferentiated.

From

the remarks ^{above, it}will be seen that Indianverbs often in-

clude within themselves meanings which iu English are expressed

by

adverbs and adverbialphrases

and

clauses.

Thus

theverb

may

express withinitself direction, manner, instrument,

and

purpose, one or all, as the verb to go

may

be represented by a

word

signifying gohome; an- other,go

away from

home; another,goto a place other than home; an- other, go

from

a place other than home; one, go

from

^this place, with reference to

home;

one, togo up; another, togo

down;

one,goaround;

and, perhaps, there will be a verb go up hill; another, go up a valley

;

another, go up a river, etc.

Then we may

have to go onfoot,to go on horseback, to go in a canoe; still another, to go for water; another for wood,etc. Distinctwords

may

beused forall these,or a fewer

number

used, and these varied

by

incorporated particles. Inlike manner, the Englishverbtobreak

may

be representedbyseveralwords, eachofwhich

will indicatethe

manner

of performing the actor the instrument with which itis done. Distinctwords

may

beused, ora

common word

varied with incorporated particles.

The

verb to strike

may

^be represented

by

several words, signifying

12 ON THE EVOLUTION OP LANGUAGE.

severallytostrikewiththefist, tostrike withaclub, to strikewiththe open hand, tostrikewitha whip,tostrikeivithaswitch,to strike with aflatin-

strument, etc.

A common word may

be used with incorporatedparti- clesor entirely differentwords used.

Mode

inan Indian tongueisa rather difficultsubject.

Modes

analo- goustothose of civilized tongues arefound,

and many

conditions

and

qualificationsappearin the verb which in English

and

othercivilized

languages appear as adverbs, and adverbial phrases

and

clauses.

No

plane of separation canbe

drawn

between suchadverbial qualifications

and

truemodes.

Thus

there

may

be a form of the verb, which shows that the speaker

makes

a declaration as certain, i. e., an indicative

mode;

another which shows thatthespeaker

makes

a declaration with doubt,i. e.,a ^dubitative

mode

^; another that he

makes

^adeclaration on hearsay, ^{i. e.,}a quotative

mode

; anotherform willbe used in

making

a

command,

^giving

an

imperative

mode

^; anotherin imploration, ^i. e., an implorative

mode;

anotherform todenote permission, ^i. c, a permissive

mode;

^{anotherinnegation, i. e., a negative}

mode;

anotherform willbe usedto indicatethatthe actionis simultaneouswith

some

otheraction,

i. e.,asimulative

mode;

anotherto denote desire orwishthat something bedone, i. e., a desiderative

mode

^; another that theaction oughtto be done, i. e.,

an

obligative

mode

^; another that action is repetitive fromi timeto time, i. e.,afrequentative

mode

; another thataction iscaused,

i. e.,acausativemode, etc.

Theseforms of the verb,which

we

are compelledto callmodes, are of greatnumber. Usually with each of

them

aparticular

modal

particle orincorporatedadverb willbe used; butthe particular particle which gives thequalified

meaning may

not alwaysbediscovered; andin one languageadifferent

word

will beintroduced, whereinanother the

same word

will beused with anincorporatedparticle.

Itis statedabovethatincorporatedparticles

may

be used^toindicate direction,manner,instrument,

and

purpose^;infact,anyadverbialquali- fication whatever

may

be

made by

an incorporated particle instead of an adverb asadistinct word.

No

lineof demarkation canbe

drawn

between these adverbialparti- cles

and

those mentioned above as modal particles. Indeeditseems bestto treatall these forms of the verb arisingfromincorporated par- ticlesas distinctmodes. In this sense, then, anIndian lauguagehas a multiplicity of modes. It should be further remarked that in

many

cases thesemodal or adverbial particles are excessively worn, sothat they

may

appear as additions orchangesof simplevowelorconsonant sounds.

When

incorporatedparticles are thus used, ^distinctadverbial words, phrases,or clauses

may

alsobe employed,

and

the ideaexpressed twice.

In an Indianlanguageitisusuallyfouud^difficulttoelaborate asystem oftenses inparadigmatic form.

Many

tenses ortimeparticles arefound incorporated in verbs.

Some

of these time particles are excessively

powell.]

VOICE — ^MODE

^TENSE.

¹³

worn,

and may

appear rather as inflections than as incorporated^par-

ticles. Usually ratherdistinct present, past, and futuretensesare dis- covered; often a remoteorancient past,

and

lessoften an immediate future.

But

great specificationof time iu relation tothe present

and

inrelation toothertime is usually found.

It

was

seen abovethat adverbialparticles cannot be separatedfrom luodalparticles. Inlike

manuer

tense particles cannot be separated from adverbial

and

modal particles.

In anIndianlanguage adverbsare differentiatedonlytoa limitedex- tent. Adverbial qualifications arefoundin theverb,

and

thusthere are a multiplicityof

modes and

tenses,

and

noplane of demarkation can be

drawn

between

mode and

tense.

From

preceding statementsitwillap- pear that a verbⁱⁿan Indiantongue

may

have incorporated with ita greatvariety ofparticles,which can be arrangedinthreegeneralclasses,

i. e.,pronominal, adverbial, audprepositional.

The

pronominalparticles

we

havecalled articlepronouns^; they serve topointout avarietyof characteristicsinthe subject, object,

and

indi- rectobjectof the verb.

They

thus subservepurposes whichinEnglish aresubserved

by

differentiated adjectives as distinct parts of speech.

They

might, therefore, with

some

propriety, have beencalled adjective particles,buttheseelementsperformanotherfunction;they serve thepur- pose which ^is usuallycalledagreementinlanguage; thatis, they

make

the verb agree with thesubject

and

object,andthusindicatethesyntac- ticrelationbetween subject, object,

and

verb. Inthissensetheymight with propriety have beencalled relation particles,

and

doubtless this function

was

in

mind when some

of the older

grammarians

called

them

transitions.

The

adverbial particles perform the functions of voice, mode, and tense, togetherwith

many

other functions that are performed in lan-

guagesspoken

by more

highlycivilizedpeople

by

differentiated adverbs, adverbial phrases,

and

clauses.

The

prepositional particles performthe function of indicating a great variety of subordinate relations, like the prepositions usedas distinct parts ofspeechin English.

By

thedemonstrative function of

some

of the pronominal particles, they areclosely relatedto adverbial particles,

and

adverbial particles are closely related to prepositionalparticles,sothat,itwillbe sometimes

difficultto say of aparticular particlewhether it be pronominalor ad- verbial,

and

of another particular particle whetherit be adverbial or prepositional.

Thus

the threeclassesofparticlesare not separated

by

absoluteplanes of demarkation.

The

use of these^particlesas parts of theverb; the useofnouns,ad- jectives, adverbs, and prepositionsasintransitiveverbs;

and

thedirect use of verbs as nouns, adjectives,

and

adverbs,

make

thestudy ofan Indian tonguetoalargeextent thestudy ofitsverbs.

14 ON THE EVOLUTION OF LANGUAGE.

To

the extent that voice, mode,

and

tense are accomplished

by

the useofagglutinatedparticles or inflections, to that extent adverbs

and

verbs are undifferentiated.

To

the extent that adverbs are found as incorporated particles in verbs, thetwo parts of speechare undifferentiated.

To

the extent that prepositionsare particlesincorporatedin theverb, prepositions

and

verbs are undifferentiated.

To

theextent thatprepositionsare affixed tonouns, prepositions

and

nouns are undifferentiated.

Inall these particularsit is seen that the Indian tonguesbelongtoa verylowtype oforganization. Various scholars

have

called attention to this feature by describing Indian languages as being holophrastic, polysynthetic, or synthetic.

The

term syntheticis perhaps the best,

and may

be usedas

synonymous

with undifferentiated.

Indian tongues, ^therefore,

may

be said tobe highly syntheticinthat theirpartsofspeech areimperfectlydifferentiated.

Inthese

same

particularsthe English language ishighly organized, asthe parts of speech are highly differentiated. Yet the differenceis

oneof degree, notof kind.

To

the extent intheEnglish languagethatinflectionisusedfor quali- fication, as for person, number, and gender^{of the} noun

and

pronoun,

and

for

mode and

tensein theverb, to that extent the parts of speech areundifferentiated.

But we

have seen that inflectionis usedfor this

purposeto a veryslight extent.

There is yet ⁱⁿ the English language one important differentiation which has been but ^partially accomplished. Verbs as usually consid- ered areundifferentiatedpartsofspeech^; theyarenouns

and

adjectives, oneorboth,

and

predicants.

The

predicant simple^is adistinctpart of speech.

The

English language has butone, the verb to be, and thisis

notalways apurepredicant, for it sometimes contains within itselfan adverbial element

when

^itis conjugatedfor

mode

and tense, anda con- nectiveelement

when

^itis conjugated for agreement.

With

adjectives and nouns thisverb ^isused asapredicant. In thepassivevoice alsoit isthus used,

and

the participles are nouns or adjectives. In

what

is

sometimes called the progressive form of the activevoice nouns

and

adjectives are differentiated intheparticiples,

and

the verb "to be"^is used as a predicant.

But

iu

what

^{is usually}denominated theactive voice of the verb, the English language has undifferentiated partsof speech.

An

examination ofthehistory of the verb tobeinthe English languageexhibitsthefactthatit iscoming

more and more

tobe used ^as the predicant;

aud what

isusuallycalledthe

common

formofthe active voiceis coming

more and

moretobe limited in its use to specialsig- nifications.

The

real active voice, indicativemode,presenttense,firstperson, sin- gular number,of the verb to eat, is

am

eating.

The

^expression

I

eat, signifies

I am

^{accustomedto eat. So, if}

^we

consider the

common

formof

powei.!..1

THE VERB

TO BE.

15

the active voice throughout its entire conjugation,

we

discover that

many

^ofits forms are limited to special uses.

Throughouttheconjugationoftheverb theauxiliariesarepredicants, but these auxiliaries, to the extent that they are modified for mode, tense, number,

and

person, contain adverbial andconnective elements.

In like

manner many

^{of thelexical}elementsofthe Englishlanguage contain

more

than one partof speech: To ascendistogoup; to descend

is to go

down; and

to departis togo from.

Thus

itis seen that the Englishlanguage isalso synthetic inthatits parts of speech are not completelydifferentiated.

The

English, then, differsin thisrespectfroman Indianlanguageonlyindegree.

In most Indian tonguesno pure predicant has been differentiated, but ⁱⁿ

some

the verb to be, or predicant, has been slightlydeveloped, chiefly to aflinn existencein aplace.

It willthusbeseenthat

by

thecriterionoforganizationIndian tongues are ofvery low grade.

Itneed but tobeaffirmed that

by

the criterionofsematologic content Indian languages areof a very lowgrade. Therefore the frequently- expressed opinion that the languagesof barbaric peoples havea

more

highly organized grammaticstructure than the languagesof civilized peoples hasits completerefutation.

It is worthy of remark that all paradigmatic inflection inacivilized

tongueis arelicof its barbariccondition.

When

the parts of speech are fullydifferentiated

aud

the process of placementfully specialized, sothat theorder of wordsin sentenceshasitsfull significance, no ^use-

fulpurpose issubserved byinflection.

Economy

inspeech ^is the force

by

which itsdevelopment has been accomplished,

and

it divides itself properly into

economy

of utterance and

economy

ofthought.

Economy

ofutterance has

had

todowith the phonicconstitution of words^;

economy

of thought has developed the sentence.

All paradigmatic inflection requires unnecessary thought. In the clause _ifheteas here, if fully expresses the subjunctive condition,

and

itisquiteunnecessaryto expressita second time

by

usinganotherform of the verb to be.

And

so the people

who

are using the English lan- guage are deciding, for the subjunctiveform is rapidly becoming obso- lete with the longlistofparadigmaticforms which havedisappeared.

Every

time the pronoun he, she,oritis used it is necessaryto think of the sexof its antecedent, though in its use there is no reason

why

sexshould be expressed,say, onetimeinten thousand. Ifonepronoun non-expressive of gender were used instead of the three, with three gender adjectives,then inninethousand nine

hundred

and ninety-nine cases the speaker wouldbe relieved of the necessityof an unnecessary thought,andintheonecaseanadjectivewouldfullyexpressit.

But when

these inflections are greatly multiplied, as theyare inthe Indian lan- guages, alike with theGreek

and

Latin, thespeakeris compelled inthe

16

ON THE EVOLUTION OP LANGUAGE.

choice ofa

word

toexpress

Ms

^{idea to}think of amultiplicity of things which have no connection with thatwhich he wishes to express.

A

Porika Indian, in saying that a

man

killedarabbit, would haveto say the man, he, one, animate, standing, in the nominative case, pur- posely killed, by shooting an arrow, the rabbit,he, the one, animate, sitting,in the objectivecase^; fortheform of averb tokill

would

have to be selected,

and

the verb changes its form

by

inflection

and

incor- poratedparticles to denote person, number, and gender as animate or inanimate,

aud

gender as standing,sitting, orlying,

aud

case^;

and

the form of the verbwould^also expresswhether thekilling

was

doneacci- dentallyorpurposely,

and

whetherit

was by

shootingor

by some

^other

process,and, if

by

shooting, whether

by bow ^and

^{arrow,orwith a}

gun

;

andthe form of the verb wouldin like

manner

have to express all of these thingsrelating tothe object; thatis, the person,number,gender,

and

case of theobject^;

and

from the multiplicityofparadigmatic forms oftheverb tokill this particular one would

have

tobeselected. Per- haps onetimeinamillionitwouldbe thepurposetoexpress all ofthese particulars,

and

in that case the Indianwould havethe wholeexpres- sioninone compactword,but^iu the ninehundred

and

ninety-nine thou- sandnine

hundred and

ninety-nine casesallof theseparticulars would havetobe thoughtofinthe^selection of the formof theverb,

when

no valuablepurpose wouldbeaccomplished thereby.

In thedevelopmentofthe English, aswellastheFrenchand

German,

linguisticevolution has notbeenin vain.

Judged by

these criteria, the English stands alone in the highest rank; but as a written language, in the

way

in which its alphabet is used, the English hasbut

emerged

from a barbariccondition.