You are on page 1of 45

The Smithsonian Institution Regents of the University of Michigan

Review: Arabic Paleography Author(s): Nabia Abbott Reviewed work(s): A Survey of Persian Art by A. U. Pope ;P. Ackerman Source: Ars Islamica, Vol. 8 (1941), pp. 65-104 Published by: Freer Gallery of Art, The Smithsonian Institution and Department of the History of Art, University of Michigan Stable URL: . Accessed: 30/01/2011 20:43
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. 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

The Smithsonian Institution and Regents of the University of Michigan are collaborating with JSTOR to digitize, preserve and extend access to Ars Islamica.


strating once again that these values rightly give Islamic calligraphy a highly honored place among the Islamic arts. It is a pleasure to turn to the volumes of plates and find so much calligraphic material so excellently reproduced. Contributors and editors alike show a keen appreciation of the significance of the historical development of Arabic paleography and calligraphy, though different ones have approached this problem differently, some having more familiarity than others with the historical, less artistic, but more difficult and fundamental phase of the study of Islamic writing. Most Arabic paleographical studies have been based largely on extant specimens, with dated ones serving as landmarks. This may be a good enough method if all that is desired is a general idea or appreciationof the main developments over an extensive period of time. The method, however, is decidedly unsatisfactory if one wishes to follow finer developments within shorter and consecutive periods, unless one has the good fortune to find an unbroken series of dated specimens in several of the different fields of Islamic writing such, for instance, as the Koranic, secretarial, and epigraphic fields. As yet no such unbroken dated series have come to light; neither are they ever likely to come to light. Therefore, if further progress is to be made, a serious effort must be made to supplement the evidence of the specimens with that of the literary sources bearing on the problem. The truth is that the great majority of such sources -Arabic, Persian, and Turkish-are as yet either largely unknown or little explored. This may be due partly to the fact that frequently the history of Arabic writing has been touched on by students interested in it less for its own sake than for what side-light it may throw on some particular problem they happen to have






opening the pages of Ars Islamica to an extensive series of comprehensive review articles on A Survey of Persian Art'-articles that may go beyond the limits of a general review to that of the presentation of new materials-gives me the welcome opportunity to kill three birds with one stone. In the following pages I shall attempt to review the paleographical materials offered in the Survey, dispose of some points raised by Arthur Jeffery in a review of my study, The Rise of the North Arabic Script and Its Kur'anic Development, and present new materials that lead to some suggestions on the development of early Islamic secular scripts. The specific sections here reviewed are to be found under the heading of calligraphy in the second volume of the Survey (Chap. 46, pp. I707-84). These consist of: A. "An Outline History," by M. Minovi, P. Ackerman, et al. (pp. I70742).

B. "Ornamental Kfific Inscriptions on Pottery," by S. Flury (pp. I74369). C. "Ornamental Naskhi Inscriptions,"by V. A. Kratchkovskaya (pp. I77084). The different contributors and the editors have succeeded, both by choice of illustrative materials and by comment, in bringing into due prominence the artistic and decorative values of Arabic and Persian scripts, thus demon1 Ed. by A. U. Pope and P. Ackerman(Londonand New York, I939), II, I707-84. Referred to hereinafter as the Survey.



on hand. These find it irksome, if not unnecessary, to delve deep into difficultand widely scattered oriental sources.2 Of the several contributors to the section under consideration, Flury alone has ignored the literary sources, and proceeded to classify the various "Kufic" scripts on the evidence of the specimens and in terms innocent of any Arabic or Persian origin and flavor. He has thereby side-stepped two difficult and thorny problems. The first is that of viewing the entire matter of the evolution and classification of scripts as the Moslem scribes and calligraphers themselves saw and understood it. The second is that of proceeding thence to the question of the identification of the various individual scripts scattered in the literary sources. Apart from this, his presentation is clear and readily understandable and usable by students who approach Arabic writing almost exclusively from an artistic and decorative point of view. Madame Kratchkovskaya's article indicates an understanding of the fundamental relationship of the angular and round or cursive scripts which she, following accepted but erroneous terminology, calls "Kfific" and "naskhl." She does, however, correctly characterize both types as springing from the same roots and as being never wholly free from a reciprocal influence. I have shown elsewhere that the terms Kufic and neskhi, originally associated each with a specific individual script, have come in time to be applied each to a large group of scripts, Kufic being applied tlo all scripts that are predominantly angular and neskhi to those that are predominantly round and cursive. Not until more progress in the identification of individual scripts, both of the angular and the cursive varieties, has been made will this erroneous and, in this instance, confusing terminology of desigCf. for instance Flury's attitude as expressed in the Survey, II, 1769, of the article under review.

nating the whole by one of its parts yield readily and completely to a more accurate one that is in keeping with the actual historical development of the numerous scripts. The use of both these terms in their all-inclusive sense has already been challenged in the field of Koranic writing, where scripts hitherto accepted as Kufic are seen to be either Meccan, ma'il or mashk and where the neskhi proper shares honors with the thuluth, both representing the cursive type of scripts.3 In the epigraphic field the term "Kufic" is still unchallenged, but the term "neskhi" is slowly yielding to thuluth in which script the great majority of cursive inscriptions are in truth written. Madame Kratchkovskaya's text creates repeatedly the impression of contradiction when speaking of the development of cursive (her naskhi) scripts. The apparent contradiction rises out of the fact that statements which are true in the development of epigraphic cursive alone create the impression of being meant for the development of cursive writing in general. The editors, aware of this, have clarifiedher position in most instances.4 However, they have overlooked one real and evident contradiction, in which, on page I 77I, she refers to an eleventh-century manuscript of Ibn al-Bawwab

(d. 423




as the first authentic

cursive (her "naskhi") extant and on the following page herself draws attention to an extant manuscriptof al-Balkhi dated 320 H. (932 A.D.) and therefore something like a century earlier than the manuscript of Ibn al-Bawwab. This date again must be understood to apply only to the manuscript or book variety of the round
3 See N. Abbott,The Rise of the North ArabicScript and its Kur'anic Developmcnt ... Oriental InstitutePublications,Univ. Chicago, L (I939), I7-37 andindexunder namesof scripts;cf. also N. Abbott,IKurrah Papyrifrom

Aphrodito in the Oriental Institute ... Studies in Ancient Oriental Civilization, Univ.Chicago, No. i - (I 938).

E.g., pp. 1770n. 6; p. '7,I ns. I and S.



script and not to the earliest use of round script in general in other fields, for example, papyri and coins, to which the author makes earlier references. The point is well taken that the ornamental cursive script is not necessarily simpler than the Kufic; but the illustrations, though they reveal some of the complexities of the ornamental varieties of the cursive, do not by any means exhaust these complexities, which in some instances render the deciphermentof a cursive inscription a more difficult task than that of the reading of an exceedingly complicated Kufic text. This limitation of specimens, however, is imposed on the author by the triple limits of her theme, which confine her to the ornamental use of epigraphic cursive scripts in Persia. In using the terms "foliated naskhi," "palmetta naskhl," and "interlaced naskhi" attention is drawn to the interdependence of the purely decorative elements in the angular and cursive scripts. Again, it must be pointed out that in the majority of cases the script referred to as neskhi is in reality the more elaborate thuluth. This brings us back to the problem of the development and identificationof the individual scripts, particularly of those in the early centuries of Islam. In the "OutlineHistory" by Minovi et al.,,an attempt has been made to sketch the history of the development of Arabic writing in the Islamic world. The importance of the subject decided the editors to include the "Outline" even when no one competent scholar could be found to undertake it. The consequent "composite" authorship of this section slows down the progress of the reader and t:he reviewer who must frequently turn back several pages seeking patiently for unobtrusive square brackets, and keeping tabs on several footnotes in order to be sure who is saying what. Gladly would the reviewer forget the "who" and concentrate on the "what"were it not very evident that all involved

are eager to see that credit goes to whom credit is due. In the following comments, therefore, whenever the name of the contributor of the particular material in question is called for it will be mentioned at the start. But before proceeding to that I pause to commend the editors for achieving a pleasing uniformity of style despite the handicap of multiple authorship. Minovi, who is largely responsible for the commendable effort to explore some of the literary sources bearing on the subject in hand, is evidently working under two serious disadvantages, viz., an insufficient grasp of the Arabic language and a better acquaintance with the calligraphic practices of medieval and modern Persia than with the Arab and Arabic-Persian practices of the early centuries of Islam. Alone or together these disadvantages have led him into several minor and major errors in reading and interpreting the Arabic sources as they have also led him to the adoption of some very doubtful theories. The following list, by no means exhaustive,takes note of such errorsas have come to light when checking his sources, the points in question having in most instances some signiricant bearing on the history and practice of Arabic writing. Page I709, line 9. The word suku4fmeans sheets or pamphlets of any writing materials, e.g., papyrus, parchment, leather, and, later, paper, and is therefore not necessarily limited to "white skin." Page I709, lines i9 ff. The reference to 'Abd al-Muttalib's note as reported in the Fibrist p. 5, lines i8 ff., cannot be traced back to Omar ibn Shabba, for in lines I5 f. of the Fihrist page, Nadim introduces other though unnamed authorities. Page I7I0, footnote 5. Minovi is in error in thinking that Ibn Khallikan's text supports the theory that the cursive scripts were subsequent to the Kufic (cf. N. Abbott "The Contribution of Ibn Muklah to the North-Arabic Script,"

Hilal al-Sabi. This last reference tells how the hand of the vizier Ibn al-Furat failed him in his penmanship so that he had to fall back on a special calligrapher- . and Arabic text pp. op. Tfimar. 1922]. continued in use until the advent of the 'Abbasids when they were restricted to the Qur'an. Journ.. io). here and elsewhere. in reading lines 24 ff. and Lit. pp. Nadim's text has been both misused and misread. then that use is more apt to have been limited to the period before the Abbasids than to the Abbasid times themselves. Kitdb al- wuzard [Leyden. Minovi does not seem to realize that beginning with line 23 Nadim is using sources other than Ibn Thawaba. The word mukarrir. pp. IKalkashandl. 8. As differentiation of bureaus and of functions within these bureaus advanced special muharrir calligraphers were employed whose duty it was "to copy out fair" or to make out the final calligraphiccopies of important documents of state (cf. If I am not mistaken.continued in use until the coming of the Abbasids. 83 f below). especially pp. lines 20-25. and P1." and the scribe of Walid ibn 'Abd alMalik. lines 8 ff. this and the supposed illustration of the Meccan-Medinan script in the Chester Beatty manuscript copy of the Fihrist (p. Furthermore. the thuluthain. Page I7I2. I have already drawn attention to the fact that the tuimarwas associated with very early Korans. 6. I32. I shall return to this question again. for it is Khalid and not Sa'd who is the scribe of Omar ibn 'Abd al'Azlz's Koran. lines io ff. Page I7II. Pages I7II f. Subh al-a'shd [Cairo. Ig93-I9]. were used for the large Korans. this does not convey the real meaning of the word. Diacritical points were introduced before the reign of Mu'awiya (cf. p. English introduction p. Fihrist. 8I-83). therefore. 70 ff. 2 2 and 3I] and references there cited). Sfill.. The association goes back to the as yet unidentified author of Minhdji alisaba.. Kalkashandi. saying here is that the use of the old scripts. Adab al-kuttdb [Baghdad. tells us that the tfimar. which implies a careful and accomplished writer. I56-59. The Rise of the North Arabic Script [pp. For it is not at all impossible that at that early stage some of the larger scripts before they were classified as yet as secular. as though they were the unbroken continuation of the previous passage seriously violates Nadim's text in equating the unnamed Koranic scripts of line 25 with the What Nadim is four scripts of lines 20-22. but mention instead the nisf. was an angular Kufic script (cf. Thuluthayn. while new scripts were put into secular use (cf. The unnamedKoranic scripts of line 25. V). Minovi. Page 171I. can refer only to the earliest list of Koranic scripts given by Nadim on page 6 of the Fihrist. p. p. is translated by the term "clean-copyist". LVI [I939]. therefore. 59. Koranic and non-Koranic.68 ABBOTT: ARABIC PALEOGRAPHY Amer. 38 ff. according to Ibn Mukla. has been misread and Sa'd confused with Khdlid ibn Abi'l-Hayyadi. when the Korans were limited to the old Koranic scripts. and Thuluth. the "earliest monumental calligrapher. cit." If these secular scripts were ever used for the Koran. I7I8 and n. 456 and 128). Fihrist. lines I4 ff. that is a calligrapher as distinguished from an ordinary though qualified scribe (cf.SemiticLang. who. I7I0) are Minovi's starting points for his spectacular announcement (p. This inability to handle the Arabic text has led Minovi to make the utterly untenable statement that "the Jalil. The Rise of the North Arabic Script. is mentioned earlier in that passage but is not included in the concluding summary. In the first place the four scripts of lines 20-22 do not include the thuluthain script. however. 4) that the extant early Koranic specimens are all either forgeries or suspect and that there was a widespreaduse of various cursive scripts in early Koranic writing. III. I. it is true. I907].

8i f. Page I717. I924]. p. cit. Kashf al-zunfin-Lexicon bibliographicum et encyclopaedicum. Arabica. page 79. there is the matter of the Koranic scripts. The relationships of the major Persian scripts to each other are. Doubtless they were masters of the Kufic script. There is nothing in Fliigel's notes (which are apparently responsible for Minovi's substitution of m'ras'f for the rasif of the text) that would suggest any change whatever for ma'il. and the few new scripts mentioned do not justify doubling the numbers from twelve to twenty-four (cf. but Nadim's text cannot be used to limit their activities to that script alone. page 8. therefore. thuluthain. but a copy of Yahya ibn al-Husain's Kitdb al-akkam. lines 5 f. footnote 3!. 2 2 ff. shandi. lines 4 ff. in which case the reference is to page 6 and not 7. pp. therefore early Korans said to be written in the Meccan-Medinan scripts were written in cursive scripts (p. not to twenty-four in each case as Minovi has it. III. 68) Minovi has persuaded himself that the djalil. Corpus Papyrorumn Austriae [-CPR. Copious and valuable materials are made available in connection with numerous famous Persian calligraphers.and attention is drawn to extant copies of their calligraphic masterpieces.] III. footnote 6. Minovi is presum- ably still quoting from the Flilgel edition of the Fihrist. It is one of the very few specimens of the angular writing in other than Koranic manuscripts (cf. I.ed. Vol." but only that they were among the "big. outstanding Kufans.)." that is. Next by violating Nadim's account of the scripts (see p. Fikrist. lines 9 ff. In this early field Minovi has proposed some theories that are extremely vulnerable. lines I-5. HIadidjl1 Khallfa. tuimar. For some obscure reason he leaves out the word ma'il substituting in its place in the text the word m'ndb'dh. op.). I [Vienna. well stated. and footnote i. Minovi's position here seems to be as follows: The illustration of the Meccan-Medinan script as found in the Chester Beatty manuscript is in cursive. Page I724. in fact their calligraphicfame must rest on proficiency in several of the early Koranic scripts. First. Page I7I2. GrohRaineri Archiducus mann.). why is this new source not indicated? And why the peculiar transliterationof both these terms? Cf. I49 f. For the treatment of the riyasi script and its derivations see pages 92-94. G. 8-9 refers to twelve main scripts and twelve variations.REVIEW OF A SURVEY OF PERSIAN ART 69 scribe for the final and fair copies of his work. Is this something new in the Chester Beatty copy of the Fihrist? If so. and thuluth. KlalkaPt." since paper was not yet in general use in the diwans in Mansuir'stime. 1835-58]. For criticismof Minovi'streat- ment of the imsak script.. 474 f. Page I714. as was also the term kdg4id. Page I717. he is on surer grounds both in his command of the Persian language and sources and in his first-hand knowledge of Persian calligraphic practices. Later.pp.. though this in time came to be associated more specifically with paper (cf. does not say that Koranic calligraphers of this list "were famous for their mastery of Kfific. The list on page 9 duplicates in part that on manuscript of al-BalkhI referred to is not a Koran as Minovi and the editors state. in the light of present knowledge of these scripts. I7I a). Ser. and 53. Fliigel [London. Fihrist. in this connection. The Page I7I3. more usable and reliable than that which deals with the earlier Arab and Persian practices. 7. all of which he erroneously assumes to have been predominantly and equally cursive . The word kirtdsin Pjjahshiyari's text should be translated "papyrus" and not "paper. This section of the "Outline History" is. I now draw attention to these. see page 97. A. When Minovi deals with the later Persian scripts. II. the term came to be used indifferentlyfor both papyrus and paper.

Gessellsch. . which he attributesto the early second century. the first of which to receive a distinctive name being the Meccan. but one need hardlv caution against drawing any conclusions from such an irregular and nonvalid specimen. instances pointing to forged documents. dated by Moritz as of the third century. e.g. From here it is easy enough to see how he is now ready to account for the apparent preponderanceof Kufic (that is angular) Korans attributed to the early centuries by wholesale forgeries of such Korans and to account for the use of the Kufic script by the forgers because of their erroneous assumption that that was the script current in the first centuries of Islam (p. written in a crude hand. The script shows no affiliation whatsoever with that of the Chester Beatty bismillah. cf. the muhakkak. deutschen morgenl. Let us examine Minovi's position. however. until around the turn of the century differentiation of scripts paralleled differentiationof functions and progressive interest in writing kept pace with the advance of the great culture that flowered in the second century of Islam. Kufan.70 ABBOTT: ARABIC PALEOGRAPHY scripts. I have both in The Rise of the North Arabic Script and in The Kurrah Papyri from Aphrodito dealt at some length with the simultaneous development of an angular and a cursive variety of the earliest Arabic scripts. Again he assumes that new Koranic:scripts developing at this time were also cursive. then there should be no difficulty in finding other specimens of this script in documents dating from these centuries or even from the third century. Careful search for and study of documents from these early centuries. "Zwei arabische Papyrus. If the Chester Beatty specimen represents the true cursive variety of the Meccan-Medinan script. starting with the supposed illustration of the Meccan-Medinan script as found in the Chester Beatty manuscript copy of the Fihrist. but is instead similar to that of many an early papyrus document. He gives as further support of his theory of wholesale forgeries. that would place it. step by step. XXIV (i 88o). d. but is considered a very early copy of the Fihrist.5 In fact one must go to the fourth. in the last quarter of the fourth century. At this point he states his belief that because of mass destruction of Islamic libraries not many early Korans. and Basran. Medinan. Neither Minovi nor the editor's note 4 on page ][ 7IO tells in what script the entire manuscript is written.. when Nadim himself completed his Fihrist. were with the coming of the Abbasids restricted to Koranic writing.and fifth-century scripts to find the peculiar alif of the specimen with its sharp-hooked bend to the left at the top." Zeitschr.. in the order indicated. I regret the lack of opportunity to see and examine this particular manuscript. cursive or otherwise. The manuscript is presumably undated. 43). though the implication is there. that the illustrative bismillah is in a different script from that of the whole. including some Korans attributed to Ali. I7I8). faulty in its haphazard comixture of angular and cursive elements. at the very earliest. Their use in the first century of Islam perforce had to cover all fields of writing. Therefore he is now ready to accept the widespread use of cursive scripts for early Koranic writing. and for the treatment of the short lams in allah. 687-9I. Again if the specimen under consideration 5Loth mentions one instance of a Koranictext oni papyruswritten on the back of a private letter. The one Koranic fragment I know of written in a decidedly early cursive script is a papyrus illustrated in Arabic Palaeography (Pl. It too has no similarity whatsoever with the Chester Beatty specimen. The text consists of suras I I I2-I4. have failed to bring to light any document that could even by a long stretch of the imagination be said to be written in the script of the bismillah of the Chester Beatty manuscript.have survived from the early centuries.

In these the alif has a bend to the right. ed. and this slant is seen to be as undoubtedly and as consistently a downward one to the left. . Flugel (Leipzig." "a finial hooked in the opposite direction. the willful destruction of public or private libraries was linked with a politicoreligious controversy of the day. that were the marked object of wrath and destruction. the alif. says nothing about a "slightly oblique terminus. 187I-72). and not copies of the sacred Koran. and it was this body of "foreign" and "dangerous"manuscripts. Minovi's chief difficulty here is again his inability to use and translate correctly the Arabic text of the Fihrist. 6. More often than not." The three characteristics that the Fihrist specifies are: An alif with a bend to the right. The Arabic passage in question runs as follows: Minovi's treatment of the passage.. denouncedby the propagandistic destroyers either as superfluous or as heretical materials. and Lit. Mackensen. LI (I935). But I found the most perfect fit in numerous Koranic manuscripts. became convinced and still remain convinced that here one has a genuine Koranic variety of the Meccan-Medinan script. am convinced that it does not. The method of my search and its conclusions have been fully detailed in The Rise of the North Arabic Script.REVIEW OF A SURVEY OF PERSIAN ART 7I does indeed represent Nadim's illustration it should therefore answer to the author's verbal description of the Meccan-Medinan script. Starting out with the Fikrist passage. Semitic Lang. together with a slightly swaying rhythm. Minovi has persuaded himself that it does fit that description. Journ. Doubtless many precious Koranic codices have been thus lost along with non-Koranic works. therefore. namely. an alif of pronouncedheight. R. But the loss of manuscripts. and (there is) in its form a slight slant. again the Fibrist does not specify the direction of the slant." Amer. It was the literature leading up to and dealing with the particular situation that was both suspected and feared. The next point in Minovi's argument is the mass destruction of Moslem libraries. and an alif with a slight slant. he soon loses track of it and ends up by describing the specimen before him instead of ascertaining the full significance of Nadim's text and checking on how closely that text does or does not fit his specimen. like a few others before me. I..7 If any Korans were the object of willful de7 Cf. on the other hand. 83 ff. I found now one now the other of these alif characteristics alone or together in several specimens and in various fields of writing. S. colored by his specimen (p. and a finial hooked in the opposite direction. here there are the high rising alifs with a slant. reads: "The Mecca and Medina scripts were characterized by an alif with a slightly oblique lower terminus. "Moslem Libraries and Sectarian Propaganda. without specifying whether this bend is at the lower or upper end." A literal translation of the Arabic reads: "As for the Meccan and Medinan (script) there is in its alif's a turning to the right and a raising high of the fingers. slanting to the right. p. I7IO). and the bend is now seen to be undoubtedly and consistently at the lower end. one to conclude that the rest of the letters are alike in the Meccan-Medinan and the KufanBasran scripts. thus leaving 6 Fihlrist." This verbal description. particularly of Koranic ones. I." and "a swaying rhythm. through willful destruction cannot have been as great as is popularly supposed. Armed with this description I searched far and wide for specimens that would fit it. The remarkablething about this verbal description is that it definitely limits the three distinguishing marks of the Meccan-Medinan script to the characteristic of one letter only.

I937). especiallyp. p. If all known collections could be brought together. If. one accepts the promised proof on faith. On the other hand this essential uniformity of the text of these specimens may mean that the Koranic text was fixed for all practical purposes and general use. the second half of the third century. 43I. . Grohmann. almost always of good script and workmanship.Baghdadhad a hundredbooksellers. for the time being.the sacredness of the book. The bulk of Koranic output.).72 ABBOTT: ARABIC PALEOGRAPHY struction in these earlier centuries they would more likely be the noncanonical versions that were tabooed by the orthodox party. the tendency has been to exaggerate the number of those that have actually survived to our own day by giving a number to single folios. ed. I73 f. "Bibliotheken und Bibliophilen in Wien schenOrient. tlle collective orthodox Koranic output must have far exceeded that of any other one book. The demand for the finer copies doubtless exceeded the supply.and fourth-century Moslem world was completely overrun by rogues engaged in wholesale forgeries and by fools who in this respect were their willing and gullible victims? Or again must one believe that these forgers were not only rogues but also fools who set out to forge scripts and documents without taking the utmost care to produce as faithful an imitation as possible? Is one to assume further that these foolish clumsy forgers went about their roguish business successfully fooling all the people all the time. Kitdb-al-Mas hif. are all factors working for the survival of a goodly number of these early Koranic manuscripts. does that necessarily prove that these specimens even as forgeries do not give a fair representation of Koranic scripts and practices of the earlier centuries? Must one believe that the third. im IslamiA. at least as far back as the time of MuThe Rise of the North ArabicScript. which frequently gave it immunity from willful destruction. dating according to some. Jeffery (Leiden. the probabilities are that the actual number of extant complete Korans or even that of complete adjza' will shrink considerably. and sorted out. and for every one of these there must be assumed at least one copy of the sacred text. ata period earlier than nowsupposed or guessed. who copiedout Koransfor a fee. separate adjza'. Ibn Abi Dawiud. 29. under these conditions. circulated freely in the private book-markets and in the shops of the warrak. . Minovi states his conviction that the great majority of extant Koranic specimens are forgeries of later centuries and promises to give the proof of this statement in a forthcoming monograph. All in all there is little reason to look upon the extant collections as suspiciously large. I30 H'.8 though the practice of buying and selling Korans was much earlier. p. had need to be both regular and continuous. A. .9 The scale of production. particularly in these centuries in which 9 Cf. The output of Koranic manuscripts. soon these became famous objects of art. and Koranic manuscripts were to be found not only in these outstanding libraries. and larger divisions of the Koran. and the durability of the parchmentson which it was usually written. 1926). but in the numerous and prosperous lesser mosques of every large Moslem city. I think. public or private. 8 Cf. sought out and treasured as such. I75. Furthermore. . and in the possession of many learned and pious citizens. 'awiya. since a complete Koran that was a calligraphic masterpiece took years to write. Ya'kiibistates that in his time." Festschriftder Nationalbibliothek (Wien. small groups of folios. pp. however. Again there was hardly a Moslem village that did not have its mosque.. Such a warrak was Malik ibn Dindr (d. This may in part explain why so far none of the early Koranic specimens studied have yielded anything but the orthodoxKoranic text.anor stationers-copyistsbooksellers. compared.

It is significant in this connection that the few unquestioned Korans of the first four centuries of Islam. cit. cit. I7I8) is to be oversuspicious. or almost all. contrary to the historical evidence of the widespread use of various cursive scripts in Qur'ans. "Arabic Writing. In a recent review of The Rise of the North Arabic Script. I7I8). II3 f. For there were in the Moslem world of the middle of the third century scribes who could identify old scripts and detect current forgeries by methods similar to those used today by writing experts in civil and criminal courts. The same may be said of the case cited by Minovi (p. 43 f 14The Moslem XXX (I940). also F. "The Contribution of Ibn Muklahto the North-Arabic Script. I7I8) "the apparent preponderanceof the Kfific.'4 He questions the probable 13 Cf. cit."2 10Cf. Again. op. op. preciatethe drawingof attention (p.I (I924). One should be reasonably suspicious of extant Korans attributed to such leading religious and political figures as Ali and Othman. namely Ibn Mu. namely (p." Ibn al-Bawwab at the request of his Bfiyid master.. Isldm ( as has been seen. The suspect." do not seem to be any different in scripts than the great majority of the angular Korans known. Abbott. 529-32. But to extend these suspicions to all extant Korans of the angular types as does Minovi (p. "The Kalkashandi. Grantof Land by Muhammad to Tamim ad-Ddrl.S. Hil&lal-Sabi. I9I3). they would not prove that for that reason alone our extant Koranic specimens are all. Moritz. The burden of proof must rest on him. His "historical evidence" rests. Survey. and who established readily the guilt or innocence of the accused.REVIEW OF A SURVEY OF PERSIAN ART 73 flourishedthe two foremost master calligraphers of all Moslem ages. I9I-98. Krenkow. B. so that there is room to welcome any diligent student in that field. forgeries."Islamica." is faulty in both its points. and are likely as not to be the workmanship of a clever Jew or an unscrupulous scribe seeking self-advancement. Cases of forged deeds and letters come now and again in one's way. Othercases are knownwhereforgeriespassedmuster paleographicallyonly among the uninitiated. pp. It is to be readily seen that there is serious need for further work in the difficultproblems of Arabic writing and scripts. I. The "preponderanceof Kuific"is an exaggeration due to failure to perceive that what is termed collectively "Kufic.II. some definitely dated. i99." in fact represents several of the minutely differentiated Koranic scripts listed by Nadim. Arthur Jeffery offers his first contribution to these problems. VI.."Encyci. 7I f. pp. Baha alDawla. or to such famous calligraphersas Ibn Mukla and Ibn al-Bawwab. op. 12E. was called upon to write in the presence of the expert who then compared this writing with that of the document involved in the forgery case. replaced in his own handwriting a missing djuz of a Koran written by Ibn Muk. and I have yet to come across any references in these sources to Koranic forgeries in general.. 11Cf. Despite careful examination none could detect the replacement. Minovi's further argument in favor of his theory of forgeries. 388. one reads.klaand Ibn al-Bawwab? Incidentally. I7I8.g. I98) to the inade- .13 It is to the credit of the Arab historians that they report these cases of forgeries as readily and openly as they do more flattering practices. But even if a few such referenceswereto show up. In presenting the above facts and views I give Minovi the opportunity and the advantage of taking them into considerationin his promised monograph. I apWorld."pp. a story involving a Koranic script of these two might be termed a first rate case of "forgery. cf. on a misreading and misinterpretingof Nadim's text.10 Paleographicallyit passed muster not because of ignorance and gullibility but because it was an excellent specimen of the "forger's" art. I20 f. 67 f. Minoviwill likely add a few more in his promised monograph.

. The script of both is well developed and regular.) respectively. I have no access to the latter's source. involves an attempt at careful and sustained study of numerous specimens. That the diacritical signs in the Korans were strokes and not dots is repeatedly demonstrated in the numerous extant early Korans including the very earliest (cf. He (pp. I-I2 page). that these terms are used in a different sense. cit.. and slant of the alif-is to trifle with evidence! Neither his assumption nor the expectation is justified. I25). and my method of approach to that identification. It is immaterial whether Yahya died in go or I29 H. 59 f. I. op. There is no reason to assume that Koranic scripts alone lagged behind in this respect. Wafaydt ala'yan wa-anbd' abnd al-zamdn ([Cairo. of this dating. paralleled by an effort to comprehend the literary sources. (720 A. I question De Slane's translation (I. I96) "seriousmisgivings"on the ground of regular and elegant writing. cit. among others. quently interchanLgeable by the different grammarians and authors. I892]. note I5I. Jeffery starts out with the assumption that the Koranic scripts of the first two centuries of Islam were generally crude or primitive. I am fully aware.D.D. thinks that to accept Nadim's characteristics of Meccan-Medinan script -characteristics which. Two Korans illustrated by Moritz in Arabic Palacography (Pls. Again 15The Rise of thleNorth ArabicScript.74 ABBOTT: ARABIC PALEOGRAPHY dating of the Oriental Institute specimens. Grohmann either accepts or suggests for these two the most likely dates of I07 H.). Arabic Palaeography and 3I-34. Since I find that I cannot agree with Jeffery in either his methods or his conclusions. can be regarded as "nothing more than scribal fantasy. p. however. and note 3 on p. P'ls. I94 f. I. I feel it necessary to state at length the basis of this dissent. since he is not the only onel of the period credited with both the diacritical and the vowel systems. coins.) therefore. 382. is credited with the invention of the nukat that differentiate one similarly formed letter from another. and Hadidji Khallfa. 30I f. no more elegant than that of the two Korans under consideration. also p. several millimeters in length.. therefore. op. Hajd may have used a different version or edition of Ibn Khallikan's text. I. and i'djm in this particular passage.. I-I2 and 3I-34) date from the first half of the second century. pp. freand sometimes very ambiguous. including the Khalif a omission of this reference to Yahya. p. these hairlike strokes. to display some outstanding and spectacular differences that should readily meet the eye. and it is no more and no less elegant than that of the inscriptions of the second half of the first century. where Yahya. Abiu Ahmad al-'Askarl. inscriptions. quacy of the references given on page 39. and rounded out with a correlationof the two. Kitdb al-taskif (cf. the identification of the Meccan script. deal only with the bend. pp. I905]. Again the script of some of the parchment Korans described in The Rise of the North Arabic Script and dated there as of first to second century is. to be used again presently. nevertheless. 6o. fail to see any point to Jeffery's (p. 52). and the following [Cairo. the diacritical points (cf." or a careless execution of what was meant to be a dot! misused the Arabic sources Jeffery has unafortunately (cf. that is. 78) of the book reviewed. It must be pointed out here that Ha1jdji Khalifa's text when compared with his reference to Ibn Khallikan. That method. To these should be added the well-known HaQjl Khallfa (III. cit.) with respect to the meaning of the words huriif. II. therefore. 359 f. height. and to a lesser degree papyri. as has already been seen."5 They. leave no room for doubt on this point. The North Arabic script was already well stylized by the end of the first century of Islam. Suli. op. is seen to differ fEromit in several points. It is difficult to see how in the face of their consistent use. 55 and 65 and referencesthere cited.) and I02 H. 75-77) and the English text (cf.. can be considered as representative of scripts and practices of at least the last decade or two of the first century.). in its turn. but he seems to expect them. and extant specimens and the literary sources both indicate that they did not. Ibn Khallikin. nukat. I54 f. (725 A.

Then follows a list of separate scripts without this significant waw. varieties of Isfahani script. Here Jeffery has failed to grasp the fact that not only large mosque copies. . then there is a clear-cut statement that it was of two kinds. and that once when he saw a Koran written with a fine.19 giving the meaning that the fairamfizi script was claimed by the Persians as their own and that it was either of recent origin or new in Nadim's neighborhood. and six varieties of Kuifan-Basranhe gives without break. 1932). ed. I33. wherea story to the same effect is told of Ali also. 17E. I8Fihrist. developing perhaps more or less in the order indicated. I70. I am afraid Nadim cannot be made responsible for six varieties of the Kufan-Basran and several varieties of the Isfahani.' were produced at this early time but that small copies as distinguished from the bigger "mothers" were also being written and used." Jefferyhas both misreadandmisinterpreted this short passage of Arabic text. pp.'7 Jeffery believes that he has further reason to question the identification of the Meccan-Medinan script by casting reflections on the scope and soundness of Nadim's informationon which the identification is largely based. to him they are individual scripts. from one of which developed the peculiar Persian type of script. It will be well to consider both of these accusations. al-Din 'Abd al-Rahmanal-Suyfiti. 19"From Persian to Arabic..REVIEW OF A SURVEY OF PERSIAN ART 75 these two early Korans show the use of diacritical strokes and verse division marks. the best solution of which has been recently offered by Sprengling. but that he is dealing. lines 6 ff.and Lit. but with others nearer to his own age. sage on Koranic script18 (p. would seem evident from the fact that after enumerating three varieties of MeccanMedinan.g. 327. Kitdb al-mukni'. I97) he offers as a suggestion. not with first century scripts.and again the two kinds are linked by the significant conjunction waw. II. he was displeased and striking the scribe he said: "Make large the Book of Allah Most High. that whenever Omar I saw a large Koran (not just a Koran) he rejoiced. as parallel with them. Journ. fairamfizi). the ummahat or "mothers. al-Itkan fi 'ulum al-Kur'dn (Cairo.the three varieties being connected by the conjunctive waw. especially for instructional and educational purposes. "and." in each instance. In the first place he has here accepted as a fact that which on another page (p. Referring to Nadim's pas15aDan!. Semitic Lang. that is. they seem to have escapedthe attentionof the reviewer. I93) Jeffery writes: "Ibn al-Nadim goes on to deal with Qur'anic scripts."Amer. Next appears a corrupt passage of considerable difficulty. Under no condition can the present Arabic text in the Fliigel edition be made to give us "six varieties of Kuafan-Basran"and "varieties of Isfahani script. together with decorative schemes for the heading of the suras that are far more elaborate than any seen on the Oriental Institute Koranic parchments. 6.. The small format of some of these is used as a further argument against their early dating. and surely no one is going to suggest that we have primitive Qur'anic Codices from Isfahan. LVI (I939). a small pen. p.15a It is true that there were many prominent men who preferred the larger scripts and larger format and objected definitely to the smaller ones. but both the preference and the objection imply the existence of the smaller ones. 16 Djalal Otto Pretzl (Istanbul. 54 and 59. this seems to have escaped Jeffery'snotice. until we get to the kairamiizi (Minovi.p."'16 The essentials of these facts and practices have been already pointed out in The Rise of the North Arabic Script." It does indeed give us three varieties of Meccan-Medinan. One is told. for instance. that is. Kitdb I900). His position is that Nadim spoke largely for his own times and is not to be trusted when speaking of earlier practices.

is to show oneself hopelessly at sea in the matter of early scripts. the Basri or the Isfahani script. as already demonstratedin the identification of the mashk and ma'il. such. and works of general information (akhbar). To use the argument. already famous as a calligrapherin Medina in the governorshipof Omar ibn 'Abd al-'Azlz. extant specimens alone come to one's aid. the name. Again.). cf." What Nadim actually said is that Khalid was appointed to copy Korans. from this Jeffery jumps to the unsupported conclusion that in Khalid's day one and the same script was used for all three! But this is not the only mistreatment that this passage has suffered at his hand.. one has that to start out with.): Failure to see and accept this fact leads one to Jeffery's negative and static position. continuous with this one. that in these early scripts one needs not look for marked and spectacular differences. he writes (pp. a little knowledge is dangerous. accounts on the one hand for the large number of script names that have come down to us.76 ABBOTT: ARABIC PALEOGRAPHY 196 f). I93 is fortunate enough to have both extant specimens and further literary sources to assure the identification. here.e. acceptance. that the identification of only a few instead of most or all of Nadim's several Koranic scripts throws doubts on these very identifications themselves." The truth is Nadim's very next passage. that a characteristic no more marked than a slight vertical slant or an extended horizontal stroke was to the mind of the early Moslem scribes distinctive enough to be the main characteristic to set apart one script from another. conveys no definite meaning and gives no tangible clue as to the probable nature of the minutely differentiated characteristics of that particular script. was Khalid ibn Abi'l-Hayyjdi. though meant to be descriptive. poetry. for instance. and a scribe of the caliph Walid I (86-96 H.7. as with the ma'il script. In others. Occasionally. is the case with script names like tadiawid and masnui'. be it the first or the fourth? What does Jeffery propose to do with the two centuries intervening. it gives no possible clue with which to work. Where the script name is merely geographic. therefore. in the first place. perception. as elsewhere. pp. Unfortunately. centuries that saw the real golden age of Islam? It is to be regretted that Jeffery's purely subjective approach has led him to read into Nadim's passage meaning that is simply not there. poetry. proves beyond doubt that he is speaking for the early Islamic period.. as does Jeffery (pp. how does the claim that the Isfahani could not have developed in the first century affect the period of the developmentof the other two major types said to be the earliest of North Arabic scripts? Why. and utilization of this fact open up the way for the identification of some of these scripts. Again.D. 95-9. and on the other for the difficulties in the way of easy and ready identification. identification of such scripts must come from descriptive references to them.19" This one factor. . 19aThis proves to be true of the secularscripts as well. The identification of these in turn confirmswhat has already been learned from the identification of the Meccan-Medinan script.g. and akhbar. when he made Nadim say that "the writing used for Codices in his [Khdlid's] day was that used also for poetry and akhbjr. I95) went astray. It seems. as with the mashk script. must these scripts be all identified with one century. as is the case with the Meccan-Medinan. in numerous instances. more perhaps than any other. namely.Where the script name itself is in some way descriptive. It tells that among those who wrote Korans in early Islam (sadr al-awwal)./705-I5 A. in this simple passage Jeffery (p. For there is nothing in Nadim's passage to justify this "nearer his own age" notion any more than there is anything in it to justify Jeffery's "varieties. for whom he copied out Korans.

the benefit of the doubt leans heavily in Nadim's favor. for instance. because of this limitation. But is there any guarantee that it was not? Until such guarantee is forthcoming. wherethis particular collectionis cited and partly describedwithout caston its genuineness. Jeffery (p. p. is not to be trusted for the earlier centuries. browsing among this rich treasure. therefore. one reads in his Fihrist that he had seen a large collection of autographs and rare books among which was a Koran written by Khalid ibn Abi'l-Hayyadi. is 20 Cf. Every section. whatever that may be. 440. This particular collection belonged to one Ibn Abi Ba'ra. that in this section he is able to quote the names of only two of the many writers of Codices we know of from the earliest period. A given Arabist may indeed have this double limitation. he states. But because among this large collection some documents were believed to have come from the hands of Ali and his sons and other famous persons. This disposes of Jeffery's first contention.REVIEW OF A SURVEY OF PERSIAN ART 77 "It is an interesting commentary on the limitations of Ibn an-Nadim's knowledge of early Kur'anic writing. It is but natural that each succeeding Islamic century up to Nadim's time should boast a larger number of distinguished calligraphers than the preceding one. was no collection of a narrow and partisan Shiite." Nadim. that because a given Arabist is unsound in his knowledge of Arabic scripts that he must necessarily be poor and unsound in the knowledge of the scripts of his own mother tongue. too. what he is giving is the names of some of the Koran writers who have achieved distinction in their field. ing any reflection . but the former does not necessitate the latter. When. and in so doing it also answers in part his second contention that Nadim. or roll. sheet. I93) says that Nadim's illustrations of the Himyarite and the Iranian scripts "do not encourage confidence that his information regarding early Arabic scripts was much more accurate. I93) that it is "quite clear" that all were of unknown origins and that therefore one has no guarantee that the Koran in question was really Khalid's." Here Jeffery has completely missed the reason for the giving of these names. Frequently. This. representing the works of a long list of well-known authors. For Nadim nowhere pretends to be giving an exhaustive list of Koran writers. was assigned by experts to its author and the ascription witnessed to by five or six fellow scholars. namely that Nadim speaks largely for his own times. Time and again in his unique and indispensableFihrist. Jeffery concludes (p. he tells of his knowledge of and familiarity with the contents of both public and private libraries. then. In thus passing judgment on NadIm.. Indeed it is only when he comes to the warrdquin that he really knows a good list of names. and scope. and immediately passes to another section on Codex writers in the 'Abbasid period. It contained materials from almost every field of intellectual activity.a Shiite who had received it from a dying friend and fellow Shiite described as a man "passionately addicted to the collection of old manuscripts. in questioning Nadim's soundness in connection with the early Arabic scripts and the description he gives of the Meccan-Medinan script. Op. size. there is no good reason to doubt his statement. both Shiah and Sunni. Nadim relates that he had seen the original manuscriptsof the authors whose works he is cataloguing." To condemn NadIm on this ground alone would be equivalent to asserting." Here I should like to ask if anyone has comparedthese illustrations with all the known specimens of these two scripts and found them utterly impossible? Can anyone be sure they actually represent NadIm's original illustrations? Suppose that the answer in both cases is an emphatic "yes.20 Again. cit. Grohmann. was astounded at its age.

it has certainly not been made use of in The Rise of the North Arabic Script. He has theorieswith respect to the latter to which he constantly seeks to subordinate the former. is forcing paleography beyond its strictly legitimate use. then. I92 f. but their error must be proved before it can be condemned. I92) of the Arabic sources known to and used by me. Nadim's very choice of the bismillah for his illustration must be seen as proof of the fact that he knew what he was talking about. pages 8-I2. he does these same things to the English text of several pages in The Rise of the North Arabic Script (pp. in the bismillah. and Safaitic scripts. Invention is one thing. In fact. in our present state of knowledge of both texts and scripts. use another. the controversial theory that the standard Kufan Koranic text seems not to have been fixed until the third century is treated (pp. being but human. these. Jeffery (pp. In all early scripts the treatment of the 1dm. On the other hand the fact that he felt called on to give an illustration at all may be indicative of the disuse of these scripts in his own day. That proof in Nadim's case on this particular question is not likely to be forthcoming. p. If indeed there is such a legend anywhere.) starts out first by stating that there is a tradition associating the "invention" of North21 Arabic writing with "a member" of Muhammad's family. I9I f.particularlythe height of this letter. could use . Again.) as though it were an established fact and used to cast doubt on the dating of extant early Korans whose texts show remarkableuniformity. namely. I fail to see where for the pre-Islamic period the term "Arabic script" would be any less inclusive! 21 Jeffery (Op. One needs but to read The Rise of the North Arabic Script. do sometimes fall into error. he seems to have the idea that a Koran written. the probable development and use of the North Arabic script in the Hejaz rests not on any one isolated factor but on several. as Jeffery seems to think. therefore. which were fully explored and clearly stated. Thus. characterizes this tradition as a legend. Nadim had ample opportunity to illustrate his point effectively. and Nadim is no exception. dealing with the probability of the simultaneous development of the North Arabic script both in Iraq and in the Hejaz.78 ABBOTT: ARABIC PALEOGRAPHY not Jeffery forgetting the man's calling of a warrak and a bibliophile? Who more than such a professional as Nadim would have the need and the opportunity to be well acquainted with old Arabic manuscripts and old writing customs and scripts? It is true that even professionals. for instance. Thamudic. follow the Meccan verse count. I am unaware of such a tradition in any objects to the term "North Arabic script" on the grounds that it may someday be seen to include the Libyanan. The distinctive characteristics of the script were definitely associated with the alif. is conditioned on that of the alif. but goes back to the time of Kusayy.. This. and to realize that the conclusion there arrived at. There were the provinces of Egypt and Syria with no particular scripts associated with them. and present the Meccan version of the text. Here. and claims that I have taken it seriously. to see how far from seriously I have taken "Muhammad's family" in this regard. And this use does not begin with 'Abd al-Muttalib. A major difficulty with Jeffery is that throughout his lengthy review he has been unable to view this problem of Koranic scripts on its own merits and apart from the parallel problem of Koranic texts. for the simple reason that neither scripts nor minor variations of the text were limited to the narrow location of their origin. with its numerousalif's and lIm's. Jeffery not only misreads and misrepresents the Arabic sources. in the Meccan script would necessarily have to be written always at Mecca. What numerous Arabic sources do associate with several members of Muhammad's family is the "use" of Arabic writing. 8-I 2). Cit.

Further studies undertaken by me since the recent publication of The Rise of the North Arabic Script have brought no new materials to light that call for any change in the major conclusions arrived at in that work. 3940.23 in connection with its No. 28-30. I728. with T.which is not unlikely. that is. and rags. therefore. leaves.Survey. . andPls. then the Koranic muhakkak may have been a somewhat more cursive script than that suggested in The Rise of the North Arabic Script. p. I am still uncertain as to what the early Koranic muhakkak script was. III.REVIEW OF A SURVEY OF PERSIAN ART 79 such scripts as appealed to them. a possibility that here one has a clue to Nadim's Koranic script. yet the script of these specimens. these seem to be closely allied to the thuluth. I921). Such materials as these would be very impractical for heavy large scripts." I find that Ibn Durusttiya24used the term tarsif in connection with letters in opposition to the term tafrlk. comparatively cursive. The letters. separation of or extension of space between the letters. rasif or murasif. Pretzl [Leipzig. and if the few non-Koranic specimens given by Minovi22 are reliable. and with Arabic Palaeography. It will be seen that though the letters are written very close together with a minimum of ligature space. is worth offering. G. one is told. parchment. in the large scripts on uniform materials before the fourth decade of Islam. I772. that is. varieties in this field. can be said to be closely. and that the specimens under discussion may indeed represent that script. II. if not most of them. It is readily to be seen by now that due largely to the inaccurate and inadequate treatment of the Arabic sources. Kit&b al-kuttab. In connectionwith building this process is sometimes referred to as tarsif from rasafa "to join well and firmly together. Cheikho (Beyrouth. 1732. particularly early in his career. ed. much as Cf. and evenly put together. One can hardly begin to look for complete or nearly complete Korans. I909-36]). wood. or for even the larger sections. The relationship of angular and cursive Koranic scripts calls for some comment at this point. II. especially of the first two. Noldeke's Geschichte des Qordns (ed. the choice in each case would depend largely on the owner's or calligrapher'spreference. These materials included stones. firmly. were for the most part poor and needy. i5. 65. and papyrus associated with the writing of the revelations were not likely to be available to Muhammad and his scribes in large quantities. There is. Even the leather. therefore. were written. and Muhammadand his community. P1. shoulder blades. I think. yet each letter has its distinct space and identity. Berstrasser and 0. L. the suggestion. and there would be no reason whatsoever to limit their use in conjunction with different verse counts and texts. 22 the stones or bricks in a well-built wall or pavement are put together. Koranic or otherwise. But there is record on what some. seems different enough to have a name of its own. In Syria the cities of Damascus and Homs had frequently a different verse count. neither Minovi's "Outline" nor Jeffery's review can be relied on for early Arabic writing. bark. since they would provide writing space for only a few words or lines on each piece and so call for a large number of bulky materials to handle and preserve. Pls. the decade of the Othmanic edition of 24 23 Pp. with no overlapping of letters or merging of one letter into the other. Furthermore. One is not told in what script Muhammad's original revelations were written. My discussion of Koranic scripts has at no time excluded the use of some small and more or less rounded. If it is related to the later muhakkak. At any rate. there were scripts not particularly associated with any one locality. for these were always expensive. 3IIB and 725C.

in its turn.27All efforts at that time to secure photographs of Koranic specimens attributed to Ibn Mukla were unsuccessful. 232 ff. P1. "What Was the Badi' Script?" Ars Islamica. and the use of the accepted Koranic scripts. where he refers to another fifth century cursive Koran dated 499 H.) and illustrated in the Survey (V. Schroeder." pp. For use among Moslems of average means small Korans would be the natural thing to expect. The script of both is unquestionablyMeccan. The first of these is a fairly large Koran of 2I by 35 cm. 30) made brief reference to a Persian-Arabic Koran written in 308 H.D. Once this became established. This same phenomenon of the gradual entry of cursive elements into angular Koranic scripts would also account in part for some of the smaller types of Koranic scripts attributed to the fourth and fifth centuries of Islam. for public mosques.D. E. Moore. (I036 A.. 388. Figs. I have indicated elsewhere that the probabilities are that the different varieties they display had their own specific names. Harrie G. These must have been carefully executed. thus leading the way to stylized and differentiated Koranic scripts. by which time conquest. 928). pretentious and well-executed Korans became the accepted and preferred ones. P1. Under these conditions the use of the current small secular scripts for the smaller Korans would tend to have been discouraged. nevertheless. 8. and that the type in general was no doubt influenced by Ibn Mukla's script reforms. other changes working for varying degrees of cursiveness were bound to follow. more by accident than by definite search. it is but natural that for the greater part. i and 5 in The Rise of the North Arabic Script. I cannot produce chapter and verse from the sources to vouch for these developments. I.D.. the Koran. op.8o ABBOTT: ARABIC PALEOGRAPHY 25 They seem to be described as "Kufic-Neskhi. the second a small one of 7 by I7. 26 Cf. They are from the collection of the late Mr. for the very act of writing the sacred text was with most an act of pious worship deserving of a reward in heaven. and for private vanity with the rich. small scripts that have been frequently . No. their scripts.. (IIo5-6 A.a common cursive tendency that makes them not far removed from this early Koranic neskhi which. for the differences in the Meccan scripts of Nos. 4 and 5. betrays here and there a tendency to stiffness verging in instances on angularity. who up to his death in was lecturer for the Canadian Pacific I939 steamship line and an amateur collector of ori25 See.5 cm. (920 A. cit. 27"The Contribution of Ibn Muklah to the North Arabic Script. expansion. i and 2) have come into my hands. and its resultant new wealth and ready availability of materials provided to a considerable extent both the motive and the means for the execution of large Korans in large scripts. The Rise of the North Arabic Script. Ahmad Mousa (Zur Geschichte der islamischent Buchmalerei in Aegypten [Cairo. 46. But large or small they would be carefully written. show. Once this change in size was admitted. Moritz. 70 ff. for show pieces. And when as a group they are compared with the earliest known genuine neskhi cursive script. Figure i in the present article. Since then. No. I93I]. some of which may be among Nadim's list of Koranic scripts. but these now reduced in size. 45b." used in these centuries side by side with other predominantly angular types. IV (I937). both simple and ornamental. would answer both a pious and an artistic desire.). though by no means identical even among themselves.) and attributed to Ibn Mukla. for instance. In connection with these so-called Kuficneskhi scripts. two such photographs (Figs. Arabic Palaeography. p. but gives no reference to any illustration of the same. namely that of a Koran26in the British Museum dated 427 H. but it seems more than likely that such a development would account for instance. yet the script of the latter displays several unmistakable cursive tendencies.

4t' rjqt~ Nk .r Ov - p~~~~~~~~~~~~~~~~~~~~~o * (~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. r -Sn ~~~~~~~~~~~~Al.4 __~~~~~~~~~~~~~~~~~~~~~I Mt l A~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~ I I~~~~~~~~~~~I I *zK~~~~~~~~~~~d 4~~~~~~~~~4 '4 ~ ~ ~ ~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~".>---~~3&i 5 tj _u- AFE_ROMN FI ISLA CIT RTCL_ET 4-7H 866 .

-As stated before. Accepted as such they would readily solve the supposed problem frequently referred to.REVIEW OF A SURVEY OF PERSIAN ART 8i ental manuscripts. the few known specimens. Mrs.Op. This would indicate that to his mind and eye the scripts of these Korans are similar to those of other Korans. India. Anderson of Peoria. 337. [870 A. but seem to be of comparatively poor execution in the original. this copy could well have been written in the late third or early fourth century of Islam. It is difficult to speak as definitely of the marginal decorations which are not only poorly reproduced in the photographs. For the present purpose it is enough to know that both angular and composite scripts continued in restricted use long after the date of the earliest known genuine Neskhi Koran. Moritz. including master calligraphers.D. Illinois. Cf.cit. of a gap of a century or more.. the older predominantly angular scripts of the first four centuries of Islam to the predominantly cursive scripts of the fifth and succeeding centuries. and under present war conditions. Representing a general script type flourishing in the fourth and fifth centuries. But again it was pointed out that calligraphic forgeries must per force seek a reasonably faithful reproductionof what they seek to forge. So far as scripts go. attributed by him to the third century and illustrated in his Arabic Palaeography. but these were on the whole considerably cruder in execution. The binding is late. India. the work itself can hardly be earlier than the latter part of the second century. among other things. I. Unfortunately. . between the angular and t. on what grounds the copy is attributed to Ibn Mukla. p. nowhere illustrated. Plates 37-42. Cursive scripts made their way early into these protocol texts. both of which represent the same notations give copy of the Koran. and 366 H. Dan S. Mr. 388. Without further tangible information. Kashmir State. it is difficult to correspondwith this Mr. 256 H. 29 The Koranic verses used in the protocol formulae were writtenfor the earliesttimes in largescriptshaving much in common with that of some extant Koranic specimens. One cannot and need not press the point of whether the wakf dates actually represent the date of the writing of these Korans or not. AMoore's the informationthat the Koran was at "Rampur. Tait of the Indian State Railways. if it is indeed that 28 Cf." and that the specimens were photographedby a Mr. one must be reasonably suspicious of manuscriptsattributed to famous persons. if possible. as far as I know.he cursive and the equally supposed sudden emergence of the lat- ter. in Koranic usage. The Rise of the North ArabicScript. Tait and through him gain more definite information as to the whereabouts of this Koran so as to ascertain. being encouraged in this assumption by the fact that despite the fourth-century wakf date Moritz himself placed the actual writing of these Korans in the third century. which numbersare seen to be of the angular type. seem to be. Until these grounds are known and tested it is impossible either to accept or deny that Ibn Mukla is the calligrapher of the copy. for permission to use these photographs. for three such Korans dated 300. I9 and references there cited. these composite specimens in which angular and cursive elements appear side by side link. Delhi. and the script of this particular copy can hardly be any later than the early fourth century. with wakf sheets dated in the fourth centuries.29 The earliest known and illustrated specimen of this kind is now believed to be from the genealogical work of Abfu 'Abd Alldh alZubair ibn Bakkar (d. It is of significant interest that the few rare manuscript specimens of angular and composite scripts used in non-Koranic fields seem to tell a parallel story so far as these scripts are concerned.]).28 Since their script has drawn no special comment from Moritz one may assume them to be of the predominantlyangular type. I am obligated to his daughter.

32 The next is a copy of Hasan ibn 'Abd Allah al-Sirafl's biographies of grammarians copied by Ali ibn Shadhan and dated 376 H. morgenl.D. Ahlwardt. I. cf. 427. VIII-IX). Nos. Oriental Series (London. Brockelmann. Margoliouth (E. however. judging by the script.82 ABBOTT: ARABIC PALEOGRAPHY late.31 I think. 32 An interesting instance of the use of "large scripts" in non-Koranic mianuscripts is associated with Tabarl's (autograph?) copy of an eighteen-volume treatise dealing with the various readings of the Koran. in which event it would have to be a copy made in the author's early life." Zeitschr. I887-99). Kitfb al-Abniya of Abul Mansfir Muwaffak ibn Ali al-Harawl written by Ali ibn Ahmadal-Tisi. 33 Abu Sa'ld al-Hasan ibn 'Abd Allah As-Sirafi.. such as the use of wavy alifs and the marked inclination toward the simpler ornamental varieties. Bi- ographiesdes grammariens de l'ecole de Basra (Paris and Beyrouth.).).Irshdd. IVI. (986 A. Verzeichnissder arabischenHandschriftender koniglichen Bibliothekzu Berlin (Berlin. it is illustrated in Wright's Facsimiles (Pls. . Lectures on Arabic Historians (Calcutta. ed. and fourth centuries. Gibb Memorial Series [Leyden. since at his death he was aged eightyfour years. I.D. Abbott. I930). It is illustrated in Wright's Facsimiles30 (P1. both types flourishing mainly in the fourth. 31 C. I02 f. and the last one is further distinguished by being written in the Persian language. It is not known whether these large scripts were of the secular or Koranic variety.) in a 'Abd Allah al-IBalkhI stiff composite script. in 447 H. also W. but considering the nature of the work. 63 ff.34 The last and seemingly latest manuscript 30Facsimilesof Manuscriptsand Inscriptions. and the other in the direction of a composite script. Yaikuft. P1. Koranic scripts would not seem to be quite improbable. (I055-56 A. "Die jiingste ambrosianische Sammlung arabischer Handschriften. and D. XIX) and attributed by him to the latter half of the second century. Finally. the cursive type won the field beginning around the 3 Cf. in a composite script that is more cursive than angular. The manuscript was recently edited by Krenkow33with three plates that illustrate the use of several varieties of both the angular and the composite angular-cursive types side by side. Margoliouth. d. these seem to have had their own smaller and comparatively rounder types used mainly for small private copies. These latter scripts have other eclectic tendencies. deutsch. This is certainly very strong evidence that the Persians took the lead in the double innovation of using Koranic scripts for non-Koranic manuscripts and in evolving and using a composite angular-cursivetype in the Koranic field itself. third.35 In recapitulating the findings made herein in the matter of Koranic scripts one sees a progression of natural and readily understood development that can be summarizedin broad and general terms somewhat as follows: First there were the two major but simple angular types. III. 1907-27]). S. 367.D. I936). W. VI. and sixth centuries. but the other three are specimens from the pens of Persians.Iakk Yahya ibn al-Husain written by 'Isa ibn in 4I8 H. I898-I902). The third manuscript in this class is a copy of Kitdb al-akkc7n of the Imam al-Hadi MllalI."pp. that one is safest in attributing it to the third century. I875-83). cf. Next came a dual development. the Kufan-Basran and the Meccan-Medinan. Geschichte Literader arabischen tur (Weimar. and referencesthere cited. current largely in the first three centuries of Islam. 368. D. S. "The Contribution of Ibn Muklahto the North-Arabic Script. fifth. Griffini. which is an old simple angular Kufic. one in the direction of ornamental angular scripts. I do not know when and by whom the first of these four was written. J. Then came a number of new types closely related to the first two and current largely in the second. it is illustrated by Griffini. of this class that I know of is a copy in Persian of the medical work. and 3o0. LXIX (I9I5).Gesellsch. pp. 79 f. 34 E. (I027 A.

the idea of Koranic scripts being sunna would tend to keep the older and simpler forms in practice.). . afforded the artistic Persian innovators the opportunity to lead the way with neskhi Korans. (I030 A. especially for the change from angular to cursive scripts. became established as a sunna or sacred practice.. This. the Koran was set apart in its scripts. Thereafter. This sketch of the development of Koranic scripts may appear too orderly and too logical. giving us the neskhi and thuluth Korans of later times. Again. it may even alter by a quarter or half century some of the initial and terminal points suggested for the major progressions. 88. (I036 A. Flury. and I77I.36and with the unique position of the Koran in the manuscript field. I have encounteredthe same obstacles to rapid progress that were met with in the Koranic scripts. cf. _53. Treated from early times as a sacred and mionumentalwork.REVIEW OF A SURVEY OF PERSIAN ART 83 fifth century. cit. The question of whether religious and political motives played a part in these progressions must be linked with the motives behind the practices in the monumental or epigraphic field. 1770 f. The early use of angular scripts similar to the epigraphicones for the Koran suggests the continued use of such scripts and their newer developments. the literary sources on the development of writing in the secular fields. and the cursive scripts of the neskhi and thuluth varieties won out in the Koranic field of the succeeding centuries. in a measure. as cursive scripts made their way in the fourth and fifth centuries into the epigraphic field in the eastern provinces of the caliphate under the Samanids and the Ghaznawids.). and IV. the continued advance of the cursive into the epigraphic field together with the supremacy of the Shiite Bfuyids. This is true despite the fact that there are more literary 38Cf.37 There are here then two tendencies thatin some respects are in conflict.. for instance. seems to have further weakened this idea. op. On the other hand. (I055-58 447-50 A. Further research may cause some modification of its details. III.D. 255.D. for calligraphic scripts and usages neither appear nor disappear overnight. enable one to speak in terms of periods of time whose unit is less than the century." Syria. as one learns from Ibn Durustfiya and others. The Turks now followed the Persian lead. already referred to above. Finally. "Le Decor epigraphique des monuments de Ghazna. but the conflict does not seem to have been an open or major one.D. pp. illustrating cursive epigraphicscripts of H. XXIV for other Ghaznawid cursiveinscriptions of 42I H. Each major progression is accompanied by a comparatively long period of overlapping.. cit. 3 20-24. p. based largely on what seem minor or trivial characteristics. pp. 37 Kitdb al-kuttdb. op.) the earliest known dated neskhi Koran. and P1. such as practicability and natural preference. Pls. including spelling rules and scripts. S.38 the idea of the Koranic scripts being sunna seems to have weakened. At any rate it is Persia that produced in 427 H. VI (I925). i8o6 ff. in too many cases obstruct the way to definite identification. Early differentiation and classification of an impressive list of scripts. with other factors. cit.. Suli. op. Kratchkovskaya. This 36 1 am in agreement with Mayer and Kratchkovskaya that one need not look for primarily religious or political reason for changes in this field. II.. and a compromise script-the composite angular-cursiveKoranic types already discussed-came to the fore. In attempting to follow. may be one reason why no complete or large sections of the Koran are known to be written in the more elaborate varieties of the ornamental angular scripts used sometimes for the sura titles. 4 f. 1Iadidj! Khalifa. but it is nevertheless a picture whose essential outline has been drawn from the sources and the specimens available. This may. its writing. Survey.

op. These scripts. and at least as many extant specimens.Beitriige zur Geschichteder Staatskanzlei im islamischen Agypten(Hamburg.I926).pp. II. I have elsewhere endeavored to show that these first-century scripts were closely associated with the Meccan-Medinan and KuficBasran practices. Two others. Ibn 'Abd Rabbihi records that two of Muhammad's own secretaries. where incidental. serving both men and women. W. Arabic secretarial art.41 whose talents must have been put to work among their neighbors. 204. both foreign and imperial.84 ABBOTT: ARABIC PALEOGRAPHY sources than for the Koranic field. however. 28 f.43for the initiative in Islamic Arabic writing in this first century was largely if not entirely in the hands of the Arabs. Kratchkovsky and V. cit. The first four "rightly guided" caliphs and Mu'awiya had all served as scribes in their earlier careers. and when professional. 39I. 52-go and I20 f. both dated and undated. ed. 41 Futuikal-bulddn. Walid ibn 'Abd al-Malik (86-96 H. entirely official and governmental.Kurrah PapyrifromAphrodito. who had his many secretaries. wrote for the people of the tribes at their watering places. p. Neither was this demand. Mugj. Egypt.lra ibn Shu'ba and Husain ibn Numair. and for the Ansar in their homes. 4I. in both its western and eastern provinces. but one that was to grow more complex with 39Cf.. with the turn of the century. for when in the reign of 'Abd al-Malik (6586 H. 43 The . marked rapidity in the few succeeding decades. were the first to become current in the Umayyad empire. the services of the private secretary and the public scribe were in continual and increasing demand among the Arabs. came into its own. for. 43. Zaid ibn Arkam and al-Ula ibn'Ukba. 40'Ikdal-farid (Cairo.42 Scripts used for royal correspondence. VI.40 Baladhurl gives an interesting list of women who could read and write. The Rise of the North ArabicScript.). Unfortunately. and for local tax notifications and requisitions. Mcik (Leipzig. The office of the public secretary was recognized as an important and honorable one. and with it Arabic writing. 44 Cf. I. p. One comes to suspect seriously that a goodly number of extant specimens represent not a specific standard or muhakkakscript. p. thus ousting in time the other languages of the empire. The specimens on the other hand are so varied and in many instances so poorly executed that to look to them for the workable and successful clue to an identificationis to hunt for a needle in a haystack. has provided numerous documents. pp. I293). lZalkashandi. even at the start. take too much for granted on the part of the present-day nonprofessional reader.acted as scribes for the people.) Arabic was made the official language of all governmentaltransactions in all of the provinces. dealing with its provincial administration in the first century.Kitdb al-Wuzard wa al-Kuttdb. were doubtless considered the best. Under his successor. The sources. I928). In these early decades of Islam the Arabic secretarial art or profession must have been as yet a simple one. therefore. 472. These documents indicate that different scripts varying in size and degree of angularity were in use for protocols. for the governor's correspondence. fine papyri and fine large scripts received special attention. J." Sogduski Sbornik. 33-39. The largest and most angular were used for the first.1934. scripts from Egypt on the one hand and from Khurasan44on the other are of these same types. ed. pp. Kratchkovskaya. i866). . but a general nondescriptcommon or mutlak writing. Bjorkman. DeGoeje (Leyden. A. to which class most of us belong. are too fragmentary. the smallest and most cursive for the last. 42 Ibn 'Abdiusal-Djahshiyari. "Le Plus ancien document arabe de l'Asie Centrale. imperial documents of these types have not survived for inspection. cf.39 From the time of Muhammad.

Proficiency in the art of penmanshipwas one of 4S Amer. or how to make and preserve ink.46 It was around Salih that the first Arabic school of secretarial science grew. who were known for their good writing and for their specific knowledge of penmanship. 46Ibid. This situation proved both discouraging and puzzling. that have come my way. perhaps because of its wider application.Journ. but the art itself. I90 H. I20.pp." shown the major role played by Salih in the transfer of the dcwdn al-khardd from Persian to Arabic. many requirements in the qualification of a scribe. particularly in Iraq. [805 A. on examination. more on how to sharpen a pen.48 Mufaddal ibn Salma. Here. Cit. was important enough to receive a separate treatment all its own. I-62). A working list of these latter. This is true despite the fact that lists of names of scripts frequently occur in these works. Sprengling. . The Iraqian school continued to lead in this field. This early type of monograph. and practicing secretaries who became known for the has attempted mastery of their art. They dwell.Semitic Lang. than on what specific features constitute the essential characteristics of a given script. This list. $alih ibn 'Abd alRahman. 325-36 (reprinted. for instance. I offer the titles from the Fihrist. The second type refers to specific treatises or monographson scripts.. until its explanation was forthcoming. 22I. a young Moslem of Persian origin. The first refers to men of the late Umayyad to be distinguished from monographs or books written on adab al-kuttab or san'at al-kitaba. and includesmost of the leadingones. including knowledge of the names and the distinguishing characteristics of the various scripts must have been acquired in the early and apprentice days of the future scribe. As a type. 175-224. 7-I6..D. for the life and careerof this remarkable man. This explanation suggested itself on the strength of two types of facts. as already stated. thoughby no means complete. 'Ikd al-farid. I9I-208.REVIEW OF A SURVEY OF PERSIAN ART In the meantime the secretarial sciences were making further headway.consists neverthelessof typical works of this type. pp. both of whom are associated with tracts or monographson the secretarial art and the training of secretaries. which deal with scripts either alone or in combinationwith some other professional art or branch of knowledge: Muhammad ibn al-Laith al-Khatib. producing in the succeeding generations prominent men who wrote more fully on the profession. they are more inclusive than works represented by the first two titles. There are some titles from the earlier centuries that indicate a treatment of both the secretarial sciences in general and the art of penmanship in particular. II. associated with Fath 48Fihrist.. The casual and frequently too fragmentary nature of the references to the different scripts can therefore now be understood. thanks to the farseeing policy of Ziyad ibn Abihi and Hajjadi ibn Yfisuf. some of them make reference to and others give indication of brief citations from the monographson scripts. was laying the foundation of his secretarial career. 47 op.pp. secretary to Yahya ibn Khalid the Barmicide (d. and Lit. The fundamentals of the art. LVI (I939). has been drawn up by Bj6rkman.]) wrotea bookcalledKitdbalkhatt wa' a recent and bril45 has liant study. Bjorkman47 a chronological list of authors and their work along this line. Among his pupils of the next generation were men of the caliber of Ibn al-Mukaffa' and 'Abd alHamid ibn Yahya.referred to usually as kitab al-khatt or kitab al-khatt wa'l-kalam. cf. to be greatly disappointingin their meager information on the art of penmanship itself. Without pretending to anything like an exhaustive list. p. "From Persian to Arabic. Some of the earliest sources he mentions which are still available prove.

277 H. Had. p. Cit. Biographies of outstanding secretaries were being written in the third century of Islam.. I58. Mif- Gift of the Lover"). The title here may be redundant. though its title seems to be nowhere recorded. Khalifa.. already referred to. to judge by references to titles in the works of later Arabic authors such as Muhammad ibn 'Abd al-Rabman in his work. [907-32 A.jl tdl al-sa'ada.5' Muhammadibn Hubaira al-Asdi. III. tutor and later secretary of Caliph Muktadir (295-320 H. a:ndreferences there cited. This was to be expected. had a work en56Hadjdji Khalifa. cit. cit. 52Ibid.D. p. cit. one comes across some brief extract from a few of these earliest materials. lists a number of these subdivisions as different 'irms. the group of books.57 In that lies one of the major difficulties. here and there. held itls own in the Arabic field and 49 Fihrist.. p.49 Muhammad ibn Yazid al-Mubarrad (207the well-known gram85 H. op.50 Ahmad ibn Muhammad ibn Thawaba (d.. and very likely did. One deals with the rules governing the forms of the letters. cf. 55Cf.. implying perhaps his devotion to his art and profession.) had a work entitled Kitdb al-kitaba wa'l-khatt. associated with the BanfuFurM. quoting Abu 'I-Khair Mustafa Tashkuprlzada (9OI-68 H.) was a good penman and wrote Kitdb al-khatt wa'l-kalam. I58.]) marian wrote KitJb al-khatt wa'l-kihjd. Bjorkman. [907-32 A. 74. op. therefore. p. [822-98 A. op. 57 Op. Within the field of calligraphy itself. One is. p. Hadidj Khalifa. 34.]). III. Kitdb al-lum'at 55 Kalkashandi and Haddji1 Ji 'ilm al-khatjl Khallfa. 51 Fihrist. IIadjijl Khallfa reports. had a Risdla fJ'l-khatt.. another with their excellent execution. cf. wrote on Al-khatt wa'l-kitdba. friend of Ibn al-Mu'tazz and secretary of the caliph Aluktadir (295-320 H.. I30. The Rise of the North Arabic Script. 50Ibid. . include one on scripts. The next source of relevant materials is the wide field of Islamic biographical literature. for Abii'l Op. III. 53Ibid. p. Brockelmann. in which case the emphasis is on the spelling. cit.56 Extant works dealing with these and other phases of writing. that deals with the secretarial profession in general.]).]) but entitled his work Tuhafat al-wdmik53 ("The made its way into the Persian and Turkish literatures. i 5o. were extremely rare even in his own day. pp. p. IJadjdj Khallfa. 247 H.86 ABBOTT: ARABIC PALEOGRAPHY ibn Khak1n (d.) undoubtedly wrote a book on scripts. 9.D. 73 and ii6. Ibn Mukla I'd.52 Ishak ibn 'Ibrahim al-Tamimi. IIL425 f. This interest in scripts and calligraphy. 296 H). was a scribe who wrote a series of monographs on every art of his trade or profession. 13..4.D. Khair. forced to approach the study of the developments of scripts and calligraphy in the early and formative centuries of Islam without the aid of such studies and monographs as were written on the subject in those very centuries themselves. as is to be readily seen from Kalkashandi's exhaustive treatment of the scripts of his day.. under the circumstances. subdivisions of the art were accorded separate treatment. p. a man devoted to the one-day caliph.D. The most usable source is. cit. Nadim records that Ahmad ibn al-Khasib. since calligraphy itself held an honored position in every field of writing in all Islamic lands. 'Abd Allah ibn alMu'tazz (d. 59. 6i. except as. and still another with the derivation of new letters from the old by the addition or subtractionof old elements or by the introduction of some new elements of change.328 H. 54Filhrist. III. I52 f.. Sa'ld ibn ILbrThim al-Tustarl. 58 of this article for Ahmad ibn al-Khasib. [I495-I560 A.54 which should.

65Ibn Khallikan states that he learned kitaba from Salim. Nadim credits him with a work Diwdn rasd'il of some one thousand folios covering every phase of correspondence. cf. Bj6rkman. Still. indeed. is difficult to say. I believe one can get a fairly accurate picture of the main paleographic andicalligraphic developments in early Islam.REVIEW OF A SURVEY OF PERSIAN ART 87 titled Tabakdtal-kuttdb. Bibliotheca Geographorum Arabicorum [Leyden. In his brief letter of instruction to the secretaries he urged on them. 66 Op. cit. p.. the author receiving his information orally. 65Kurd 'All. I913).. and Ibn Mukla. cit. xxii). with some slight variation. Still later.. cf. including Mu'awiya. Sprengling. p. however.) one of Sdlih's pupils and secretary to the caliphs Sulaiman and Marwan II. cit. cit..).62 Whether Mu'awiya deserved this credit any more than Ali. Rasd'ilal-BulaghJ' (Cairo. p. p. 58 and n. I... the mawla of Hisham ibn 'Abd al-Malik. I74. IOI-5 H.64 there are a few incidents recorded that would seem to indicate his interest in penmanshipas such.). literally dozens of times in Sfili's.. Ibn 'Abdfus. Whether this included a separate section on Adab al-kuttdb is not stated. Survey. 7. II. K~udamaibn Dja'far. "Muhammad ibn 'Abd al-Rahman on Calligraphy. I908) is drawn from this class of literature. I suffer more keenly from the nonavailability of all known early sources." Studia Semitica et Orientalia (Glasgow. 7I via Habib.-iscarded. I920). p.. but with biographicaltreatment that is disappointingly short and inadequate. Such works. 'Ikd. i. Kitalb al-kharddi wa-san'at al-kutt&b (cf. op. 20."but I believe "secretarial art" is what is meant here. op. . I7IO. cit. Bjdrkman. op. 64 Cf. 30. 8. I97.61 The tradition must be. 6-i6. Bj6rkman. p. Ibn Khordadhbeh. Again though $alih ibn 'Abd al-Rahman was the leading secretarial figure in the last quarter of the first century (he was put to death under I have so far come Yazid II. Kitdb al-masalik wal-mamdlik. I889). 63 Cf. 62 Cf. across no reference to his penmanship in particular. however. which in a few instances yield very valuable information. as are available to me are generally conceded to be typical of their kind. 6i. I75. there is no denying the fact that around him grew the first secretarial school in Arabic. op. Ibn 'Abduis. Kurd 'All. 69 f.59 While this situation regarding source materials for the earliest centuries is. Abui Hilul al-'Askarl. where an edition and translation of this work are promised). and that his pupils and successors readilyconceded his superiority. op. Huart's information for his Les Calligraphers et les miniaturistes de l'orient musulman (Paris. Kitdb al-kuttib (cf. among other things. I889]. first in Islam to give instruction on writing. one that calls for much labor with the promise of comparatively small results. VI. cit. since the particular instruction. pp. p. Persian and Turkish literature produced works of short biographical notices of famous and outstanding calligraphers. p. E. Adab al-kuttdb. Bj6rkman. Huart. cit. 'Abd al-Hamid ibn Yahya. II. Robertson (tr. 36.. Khatt wa-khattdtdn (Constantinople. pp.60 Popular tradition credits Ali with being the 58Filrist.. proficiency in penmanship. 59Most of C.63 With 'Abd al-Hamid ibn Yahya (d. 4 and 58. and Minovi. though pressed for time. frequently present a list of names that is surprisingly long. it is nevertheless one that is not as hopeless as it may at first seem. nor to any reference indicating that he wrote on the subject as such. 60I regret particularly the nonavailability of Abd Allah ibn 'Abd al-'Azlz al-Baghdadi. however. and n.51 This and similar early works are either lost or have been incorporated in later Arabic biographical dictionaries. Op. cit. Given time and all the known sources. pp. cit. Cit.op. He is cited. II.66 In another instance 'Abd al-Hamid gave instruction as to pen sharpening 61 Cf. 22I. These.. op. I24. is credited to several others.. 72.. De Slane renders kitaba as "penmanship. I32 H.. op. For the present. I902). p. p. Kitdb al-sina'atain al-kitdba wa al-shi'r (Constantinople..

From this period of the third century have come most of the earliest sources and the first series of lists of scripts.D. The first of these is to be found in al-Baghdadi's Kitdb al-kuttdb. among them the nisf. 7 f. 76 See p. in time.. p. p. 175. cit. 94 f. and djawanih . It is well to stop here to take note of the lists of scripts that have come down from about this time. Fihrist. created the demand for more and better penmen. 33 f. I 54H. But even if 'Abd al-Hamid himself did not devote his talents to writing on penmanship as such. d. 7. p. For. 69Fihrist. op. 8. . 87. cit.) Hamid.88 ABBOTT: ARABIC PALEOGRAPHY so as to produce good penmanship. Huart. op.. in the period beginning with Slih ibn 'Abd al-Rahman and ending with al-A.) it was Kutba in part contemporary of 'Abd al(d. kisas. and references there cited for this and the following statements. 7 f. p. Op. So far. Bjorkman. 71 Fihrist. according to Ibn Thawaba (d. Risalat al-'Adhrd'. 77 See p. op. "Arabic Papyri of the Reign of xa'far al-Mutawakkil 'ala4lh.7' Some of his pupils in turn surpassedhim in carrying on the tradition of good penmanship. pp.. the march of events which made a secretarial school such as his possible. 7 f. 78 Kurd'All. op. ghub&ral-hilya. pp. i8o. yet added something to the art. the golden age of Islam. Abbott... [825 A. Thereafter. 92 (1938). Nadim lists fifteen of Ishak's pupils including men of many nationalities and one woman.. I76-93.. 75 The Rise of the North ArabicScript. op.. n. 68A. the last of which coincided with the reign of Ma'mfin. aDahhak ibn 'Adiilan (d. cit. cit. secretary to the caliphs Mansfir and Mahdi (I36-58-69 H. 8. cit.75 Thus.p..76 The second and largest is an extract from Ibn Thawaba preserved by Nadim. Chief among these were the brothers Ibrahim al-Sidizi (d. the art of writing and penmanship shows a rapid and progressive advance. 7372The Rise of the North Arabic Script. op.p. pp. Pihan. p. morgenl.. The Rise of the North Arabic Script. I856'). op. 33 and n.. Bj6rkman. and for some regulation on penmanship. cf. cit. p.] ).72 The first came. 3 I. 86 and Bj6rkman. to be 67Ibn 'Abduis. The momentum gained in this period kept the progress up for the next half century or so. 277 H. .67Pihan in an undocumentedessay68 creditshim with perfecting the Kufic in the time of the Umayyads. cit.. 6o and Bj6rkman. pp. confusion and neglect seem to have gained sway until Ibn Mukla came to the rescue of Arabic scripts.69 Kutba's conH.des persanset des turcs (Paris. 200 H. p. ) temporary. I secretary of Saffah.secreand contemporaryof Yuisuf. Gesellschaft.p. 59.]) and Yuisuf (d. 82. mu'amarat.70 With Ishak ibn Hamad. cit. p. 59. pp.73 Yuisuf in his turn became associated with the invention of the riyasi credited by Nadim to Fadl Dhui'l-Riyasatain. cit. I have come across no such statement in the early sources. who is said to have been the first Arabic penman and to have differentiated the scripts and classified the leading ones. ii6.D.. p." Zeitschr. The sources present several lists from about the middle of the third to the middle of the fourth century. 72. Ibn Khallikan. cit. Fihrist.. For Ibrahim ibn al-Mudabbir see N. Notice sur les diverse genres d'Jcritures ancienneet moderne des arabes.Cit.hwalcovering about a century and a half. op. 72. 70 Fihrist. Robertson. I83 f. graded and proportionedscripts. 7. Bj6rkman. especially pp. op. P. 74 Robertson. II. 73.78 73 Robertson. [8I5 A. p. op. thuluth khafif musalsal. all of whom wrote the genuine and well-balanced. already credited by some to Kutba. deutsch. 9. cit.. p.. I77. that is.74 Ibrahim has the further distinction of being the master who came to be credited later with the classification and invention of numerous scripts. tary of Ma'muin. 5. associated with the invention of both the thuluth and the thuluthain scripts." A third comes from IbrThim ibn al-Mudabbir. p. one comes to what might be termed definitely the first organized school of penmanship in Islam. op. though second in penmanship to Kutba.). 2 I0 H.

TABLE I LISTS OF SCRIPTS FROM THE THIRD AND FOURTH CENTURIES OF THE HIJRA Baghdaddi (active in 255-56 H. 277 H.i'at Nardiis Futfihat Musalsal Nardjjis Musalsal Ghubar Mudammad Hawa'iDi Muhaddab Mashk Ghubar Mudammadj Lazaward Muwashsha'| Muwalla' ? . TWELVE PRIMARYSCRIPTS80 Dijalil Dijalil Sidiillat Sid jillat Sid jillat Sidjillat Tfirmr Thuluthain Dibadd Tumar Thuluthain Zanbfir Haram Mu'amarat 'Uhfid Nisf (al-mumsak) Kisas Adiwiba Mudabbadi Kutub al-mulfik or Djalil?) (Tfumar Dibadj Thmar Thuluthain Mu'amarat Nisf Kisas 'Uhifid Amnanat Kisas Mu'amarat 'Uhud Amanat I B. 279 H.) A.)79 Nadim (wrote 377 H. TWELVE DERIVED SCRIPTS Sam!'! Ashriya Khafif al-nisf Khirf 4d Khirfadikhafif Mufattah (al-nisf) Khafif al-nisf Riyasi (mudawwar al-kabir) (Futuhat?) al-nisf Saghj.) Ibn Thawaba ( Riyasi (al-kabir) Nisf riyasi Thuluth Khafif al-thuluth Rika' Nardjis Tliuluth Khafif al-thuluth MudawwarsagLhr (Dmi') (Missing thuluth?) Khafif al-thuluth Rik' (used for tawkil'at) Tawk.) JIbn al-Mudabbir (d.

muhiiknmmt. cit. For lists from later times cf.. cit. loc. also Habib.I IIUIULldlall IIUIULIIllldIIJ 1L Zanbuir| Haram Mu'amarat Nisf Kisas Mu'amarat 'Uhfid Nisf (al-mumsak) Kisas Adjwiba 'Uhifid Amnanat Kisas Mu'amarat 'Uhud Amanat I B. murasalit. op. 8. p. iOi f. apparently quoting the Fihrist. cit. loc.. op. Robertson. for further differences in this list. see pp.. .. 20. 7I and 63 f. cit. op. e. I50. cit. footnote 3." meaning "reed pen" for its other meaning of "script" in the case of kalam al-mnurabba' and kalam al-muharrafal-Kiifi. op. cf. Habib. nawd'ib.. cit. Bjbrkman. III. has mufattah in the place of nisf and khirfi&j (his hirfdda) for a4jwiba.i'at Nardiis Futfihat Musalsal Nardjjis Musalsal Gubar Mudammad Hawa'i0i Muhaddab Mashk Ghubar Mudammadj Lazaward Muwashsha' Muwalla' ? Musahham Munamnam J Wash! Murassa' Nussakh (Neskhi) Muhakkak Manthuir Mukatabat Biyad 79Only definite names of scripts are included. TWELVE DERIVED SCRIPTS SamN' Ashriya Khafif al-nisf Khirf 4d Khirfadikhafif Mufattah (al-nisf) Khaflf al-nisf Riyasi (mudawwar al-kablr) (Futuhat?) Saghjral-nisf Riyasi (al-kabir) Nisf riyasI Thuluth Khafif al-thuluth Rika' Nardjis Thululth Khafif al-thuluth MudawwarsagLr (fljmi') (Missing thuluth?) Khaflf al-thuluth Rik' (used for tawki'at) Tawk. HaljdQl Khalifa. documents with no script names associated with them have been left out. 80 Hadjdjl Khallfa. kutub al-ikhwan. pp.g... has misread "kalam. p.

According to the summary in the Fihrist. The following diagrammaticalrepresentation of these relationships seem to me to be the most probable.87and nisf. 8 f. op.. 8. p. cf. This would call for an emendation of Nadim's text replacing the word "thuluth" on page 8. with the "four pens" of page 8.83 Ibn Thawaba. The difficulty is that both of our primary authors. to assume that Kutba's four scripts were djall. and "father of all pens. only a hundred years after its author's death. lines 3I. 89 p.86 If the above relationship of the scripts is correct. they were diali1. II.84 Following this summary I credited him with these four in The Rise of the North Arabic Script. and thuluth. also Huart. dlbadi. timar. thuluthain and thuluth. .89 who outlived Kutba by some forty-six years. cit. p. Nahhas' text or information on the other hand comes through authors writing some five hundred years later. I7IO- See also p.. siaiillat. pp.8' The fourth and last list is given us by Nadim based on early sources "other than Ibn ThawTba"8'(Table I). cit. it will be noticed. perhaps edited or abbreviated. line I2. cit. cit. does not give a definite list of scripts. thuluthain. it must be noted.D. op.85 Minovi. are known from later sources only. op. But this too does not solve the difficulty. i6. Ibn Durustuiya (p. Nadim gives Ibn Thawaba's text. is nevertheless the longest and best organized.. I have given it much thought. P. 90 See p.REVIEW OF A SURVEY OF PERSIAN ART 89 Ibn al-Mudabbir. page 7.pp. [950 A. op. 76. 73. timar. nisf. but Ibn al-Mudabbir's list alone is used here because he is the earliest of these three authors." This in turn draws attention to the odd order of these scripts. op. 83 It will remainto be seen if Fuick's promisededition of the Fihrist will give a clearer version of this passage. the dialil itself being the primary pen. 82 Fihrist. 81 A similar procedureis followed by Mliin his Adab al-kuttdb(cf. 20 f. Bj6rkman. as the following page indicates (Table II). 88. lines 2 If. we know from the other lists. for the invention of both the thuluth and the thuluthain is associated by Abt Dia'far al-Nahhas (d. 37. certain scripts are associated. Robertson. the distinction itself being more implied than stated in clear cut terms. takes the four to be djalil. 338 H.This last may be a minor point of order which should be ignored. to avoid all these difficulties.])88 in his SinJ'at al-kuttdb with IbrThimal-Sidizi. Bj6rkman. lines 2 If.. and tfimar? These four. and have tried to read it in several possible ways in order to get at the interrelationships of the twenty-four scripts it is dealing with. Ibn Thawaba's text as preserved by Nadim is one of the most complicated pieces of unpunctuated Arabic I have come across. it is true. it would seem that the four should be djal1l. Ibn Thawaba and Nahhlas. A study of the lists shows that the list of Ibn Thawaba. but speaks of kinds of documents with some of which. 87 88 86 Survey. which now places the smaller nisf ahead of the larger thuluthain.. and his documents and scriptsshouldbe compared with these lists.. 72. does not state the principle on which the distinction is drawn between his two groups of twelve scripts each. III. though by no means free from all difficulties.op. The first concerns the four pens that Kutba is said to have developed. 85p." Tables I and II raise a number of questions. ii f. 74) has a partiallist of documents and scripts. are among the earliest primary pens. labib. with "thuluthain. cit. In a strict sense all but the djalil are "derived" scripts.). although of the same period as those of Baghdddl and Ibn al-Mudabbir. But this particular element of 84 P. cit. 92.90 Is it possible. but on the other hand I do not see how one can avoid equating the "four pens" of Fihrist. tfimr. Kalkashandi. for an unstated reason.

has. is not definitelyspecified as beingfrom the haram. however. Ibid. 94 93 Ibid. The derivationof the nisf. closer to the events recorded. does know of the secretarial school of Ishak ibn Hammad and of Yfisuf the brother of IbrThim al-Sidizi. lines I 7 f. Nisf al-mumsak93 'Uhud KhafJf (al-nisf) Vardiis [Thuluth?] Mufattak (al-nisf) Riydsi=al-mudawwar' al-kabir Al-mudawwar'al-saghgr (jami' pen) I Khaftf al-thuluth 1?ikd'(used for Tawki'at) 94 time can lead neither here nor there. lines 8 f.) or Ibrahim (d. lines 5 and I2 f. 8. and credit that belonged to the former went instead to the latter. The Roman numerals indicate the four leading p.) invented the thuluthain and the thuluth. 205 H.but since it followsafter 'uhild. 92 Fihrist. 8. it seems that one must leave open the question of whether Kutba (d. Is it possible that in the course of time the contribution of the older penmen. 92 f.. was overshadowed by the fame of their successors. p. He 90a The derivedscriptsare in 91 I italics. scripts. Ibn Thawaba died some sixty years earlier than Nah. such as Kutba and Dahhak. and 98. which is so specified.90 PALEOGRAPHY ARABIC ABBOTT: TABLE II AND DERIVED SCRIPTS90a OF PRIMARY RELATIONSHIP I. He is.95 There is. unlike that of the 'uhid.? The case of Ibn Mukla who received credit for the work of his predecessors is an instance in kind. 95 See pp.this derivationis to be preferred to one from either the mu'amaratalone or from the haramand mu'amar&t combined. Dja1i191 Dlbadd Sidiillat Ashriya Tfiumr III. Thuluthain Mu'amarat Haram Zanbfir II. it . 8. therefore.. In any case. p. another element of time. Though the time differencebetween the two is just short of half a century. but he credits him with no specific invention. I54 H. add AKirf kajif al-tumar92 SamN'i Kkir adi khaf f l l Adiwiba Kisas IV.

III. Again.the scripts in the first group)are frequently mentioned without the use of their qualifying adjectives. tfimar-kabir. uses the term 98Fihrist. 7. There is here." therefore. namely a smaller lighter variety of the original tuimar.nisf. pp.. The "it. Furthermore. lines i f.6 Realizing this to be the general use of these terms. cit. p. 96Robertson. Kalkashandi . I have been for some time puzzled with the term al-thuluthain al-saghir al-thakil. in turn. namely the original tfimdr. and al-tulmar alkabir al-thakil refer to one script only. that is.REVIEW OF A SURVEY OF PERSIAN ART 9I is for the history of scripts an important halfcentury. but the rest of the paragraph mentions no scripts derived from the mufattah. tfimar.q498 The different forms and voices of the verb kharadia as used to indicate that a given script "derives from" or "is derived from" another usually implies direct derivation from the older script. for is also called khirexample the khafif al-tfumar fa. Sometimes this variety has a distinctive name all its own. however. The next importantquestion is that of terminology used to explain the relationships of the scripts.op." and their derivatives. of the thuluthain is. On page 8. p. 32. it would seem that the mufattah is a case of indirect derivation from the thuluthain through two intermediate links-the nisf and the haram. op. then.. 8. It must be also in this sense of indirect derivation that the thuluthain and the riyasi are said to be derived from the dialil. In line 5 one is told the nisf has two varieties instead of the three of line I3. the terms khafif al-tuimarand saghIr al-tilmar represent one script.'00 The reader should keep these factors of terminology in mind. the terms. cit. in turning to the following problems of derivations that are found in this difficult and complicated passage of the Fihrist. is followed by the thuluthain al-sagh-iral-thakil and its three derivatives. It is in this sense that one is told that all the twenty-four scripts derive from the "four scripts" and the four in turn derive from the dialil.. tuimar thakil. Mention of the djalil and its two derivatives has been followed by the tuimaral-kabir and its two derivatives. Cit. used here in contrast with the kabir of the tiimar emphasizing the fact that the thuluthain is a smaller script than the tuimar. The rest of the paragraphmentions no scripts derived directly from the nisf itself. lines 2I f. and riyasi. Therefore. and this. line 29." The "it" could refer back to the mufattah. 72. 97Fihrist. In line I3 one reads that "three pens derive from it. thuluth. Note here that the diali is itself one of the four that are said to derive from it. i6. a case of indirect derivation from mukhtasaral-tiimar in the same sense as khafif altumar.97where usage calls for kabir instead of saglir.99 If the diagram is correct. but does mention two-the riyasi and nardiis-as coming directly from the khafif al-nisf. 99Fihrist. in compilinga list of scripts has overlookedthis. lines 8 f. 8. that it is derived from thakil al-nisf almumsak or the heavy. I believe.p. cf. Robertson. The terms kabir versus saghjlirand thakil versus khafif. that is. 100Kalkashandi. must refer back to the thakil al-nisf al-mumsak. restrained nisf. it is stated that the mufattah is derived from the thuluthain and in lines I2 f. "large" versus "small" and "heavy" versus "light" are in constant and interchangeable use. the main nisf.. Perhaps the apparent inconsistency is to be explained thus: One is here dealing with a passage which is enumerating the "four scripts. The Rise of the North ArabicScript. where these terms are used for indirect derivation involving one or more links between the script derived and the one derived from. for instance.op. 63 f.. The sag. and has therefore increasedthe numberof derived scripts. There are examples. note especially those under tiimar. p. This gives us a group of large and heavy pens in contrast with a group of small and light ones.

failing to realize that the riyasi and the riyasi al-kabir referred to one and the same pen. 105 Robertson. currentfor common purposes.. lists only five scripts. 73. Muhammad ibn 'Abd al-Rahman states that the "riyasi leans to the muhakkak and the naskh. it dissolves the apparent inconsistency in the number of scripts derived from the nisf. Its presence at this point solves more than one difficulty involved in the text. op. p. can still be considered as derived from khafif but states that it branched out into several pens. It 101 Cf. where riyashi is evidently a misreadingor misprinting of riyasi. the "two" referring to those derived directly from the nisf and the "three" to those derived indirectly from it through the nisf khafif.105 It is evident. where. directly or indirectly. in contrast with the light nonofficialand poorly executed scripts. The Rise of the North ArabicScript. if the diagram is right. nardiis. There the phrase riydsi-mulakkak means the "standard. 72. The first. 104 and derives them all kisas. which. and khafif al-nisf (= sagIr al-nisf). One learns further from Abu Dia'far alNahhas (d. that coming from Ibn Thawaba. clashes violently with information from the Fihrist and other sources. however. as derived from it. khafif al-thuluth musalsal. p. mu'amarat. According to the Fihrist the muhakkak developed at about the time when the Abbasids came to power. but indirectly through the link of the thuluth itself. muhakkak. whereas the riyasi did not develop until the time of Ma'mfin. p. I717) to mean that the riyasI had thirteen other scripts derived from it. 338 H. ghubar. Minovi.92 ABBOTT: ARABIC PALEOGRAPHY the nisf proper through the link of its khafif variety. mu'amarat. 103Kalkashandi. beginning with that taken from Ibn Thawaba which derives four of these twelve scripts. about threequarters of a century later. musalsal. The second list has nothing about its derivation. clashes at more than one point with the information from the Fihrist. III. is the missing thuluth. the mudawwar al-sag-ir. op. cit. But cf. from the nisf and so places these four on a par with the riyasi itself. and hawa'idlil from the thuluth.kisas. Then follows a list of thirteen pens beginning with the riyasi itself. 49. 99. op.''12 which would II. The text as preserved by JKalkashandl lists these scripts as nisf. but see p. It is to be noted that both of Nadim's lists of scripts mention the riyasi pen. cit. That script. 84. seem to imply some sort of similarity and to reflect the implied priority of the latter two. I venture to suggest. 104 Text has djawdnihk. therefore. as line I7 of the text records.. rika'. 31. it supplies the thuluth pen itself which is certainly called for in the list. But the third script of line I3 is still lacking. At any rate the earlier script. has interpreted this passage (Survey.op. wherethis interpretation was taken. From the unpunctuated text alone such an interpretation (but with twelve instead of thirteen) is possible. . thuluth-khafif. that 102 Robertson. after visualizing the diagrammatical representation of the twenty-four scripts. cit. abovep..10' This interpretation I now realize. cannot be derived from the later riyasi. it provides the most logical direct derivation of khafif al-thuluth. finally it places the thuluth among the second group of twelve scripts and so eliminates it as one of the "big four. and hawa'idjj 103 but does not specify which of these were derived from the thuluthain and which from the thuluth. al-hilya. Siill. ghubar al-halba. i6. cit. the thuluth also belongs. But Nahhas' text." leaving that place of honor to the more appropriate thuluthain. First of all. as preserved by Muhammad ibn 'Abd al-Rahman and translated by Robertson.) that al-Ahwal learned the thuluthain and the thuluth from IbrThim alSidjzl and derived from them several other scripts. p. gives its derivation and mentions only one script.official"script.

. op.. prime minister of Ma'mfin. by Mahmfid Sidki in 1908 in the then Khedivial Library from a copy belonging to the library of Ahmad Beg Timur.]). op. I-7. Cit. One is now ready to eliminate gzubar al-halba (or hilya) from the list of scripts supposedly evolvedl from the riyasi. 108 Kalkashandi. IX (I897). the mudawwaral-sagh r of Ibn Thawaba. Among the first authors. and nisf-are definitely assigned to the thuluthain by Ibn Thawaba. cit. 9 and note 5 where credit is given to "Dhfi'l-Riyasatain and the scribes. III. Whether these four belong with the riyasi group cannot be positively ascertained from the Fihrist passage alone. has collated the first few pages with a Berlin manuscriptof the same title. p.'07 Others give this 108 and still others to credit to Yfisuf al-Sicdjzi al-Ahwal. but with the author's name missing." Cf. Op. so far as I know. to mention the riyasi is Tamim ibn al-Mu'izz ibn Badis (398-453 H. The Rise of the North Arabic Script. p. separated from each other and scattered among scripts that do not belong with the riy&si group. 33 f.. I50. Early characterizations of the script are.washi.'10 The first chapter of this work deals with al-kalam wa 'i-khatt. presumably for Moritz. cit. The manuscript was copied. Robertson. That leaves khafif al-thuluth. The final answer must come from other sources. it would seem that these four likewise do not belong with that group. III OrientalInstitute. mukatabat. How then is one to initerpretthis passage? There is a clear-cut statement that the riyasi branchedout into several pens. in his 'Umdat al-kuttdb wa'uddat dhawi 'l-albdb. Nadim gives the credit to Dhu'l-Riyasatain al-Fadl ibn Sahl.REVIEW OF A SURVEY OF PERSIAN ART 93 Nahhas' text has suffered some change somewhere in its long journey of about five hundred years. and hawa'idii to be assigned to the thuluth. well executed and generally pleasing to the eye. which was acquired with the Moritz collection and which now bears the Oriental Institute number A I2060. pp. Ahlwardt. Seven of the remainingeleven have been definitely eliminated as not belonging with this group. pp. op..the problem need cause no further delay. bring together what little other information I have gleaned about the riyasi script. may have some information. cf. most likely Moritz himself. To these two may be added a third. First there is the question of its inventor. IJabib. leaving the manthuir. naturally begin with the original riyasi itself. i6.neither of whom are availableto me.D. Cit. 73 f. of the twelve scripts that follow the nis--riyas! alone belongs definitely with the group. Three of the seven-mu'amarat.'09 The probabilities are that either Yfisuf or his pupil al-Ahwal invented the script. op."'1 The 107 Fihrist. Op.106 For the time being all one knows of the riyasi group from this passage is the riyasi itself and the nisf riyasi. all very general and indicate that it was a good round pen.. Seven scripts are to be divided between the thuluthain and the thuluth. and biyad. Had. No. III. Consideringtheir position in the list. making it the official script for his day.a work devoted to the tools and accessories needed for the calligrapher'sart. 37. cit. pp. [I007-6I A. kisas. which is here called the riyasi al-kabir. the Zaiyarid prince of Tunis. gJhubaral-hilya. 109Huart.ldii Khallfa.. 72. This is perhaps as good a place as any to 106 Baghdadi and Abii Hilal al-'Askarl. after Nadim. The Oriental Institute is fortunate in possessing a copy of this work. A I2060. Since the three men are contemporary.. and does. I0207. If one controls both versions of the text by what is already knowni from the earlier account of Ibn Thawaba one gets some interesting results.. cit. 110 Cf. musalsal. p. Someone. The riyas1 group of pens would. It must be clear by now that one cannot consider the riyasi as the source of all the twelve scripts listed after it. but that Dhu'l-Riyasatain improved and approved it.

so far as time is concerned. except that in his time it was still considered among the great and good pens.op. He elaborates further that a given document written in the tfimar takes a certain amount of time. as for the riyasi. nisf. thuluthain and thuluth.94 ABBOTT: ARABIC PALEOGRAPHY emphasis is on how to select.12 In another section of the manuscript dealing with fancy pens. Gibb Memorial Fund. new ser." 114 a phrase not frequently met with in his work..RAwandwho is himself supposed to have mastered seventy varieties of scripts (ibid. cit. 34.116One must also keep in mind that the riy&sIsaghjr was a smaller round pen used generally for accounts.. . 1t17Fihrist. muim. and that "the riytasi is heavier than the nisf by a sixth. The Tauqi'dt. 443 f. partial list of scripts he ends this brief section with "Allah most High knows best. Rise of the North ArabicScript. prepare. xv) mentions only four in this chapter devotedto calligraphy-thuluth. for the letters ra. the author lists the five great (djalila) pens in this order: tiimar. nisf. however. and nfun. and characterizes the last as the lightest. The question rises as to the relationship of Nadim's second list to the one that he gave from Ibn Thawaba. (London. and poetry and called. for the Rydsi leans to the Muhaqqaq and the Naskk.'13 One wonders why he does not go on to calculate this time which is clearly a half plus a sixth. diami'. riy's-. and thuluth pen takes two-thirds. ioO for further reference to this list of p. got their names. 116Rahat al-sudiur. 192I). muhakkak.give considerable attentionto the khatt al-mansuib. These are referred to as such collectively. one-half. p. feeling that some explanation of this was needed. it must be mentioned that after giving a 112 Ibid. He then states that in the matter of size and heaviness the thuluthain is next to the tuimarfrom which it is derived. To do him justice. ed.and neskhi. but the time cannot be so mathematicallyproportionedfor either a group of scripts or a group of scribes each of whom sets his own pace for his artistic accomplishment. it is to be readily seen that the first three come from about the same time and present therefore many duplications. At some time in its career the riyasi pen seems to have been erroneously associated with the tawkl' pen. One therefore learns little from this author about the riyasi a class with the thuluthain. 72 f. 64 f scripts."17 Of the four lists of scripts. including the riyasi. He does. and sharpen the reed pens used for good and fine scripts. Perhaps the reader is already impatient of this talk about time being the one significant element in determining the names of artistic scripts! The heavy large scripts do necessarily take more time to write than the smaller lighter ones. That these 114 See p. p. its time is long.. pp. inclines to the class of 'coiled' pens and that is a feature of it. rika'. however." Then. The pen with which the riyasi is written is characterized as the thickest of pens (aghlaz al-aklam). It is clear that the author does not know how and why these scripts. In a few instances the three lists supplement each other." 115 Some idea of how the neskhi and the muhakkak were related is to be gained from Rawandi's description of some of the letters that have forms common to both scripts. kdf. except for the riyasI (manuscript has riyashi) which is referred to by name. and so places the riyasi. and a third of the time respectively. There is no eclipse and no abasing (of letters). for that reason. II. on the other hand. 8. which gives two-thirds. traditions. illustrating the proportioned letters by diagrams.. Muhammad ibn 'Abd al-Rahman says: "This is not so. pp. Muhammad Iqb5l.p. or collective script. and it is large-headed throughout. but that same documentwritten in the thuluthain. 115 Robertson'stranslation. he adds that he means the time element. cf. 3. 113 Ibid.

but that distinctive and artistic variations do characterize the various groups of scripts such as the Koranic scripts. . or g4ain. based on this and other sources. III. the additional scripts of Nadim's second list. p. The date of the sources on which Nadim drew for his second list is unknown. CPR. which is devoted to the forms of the letters of the alphabet in general and in a few specific scripts. are grouped in twos comprising an idea and its opposite. particularly those included in Ibn Thawaba's twelve primary scripts. since it is not known how completely representative each of the two lists is. These elements. they may well be as early as the third century. must be treated each on its own merits. Ibn Durustfiya has preserved a most significant passage in the matter of calligraphic nomenclature. Therefore. but I believe it is also a dangerous one. the closer one comes to the probable identification of several of the oldest of these scripts. To discover and evaluate the role of these calligraphic factors a more intimate acquaintance with calligraphic terminology is needed. e. kdf. I. as in the case of the riyasi script. the open and blinded have reference to the "eyes" or loops of such letters as fd.g.. The list follows: Thakil wa-khafif Imsak wa-sarrl' Dialil wa-daklk Idgham wa-tabyin Fath aw ta'wlr Kasr aw ta'llk Taswiya aw tahrif waTafrlk al-huruif al-sutuir diam' aw tarsIf al-huruif wa-tab' id al-sutfir Heavy and light Restrained and speedy Large and small. so far as the element of time is involved. and Grohmann. khutfit al-warrakin or book-copyist-hand. This occurs in his chapter eleven. A glance at these lists and at the diagram in Table II shows that somethinglike two and onehalf to three dozen scripts had evolved by the middle of the third century for secular use. I. Function and size of a script play an important role in the name and classification of that script. op. The more one knows of these factors. I5.REVIEW OF A SURVEY OF PERSIAN ART 95 two are not meant to be purely supplementary is clear from the many cases of duplications. cit. 64 begins by emphasizing the fact that the fundamental forms of the letters in all scripts are alike. broken letters are those cut off in preparationfor a ligature. Pt. pp. Is one to conclude that Nadim in the second list omitted scripts from the former which were no longer favored. applicable to a given script as a whole. Siili. This list is further supplemented with terms applicable to the individual letters. The contexts clarify the significance of some of these seemingly obscure terms.. more truly calligraphic in nature. But there must have been other factors. and that many of these scripts were closely related to each other. It may be significant in this connection that the definite monographs on scripts found so far in the Fihrist are all from the third century.. and khutfit al-kuttab or secretarial scripts. and included new ones that had come to the fore? This is a tempting proposition. generous and stingy"9 Merged and distinct Open or blinded Broken or ligatured Straight or oblique or curved Far-apart letters with close-togetherlines or close-together letters with far-apart lines."18 It is to be noted that the passage 118Kitdb al-kuttdb. Then follows a list of calligraphic elements which must be considered as the raw materials from which the distinctive element or elements of a given script are to be drawn. and straight or oblique (or curved) usually apply to the alifs which are either straight or 119 Cf. or the merging of two letters together instead of giving each one its due space or form. that also played an important role. 70. the merging of the three teeth strokes of a sin or shin instead of writing each tooth distinctly. f. medial 'ain.

123 Kitab al-kuttib. with what has preceded. lines I2 f. Those missing are the tfimar. which is not only one of the twelve primary scripts but also one of the principal four of Ibn Thawaba's list. dlbjdi. p. i. or it may have reference to good straight strokes and lines. and the lines close together. the hand moves (freely) in it without striving for (complete?) perfection.'25I. also The Rise of the North ArabicScript.'20 Unfortunately.. the letters should be drawn together. Whether the tfimar and rasa'il scripts are identical is difficult to say with certainty. It has already been noticed that Ibn Thawaba speaks of the nisf as kalam al-thakl1 al-nisf al-mumsak. He does. on the other hand. therefore. seems difficult to account for. but from its place among the list the rasa'il seems to be one of the larger scripts. seem to me to call for the assumption that they were known in other early sources. neither smaller nor larger. thuluth: 123 It is (derived) from the first original (dialll?). Sidiillat: Thle opposite (of the preceding). be the mukatabatof Nadim's second list. however. p. cf. The construction of the letters is on the same pattern (as the rasa'il?). 124 It .e. and some not. Athlath. The omission of the nisf. other choices and practices. pp. This specification can be interpreted in two ways. and that Ibn Durustiuya'skhafif is. 32 f. 8. the scripts that Ibn Durustfiya describes as imsak and khafif.'2' Rasa'il (tfimar?): The letters should be far apart. which appear only in Ibn Thawaba's list. or it may mean that the jotting of notes and comments on the margins of these documents is not permissible. The renditionwhich follows is not a literal translation. Mu'amarat: Its letters are the same as the thuluthain onlv smaller. Extension and separation should be avoided. therefore. Ibn Durustuiya does not undertake to give a complete and systematic description and classification of the known major scripts in the light of the above terms. none other than nisf khafif. This identification would explain why the illusmay. zanbulr. some in accord. That leaves only the nisf to consider. It is to be noted that Ibn Durustuiya has mentioned here. Kitdbtal-kuttdb. which is the last in the chapter. The passage. on a par with the khafif al-nisf. 71-73. p. Adjwibal22 and mufattah: They are between the original script (dialil?) and the thuluthain. venture to suggest that Ibn Durustiuya's imsak is none other than Ibn Thawaba's nisf al-mumsak. give several very instructive applications of some iof the above principles for the following scripts. and the lines be far apart. There is no ta'llk (wa la yu'allak) in the sidiillat or djalil. 'Uhfid: Its alifs differ from those of the thuluthain. 72.96 ABBOTT: ARABIC PALEOGRAPHY present some slight curvatures at the end. directly and indirectly. On the other hand. kisas and nisf.The riyasi is conspicuous by its absence from this list. but perhaps under different or related names. 121 Kitcb al-kuttib.haram.g.. giving them considerable attention and singling them out for special illustrations. 74. concludes with the information that the scribes have.124 The absence here of the zanbuir and haram. 122 Text has adiwira which is here corrected to adjwibafrom Nadim'slist via Ibn Thawaba. seven of Ibn Thawaba's twelve primary scripts. Significant for this identification is the fact that Ibn Durustuiya mentions the mufattah which is known from Ibn Thawaba to be a variety of the nisf proper. It may mean that irregular ligatures are not tolerated in these great scripts. Why the dlbd4j and the kisas were overlooked is not easily explained. 120E.pp. 125 Filrist. in addition to these. does not surprise one.

129I know of no specimens of the dialil. as in the case of Koranic scripts.huluthas it is in the still later work of Kalkashandi the nisf should be larger and not smaller (as it is in the illustrations) than the thuluth. a comparatively humble member of the original group. p.128a With this contribution of Ibn Durustuiyato the knowledge of old scripts. in referring to the imsak and khafif. pp. size.. since that was the script used for foreign imperial correspondence. and has accepted as genuine the illustrations found in a copy of Ibn Durustfiya's work dated 633 H. Their differences among themselves.128 For if one is to judge by the size of the t. 128It is assumed.). Among the leading scripts that have survived to the present day the thuluth. such differencesas Ibn Durustuiya himself drew attentionto in his pp. so minute that to the uninitiated in the developnments of early Islamic calligraphy. the khafif of any scripts is. 128aE. the different types of documents with which each of these individual scripts is usually associated.'27 Inasmuch as no distinction of size is maclebetween the imsak and the khafif.TheRiseof theNorth. For the essential difference between any script and its khafif variety is primarily one of slight reduction in size together with some minor peculiarities in connection with the formation of a few of the letters. Ibn I)urustuya. representingadequately over a reasonableperiod of time. but a great number of them. and were written presumably in the djalll (or tfimar?) pen. since. I am inclined to suspect that the size of Ibn Durustfiya's original illustrations has been tampered with. aside from function. cit.REVIEW OF A SURVEY OF PERSIAN ART 97 trations of the alphabet in 1)oththese scripts are so nearly identical. one has come.g.II. seems now to be the best known representative.126 The next question is how true to the author's original illustrations are the ones which have survived. one needs specimens. It is. as already statedl. 7I f. The closest approach to it from the later sources is Kalkashandi's reference and illustration not of the dialil but of the tfimar and the '29$uh. Fihrist. I believe. These documentscame from the famous pens of al-Ahwal and Sulaiman ibn Wahb. II.since nothing is said to the contrary. respectively. Again.32 f. pp. (I23 6 A. cellory scripts. op. the closest so far to realizing the general nature of these more irnportantearly chan- 127Survey.. What. if my identification of nisf and imsak holds. for instance. cf. . Before the further step of specific identifications can be attempted. has failed to see the essential likeness between them. Their calligraphic beauty was so remarkable that they were considered worthy of public display. And if the size has suffered some change the probabilities would be that some of the finer distinctions in connection with the letters have also been lost sight of. they seem to be trifling. cit. 69 f. are the chances of unearthing or bringing to light out of some dark and dusty repository in the various museums and libraries documents of early Islamic imperial correspondencewith foreign powers? One wonders what was the ultimate fate of two such documents sent by the caliphs Ma'muinand Mu'tamid to the emperor at Constantinople. Minovi. are. Script. and a varying degree of cursiveness. perhaps. the size of the imsak illustrations should be larger than that which the extant manuscript presents. p. not one or two. that the printedillustrations are of the same size as those of the manuscript. Arabic 126Cf. too much to expect that such early documents are readily available in the many fields involved. 45. Minovi in Survey. I713. They are different as a group from the Koranic scripts and the book hand. 7. op.D. I7I2. somewhat smaller in size and lighter in stroke than the script from which it is derived.

3).now and again. (908-24 A. Vol. Oriental Nos. Cit. i6. . representing the mu'amarat ence between the differentimperiallords and vassals. op. I.cit. and gliubar allilya. are here lig132 Kurd'All. 136Ibid. 32b. 131 Cf. 28a. are small. 27b and d.132 Nahhas carries one much further. 26a. Many of the specimens are confined to a narrow date limit in the third century beginning with the 240's. I2e. 29a andf. x85in the text. PIs. namely that of Baghdadi and that of Ibn al-Mudabbir.98 ABBOTT: ARABIC PALEOGRAPHY the second of the prinmukhtasaral-tuimar. I4023. and I4039 are such papyri fragments. 133 Op. 73.larger papyrusfragment than what is met with in Koranswrittenin the InstituteArabicPapyri thuluthscript. perhaps the thuluthain or the mu'amarat. In his illustrative volume accompanying his work on Arabic protocols'35 I found numerous third-century examples of what I believe to be the musalsal script. the tfmuar largestpen in use.36c. but are placed by Grohmann mostly in the second half of the third century.. 28a. III. 27d. p. 53-6i.which is No. One would expect the old djalil to inclineless to roundedand cursiveelementsthan the tulmar. III. and d.D.'39A careful study of these many specimens shows that the letters within a word or more are joined together. provided. Pts. documents Againthe chancesfor discovering the correpresenting writtenin the thuluthain. c. of the caliphswith their governors respondence or with the vassal kings. and waw. and34. 139 Ibid. The passage from Kalkashandi records further that the musalsal was a pen of joined letters (muttasil al-huruif) with none of its letters separated from the rest. square. One has a working knowledge of only a few that survived into the later centuries and that received attention at the hands of such authors as Muhammad ibn 'Abd al-Rahman and Kalkashandi. 27b. that is. In this I was fortunate.placed withindatelimitsof 299-3I 2 H. Op. III. 3. i6b. PIs.P1.. the "linked" or "chained" pen cannot be linked (together) except by (the use of) a pen cut cit. 2-3. so also are written in the chancesof obtainingdocuments correspondscript. thanks to Grohmann's publications (Fig. 33a.alil to characterize the in his day was apparently as a large pen. cit. normally final or separate. I4026.136 with one likely from the early fourth century... whose letters according to Ibn Durustuiyaare the same as those of the thuluthain.Pt.. The last source adds one further piece of information to the effect that the musalsal. tawkl'. These include the neskhi. I83. I soon realized that with such a definite and clear-cut characterization the identification of the musalsal script should be comparatively easy..'38 A few of the specimens are reproduced in color showing the use of several colored inks.) 138Ibid. i3a. p. it is stated more wal derivedit from the thuspecificallythat al-Ah. but one wonders if the specimens represent any of these early scripts. 133 but in his text as preserved by Muhammad ibn 'Abd al-Rahman. The lists given herein present a large number of scripts yet to be considered.. CPR. extant specimens were to be had. rika'. rd.'37 Several are undated.comeacrossa writtenin large script. 135 Grohmann. It will be seen that this script is mentioned in two of the earliest lists. and 3oa. of course.13' I believe the musalsal may now be added to this list. it is to be noted that Kalkathe tiimar shandi uses the term d.Pls.. index for the differentscripts. Onedoes.134This derivation I am inclined to accept. only smaller. From his text as preserved by IKalkashandl one learns that the musalsal was derived by alAhwal from either the thuluthain or the thuluth.130 cipal four scripts. iib. Protokolle. even such letters as ddl. luth.33a. 134 Robertson. The Rise of the North Arabic Script. 36e and f. 28d. 137 Ibid.. 130 op.

. XI (I9I4). 96. But since frequently a very small clue leads to bigger things. and Ibn al-Sairafl. above p.Bull. Biyad: This is probably an elegant script used by the muharrir or mubaiyad. Grohmann. franzaisd'archkol.p. Some specimens bring out the "chain" or "link" idea better than others. It is more of a common or mutlak as against the normal standard or muhakkak writing that such samples might represent. I venture to suggest. a special calligrapher-scribe.. The script. it was used in slave and real-estate transactions. I934). the letters and their relative position in the word being the same in both readings. In that case it 144Filrist.p. 7. included in Ibn Thawaba's twelve primary scripts. 176-83. shows both these scripts to be derived indirectly from the thuluthain. Bjorkman. 96. pp. if they are to be connected with the musalsal at all. I. ljandh op. 27-29. I26. Here and there one comes across documents written in irregularly ligatured script.'47 umuir al-nas wa-hawa'idjuhum" in connection with the early practice of giving public hearings to the people's complaints and petitions. may have been one used in legal documents. [I070-II47 seems promising. They are usually too fragmentary and incidental to be of direct use in the history and identification of a given script.'4' The script of none of these. ArabicPapyriin the Egyptian Library (Cairo. I 26.]) is not available work (463-542 H. Papyri documents of this type have come down to us. however. Kalkashandi. PIs. 142 Kitdb al-kuttab.l~awa'idj. is.ICanfndiwdn al-rasd'il. III. Irregular ligatures are familiar in the writing of the bismillah in all kinds of documents. As has already been seen. But replies to what? To legal petitions? I find another peculiar combination. 34-35 for the musalsal numbers in these plates. 147Fikrist. Nahhas mentioned a djawdnihl script. Hawa'idil: Baghdadi's list alone mentions this script. therefore. cit. I offer these probable clues with a few suggestions of my own to those who may be in a position to use them effectively. the same as Ibn Durustuiya'sa 1wira.. 43Cf. It is difficult to tell where other genuine samples of the musalsal scripts are to be looked for. either to discard or to build on further: Adiwiba: This script. IX. I32. now to another. this. tr.diawabat al-futh. to me. XX. p. instit. The full Arabictext of SairafI's A. a comparisonof the scripts of a sufficient number of these should help toward an identification.'44 Ashriya: This is mentioned by Ibn Thawaba as derived directly from the sidjillat. I03. op. orientale.. 141 E. The script was evidently used for some official replies.awdnihas a possibleplural the scriptwithkalamal-djan5h.. there cited. Pl.42 which I believe to be a misprint for adjwiba.for kalamalthere wouldthen be duplication is the same as al-h Op. cit. one comes across the phrases "talb al148 . His chapteron calligraphy 146 If one could accept d. have reference to writing in white ink or paint. 68. I04. it is coupled with the mufattah by this author and placed between the original Table II. can be considered as stylized to conform with the requirements of a genuine or standard musalsal. cf. The lists given still present more than a dozen scripts that are little more than names at present. and references op.who wrote out the final and fair copies of important documents. Henri Masse.g. ghlubar ilya. A. p. however. script (djalil?) and the thuluthain..D. note i.'40 Some of the ligatures appear like shallow festoons connecting the letters which themselves approach the form of links in a fancy chain. cit. cit. 145Cf. 42. however. and II4. is a misprint of hawa'idi . that is.'45 It may. I think. p. of djan4handso associate of scripts.143 140 Ibid. In the course of this study I have come across scattered references now to one. Cf. 148Bjbrkman.'46 Other references seem to favor the suggestion. . also Kalkashandi.REVIEW OF A SURVEY OF PERSIAN ART 99 atured.

"such as khafif al-thuluthain. Faulmann. I84.154 Muhaddabor muhaddath: One of these of the other.ioo ABBOTT: ARABIC PALEOGRAPHY would belongwith the sidi llat group. IHabib. 155 See n. I suggestnext that the hawd'iji of Baghdadi'slist may be the same as the sami'l of Ibn Thawaba'slist. a sort of floriated script affecting the letters themselves or is it any of the other scripts on a floral background? Is the murassa' a special kind of script written perhaps in gold or jewel-colored ink. Still uses the term manthuir way in oppositionto magdij'.. the second is met with in later lists. 49.p. I have come to it. 423. the Kuifl. munahnam(munamnam). 50 'Umdatal-kuttdb 65. I. he adds that there in size and wereothersbelowthese. perhapssomething or even musalsite directionfrom mudammadi in a general sal. 153E. p." wash! or "embellished. 528.. cit. p. for instance. 93. 7I. I50. p. p.. 154 Cf. for instance..op. e.g. kalam al-saml'i would as the scriptused to then have to be interpreted write out the decisionsresultingfrom the hearing givento publicpetitions.'52 It is usually It mightproveto be in later'amarat.ghubar sidjillat. 152 Adab al-kuttab. saghiiralnisf. op. Geschichteder Schrift (Wien.op. Op.. P..g.149 referred work already in is Ibn Badis' the five scriptshe considered Havingmentioned as the good and great ones. Robertson. = OrientalInstitute A I2060. above. 20.153 mentioned 149 Cf. One wonders what. or is it any of the wellknown scripts written on an illuminated background? Or is it both a special illuminated script on an illuminated or ornamental background? size of the script. The wordsseemsto be a misprint first is foundin our earliestlist. also Sili. Khirfadi: This scriptseemsto have been of size. Muhit al-mukit (Beirut.E. Cit.155 damadia. This acrossonly one otherearlyreference to.and Allah Most High knowsbest. Clearly. K. and Busti. p. these deal with ornamental writing.. i88o). I870).if one is to be guidedby the essentialmeaningof the verb The scriptis metwithin laterlists." seems to be descriptive of the generous iant" 151 relatedto the laterTurkishkalamrasdin which all the letters are written without ligatures.. op. and for the presentlet the referencego at that. 153aCf." its meaningof "ample. EIabib. for instance. Does the ornamentationin this group extend to the script itself or is it something in the background and. to judge from its positionin considerable the diagram and fromthe fact that it is equated The wordkhirfadiin with the khafif al-tuimar.. An Arabic-English Lexicon. Igamais al-mu4it (Bulaq.. If I am rightso far. p. cit. Manthfir: This is some sort of a spreador in the opposcatteredscript. I.. I883-85). . presumably rank. There is a last group of scripts to consider. or "luxur"abundant.-53" Mudmadior mudammadi:This seems to have beena firmand compact script." and murassa' or "bejeweled" script. W. cit. 20. haram and it is al-halba. Cit. 7I. and lists. Fairuzabadi. op. IHa-djdji Uabib. The script is includedin later Khallfa.. above. cit. p. 50. cf. p. is the characteristic feature of a zanbfir or "wasp. op. supplementary to any given script? Or does it apply to both script and the background? Is the nardjis. op. I53 156 Robertson. 20. Lane. therefore. namelythat of Baghdadi."'IO I too must concedehere that Allah knowsbest." and nardjis or "narcissus" script. The names of these have a figurative and fanciful connotation rather than a literal and practical one. washi.156 It is hardlypossiblethat these are two differentscripts. cit. Haram: Outsideof the Fihrist referenceto this script in Ibn Thawaba'slist... 15'Bustin!. or again of a dlb4di or "tapestry. III.

and has made several valuable whichhave been incorporated in the suggestions followingparagraphs. The problemhere is to ascertainif the kalams mentioned in the passagereferto reeds or to scripts." Here again though it is possible to interpret kalam al-lazaward as a script. P . Hamd Allah Mustawfii. OP. is nevertheless of considerable significance for ornamental writing. in which Ibn alMudabbir's Risalat al-'adhrd is included. The difficulty with the word Kufi as a script is soon removed when one realizes that Kufa produced large reeds of excellent quality. iqio). what Ibn al-Mudabbir states is that just as the musalsal script cannot be written except with a reed pen cut straight or square. The list beginsin a section dealingwith the description and use of the different reedpens as such.'59 According to the dictionaries a kalam muharraf is a pen cut obliquely. It may be of interest to note here that Steingass' Persian-English dictionarv zives 158 Op. 37. The second unit beginning with wa-amma reads: "And as for kalam al-ldzaward.Gibb Series (Leiden-London.In 157Kurd ' not likely to be easily available. I.I give here Ibn alMudabbir's157 Arabictext as it appearsin that edition: zs6M . L. . 8. The passage. Le Strange. has kindly given this passage his attention. Iv3. .REVIEW OF A SURVEY OF PERSIAN ART IOI Ibn al-Mudabbir gave a list of kalamnames whichraises furthersimilarquestions.. did. 161 Op. P. it seems best to treat this passagein four successive units each beginning with wa-ammd. p. Taking this last phrase alone one can consider it as referring to a variety of Kufic script. for a straight. To do this. Cit. or if some referto reed pens. just as the kutub al-mulfik and the sidiillat could not be well made except with kalam al-muharraf al-Kuifl. 160 Op. p.thoughsomewhatambiguous.Nuzhat al-kulfib.. ^. using the terms mustawl and muharrafor istiwa and tahrif. I -6 U 51. that is. 93. Ibn Durustfiya'60 and Sflli 6" both lay stress on the final shaping of the pen point. G. indeed. loan of the book. .. 4&1l 6W ). Therefore. He specifies further that the musalsal required a square-cut pen. so also the script of kutub al-mulfik (which must be either the djalil or tulmar) and the sidiillat script cannot be well written except by a large Kufan reed pen cut obliquely. especially the final shaping of the pen point.and still othersto both.158But taking it in its context it can have reference only to the reed pen and the way in which its point is shaped and is the one relied upon and resorted to in the nawa'ib and muhimmat.54. the first of these passages Ibn al-Mudabbir states the general principle that the final cutting of the point of the pen depended on the size of the script that the scribe wished to use. others to scripts. Since Kurd 'All's edition of Rasd'il al-bulagka. 159 Cf.. as Bj6rkman. Cit. it is equally possible to interpret it as a reed.. to whom I am indebtedfor the Sprengling. with emphasis on the selectionand sharpening of the reeds. cit. a square cut as against an inclined or obliquely cut point.

But I have seen many scribes prefer the nardjis script because of its compactness and homogeneity.. there is in this passage a reference to the nardiis and lazaward scripts. And as for beauty or excellence in penmanship there is no limit to it.163 Ibn al-Mudabbir himself has nothing more 162There is also the possibility that a given reed pen mightbe designedby the script for whichit is used. some of which are inclined to yellow (nardjis) streaked sometimes with light brown. and pliability of a reed. 163 Cf. the passage continuing to describe the nardjis and lazaward reed pens and to compare their qualities and the manner of sharpening their matter how minute. It is igo for passingreferenceto the form and beautyof penmanship. Continuingwith the Arabic text.'62just as some were named after the size of the sheet on which they were written. The fourth and last wa-amma should start a new paragraph here. These qualitative differences. But as in the case of the lapis lazuli. from the lists of scripts. and as compared with the lazaward it (nardjis) is simpler or more open and straighter of letters. it will be noticed. which. then why not a lazaward script also? This would indicate that scripts were sometimes named after a special kind of reed pen with which they were written.I02 ABBOTT: ARABIC PALEOGRAPHY ladiaward kasJzsdan. then the paragraphingis faulty as it stands.. much as the Spencerian steel nib was so namedafter the Spencerian handwriting or vice versa. therefore. that there were the nardiis and the dibadi (the mudabbadi of this passage). starts off with a new paragraph."to write. Such color and qualitative differences may also be due to some special treatment-see below-of the reeds preparatory to use. therefore. If." This second interpretationwould seem to be preferablewhen one remembers that the section is dealing presumably with reeds and not with scripts. would be of great importance to the penman who must watch for the strength. . since the sentences which follow refer definitely to fine or calligraphic writing in general. The new paragraph should begin after the word "Kdfl.there is reference now to kalam al-nardjis and kalam al-lazaward. the nardiis and lazawardin this passage are reeds. The different colored reeds in turn may involve qualitative differences due to the kind of reed and the stage of its growth. and others inclining to reddish browns and deep lapis lazuli blues. one wonders what a nardfis or "narcissus"reed is. But at this point one begins to wonder what a lazaward or "lapis lazuli" reed could be. with which the rest of the passage under consideration must now be associated. The text is such that both of these can be interpreted as either a reed pen or as a script. then the list beginning with the third wa-amma would also most probablyrefer to reeds of minute if not fastidious differentiation. and munamnam. and ending with the fourth and last wa-amma by emphasizing the fact that there is no limit to fine and excellent penmanship per se. as Sprenglingpointed out." indicating a transition from the subject of cutting a reed pen to that of fanciful ornamentationand scripts. The text would then be progressing from the nardljisand lazaward with the next wa-amma to even more fanciful type or types. and others after the type of document for which they were used. for instance. But one cannot so readily discard the possibility that some or all of these kalams may in the field of fine writing refer also to scripts as well as to reeds. The passage would then seem to read: "As for the lazaward script it is relied upon and resorted to in the nawa'ib and muhimmat. his.. If the first is meant then. consistency.and musahham (scripts?). If. the paragraphing at this point is uncalled for. and mudabbadi. And as for the use of the muwashsha'. that depends on the measure of the elegance of style in the writing of the scribe and on the beauty of his pen. Those familiar with reed pens suggest that the reference here is to the color of the reeds.and muwalla'.

mawshl. the dictionaries. and musahham can be used to describe one instead of five different objects. IX. pp. 166 Ibid." But the Arabic is more enlightening for it reads: 170 4JV" 'i It In connection with muwashsha' we read again of striped or figured garment. 167OrientalInstitute A I2060 p. pp. as has been seen. leather and leather binding. cleaned. already referred to. 165Adab al-kuttdb.muwashsha'.164 that is. 61-64 on pens. cf. "Concerningpaper making and the tempering and forming of pens for fine ornamental work. using them as synonyms in connection with one and the same script. and though I have so far found no direct connection with the term muwashsha'. smoothed. 169 Murtadaal-Zabidi. ornamental processes. on the other hand. The pen is then ready for use by the calligrapher employing colors as he desires. and 46-48. Sili. which is to be cut to a given size. one finds. 170Lane. 39 ff.166 If these two accounts are taken together. VIII. where the Arabic reads: 171 W (1 5o wLuoI i) Again one finds muwalla'to mean markedwith oblong stripes in black and white or other colors. Starting with musahham. The praise of the reed pen itself. They seem to imply general ornamentation involving hasmunahnam-likely a misprint. among other things. as already indicated. writing in gold and silver inks. dipped into some sort of a solution to acquire a thin coating. V. cit. cit." The section dealing with the pens gives minute instructions as to the selection of the reed.. Ibn Badis' list.I do note that washi'a and wali'a are grouped together. 17 Lane. and T&djal-'aris. seemingly. hair and feather brushes and all tools needed for gold and silver.172 It seems as if the terms washi or muwashshi. 71 f. or specific ornamental scripts. . 352 f. 65. op. His chapter eleven is entitled Fi 'amal al-kaghid wa-tawskiyat alaklam wa-nakshiha. He assigns entire chapters to such topics as black inks. be they reed pens. Of the rest of the sources available to me Ibn Badis and Sfili hold out some little hope of help. or muwashshi are freely used largely in a figurative sense and associated at times with jewels and floral terms.Tddjal-'arus(Misr. muwalla'. short close-together lines. includes washi and munamnam. the manuscript o00.. cit. mudabbadj. 15 f. quotes a number of verses in which the terms washi. The terms are used both in praise of the penmanship of a given scribe'65 and. and Tddj al-'arus. op. tested out in writing with red ink. term washi or muwashshi does not occur. p. V. and so forth. 168Op. dried in the sun.. p. cit.169 One still wonders if Ibn al-Mudabbir's muwashsha'. and the smoking repeated till the workman is satisfied that the end desired is achieved. via Lane. I88o-90). and note i. 543. 47. smoked in some fumes rising from specially treated red-hot bricks. and musahham are to be considered more or less as synonyms of his munamnam. In Ibn al-Mudabbir'slist one notes that the 164Oriental Institute A I2060. such as the wind produces in the sand. they involve also the presence of several colors. the making of glue to use in application of gold leaf. 172Lane. colors and how to mix them. and Tddj al-'arus. couple munamnam and muwashshi. it becomes necessary to keep in mind both a special kind of reed pen and some sort of ornamental script or scripts. op.REVIEW OF A SURVEY OF PERSIAN ART Io3 to say in this connection..167Suli's verses. 553 and 543.. pp. devotes his pages largely to the tools of calligraphy. colored inks. Such a pen might well be called muwashshi and connected with the script washi.'68 The dictionaries confirm this usage and explain further that the terms signify. munamnam. define it in connection with a garment "marked with stripes or lines like arrows. 85.muwalla'.andX.

The early use of color and of ornamental scripts is not to be questioned. otherslong like arrows. like fine lines in the sand. A carefulstudyof a representative group of sura headingsin the early Koransmay shed somemuchneededlight on the subject. Someof thesepresent wondersagain if there are not here five specific ornamental scripts with minute differentiation. namely. The Protocol texts to whichI have referred beforeshowsome of the earlyscriptsincluding Kuficand musalsal writtenin different colors.whenit is remembered that both in the Koranic and secular scripts knownto us. utilizing such materialsas were available. On the otherhand.someshort.or moreheavylines or "stripes" the lines of script.must wait on furtherpatientsearchfor and intelligent and sympathetic of the early Moslem understanding calligrapher's point of view.I04 ABBOTT: ARABIC PALEOGRAPHY different colorsandstripesor strokes. There still remainsmuch to be done before many of the problemsof Islamic paleography and calligraphy can be fully solved. the fourth verb form in connection with someof thesewords. In the meantime one must be on the lookout for more literary references bearingon these termsin their connection with both reeds and scripts. But the present article. 27 ff. . has or has not anythingto do with the groupof wordsunderdiscussion is difficult to say at this point. Along this line there is the furthersuggestionto keep in mind. III. The final step. Thereremains the pleasantduty of acknowledgingmy indebtedness to ProfessorSprengling for the many valuablesuggestions made in the course of readingthe entire article in manuscript.both Koranicand secular. two. has already exceeded its originally intendedlimits. three. One must also look for possiblespecimens that may be of furtherhelp. Pt.173 underlining Whetherthis 173 CPR. PIs. the differentiating characteristics are frequentlyvery minute. 3.the rapid and definiteidentification of the many simpleand ornamental scripts involved.