You are on page 1of 4

Decipherment of the Earliest Tablets Author(s): Denise Schmandt-Besserat Reviewed work(s): Source: Science, New Series, Vol.

211, No. 4479 (Jan. 16, 1981), pp. 283-285 Published by: American Association for the Advancement of Science Stable URL: . Accessed: 31/01/2012 22:39
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . 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

American Association for the Advancement of Science is collaborating with JSTOR to digitize, preserve and extend access to Science.

crossarrn, is counterweighted, and has Decipherment of the Earliest Tablets proved sturdy and stable in field studies Abstract. The first signsof writing werecrudely impressed on clay tablets.These of corn (6). and standfor clay tokensusedfor recording priorto We field-tested three prototype in- signs arefound to represent Therecentdecoding of a seriesof tokensmakesit possibleto identify the strumentsduring the 1979 growing sea- writing. landmeasure, animal numeration, andotherecoson with satisfactory results. The tests signsas unitsof grainmetrology, were conducted on rangelands in west nomicunits. Texas and on alfalfa, corn, soybeans, and winter wheat at the Beltsville AgriJoranFribergrecentThe recent identification of a system of cles, andtriangles. cultural Research Center, Maryland. clay tokens used for recording/counting ly documentedthe fact that the SumeriThirty instruments were tested during as a progenitorof writinginvites the re- ans and the Elamites used the same systhe 1980 growing season at locations evaluation of the earliest written docu- tem to record grain metrology (3, pp. throughoutNorth America. Some of the ments, known as tablets (1, 2). This re- 10 and 20): (i) the most basic unit (about collected data appear in Fig. 3. port deals with the first series of tablets 6 litersof grain),called the banin Sumer, In addition to the demonstrateduses which are characterizedby crudely im- was representedby a wedge; (ii) a un'it with vegetation, hand-held radiometers pressed signs (Fig. 1). These tablets are six times larger, the bariga,was reprehave potential application in any field usually referred to as "numerical tab- sented by a circle; (iii) a fraction of the where radiometricmeasurementsin the lets," suggestingthat they yield only nu- banwas shown as a triangle.I postulate 0.3- to 2.5-,umregion are of value. The merical notations. They will be called thatthe shapes of the signs used for grain spectralrangeof the device describedin here "impressedtablets," and, in !ightof metrologyderive from the tokens in the this reportmay be reset by changingin- new evidence drawnfrom the token sys- shape of cones, spheres, and triangles. terterencefilters or detectors. tem, I will propose a new decipherment As a consequence, Tconsider th small COMPrON J. TUCKER of the earliest stages of writing. cone to representthe most basic unit of WILLIAM H. JONES, WILLIAM A. KLEY The tablets and the signs. About 200 grain. Its shape may be viewed as deGUNNAR J. SUNDSTROM impressed tablets are reported in the rivingfrom the representationof a deep NASAlGoddard Space Flight Center, literature, coming from excavations in bowl. The sphere will stand, accordingGreenbelt,Maryland20771 Iran (Susa, Chogha Mish, Godin Tepe, ly, for the second most basic unit of Tepe Sialk, and Tall-i Ghazir), in Iraq grain, of larger size. Its shape may be References and Notes (Uruk and Khafaje), and in Syria (Ha- suggestiveof a bag of grain. Largecones 1. L. F. Silva, R. Hoffer, J. Cipra,Proc. Seventh buba Kabira and Jebel Aruda). They and spheres represent still larger units Int. Symp. Remote Sensing Environ.(1971), p. date from about 3150 to 2900 B.C. of grain metrology, whereas the plain 1509;R. W. Leamer, V. J. Myers, L. F. Silva, Rev. Sci. Instrum.44, 611 (1973);L. D. Miller, Eighteen differentsigns can be identi- trianglestands for a fractionof the basic R. L. Pearson,C. J. Tucker,Photogramm. Eng. fied on the impressed tablets which, as unit. RemoteSensing 42, 569 (1976);E. J. Brack, R. W. Tinker,G. T. St. Amour,Can. Agric. Eng. illustratedin Fig. 2, I interpretas repreThe impressionof these tokens on the 19, 78 (1977);F. M. Zweibaum and E. W. Chappelle, SPIE J. 180, 242 (1979). sentations of tokens. I view, in particu- tablets had the same meaningas the to2. C. J. Tucker, Remote Sensing Environ. 6, 11 lar, the deep circular markings(Fig. 2, kens themselves, and, as a consequence, (1977). 3. Photogramm.Eng. RemoteSensing 44, columns 1 and 2) as standingfor spheres the two small circular impressions and 1369(;978). and the shallow circular markings(Fig. the three wedges on the tablet in Fig. 1 4. W. Collins,ibid, p. 43. 5. M. Methy, Oecol. Plant. 12, 395 (1977) B. F. 2, column 8) as standing for disks. I may be read "two 'large' and three Robinson, M. E. Bauer, V. P. de Witt L. F. Vanderbilt,SPIE J. 196, 8 (1979), R. D. Jack- equatethe short wedges (Fig. 2, columns 'small' measures of grain." Land meason, P. J. Pinter, S. B. Idso, R. J. Reginato, 9 and 10)with cones and the long wedges sures in Sumer were calculatedin terms Appl. Opt. 18, 3730(1979);J. K. Aase and F. H. Siddoway,Agron. J. 72, 149(1980);B. N. Hol- (Fig. 2, column 16) with cylinders. of the seed ratio necessary for sowing ben and C. O. Justice, Photogramm.Eng. ReThe recent decoding of the meaningof (4). It is, therefore, not surprising to find mote Sensing 46, 1191(1980). 6. D. S. Kimes, B. L. Markham, C. J. Tucker,J. these tokens (2) allows me to proposethe in Friberg that the cones and spheres E. McMurtrey,NASAlGoddardSpace Flight following deciphermentfor the tablet in were also used as units for land measureCent. Tech. Memo., in press. 7. G. S. Birthand G. R. McVey,Agron.J. 60, 640 Fig. 1: two ''large" measures and three ments (3, p. 46). In this case the multi(1968). ples of the standardSumerianunits, the 8. R. L. Pearson and L. D. Miller,Proc. Eighth ''small" measuresof grain(equivalentto Int. Symp. Remote Sensing Environ.(1972), p. about 90 liters of grain). An iku explanation (3258 m2) and the bur (63504 m2), 1357. 9. , C. J. Tucker,Appl.Opt. 16, 416( 1976). follows. were expressed by punched cones (Fig. 10. The device was sold by Exotech, Inc., GaithersThe meaning of the wedges, deep cir- 2, column 11) and spheres (Fig. 2, colburg,Md. 11. C. J. Tucker,unpublished data umn 4) and a fraction by an incised 12. B. N. Holben and C. J. Tucker,Photogramm. sphere (Fig. 2, column 3). Eng. Remote Sensing, in press. 13. C. J. Tucker,J. H. Elgin,J. E. McMurtrey, C. J. When, how, and why certain signs Fan, RemoteSensing Environ.8, 237 (1979);C. J. Tucker,J. H. Elgin, J. E. McMurtrey, Photospecificallyused for grainand field meagramm. Eng. Remote Sensing 45, 643 (1979) surementscame to be used for abstract Int. J. RemoteSensing 1, 69 (1980);C. J. Tucker, B. N. Holben,J. H. Elgin,J. E. McMurtrey, numbers are questions of fundamental Photogramm. Eng. Remote Sensing 46, 657 importance for the developmentof math(1980);B. N. Holben, C. J. Tucker, C. J. Fan, ibid., p. 651. ematics, but they are beyond the scope 14. C. J. Tucker, W. H. Jones W. A. Kley, G. J. of this report. In the texts of the Uruk Sundstrom, NASAlGoddariSpace Flight Cent. Tech. Memo. 80641 (1980). IVa period of 3100 B.C., pictographsin 15. Theextensionpole was designedby P. Leone, J. the form of an ear of barley and of a Humphreys,and K. Kirks of NASA Goddard Space FlightCenter. schematizedfield are added next to the 16. We thankD. Friedman,L. Miller,D. Smith, L. impressedsigns in accounts dealingwith Walter,J. Knudson,V. Salomonson, T. Ashton, and R. Gilbertfor thirassistanceand support. Fig. 1. Impressedtablet (Gd 73.392)fromGo- barley and land to specify that the quanSupported by the Department of Agriculture and din Tepe, Iran. [Courtesy T. Cuyler Young, NASA's Office of TechnologyUtilization. Jr., Royal OntarioMuseum, Toronto, Cana- tities referredto these commodities,thus 11 Jllly 1980 suggestingthat the process of acquisition da]
SCIENCE, VOL. 211, 16 JANUARY 1981
003s807S/81/011s0283$00.S0/0 Copyright 42)1981 AAAS 283

steiIl's translationof "slab, block, total circle" (5) (Fig. 2, column 8) becomes suspect, and the correct meaningof the plain disk may be providedby Friberg's study. Accordingto his results, a special numerationsystem was used in Elam to keep track of animals. The signs used were alternativelya wedge standingfor one animal(Fig. 2, column 16), a circular 1. Spheres shape for ten (Fig. 2, column 8), and a double wedge, apex to apex, for 100 tfig. 2, column 13) (3, p. 21). I interpret the wedge used in the animalnumeration system as being the long wedge, which represented the cylinder of the token system I also postulate that the circular markingused to representten animalsis Proposed Unit of grain Unit of grain Unit of Isnd Unit of land ? ? translation nieasurel measure nceasure measure the shallowcircularmarking which stood for a disk (Fig. 2, column 8). It is imporPictograph lilrd mill. tant to note here that the double cones ATU 897 placedapex to apex which appearon one 907 898 781 tabletfrom GodinTepe (6) and standfor Transiation. 10 3600 10 2heep 100 animals, represent an importantdeFalkenstein 1 2 3 4 5 Y parturefor writing. The sign is not the straightforward representationof a to11.Disks 111. Cones ken. The symbol is doubledand exploits for morphologythe new possibility proTokens vided b-ytwo-dimensionaldesign. The remainingsigns. The deep circular markings standing for the spheres and the short wedges standingfor cones are by far the most-frequently used signs (Fig. 1). They occur, respectively, on Proposed Unit of grain Uni1 of graln Unit of Iand 88 and 69 tablets. The readingof these Unit of iand tOO translation 1O animais measure measure m3asure measure animais two SigIlSalone provided, therefore, a Pictograph translationfor 85 percent of the tablets Illrd mil. D The long wedges standingfor cylinders axld the shallow circles standing for ATU 753 892 890 aO6 918 disks are next in frequency and appear Transiation Siab. tOttli after circie 1 6 6 Fraction on 30 and 15 tablets, respectively. The remaining signs are rare. An attempt 8 9 1O 11 12 1-3 has been made in Fig. 2 to correlate them to known Sumerian pictographs V. i3iconoids Vl.Ovoids Vll: CylAnders IX. Trlangles suggesting such translations as "oil," "sweet," and i'fat tail sheep." The decipherment of the impressed Tokens tablets has three major consequellces. First, it makes available a body of previously enigmatic data and reveals that the impressed tablets bear accounts of foodstuffsconsisting mainlyof grainand small stock. Second, it provides an insight into the first use of writing. The smallamountsof food dealt with suggest Proposed Unit ot graln translation ? ? 1 animal messurement ? that the impressed tablets, like the later pictographic tablets, were receipts of Pictograph Illrd mill. taxes paid by individuals.Third, they ilt3 lustratethe last stage of qualitycounting which xnathematicians have conjectured ATU 428 733 9 preceded the acquisition of abstract Translation numbers. after Good/sw eet Oli 6 Faikenstein DEN7ISE SCHMANI)T-BESSERAT 15 17 CenterAor MiddleEasternStudies, Fig. 2. Chartshowingthe relationship between tokens, impressedsigns, and.pictographic signs Universityof Texas, lsee (S) for an explanationof ATU numbers]. Austin 78712
Impressed 0 t e e e f

of abstractnumberswas then alreadyunder way. Themeaningof the long wedges, shallow circles, and double wedges. In the token system, a series of disks bearing distinctivepatternscan be matchedwith pictographs translated as sheep, ewe,

lamb, wool, cloth, and garment. This demonstrates that concepts of related meaningwere representedby variations of markingsplaced on a type of token. One may, therefore, expect that the plain disk was a common unit in the series, which acted as a root. Falken-










season has been maintaihedsince 1976 by calibrationstandards. The values in Table 1 are averages over mAny measurements and represent our best estimates of concentrationsin the antarctic troposphereduringJanuaryof each year. Whenever possible, the concentrations observed in Antarctica were compared with measurementsmade by P. Fraser, CommonwealthScientific and Industrial ResearchOrganization,Cape Grim,Tasmania ( 43S), and measurements made elsewhere in the Southern HemiAtmospheric Trace Gases in Antarctica sphere. The antarctic measurements AWbstract. Tracegases have been measured, by electrota-capture gas chromatogra- have always been consistent with these phy and gas chromatography-mass spectrometrytechniques,at the South Pole (SP) additionalmeasurements(2). in Antarcticaand in the U.S. Pacific Northwest (PNW) ( 45N) duringJanuRry of On the basis of the data given in Table each yearfrom 1975 to 1980. These measurementsshow that the concentrationsof 1, we have estimatedthe increases in the CCI3F,CCI2F2, and CH3CCI3 have increased exponentiallyat 3ubstRntial rates. The concentrationsof CC13F,CC12F2, CC14, concentrationof CCI3Fincreased at 12 percent per year at the SP and at 8 percent CH3CC13, and N20 and considered the per year in the PNW; CCI2F2 increased at about 9 percent per yeRr at both loca- changes in the north/south (N/S) ratio. tions, and CH3CCI3 increased at 17percent per year at the SP and 11.6 percent per The results of these calculations are year at the PNW site. Thereis some evidence that CCI4 ( 3 percent per year) and summarizedin Table 2. We found that N2O (0.1 to 0.5 percent per year) may also have increased. Concentrationsof tine the time series data are appropriately othertrace gases of importancein atmosphericchemistryare also being measuredat representedby the (exponentialgrowth) these two locations. Results of the measurementsof CHCIF2 (F-22), C2CI3F3 (F-113), expression SFs, C2-hydrocarbons, and CH3CI are reportedhere.
References and Notes (Ausgrabungender Deutschen Porschunggemeinschaftin Uruk-Warka 2, No. 111, Berlin, 1. D. Schmandt-Besserat, Sci. Am. 238, 50 (June 1936),p. 260. 1978). l>, 6. Gd 73.291, unpublished. 2. _ , Technol.Culture21 (III), 357 (1980). 7. This researchwas supported by a grantfromthe 3. J. Friberg,TheThirdMillennium Roots of BabyNational Endowment for the Humanities. I l(nian SMathematics: Methodfor the DecipherthankP. Amiet at the Musee du Louvre, Paris, ment, throughMathematicaland Metrological and T. CuylerYoung, Jr., at the Royal Ontario Analysis,of Proto-Sumerian and Proto-Elamite Museum,Toronto,for allowingme to study the Semi-Pictographic Inscriptions(ChalmersUrxitablets from Susa and Godin Tepe; M. Irwin, versity of Technology and th University of who edited the report; and E. Simmons, who C;oteborg, Goteborg,Sweden, 1978-79). drew Fig. 2. 4. W. W. Hallo, Bibl. Orient.33, 38 (1976). 5. A. Palkenstein, Archaische Texte aus Uruk 27 June 1980,revised 7 November 1980


January1980markedthe sixth year of tance in the years to come (1). We meaour trace gas measurementsprogramat sured these halocarbonsand other trace the South Pole (SP). We report here the gases, using establishedelectron-capture data obtained over these 5 years, esti- gas-chromatography techniques (3). mate the average annualincreases in the The internal consistency in the absoconcentrationsof CC13F (F-11), CCl2Fa lute concentrations reported for each (F-:l2),CC14, CH3CC13, and N20, and report the consistency of our measureme]ntsof CC13F,CCl2Fa,and CH3CC13 17 \ _ with the emissions estimates. In January F-1 1 1980 we also measured CHC1F2 (F-22), \SP C2(n13F3 (F-113), CH3C1, SF6, CH4, C0, 15 \ _ C2lI2(acetylene), C2H4(ethylene), and C2lI6(ethane). These data are also included. 13 We have maintainedthe objective of quantifyingthe global trend toward increasing atmospheric concentrations of long-livedgases such as CC13F, CC12F2, andl CH3CC13, which are releasedas a rePNW sult of humanactivities, primarilyat latitudes above 30N, and which may adversely affect the future global environ7 ment (1). In order that the picture be more complete, the trends observed in the antarcticatmosphere are contrasted to those observed at (remote) Pacific Northwest (PNW) ( 45N) sites during 12 the same period. The measurements of- N20, CC13F, CC12F2, CC1L4, and CH3CC13 made every 10 Janruary from 1975 to 1980 are reported in 1[ able 1 (2). Concentrations for a variety of trace gases, not previously mea8 sured in Antarctica, are also shown in 1975197619771978Table 1, includingCHC1F2 (F-22), which 1977 1978 1979 1980 is likely to gain environmental imporTime
1 1


U) S D


where c is the concentrationand ,8 is regardedas a constant, thus reflectingthe average rates of atnlospheric increase duringthe years when observationswere made. In Table 2, we estimated,8 by using nonparametric statistical methods based on the Theil statistic (4); the estimate of the (average) value of ,8 is denoted ,8. The ,8 obtained by this method is less sensitive to gross errors than the usual classical least-squares estimates. The numbers we obtained for,8 were multiplied by 100 percent to convert them to percentage increases per year. Similarly, we estimated a distributionfree approximate90 percent confidence interval for,S, using the Theil test (4). The results are reportediti Table 2 as pL (lower limit) and ,Bu(upper limit). The concentrations of CC13F,CC12F2,and CH3CC13 increasedat about 10, 9, and 13 percentper year, respectively, on the average. The data in Table 1, for 6 years, also indicate that the rate of atmosphericaccumulation of CC13Fand CC12F2 (denoted ,x3) may be slowing down. Such an observation would be consistent with, indeedrequiredby, the leveling offin the global emissions since about 1975(5). A leveling off in the emissions of these

Fig. 1. Rates of growth of CCl3F (F-l l ) and CCl2F2(F-12) during overlapping 3-year periods from 197S to 1980.

SCl:ENCE,VOL. 211, 16 JANUARY 1981

003s8075l81/011s0285$00.50/0 CopyrightK)1981AAAS