7/30/2019 Bloomfield Sentence and Word http://slidepdf.com/reader/full/bloomfield-sentence-and-word 1/12 Sentence and Word Author(s): Leonard Bloomfield Reviewed work(s): Source: Transactions and Proceedings of the American Philological Association, Vol. 45 (1914), pp. 65-75 Published by: The Johns Hopkins University Press Stable URL: http://www.jstor.org/stable/282688 . Accessed: 15/03/2013 11:32 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected]. . The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to Transactions and Proceedings of the American Philological Association. http://www.jstor.org This content downloaded on Fri, 15 Mar 2013 11:32:35 AM All use subject to JSTOR Terms and Conditions
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Author(s): Leonard BloomfieldReviewed work(s):Source: Transactions and Proceedings of the American Philological Association, Vol. 45 (1914),pp. 65-75Published by: The Johns Hopkins University Press
Stable URL: http://www.jstor.org/stable/282688 .
Accessed: 15/03/2013 11:32
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp
.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
.
The Johns Hopkins University Press is collaborating with JSTOR to digitize, preserve and extend access to
Transactions and Proceedings of the American Philological Association.
http://www.jstor.org
This content downloaded on Fri, 15 Mar 2013 11:32:35 AMAll use subject to JSTOR Terms and Conditions
THE first ask of the linguistic nvestigator s the analysisof a language into distinctive ounds, theirvariations,andthe like. Whenhe has completedthis,he turns o the analy-
sis of the semanticstructure,to whatwe call the morphol-ogy and syntax of the language, its grammatical system.The method generallypursued in this semantic analysis isadmittedly makeshift: we adhere to the process of syn-theticdescriptionwhichhas been developedout ofthepracticeof the Alexandrine and Roman grammarians. Taking thesingle word as our unit,we name the big classes of words(parts of speech) and then describe the inflection f each;there follows a hasty survey of such mattersas derivationand composition; finallywe discuss the uses and inter-relations of the various inflectedwords in the sentence
(syntax).This procedure s a makeshift, or it has long been recog-
nized that the first nd original datum of language is thesentence, that the individual word is the productof a the-
oretical reflectionwhich ought not to be taken forgranted,and,further,hatthegroupingofderivedand inflectedwordsintoparadigms, nd the abstraction froots, tems,affixes, rotherformative rocesses, s again the resultof an even morerefinedanalysis. It needs but little scientificreflection omake us realize that the grammarianought by no meansto extract such products withmagic suddenness, live and
wriggling,out of the nafve speaker's hat. This has longbeen recognized. Wilhelmvon Humboldtbegins his discus-sion of polysynthetic anguages (Uber die Verschiedenheitdes menschlichen prachbaues, , paragraph 17) by saying:"Wenn man,wie es ursprunglich ichtigerst,da jede, nochso unvollstandigeAussage in der Absicht des Sprechendenwirklich inen geschlossenenGedanken ausmacht,vom Satze
This content downloaded on Fri, 15 Mar 2013 11:32:35 AMAll use subject to JSTOR Terms and Conditions
ausgeht,so zerschlagenSprachen,welche sich dieses Mittels
bedienen, die Einheit des Satzes gar nicht, . . . " Increased
psychologic understandinghas only confirmedthis greatscholar's intuition. Since we have learned to distinguishbetween an investigator'sogical analysis afterthe fact and
theactual psychicoccurrence, nd to observethe latterwith-
out confusing twithsuch logical analysis,we find t obviousand easily provedthat in most of our speaking we are con-
scious of the whole sentence only, not of the words into
which it may be divided. The experiment s easily made:one asks a speaker to tellhowmanywordshe has used in the
casual sentence ust spoken. The answer, f it comes at all,will be surprisinglyong in preparing,- and this withour
ceaseless training, hroughoutour reading and writing, nthis formof linguistic nalysis. I need hardlyreferto the
fact,so well illustratedby Brugmann Grzundriss,I2, I, I ff.)
that in some cases we do not even upon reflection ucceedin makinga divisioninto words: shall a German writeEskomintZg Stande in two, three,or fourwords? Shall we
write n stead ofas twowordsor as three? In as mucl as in
one, two,three,or fourwords? We have many nstancesof
the writing f uneducated people (who lack the practice of
copiousreading)in which theword-divisions entirelywrong.
Hence we repeat to-day n more decided termsthe quoted
thispitfall,-that some of the currentdoctrineof linguisticscience is a transference f ourown process of analysis into
ourbeliefsabout the course of linguistichistory, nd, as ourprocess of analysis is, admittedly, ot in harmonywith thefacts of speech,but, n a sense,diametricallypposed tothem,the transferencemay (and, I think, ometimesdoes) lead tofalse conclusions.
A bit of the older historyof our science well illustrateswhat I mean. The personalverb-formsf the Indo-European
languages were easily analyzed, as soon as people began toreflect pon such things, ntopersonalendingsattached to astem e.g. &w-,ut&&-' &saz-at,older andWest Greek&8&-Tt).
Proceeding from hisanalysis and takingforgrantedthat itrepresented historic ynthesis, opp identified he personalendingswithold formsof the personalpronouns. Similarlyhe saw in the i-suffix f the futureand optative the root(itself, of course, the creatureof a similiaranalysis) of theverb ire. These theories were given up not only becauseBopp's specificexplanations were in conflictwith the ascer-tained sound-developmentsf the languages concerned,butalso because we realize thatBopp was inspiredchiefly y thefeeling that our analysis of forms s necessarilyin accordwith theirhistoric rigin, and we know now that this f el-ing was wrong. When scholars to-dayspeculate upon theoriginof the personal verb-inflectionheyturnratherto anadaptation-theorynd suppose that the endings of theseformshave come, more or less accidentallyto theirpersonalmeaning; so, for nstance,Hirt,L.F.XVII, 6. That is to say,thegrammatical nalysisofa givenstage of a language mustnot lead us into thinking hat the forms re the result of acorresponding onglutination.
Bopp's error s a thingof the past,' but if a genius likeBopp could fall into such an error, t is obvious thatwe, too,need the correctiveof an occasional analysis in the psycho-logically ustified irection, roceedingfrom he sentence, he
1The Editor remindsme of Professor ay's articles; needlessto say, I donot agreewiththeirtendency. It is fair to add, however,hat I know Fay's" Return oBopp" onlyfromhesummaryn Idg.Jahrb.ni.
This content downloaded on Fri, 15 Mar 2013 11:32:35 AMAll use subject to JSTOR Terms and Conditions
concretedatum,to the less and less explicit articulations nthe sequence of speech.
A seriouserror hat has outlivedthe agglutination heoryis our definition f the sentence. The ancients,forwhomgrammar was an ancillary discipline of logic, necessarilylooked upon the sentenceas a combinationbuilt up out ofwords. Dionysios Thrax2 defined the sentence as vre?s
this: Oratio est ordinatio dictionumcongrua, sententiam
perfectamdemonstrans. It is Wilhelm Wundt who, in hisVoikcrpsychologie, , 2, 234 ff., first showed that, when we
understand the psychologyof the thing,this definitionstopsy-turvy: sentence, ays Wundt, s the linguistic xpres-sion of the voluntary nalysis of a total experience into itsparts,whichthenstandin logical relation o one another.
It has beenobjectedthatthisdefinitionoes notdistinguisha sentencefrom word, uch as Tpt'wOVW,which also involvesan analysis of the total experiencewhichit expresses. Weare face to face, then,with the problemof distinguishingbetween the analysis made by a speaker who says Tpeds
7o'Saq et'ov and that made by one who says Tpti'ovs,--
between sentence and word. Meanwhile we cannot retain
even Wundt's definitionof the sentence,for it implies an
articulationof the sentence into parts which we have no
rightto look upon as essential or universal. The assump-tion that every sentence must break up into two or more
independent - and logically articulated !- components is
Einleitung in die Altertumswlsscnsclzaft2 1 5 f.) as a ves-
tige ofthe old rationalizing iew,accordingto which it was
built up out of such components. Kretschmerdefinesthe
sentence as the linguistic xpressionof an affect,that is,of an up-and-downfemotional olumeand tension. Perhapswe shoulddo better o say thatthe sentence is the linguisticexpressionof an affect nvolvinga single total experience,for an affectof higherordermay be accompanied by the
2 TiXvm7,ar. II. 3 De ArkeGramm.I, 15.
This content downloaded on Fri, 15 Mar 2013 11:32:35 AMAll use subject to JSTOR Terms and Conditions
utterance of a succession of sentences, each of whichcorrespondsto a subordinateup-and-downmovementof the
emotional curve. What we most need is not, however, adefinition f the sentence or of theword, we have a verydecided naYvefeelingforthese units, but ratheran under-standing of the difference etween a succession of words,such as Tpe's vro"asge'xov nd what we feel to be a single in-flected,erived,rcompounded ord, uchas Tpw7roVS.
The phonetician,first f all, tells us that physical differ-
ence there s none. His ear tells him,and the difficultiesforthographic eparationabove referred o proveto him,thatthere are in an utterance no pauses to indicate its structure.We have the proof in ourselves wheneverwe hear peoplespeaking a language which we do not understand, or it isthen beyond our power to find heword-divisions. What isit,then,thatenables us to analyze utterances ntowordsandmorphologic lements
To begin with,it is not any reflection f the speaker's.Even people who have studied language and maybe to anabnormalextentconsciousof the facts of speech,uttermanysentences every day without the least reflection pon theiranalysis. As a writer n logic puts it,we oughtto write llour sentenceswithhyphensbetweenthewords; a phoneticianwould say thatwe ought not to indicate theword-divisiont
all. The divisionof the sentence is not a reflective ne; itis a matter f implication,nd is due to the associationalcon-nections of the partsofthe sound-sequencewhichconstitutesthe sentence, as it were, to their connotation. A Latinsentence such as exibant is, like every sentence,primarilyand so faras any logical reflectionn the part of speaker orhearermay be concerned, unit. The various parts of this
sound-sequence,however,have been heard and utteredbythe speaker (or the hearer) in other sentences and have, inthese other earlier occurrences, lways correspondedto anelement of meaningwhichis present also in this new expe-rience accompanyingthe sentence exibant. All these pastoccurrencesof parts of the presentsound-sequenceexerciseupon the latterthe subtle force knownto psychologistsas
This content downloaded on Fri, 15 Mar 2013 11:32:35 AMAll use subject to JSTOR Terms and Conditions
simultaneousassociationor fusion: theygive them a toneofrecognitionwhichwe,for our purposes,may speak of as their
meaning. Thus the first art of the sound-sequenceexibantowes itsvalue to earlier utterances heard and spoken fromchildhoodon), such as excessit, :r7igit,nd the like; the nextsound, to such as abfrcm, re(llmtls, and the like; the nextsound,-b-, as occurredalso in numerousutterances, uch asregUbat,videbit, ondbitur,n all ofwhich it correspondedtoa vague notionofcontinuityf action,past or future; the-a-
has occurred also in regebat, ram,fucrat, parallel with asemantic elementof past time; the-nt, inally,s one of themostfamiliar ound-successionsn the language,and has beenheard and spoken innumerabletimes in sentences that ex-pressed an event in which more than one actor, includingneitherspeaker nor hearer,performed n action or was thegoal (object) of an action, e.g. dolent, onaztur,de1ecta;itur,
exeuint,nd so on. Now, thoughall this dissection s fartoo
clumsyto do justice to the intanigiblemplication-valueshatare immediatelynd automaticallynvolvedin the speakingor hearing of the sentence exibant,yetwe can be surethatthe meaning of this sentence to a Latin was due to thesevery ssociations, orwe knowthat n language the sentenceswhicha speaker mayutterare not confined o thosewhichhe has actuallyheardbefore,but may consistof entirely ew
combinations f the habitualspeech-elements. A speakerofLatin who happened never to have heard the formextbatcould use it, and use it without the slightestconsciousnessof innovation, ince he had manytimesheard exabant, nizu-bat, arn4bant, nd so on. In other words, we may, veryclumsily, ndicate the associational values in the sentencecxfbant y dividing t into ex-ib-a-nt.
If, now, we take the corresponding English sentence,Yejuwrbowizydwt,e find similar associational habituation fthe different arts of the sound-sequence. Yejhas occurredin ;Yejddn jdit, eseJdsow,and the like, where also there was
a thirdperson plural actor; wr in such expressionsas wj-wrwejtiy, juwwi-Vr, and so on; gow in letsg5w, downtgzw,
and the like; iy,expressiveof continued ction, nhizjrajtiy,
This content downloaded on Fri, 15 Mar 2013 11:32:35 AMAll use subject to JSTOR Terms and Conditions
almw4jtiy,and so on; awt, in value like the Latin ex-,inkomanawt,hararndwt, ndmanyotherutterances; and here,
as in the Latin sentence,these parts are in ordinary peechby no means drawn into the focus of the attentionor ex-plicitlydistinguished, ut are rather,by the associative effectof theirearlieroccurrences, one mightalmost say,mutely,- symbolic of the meaning. Their utterance in certainsituations of experience,and the reproductionof a corre-sponding meaningwhenevertheyare heard, is a matterof
habit, not of explicit greemnentr reflection.There are, however,occasions when we utter such a sen-tencewitha full and explicitinsistenceupon some one partof it,and thusshow, consciousness of its division ntopartsand try,indeed, to arouse the same consciousness in thehearer. Suppose thatan element of the situation s in doubtor in question,for instance,the identity f those who wentout. Then we say
;ejwrgowiyawt. Here it is no longer
the mere implicitassociationalvalue of the sound-groupYejthat lends it meaning; our attention,ike a vivid spotlight,focuses this part of the utterance, ingling it out from therest; and thehearer's attention, y the loud tone and otherphoneticfeatures, s drawn to it. We maysimilarly, f thetime ofthe occurrencebe in question, ccent thewr and sayYejwr gowiyawt, nd so on. One element,however, fthose
found by analysis to make up this sentence, we cannot soemphasize, namely the sound-group y, expressive of conti-nuityof action. Its associational value is clear,but apper-ceptivevalue it can never have: it never falls into the focusof the attention. Besides this habit of never clearly consid-eringthe element y,we have another imiting ts use: it isspoken -after n elementexpressive of action, such as gow or
rajt,to which t lends themeaning of continuity,nd it neveroccurs in any other connections. This, moreover,s trueofall the parts of the Latin sentence whichwe have examined,exzbant.4
4 The sound-group ex-, to be sure, does occur in other connections, such as exurbe venif,but it has then a differentalue, expressing spatial relationwith regardto an object, not direction ofmovement: it is then a preposition, not an adverb.
This content downloaded on Fri, 15 Mar 2013 11:32:35 AMAll use subject to JSTOR Terms and Conditions
This is the difference etween a formative lementand aword,of course: both recur as the expressionof a constant
element of meaning, but the formative lement s bound tocertainpositionswith regardto the otherelements,whiletheword mayoccur in all kinds of connections; and, above all:while bothoccur usuallyas associativelydeterminedpartsofa sentence, he word maybe focused by the attention clearlyapperceived),while the formative lementnever rises to thisexplicitrecognition.
It is a commonplaceofpsychology hat, f these twoformsof the structure f experience, he associativeor passive,andthe apperceptive or active, the former s the primary ndusual one, the latter the more developed and rarer. Thegreaterpredominance f associational processescharacterizesfor us the mental habits of savages (sympatheticmagic,andthe like; cf.,in connectionwith language, Levy-Bruhl,Lesfonctiousmentalesdans les socie,'s infirieures, nd Marrett,Anthropology;Jespersen'sProgress n Language is, ofcourse,familiar), he courseof dreams,and morbidmental processes(Wundt,Gruzndrisser Psychologie). Opposed to all these,the higherphases of mental life,such as sane thinkingorscientific easoning,are characterizedby the frequentandunhesitating esort,whenever he occasion demands, o apper-ceptivefocusingofpartsof an experience.
It would be strangeif linguistichistory,s the agglutina-tiontheory ssumed,showedus a retrogressive evelopment,-a developmentfrom formsof speech which allowed notonly of associative but also of occasional apperceptivedis-tinction, oward formswhich moved onlyand always in thedimrealmof associativereminiscence. As a matterof fact,linguistichistory,whereverwe know t,showsus progress n
thedirection rom ssociative towardapperceptivestructure.Where in Old English one saidgad6u't,we express by a sepa-rateword both the actor and the tense: ;Yej a:r gowiy awt;where n Latin one said Romam it orRomamvadit,one usesin French a separatewordforthe direction nd anotherforthe actor: il va a' Rome.
The differences,n this respect, between Latin or Old
This content downloaded on Fri, 15 Mar 2013 11:32:35 AMAll use subject to JSTOR Terms and Conditions
English and the modernlanguages are of interestbecauseof the accessibilityof the historic relation and all that it
implies,but the structure f Latin or of Old English is notso widely different romthat of our speech. If actor, ac-tion,and tense are there expressed in one word,we find nother languages not only these elements,but also objects,directand indirect, nd other featuresof the experience, llexpressed without the possibilityof a single apperceptivearticulation,hat s, in one word. Thus in the Fox language
(Jonesand Michelson,Bulletin40, U. S. Bureau ofEthnology),pyit(`kwdwdwa, He brings home a wife' (pydte 'hither,home,' 'kzwvwlong hair,woman,' a 'her,' wa 'he'), nimd-winA utAmawawa, 'I shall go and ask him for it' (ni ' I,in future action,' mdwi 'go,' nAtut 'ask,' Amaw 'it, assecondaryobject,'d and wa, both referringo animate thirdperson).
It is interestingo noticethatthe first ndmost importantdivisionwhichlogical reflection as always demanded of the
sentence,namely,that into subject and predicate, s one ofthe rarest, nd,wherewe knowthehistory, ne of the latest,to receivea correspondingword-divisionn the sentence: inLatin, Greek,and Sanskritsubject and predicateare usuallyboth expressed n theverb-form: n Slavic and most of theRomance languages both possibilities re open (Italian cantaor ella canta,Polish spiewa or ona 4piewa).
In our anguageswe have in some cases the choicebetweenthe twomethodsof expression,one by a single wordand oneby a succession of severalwords. Thus, we mayspeak ofahorse-tamerr a tamer fhorses. The former indofexpres-sion obviouslyanalyzes itself into the elements horse andtamer,and linguisticscholars,taking for granted that our
analysis correspondsto the historicoccurrence, re wonttoassume thatsuch compoundwords are the productof a coa-lescence of independentwords. This assumptionmeetswitha verysignificant ifficulty: he fartherback we go in trac-ing thehistory f our languages,the less resemblancedo theparts of such compoundsbear to the individualwordsfromwhich it is supposed that theywere derived. Thus in An-
This content downloaded on Fri, 15 Mar 2013 11:32:35 AMAll use subject to JSTOR Terms and Conditions
cientGreekthe partsof theword i7rwo'a,juos,iffer rom nyindependently ccurring orms; 'w7ro-iffers rom nyactual
formof '77ros(, and -8ajuo'stamer,' is a type of formationrarelyfoundoutsideof compound words. Similarly, he -pt-of rpt'7rov does not occuras an independentword. Anyonewho reads Brugmann'ssectionon noun-compoundsGrund-
riss, II2, 1, 49 ff.) or the second volume of Wackernagel'smonumental anskritGrammarwillbe impressedby the end-less deviations, xceedingall possibilities f accidentalor sec-
ondarydevelopment, f composition-stemsrom ndependentwords, e.g. Greek lcv3po";cv&dvetpa, raTrp 7raTpo'. The most
is so obtrusivethat it has given rise to the supplementarytheory hat these compounds go back to a time postulatedOd hoc) when uninflected tems were used as words,andused,the compounds compel one further o assume, in the
value ofany and everycase-relation. So Brugmann Grun-
driss, I2, I, 78); upon this theoryJacobihas builthis specu-lations in Compositmi indNebensats. Needless to say that
the whole assumptionthat compoundwords are historicallythe resultof a coalescence has no othersupportthanthe cir-cumstancethatwe analyze them nto elementsmore or lesscloselyresembling inglewords, exactly as Bopp analyzedout of the personalverb-formsertainelementsmoreor less
closelyresembling ersonalpronouns. In neithercase doesthe analysis justifya historicalassumption. Quite on the
contrary,hefarther ack we go intohistory,he less do the
elements f compoundsresemble inglewords: we have everyreasonto believe thatthe compoundwordsof the Indo-Euro-
pean languagesrepresent n oldertypeof formationnwhich
meanings that are now usually expressed in several words
were stillmergedinto one word whose divisionshad onlyanassociative identity, a word comparable to the formationsof the American anguages. The possibility f breakingupthe sentence ntothose smallerunitswhichwe are accustomedto lookuponas correspondingosimplewordswas of laterde-
velopment,xactly s thepossibilityfseparating ctorfrom c-
tionnLatinorGermanic peechhas developed n historicimes.
This content downloaded on Fri, 15 Mar 2013 11:32:35 AMAll use subject to JSTOR Terms and Conditions
The compoundword remained in use where its meaninghad undergone transferenceor specialization and differed,
accordingly, romthat of the now more favoredcollocationof simple words. This accountsforthe persistenceof suchtypesas the so-calledexocentric ompounds, /cV7rEposI ' hav-ing quick wings,' English long-noseone who has a longnose,' and, in general, for our habit of using compoundswhere we mean somethingmore specificthan what would beexpressed by a collocation, .g. blue-bird,s opposed to blue
bird.To recite the evidence for this view would be to tell theentire storyof compound words in the Indo-European lan-guages.5 So much,however, s certain, hat,here as elsewhere,the course of linguistichistoryhas been fromassociationalarticulationof the utterance toward apperceptivestructure;and that the grammarian'sdissection of words, though ofinfinite racticalvalue, mustnot mislead us into thinking hatlanguage is really a pasting together, y means of hyphensor a similar agency, of the elements which this dissectionmay reveal.
6 An interestingask,which have undertakennd hope somedayto finish.