ebook img

Of Glyphs and Glottography 1. The earliest writing - Archimedes PDF

29 Pages·2006·0.65 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Of Glyphs and Glottography 1. The earliest writing - Archimedes

DRAFT2006-04-01,toappearinLanguage&Communication Of Glyphs and Glottography * MalcolmD.Hyman MaxPlanckInstitutefortheHistoryofScience Boltzmannstraße22,D-14195Berlin,Germany Abstract: Thetermglottography isfamiliarfromPulgram(1976)andisequivalenttoGelb’s(1952) phonography.Thetermisproblematic, sinceitisrarelymadeclearhowawritingsystemis heldtorepresentaspokenlanguage,andhowglottographicsystemsexistalongsidenon- glottographicones.Thispaperexaminesglottographyfromtheparticularperspectiveof archaicorproto-cuneiform(firstattestedca.3300B.C.E.).Afterclearingupsomefundamental confusions,weofferfoursymptomsofglottographicwriting.Wethenconsidertherelation betweenglottographicandnon-glottographicwriting,associatetheformerwiththeactof readingandthelatterwithverbalizing,andconcludebysuggestingthatglottographyisbest viewedasasinglesubsystem ofwrittenlanguage. Keywords:proto-writing;proto-cuneiform; glottography;typologyof writingsystems Thiswritin’iswrotenrotten,ifyouhappento askme. —Popeye,in‘PopeyetheSailorMeetsAliBaba’sFortyThieves’(1937) 1. The earliest writing: an introduction WritingarosearoundtheendofthefourthmillenniumB.C.E.(ca.3300)insouthernBabylon 1 (modernIraq). Theearliestwritten documentsareclaytabletsimpressed withnumerical notationsandsealingsthatlikelyindicatedinstitutionalcontexts.Byca.3200asystem knownasarchaiccuneiformorproto-cuneiformhaddeveloped;thissystemistheprecursorto thecuneiformwritingthatdevelopedlaterasavehicle fortheexpressionofnumeroustexts inSumerian,Akkadian,andvariousotherlanguages.Anothergroupofarchaicdocuments fromaslightlylaterdate(ca. 3100–2900)comesfromearlysettlementsin(contemporary) 2 Iran; thewritingofthesetabletsisinanumberofrespectssimilartoproto-cuneiform,butit 3 neverthelessrepresentsadistinctsystem. *Tel.:+493022667129;fax:+493022667299; email:[email protected] 1 DRAFT2006-04-01,toappearinLanguage&Communication Thesearchaic textsbelongtoarestrictedrangeofTextsorten,beingconstitutedalmost exclusivelyofadministrativedocuments(recordsassociatedwiththeproductionand transferofagriculturalcommoditiesandmanufacturedgoods;recordsofthedisbursement offoodstuffstoworkers;etc.)andso-called‘lexicallists’ (thematicallyarranged enumerationsofdesignationsforsuchcategoriesasprofessions,places,animals,plants,and 4 manufactureditems). Lexicallistsapparentlyserved thepurposeoftransmittingthe conventionsofthewritingsystemandlaterevolvedintothefirstbilingual(Sumerian- Akkadian)dictionariesinthesecondandfirstmillennia. Althoughsignificantprogresshasbeenmadeinthedeciphermentandunderstandingof thesetextssincetheirpublication(startinginthefirstpartofthetwentiethcentury),crucial aspectsoftheirnatureremaincontroversial.Whilesomescholarshaveregardedarchaic cuneiformaspossessing(atbest)‘weakconnectionsto[spoken]language’(Damerow,1999, 5 2),othershaveseeninthetextscleartiestospokenlanguage. Particularlyvexedis‘the Sumerianquestion’: wasproto-cuneiforminvented(if thatistherightterm)byspeakersof Sumerian(alangueisoléeof theagglutinativetypeknownfromlater cuneiformdocuments) 6 orbyspeakersofsomeotherlanguage? Researchersaddresstwo sourcesinattemptingto answer‘theSumerianquestion’:(1)possible clueswithinthecorpusofproto-cuneiform tabletsthatunequivocallyindicateSumerianphonology,morphology,syntax,orlexicon;(2) evidencewithinattestedSumerianforthepresenceofasubstratelanguage,indicativeofan 7 aboriginalpopulation. Inwhatfollows,Iamnotconcernedwiththe‘Sumerianquestion’ perse.Rather,Iwishtoexplorethecriteriabywhichweidentifycertainwritingas glottographic(i.e. representingspokenlanguage),andtoanalyzetherelationbetweenwriting andspeechinbothpastandpresentcontexts. 2. Signal confusions 8 9 10 Sampson(1985, 29)divideswriting into glottography,whichrepresents spokenlanguage, 11 andsemasiography,whichrepresentsideationalcontentdirectly. Yetsemasiographyisnota 2 DRAFT2006-04-01,toappearinLanguage&Communication felicitousspecies:instancesofsemasiographyareunifiedonlybythefactthattheydonot notatespokenlanguage(thenotchesonthehandleofakiller’sknife,anilliteratepirate’s treasuremap,theicononthepapershredderthatwarnsmetokeepmynecktieatasafe distance,brandmarksontabletsofEcstasy,adiagramofafinitestateautomaton,Wilkins’ 12 ‘realcharacters’). AsPlatofamouslyobserved,thedichotomybarbarian/Greekleadspeople totheerroneousbelief thatthereisanaturalclassofbarbarians,whoexemplifyasharedset offeatures(Politicus262d–e).Iproposethatthetermsemasiographyisasvacuousasbarbarian —andthatbotharisefromunexaminedprejudices.Itisbetterthatweemploythesimple designationnon-glottographic. Certaintypologists(e.g.Hill, 1967;cf.Coulmas,1996)denythepossibilityofnon- glottographicwriting. Forthem,allwritingisfundamentallylinguistic—evenso-called 13 BildschriftandSachschrift(cf.Gelb,1952,4–6). SinceIdonotwishtogetboggeddownhere inwhatismeantbytheterm‘linguistic,’Iproposesimplytodefinewritingasnon- glottographicif (1)itcanbereadwithsimilarfacilitybyspeakersofdifferentlanguages, or (2)itsreadinghasthecharacterofparaphrase(i.e.twodifferent‘readings’arelikelyto employsignificantdifferencesinwordchoiceorsyntactic construction). Insomeaccountsofnon-glottographicwriting,however,spokenlanguagesneaksinbythe backdoor.Witness:‘Graphsofthissort[i.e.non-glottographic]mapdirectlyontotheworld ofrealthings,andreadingthemisatwo-stepprocessthatproceedsfirstfromGRAPH toTHING orACTION[…]andthenfromTHING/ACTIONtoNAME ofthing/action,thatis,toLEXICALMEANING’ (Boltz,2000,1–2). Theclaim(behindwhich oneistemptedtoseeanaïveHumboldtianism orbehaviorism)isthattoread suchsigns,wemustmapthemontoverbalconcepts.WhenI receivemyshirtbackfromthedrycleaner,Iseetheyellowsmileyfaceonthehanger;I picturearealface[cf.Boltz’s‘worldofrealthings’]—smiling;Ithinkofthewords‘smile’ (andperhaps‘Haveaniceday!’). Thisscenarioissurelytobediscounted;rather,thehappy faceleadsmebypriorassociationstoaparticularpsychologicalstate,withperhapsboth cognitiveandemotionalcomponents;verballanguageisnotneeded(whichisnottosay 3 DRAFT2006-04-01,toappearinLanguage&Communication thatwedon’toftenverbalizewhenconfrontedwithsignsorthings).Also—andthisisa pointtowhichIshallreturn—theverbalizationsthatmayarisewhenweareconfronted withthesignarenot predictable andnotfixedinrelationtothesign. Iemphasizethedividebetweennon-linguisticsignsandlinguistic conceptsbecause scholarsofearlywritingfindthemselvesalltoooftenintheprison-houseoflanguage. Considerthefollowingstatement,fromareputabletextbookofAkkadian: Functionally,signsintheirearlieststagewerelogographic,i.e.they representedoneormorespecificwords.Apurelylogographicwriting systemispersenotboundtoanyspecificlanguage:givenasequence ofpicture-signsrepresentingTHREE,SHEEP,GIVE,TEMPLE,agivensocial contextmightenableoneto‘read’(addingformalmodifications unexpressedinthewritingsystem)inEnglish‘threesheepweregiven tothetemple’orinFrench‘onadonnétroisbrebisautemple’as adequatelyasonecould‘read’suchatextinSumerian.(Caplice, 1988, 5) Theconfusionhereisremarkable.Theauthoridentifiesalogogramascorrespondingtoone ormorewords,butthendeniesthatlogographicwritingisglottographic!Ifproto-cuneiform exhibitsthelooseconnectionclaimedherebetweenasequenceofsignsandanutterancein aparticularlanguage(andsomeevidencesuggeststhatitdoes),thenproto-cuneiformisnot glottographic.Outsidelanguagetherearenowords(and,hence,nologographicwriting). 3. Symptoms of glottography InthissectionIidentifyfourcategoriesthatmightberegardedassymptomsofthe dependenceofawritingsystemonspokenlanguage.Iintroducetheseingeneraland summarizetheirapplicabilitytothecaseofproto-cuneiform. 3.1 Phonography 14 Byphonography,wemeanthenotationofspeechsounds. Obviouslysyllabariesand alphabets(includingdefectivealphabetsthatrecord onlyconsonantsounds;theseare 15 sometimescalledabjads)arephonographic. Sotooarethehybridsemi-alphabetic scripts usedinSouthandSoutheastAsia(Vaid,2002;Sproat,2000,45).Butasystemsuchas Chinese—inwhichcharacters(字zi4)representmonosyllabicmorphemes—alsopossesses 4 DRAFT2006-04-01,toappearinLanguage&Communication aphonographicaspect.InChineseorthography,manycharactersarecomplexesgenerated fromasemanticradicalandaphoneticcomplementthatindicatesthepronunciation.Thus 楓feng1‘maple’isanalyzedas木mu4‘tree’+風feng1‘wind’.AndsoaChinese reader recognizesthat楓referstoatree (radical ontheleft)thatispronounced likethewordfor wind(phoneticontheright).Thephoneticcorrespondenceneednotbeexact.Thatsuch phoneticcompoundingisrecursiveaccountsforitsastonishingproductivityinChinese 16 writing(80%ofcharactersemploythedevice );thusthephoneticof 楓(i.e.風‘wind’) is furtheranalyzedas虫hui3‘insect’+凡fan2‘all’(Mattingly&Hsiao,1999;Sproat,2000,49– 50;cf.Martin,1972;Yin,1994,21–25).(Admittedlythisanalysisisphoneticallyand semanticallylessperspicuousthantheformer;theevolutionofwritingsystemstendsto introducesomemeasureofopacity;cf.Halliday, 1981,128.) Thereislimitedevidenceforphonographyintheearliestproto-cuneiform,fromtheUruk 17 IVa/IIIperiods(ca. 3200–3000B.C.E.). Thistakestwoforms.First,putativerebusreadings, suchasTI‘arrow’(pictogramofabowandarrow) usedfortheSumerianhomophone‘life, tolive’,inarchaicspellings(EN E TI )ofthecommonformula(d)En.lil .ti‘mayEnlil give a 2a a 2 18 life’. Second,putativephoneticcomplements,similartothosejustdiscussedforChinese: 19 e.g.GA ×EN readMEN‘crown’;EZEN×ENreadEZEN ‘feast’ratherthanthevariantfor 2 x theformersign, BAD ‘wall,enclosure’(Steinkeller,1995,694–695;Glassner,2003,145). x Third,putativesyllabicspellings:e.g.MAŠ+GAN fortheAkkadianloanmaškanu 2 20 ‘threshingfloor;smallagriculturalsettlement’(Steinkeller,1995,695;Glassner, 2003,161); GAL+KIN forKINGAL‘headoftheassembly’(Glassner,2003,146;N.B.thereversed 5 21 spelling). 22 Notoneoftheseexamplesisuncontroversial. Evenamongscholarswhoacceptthe evidenceforphonographyinthearchaicscript,itisacknowledgedonlythat‘thereseemsto 23 beasmallamountofphoneticwritinginproto-cuneiform’(Cooper,2004,80). 3.2 La chaine de la parôle 5 DRAFT2006-04-01,toappearinLanguage&Communication Althoughnaturallanguagesyntaxcompriseshierarchical relationships(representedbytrees ingenerativegrammar),speechisrealizedasalinearsequenceofwords(orofmorphemes). Weexpectaglottographicwritingsystemtoencodespeechasastringofgraphemes.This stringmustbeaccommodatedtothetwo-orthree-dimensionalsurface uponwhichitis 24 inscribed(e.g.bymakingprovision forline-breaks); butthedominantprincipleisthat graphemesarearrangedin anorderthatcorrespondstotheirlicensingelementsinthe 25 temporalchainofspeech. (Thoughthelettersofane.e.cummingspoemsprawldrunkenly downthepage,theirarrangementisdeterminedinlargepartbyacodethatexistsin additiontotheglottographicone—Howtoreproducethearrangementorally?) Proto-cuneiformtabletsexhibitastructurethatbearsvirtuallynoresemblancetonatural- languagesyntax.‘Syntax[…]wasscarceornon-existent.Individualentriesandsummary remarkswereenclosedincases(rectangularboxes)ontablets,andthesecasescouldbe arrangedinverticalcolumns, butwithincasestheplacementofsignswasarbitrary,except thatnumeralscamefirst’(Cooper,2004,81).‘Thepositionofthesignswithinthetextswas determinedtoalesserextentbycriteriarelatingtolanguagesyntaxthanbythe“tablet format.”Thislayoutwasspecifictoparticulareconomictransactions;itdeterminedhow andwhereagivenbitofinformationwastobedisplayed.[…]Thevarioustabletsfromthe earliestphasesofwritingthereforebearcloserresemblancetosuchmoderndocumentsas punchedcards, dockets,clearingchecks,balancesheetsormanyotherformalizeddata carriersthantoindependentlyandfreelycomposedmanuscriptsinthemodernsense’ 26 (Nissenetal.,1993,30). Lexical lists(which comprisemorethan10%oftheproto-cuneiformcorpus)followa columnarformatwhereeachcasecontains(1)thesignN ,whichhasthevalue‘1’inthe 1 sexagesimalnumericsystem(thesignseemsheretoservemuchthesamepurposeasthe modernbulletindemarcatingthebeginningofanentryinalist),and(2)oneormoresigns correspondingtoadesignationofthecategorywithwhichthelist isconcerned(e.g.a 27 profession,nameofacity, typeofanimal,etc.). 6 DRAFT2006-04-01,toappearinLanguage&Communication Aphenomenonconcomitantwiththe(neartotal)absenceofnaturallanguagesyntaxisthe absenceofgrammatical functionwordsandmorphs.Signsonadministrativetablets correspondgenerallytonumericnotations,saliententitiessuchasagriculturalproductsand disbursements,andadministrativetitles/names.Therelevantactionsaregenerallytobe inferredfromthedocumenttype.OnlybytheUrperiod(ca.2800)arethereclearlyreliable indicationsofSumerianfunctionmorphs(Cooper,2004,80). 3.3 Formal characteristics Whatistheformofglottographicwriting?Inasuperficialsense,itisaline:left-to-right(for English),right-to-left(forHebrew),top-to-bottom(forMongolian), boustrophêdon(i.e. alternatingdirection:left-to-right,thenright-to-left,etc.)inthecaseofseveralpaleographic 28 traditions. Sometimeswritingfollowsadifferentcurveinordertoaccommodateitself to itsmaterial:thusaninscriptionaroundtherimofaplatewill followthecircumference. Onlyrarelydoesglottographytransfigureitselfcompletely,intosomethinglike Apollinaire’sCalligrammes,orthecalligraphic rendition ofsūra105oftheQur’an(al-Fīl)that takestheformofanelephant. Letusbeginwithontogeny.Howdochildrenwrite,beforetheyhavebeentaught?Atthe firststage,thereisundifferentiatedwriting(Tolchinsky, 2003,56–62;Gibson,1972,12). Childrenatthisstagearecapableofdistinguishingbetween(whatconstitutesfortheir culture)writingandnon-writing.Whentheywrite,words(orothersalientlinguistic units) maynotbeidentifiable; yetwefindsomecriticalfeatures:‘linearity, presenceof 29 distinguishableunits, regularityofblanks,anddirectionality’(Tolchinsky,2003,58). These featuresdifferentiatewrittenlanguagefrompictures; similarly,formalconfigurations distinguishwrittenlanguagefromnumericnotations.Childrenwhohavenotyetlearnedto readwill rejectstringsofrepeatedtokensas‘goodforreading’(e.g.AAAA),butwillallow themaspotentialnumericalnotations‘goodforcounting’(e.g.2222)(Tolchinsky,2003,152– 157). 7 DRAFT2006-04-01,toappearinLanguage&Communication Formreachesmuchdeeper,however.Inthe1930sand40sGeorgeKingsleyZipfdiscovered anumberoffactsaboutthedistributionofitemsinlanguage.Bestknownis‘Zipf’sLaw,’ whichstatesanexpectedrelationbetweentherankorderofawordinatextanditsfrequency. Aword’srankisitsmeasureoffrequencyincomparisonwithotherwords(sothattheword withrank1isthemostfrequentwordinthetext;thewordwithrank2,thesecondmost frequent;etc.);aword’sfrequencyisthetotalnumberofoccurrencesofthewordinthetext (orsample).Zipf’slawpredictsthatinatext/sampleofsufficientlength,thefrequencyof wordswill followaharmonic series, suchthatthewordwithrank2occursapproximately halfasoftenasthatwithrank1,thewordwithrank3occursapproximatelyonethirdas oftenasthatwithrank1,andsoon.Thustherank-frequencypointswilllieonornearthe hyperbolic curverf=C,whererisaterm’srank,fitsfrequency,andCisaconstant.When plottedondoublylogarithmicscales, thiscurvewill appearasalinewithslope=-1(Zipf, 1949,19–55). Thepredictionhasrepeatedlybeenshowntoholdtrueforwordsintext samplesdrawnfrommanylanguages,periods,andgenres.Asanillustration,FIGURE1plots 30 therank-frequencyrelationinaLatintextofthe16thcentury. Theproto-cuneiformcorpusconsistsofca.5820tablets, whichcontainca.42000signs, 31 excludingnumericsigns(Englund,1998, 65). Thesetokensrepresent 1618signtypes provisionallyidentifiedasdistinct,althoughtheworkondecidingwhich signforms 32 representmereallographsandwhich aregraphemicallydistinctisbynomeanscomplete. Ifwetreattheentirecorpusasasinglesample,wecanplotsignsbyrank-frequency,aswe 33 didabove(FIGURE2). It isimmediatelyevidentthattheplotdifferssubstantiallyfromthat expectedforwordsinanaturallanguage;inparticular,asrankincreases,frequency decreasesconsiderablymoreslowly.Thisresultisnotsurprising,as(1)thereisnoreasonto supposethatproto-cuneiformsignsrepresentwords, and(2)ithasnotbeendemonstrated thatproto-cuneiformisinfactglottographic. Ontheotherhand,thedistributionofproto-cuneiformsignsbroadlyresemblesthe distributionofmorphemesinanaturallanguage(seeZipf,1949,87–97). Bycomparison,we 8 DRAFT2006-04-01,toappearinLanguage&Communication shalllookataclassical Chinesetext.TheChinesewritingsystemisoverwhelming morphosyllabic, inthatitscharacters(字)typically representmonosyllabic morphemes (Ramsey,1989,58–59;Yin,1994,311–312;cf. supra,§3.1).Thustherank-frequency distributionofChinesecharactersisverysimilartothatofnatural-languagemorphemes 34 (FIGURE 3). Thesimilaritybetweentheproto-cuneiformandChinesedataisstriking.Yetof course,wecannotconcludethatproto-cuneiformsignsrepresentmorphemesinanatural language;all wecanconcludewithsafetyisthattheirdistributionandfunctioningisin certainrespectsakintonaturallanguagemorphemes. Oneaspectofthisfunctional similarityisthatsignsarecombinedtoformnames/titles,suchasKUŠIM(KU+ŠIM), apparentlyanadministrativeofficialnamedoneighteentabletsoftheErlenmeyercollection 35 (Nissenetal.,1993,36–46). Likewise,qualifieddesignationsinvolvesigncombinations wherethecomponentsmightbethoughtofas‘morphemic’(regardlessofwhetherwetake thearchaictextsasbeingglottographicorasbeingcomposedinapurelywritten Sondersprache):e.g.KUR+AMAR‘bullcalf’, SAL+AMAR‘heifercalf’,AB +GA‘milkcow’, 2 AMAR+GA‘suckling(calf)’(Nissen etal.,1993,89). 3.4 Mediation Inapassagethathasbeenquotedtothepointoftedium,Aristotledeclaresthatwritten wordsarethesignsofspokenwords,which inturnarethesignsofpsychological states(De interpretatione16a4–6).That,ofcourse,isastatementaboutglottography. Non-glottographic writinglacksthemediationofspokenlanguage;itsnotationsarethemselvessignsof 36 psychologicalstates.Proto-cuneiformsignsdepictthings, butbeyondtheobjectthatis moststraightforwardlydepicted, thereliesasemanticconstellation:‘apictogramcanrelate tootherrealities, attachedtothesameobjectbymentalprocessesthataremoreorless foundedinreality, ordownrightconventional,andperfectlywellknownandutilizedinthe 37 representationofart’(Bottéro,1992,76). Newsignscanbecreatedfromexistingsignsby producingtableaux;forinstance,SAG‘head,person’+NINDA‘rationingvessel’(a pictogramofabeveled-rimbowl)yieldsGU ‘disbursement(offoodrations)’, which comes 7 9 DRAFT2006-04-01,toappearinLanguage&Communication eventuallythroughtheineluctableprocessofsemanticchangetomeansimply‘eat’(Nissen etal.,1993,14–15).Writing,however,differsfromthefigurativeartsinthatiteschews particularsandseeksratherthegeneric,theuniversal,theuniform,theconventional (Bottéro,1992,77).Whenasignisdrawnagain,itisnotanewpicture,butratheranother instanceofthesamesign.Justastheformofthesignbecomesstereotyped,sodoesits semantics.Thesignisnownolongerinterpretivelyopen-endedinthefashionofapicture; rather,itbecomeslimitedtoafewconventionalusages(onepresumesthatthelexical lists playedaroleinthisprocess). Thisgeneralizationofformandfunctionallowsusto distinguishnon-glottographicwritingfrompicturesandothermanifestationsofthe figurativearts,andallowsustospeaknotincoherentlyofnon-glottographicwriting systems. Thelastpoint,however, isevidentlycontroversial. Unger&DeFrancisassert,bluntly,‘[a] semasiographicwritingsystemisacontradictioninterms’(1995,53). Theirargumentisnot thatnon-linguistic signsareimpossible perse,butratherthattheycannotserveasthebasis foracommunicativesystembecause:(1)picturesarenotself-explanatory, and(2)awriting systemcouldnotbedevelopedwithoutrecoursetospokenlanguageforthepurposeof 38 descriptionandexplanation. Totheformer,wemayquotePettersson: ‘DeFrancisisof courserightinthatpicturesarenotunderstandablebythemselves—ittakesacultureto 39 explainthem,namelytheverycultureinwhich theywereproduced’(1996,51). Tothe latter,IshoulddoubtthatanyonebelievestheUrukscribesdidnotspeakamutually intelligible languageamongstthemselves.Butitissenselesstoassertthatrecoursetoa spokenlanguagefortheexplanationandteachingofawritingsystemmakesthatsystem glottographic;weshouldthenbeobligedtoclassifyvirtuallyalltechnologyaslinguisticin nature,becausehumansrelyonspokencommunicationinitscreationanduse. 4. Glottographic and non-glottographic writing Theevidenceavailable atpresentseemsinsufficienttomakeacleardiagnosisof 10

Description:
Apr 1, 2006 Popeye, in 'Popeye the Sailor Meets Ali Baba's Forty Thieves' (1937). 1. The earliest writing: an introduction. Writing arose around the end of the fourth millennium B.C.E. (ca. spoken language for the explanation and teaching of a writing United States Department of Agriculture to
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.