by Arash Zeini
The need for even a short Pahlavi-English dictionary has been obvious for many years, particularly to would-be students of the language. The glossarial part of Professor H. S. Nyberg's Manual of Pahlavi is still eagerly awaited, five years after the publication of Part I, although the 'archaistic' tradition of transcription followed therein is no longer universally acceptable. With the partial material at present available, every serious student still has to compile his own dictionary, a formidable task with Pahlavi and made more so by the plethora of different systems of transcription in print. The present dictionary, admittedly introducing yet another transcription, is intended first to relieve the stranger to Pahlavi of the necessity of covering what is now well-travelled ground - the commonest 4,000 simple words - without a unified map, and secondly to urge the practical value of this 'Sasanian' transcription for his journeying further. To have included etymologies and textual references would have changed the emphasis of the work, turning it from a map to a historical atlas-cum-gazetteer, besides retarding its appearance unduly. Such a complete dictionary of Pahlavi is nowhere within sight and must remain so until many more texts are adequately edited.
For his good counsel on what to include in the dictionary I am most grateful to my friend Dr. P. Gignoux of Paris, for without him many common words would have slipped through my net. Most of the labour of compiling the English index, which will surely be appreciated, was undertaken by my wife. The debt I owe to the teaching of the late Professor W. B. Henning, in connection with Pahlavi and all things Iranian, cannot be expressed in words.
It is again a pleasure to acknowledge the support of the School of Oriental and African Studies in financing this publication, and the customary skill of the printers.
D.N.M.
October 1969
ALMOST every published list of Pahlavi words contains a number of entries of doubtful nature, the uncertainty lying either in their form, their meaning, or even, occasionally, their very existence. The purpose of this dictionary is to provide the student with a representative vocabulary of Pahlavi in which such uncertain words have been reduced to a minimum and marked. It differs from its predecessors in that the Pahlavi words are given in a phonemic transcription representing, as far as it is deducible from the evidence, the pronunciation they would have had in the third century of our era, the period of the rise of the Sassanian empire. The reasons for this departure from the 'archaistic' tradition of Pahlavi studies I have given elsewhere. (1) If one reason stands out it is that of practicality, as most corroborative evidence comes from this period.
To reduce the number of doubtful words it has been necessary to concentrate on the simpler vocabulary of such texts as the Karnamag i Ardaxsir, Dadestan i Menog i Xrad, Arda Wiraz namag, Bundahisn, and the Pahlavi Texts edited by Jamasp-Asana. Although I have also drawn upon the Pahlavi versions of scriptural texts (Vendidad, Yasna, and Vispered), most of their very special vocabulary has been omitted. So have the many nonce-words and ad hoc compounds of religious texts such as the Denkard, since it is generally a simple matter to analyse Pahlavi compounds, given the elements and provided that the rules of their construction are understood. Those words calling for inclusion whose form or meaning remain doubtful are marked with an asterisk.
By its nature the dictionary cannot claim to contain the entire vocabulary of any particular text. Nor is there any express intention of presenting new solutions to the many problems remaining in the Pahlavi texts. Rather the readings of many scholar have been adopted, eclectically, and grateful acknowledgement is hereby made to all whose discoveries have been assimilated within. Nevertheless, I must retain responsibility for the choice as a whole and for readings and interpretations unclaimed or unpublished elsewhere.
The script of the Pahlavi books derives ultimately from that of the official Aramaic of the Achaemenian empire. In the course of time, however, a high degree of ambiguity has developed in the script. Table I shows the development of the letters from Imperial Aramaic (with the transliteration and names of letters commonly used by Semitists), through the distinct forms used in the Parthian and Persian inscriptions of the third century A.D., and the archaic manuscript forms of the Psalter fragment (about seventh century),(2) to those of the Zoroastrian books. In addition the multiple values of the letters listed, some of them may present combinations of letters in certain contexts, e.g. ***
IMAGE INSERT (table I)
Since the ambiguities of the script make a mechanical transliteration of one letter by one roman symbol of no practical value, a certain amount of interpretation is essential. This interpretation can be based, either directly or by the use of analogy, on the less ambiguous evidence of the above-mentioned earlier documents of the language. With this help it is possible to go far in distinguishing between the original values ***, ***, dy, etc., of ***, n, w, r of ***, and so on. But it is neither necessary nor desirable to go beyond this by attempting to give closer 'phonetic' values to the letters, such as r for written l, or x for written ***. For example, *** represents xwar 'sun', but to transliterate it *xwr entails both confusion with the true letter r *** and the introduction of even greater ambiguity into the interpretation of ***. It is better to accept that the scribes wrote r and l almost indiscriminately for both phonetic values, *** for both Iranian h and x, etc., and to transliterate accordingly.
It is common practice to distinguish the Semitic ideograms in Pahlavi, from the Iranian elements, by transliterating them in capital letters. Thus the same letters *** are transliterated *** when they stand for rahi1g 'child', but *** when they represent the ideogram for du1r 'far'. However, as the ideograms are often much distorted and disguised forms of the original Aramaic words, it is sufficient for practical purposes to transliterate them by means of a simplified system.
The system adopted hereafter in this book, for both ideographic and Iranian spellings, is designed to avoid the use of diacritics as far as possible and also to parallel the system now conventional for the representation of the related Manichaean Middle Persian and Parthian script. Each letter is throughout given its basic Iranian value(s), thus c and C for Aramaic ***, h and H for Aramaic ***. A represents the capital form of *** (***). There are three letters which occur only in ideograms: Q (in the word QDM = abar; otherwise it is replaced by K), E, and O. E stands for the Aramaic *** (when it is not replaced by ***) and O for Aramaic ***ayin, the roman letters being descendants of these Semitic forms respectively. Thus OLE (= ***) is a form of shorthand for an Aramaic original ***lh, HCDLWN- (=***) indicates a corruption of the Aramaic root ***. Purists will find no difficulty in restoring the forms with diacritics.
Pahlavi letters are often distorted, or joined irregularly, in such a way that they coincide with other letters, so producing even greater ambiguity. The commonest distortions are here noted by under- or over-lining the letter value intended. Thus:
IMAGE INSERT(representation)
The full scheme of transliteration is given in Table II. The order is still that of the Aramaic alphabet, taking account of the first value of each letter.
IMAGE INSERT (table II)
Notes
IMAGE INSERT (notes)
The orthographic principles followed, in seemingly haphazard permutation, in the writing of Pahlavi, besides the ideographic, include the phonetic (within the limits of the alphabet), the historical, the pseudo-historical, and others so indeterminable as scarcely to merit the name. The establishment of the original spelling of a word, therefore, very seldom reveals its pronunciation. To determine this it is necessary to examine all the available comparative evidence.
Since the discovery of the Iranian material written in the near-phonetic Manichaean script it has become possible to establish a hypothetical phonemic system for Middle Persian and Parthian of the third century A.D. far more accurate than before, when only the remoter evidence of Old Iranian, New Persian, and loanwords in other languages (particularly Armenian) was available. A description of this phonemic system has been given in my article cited above. Within its terms all the entries in this dictionary are transcribed.
The system comprises the following phonemes:
|
i1 |
e1 |
a1 |
o1 |
u1 |
|
i |
(e) (1) |
a |
(o) (1) |
u |
|
p |
t |
c8 |
k |
|
|
b |
d |
j8 |
g |
|
|
f |
s |
s8 |
x |
h |
|
|
z |
z8 (2) |
***(2) |
|
|
m |
n |
|
|
|
|
w |
r l |
y |
|
|
(1) The phonemic status of e and o is doubtful and they are only marked when there is direct evidence of their occurrence.
(2) Only in non-Persian, i.e. 'learned' or loaned, words.
The phonetic realizations of these phonemes can only be supposed to have been similar to those of the corresponding sounds of present-day Western Iranian speech. The only allophonic variation of importance concerns the voiced stops b , (d ?), g , and affricate j8 . Between a vowel and a following consonant, in the same morpheme or close juncture, these almost certainly became the fricatives [***] respectively. Thus abd , tigr , duj8de1n represent pronunciations [***]. How far, or when, this development affected the same phonemes elsewhere in postvocalic position is uncertain. It is sufficient to regard [***] as possible realizations of postvocalic b , d , g , j8 .(4)
TRANSCRIPTION
In seeking an unknown word, or one known only in another transcription, it may be necessary first to find it in the Pahlavi key. The English-Pahlavi index is intended to serve as an aid to memory, and for tracing synonyms, etc., but is emphatically not an English-Pahlavi dictionary.
The main entries, with two exceptions, are arranged throughout in the following order:
|
a |
b |
c8 |
d |
e |
f |
g (y) |
h |
i |
j8 |
k |
l |
m |
|
n |
o |
p |
r |
s |
s8 |
t |
u |
w |
x |
y |
z (z8) |
|
The exceptions are:
(i) abstract nouns ending in -i1h [ -yh ], which immediately follow the words from which they are derived, e.g. stu1ri1h after stu1r , before sturg ;
(ii) verbal nouns formed from the present stems by the suffix -is8n [ -s8n' ] and their derivatives in -is8ni1g [ -s8nyk' ], -is8ni1h [ -s8nyh ], which are all listed under the verb.
Verbs are quoted in the infinitive form, followed by the present stem, except when either of these is not certainly attested. Cross references are made where necessary. Not all verbal derivatives are included if they are regularly deducible from the 'principal parts' of the verbs given, e.g. actor nouns in -da1r/-ta1r [-t***l], and past participles in -dag/-tag [-tk'], corresponding to infinitives in -dan/-tan [-tn']; the abstract nouns in -i1h formed from these (e.g. -da1ri1h, -dagi1h); agent nouns in -a1g [-***k'] formed from present stems, etc.
Pahlavi words are always given in bold type. The few entries in italics are concessions to another transcription, made for reference.
P before a main entry indicates that the word is found only in texts of Parthian origin, namely the Aya1dga1r i1 Zare1ra1n and Draxt i1 Asu1ri1g.
An asterisk denotes that the transcription or reading of the word is unconfirmed and remains doubtful.
Within square brackets the following information is given:
(i) The transliteration of the Pahlavi spelling. In the case of ideograms this is followed immediately by an indication of the Aramaic (A) origin of the spelling. With verbs this can seldom be given more closely than by the form of the Aramaic root.
When the Pahlavi spelling of the present stem of a verb coincides with part of that of the infinitive, however much they differ in transcription, the transliteration is given in abbreviated form, with a hyphen. Thus s8kwp-tn' denotes 'infinitive s8kwptn' (= s8kuftan ), present stem s8kwp- (= s8ko1f- )', YMYTWN-tn' denotes 'infinitive YMYTWNtn' (= murdan ), present stem YMYTWN- (= mi1r- )'.
(ii) Evidence supporting the transcription.
For 'learned' words taken from the Avesta the form which the Pahlavi represents, however corruptly, is quoted (as '= Av.').(5) In such cases the Pahlavi transcriptions are more hypothetical than usual, since they can be no more than probable representations of the contemporary pronunciation of these words.
Justification of the transcription of each word would require the discussion of a mass of etymological and comparative evidence far beyond the scope of this dictionary. Only that evidence is included, therefore, which is most closely related and lends itself to concise and regular presentation.
After the vertical stroke the corresponding forms from Manichaean Middle Persian (M) and/or Parthian (P) are given wherever possible. The appearance of only a Parthian form does not necessarily imply that the Pahlavi word is of non-Persian origin. It may be that the identical Middle Persian form is not attested in Manichaean script.
Later forms of the word are quoted from Judaeo-Persian (J - from whatever source) and the earliest New Persian (N - rarely in their Arabic guise, when this has displaced the Persian, Ar.).
When the forms presented are not directly comparable, being from a different dialect or in some way modified (e.g. by popular etymology), they are enclosed in parentheses.
EXAMPLE
IMAGE INSERT (page xviii)
dlidan, wen- [HZYTWN-tn' < A -/lzzy; dytnl, wyn- I M dydn, wyn-, N / I, bin-] see.
, ,isn: sight, seeing. , ,isnig : visible.
wen-, , ,isn(-ig), v. didan.
(I) -\\~ is an ideographic spelling of wen-, present stem of didan. It consists of the letters H, Z (reduced to a shape of G), Y, T, W, N, being a (normal) PahlaVi development from
the Aramaic root lzzy. The infinitive (3) has the 'phonetic com- plement' -tnl, for the -daD of didan.
(7) \\~ dytn' spells the infinitive didan, and (4) -,., wyn- the present stem wen-.
In Manichaean Middle Persian the corresponding forms are spelt dydn, wyn- respectively, indicating the same pronunciation. In New Persian the infinitive is still didan, but the present stem
has become, regularly, bin-.
The verbal noun wenisn may be written part ideographically (2) HZYTWNln' or 'phonetically' (6) wynlnl. The derived adjec- tive wenisnig is attested in the spelling (5) wynInykl (but could occur as HZYTWNlnyk').
A Aramaic
Ar. Arabic
astr. astrology, astronomy
Av. Avestan
cal. calendar
intr. intransitive
J Judaeo-Persian
lit. literally
M Manichaean Middle Persian
N (early) New Persian
P Parthian
pl. plural
sg. singular
tr. transitive
~ (the same as) the main entry
* doubtful transcription, meaning
I separates alternatives
() variant form, spelling
(N...) N, etc., form not directly comparable
= M the same spelling in M, etc.
1 'Notes on the transcription of Pahlavi', Bulletin of the School of Oriental and African Studies, xxx, pt. 1 (1967). 17-29.
2 See F. C. Andreas, Bruchstuecke eine.r Pehlevi-Uebersetzung der Psalmen. Aus dem Nachlass herausgegeben van Kaj Barr. Sitzungsberichte der Preussischen Akademie der Wissenschaften (Berlin), 1933, i. 91-152, II plates.
3 But see Kaj Barr, 'Remarks on the Pahlavi ligatures *** and ***, Bulletin of the School of Oriental Studies, viii, pt. 2 (1936), 391-403.
4 This [***] is an allophone of the Persian phoneme g, while *** of tho transcription is a non-Persian phoneme. Parthian also had a phoneme z8, replacing Persian j8.
5 In Bartholomaean transcription (Altiranisches Woerterbuch) except for ***, ***, w, replacing ***, w, v respectively.