Frahang i Pahlavik.

• Author: Shaked, Shaul

THE LATE H.S. NYBERG'S WORK on the Frahang i Pahlavik (abbreviated here FrPhl) was pursued over many years, and its publication has been eagerly awaited for a long time. It has now appeared, many years after the author's death, thanks to the efforts of his former students, the Iranian scholar Bo Utas and the Semitist Christopher Toll. They are to be congratulated for the energy and time invested in this work of love, but some disturbing questions cannot be avoided. As the manuscript was not ready for print, it was decided to publish the extant notes while supplying the commentary to the text from lecture notes taken by students of Nyberg. It is doubtful whether this does justice to the late master. If he had lived to put his hypotheses and conjectures on paper, he might have wanted to reconsider some of them. They do contain a great many untenable, sometimes fantastic, speculations. It is arguable that in deference to the memory of a great scholar it might have been better to leave the incomplete manuscript unpublished. The user of the book must at any rate be warned that many of the statements, however confidently expressed, should be treated with some caution. This applies to the Semitic as well as to the Iranian aspects of the work.

The book is extremely complex to work on, and its layout as printed does not make its use any easier. The text bristles with problems of reading and interpretation, in addition to problems of manuscript transmission. It might have been a useful service to indicate which ideograms are actually encountered in epigraphic texts or in Book Pahlavi, and which figure only in the Frahang i Pahlavik. It must, however, be obvious that the occurrence of ideograms in the extant literature is not the only yardstick for deciding whether they are genuine. Ideograms that correspond to known Aramaic words are undoubtedly also part of the repertory, even though they are still unattested in Pahlavi outside the Frahang; and others may still turn up in Aramaic literature, and should not all be dismissed as spurious.(1) It does not seem plausible that recent scribes could or would invent Aramaic forms at the time in which the book was compiled. The store of ideograms listed here, with all their numerous corruptions and incongruities, must have largely come down in school traditions from an earlier date, when they were still in use. They were most probably diligently memorized by students over many centuries.

There is a problem with the history of the FrPhl. It exists only in very late manuscripts, from the seventeenth century onwards, with a great many errors of transmission and many variants, which make it almost as fluid as oral transmission. The book may have been treated to some extent as an open-ended notebook, in which copyists and users could make additional remarks, or introduce their own lists. This would account for the lack of homogeneity in its composition, for the fact that genuine ideograms are listed side by side with variant spellings of the same Iranian word, and other similar features.

Another problem that besets the exploration of the ideograms is that of the existence of several variant letters used to convey certain identical sounds, and of the frequent interchange of letters.(2) Here is a list of the most prominent categories of corruption attested in ideograms:

/T/ and /D/ interchange in the Aramaic words used as ideograms, as they sometimes do in Middle Persian words. Some clear examples are: 7:21 GLLTA(3) = pwst 'skin', for Aramaic gld ; 7.6 KWTYNA(4) = stl 'mule', for Aramaic kwdn; 23.3 SDRWNtn': STRWNtn' = plyst tn' 'to send'.

/P/ and /B/ also interchange in some cases. Examples: 1:16 KWKBA : KWKPA = st lk', for Aramaic kwkb ; 11.2 PLBAY, PLBA = zywndk 'living being', the Aramaic word that underlies the ideogram being perhaps (if Geiger 1912:305 is right) bry, bryt. Similarly, 11:19 LP(YD)A: LPYA = lytk' 'boy', for Aramaic rby (cf. Geiger 1912:305); 7:20 GWBTA: GWPTA(5) = pnyl 'cheese'. 10:31 A CTPH(6) = ngwst' 'finger', for Aramaic sb t, where transpositions of letters also occur.(7)

/M/ turns up for /N/ in 19:2 SWMAHL = gwlbk 'cat', for Aramaic swnr, and /N/ seems to turn up for /M/ in 1:17 TTLWNtn' = w lytn' 'to rain', where Nyberg is probably right in assuming the ideogram is a graphic distortion of YNTLWN--from the Aramaic root mtr.

/S/ occurs instead of /S/ in some ideograms; cf. 1:11 SMSYA |"corrected" by Nyberg to SMSA~ = xwl 'sun', for Aramaic sms ; 2:11 STRA = l k' 'side', for Aramaic str ; 17:3 XWBSYA = zynd n 'prison', for Aramaic hbws. /S/ occurs instead of /S/ in 10:36 KLZDH (an error for KLSH): GRSH = skmb' 'belly', for Aramaic krs or krsh.

/N/ is sometimes changed to /L/, since the other value of /n/ is /r/, and it is then identical to the usual phonetic value of /1/. This is attested in the word 7:7 KLYA(8) : KYNA: gwspnd' 'sheep'.

It is possible that on one or two occasions an assimilation of /D/ took place before /N/, unless they are matters of simple graphic corruption. One example is 10:10 AWNYA = gws 'ear', from Aramaic wdn.

The writings discussed above, in which phonetic hypercorrection is involved, i.e., /P/ : /B/ ; /T/ : /D/, operate in the same way as they do in the Iranian words in the Pahlavi script. This fact points to a mode of transmission of the ideograms that must have been partly oral. The spelling of the ideograms thus became influenced by a scribal tradition of pronunciation. This is also supported by the substitution of /N/ and /M/, and of /N/ and /L/, in addition to the fact that in a word such as 19:13 SGDH = nm c 'adoration', the first letter shows a prothetic vowel probably added by the Iranian scribes to a word beginning with two consonants. If the interchange of /S/ for /S/ came, as suggested by several scholars, under the influence of Arabic, there are a number of words where Arabic cannot explain this substitution. The graphic similarity of /S/ and /S/ may account for this change, or perhaps the phonetic similarity between the sounds.

These types of change should be kept separate from the many cases of purely graphic, or orthographic, corruption. A short list of such phenomena may be given:

Reduplication of letters is fairly common, as in 7:21 GLLTA.(9) Another group of typical reduplications is the intrusion of an /M/ before final /H/, these two letters being graphically very similar: 10:35 LBBMH = dl 'heart', for Aramaic lb or lbh; 10:33 GBMH = pwst 'back', for Aramaic gb or gbh.

The inversion of the order of letters is attested in KLYA : KYNA 'sheep' (cf. below); 7:16 TYBA = xwk' 'gazelle', for Aramaic tby ; 17:3 XWBSYA = zynd n 'prison', for Aramaic hbws ; 4:9 KXMA = lt' 'flour', for Aramaic qmh ;(10) 7:10 XLLN = mes 'sheep, ewe', for the plural of Aramaic rhl.(11)

The change of /B/ to what looks like /G/, /D/, or /Y/ can be explained as a graphic contraction, by the fact that /B/ was often joined to the following letter. E.g., 21:11 YNSBWNtn': YNSDWNtn' = ysttn' 'to take'; 20:21 YDLWNtn' (i.e., YBLWNtn') = bwltn' 'to carry'; 7:18 XLYA = syl 'milk', for Aramaic hlb (this...