Cuneiform Digital Library Journal 2011:1 <http://www.cdli.ucla.edu/pubs/cdlj/2011/cdlj2011_001.html> © Cuneiform Digital Library Initiative ISSN 1540-8779 Version: 3 March 2011 The Earliest Contributions to the Decipherment of Sumerian and Akkadian Kevin J. Cathcart Campion Hall, Oxford Emeritus Professor, Dublin §1. Introduction cise details. Strange as it may seem, it is a popular book §1.1. Among Edward Hincks’ papers in the Griffi th on the discoveries at Tell Mardikh (ancient Ebla) by C. Institute, Ashmolean Museum, Oxford, there are three Bermant and M. Weitzman (1979) that presents a use- drafts of an undated lecture describing the fi rst steps ful, though imperfect, account to a wider readership. It taken in the decipherment of Old Persian (Cathcart is to their credit that the authors went to the trouble of 1983).1 The main aim of the lecture is to describe Georg examining the relevant A. H. Layard-Rawlinson corre- Friedrich Grotefend’s procedure in the decipherment spondence in the British Library. My own contribution because, Hincks says, “few persons seem to have correct has been to publish as much of Hincks’ correspondence ideas of the nature of his proceedings” (Cathcart 1983: as I could locate and to present the details of his exten- 30). Hincks pays due attention to Carsten Niebuhr’s sive publications, some of which are found in obscure copies of the inscriptions and his remarks on them, and and forgotten journals. he praises the progress made by the Danish philologist Rasmus Rask. Unfortunately, Hincks and Henry Cres- §1.3. I introduce this discussion of early contributions wicke Rawlinson never wrote full accounts of the steps to cuneiform decipherment with a passage from the in- they took in the decipherment of Akkadian, but, as we troduction to The Literature of Ancient Sumer, edited by shall see, much can be learned from Hincks’ publica- Jeremy Black and colleagues (2004: li-lii), which reads tions and correspondence (Cathcart and Donlon 1983; as follows: Cathcart 2007-2009). The Rosetta stone of cuneiform writing is the Bisutun, or Behistun, inscription, an enormous trilingual rock relief §1.2. Until recently, there were no satisfactory accounts carved inaccessibly high up a mountainside in modern of the decipherment of Akkadian and the cuneiform Iran near the border with Iraq. The fi rst serious attempt script in which it is written. Publications on decipher- to examine it was made in 1835 by a British diplomat, ment are generally very good on the deciphering of Old Henry Rawlinson who was in Persia as military adviser to Persian (Friedrich 1966; Gordon 1968; Pope 1999) but the governor of Kurdistan. He managed to climb the cliffs they are wholly inadequate for Mesopotamian cunei- several times in order to make a papier-mâché copy of the form (cf. Daniels 1994: 54 n. 1). There is an informed then unintelligible cuneiform. Within two weeks, and with the help of work that the German scholar Georg Friedrich account of the decipherment of Akkadian in R. W. Rog- Grotefend had done on inscriptions from the Persian capital ers, History of Babylonia and Assyria (1915), but inevi- Persepolis, Rawlinson was able to establish that one of the tably it is dated and we have had to wait for the recent three inscriptions used a 42-letter cuneiform alphabet and contributions of Peter T. Daniels (1994, 1996) and Mo- started with a description of the Persian king Darius which gens Trolle Larsen (1996, 1997) for fuller and more pre- was almost identical to that given by the ancient historian Herodotus in his Histories. Within three years, after a re- 1 The three manuscripts of Hincks’ lecture are numbered turn to the monument and using his knowledge of Middle 554, 555 and 556. The text published in Cathcart 1983, and Modern Persian, Rawlinson had deciphered 200 lines 30-41 is from MS 554. Cuneiform Digital Library Journal 2011:1 page 1 of 12 of the Old Persian inscription – an account of Darius’s rise ternational enterprise. to power – and presented his work to the Royal Asiatic Soci- ety in London and the Société Asiatique in Paris…In 1844 §1.6. The work done on Elamite by Nils Ludwig Rawlinson and three colleagues again climbed the cliffs at Westergaard (1844, 1845) and Hincks (1846) was not Bisutun, now making a complete papier-mâché mould. Us- based on Rawlinson’s paper-casts from Behistun, but on ing this copy, and working on the assumption that all three texts which Westergaard himself had copied at Perse- inscriptions told essentially the same story, Nils Westergaard polis and Naksh-i-Rustam. It was only in 1853 that and Edwin Norris managed to decipher the second. Its script Norris, using Rawlinson’s casts and notes, published the used 131 characters and the language, Elamite, turned out Elamite text from Behistun, though like Rawlinson he to be an isolate, related to no other known. But the third version of the inscription, which was by far the most com- called it Scythic. Norris graciously acknowledged that plex, remained a mystery…Its decipherment became an in- about half of the characters in his list had already been ternational enterprise to which many scholars contributed. deciphered by Westergaard or Hincks (Norris 1853: 47). A more thorough study of Elamite was published §1.4. This account is not accurate in a number of ways by A. H. Sayce in 1874. and my remarks here are simply a way of showing that accuracy is of the essence in accounts of decipherment. §2. Rawlinson and Behistun The inscriptions which Rawlinson copied in 1835 §2.1. The main Babylonian text at Behistun was copied were the trilingual inscriptions at Mount Elwand near by Rawlinson in 1847 but it was not published till the Hamadan.2 At this time Rawlinson visited Behistun but end of 1851. Therefore, contrary to widespread belief, he did not copy the Old Persian inscriptions there until it played no role in the crucial fi rst phase of the deci- 1836 and 1837. So it was on the basis of the Mount El- pherment of Akkadian (Daniels 1994: 50). To describe wand texts that he took his fi rst steps, inspired of course the Behistun inscription as the “Rosetta Stone” of cune- by Grotefend’s breakthrough of 1802. His translation iform decipherment is misleading when we are dealing and transcription of the fi rst two paragraphs of the Old with the decipherment of Akkadian. Indeed, Rawlinson Persian text in the Behistun inscription were commu- recognized that if the term “Rosetta Stone” is used, it nicated to the Royal Asiatic Society in January 1838 should be applied more broadly. He puts it this way: “As and he made further progress after studying Burnouf’s the Greek translation, then, on the Rosetta Stone fi rst Commentaire sur le Yaçna (1833), a pioneering work on led the way to the decipherment of the hieroglyphic Avestan. writing of Egypt, so have the Persian texts of the tri- lingual cuneiform tablets served as a stepping stone to §1.5. Rawlinson had also seen the contribution of the the intelligence of the Assyrian and Babylonian inscrip- Danish philologist Rasmus Rask, who, in 1826, cleverly tions” (Rawlinson 1850: 403). He knew that Grotefend gave the values m and n to two signs and as a result of and Hincks had used the Persepolis and Naksh-i-Rus- this he was able to recognize the genitive ending –anam tam inscriptions and he quickly realized that the Behis- in the phrase “king of kings.” This was the termination tun inscription was not the key that he had hoped for. of the genitive plural in Sanskrit and the conclusion He writes (1850: 408): that the language (Old Persian) was related to Sanskrit If the Behistun inscription had been recovered in as perfect was an important breakthrough. Rawlinson was also in a state as the less celebrated record at Naksh-i-Rustam, all communication with Norwegian-born Christian Las- the essential diffi culties of decipherment would have been sen, Professor of Indian Languages and Literatures at at once overcome…Unfortunately, however, the left half, Bonn. He had to re-write more than once the report he or perhaps a larger portion even, of the tablet is entirely sent to London in 1839 (Daniels 2009). In September destroyed, and we have thus the mere endings of the lines 1844 he recopied the Old Persian inscription at Behis- throughout the entire length of the inscription; the fragments tun and made paper-casts of the Elamite text and some which in several of the most interesting passages are alone legible, being not only insuffi cient to resolve diffi culties, but of the Babylonian. He prepared his translation of the sometimes actually affording of themselves fresh enigmas for Old Persian in 1845 and by this time he had received solution. Lassen’s latest work on the inscriptions (1844). So it was really the decipherment of Old Persian that was an in- §2.2. So Cyrus Gordon (1968: 60) is exaggerating when he speaks of “the squeeze of the Babylonian version that was destined to open up the whole fi eld of Assyriology.” 2 See Rawlinson’s “Memoir” (1846: 1-18) and the recent, reliable and detailed presentation of Rawlinson’s work These preliminary remarks should be suffi cient to show on the Old Persian inscriptions by Daniels (2009). page 2 of 12 Cuneiform Digital Library Journal 2011:1 that precision in detail is essential for any satisfactory Writing” (1846: 131) Hincks announced that he had description of decipherment. made some progress in reading the inscriptions in the “third Persepolitan” or Babylonian writing and that he §2.3. When Rawlinson published the fi rst part of his was able to show that the Assyrio-Babylonian language, impressive work, The Persian Cuneiform Inscription at as he called it, appeared to have much in common with Behistun, Decyphered and Translated, with a Memoir in the Semitic languages. This important statement reads 1846, the semi-syllabic nature of the script was not yet as follows: known to him. However, in August 1846 he sent a note I have applied myself to the third Persepolitan writing, from Baghdad announcing that he had discovered it. which agrees in character, and, to a great extent at least, Some months later he learned that this discovery had in language, with the Babylonian inscriptions, and to the already been made by Hincks (1846), who announced Assyrian writing in Schulz’s inscriptions. Having as yet a it in his paper “On the First and Second Kinds of Perse- very scanty supply of data, I have not been able to prepare politan Writing,” which was read at a meeting of the alphabets of either of these modes of writing. I have, how- Royal Irish Academy in Dublin on 9 June 1846 and ever, ascertained that they both agree in principle with the published in the Academy’s Transactions.3 Hincks had second Persepolitan. In both, some of the characters represent already published articles on ancient Egyptian (Cath- elementary sounds and some combinations. In both, two or more characters are used to represent the same sounds. In cart and Donlon 1983: 330-332) but there is nothing both, no vowel is omitted, but vowels and consonants are in his correspondence before 1846 to suggest that he repeated in two consecutive characters. The number of el- was studying cuneiform inscriptions. However, it is not ementary characters is greater in both these modes of writing surprising to fi nd that he was trying to decipher cu- than in the second Persepolitan. In the latter, a single vowel neiform, because he would have relished the challenge. was rarely expressed after a syllable terminating with the Sayce (1882: 378) said many years later that no problem same vowel, but this was commonly done in the Babylonian in decipherment ever seemed to baffl e him. Rawlinson and Assyrian, in which, of course, the simple vowels were of must have been surprised by Hincks’ long favorable re- much more frequent occurrence. In the second Persepolitan, view (1847d) of his Behistun volume. It is clear from m was written by w, but in the Babylonian by b, which this review, wonderfully entitled “Some Passages of the accounts for the same name being written Berodach in the Life of King Darius, the son of Hystaspes, by Himself,” Second Book of Kings, and Merodach in Isaiah. I have that Hincks already knew Old Persian well. He praised found the name of Babylon in the inscription on a piece of baked clay, shaped like a barrel, brought from the ruins, Rawlinson’s work very highly, especially his translation and in those on a few of the bricks. I have also found the and interpretation, but he thought his contribution to name of Nineveh on the bricks brought from that place. deciphering the Old Persian language was small. It is Both the Assyrian and Babylonian languages appear to have generally accepted that Rawlinson made a remarkable much in common with the Semitic languages; but some of contribution to our understanding of the language and their roots are common to them with the language of the he has rightly received high praise for his edition and second Persepolitan inscriptions, with which also they have translation of the Old Persian inscription at Behistun. many characters in common. I have found it to be a general rule, though it admits of some exceptions, that where a char- §3. Mesopotamian Cuneiform acter occurred in two or more alphabets, it had the same §3.1. It is time to turn to what is sometimes called Mes- value, or nearly so, in all of them. Thus, the pa of the second opotamian Cuneiform, the writing system used for Ak- Persepolitan is pa in Assyrian, and ba in Babylonian; and kadian, Elamite and Urartian as distinct from the cunei- so in other instances. The fi rst Persepolitan alphabet, on the contrary, had nothing in common with any of the others. form writing system used for Old Persian. Old Persian has 43 signs (36 phonetic signs and 7 logograms). The §3.2. Although this statement, written at the beginning fi rst decipherers quickly noticed that the cuneiform of June 1846, was not published until December 1846, script used to write Elamite (Median or Scythic as they there is similar information in the letters which Hincks called it) had more than 100 characters and the one to wrote to learned societies and to the editor of the Liter- write Babylonian had many more. In a postscript to his ary Gazette in June and July (Cathcart 2007: 137-148). article “On the First and Second Kinds of Persepolitan Letters reporting his progress in decipherment were read out at the meeting of the Royal Society of Literature on 11 June 1846. In one of them Hincks wrote: “As to my 3 Hincks (1846: 20) tells us that the body of his paper Babylonian and Assyrian deciphering, I am not aware was written at the beginning of May 1846, a postscript added at the beginning of June, and the notes added at that any thing in a right direction has yet been done the end of August. Cuneiform Digital Library Journal 2011:1 page 3 of 12 by others. I feel confi dent of having mastered the great garded the writing system as partly consonantal and diffi culty of making a commencement in each of these.”4 partly syllabic and he had already discovered that some characters had more than one value. He was confi dent §3.3. In the beginning, the values of the Babylonian and that he was on the right track and would make great Assyrian signs could only be worked out by comparing progress. Rawlinson, however, told Renouard: “That we the way proper names were written but Hincks quickly shall ever obtain the same insight into the Babylonian recognised how the signs were used. Consonant-vowel language that we now have of the Achaemenian Persian, and vowel-consonant signs could be combined to rep- I doubt extremely” (Cathcart 2007: 159). He wrote in resent a consonant-vowel-consonant syllable. Hincks a similar vein in February 1847: “The day however is went a step further by reading “Nebuchadnezzar, king still, I think, far distant in which we shall be able to of Babylon” on various bricks and he also found this read and understand independent Babylonian and As- name in the East India House inscription. In a letter syrian inscriptions, for we want the grand desideratum to the Literary Gazette he showed that he had not been of language: and unless it lurk in the old Egyptian or merely conjecturing, as his predecessors had done, but Aethiopic, I am sure I know not where to look for it” had proceeded on the surest grounds (Cathcart 2007: (Cathcart 2007: 190). 139). Much more important, however, was his an- nouncement in the Literary Gazette for 25 July 1846 §4. Rawlinson and Hincks that a fragment of the Ker Porter inscription was a tran- §4.1. Rawlinson claimed that many of Hincks’ 76 val- script in “cursive” characters of part of the great inscrip- ues drawn up in December 1846 and published in 1847 tion of the East India company (Cathcart 2007: 146). were incorrect (Hincks 1847b: 245). “I am much re- This discovery is published in some detail in Hincks’ lieved for he has not above a dozen correct identifi ca- paper “On the Three Kinds of Persepolitan Writing, tions,” he wrote to Layard (Larsen 1996: 181). Hincks and on the Babylonian Lapidary Characters” (1847b). was soon aware that some of his values were either Hincks established the equivalences between what he wrong or only partly correct and he submitted another called the “cursive” and the “lapidary” signs (Daniels paper to the Royal Irish Academy a month later, that is, 1994: 38). By “cursive” he means the signs he has found in January 1847. This paper contained a revised list of on clay tablets and by “lapidary” he means those found values (Hincks 1847c: 252). It also included the fi rst- on bricks and on the great East India House inscription. ever table of cuneiform numerals. I have recently ac- Concerning this discovery, Rawlinson wrote to George quired an unpublished letter by Hincks, dated 1st Janu- Cecil Renouard5 on 27 October 1846: “I am indebted ary 1847 and probably addressed to Humphrey Lloyd, to him [Hincks] indeed for a most notable discovery, the President of the Royal Irish Academy. It deals with one in fact which has proved of more use to me even the plate being prepared by the engraver to accompany than my Behistún key” (Cathcart 2007: 158). Hincks the publication of his December 1846 paper (Hincks provided a plate with these signs, showing the equiva- 1847b) and explains the nature of his new paper with lences he had worked out. He and other scholars now “rectifi cations” or revised list of values, which was going had access to a much larger corpus of texts. to be read at the Academy on the 11th January 1847 (Hincks 1847c). I publish the letter here. §3.4. With regard to the language of the texts, we saw in Dear Sir, the extract above that Hincks stated without hesitation that it was Semitic. Since then he had identifi ed and I fear you will think me beyond measure troublesome in reference to this Babylonian alphabet. I fi nd I have made a published the fi rst person pronoun anæku and therefore considerable number of mistakes in that which is in the en- had read the fi rst Akkadian word that was not a proper graver’s hands & which is described in my last paper. I now name (Hincks 1847b: 247). In these early days he re- send a short paper, stating the nature of the rectifi cations which I have made (which remove many diffi culties) & will if possible send the alphabet belonging to it tomorrow; but, as this post is sometimes late in Newry for the Belfast 4 Short extracts of the letters can be found in the Literary mail, my letter might not arrive in time for the Council on Gazette, no. 1535 (20 June 1846), 561-562. The extract Monday; so I think it best to send the paper itself today. given here is on p. 562. You will observe that I do not mean to cancel the former 5 George Cecil Renouard (1780-1867), Church of plate, which is valuable not only as a fi rst attempt but as England clergyman and scholar, was one of Hincks’ giving with (I believe) perfect correctness the correspondence most frequent correspondents. He was Professor of between 76 Third Persepolitan characters & the Babylo- Arabic at Cambridge for some years. page 4 of 12 Cuneiform Digital Library Journal 2011:1 nian lapidary ones. I do not give any lapidary characters in murdered by a local chieftain in 1829, but his copies of the new plate; but increase the number of cursive ones – & the inscriptions were retrieved later and published in add the numerals & combinations of them expressing large Journal Asiatique (Schulz 1840). This is the publication numbers. Part of the last paper – viz. the list of Babylonian that Hincks worked from. Thirty-nine of the inscrip- characters & the transcriptions will be suppressed; & the tions are in Urartian and three are trilingual like those at new & much improved list with a few transcriptions will Persepolis. Hincks called the language Assyrian, as dis- be sent to conclude the present paper. This, however, I shall tinct from Assyrio-Babylonian and Babylonian. There not be able to accomplish till next week; but will have it in is a striking passage about the inscriptions in Hincks’ Dublin in time for the 11th. letter of the 23 August 1847, which reads as follows I hope you received the copy of my papers which I sent to (Cathcart 2007: 206): Mr. Clibborn for you. The characters in the Van inscriptions are clearly derived With many thanks for your kindness, believe me to remain from the Babylonian ones; but I have ascertained that in the mode of using them there is a very material differ- Yours very faithfully ence. While the generality of the Babylonian characters are Edw. Hincks used, like the Hebrew and other Semitic ones, to express consonants in which no particular vowels inhere, the Van §4.2. Hincks elaborates on the contents of this letter in characters present a complete syllabary, its vowels being all the introductory pages of the published article with the expressed either by separate characters or by syllabic signs, in which they inhere. The mode of reading the language of rectifi cations (1847c: 249-250). He did not expect his these inscriptions is consequently much more defi nite than previous “alphabet” to prove “perfectly correct." He had in the case of any other species of cuneatic writing, with the noticed, for example, that the number of dentals was exception of the fi rst Persepolitan; and, what surprised me too small. Larsen (1996: 181-182) has made an evalua- not a little, the language of the inscriptions agrees with that tion of the values in Hincks’ fi rst list of December 1846 of the last-named inscriptions in being Indo-Persianic. Its and concludes that “twenty-three of the values were en- resemblance to the Sanskrit is, indeed, in some respects, clos- tirely correct, and a number of others were ‘nearly’ cor- er than that of the ancient Persian; though it is curious that rect, in the sense that Hincks had the right consonant in some of its grammatical forms are more akin to the Greek. a syllable but was wrong about the vowel. He gave the To all who are engaged in philological and ethnological value ra for a sign which should be read ri, nu for ni and pursuits it must be of the highest interest, as the oldest mem- so forth.” Rawlinson’s evaluation of Hincks’ readings is ber – of the eastern branch at least – of this widely-diffused so wide of the mark that Larsen, quite rightly, questions family of languages. the reliability of the basis for Rawlinson’s readings. §4.5. On the 19 October 1847, Norris wrote (Cath- §4.3. An indication of Hincks’ astonishing energy and cart 2007: 213): “I have read your announcement of linguistic genius can be highlighted by the observation the Sanskritish character of the Van language, with very that in addition to all his work in cuneiform decipher- much surprise, as my own researches with that idea ment during 1846, he wrote a hundred-page article on were quite fruitless; I wait for your discoveries on that “An Attempt to Ascertain the Number, Names, and head with great impatience; I have had no leisure to Powers, of the Letters of the Hieroglyphic, or Ancient go over the fi eld myself, since I heard your announce- Egyptian Alphabet; Grounded on the Establishment ment.” Hincks’ conviction that he was dealing with an of a New Principle in the Use of Phonetic Characters” Indo-European language was wrong, but he correctly (Hincks 1847a). Anyone interested in the present order recognized that the language was not Semitic. William of the ancient Egyptian alphabet in the Egyptian gram- Henry Sykes, an army offi cer in the East India Com- mars we use today and the relationship between the Se- pany, told Hincks in October 1847 that he had seen mitic languages and ancient Egyptian should read this a letter in which Rawlinson expressed an opinion that article (see Ray 1994: 59-61). “the Van inscriptions were of Semitic origin, by a people who came originally from Egypt via Phoenicia” (Cath- §4.4. In letters to the Literary Gazette dated the 21 June cart 2007: 215). and the 23 August 1847 (Cathcart 2007: 195-196, 204-206), Hincks announced the progress he had made §4.6. In December 1847 a paper by Hincks “On the in reading the Van inscriptions, which were found by Inscriptions at Van” was read at a meeting of the Royal Friedrich Edward Schulz in the 1820s near Lake Van Asiatic Society and a supplementary memoir was read in Armenia (modern south-eastern Turkey). Schulz was in the following March. Publication followed quickly in Cuneiform Digital Library Journal 2011:1 page 5 of 12 1848.6 In Daniels’ view (1994: 39-42), although this ar- why signs had more than one phonetic reading and had ticle deals with Urartian and not with Akkadian, it was logographic as well as phonetic readings. In a footnote the most important of Hincks’ papers. Hincks seems to he makes the following invaluable and far-reaching ob- have promised another paper on the Van inscriptions servation (Hincks 1850: 19 n.*): in November 1856, because on 9 January 1857 Norris It is possible, too, that the word from which the phonetic wrote (Cathcart 2009: 3): value is derived may be one belonging to a different lan- I have been for some time impatiently looking for your guage. I will, in a subsequent section, produce an instance, promised paper in continuation of your labours on the Van in which I believe that the ordinary phonetic character, monuments. I hope you will let us have it soon. Our next namely pa, the value of , was adopted from a foreign meeting is on the 17th when I should like much to read it language. This mode of proceeding may be illustrated by to the Society. I hope you have not given up the intention of some of the abbreviations used in Irish manuscripts. The writing on the subject, for if you do not do any thing with Irish letters were at fi rst chiefl y used in the copying of Latin these Van Inscriptions, nobody will, for half a century, I am texts. In Latin manuscripts, the letter s with a peculiar sure. mark, which may be represented by s', was used to express the word sed. In the course of time the same mark was used in Irish manuscripts to express acht, the Irish equivalent of §4.7. Among Hincks’ papers in the Griffi th Institute at sed. And by a further progress, it was used to express this Oxford, there are three drafts of a paper on the Van sound, when it no longer signifi ed “but”, but was a portion inscriptions (Hincks MS 571). Two of them were writ- of a word of totally different meaning. Thus, ts' was used ten in 1856-1857 and the third was written in 1861, for teacht, “to come”. See O’Donovan’s Irish Grammar, p. but Hincks did not publish anything more on the Van 430.8 inscriptions. Norris’ worry that nothing would be done with the Urartian inscriptions for a long time was jus- §5.2. The decipherment of the “different” and “foreign” tifi ed and a quarter of a century passed before Sayce language, which would eventually be identifi ed as Su- (1882) published “The Cuneiform Inscriptions at Van, merian, would become another challenge for Hincks Deciphered and Translated.” during the following years. §5. Hincks’ Syllabaries §5.3. We know from Norris’ letter to Hincks dated 18 §5.1. The insights that Hincks obtained from his study March 1850 (Cathcart 2008: 14) that Rawlinson was of the Van inscriptions helped him to arrive more rapid- studying the paper “On the Khorsabad Inscriptions.” ly at a clear understanding of the system used in writing Hincks’ fi ndings seem to have upset Rawlinson, who Mesopotamian cuneiform. By the end of 1847 Hincks told Layard that the paper was “almost as wild and un- realized that full syllabic writing was also found in As- intelligible” as his previous contributions; anything that syrian and Babylonian texts. Every character that was was correct was due to a “few fortunate hits” (Larsen not a logogram or determinative stood for a full syllable, 1996: 225). But it is likely that Rawlinson was deeply never for one consonant alone.7 By 1849 Hincks could worried by Hincks’ insights, which were very clearly demonstrate that the signs had more than one reading. in confl ict with the views he had expressed in his lec- In his paper “On the Khorsabad Inscriptions” (1850), tures to large audiences at meetings of the Royal Asiatic one of his most important publications, he explains Society on the 19 January and the 16 February. These the nature of logograms (he calls them “ideograms”), lectures went to press in late February before Rawlin- including composite logograms. He has also worked son had seen Hincks’ Khorsabad paper. In one of the out the principle of homophony, explaining to his read- lectures he said (1850: 404-405): ers that in some cases two or more outwardly distinct signs expressed the same syllable. Hincks had come to I will now offer a few remarks on the nature and structure of these alphabets. That the employment of the cuneiform grips with the writing system but he wanted to know character originated in Assyria, while the system of writing to which it was adapted was borrowed from Egypt, will 6 Franz Bopp, Professor of Sanskrit and Comparative hardly admit of question…the whole structure of the As- Literature at Berlin, was particularly pleased to receive syrian graphic system evidently betrays an Egyptian origin. Hincks’ Van paper. See his letter of 22 August 1848 in Cathcart 2007: 253-256. 7 Hincks clarifi ed the date of this realization ten years later 8 Hincks took the example from J. O’Donovan, A in a paper read at the meeting of the British Association Grammar of the Irish Language (Dublin, 1845), p. 430. for the Advancement of Science held at Dublin in 1857 O’Donovan is referring to the use of a superscript hori- (1858: 135). zontal stroke above the letter. page 6 of 12 Cuneiform Digital Library Journal 2011:1 The alphabet is partly ideographic and partly phonetic, and §5.6. At a meeting of the British Association for the the phonetic signs are in some cases syllabic, and in others Advancement of Science at Edinburgh in July 1850, literal. Where a sign represents a syllable, I conjecture that Hincks delivered a lecture “On the Language and Mode the syllable in question may have been the specifi c name of of Writing of the Ancient Assyrians.” Unfortunately, the object which the sign was supposed to depict; whilst in only the handout and a short summary were published cases where a single alphabetical power appertains to the (Hincks 1851). In the summary it is stated that Hincks sign, it would seem as if that power had been the dominant “maintained, in opposition to all other writers, that the sound in the name of the object. In this way, at any rate, characters had all defi nite syllabic values; …though the are we alone, I think, able to account for the anomalous language of the Assyrians was Semitic, their mode of condition of many of the Assyrian signs, which sometimes represent phonetically a complete syllable, and sometimes writing was not.” Hincks was re-affi rming to his audi- one only of the sounds of which the syllable is composed. It ence, which included Rawlinson, the views which he cannot certainly be maintained that the phonetic portion had laid out in his previous papers. of the alphabet is altogether syllabic, or, that every phonetic sign represents a complete and uniform articulation. §5.7. In a paper presented to the Royal Irish Academy in May 1852, “On the Assyrio-Babylonian Phonetic §5.4. As early as May 1846 Hincks had already pointed Characters,” a paper in which he listed hundreds of out that the syllable tash (in Elamite), for example, could signs with their values when it was published later that be written ta-ash and not ta-sh (Hincks 1846: 125-126; year, Hincks (1852: 295) presented a very clear and sat- see Daniels 1994: 36). It is astonishing that Rawlinson isfactory assessment: (1851: 4) insisted on maintaining the following view in It has been assumed by all other investigators, that the mode the memoir that accompanied his Babylonian text from of writing used in the Assyrio-Babylonian inscriptions was Behistun: contrived with a view to represent the words of the language In the articulation, kat, for example, which commences the of those inscriptions. This language is unquestionably of name Katpatuka (for Cappadocia), and which is composed the family commonly called the Semitic; and it is therefore of the two characters ka, and at, either one or taken for granted that the characters used in the inscriptions the other of these signs must represent a simple letter rather represent Semitic letters. I can have no doubt whatever that than a syllable; and as this peculiarity of expression pervades this is a mistake; and moreover, that it is one of so serious a the whole Assyrian alphabet, I think I am justifi ed in still nature as to render it impossible for those who labour under adhering to the statement which I announced last year, that it to attain any accurate knowledge of the grammar of the the phonetic signs were in some cases syllabic, and in others language. I am myself fully satisfi ed, and I hope in the pres- ent paper to satisfy all who will take the trouble to follow literal. my arguments, that the characters all represent syllables, and that they were originally intended to represent a non- §5.5. In Rawlinson’s edition of the Babylonian text, in- Semitic language. Instead of the vowels being unrepresented, stances abound of cuneiform characters transcribed by or only represented by points, as in all Semitic writing that single letters. The fi rst person singular pronoun anæku, was fi rst applied to a Semitic language, we have in the identifi ed by Hincks in 1846 (1847b: 247), is written cuneatic inscriptions every vowel defi nitely expressed. The anak by Rawlinson, who compares Egyptian anok. Yet Semitic language appears in a disguise similar to what the in the accompanying edition he has writings anaku, Maltese does in Roman letters, or the Punic in the well- ank, anak. What is one to make of Rawlinson’s warning known passage of Plautus. (1850: 420) that the system of the Assyrian writing is Again it has been taken for granted, that the only method “in the last degree obscure, and the language in which of ascertaining the value of the characters is the analysis of the writing is expressed, unintelligible, except through known proper names. It appears to me, however, that, the the imperfect key of the Behistun translations, and the characters representing what I have just stated that they do, faint analogies of other Semitic tongues”? He speaks of this method can only lead to approximate, as distinguished “the Assyrian alphabet, with its many imperfections, its from accurate, knowledge. The way by which I have sought inconvenient laxity, and its cumbrous array of homo- to obtain accurate knowledge is by analysing verbs and phones” (1850: 407) but a year later he says that the nouns, especially such as have three radicals, of which none laxity has “either disappeared under a more rigid exami- is liable to be omitted or altered. I assume two principles: nation, or has yielded to the solution of one character fi rst, that the characters which occur in different infl exions being qualifi ed to represent several dissimilar sounds” of the same root, if they be not the same, must contain the same consonant differently combined with a vowel; secondly, (1851: 3). that characters which occur in like forms of different roots, contain the same vowel in the same position, differing only Cuneiform Digital Library Journal 2011:1 page 7 of 12 in the consonant. The former principle shows which char- §6.3. We have seen that as early as 1849 Hincks had acters express different functions of the same consonant; the drawn the conclusion that the language of the inventors latter shows which are like functions of different consonants. of cuneiform writing was not Akkadian. During the fol- lowing years he had to remind scholars that he had long §6. Origins of Babylonian Cuneiform held it as an established fact, that the Assyrians derived §6.1. The years 1852 and 1853 were signifi cant for many of the values of their characters, as well as their sys- Hincks. Layard was preparing his Discoveries in the tem of writing, from a people who spoke a non-Semitic Ruins of Nineveh and Babylon (1853) and decided that language. In 1856 he recalled that at the meeting of Hincks was the best person to help him with the in- the British Association at Edinburgh in 1850, when he scriptional materials which he had brought back from announced that the Assyrian and Babylonian characters the Near East. He visited Hincks in Ireland for nearly represented syllables, and not letters, as Rawlinson had three weeks in September-October 1852, showing him previously supposed, he accompanied the announce- his inscriptions and drawings. Highly impressed by ment with what he considered “a necessary consequence Hincks’ knowledge and energy, Layard convinced the of this newly discovered fact, that this mode of writing Trustees of the British Museum to employ him and must have originated with a non-Semitic people, as no Hincks signed a contract for a year from 1 May 1853. Semitic people could have invented a system of writing The brilliant decipherer was naturally very pleased to so uncongenial to their language” (Hincks 1856a: 132 = have access to inscriptions. Among the signifi cant dis- Cathcart 2008: 297).9 He pointed out that on that oc- coveries that he made was a text which he identifi ed casion Rawlinson, who was present at Hincks’ lecture, as a syllabary. In a long footnote in his article “On the represented this to the audience as a ridiculous opinion. Assyrio-Babylonian Phonetic Characters” (1852: 342) As we saw earlier, Rawlinson himself believed that the Hincks says that he had examined a fragment of a text writing system “evidently betrays an Egyptian origin” which had the accession number K(uyunjik) 62. He (1850: 404). writes: Among some inscriptions from pieces of terra cotta in the §6.4. Hincks applied himself with zest to the study of British Museum, which Mr Layard recently showed me, this non-Semitic language attested in bilingual tablets. was one which I recognized as an Assyrian syllabarium. By 1856 he was able to send Hermann Brockhaus, edi- Unfortunately it is but a fragment; but enough remains to tor of Zeitschrift der deutschen morgenländischen Gesell- show its nature. It contains parts of four columns, each of schaft, an article containing nine bilingual texts (Hincks which is divided by ruled lines into three series. That in 1856b). He even provided a German translation! the middle contains the characters to be valued; that on the All but one of these texts have been edited by Benno left contains the values; and that on the right contains the Landsberger (1937). The entry for 11 January 1856 in plural form, or the value which the character would have Hincks’ diary reads: “All morning at Magyar and Mon- if the plural signs were added. This syllabary, which will golian, trying to understand the general principles of probably be speedily published by the authorities at the Mu- this class of language, with a view to compare them with seum, establishes a number of points on which doubts may that of the bilingual tablets.” This was typical of him. yet linger in some minds. First, it proves that the characters are syllabic; secondly, that many values belong to the same Only when he had established the facts for himself was character. he going to disagree with Rawlinson’s views that Sume- rian was related to Mongolian and Mandshu. In his let- §6.2. Daniels (1994: 48; 1996: 147), the fi rst to high- ter to Brockhaus (the English original is in the library at light the importance of this discovery, explains its sig- Göttingen; see Cathcart 2008: 300-301), he expresses nifi cance for the progress of decipherment. This frag- mentary text is the one known to modern scholars as Syllabary A (Hallock 1955). Although Hincks thought 9 It is important to mention here the letter by Hincks which was published in February 1856 with the title it would be published quickly, the tablet was not pub- “Are There Any Assyrian Syllabaries?”: A Letter to the lished by Rawlinson and Norris until 1866, the year in Editor, Monthly Review 1, 130-132. It was a reply to which Hincks died. In a letter to the Literary Gazette on claims made in “Colonel Rawlinson’s Researches,” 24 July 1854 (Cathcart 2008: 251), Hincks says that Monthly Review 1 (January 1856), 44-47. This article he has identifi ed other tablets of a similar kind, includ- was probably written by Rawlinson’s friend William Sandys Wright Vaux, who worked in the British ing K 110, today known as Syllabary B (Landsberger Museum. The periodical Monthly Review ceased publi- 1955). The identifi cation of these tablets and the study cation after two years. Hincks’ letter is republished in of the texts extended the reading of signs enormously. Cathcart 2008: 295-297. page 8 of 12 Cuneiform Digital Library Journal 2011:1 the view that in some respects Sumerian (he and Rawl- fi ne scholars like Renouf had real diffi culty. S. A. Pallis inson called it “Akkadian”!) resembled the Turanian lan- (1956: 155) thinks that Rawlinson was the discoverer guages. By Turanian he means languages like Turkic, the of Akkadian polyphony, but the example he gives, “the association with Sumerian being made on the under- vowel sound a also has the ideographic value ‘son’” (see standing that the languages were agglutinative. In the Rawlinson 1850: 405 n. 2), is already found in Hincks’ nineteenth century, Friedrich Max Müller became the paper “On the Khorsabad Inscriptions” (1850: 20). In celebrated champion of a vast language family known as September 1863 Norris, having read Hincks’ paper on Turanian but the term came to be used more specifi cally polyphony where there was a reference to his research for the language group that today is called Uralo-Altaic. on the Akkadian verb, pleaded with Hincks to publish a paper on the verbal system in the Journal of the Royal §6.5. If Hincks had lived ten years longer, he would Asiatic Society (Cathcart 2009: 203-204). In the fi nal have been quite shocked by the bitter and often nasty year of his life Hincks agreed to publish, with fi nancial hostilities that broke out between Joseph Halévy and help from Norris, “Specimen Chapters of an Assyrian Jules Oppert. In 1874 Halévy published his theory that Grammar” (1866) in which he sets out his analysis of the Sumerian language and people never existed. Op- the Akkadian verb. pert, who had proposed Sumerian as the correct name for the language in 1869, took exception to this theory. §7. Concluding Remarks The ensuing controversey drew in other scholars and §7.1. Rawlinson rarely acknowledged Hincks’ contri- irrelevant chauvinism bedevilled the exchanges. The butions to the decipherment of cuneiform. In a foot- whole affair is discussed at length by Jerrold Cooper note he once wrote (1850: 448): “I bear a most willing (1991). When Hincks postulated that cuneiform had a testimony to the great sagacity which he has brought non-Semitic origin in 1849, he could scarcely have be- to bear on this and many other points connected with lieved that his discovery would give rise to such rancor. the cuneiform inscriptions, and which very frequently has rendered him independent of data.” In letters to §6.6. On 28 December 1861 Hincks wrote to Henry Layard on the other hand, he usually dismisses Hincks Fox Talbot (Cathcart 2009: 114): “I am surprised at and his work. When Hincks was employed for a year your speaking of the Chaldaean inscriptions as Hamitic by the British Museum, Rawlinson complained bitterly and diffi cult to decipher. I fi nd them almost as easy to to the trustees that he found himself “supplanted by a translate as the Assyrian and they have not the slight- gentleman, who, however it may be sought to disguise est resemblance to the Egyptian. They are in an agglu- the fact, is notoriously indebted in great measure to tinated or Turanian language.” Talbot meekly replied my published papers for his present undoubted profi - (Cathcart 2009: 115): “The term Hamitic was used by ciency” (Larsen 1996: 335). A serious evaluation of the me conventionally; certainly not as implying any resem- materials at our disposal, however, supports Julius Well- blance with the language of Egypt. The term I myself hausen’s view (1876) that one is not saying too much, if selected was Proto Chaldean, but I found that Hamitic one calls Hincks the true decipherer of Assyro-Babylo- and Accadian had been employed by others, and I chose nian cuneiform. There can be little doubt that anyone Hamitic as the shortest. If it is a bad name I am ready to who took the trouble to study carefully Hincks’ papers adopt any other. I believe Rawlinson coined it.” Around published from 1846 to 1852 would have made a good this time Hincks and Rawlinson began to use the term beginning in cuneiform research. Old Chaldaean. §7.2. Rawlinson is best remembered for the editing and §6.7. It is not my brief here to describe the progress of publication of the fi ve volumes of Cuneiform Inscrip- cuneiform research in the late 1850s and 1860s, but I tions of Western Asia (1861-1884) with considerable must mention Hincks’ article “On the Polyphony of the help from Edwin Norris and other scholars. Two other Assyrio-Babylonian Cuneiform Writing” (1863) which fi gures, Jules Oppert and Henry Fox Talbot, are mak- was published in the form of a long letter addressed to ing their presence felt when the fi rst critical phase of the Egyptologist Peter le Page Renouf, Professor of An- decipherment has been accomplished in 1852. Oppert’s cient History at John Henry Newman’s Catholic Uni- contributions to the elucidation of Akkadian and Su- versity of Ireland in Dublin, who later became Keeper merian merit a detailed study. Only then shall we be of Egyptian and Assyrian Antiquities at the British able to assess his role in the next phase of decipherment. Museum. This important article is an extended discus- He certainly deserves a place next to the genius Hincks sion of an aspect of cuneiform writing with which even and the hard-working Rawlinson. Cuneiform Digital Library Journal 2011:1 page 9 of 12 BIBLIOGRAPHY Bermant, Chaim & Weitzman, Michael 1979 Ebla: An Archaeological Enigma. London: Weidenfeld & Nicholson. Black, Jeremy, et al. 2004 The Literature of Ancient Sumer. Oxford: Oxford University Press. Burnouf, Eugène 1833 Commentaire sur le Yaçna. Paris: Imprimerie Royale. Cathcart, Kevin J. 1983 “Edward Hincks (1792-1866) and the Decipherment of Cuneiform Writing,” Proceedings of the Irish Biblical Association 7, 24-43. 2007-2009 The Correspondence of Edward Hincks. 3 vols. Dublin: University College Dublin Press. Cathcart, Kevin J. & Donlon, Patricia 1983 “Edward Hincks (1792-1866): A Bibliography of his Publications,” Or 52, 325-356. Cooper, Jerrold S. 1991 “Posing the Sumerian Question: Race and Scholarship in the Early History of Assyriology,” AuOr 9, 47-66. Daniels, Peter T. 1994 “Edward Hincks’s Decipherment of Mesopotamian Cuneiform,” in K. J. Cathcart, ed., The Ed- ward Hincks Bicentenary Lectures. Dublin: Department of Near Eastern Languages, University College Dublin, pp. 30-57. 1996 “Methods of Decipherment,” in P. T. Daniels & W. Bright, eds., The World’s Writing Systems. New York & Oxford: Oxford University Press, pp. 143-159. 2009 “Rawlinson, Henry ii: Contributions to Assyriology and Iranian Studies,” Encyclopaedia Iranica. http://www.iranica.com/articles/rawlinson-ii Friedrich, Johannes 1966 Entzifferung verschollener Schriften und Sprache. 2nd ed. Berlin & New York: Springer-Verlag. Extinct Languages. New York, 1957, is a translation of the 1st edition. Gordon, Cyrus H. 1968 Forgotten Scripts: The Story of their Decipherment. London: Thames and Hudson. Grotefend, Georg F. 1815 “Über die Erklärung der Keilinschriften, und besonders der Inschriften von Persepolis,” in A. H. L. Heeren, Ideen über die Politik, den Verkehr und den Handel der vornehmsten Völker der al- ten Welt. Vol. 1. Göttingen: Vandenhoek & Ruprecht, pp. 397-433. Eng. trans. 1833: “On the Cuneiform Character, and particularly the Inscriptions at Persepolis,” in Historical Researches into the Politics, Intercourse, and Trade of the Principal Nations of Antiquity. Oxford: David Al- phonso Talboys, pp. 313-360. Halévy, Joseph 1874 “Observations critiques sur les prétendus Touraniens de la Babylone,” JA, 3rd ser., 4, 461-536. Hallock, Richard T. 1955 “Syllabary A,” in MSL 3. Rome: Pontifi cium Institutum Biblicum, pp. 3-45. page 10 of 12 Cuneiform Digital Library Journal 2011:1