Ana SadiLabnani lu allik
Beitrlge zu altorientalischenund mittelmeerischen Kulturen
Festschriftftir Volfgang Rollig
Herausgegeben von B eate Pongratz- Leisten Flartmut Krihne Paolo Xella
/Ycftr LqV Sonderdruck
1997
VerlagButzon & BerckerKevelaer
NeukirchenerVerlag Neukirchen-Vluyn The Man Without a Scribe and the Questionof Literacy in the Assyrian Empirex
SimoParpola, Helsinki
K 652 (ABL 151) is a shortbut difficult Neo-Assyrianletter hitherto be- lievedto refer to the makingof a royal imageand relief figuresfor Assyrian governmentofficials. This interpretationderives from E,.Klauber, who in his studyAssyrisches Beamtentum (Leipzig, 1910),p. 103f, renderedthe words la-a-ni and a-su-mu occurringin the text as ,,Bild" and ,,Relief' respectivelyand comparedthem with the inscribedstelae of Assyriankings and governorsactually found at Assur and other Assyrian sites. At first glance, this interpretationmakes senseand has, in fact, been generally accepted.Later translationsof the letter (L, Waterrnan,RCAE [1930],no. 151; R, Pfeiffer, SZI [935], no. 143; ChicagoAssvrian Dictionary, B [1965]213, Al2 [1968]348 f, andL [973] 79) differfrom Klauberonly in minor points:
[Waterman] ,,To the king my lord, your servantSinna'id. May it be well with the king my lord. ,,In regard to that which the king my lord has written me, he should surnmon (i-ie-si-a)a scribe.Let him designan image(la-a-ni) of the king, a relief figure (a-su-mu)for the governorof the city of Arrapha,a relief figurefor ..."
fPfeiffer] ,,To the king, my lord, your servantSin-naid. Greetings to the king my lord. ,,Thesecretary is giving orders(i-ie-si-a) concerning what the king my lord has written me, that they shouldmake an imageof the king, a bas-relieffor the pre- fectof Arrapkha,a bas-relieffor..."
[CAD] .,Asto what the king my lord haswritten me, (,..) let themdraw an image of the king. let them... a relieffigure for the governorof Arrapha,let them... a relieffigure for PN." See also AHU'. p. 77, s.v. " ,,Bildstele(?),
* It givesme greatpleasure to dedicatethis articleto WolfgangRollig, a long-time memberof the advisory committeeof the Neo-AssyrianText CorpusProject, whose work has always been a sourceof inspirationto me. Abbreviationsare thoseof the ChicagoAssyrian Dictionarv andAkkadisches Handwrirterbuch, with the following additions: KAI : H. Donner and W. Rollig, Kanaanriische und aramciische Inschriften(2nd ed.,Wiesbaden, 1966); NL : H.W.F. Saggs,,,The Nimrud Letters," Iraq l7 (1955)2l ff, etc.,cited by textnumbers. 316 S,Parpola
This consensusdoes not mean, however, that the letter has necessarily been correctly understood.Quite the contrary. All the publishedtranslations suffer from numerous flaws making them totally obsolete today.r In particu- lar, the idea that the letter would refer to the making of a royal image and relief figures has to be abandoned.Thc word lanu did not mean ,,image" in Neo-Assyrran,'nor has the word a-su-mu (a hapax legomenon)any connec- tion to ,,relief figures" except its assonanceto tsabylonianasumitttt ,,stela."3 Overall, the text contains many unusual and problcmatic features, some of which (like thc form i-ie-si-a in line 7) are anomalouscnough to raisc the suspicion that something is wrong with the cuneiform copy on which the translationshave been based. Collation of the original confirms this suspicion.Instead of i-ie-si-a, thc tablet actually readsi-ie-e!-a. This word also occurs in rev. 3. anclin general the reverse of the tablet, indicated as largely damaged in the copy, turns out to be completely readable.oThe allegedlct-a-ni in obv, 7 acruallyrcacls la-o-sil.
' ana bEt (obv. 5) does not mean ,,in regard to" (waterman), ,,concerning" (pfeilfer) or ,,as to" (cAD), but ,,where(to)", rarely ,,when" (see the examples cited in n. I I below); iipurannlni (obv. 6) does not mean ,,wrote me" but ,,sentme" (see n. I l. and note also SAA l0 316 s.2 and 318 s.l,,,l have gone where the king, my lord, sent me"); i-ie-si-a (obv. 7) can under no circumstancesbe taken as a form of sasu ,,to call, summon," cf. simply cAD Stz 11992) under iasi; li-is-pu-ru (obv. 9) cannot be read ltsiru ,,let them draw" (CAD) since the sign su did not have the value /sir/ (seevon Soden,Akk. Syll., no.2I3); reading li-is-sfr-ru(Klauber, Water- man, P1-eiffer)is out of the question since neither eseru A ,,to collcct" nor esEru B ,,to shut in" makes sense in the context and since neither verb is attested in Neo- Assyrian (cf. already S. Ylvisaker, LSS 516ll9l2l. p. 3l) For la-a-ni and a-su-mu see presently. t l*, is well attested in Neo-Assyrian, but only in the meanings ,,body, staturc. figure,height"; see SAA 2 5 iv 16,6:610;SAA 3 13:12.15,23r.5,38 r. 10,39:10; sAA 5 156r.3; SAA 9 9.14.15;sAA l0 349:14;ABL 1078r. l;ADD 310:5anct 312:4; crN 3 95 r. 11.17.22.26;ND 2082',4(traq l6 34), Ass. 8476h r.2 and 9687:4. The Neo-Assyrian term for,,image" was salrnu, cf-.e.g. lsat-mlu.ia RvR ILUGAIL ,,image of the queen mother," cT 53 921 r.17, -cal-mu-LtJGAL.,royal image," CT 53 4l:14.16 and r.2, and 2 sslt.-fmu-.Lucnll.vES-ni ,,two ro,u-al images," ibid. l2; also, written NU MANMrs-ni, cr 53 l8:6; Nu-nanNvEs, ABL ll94:13, cr 53 516:2; 1Nu1-LUGAT-Mrrs-ni,ABL l09g:ll; [N1uLUGAL-rzi, ABL 951:19;NU-MAN, sAA 7 62 i 14, 1i 12, iii 10, r. i 10. The readingla-a-ni involves fwo particular difficulties rendering it suspect:the ,,overhanging" -i, which is incon- gruous with the rules of Neo-Assyrian phonology and morphology, and the separa- tir:n from LUGAL in the next line, which would be inexplicable had ,,royal image" been meant by the writer. 3 See below. The attestedNeo-Assyrian forms of the word are NA4.as-mit-tu, SAA 3 29 r.4; NAo.ui-me-la,Scheil rn II 60, ui-me-le, ND 2774:5 (rraq23 p|.26); and,u- su-mit-tu,SAA l0 221 r.25. a I collated the reverse of the tablet in November, 1966, the obverse (from photo) in the early eighties. All the new readings indicated with exclamation marks in the The Man Withouta Scnbc 311
On the other hand, the hapax legomenona-su-mu in obv. l0 and l2 as well as the difficult li-is-pu-nr in obv. 9 and rcr,. 4 turn out to be correctly copied, and on the whole thc copy proves reasonablyaccuratc. Thus, colla- tion does not eliminate the unusual charactcrof the text but rather empha- sizes it. Like the deletedi-ie-si-a and la-o-ni (and the obscurea-su-ntu and li-is-pu-ru), the new word forms i-.ie-e-tt and,la-a-si are not attested any- where else. The difficultics posed by thesc u ords quickly disappear,however, when it is realized that in spelling thr-nr ihc uritcr did not fcrllow the standard (s) Neo-Assyrian orthographv. u'herr- uraphrc stands for spoken [5] and (s): vice versa,tbut ratherthe Necr-Bab',itrnran\\'stcll. u'here [s] and <5> : : tS] Thus i-ie-e-o corresprrnd:.tr rtrrnrsl Nc'o-Assyrian translttcratlon'.:.r'f; .Itca-.:r.i : --:-t- i- ', . ival I ri 15|1to thank the Trustc.'s r.,i'iirr Il::.-.:. \1-.>r-::: :lJr.:Jnr().iucr-d and lirr the 33nl:.:l-11 . :-r...:: .--.::l '.crh *rrir-q.iand the n i3irnp -1 Li :--ra-..-,1.S t.t ^ l:l l-. i-i--'.-g-,i.rbrti :(): l-l and S.\.\ lri_1-<-l_< 1ttli. t-st-iu and l-s--st-i..1passlm: l-sc-e-A-.1..*tth )r)u." \L la'i[) (lraq 1.. p1 9i.:-,-.,t-e-krt.C-I -i3 908 r. 2. i-se-ka and rs-.ie-A'apassirn r-[.se-e]-.irr.,u'ithhrm." \L i 9 tlracl i7 pl 4). i-se-e-ii,SAA I 29,31,35r.3.124'22 andr 12.SAA l0 369i:. '\BL,i,i7r f. is- se-e-iu,SAA I 4:13and 13 r.9, SAA l0 97 r.8 and353 r 7. ABL 951r -16 8. r- se-iu and is-se-sripassim,' i-se-e-ni,,with us". ABL 621* r. 1.1.t-se-nt and r.s-.se-ni passim;etc. The frequentspellings with i- imply that the geminate figure,"e-su-mu, is in realityjust a variantof Transliteration Translation I a-na LUGAL BE-ital lTo the king, my lord: your servant md36_r1r 2 ARAD_ta Sin-na'di. Good health to the king, my a J (lu)DI-mu a-naLUGAL lord! 4 be-li-ia /i-is) and thereforemuch used by professionalscribes, but using a CV-VC sign combinationinstead did the sametrick and helpedreduce the numberof cuneiform graphemesto be mastereddrastically. 'cf., e.g., SAA | 99 r. 17,iim-mu LU.qur-bu-teium-mu LtJ.ia--EN.NLfN[lit-fli-ka ,,Let either a bodyguardor a guard come"; similarly ABL 556 r. 17. For further examplesof disjuncriveiummu... iummu,,either... or" seeSAA | 4l r.5 f,4g:13f. 139:4ff, 220:5f and r. 8 ff, sAA 5 213:9ff (quotedin n. g), sAA I 0 152:7, 194 r. 14,362:6,ABL 1056r.4ff, etc.The prothetica- attachedto su-muis certainly due to analogywith ktma,,when, i1," which is often found with this prothetica- especiallyin lettersfrom Der andBabylonia (e.g., ABL 800:9,861s. l, 1063r. ll, ABL 1296r.6, cr 53 77 s.4,716:5t note, in themeaning ,,if," ABL 2ll r. 15 ff, a ki-ma ina 5A a-bi-tean-ni-te qur-ba-ku,,if I am involvedin this matter(let the king punishme)"; also ADD 102 r.5). Note alsoa-ki-i besideki-i,,as, if, whether; (passim);a(m)-mar besidemer ,,asmuch as" (passim),etc. Occasionalspellings of iummuwith the MU signin Assur(TCL 9 628 r, l, FWA l07q r. 2, Ass. g57Il r. 2, 9634:7,9644d :6, 9644e'.7,9661d:7, 9661kJ, l3846adr. 2) imply a mergerwith sumu,,name"and thus a variantpronunciation [su:mu] beside the normal [summu]. "'cf. o-ro-si and ui-ie-si-ilb] for normal 5 a-nabe-et -sI have no scribe where the king sent 6 LUGAL i!-pu-ra-ni-ni me to. 7 LIJ*.A.BAi-ie-el-a 8 la-a-si: 9 LUGAI- li-ts-pu-ru 9 Let the king order either the governor l0 e su-nltt.l-l,tLU*.1:N.NAM of Arrapha or A55ur-belu-taqqin to I I ia URU :trrap-ra-ap-ha sendme one. e.l2 |e) su-rtttt r. I tot_,,1rn.7-i_irrrlULAL! 2 r1' -grt LL'*i, A.BA . J^ 'la.l --tc-c -,1 -!)Lt-t'u 4 [/r]-t;' Thus the ntcssagr-of the letter actually turns out to be very simple: the sender sirrpi'' inlorms the king that he didn't have a scribe and that he neededsn.- \\-har is interestingis that this circumstancedid not prevent him from puttrns hr. ntessageinto writing and drafting the presentletter. Lines 5 f and I ii iniplr that the sender was on a mission in the Zagros moun- tains.rru'here reople knowing Assyrian cuneiform certainly did not grow on trees,'tHenc.- he must have either written the letter himself or had someone " Fo. irn.'>j I ;i SAA 5 226:5 f, ina E LUGALbe-li ina UGULU.GAL.MES ii-pur-an- ni-ni ..u hcn ::re king sent me to the magnates(in Media)"; CT 53 14l'.6 ff be-et LUGAL[re--.: ,r-rrirKUR GAL KAS.LULii-pur-an-ni-ni,,when the king my lord, sent me to the land rrt'thechief cupbearer",ABL 992:13, rcUnbe-etLUGAI be-li ii-ku-ni- ni ..(the klns. mr' lord, knows) the land where the king, my lord, stationed me" (refernng tc Phoenrcia). Note further, ref-erringto missions in the Zagros area: NL 4l r i i riraq lt) pl. 38), a-ki LUGAL be-li ina UGU LU.LUI-.MIISii-pur-lanl-ni-ni ,,when the krng. rny lord, sent me to (catch) the criminals, (I went there)"; SAA 5 227.21 f. i:;-i Lt-ma-ai-li-kan-a-ni E Lucet be-li i3-pur-iu-u-ni ,,as he now came, ',i (going i 1p irere the king, my lord, sent him"; and SAA 5 198 r. 3 ff, lbe-et) LUGAI- tu1-ta-me-S[i] be-li j)i-pttr-rtt-ni-ni ,,1 have set out [to where] the king, my lord, " sentme The lack of a blessing formula in our letter dates it to the reign of Sargon II. A55ur- belu-taqqin (rev. l) is known to have served at this time as a governor in or in the viciniq'of the Transtigridian city of Meturna (Tell Haddad on the Diyala), 160 km SE oiAnapha (Kirkuk); see AtsL 455 r.5 and the other letters lrom or mentioning A55ur-belu-taqqin(ABL 212,887,1057, CT 53 64, ABL 438,638, 1296,CT 53 6 and 244), The fact that the desired scribe was to be supplied either by the governor of Arrapha or by A55ur-belu-taqqin but not, e.g., the governor of Der (Tell Aqar), implies that the sender was located somewhere in the upper reaches of the Diyala river. 12Occasional letters from vassal rulers (see SAA 5164-168) show that Assyrian cuneiform was used at least in some Zagros principalities under direct Assyrian influence. The level of scribal competence evidenced by these letters is low, indicat- ing that the use of the script was limited to the vassal court only. 320 S. Parpola in his retinue write it.r3 Whichever the case, the writer evidently was an ordinary Assyrian administrator not generally thought to have been able to read or write. r4 This explains the many unusual spellings, word forms and phrases occurring in the text. Not being a professional scribe, the writer simply was not able to adequately follow the standard spelling and phrasing conventions of the Neo-Assyrian royal correspondence.At the same time it is clear, however, that he was by no means a mere dilettante either but must have had considerableprevious writing experience.The tablet, expertly moulded, has the typically Neo-Assyrian lcttcr fbrmat,r5 the introductory formula agreeswith contemporary conventions, and the signs are drawn and distrib- uted over the lines with a skill that can be acquired only through long and repeatedpractice. If an ordinary Assyrian government official possessed such writing skills, there is every reason to believe that literacy in the Assyrian empire was far more widespread than hitherto assumed.toAfter all, mastering the elementsof cuneiform script (the basic syllabogramsand the most important logograms) does not take longer than a semestertoday. This ler,el of compe- tence is sufficient for writing and reading simple texts, and it is the level ''' The forms of the signs and severalorthographic details agreeing with standard Assyrianconventions (in the first placethe spellingof the city nameAnapha with the liuuu sign)betray the Assyrianorigin of thewriter. It couldbe objectedthat he might aswell havebeen a local scribetrained in Assyria;but in that casethe sender would (contraryto the wordingof the letter)have had a scribeat his disposal! 'o See,e.g., A.K. Grayson,Cambridge Ancient Historv,2nd ed., llll2 (1991),p.202. ,,Thereseems to have been no training programmefor potentialbureaucrats nor werethe officialsliterate, since an army of scribesbolstered up theentire system." 't Thetabletmeasures Ll x3,l x6.2 cm anclthus displays the standard l:2:4egirru ratio betweenits thickness,width and length.See S. Parpola, JNES 42 (1983)2 n, 5 and LAS I (1970) 331-341,and K. Radner,,,The Relation Between Format and Contentof Neo-AssyrianTexts," in R. Mattila,ed., Nineveh 612 BC. The Glon,and Fall of theAssvrian Empire (Helsinki, I 995),63-77 , especially7l f. 16 The current scholarly consensusseems to be that after relatively wide-spread literacyin the early secondmillennium BC (seen. l9) therewas a,,reversal"in the developmentof the cuneiformscript, which eventually,,led to literacybecoming the prerogativeof a restrictedclass of highly trainedspecialists [in] the late secondand earlv first millennia" (M.T.Larsen, ,,TheMesopotamian Lukewarm Mind: Reflec- tions on Science,Divination and Literacy," in F. Rochberg-Halton,ed., Language, Literature, and Historv: Philological Studies Presented to Erica Reiner [New Haven, 1987], p. 220, with referenceto earlier opinions);see also J.S.Cooper, InternationalEnc.vclopaedia of CommunicationsI (Oxford, 1989), 442, and I. Oates,Babvlon (rev. ed., London. 1986). 163. The Man Without a Scnbe 321 evidencedin our letter." Naturally, full masteryof the cuneiformwriting systemtakes a iifetime and can be attainedonly by a small numberof pro- fessionals.But sucha masterywas not necessaryin everydaylife. The level of literacy evidencedby the presentletter was within the reach of every affluentAssyrian family, and therecertainly was no prohibitionagainst it.r8 The extentto which this potentialwas actuallyrealized of courseremains a matterof conjecture,but it would standto reasonthat elementaryliteracy was mandatoryat leastfor public and stateoffices, as later in Greeceand Rome."'I submit that the allesed..drastic" second-millennium chanse in " The letter contains a total of 40 diff'erent graphemes, of which 26 (: 65o/o)are phonetic and 14 ideographic. The phonetic signs (a e i u; ba ha ia ka la na ra ia, be ie, li ni si, mu pu ru su; ap et is ii) account for 70.5o/oof all sign occurrences.The ideoqrams include numbers, determinatives and logograms for common nouns, names.and name elements(e.nn ,,scribe,"ARAD,,servant," BE(-li),,lord," DI ,,well- being." EN\Alvl ,,governor,"LUGAL,,king"; 30. ai-iur, arrap, nN,lnl). Even though the sample is very limited, the total absence ol'CVC-type syllabograms on the one hand, and the presence of CV-VC type spellings (be-et, li-is, ra-ap) on the other, indicates that it is representative(cf. n. 8 above). Supposing that the writer's com- plete syllabary included all the commonly uscd Neo-Assyrian CV and VC graph- emes and a proportional number of additional logograms, it would have consisted of a total of 112 signs, 79 oI them syllabogramsand 33 ideograms.(In actual fact, the syllabary may well have been more reduced and still perfectly functional). This compareswell with such syllabic systems as Japanesehiragana and Linear B. To put the matter in perspective, it may be noted that cven the syllabary of such an expert scribe as Mar-Issar, attested in an cxtensive correspondence(SAA l0 347-370), doesnot include more than 225 graphemes( 170 syllabograms+ 55 ideograms). '*An anonymous letter to f]sarhaddon (ABL 1245) denouncesa goldsmith who ,,like the king and the crown prince has bou-shta Babylonian, settledhim in his own lrouse.and taughl (liginntt iqtibil his son in exorcism; they have even explained to him extispicy'omens. and he has even studiedgleanings from Enuma Anu Enlil, and this right befbre the king. my'lordl" lobr'..1-12).However, it is clear that the man was not accused fbr teaching his son to read and write but fbr acquiring knowledge in magic, extisprc,v and astrology' subjects potentially highly dangerous to the - monarchy (see SAA l0 179 and m1'remarks in Iraq 34 [972]32) without the king's permission. To ludge fiom a contemporary letter, learning to read and write cuneiform was not considered a ditflcult task in this period - not at least by the writer, the princess Serua-etirat,who degradesher sister-in-law by hinting she was involved in such childish exercises (ARL 308). A letter written by the above-men- tioned goldsmith's son (ABL 847) has actually been preservcd, and it displays very good writing competence. ''Inhisbook The Otd Assyrian City State and lts Colonies (Copenhagen,1976), p. 305, M.T. Larsen writes: ,,There are indications that a great many Assyrians knew how to read and write so the need for privately employed scribes may not have been so great. The system of writing was highly simplified with only a limited number oI syllabic signs and quite I-ew logograms, and many of the outrageously hideous pri- vate documents constitute clear proof of the amateurishnessof their writers. We know lbr certain that some of the sons of important merchants were taught the 322 S, Parpola Mesopotamianliteracy (see n. l6) actually never took place,and that the levelof literacyin first millenniumMesopotamiawas at leastas high (if not higher)as in earliertimes.2" APPENDIX: Additional PhilologicalNotes on K 652 2. The readingof the sender'sname is basedon the syllabicspellings tnai-iur-na-d-di (ABL 94I:9), mA.tona-a'-di (ADD 22:3), meN-na-a'-di (SAA 6 325r.23) andtnna-a'-di-otNcrn(CT 53 38 r. 3 andCTN 3 102ii 9). 3. The supplied/u is (with the exceptionof CTN 3 l:4) otherwisemiss- ing from the salutationonly in lettersfrom the Balih and Habur area(see SAA | 214:3,222:3,223:3and 224:3), where its omissionis surelyinten- tionaland due to Aramaicinfluence (see KAI no. 233:l). 5, a-nabt-et is elsewhercregularly spelled with tbeina sign.For a-na: c.) €bo G) .!J o0 & F l-{ j C) o i (-'! o +r l-( C) ca o o) t-.( F c) +-) (1.< 0.) o & a q) {-)(.) a t F h ()cn +-l l< j o O g Cg t-r b0 o {-ro A t+{ (l) a c.i L{ \n () \o ,o M v