Thelatteretymologyisconfirmedbythewordsboulesthai，boule，aboulia，whichallhavetodowithshooting（bole）：andsimilarlyoiesisisnothingbutthemovement（oisis）ofthesoultowardsessence。Ekousionistoeikon——theyielding——anagkeiseanagkeiousa，thepassagethroughravineswhichimpedemotion：aletheiaistheiaale，divinemotion。Pseudosistheoppositeofthis，implyingtheprincipleofconstraintandforcedrepose，whichisexpressedunderthefigureofsleep，toeudon；thepsiisanaddition。Onoma，aname，affirmstherealexistenceofthatwhichissoughtafter——onoumasmaestin。Onandousiaareonlyionwithaniotabrokenoff；andoukonisoukion。’Andwhatareion，reon，doun？’Onewayofexplainingthemhasbeenalreadysuggested——theymaybeofforeignorigin；andpossiblythisisthetrueanswer。Butmereantiquitymayoftenpreventourrecognizingwords，afterallthecomplicationswhichtheyhaveundergone；andwemustrememberthathoweverfarwecarrybackouranalysissomeultimateelementsorrootswillremainwhichcanbenofurtheranalyzed。Forexample；thewordagathoswassupposedbyustobeacompoundofagastosandthoos，andprobablythoosmaybefurtherresolvable。Butifwetakeawordofwhichnofurtherresolutionseemsattainable，wemayfairlyconcludethatwehavereachedoneoftheseoriginalelements，andthetruthofsuchawordmustbetestedbysomenewmethod。Willyouhelpmeinthesearch？
　　Allnames，whetherprimaryorsecondary，areintendedtoshowthenatureofthings；andthesecondary，asIconceive，derivetheirsignificancefromtheprimary。Butthen，howdotheprimarynamesindicateanything？Andletmeaskanotherquestion，——Ifwehadnofacultyofspeech，howshouldwecommunicatewithoneanother？Shouldwenotusesigns，likethedeafanddumb？Theelevationofourhandswouldmeanlightness——heavinesswouldbeexpressedbylettingthemdrop。Therunningofanyanimalwouldbedescribedbyasimilarmovementofourownframes。Thebodycanonlyexpressanythingbyimitation；andthetongueormouthcanimitateaswellastherestofthebody。Butthisimitationofthetongueorvoiceisnotyetaname，becausepeoplemayimitatesheeporgoatswithoutnamingthem。
　　What，then，isaname？Inthefirstplace，anameisnotamusical，or，secondly，apictorialimitation，butanimitationofthatkindwhichexpressesthenatureofathing；andistheinventionnotofamusician，orofapainter，butofanamer。
　　Andnow，Ithinkthatwemayconsiderthenamesaboutwhichyouwereasking。Thewaytoanalyzethemwillbebygoingbacktotheletters，orprimaryelementsofwhichtheyarecomposed。First，weseparatethealphabetintoclassesofletters，distinguishingtheconsonants，mutes，vowels，andsemivowels；andwhenwehavelearntthemsingly，weshalllearntoknowthemintheirvariouscombinationsoftwoormoreletters；justasthepainterknowshowtouseeitherasinglecolour，oracombinationofcolours。Andlikethepainter，wemayapplyletterstotheexpressionofobjects，andformthemintosyllables；andtheseagainintowords，untilthepictureorfigure——thatis，language——iscompleted。NotthatIamliterallyspeakingofourselves，butImeantosaythatthiswasthewayinwhichtheancientsframedlanguage。Andthisleadsmetoconsiderwhethertheprimaryaswellasthesecondaryelementsarerightlygiven。Imayremark，asIwassayingabouttheGods，thatwecanonlyattaintoconjectureofthem。Butstillweinsistthatoursisthetrueandonlymethodofdiscovery；otherwisewemusthaverecourse，likethetragicpoets，toaDeusexmachina，andsaythatGodgavethefirstnames，andthereforetheyareright；orthatthebarbariansareolderthanweare，andthatwelearntofthem；orthatantiquityhascastaveiloverthetruth。
　　Yetallthesearenotreasons；theyareonlyingeniousexcusesforhavingnoreasons。
　　Iwillfreelyimparttoyoumyownnotions，thoughtheyaresomewhatcrude：——theletterrhoappearstometobethegeneralinstrumentwhichthelegislatorhasemployedtoexpressallmotionorkinesis。（Ioughttoexplainthatkinesisisjustiesis（going），fortheletteretawasunknowntotheancients；andtheroot，kiein，isaforeignformofienai：ofkinesisoreisis，theoppositeisstasis）。Thisuseofrhoisevidentinthewordstremble，break，crush，crumble，andthelike；theimposerofnamesperceivedthatthetongueismostagitatedinthepronunciationofthisletter，justasheusediotatoexpressthesubtlepowerwhichpenetratesthroughallthings。Thelettersphi，psi，sigma，zeta，whichrequireagreatdealofwind，areemployedintheimitationofsuchnotionsasshivering，seething，shaking，andingeneralofwhatiswindy。Thelettersdeltaandtauconveytheideaofbindingandrestinaplace：thelambdadenotessmoothness，asinthewordsslip，sleek，sleep，andthelike。Butwhentheslippingtongueisdetainedbytheheaviersoundofgamma，thenarisesthenotionofaglutinousclammynature：nuissoundedfromwithin，andhasanotionofinwardness：alphaistheexpressionofsize；etaoflength；omicronofroundness，andthereforethereisplentyofomicroninthewordgoggulon。Thatismyview，Hermogenes，ofthecorrectnessofnames；andIshouldliketohearwhatCratyluswouldsay。
　　’But，Socrates，asIwastellingyou，Cratylusmystifiesme；Ishouldliketoaskhim，inyourpresence，whathemeansbythefitnessofnames？’Tothisappeal，Cratylusreplies’thathecannotexplainsoimportantasubjectallinamoment。’’No，butyoumay"addlittletolittle，"asHesiodsays。’Socrateshereinterposeshisownrequest，thatCratyluswillgivesomeaccountofhistheory。Hermogenesandhimselfaremeresciolists，butCratylushasreflectedonthesematters，andhashadteachers。CratylusrepliesinthewordsofAchilles：’"IllustriousAjax，youhavespokeninallthingsmuchtomymind，"whetherEuthyphro，orsomeMuseinhabitingyourownbreast，wastheinspirer。’Socratesreplies，thatheisafraidofbeingself-deceived，andthereforehemust’lookforeandaft，’asHomerremarks。DoesnotCratylusagreewithhimthatnamesteachusthenatureofthings？’Yes。’Andnamingisanart，andtheartistsarelegislators，andlikeartistsingeneral，someofthemarebetterandsomeofthemareworsethanothers，andgivebetterorworselaws，andmakebetterorworsenames。Cratyluscannotadmitthatonenameisbetterthananother；theyareeithertruenames，ortheyarenotnamesatall；andwhenheisaskedaboutthenameofHermogenes，whoisacknowledgedtohavenoluckinhim，heaffirmsthistobethenameofsomebodyelse。Socratessupposeshimtomeanthatfalsehoodisimpossible，towhichhisownanswerwouldbe，thattherehasneverbeenalackofliars。Cratyluspresseshimwiththeoldsophisticalargument，thatfalsehoodissayingthatwhichisnot，andthereforesayingnothing；——youcannotutterthewordwhichisnot。
　　Socratescomplainsthatthisargumentistoosubtleforanoldmantounderstand：SupposeapersonaddressingCratylusweretosay，Hail，AthenianStranger，Hermogenes！wouldthesewordsbetrueorfalse？’I
　　shouldsaythattheywouldbemereunmeaningsounds，likethehammeringofabrasspot。’Butyouwouldacknowledgethatnames，aswellaspictures，areimitations，andalsothatpicturesmaygivearightorwrongrepresentationofamanorwoman：——whymaynotnamesthenequallygivearepresentationtrueandrightorfalseandwrong？Cratylusadmitsthatpicturesmaygiveatrueorfalserepresentation，butdeniesthatnamescan。Socratesargues，thathemaygouptoamanandsay’thisisyearpicture，’andagain，hemaygoandsaytohim’thisisyourname’——intheonecaseappealingtohissenseofsight，andintheothertohissenseofhearing；——mayhenot？’Yes。’Thenyouwilladmitthatthereisarightorawrongassignmentofnames，andifofnames，thenofverbsandnouns；andifofverbsandnouns，thenofthesentenceswhicharemadeupofthem；andcomparingnounstopictures，youmaygivethemalltheappropriatesounds，oronlysomeofthem。Andashewhogivesallthecoloursmakesagoodpicture，andhewhogivesonlysomeofthem，abadorimperfectone，butstillapicture；sohewhogivesallthesoundsmakesagoodname，andhewhogivesonlysomeofthem，abadorimperfectone，butanamestill。Theartistofnames，thatis，thelegislator，maybeagoodorhemaybeabadartist。’Yes，Socrates，butthecasesarenotparallel；forifyousubtractormisplacealetter，thenameceasestobeaname。’Socratesadmitsthatthenumber10，ifanunitissubtracted，wouldceasetobe10，butdeniesthatnamesareofthispurelyquantitativenature。Supposethattherearetwoobjects——CratylusandtheimageofCratylus；andletusimaginethatsomeGodmakesthemperfectlyalike，bothintheiroutwardformandintheirinnernatureandqualities：thentherewillbetwoCratyluses，andnotmerelyCratylusandtheimageofCratylus。Butanimageinfactalwaysfallsshortinsomedegreeoftheoriginal，andifimagesarenotexactcounterparts，whyshouldnamesbe？iftheywere，theywouldbethedoublesoftheiroriginals，andindistinguishablefromthem；
　　andhowridiculouswouldthisbe！CratylusadmitsthetruthofSocrates’
　　remark。ButthenSocratesrejoins，heshouldhavethecouragetoacknowledgethatlettersmaybewronglyinsertedinanoun，oranouninasentence；andyetthenounorthesentencemayretainameaning。Bettertoadmitthis，thatwemaynotbepunishedlikethetravellerinEginawhogoesaboutatnight，andthatTruthherselfmaynotsaytous，’Toolate。’
　　And，errorsexcepted，wemaystillaffirmthatanametobecorrectmusthaveproperletters，whichbeararesemblancetothethingsignified。I
　　mustremindyouofwhatHermogenesandIweresayingabouttheletterrhoaccent，whichwasheldtobeexpressiveofmotionandhardness，aslambdaisofsmoothness；——andthisyouwilladmittobetheirnaturalmeaning。
　　Butthen，whydotheEritreanscallthatskleroterwhichwecallsklerotes？
　　Wecanunderstandoneanother，althoughtheletterrhoaccentisnotequivalenttotheletters：whyisthis？Youreply，becausethetwolettersaresufficientlyalikeforthepurposeofexpressingmotion。Well，then，thereistheletterlambda；whatbusinesshasthisinawordmeaninghardness？’Why，Socrates，Iretortuponyou，thatweputinandpulloutlettersatpleasure。’Andtheexplanationofthisiscustomoragreement：
　　wehavemadeaconventionthattherhoshallmeansandaconventionmayindicatebytheunlikeaswellasbythelike。Howcouldtherebenamesforallthenumbersunlessyouallowthatconventionisused？Imitationisapoorthing，andhastobesupplementedbyconvention，whichisanotherpoorthing；althoughIagreewithyouinthinkingthatthemostperfectformoflanguageisfoundonlywherethereisaperfectcorrespondenceofsoundandmeaning。Butletmeaskyouwhatistheuseandforceofnames？
　　’Theuseofnames，Socrates，istoinform，andhewhoknowsnamesknowsthings。’Doyoumeanthatthediscoveryofnamesisthesameasthediscoveryofthings？’Yes。’Butdoyounotseethatthereisadegreeofdeceptionaboutnames？Hewhofirstgavenames，gavethemaccordingtohisconception，andthatmayhavebeenerroneous。’Butthen，why，Socrates，islanguagesoconsistent？allwordshavethesamelaws。’Mereconsistencyisnotestoftruth。Ingeometricalproblems，forexample，theremaybeaflawatthebeginning，andyettheconclusionmayfollowconsistently。
　　And，therefore，awisemanwilltakeespecialcareoffirstprinciples。
　　Butarewordsreallyconsistent；aretherenotasmanytermsofpraisewhichsignifyrestaswhichsignifymotion？Thereisepisteme，whichisconnectedwithstasis，asmnemeiswithmeno。Bebaion，again，istheexpressionofstationandposition；istoriaisclearlydescriptiveofthestoppingistanaiofthestream；pistonindicatesthecessationofmotion；
　　andtherearemanywordshavingabadsense，whichareconnectedwithideasofmotion，suchassumphora，amartia，etc。：amathia，again，mightbeexplained，aseamatheoiontosporeia，andakolasiaaseakolouthiatoispragmasin。Thusthebadnamesareframedonthesameprincipleasthegood，andotherexamplesmightbegiven，whichwouldfavouratheoryofrestratherthanofmotion。’Yes；butthegreaternumberofwordsexpressmotion。’Arewetocountthem，Cratylus；andiscorrectnessofnamestobedeterminedbythevoiceofamajority？
　　Hereisanotherpoint：weweresayingthatthelegislatorgivesnames；andthereforewemustsupposethatheknowsthethingswhichhenames：buthowcanhehavelearntthingsfromnamesbeforetherewereanynames？’I
　　believe，Socrates，thatsomepowermorethanhumanfirstgavethingstheirnames，andthatthesewerenecessarilytruenames。’Thenhowcamethegiverofnamestocontradicthimself，andtomakesomenamesexpressiveofrest，andothersofmotion？’Idonotsupposethathedidmakethemboth。’
　　Thenwhichdidhemake——thosewhichareexpressiveofrest，orthosewhichareexpressiveofmotion？……Butifsomenamesaretrueandothersfalse，wecanonlydecidebetweenthem，notbycountingwords，butbyappealingtothings。And，ifso，wemustallowthatthingsmaybeknownwithoutnames；
　　fornames，aswehaveseveraltimesadmitted，aretheimagesofthings；andthehigherknowledgeisofthings，andisnottobederivedfromnames；andthoughIdonotdoubtthattheinventorsoflanguagegavenames，undertheideathatallthingsareinastateofmotionandflux，Ibelievethattheyweremistaken；andthathavingfallenintoawhirlpoolthemselves，theyaretryingtodragusafterthem。Foristherenotatruebeautyandatruegood，whichisalwaysbeautifulandalwaysgood？Canthethingbeautybevanishingawayfromuswhilethewordsareyetinourmouths？Andtheycouldnotbeknownbyanyoneiftheyarealwayspassingaway——foriftheyarealwayspassingaway，theobserverhasnoopportunityofobservingtheirstate。Whetherthedoctrineofthefluxoroftheeternalnaturebethetruer，ishardtodetermine。Butnomanofsensewillputhimself，ortheeducationofhismind，inthepowerofnames：hewillnotcondemnhimselftobeanunrealthing，norwillhebelievethateverythingisinafluxlikethewaterinaleakyvessel，orthattheworldisamanwhohasarunningatthenose。Thisdoctrinemaybetrue，Cratylus，butisalsoverylikelytobeuntrue；andthereforeIwouldhaveyoureflectwhileyouareyoung，andfindoutthetruth，andwhenyouknowcomeandtellme。’Ihavethought，Socrates，andafteragooddealofthinkingIinclinetoHeracleitus。’Thenanotherday，myfriend，youshallgivemealesson。
　　’Verygood，Socrates，andIhopethatyouwillcontinuetostudythesethingsyourself。’……
　　Wemaynowconsider（I）howfarPlatointheCratylushasdiscoveredthetrueprinciplesoflanguage，andthen（II）proceedtocomparemodernspeculationsrespectingtheoriginandnatureoflanguagewiththeanticipationsofhisgenius。
　　I。（1）Platoisawarethatlanguageisnottheworkofchance；nordoeshedenythatthereisanaturalfitnessinnames。Heonlyinsiststhatthisnaturalfitnessshallbeintelligiblyexplained。Buthehasnoideathatlanguageisanaturalorganism。Hewouldhaveheardwithsurprisethatlanguagesarethecommonworkofwholenationsinaprimitiveorsemi-
　　barbarousage。How，hewouldprobablyhaveargued，couldmendevoidofarthavecontrivedastructureofsuchcomplexity？Noanswercouldhavebeengiventothisquestion，eitherinancientorinmoderntimes，untilthenatureofprimitiveantiquityhadbeenthoroughlystudied，andtheinstinctsofmanhadbeenshowntoexistingreaterforce，whenhisstateapproachesmorenearlytothatofchildrenoranimals。Thephilosophersofthelastcentury，aftertheirmanner，wouldhavevainlyendeavouredtotracetheprocessbywhichpropernameswereconvertedintocommon，andwouldhaveshownhowthelasteffortofabstractioninventedprepositionsandauxiliaries。Thetheologianwouldhaveprovedthatlanguagemusthavehadadivineorigin，becauseinchildhood，whiletheorgansarepliable，theintelligenceiswanting，andwhentheintelligenceisabletoframeconceptions，theorgansarenolongerabletoexpressthem。Or，asothershavesaid：Manismanbecausehehasthegiftofspeech；andhecouldnothaveinventedthatwhichheis。Butthiswouldhavebeenan’argumenttoosubtle’forSocrates，whorejectsthetheologicalaccountoftheoriginoflanguage’asanexcusefornotgivingareason，’whichhecomparestotheintroductionofthe’Deusexmachina’bythetragicpoetswhentheyhavetosolveadifficulty；thusanticipatingmanymoderncontroversiesinwhichtheprimaryagencyofthedivineBeingisconfusedwiththesecondarycause；andGodisassumedtohaveworkedamiracleinordertofillupalacunainhumanknowledge。（CompareTimaeus。）
　　NeitherisPlatowronginsupposingthatanelementofdesignandartentersintolanguage。Thecreativepowerabatingissupplementedbyamechanicalprocess。’Languagesarenotmadebutgrow，’buttheyaremadeaswellasgrow；burstingintolifelikeaplantoraflower，theyarealsocapableofbeingtrainedandimprovedandengraftedupononeanother。Thechangeinthemiseffectedinearlieragesbymusicalandeuphonicimprovements，atalaterstagebytheinfluenceofgrammarandlogic，andbythepoeticalandliteraryuseofwords。Theydeveloperapidlyinchildhood，andwhentheyarefullgrownandsettheymaystillputforthintellectualpowers，likethemindinthebody，orratherwemaysaythatthenobleruseoflanguageonlybeginswhentheframe-workiscomplete。
　　Thesavageorprimitiveman，inwhomthenaturalinstinctisstrongest，isalsothegreatestimproveroftheformsoflanguage。Heisthepoetormakerofwords，asincivilisedagesthedialecticianisthedefinerordistinguisherofthem。Thelattercallsthesecondworldofabstracttermsintoexistence，astheformerhascreatedthepicturesoundswhichrepresentnaturalobjectsorprocesses。Poetryandphilosophy——thesetwo，arethetwogreatformativeprinciplesoflanguage，whentheyhavepassedtheirfirststage，ofwhich，asofthefirstinventionoftheartsingeneral，weonlyentertainconjecture。Andmythologyisalinkbetweenthem，connectingthevisibleandinvisible，untilatlengththesensuousexteriorfallsaway，andtheseveranceoftheinnerandouterworld，oftheideaandtheobjectofsense，becomescomplete。Atalaterperiod，logicandgrammar，sisterarts，preserveandenlargethedecayinginstinctoflanguage，byruleandmethod，whichtheygatherfromanalysisandobservation。
　　（2）ThereisnotraceinanyofPlato’swritingsthathewasacquaintedwithanylanguagebutGreek。YethehasconceivedverytrulytherelationofGreektoforeignlanguages，whichheisledtoconsider，becausehefindsthatmanyGreekwordsareincapableofexplanation。Allowingagooddealforaccident，andalsoforthefanciesoftheconditoreslinguaeGraecae，thereisanelementofwhichheisunabletogiveanaccount。
　　Theseunintelligiblewordshesupposestobeofforeignorigin，andtohavebeenderivedfromatimewhentheGreekswereeitherbarbarians，orincloserelationstothebarbarians。Socratesisawarethatthisprincipleisliabletogreatabuse；and，likethe’Deusexmachina，’explainsnothing。Henceheexcuseshimselffortheemploymentofsuchadevice，andremarksthatinforeignwordsthereisstillaprincipleofcorrectness，whichappliesequallybothtoGreeksandbarbarians。
　　（3）Butthegreaternumberofprimarywordsdonotadmitofderivationfromforeignlanguages；theymustberesolvedintothelettersoutofwhichtheyarecomposed，andthereforethelettersmusthaveameaning。Theframersoflanguagewereawareofthis；theyobservedthatalphawasadaptedtoexpresssize；etalength；omicronroundness；nuinwardness；rhoaccentrushorroar；lambdaliquidity；gammalambdathedetentionoftheliquidorslipperyelement；deltaandtaubinding；phi，psi，sigma，xi，windandcold，andsoon。Plato’sanalysisofthelettersofthealphabetshowsawonderfulinsightintothenatureoflanguage。Hedoesnotexpressivelydistinguishbetweenmereimitationandthesymbolicaluseofsoundtoexpressthought，butherecognisesintheexampleswhichhegivesbothmodesofimitation。Gestureisthemodewhichadeafanddumbpersonwouldtakeofindicatinghismeaning。Andlanguageisthegestureofthetongue；intheuseoftheletterrhoaccent，toexpressarushingorroaring，orofomicrontoexpressroundness，thereisadirectimitation；
　　whileintheuseoftheletteralphatoexpresssize，orofetatoexpresslength，theimitationissymbolical。Theuseofanalogousorsimilarsounds，inordertoexpresssimilaranalogousideas，seemstohaveescapedhim。
　　Inpassingfromthegestureofthebodytothemovementofthetongue，Platomakesagreatstepinthephysiologyoflanguage。Hewasprobablythefirstwhosaidthat’languageisimitativesound，’whichisthegreatestanddeepesttruthofphilology；althoughheisnotawareofthelawsofeuphonyandassociationbywhichimitationmustberegulated。Hewasprobablyalsothefirstwhomadeadistinctionbetweensimpleandcompoundwords，atruthsecondonlyinimportancetothatwhichhasjustbeenmentioned。Hisgreatinsightinonedirectioncuriouslycontrastswithhisblindnessinanother；forheappearstobewhollyunaware（comparehisderivationofagathosfromagastosandthoos）ofthedifferencebetweentherootandtermination。ButwemustrecollectthathewasnecessarilymoreignorantthananyschoolboyofGreekgrammar，andhadnotableoftheinflexionsofverbsandnounsbeforehiseyes，whichmighthavesuggestedtohimthedistinction。
　　（4）Platodistinctlyaffirmsthatlanguageisnottruth，or’philosophieunelanguebienfaite。’Atfirst，SocrateshasdelightedhimselfwithdiscoveringthefluxofHeracleitusinlanguage。Butheiscovertlysatirisingthepretenceofthatoranyotheragetofindphilosophyinwords；andheafterwardscorrectsanyerroneousinferencewhichmightbegatheredfromhisexperiment。Forhefindsasmany，oralmostasmany，wordsexpressiveofrest，ashehadpreviouslyfoundexpressiveofmotion。
　　Andevenifthishadbeenotherwise，whowouldlearnofwordswhenhemightlearnofthings？ThereisagreatcontroversyandhighargumentbetweenHeracleiteansandEleatics，butnomanofsensewouldcommithissoulinsuchenquiriestotheimposersofnames……InthisandotherpassagesPlatoshowsthatheisascompletelyemancipatedfromtheinfluenceof’Idolsofthetribe’asBaconhimself。
　　Thelessonwhichmaybegatheredfromwordsisnotmetaphysicalormoral，buthistorical。Theyteachustheaffinityofraces，theytellussomethingabouttheassociationofideas，theyoccasionallypreservethememoryofadisusedcustom；butwecannotsafelyarguefromthemaboutrightandwrong，matterandmind，freedomandnecessity，ortheotherproblemsofmoralandmetaphysicalphilosophy。Fortheuseofwordsonsuchsubjectsmayoftenbemetaphorical，accidental，derivedfromotherlanguages，andmayhavenorelationtothecontemporarystateofthoughtandfeeling。Norinanycaseistheinventionofthemtheresultofphilosophicalreflection；theyhavebeencommonlytransferredfrommattertomind，andtheirmeaningistheveryreverseoftheiretymology。Becausethereisorisnotanameforathing，wecannotarguethatthethinghasorhasnotanactualexistence；orthattheantitheses，parallels，conjugates，correlativesoflanguagehaveanythingcorrespondingtotheminnature。Therearetoomanywordsaswellastoofew；andtheygeneralizetheobjectsorideaswhichtheyrepresent。Thegreatestlessonwhichthephilosophicalanalysisoflanguageteachesusis，thatweshouldbeabovelanguage，makingwordsourservants，andnotallowingthemtobeourmasters。
　　Platodoesnotaddthefurtherobservation，thattheetymologicalmeaningofwordsisinprocessofbeinglost。Ifatfirstframedonaprincipleofintelligibility，theywouldgraduallyceasetobeintelligible，likethoseofaforeignlanguage，heiswillingtoadmitthattheyaresubjecttomanychanges，andputonmanydisguises。Heacknowledgesthatthe’poorcreature’imitationissupplementedbyanother’poorcreature，’——
　　convention。Buthedoesnotseethat’habitandrepute，’andtheirrelationtootherwords，arealwaysexercisinganinfluenceoverthem。
　　Wordsappeartobeisolated，buttheyarereallythepartsofanorganismwhichisalwaysbeingreproduced。Theyarerefinedbycivilization，harmonizedbypoetry，emphasizedbyliterature，technicallyappliedinphilosophyandart；theyareusedassymbolsontheborder-groundofhumanknowledge；theyreceiveafreshimpressfromindividualgenius，andcomewithanewforceandassociationtoeverylively-mindedperson。Theyarefixedbythesimultaneousutteranceofmillions，andyetarealwaysimperceptiblychanging；——nottheinventorsoflanguage，butwritingandspeaking，andparticularlygreatwriters，orworkswhichpassintotheheartsofnations，Homer，Shakespear，Dante，theGermanorEnglishBible，KantandHegel，arethemakersoftheminlaterages。Theycarrywiththemthefadedrecollectionoftheirownpasthistory；theuseofawordinastrikingandfamiliarpassagegivesacomplexiontoitsuseeverywhereelse，andthenewuseofanoldandfamiliarphrasehasalsoapeculiarpoweroverus。ButtheseandothersubtletiesoflanguageescapedtheobservationofPlato。Heisnotawarethatthelanguagesoftheworldareorganicstructures，andthateverywordinthemisrelatedtoeveryother；
　　nordoesheconceiveoflanguageasthejointworkofthespeakerandthehearer，requiringinmanafacultynotonlyofexpressinghisthoughtsbutofunderstandingthoseofothers。
　　Ontheotherhand，hecannotbejustlychargedwithadesiretoframelanguageonartificialprinciples。Philosophershavesometimesdreamedofatechnicalorscientificlanguage，inwordswhichshouldhavefixedmeanings，andstandinthesamerelationtooneanotherasthesubstanceswhichtheydenote。ButthereisnomoretraceofthisinPlatothanthereisofalanguagecorrespondingtotheideas；nor，indeed，couldthewantofsuchalanguagebefeltuntilthescienceswerefarmoredeveloped。Thosewhowouldextendtheuseoftechnicalphraseologybeyondthelimitsofscienceorofcustom，seemtoforgetthatfreedomandsuggestivenessandtheplayofassociationareessentialcharacteristicsoflanguage。ThegreatmasterhasshownhowheregardedpedanticdistinctionsofwordsorattemptstoconfinetheirmeaninginthesatireonProdicusintheProtagoras。
　　（5）Inadditiontotheseanticipationsofthegeneralprinciplesofphilology，wemaynotealsoafewcuriousobservationsonwordsandsounds。
　　’TheEretrianssaysklerotesforskleroter；’’theThessalianscallApolloAmlos；’’ThePhrygianshavethewordspur，udor，kunesslightlychanged；’
　　’thereisanoldHomericwordemesato，meaning"hecontrived"；’’ourforefathers，andespeciallythewomen，whoaremostconservativeoftheancientlanguage，lovedthelettersiotaanddelta；butnowiotaischangedintoetaandepsilon，anddeltaintozeta；thisissupposedtoincreasethegrandeurofthesound。’Platowasverywillingtouseinductivearguments，sofarastheywerewithinhisreach；buthewouldalsohaveassignedalargeinfluencetochance。Norindeedisinductionapplicabletophilologyinthesamedegreeastomostofthephysicalsciences。Forafterwehavepushedourresearchestothefurthestpoint，inlanguageasinalltheothercreationsofthehumanmind，therewillalwaysremainanelementofexceptionoraccidentorfree-will，whichcannotbeeliminated。
　　Thequestion，’whetherfalsehoodisimpossible，’whichSocratescharacteristicallysetsasideastoosubtleforanoldman（compareEuthyd。），couldonlyhaveariseninanageofimperfectconsciousness，whichhadnotyetlearnedtodistinguishwordsfromthings。Socratesrepliesineffectthatwordshaveanindependentexistence；thusanticipatingthesolutionofthemediaevalcontroversyofNominalismandRealism。Heisawaretoothatlanguagesexistinvariousdegreesofperfection，andthattheanalysisofthemcanonlybecarriedtoacertainpoint。’Ifwecouldalways，oralmostalways，uselikenesses，whicharetheappropriateexpressions，thatwouldbethemostperfectstateoflanguage。’Thesewordssuggestaquestionofdeeperinterestthantheoriginoflanguage；viz。whatistheidealoflanguage，howfarbyanycorrectionoftheirusagesexistinglanguagesmightbecomeclearerandmoreexpressivethantheyare，morepoetical，andalsomorelogical；orwhethertheyarenowfinallyfixedandhavereceivedtheirlastimpressfromtimeandauthority。
　　Onthewhole，theCratylusseemstocontaindeepertruthsaboutlanguagethananyotherancientwriting。Butfeelingtheuncertaingrounduponwhichheiswalking，andpartlyinordertopreservethecharacterofSocrates，Platoenvelopesthewholesubjectinarobeoffancy，andallowshisprinciplestodropoutasifbyaccident。
　　II。Whatistheresultofrecentspeculationsabouttheoriginandnatureoflanguage？Likeothermodernmetaphysicalenquiries，theyendatlastinastatementoffacts。But，inordertostateorunderstandthefacts，ametaphysicalinsightseemstoberequired。Therearemorethingsinlanguagethanthehumanmindeasilyconceives。Andmanyfallacieshavetobedispelled，aswellasobservationsmade。Thetruespiritofphilosophyormetaphysicscanalonecharmawaymetaphysicalillusions，whicharealwaysreappearing，formerlyinthefanciesofneoplatonistwriters，nowinthedisguiseofexperienceandcommonsense。Ananalogy，afigureofspeech，anintelligibletheory，asuperficialobservationoftheindividual，haveoftenbeenmistakenforatrueaccountoftheoriginoflanguage。
　　Speakingisoneofthesimplestnaturaloperations，andalsothemostcomplex。Nothingwouldseemtobeeasierormoretrivialthanafewwordsutteredbyachildinanylanguage。Yetintotheformationofthosewordshaveenteredcauseswhichthehumanmindisnotcapableofcalculating。
　　Theyareadroportwoofthegreatstreamoroceanofspeechwhichhasbeenflowinginallages。Theyhavebeentransmittedfromonelanguagetoanother；likethechildhimself，theygobacktothebeginningsofthehumanrace。Howtheyoriginated，whocantell？Neverthelesswecanimagineastageofhumansocietyinwhichthecircleofmen’smindswasnarrowerandtheirsympathiesandinstinctsstronger；inwhichtheirorgansofspeechweremoreflexible，andthesenseofhearingfinerandmorediscerning；inwhichtheylivedmoreincompany，andafterthemannerofchildrenweremoregiventoexpresstheirfeelings；inwhich’theymovedalltogether，’likeaherdofwildanimals，’whentheymovedatall。’
　　Amongthem，asineverysociety，aparticularpersonwouldbemoresensitiveandintelligentthantherest。Suddenly，onsomeoccasionofinterest（attheapproachofawildbeast，shallwesay？），hefirst，theyfollowinghim，utteracrywhichresoundsthroughtheforest。Thecryisalmostorquiteinvoluntary，andmaybeanimitationoftheroaroftheanimal。Thusfarwehavenotspeech，butonlytheinarticulateexpressionoffeelingoremotioninnorespectdifferingfromthecriesofanimals；
　　fortheytoocalltooneanotherandareanswered。Butnowsupposethatsomeoneatadistancenotonlyhearsthesound，butapprehendsthemeaning：orwemayimaginethatthecryisrepeatedtoamemberofthesocietywhohadbeenabsent；theothersactthesceneoveragainwhenhereturnshomeintheevening。Andsothecrybecomesaword。Thehearerinturngivesbackthewordtothespeaker，whoisnowawarethathehasacquiredanewpower。Manythousandtimesheexercisesthispower；likeachildlearningtotalk，herepeatsthesamecryagain，andagainheisanswered；hetriesexperimentswithalikeresult，andthespeakerandthehearerrejoicetogetherintheirnewly-discoveredfaculty。Atfirsttherewouldbefewsuchcries，andlittledangerofmistakingorconfusingthem。
　　Forthemindofprimitivemanhadanarrowrangeofperceptionsandfeelings；hissensesweremicroscopic；twentyorthirtysoundsorgestureswouldbeenoughforhim，norwouldhehaveanydifficultyinfindingthem。
　　Naturallyhebrokeoutintospeech——liketheyounginfanthelaughedandbabbled；butnotuntiltherewerehearersaswellasspeakersdidlanguagebegin。Nottheinterjectionorthevocalimitationoftheobject，buttheinterjectionorthevocalimitationoftheobjectunderstood，isthefirstrudimentofhumanspeech。
　　Afterawhilethewordgathersassociations，andhasanindependentexistence。Theimitationofthelion’sroarcallsupthefearsandhopesofthechase，whichareexcitedbyhisappearance。Inthemomentofhearingthesound，withoutanyappreciableinterval，theseandotherlatentexperienceswakeupinthemindofthehearer。Notonlydoeshereceiveanimpression，buthebringspreviousknowledgetobearuponthatimpression。
　　Necessarilythepictorialimagebecomeslessvivid，whiletheassociationofthenatureandhabitsoftheanimalismoredistinctlyperceived。Thepicturepassesintoasymbol，fortherewouldbetoomanyofthemandtheywouldcrowdthemind；thevocalimitation，too，isalwaysinprocessofbeinglostandbeingrenewed，justasthepictureisbroughtbackagaininthedescriptionofthepoet。Wordsnowcanbeusedmorefreelybecausetherearemoreofthem。Whatwasonceaninvoluntaryexpressionbecomesvoluntary。Notonlycanmenutteracryorcall，buttheycancommunicateandconverse；theycannotonlyusewords，buttheycanevenplaywiththem。Thewordisseparatedbothfromtheobjectandfromthemind；andslowlynationsandindividualsattaintoafullerconsciousnessofthemselves。
　　Parallelwiththismentalprocessthearticulationofsoundsisgraduallybecomingperfected。Thefinersensedetectsthedifferencesofthem，andbegins，firsttoagglomerate，thentodistinguishthem。Times，persons，places，relationsofallkinds，areexpressedbymodificationsofthem。
　　Theearliestpartsofspeech，aswemaycallthembyanticipation，likethefirstutterancesofchildren，probablypartookofthenatureofinterjectionsandnouns；thencameverbs；atlengththewholesentenceappeared，andrhythmandmetrefollowed。Eachstageintheprogressoflanguagewasaccompaniedbysomecorrespondingstageinthemindandcivilisationofman。Intime，whenthefamilybecameanation，thewildgrowthofdialectspassedintoalanguage。Thenarosepoetryandliterature。Wecanhardlyrealizetoourselveshowmuchwitheachimprovementoflanguagethepowersofthehumanmindwereenlarged；howtheinnerworldtooktheplaceofouter；howthepictorialorsymbolicaloranalogicalwordwasrefinedintoanotion；howlanguage，fairandlargeandfree，wasatlastcomplete。
　　Sowemayimaginethespeechofmantohavebegunaswiththecriesofanimals，orthestammeringlipsofchildren，andtohaveattainedbydegreestheperfectionofHomerandPlato。Yetwearefarfromsayingthatthisoranyothertheoryoflanguageisprovedbyfacts。Itisnotdifficulttoformanhypothesiswhichbyaseriesofimaginarytransitionswillbridgeoverthechasmwhichseparatesmanfromtheanimals。
　　Differencesofkindmayoftenbethusresolvedintodifferencesofdegree。
　　Butwemustnotassumethatwehaveinthiswaydiscoveredthetrueaccountofthem。Throughwhatstrugglestheharmonioususeoftheorgansofspeechwasacquired；towhatextenttheconditionsofhumanlifeweredifferent；
　　howfarthegeniusofindividualsmayhavecontributedtothediscoveryofthisasoftheotherarts，wecannotsay：Onlyweseemtoseethatlanguageisasmuchthecreationoftheearasofthetongue，andtheexpressionofamovementstirringtheheartsnotofonemanonlybutofmany，’asthetreesofthewoodarestirredbythewind。’Thetheoryisconsistentornotinconsistentwithourownmentalexperience，andthrowssomedegreeoflightuponadarkcornerofthehumanmind。
　　Inthelateranalysisoflanguage，wetracetheoppositeandcontrastedelementsoftheindividualandnation，ofthepastandpresent，oftheinwardandoutward，ofthesubjectandobject，ofthenotionalandrelational，oftherootorunchangingpartofthewordandofthechanginginflexion，ifsuchadistinctionbeadmitted，ofthevowelandtheconsonant，ofquantityandaccent，ofspeechandwriting，ofpoetryandprose。Weobservealsothereciprocalinfluenceofsoundsandconceptionsoneachother，liketheconnexionofbodyandmind；andfurtherremarkthatalthoughthenamesofobjectswereoriginallypropernames，asthegrammarianorlogicianmightcallthem，yetatalaterstagetheybecomeuniversalnotions，whichcombineintoparticularsandindividuals，andaretakenoutofthefirstrudeagglomerationofsoundsthattheymaybereplacedinahigherandmorelogicalorder。Weseethatinthesimplestsentencesarecontainedgrammarandlogic——thepartsofspeech，theEleaticphilosophyandtheKantiancategories。Socomplexislanguage，andsoexpressivenotonlyofthemeanestwantsofman，butofhishighestthoughts；sovariousaretheaspectsinwhichitisregardedbyus。Thenagain，whenwefollowthehistoryoflanguages，weobservethattheyarealwaysslowlymoving，halfdead，halfalive，halfsolid，halffluid；thebreathofamoment，yetliketheair，continuousinallagesandcountries，——liketheglacier，too，containingwithinthematricklingstreamwhichdepositsdebrisoftherocksoverwhichitpasses。Therewerehappymoments，aswemayconjecture，inthelivesofnations，atwhichtheycametothebirth——asinthegoldenageofliterature，themanandthetimeseemtoconspire；theeloquenceofthebardorchief，asinlatertimesthecreationsofthegreatwriterwhoistheexpressionofhisage，becameimpressedonthemindsoftheircountrymen，perhapsinthehourofsomecrisisofnationaldevelopment——amigration，aconquest，orthelike。Thepictureofthewordwhichwasbeginningtobelost，isnowrevived；thesoundagainechoestothesense；menfindthemselvescapablenotonlyofexpressingmorefeelings，anddescribingmoreobjects，butofexpressinganddescribingthembetter。Theworldbeforetheflood，thatistosay，theworldoften，twenty，ahundredthousandyearsago，haspassedawayandleftnosign。Butthebestconceptionthatwecanformofit，thoughimperfectanduncertain，isgainedfromtheanalogyofcausesstillinaction，somepowerfulandsudden，othersworkingslowlyinthecourseofinfiniteages。Somethingtoomaybeallowedto’thepersistencyofthestrongest，’to’thesurvivalofthefittest，’inthisasintheotherrealmsofnature。
　　Thesearesomeofthereflectionswhichthemodernphilosophyoflanguagesuggeststousaboutthepowersofthehumanmindandtheforcesandinfluencesbywhichtheeffortsofmentoutterarticulatesoundswereinspired。Yetinmakingtheseandsimilargeneralizationswemaynotealsodangerstowhichweareexposed。（1）Thereistheconfusionofideaswithfacts——ofmerepossibilities，andgeneralities，andmodesofconceptionwithactualanddefiniteknowledge。Thewords’evolution，’’birth，’’law，’
　　development，’’instinct，’’implicit，’’explicit，’andthelike，haveafalseclearnessorcomprehensiveness，whichaddsnothingtoourknowledge。
　　Themetaphorofafloweroratree，orsomeotherworkofnatureorart，isofteninlikemanneronlyapleasingpicture。（2）Thereisthefallacyofresolvingthelanguageswhichweknowintotheirparts，andthenimaginingthatwecandiscoverthenatureoflanguagebyreconstructingthem。（3）
　　Thereisthedangerofidentifyinglanguage，notwiththoughtsbutwithideas。（4）Thereistheerrorofsupposingthattheanalysisofgrammarandlogichasalwaysexisted，orthattheirdistinctionswerefamiliartoSocratesandPlato。（5）Thereisthefallacyofexaggerating，andalsoofdiminishingtheintervalwhichseparatesarticulatefrominarticulatelanguage——thecriesofanimalsfromthespeechofman——theinstinctsofanimalsfromthereasonofman。（6）Thereisthedangerwhichbesetsallenquiriesintotheearlyhistoryofman——ofinterpretingthepastbythepresent，andofsubstitutingthedefiniteandintelligibleforthetruebutdimoutlinewhichisthehorizonofhumanknowledge。
　　Thegreatestlightisthrownuponthenatureoflanguagebyanalogy。Wehavetheanalogyofthecriesofanimals，ofthesongsofbirds（’man，likethenightingale，isasingingbird，butiseverbindingupthoughtswithmusicalnotes’），ofmusic，ofchildrenlearningtospeak，ofbarbarousnationsinwhichthelinguisticinstinctisstillundecayed，ofourselveslearningtothinkandspeakanewlanguage，ofthedeafanddumbwhohavewordswithoutsounds，ofthevariousdisordersofspeech；andwehavetheafter-growthofmythology，which，likelanguage，isanunconsciouscreationofthehumanmind。Wecanobservethesocialandcollectiveinstinctsofanimals，andmayremarkhow，whendomesticated，theyhavethepowerofunderstandingbutnotofspeaking，whileontheotherhand，somebirdswhicharecomparativelydevoidofintelligence，makeanearerapproachtoarticulatespeech。Wemaynotehowintheanimalsthereisawantofthatsympathywithoneanotherwhichappearstobethesouloflanguage。Wecancomparetheuseofspeechwithothermentalandbodilyoperations；forspeechtooisakindofgesture，andinthechildorsavageaccompaniedwithgesture。Wemayobservethatthechildlearnstospeak，ashelearnstowalkortoeat，byanaturalimpulse；yetineithercasenotwithoutapowerofimitationwhichisalsonaturaltohim——heistaughttoread，buthebreaksforthspontaneouslyinspeech。Wecantracetheimpulsetobindtogethertheworldinideasbeginninginthefirsteffortstospeakandculminatinginphilosophy。Butthereremainsanelementwhichcannotbeexplained，orevenadequatelydescribed。Wecanunderstandhowmancreatesorconstructsconsciouslyandbydesign；andsee，ifwedonotunderstand，hownature，byalaw，callsintobeinganorganisedstructure。Buttheintermediateorganismwhichstandsbetweenmanandnature，whichistheworkofmindyetunconscious，andinwhichmindandmatterseemtomeet，andmindunperceivedtoherselfisreallylimitedbyallotherminds，isneitherunderstoodnorseenbyus，andiswithreluctanceadmittedtobeafact。
　　Languageisanaspectofman，ofnature，andofnations，thetransfigurationoftheworldinthought，themeeting-pointofthephysicalandmentalsciences，andalsothemirrorinwhichtheyarereflected，presentateverymomenttotheindividual，andyethavingasortofeternaloruniversalnature。Whenweanalyzeourownmentalprocesses，wefindwordseverywhereineverydegreeofclearnessandconsistency，fadingawayindreamsandmorelikepictures，rapidlysucceedingoneanotherinourwakingthoughts，attainingagreaterdistinctnessandconsecutivenessinspeech，andagreaterstillinwriting，takingtheplaceofoneanotherwhenwetrytobecomeemancipatedfromtheirinfluence。Forinallprocessesofthemindwhichareconsciouswearetalkingtoourselves；theattempttothinkwithoutwordsisamereillusion，——theyarealwaysreappearingwhenwefixourthoughts。Andspeechisnotaseparatefaculty，buttheexpressionofallourfaculties，towhichallourotherpowersofexpression，signs，looks，gestures，lendtheiraid，ofwhichtheinstrumentisnotthetongueonly，butmorethanhalfthehumanframe。
　　Themindsofmenaresometimescarriedontothinkoftheirlivesandoftheiractionsaslinksinachainofcausesandeffectsgoingbacktothebeginningoftime。Afewhaveseemedtolosethesenseoftheirownindividualityintheuniversalcauseornature。Inlikemannerwemightthinkofthewordswhichwedailyuse，asderivedfromthefirstspeechofman，andofallthelanguagesintheworld，astheexpressionsorvarietiesofasingleforceorlifeoflanguageofwhichthethoughtsofmenaretheaccident。Suchaconceptionenablesustograspthepowerandwonderoflanguages，andisverynaturaltothescientificphilologist。Forhe，likethemetaphysician，believesintherealityofthatwhichabsorbshisownmind。Nordowedenytheenormousinfluencewhichlanguagehasexercisedoverthought。Fixedwords，likefixedideas，haveoftengovernedtheworld。Butinsuchrepresentationsweattributetolanguagetoomuchthenatureofacause，andtoolittleofaneffect，——toomuchofanabsolute，toolittleofarelativecharacter，——toomuchofanideal，toolittleofamatter-of-factexistence。
　　Oragain，wemayframeasingleabstractnotionoflanguageofwhichallexistentlanguagesmaybesupposedtobetheperversion。Butwemustnotconceivethatthislogicalfigmenthadeverarealexistence，orisanythingmorethananeffortofthemindtogiveunitytoinfinitelyvariousphenomena。Thereisnoabstractlanguage’inrerumnatura，’anymorethanthereisanabstracttree，butonlylanguagesinvariousstagesofgrowth，maturity，anddecay。Nordootherlogicaldistinctionsorevengrammaticalexactlycorrespondtothefactsoflanguage；fortheytooareattemptstogiveunityandregularitytoasubjectwhichispartlyirregular。
　　Wefind，however，thattherearedistinctionsofanotherkindbywhichthisvastfieldoflanguageadmitsofbeingmappedout。Thereisthedistinctionbetweenbiliteralandtriliteralroots，andthevariousinflexionswhichaccompanythem；betweenthemeremechanicalcohesionofsoundsorwords，andthe’chemical’combinationofthemintoanewword；
　　thereisthedistinctionbetweenlanguageswhichhavehadafreeandfulldevelopmentoftheirorganisms，andlanguageswhichhavebeenstuntedintheirgrowth，——lamedintheirhandsorfeet，andneverabletoacquireafterwardsthepowersinwhichtheyaredeficient；thereisthedistinctionbetweensyntheticallanguageslikeGreekandLatin，whichhaveretainedtheirinflexions，andanalyticallanguageslikeEnglishorFrench，whichhavelostthem。Innumerableasarethelanguagesanddialectsofmankind，therearecomparativelyfewclassestowhichtheycanbereferred。
　　Anotherroadthroughthischaosisprovidedbythephysiologyofspeech。
　　Theorgansoflanguagearethesameinallmankind，andareonlycapableofutteringacertainnumberofsounds。Everymanhastongue，teeth，lips，palate，throat，mouth，whichhemaycloseoropen，andadaptinvariousways；making，first，vowelsandconsonants；andsecondly，otherclassesofletters。Theelementsofallspeech，liketheelementsofthemusicalscale，arefewandsimple，thoughadmittingofinfinitegradationsandcombinations。Whateverslightdifferencesexistintheuseorformationoftheseorgans，owingtoclimateorthesenseofeuphonyorothercauses，theyareasnothingcomparedwiththeiragreement。Herethenisarealbasisofunityinthestudyofphilology，unlikethatimaginaryabstractunityofwhichwewerejustnowspeaking。
　　Whetherweregardlanguagefromthepsychological，orhistorical，orphysiologicalpointofview，thematerialsofourknowledgeareinexhaustible。Thecomparisonsofchildrenlearningtospeak，ofbarbarousnations，ofmusicalnotes，ofthecriesofanimals，ofthesongofbirds，increaseourinsightintothenatureofhumanspeech。Manyobservationswhichwouldotherwisehaveescapedusaresuggestedbythem。Buttheydonotexplainwhy，inmanandinmanonly，thespeakermetwitharesponsefromthehearer，andthehalfarticulatesoundgraduallydevelopedintoSanscritandGreek。Theyhardlyenableustoapproachanynearerthesecretoftheoriginoflanguage，which，likesomeoftheothergreatsecretsofnature，——theoriginofbirthanddeath，orofanimallife，——
　　remainsinviolable。Thatproblemisindissolublyboundupwiththeoriginofman；andifweeverknowmoreoftheone，wemayexpecttoknowmoreoftheother。（CompareW。Humboldt，’UeberdieVerschiedenheitdesmenschlichenSprachbaues；’M。Muller，’LecturesontheScienceofLanguage；’Steinthal，’EinleitungindiePsychologieundSprachwissenschaft。’……
　　Itismorethansixteenyearssincetheprecedingremarkswerewritten，whichwithafewalterationshavenowbeenreprinted。Duringtheintervaltheprogressofphilologyhasbeenverygreat。Morelanguageshavebeencompared；theinnerstructureoflanguagehasbeenlaidbare；therelationsofsoundshavebeenmoreaccuratelydiscriminated；themannerinwhichdialectsaffectorareaffectedbytheliteraryorprincipalformofalanguageisbetterunderstood。Manymerelyverbalquestionshavebeeneliminated；theremainsoftheoldtraditionalmethodshavediedaway。Thestudyhaspassedfromthemetaphysicalintoanhistoricalstage。Grammarisnolongerconfusedwithlanguage，northeanatomyofwordsandsentenceswiththeirlifeanduse。Figuresofspeech，bywhichthevaguenessoftheoriesisoftenconcealed，havebeenstrippedoff；andweseelanguagemoreasittrulywas。Theimmensityofthesubjectisgraduallyrevealedtous，andthereignoflawbecomesapparent。Yetthelawisbutpartiallyseen；thetracesofitareoftenlostinthedistance。Forlanguageshaveanaturalbutnotaperfectgrowth；likeothercreationsofnatureintowhichthewillofmanenters，theyarefullofwhatwetermaccidentandirregularity。Andthedifficultiesofthesubjectbecomenotless，butgreater，asweproceed——itisoneofthosestudiesinwhichweseemtoknowlessasweknowmore；partlybecausewearenolongersatisfiedwiththevagueandsuperficialideasofitwhichprevailedfiftyyearsago；partlyalsobecausetheremainsofthelanguageswithwhichweareacquaintedalwayswere，andiftheyarestillliving，are，inastateoftransition；
　　andthirdly，becausetherearelacunaeinourknowledgeofthemwhichcanneverbefilledup。Notatenth，notahundredthpartofthemhasbeenpreserved。Yetthematerialsatourdisposalarefargreaterthananyindividualcanuse。Suchareafewofthegeneralreflectionswhichthepresentstateofphilologycallsup。
　　（1）Languageseemstobecomposite，butintoitsfirstelementsthephilologerhasneverbeenabletopenetrate。Howeverfarhegoesback，heneverarrivesatthebeginning；orrather，asinGeologyorinAstronomy，thereisnobeginning。Heistooapttosupposethatbybreakinguptheexistingformsoflanguageintotheirpartshewillarriveatapreviousstageofit，butheismerelyanalyzingwhatneverexisted，orisneverknowntohaveexisted，exceptinacompositeform。Hemaydividenounsandverbsintorootsandinflexions，buthehasnoevidencewhichwillshowthattheomegaoftuptoorthemuoftithemi，thoughanalogoustoego，me，eitherbecamepronounsorweregeneratedoutofpronouns。Tosaythat’pronouns，likeripefruit，droppedoutofverbs，’isamisleadingfigureofspeech。Althoughalllanguageshavesomecommonprinciples，thereisnoprimitiveformorformsoflanguageknowntous，ortobereasonablyimagined，fromwhichtheyarealldescended。Noinferencecanbedrawnfromlanguage，eitherfororagainsttheunityofthehumanrace。Noristhereanyproofthatwordswereeverusedwithoutanyrelationtoeachother。Whatevermaybethemeaningofasentenceorawordwhenappliedtoprimitivelanguage，itisprobablethatthesentenceismoreakintotheoriginalformthantheword，andthatthelaterstageoflanguageistheresultratherofanalysisthanofsynthesis，orpossiblyisacombinationofthetwo。Nor，again，arewesurethattheoriginalprocessoflearningtospeakwasthesameindifferentplacesoramongdifferentracesofmen。