Maisonneuve & Larose

The Making of a Textbook

Author(s): William Smyth
Source: Studia Islamica, No. 78 (1993), pp. 99-115
Published by: Maisonneuve & Larose
Stable URL: http://www.jstor.org/stable/1595608
Accessed: 22-12-2015 18:00 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/1595608?seq=1&cid=pdf-reference#references_tab_contents

You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp

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 [email protected].

Brill and Maisonneuve & Larose are collaborating with JSTOR to digitize, preserve and extend access to Studia Islamica.

http://www.jstor.org

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
 THE MAKING OF A TEXTBOOK

There has been a tendencyto discount the work of Islamic schol-

ars between the thirteenthand nineteenth centuries.(1) Muslim
and non-Muslimhistorians alike have tended to characterize this
as a stagnant period duringwhich the ulema were relunctantto go
beyondthe intellectuallimitsthat theirpredecessors
had set for
them(2). In making their case historians have often pointed to
the formats for academic writing that were used in the pe-
riod. Medieval scholars did not write what we would consider
today to be "original" works, but focused instead on the orga-
nization, classification and analysis of what earlier authors had
presented before. Some wrote commentaries,while others com-
posed encyclopedias and textbooks that would be of particularuse
to students in the madrasah.

(1) I have no hard and fast dates in mind here,but mean only to describewhat is
sometimes referredto as the "post-classical" or "later medieval" period. Is has
been generallyheld that the intellectualvitality of Islamic Civilization has declined
since the twelfthcentury,and this is often linked to the Mongol invasions of the
central Islamic lands in the thirteenthcentury.
(2) Consider,forinstance, these remarksof S. Dayf in his al-Baldghah: Tatawwur
wa Ta"'rfkh(The [Study of] Eloquence: Development and History): "We findthat
this very same phenomenon-of repetition,sterilityof mind and lifelessness-pre-
vailed among those who studied eloquence after 'Abd al-Qihir al-Jurjini and al-
Zamakhshari. They did not come up with anything new in their studies of elo-
quence, and they restrictedtheir work to summarizing what [al-Jurjini and al-
Zamakhshari] had written. [In some cases] these later scholars would not even
expand their background by reading al-Zamakhshari's Kashshdf,so that it might
improve their summary of al-Jurjainiswork. But whether [these later scholars]
wrote summaries on al-Jurjini with or without consulting al-Zamakhshari, they
would seldom add anythingnew to the subject except forabstruse details that they
had picked up fromphilosophy and logic" (Cairo, 1965, p. 272).

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
100 WILLIAM
SMYTH

I known of no standard term fromthe Islamic Middle Ages that

corresponds to textbook(3), but it is clear enough that certain
works became standard for instructionin the medieval universi-
ties. For rhetoric ('ilm al-baldghah) the standard text was
Talkhfs al-Mifidh (The Abridgment of the Key) by Khatib
Dimashq al-Qazwini '(724/1325). The work's title, of course,
points to the fact that it is an abridgmentof anothertext, namely,
Muhammad al-Sakkaki's (626/1229)Mifidh al-'Ulhim(The Key of
the Disciplines), and this link between the and the Miftdh
represents the beginning of a short chain of textbooks
Talkhs. on rheto-
ric. Although he makes no explicit referenceto it, al-Sakk~ki
bases much of his work on Fakhr al-Din al-Razi's (606/1209)
Nihdyal al-ifjdzffDirdyal al-I Jdz(The Ultimate in Concision Con-
cerning the Inimitability of Koranic Style). Al-Razi, in turn,
explains that his work is a pandect based on 'Abd al-Qahir al-
Jurjini's (471/1078) two rhetorical compositions,namely, Dald'il
al-I'jdz (Proofs for the Inimitabilityof Koranic Style) and Asrdr
al-Baldghah (The Secrets of Eloquence)(4).
It is al-Razi who explains most clearly the textbook writer's
motivation and purpose when he introducesNihdyal al-ijaz. The
author begins by praisingthe individual insightsof his predecessor,
al-Jurjani, but criticizingthe organizational structureof 'Abd al-
Qahir's two works.
[Althoughal-Jurjnli]derivedthe principles(usdl) categories(aqsdm),
conditions(shard'iI)and rules(ahIkdm)forthissubject,he neglectedclose
considerationoftheprinciples and chapters(abwdb)...[So] I tendedto the
task of organizing
[thematerialin Dald'il al-I'jdz and Asrdral-Baldghah]
withcarefulrevisionand to redactingit withcnsideredclarification.I
arrangedthe[ir]wondrousconclusions in each chapterbymeansofdefinite
subdivisionsand I collectedthe variouspointsof the discussionin [the
format]of [well]reasonedrules.[NI p. 4]
Al-Razi goes on to claim in another section of his forewordthat
the structurehe provides in Nihdyal al-Ijdz is essential to under-
standingal-Jurj~ni,and withoutit the ideas in the Dald'il and Asrdr

(3) There are, however, terms for specific kinds of textbooks/e.g. mulakhkhas
(abridgment), (versification).
(4) In the course
lanz.m of the paper I will use the followingabbreviations: DI for
Dalad'il al-I jdz, AB forAsrdral-Baldghah,NI forNihdyal al-Ijdz ffDirdyal al Id'z,
MU forMifldh al-'Ulatmand TM for al-Mifdh. The bibliographicalinfor-
mation is as follows: DI, ed. M. R.Talkhs.
Rida', sixth edition, Cairo, 1960; AB, ed.
H. Ritter, Istanbul, 1957; NI, Cairo, 1317; MU, ed. Nu'aym Zarzfir,Beirut, 1984;
TM, ed. 'Abd al-RahmAn Barqfiq, Cairo, 1982 (reprint).

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
 THE MAKING OF A TEXTBOOK 101

would fall beyond the reach of students. Thus, in very rough

terms, al-Rdzi appropriates for the textbook writerthe work of
transmittingthe original author's ideas by supplying them with
apparent and intelligible form. This entails adding, as he says,
proper "chapters," "subdivisions" and "well reasoned
rules." Such is indeed the work of textbook writing,but it begs
the question: to what extent do such additions affectthe material
whose transmission they are intended to facilitate? To what
extent do the textbook writers appropriate their subject and in
some way make it their own?

The Subject.
The subject .of our textbooks, namely rhetoricor literally "the
study of eloquence" ('ilm al-baldghah), had not received its final
definitionby the time of al-JurjAni. The linguisticformsin which
eloquence was manifestwere a concernto various disciplines. Ibn
Rashiq (463/1070),forinstance, a North Africancontemporaryof
al-Jurj~ni,made poetic conceits (badF') one topic in the far rang-
ing literarydiscussion in al-'Umdah ffMahlysinal-Shi 'r wa Addbihi
wa Naqdihi (The Crux of the Matter Concerningthe Fine Points of
Poetry, its Practice and Critique). Scholars like al-Baqillani (403/
1013) and al-Rummani (384/994),were concerned with certain of
these conceits that demonstratedthe excellence of Koranic style in
their works on i'jdz al-qur'dn (the inimitabilityof the Koran). A
clear distinctionbetween literal and metaphorical was, of course,
essential forlegal scholars, and so we findattempts to establish it
in usuil al-fiqhworks. Finally, all of these issues fell generally
withinthe expertiseof the grammariansince gramatical structures
remained the basic tools of eloquence(5).
Al-Jurjini was himselfa grammarianwho was interestedin the
issue of Koranic inimitabilityand specificallyin the position that
i'jdz al-qur'dn was attributable to the sound (lafz) of Koranic
style. Al-Jurjaniattacks the "lafz position" and argues that elo-
quence lies not in the sound of an expression but in the way that
concepts (ma'nd/ma'dnf) are interconnected. In as much as these
interconnectionsare manifestin the grammaticalstructuresof lan-
guage (their "nazm," as al-Jurjani explains), it is these structures
that are the proper focus forinvestigationsof eloquence, whether

(5) There is a good overview of works on 'ilm al-baldghahin the introductionto

G. J. H. Van Gelder's Beyond the Line (Leiden, 1982, pp. 1-14). For a more
comprehensivetreatment,see S. Dayf's al-Baldghah (op. cit.).

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
102 WILLIAMSMYTH

the subject is the Koran or literaturein general (DI passim; e.g.

p. 81).
In refutingthe possition al-JurjAnirefersalso to metaphor-
ical expression in laf.a number of contexts. He disputes, for
instance, the evaluative implicationsof the notion prevalent in his
day that "sound is the ornament of meaning" (al-lafz zaynah lil-
ma 'dnf). Al-JurjAnipoints out that this sort of aesthetic princi-
ple-in which lafz ratherthan ma 'nd is the artisticobject-is com-
pletely inadequate for understandingthe eloquence of figurative
expression. As an example he cites the phrase, "I have a gaunt,
young camel" (mahzal al-fdsil) in which the speaker illustrateshis
generosityby describing his animal's poor condition.(6) Al-Jur-
jini explains that this sort of expressionuses a "first"meaning(i.e.
the camel's leanness) to referto a "second" one (the speaker's gene-
rosity). Al-Jurjnitmaintains that the eloquence of the expression
lies in the manner in which the idea (ma'nd) of the gaunt, young
camel is chosen. Accordingly, the sound of the phrase alone
(mahzdl al-fdsil) possesses no rhetorical ornament (DI pp. 307-
308).
(7)
Al-Jurjaniconsiderssyntax and figurativespeech in both Dald'il
al-I'jdz and Asrdr al-Baldghah, but the formeris primarily con-
cernedwith the subtle nuances of Arabic grammar,while the main
subject of the latter is the "hidden meanings" (the secretsor asrdr)
of similes and metaphors. Although the starting point for the
discussion is the inimitabilityof Koranic style, it becomes clear
fromthe outset that al-Jurjini is concerned with eloquence as a
general linguisticquality. He ranges in his writingover the entire
Arabic literary corpus, and like other works on baldghah, the
Dald'il and Asrdr have their impact on a number of disciplines.

The UltimateAbridgment.
We need only the title to know that al-RAzi chooses to collect
the material fromal-Jurjini's works under the rubric of i'jdz al-
qur'dn. The author emphasizes his choice by using discussions of
the i jdz doctrineto framehis presentation. Thus, Nihdyatal-Ijdz

(6) This is to say that the camel is lean because it receives no nursing,and this is
due to the factthat the speaker has slaughteredits motherin orderto feedhis many
guests.
(7) Al-Jurjini calls the referenceto an second or intended meaning (ma'nd) by
means of a second one the "meaning of the meaning" (ma'nd al-ma'nd), and it is
his basic descriptionof figurativelanguage.

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
 THE MAKING OF A TEXTBOOK 103

opens with an introduction (muqaddimah) in which al-RJzi re-

views the various arguments that support the stylistic notion of
i'jdz, affirmingin the end al-Jurjini's position that the eloquence
of the Koran makes it inimitable(DI p. 7). The author returnsto
the topic of i'jdz in the last chapter of Nihdyat al-Ijdz where he
considers several issues concerning Koranic style(8) (DI pp. 160-
168).
Al-RFzi's arguments in the introductionand conclusion do not
come fromal-JurjAni,but fromworks that refermore directlyto
the i dz issue. (9) For al-Jurjini i 'dz is a prelude to his discussion
and little more. In the short forewordto the Dald'il he actually
identifieshis subject as grammarand goes on in the early sections
of the work to defend the study of poetry (DI pp. 15-28). Al-Jur-
jini comes eventually to the topic of ijdz, but considersthe partic-
ular case of the Koran only long enough to tie i jdz al-qur'dn to
eloquence. Afterthis the discussion turns to general features of
literaryquality and its connectionto grammaticalstructures. Al-
though he cites the Koran as an example of fa.dhah oftenenough,
al-Jurjini does not refer to any instance of eloquence that is
unique to the Koran. There is no referencein the Dald'il to crit-
icism of the i jdz doctrine,and what is perhaps more to the point,
there are no referencesat all to Koranic inimitabilityin Asrdr
al-Baldghah. Accordingly,al-Razi's introductionand conclusion
to Nihdyal al-Ijdz representsomethingof a shiftin emphasis.
In the main body of his work al-RAzi is concerned,like al-Jur-
jAni,with the general quality of eloquence, and he divides the work
accordingly. Afterdemonstratingthat i'jdz al-qurdnis attribua-
ble to eloquence al-RJzi claims that fasdhah itself may be attri-
buted to two differentcomponents of language: single elements
(mufraddial-kaldm) and composite ones (jumal). Thus, he splits
al-JurjAni'smaterial into two sections, namely one that is con-
cerned with the single component(i.e. the word as it stands alone)
and another concerned with the composite one (the phrase or sen-
tence). In the firstsection the greater part of his presentationis

(8) These are 1) the aspects of i 'z in i jdz in Saral al-Kawthar,2) whetherthere
is any contradictionin the existence of both evident and obscure passages in the
Koran and 3) the criticismthat the sacred text is inconsistent,prolix and repeti-
tious.
(9) In I'jdz al-Qur'dn (Cairo, 1964), forinstance, al-Baqillini arrangeshis presen-
tation in a similar way; he advances arguments to "prove" the idea of i'jiz and
defend it fromthose who attack it. There is nothinglike this in al-Jurjini's pre-
sentation.

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
104 WILLIAM SMYTH

devoted to figurativespeech,(10)while he considersArabic syntax

in the second. ("1)
The division of Nihdyal al-Ijdz is based on al-JurjAni'stwo argu-
ments that, on the one hand, faqdhahis manifestin grammatical
structures,and on the other,it is evident in the ma 'nd process of
figurativelanguage. It is importantto note, however, that 'Abd
al-QAhirmakes no all-inclusiveand definitivestatementregarding
literary quality. He outlines these two particular positions in
order to refutethe notion that the of Koranic style makes it
inimitable. But al-Jurjini neitherlaf. links the two statements nor
suggests that they offer an exhaustive definition of elo-
quence. We must again remember that these two arguments
come in Dald'il al-I jdz where the author focuses on fasdhah and
the idea that one expression is "superior" to another (mazlyah).
Al-Jurjfni's focus in the Asrdr lies, as the title implies, more in
understandingthe text than evaluating it; although aesthetics and
hermeneuticsare important in both works, we may nevertheless
discern a differencein emphasis.(12) This differenceis diluted in
the structureof Nihdyal al-Ijaz where al-RAzi organizes material
fromboth works in an evaluative context.
In consideringsingle elements al-RAzi makes an interestingdis-
tinctionwhich seems to contradictal-Jurjdni'sargumentthat elo-
quence should not be attributedto lafr. Althougheloquence may
not be a "quality" of (al-faqdhah ld yajaz an takan sifah li-
laf.
(10) The distinction between mufraddtal-kaldm and jumal comes on page
eight. Al-RAzi lists figurativespeech as "single elements" because they involve
the ma'nd al-ma nd method of significationconsideredabove. Thus, a singlesound
(laft) signifiesa ma'nd that is not its literal referent,but one related to it. The
ma'nd of the gaunt camel, forinstance may be considered"related" to the ma indof
hospitality. Whatever relationship-or series of relationships-might be establis-
hed between ma'dnF,the speaker will only us one lafz, and so the literal linguistic
component will always be a single element.
(11) Al-Razi goes througha long chain of divisions and subdivisions in order to
arrange the presentationin Nihdyal al-Ijdz. His discussion of figurativelanguage
forms four out of five "principles" (qawd'id) of the second (out of two) parts
(aqsdm) of the firstsection. These four principles representapproximately two-
thirdsof the firstsection. In the second section his discussion of syntax comprises
four out of the six chapters (abwdb); this is also about two-thirdsof the total
section.
(12) Although it sounds a bit simplistic I do believe the titles themselves are
indicative. In Dald'il al-I jdz al-Jurjani is concerned throughoutwith hermeneu-
tics as a tool for determiningthe superiorityof one expression to another. In
Asrdr al-Baldghah, on the other hand, the emphasis seems to fall more on the
signifyingfunctionof figureslike metaphor rather than relative quality.

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
 THE MAKING OF A TEXTBOOK 105

lafz), al-RAzi explains that we do findinstances of literaryquality

that are the result of lafz (aqsdm al-mazdyd al-hdSilah li-l-kaldm
bi-sababal-alfdz). The author dwells at some lengthon the differ-
ence between an inherentquality and a consequent one, but his
real reason formaking the distinctionis to introducethe badf' into
the presentation (NI p. 21). Whereas al-Jurjini wrote at great
lengh on those figuresof speech that interestedhim (i.e. metaphor,
simile), he did not make the effortto include those that were not
part of his system and al-RJzi aimed to provide a place for
them.(13)
Poetic conceits have an importantplace in 'ilm al-baldghah. It
was the poet and caliph, Ibn al-Mu'tazz (904), who firstdevoted an
entirework to them in Kildb al-Badf', and the practice of describ-
ing eloquence solely in terms of the conceits eventually extended
to works on the Koran, like Ibn Abi al-Isbai"s (654/1356) Bad'"
al-Qur''dn. Al-R1zi acknowledges this by working the bad'" into
Nihdyal al Ijdz, but in so doing he again distorts al-JurjAni's
ideas. 'Abd al-QAhirwas certainlyaware of the badf"manvals, but
thereis a marked differencebetween his approach to style and that
of a bad'" author like Ibn Rashiq. The bad'" method is a taxo-
nomic process of identificationand collection in which the critic
often analyzes a figureonly in so far as he needs to distinguishit
fromanother figure. The complete semantic functionof the figure
(the complete determinationof its "asrdr"), which is paramount
for al-Jurjani, is not so important.(14)

The Key to the Issue.

By adding his particular focus to al-JurjAni'sdiscussion of style
al-RAzi dilutes what the ealier author intended to empha-

(13) Al-Rizi includes the badf' in both the firstand second sections under several
headings. Concerningthe author's source forhis presentationof the bad[' see my
"Early Persian Works on Poetics and their Relationship to Similar Studies in Ara-
bic," Studia Iranica, 18, 27-53. Al-Jurjini offersimplicit criticism of the bad['
approach in Asrdral-Baldghah(p. 26) when he criticizesa previous author's simplis-
tic analysis of metaphor. He complains that the figureis simply identifiedbut its
full significanceis not addressed. One editor of the text, H. Ritter, identifiesthe
author, which does not, as Qudsma b. Ja'far (940/948)since the line and
al-Jurjfni
explanation comes fromhis Kitdb Naqd alShi'r.
(14) The discussion of tashbfh(simile) in some of those manuals is instruc-
tive. In his Hiadd'iq al-Sihr, forinstance, Rashid al-Din al-Watwat (1182), distin-
guishes between eight differentkinds of simile according to formal features like
reverse comparisons and comparing two objects to the same subject. There is,
however, no consideration of the informationa simile is used to communicate.

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
106 WILLIAM SMYTH

size. This dilution of al-JurjAni'sauthorial subject is much more

advanced in Miftdhal-'Uldm. There is no question but that al-
Sakk~ki relied on both al-Jurjani and al-RAzi in composing the
Miftdh,and yet he makes no referenceto eitherwhen introducing
the work. Al-SakkAkistates only that he will referto the worksof
his "predecessors" (qarrart md sadaft min drd' al-salaf).(15) In
the entire Miftdh there is no mention of al-RAzi and only two
isolated referencesto al-Jurjini. The farmore serious blow to the
integrityof al-JurjAni'svision comes, however, in the complete
contents of al-Sakkiki's work. The discussion of style and figu-
rative usage that originateswith'Abd al-Qahir comprisesjust two
out of nine sections fromMiftdh.al-'Uldm. Whereas al-R11zimay
have put al-Jurjini's ideas to a more specificpurpose, al-Sakkaki
incorporatesthem within a new whole.
Al-Sakkaki followsal-Rfzi in closing the Mifldh.with a general
discussion of i'jdz al-qur'dn (MU pp. 578-602).(16) He opens the
work,(17)however,by locating his topic withinthe context of adab,
which in its medieval sense can cover practicallyall literaryknow-
ledge fromhandwritingto history. The author offersno defini-
tion for adab, but claims, simply enough, that the student's first
priorityin applying literaryknowledgeis to avoid mistakes in Ara-
bic usage (ihtirdz 'an al-khata' fi kaldm al-'arab). The occasions
forsuch mistakes,he explains, are three:the single word (mufrad),
the compositionof grammaticalstructures(Ia'lif) and the manner
in which those structurescorrespondto authorial intention(kawn

(15) MU. p. 6. Al-Sakkaki makes various referencesto the salaf but never
explains who they are. For a discussion of al-Sakkdki's sources see A. Matlib's
Al-Baldghah 'Inda al-Sakkdkf(Cairo, 1964, passim); Matlib also considers al-Sak-
kiki's specificreferencesto al-Jurjini (p. 194).
(16) Al-Sakkaki does not, however, consider the same topics as al-Razi in his
discussion of i'cjdzand its detractors.
(17) MU does not have an actual introduction(muqaddimah), but only a "fore-
word", which the author seems to call a fdtih.ah. This seeminglyminorpoint has,
I believe, some significance. Scholarly works from this period almost always
include some general foreword(which is given various names) in which the author
names his subject. They do not, however, always have an methodological intro-
duction--or prolegomena, and the fact that al-Razi and al-Qazwini both include
introductionsis a clear organizational addition. Al-Sakkaki does not make this
kind of addition because he runs into so much trouble simply naming his sub-
ject. For al-Razi and al-Qazwini this was a simple task since they based their
worksdirectlyon an earliertext in which the subject was readily identifiable. But
al-Sakk~ki "creates" his subject, and so, must take some time (five pages in my
edition) in his forewordto explain what it is. This leaves no place foran additional
prolegomena.

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
 THE MAKING OF A TEXTBOOK 107

al-murakkabmuldbiqanli-md yajib an yalakallam lahu). Accord-

ingly, al-Sakkdki includes in the Mifidh.the disciplines ('ulam)
that treat these issues. He begins with morphology (sarf) in
orderto considerthe formof single words; he then moves to gram-
mar (nah.w) where he describes how those words are joined
together and ends with syntax (ma'dnf) and figurative usage
(baydn) under which headings the correlation between intention
and expression is discussed (MU, p. 8).
Al-Sakk~ki acknowledges that philology ('ilm al-lughah) is also
relevant to avoiding mistakes, but he declines to cover it. He
does, however, promise to consider several other disciplines which
are closely related to his fourmain subjects. These are derivation
(ishtiqdq), definition(hadd), deduction (istidldl), prosody ('arld)
and rhyme (qawdfi), and their relationship to the rest of the
is explained in the followingparagraph.
Mifidh.
So I includedin mybookthoseaspectsofadab that I considered neces-
sary except forphilologysince it is the [basic] element[in] the derived
aspects[ofadab.] I putin 'ilmal-sarfin itsentirety,and it is notcomplete
without'ilmal-ishtiqdq... I included'ilmal-nahwin itsentirety, and it is
[only] complete when it includes Jim al-ma'adnfand 'ilm al-
baydn... Since 'ilm al-ma'ant is [only]completewith 'ilm al-haddand
I saw no escape fromgrantingthem [a place] as
'ilm al-istidldl(18)
well. In as muchas theapplicationof 'ilmal-ma'adnand 'ilmal-baydnis
based on usingthe genresof poetryand prose- and I realizedthat
someoneconcernedwithpoetryrequires'Arddand qawdff- I movedthe
reinsof my pen to includethem.(MU p. 6)
Al-Sakkaki's rather un-ambitious definitionof adab should not
mislead us. It is true enough that the goal in studyingthe Mifldh.
is the ability to express oneself without error. But in order to
impart the required knowledgeto the student the author sets him-
self the task of organizingthe faculties of expression within a co-
herent system. This is to say that al-Sakkaki is concerned not
only with individual disciplines, like grammar and prosody, but
also to determinethe totalityof semantic functionin language and
to distinguishits various aspects. In offeringthe reader the abil-
ity to express himselfclearly al-Sakk~ki promises to describe the
semantic whole of Arabic and organize it into well defined dis-
ciplines.
It is his concern forcompletenessthat makes al-Sakkaki include
disciplinesthat are, as he explains it, auxiliary to the task of avoid-

(18) It is forhis attemps to integraterhetoricwith logic that al-Sakkiki is proba-

bly best known. Matlfibconsiders this in Al-Baldghah,pp. 159-163.

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
108 WILLIAMSMYTH

ing mistakes. As we saw in the paragraph quoted above, al-Sak-

kaki claims that exigencies inherentin the four main subjects of
the MifIdh.require that he include more material. Morphology,
forinstance, "is not complete withoutderivation;" "syntax is only
complete with definitionand deduction;" "the use of 'ilm al-
ma'dni and 'ilm al-baydnis based on... poetry and prose." It is
again clear that the author conceives of his subject as a coherent
whole in which one part demands another, and no more or less
than necessary should be included.
Distinctions between the individual parts are forthe most part
self evident since seven of the nine subjects (derivation,grammar,
definition,deduction, prosody and rhyme) were established and
well defined disciplines (i.e. 'uldm) when al-Sakkaki wrote. For
syntax and figurative usage, however, the case was different
because these topics were common to grammar,rhetoricand usyil
al-fiqh. As we noted at the outset al-Jurjini's treatmentof these
subjects straddles a numberof disciplines. The Dald'il and Asrdr
are not really grammaticalstudies, but they focus on grammatical
devices, and their author is, after all, best known as a grammari-
an. At the same time, the formerwork purportsto offerproofs
forthe i'jdz doctrine,while the latter refersto rhetoricin its title
and considers forthe most part poetry in its presentation.
Since al-Razi places Nihdyal al-Ijdz within the context of i 'dz
(an issue not a discipline)the question of where his topic fitswithin
the Islamic academic frameworkdoes not arise. Al-Sakkdki,how-
ever, begs to ask and indeed answer this question by placing his
arrangementof al-Jurjani's ideas within the comprehensivecon-
text of the Mifidh. We can see this fromthe outset in the way
that al-Sakkaki makes the subjects of syntax and figurativeusage
into uldm. By referringto the subjects as 'ilm al-ma dni and
'ilm al-baydn,al-Sakkfki grants them at least the appearance of
definitionand integritythat the other "uldmin the Mifidh actually
possessed.
The specialized use of ma 'dnfgoes back clearlyto Dald'il al-I 'dz
where al-Jurjini refersthroughoutthe work to the featuresof syn-
tax he discusses as ma'adnfal-nahw. The specialized use of baydn,
on the otherhand, is more difficultto link to'Abd al-Qfhir because
the work(meaning literallysomethinglike elucidation) had been so
long and so generallyassociated with literaryexpression.(19) Al-

(19) Al-Jurjini refersobliquely to the Koranic referenceto baydnin the opening

of Asrdr al-Baldghah (p. 2), but says nothing about figurativelanguage in parti-

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
 THE MAKING OF A TEXTBOOK 109

Jurjini makes some general comments on 'ilm al-baldghahat the

opening of Asrdr al-Baldghah, but makes no mention of 'ilm al-
baydn. The firstactual referenceto 'ilm al-ma'dnf and 'ilm al-
baydn occurs in al-Zamakhshari's (538/1144) Koran commentary,
al-Kashshdf,where the author specifiesthat the two disciplines are
central to appreciating the message of revelation. He does not,
however, offerany definitionfor the two 'uldm.(20)
Although the specialized association of ma'adnfwith syntax and
baydnwith figurativeusage was probably "in the air" by the thir-
teenth century,the firstdefinitionsforthe two disciplinesseem to
come in the Mifldh. Al-Sakk~ki begins to address this issue when
he outlines the three mistakes possible in Arabic usage. He notes
that the second of these, mistakes in sentence or phrase composi-
tion (ta'lff), is addressed by grammar,while the third,the corres-
pondance between expression and intention, is addressed by al-
ma'dnf and al-baydn. He thus implies that authorial intention
plays some role in the latter two disciplinesthat it does not play in
grammar.(21)
The distinctionbetween 'ilm al-ma'dnf and 'ilm al-baydnlies in
the differencebetween literal and figurativeusage. This is in
essence the same division that al-Razi makes when splitting
Nihdyat al-Ijdz into two sections, but the differentways in which
he and al-Sakkdki approach this distinction is instructive. For
usageis a givenprinciplethathe citesin orderto
al-Rdzifigurative

cular. Indeed, al-Qazwini notes in his own commentaryon the TalkhTs,namely

that 'ilm al-baydnis often used to referto all of 'ilm al-baldghah,not just
al-Iddh., usage (ed.: M. Khaffji, sixth edition, Cairo, 1985, p. 83).
figurative
(20) Al-Kashshdf 'an Hadd'iq al-Tanzfl wa 'Uydn al-Aqdwil fiWujdh al-Ta'wfl,
Cairo, s.d., p. 4. In his al-Baldghah S. Dayf suggests that al-Zamakhshari is the
firstto so distinguishbetween 'ilm al-ma'dnf and 'ilm al-baydn(op. cit., p. 222).
(21) Al-SakkAkielaborates on this distinctionin the main body of al-Mifldh. In
introducinggrammarhe explains that the disciplineteaches the student to conform
his speech to normative usage (MU p. 73), while he claims in the section on al-
ma'dnf and al-baydnthat these disciplines consider how one matches an expression
to a particular situation (MU p. 161). This is to say that grammarcan only teach
the student to match his nouns to his verbs and his declension to his conjuga-
tion. It gives him no advice on whetherhis statement will demean his subject or
impress his audience, but only prevents the speaker fromcontraveningthe Arabic
paradigms. A knowledge of al-ma'dnf and al-baydn, on the other hand, goes
beyond normativeusage. The two disciplinesexplain how to us language purpose-
fully and persuasively and so, in al-Sakkiki's definition,to "make speech appro-
priate to what the situation demands. Accordingly, al-ma'dnf and al-baydn
emphasize the rhetoricalaspect of grammarinasmuch as they link an expressionto
its social context.

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
110 WILLIAM
SMYTH

make a sub-division. As we have seen, the author's main princi-

ple of organization is the idea that eloquence is attribuable to ei-
ther single words or phrases. Moving to single words, he explains
that referencein this regard may be either literal or figurative(NI
p. 8) and devotes the firstpart of Nihdyal al-Ijdz to the latter,
consideringsimile, metaphor and metonymy. It is importantto
note here that al-RAzi treats these topics individually and makes
little effortto link them or to describe figurativeusage generally.
Simile, metaphor and metonymyalso formthe main topics in
'71mal-baydn,but foral-Sakk~ki a general definitionof figurative
usage is essential. The author describes 'ilm al-baydn as "the
knowledge of signifyinga single concept (ma end) in a variaty of
ways ...," and goes on to note that this manner of significationis
only possible in figurativeusage (MU pp. 329-332).
Literal usage, he explains, depends on the fixedrelationshipbet-
ween laft (the word or signifier)and ma'nd (the word's meaning or
literal referent)in which no variety is possible; the sound, "kalb,"
forexample, may only referin Arabic to the concept of dog. In a
figurative expression, however, there is potential for variety
because of the ma'nd al-ma'nd manner of significationwhich al-
Jurjfni outlines. For instance, in the expression considered
above, "I Have a young, gaunt camel," the speaker uses an inter-
mediate, conceptual referent(i.e. the gaunt camel) to referto the
intended referent(i.e. the generosityof the speaker). Al-Sakkfki
explains that this intermediatereferentmay be one or many and
further,that it is possible to referto the same intended referent
with various combinations of intermediate referents. Accord-
ingly,one may express the same meaning,as al-Sakkfki says, "in a
variety of ways." (22)
Such variety is possible in 'ilm al-baydnwhile in 'ilm al-ma
'nn,
it is not. Accordingly,al-Sakkfki uses his major division of al-
Juj-ni's ideas to focus on differentprocesses of signification. The
distinctionbetween literal and figurativedoes not originate with
any of these authors, but each one uses it for his own pur-
poses. As we have seen, al-Jurjani considers ma'nd al-ma nd in

(22) The variety, which is a detail added by al-Sakkiiki, lies essentially in the
options that these "intermediatemeanings" represent. Anotherpopular example
in this literaturis the phrase, "many of ashes," which also describes generosityby
suggestingthat the ashes are evidence of many meals given to guests. One might
consider this phrase another "way" (figurativeas opposed to literal) to express the
same intended meaning or, in al-Sakkiki's words, the "single concept."

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
 THE MAKING OF A TEXTBOOK 111

order to offerfurtherrefutationof the argument,and al-Rdzi

laf.so that he might make
refersto the literal-figurativedistinction
furthersubdivisions in his presentation. Al-Sakk~ki, however,
takes special pains to describe figurativeusage and draws the very
important line between "ilm al-ma 'dnTand 'ilm al-baydn on the
basis of it. In this way he draws the student's attentionto differ-
ences between semantic processes and gives this a special promi-
nence. He makes the mechanics of significationthe subject of
Miftldhal-'Ulum, not the attributes of eloquence with which al-
Razi is concerned in Nihdyal al-Ijdz.
The Last Word.
Of the fourauthors under considerationonly al-Qazwini wrote in
an Arab environment. Although his name suggests a Persian
milieu, al-Qazwini's familyhad moved to Mosul in the generation
beforeKhatib Dimashq, and our author lived his life in Damascus
and Cairo.(23) Al-Jurj~ni, al-R~zi and al-Sakkiki, on the other
hand, were all part of the Eastern school of textual studies. Mem-
bers of this school possessed a knowledge of Arabic that was unas-
sailable, but they were not Arabs, and the Easterners were some-
times criticized for their subtle, but dry analysis. (24) It is not
insignificantthen that al-Sakkiki, a Turk and perhaps the driestof
the lot, makes 'ilm al-baldghahinto logical semantics,while it falls
to the Arab, al-Qazwini, to reclaim it forrhetoricand literature.
Although he makes the modest claim of "abridging" (lalkhts)
al-Sakkaki's work, al-Qazwini ignores the comprehensivegoals of
Mifldhal- 'Ulm. He notes in the forewordto the Talkhfsthat his
subject s 'ilm al-baldghah,which is, oddly enough, a discipline to
which al-Sakk~ki makes no reference.(25) Al-Qazwini dispenses
with the morphology and the grammar, the deduction and the
prosody,and takes as his subject "ilmal-ma 'dnfand "ilmal-baydn,
asserting that these two disciplines are the basis for "ilm al-
baldghah. The author incorporatesal-Sakkiki's method of defin-

(23) Carl Brockelmann, Geschichleder arabischenLitteratur,Leiden, 1937-1949,

v. II, p. 26; supplement II, p. 15.
(24) The "Eastern School" of rhetorical studies was comprised of the close,
grammatical analysis for which all these texts are good examples. A correspon-
ding "Western"--or Arab school (the Eastern School was composed of Turks and
Persians forthe most part)-focused more on the badf'. See S. Dayf, Al-Baldghah
(op. cit., passim).
(25) Althoughal-Sakkiki devotes a lot of attentionto the place of 'ilm al-ma'dnf
and 'ilm al-bagn in the 'ulum structure,he makes no referenceto 'ilm al-baldghah.

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
112 WILLIAM
SMYTH

ing the two disciplines,but he adds another 'ilm to the system: he

adds 'ilm al-badf', making it the finalelement in a tri-partitefor-
mulation of 'ilm al-baldghah.(26)
As we consideredabove, al-Rdzi founda way to incorporatethe
badf' into Nihdyal al-Ijdz even though al-Jurjini had ignored the
subject. For al-Sakkfki, as well, it had been importantto include
the badf'. Although he does not referto the subject in his intro-
duction to the Miftdh,the author notes at the end of the section on
'ilm al-baydnthat he will at that point consider additional topics
which relate to the "ornamentation" (lahssn) of speech. He calls
these "the finepoints of expression" (mahldsinal-kaldm) and, char-
acteristicallyenough, divides them into two groups: "those which
relate to sound" and "those which relate to meaning" (MU p. 423).
It is this section that formsthe basis foral-Qazwini's discussion of
the badf'.
In spite of this precedent forconsideringthe badf' in this way,
the fact that the Talhkfs refersto this section as "ilm al-badf'
remains significant. Although al-Rdzi and al-Sakkfki included
the badf' in their presentations,the subject clearly held a sort of
"second-class" status forthem. This is particularlyevident in the
Miftdhwhere al-Sakkfki specifiesthat the badf' involve the orna-
mentationof expression, an obviously inferiorrole within the
semantic system he sets up. While al-Qazwini does not actually
"upgrade" the definitionforthe badf' that al-Sakkaki establishes,
he does consider the conceits at greater length than his prede-
cessors.(27) Al-Qazwini's modest expansion of the section on the
badf' is all the more apparent because of the reduced formatof the
Talkhts. It lends the sense that 'ilm al-baldghahwas a truly tri-
partite discipline and that the badf', which had always been a

(26) Bahi' al-Din al-Subki (1372), a later commentator,claims that al-Qazwini

makes such significantchanges in the Miftdh that he misnames his work. The
al-Subki argues, is no mere summary('Ards al-Afrd.h, printedwith a num-
Talkhs.,
ber of commentaries in Shurdh al-Talkhfs, Cairo p. 63). The total contents of
Talkhis al-Miftdhare: foreword,introduction, 'ilm al-ma'dnf, 'ilm al-baydn, 'ilm
al-badf' and conclusion.
(27) Al-Qazwini refersin his forewordto 'ilm al-baldghah"and its consequent
topics" (lawdbi'uhd; TM p. 21). In the discussion of baldghah and fascdhahhe
claims that the directly relevant disciplines are 'ilm al-maadnfand 'il al-baydn,
while 'ilm al-badi' refersto "manners of ornamentation" (wujdh al-lahsfn; TM
p. 37). Accordingly,al-Qazwini's definitiondoes not differfromwhat we findin
MU.

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
 THE MAKING OF A TEXTBOOK 113

popular topic among litterateurs,held a place equal to that of

al-ma'dnf and al-baydn.
The increased prominenceof the badf' shiftsal-Qazwini's confi-
guration of 'ilm al-baldghah so that the discipline as a whole
inclines more towards the concerns of poetry and belles
lellres. The author gives furtherdefinitionto this new configura-
tion in the way he begins and ends the TalkhTs. Aftergenerally
introducinghis subject al-Qazwini devotes a separate "introduc-
tion" (muqaddimah) to definingbaldghah and fasdh.ah,the two
standard termsforeloquence and verbal art. Al-Rdzi and al-Sak-
kAkihad previouslyestablished definitionsforthese two terms,but
both autors effectivelybury the definitionswithin their main
texts. They give the topics neitherfocus nor prominence,and it
is, therefore,significantthat al-Qazwini uses the definitionof elo-
quence to open the TalkhTs(TM pp. 22-37). He thus remindsthe
reader that the specificmethods he describes (al-ma'dnf, al-bdyan
and al-badf') are intended to serve the investigationof eloquence,
and that the purpose of 'ilm al-baldghahis to distinguish state-
ments that move the audience fromthose which do not.
The muqaddimah to TalkhTsal-Mifldh serves clear notice that
al-Qazwini is not interestedin logic and semantics. We can also
see that he has little interestin i 'jdz al-qur'dn since this topic has
largely disappeared from the discussion.(28) Thus, al-Qazwini
effectivelystrips 'ilm al-baldghahfromsome of its ideational asso-
ciations in the way he presentshis material. There is furtherindi-
cation of this in the conclusion to the Talkhis where the author
considers two main topics: the poet's appropriation of common
themes (sariqdl, iqlibdsand ladmfn)(29) and the structuralelements
of poetry,namely the beginnings,endingsand transitionalsections
of poems (TM pp. 408-435). The section works in some ways like
an appendix ratherthan a summingup since al-Qazwini uses it to

(28) There is a referenceto i'jdz in the forewordto TM, but this is very brief:
"Since 'ilm a-baldghah...reveals the inimitable quality in the arrangement(nazm)
of the Koran..." (p. 21). The referenceto i'jdz al-qur'dn was fairlystandard in
works dealing with rhetoric after Abfi Hilil al-'Askari's (395/1004) Kildb al-
Sind'atayn. Accordingly,the referencein TM cannot be considered to locate the
author's argument in the same way that his introductiondoes.
(29) Sariqah is often translated as plagiarism (e.g. G. von Grunebaum's "The
Concept of Plagiarism in Arabic Theory," Journal of Near Eastern Studies, 3, 234-
253) which gives it a negative connotation. The discussion is the khdtimahof the
Talkhis, however describes the topic in less evaluative ferms,and so "appropria-
tion" seems a more useful translation.

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
114 WILLIAM
SMYTH

introduce new topics to the discussion. Although al-Jurjdnihad

devoted a section in the asrdr to literary appropriation (AB,
pp. 313-324), neither al-Rdzi nor al-Sakkaki considered it neces-
sary to fitthe topic into their works. Al-Qazwiini, for his part,
"resurrects" the topic of poetic appropriation. He drastically
reduces the scope of'Abd al-Qahir's presentation,but gives it an
added prominenceby using it to conclude al-Mifidh.(30)
Although his khdtimahis short and clearly derivative, it never-
Talkhs.
theless moves 'ilm al-baldghah further in the direction of
adab. This is not the adab of linguistic competence with which
al-Sakkfki was concerned, but the adab of poetry, al-Jahiz and
belleslellres. Although there is no actual referenceto adab (much
less a definition)in the TalkhTs,the organization of the work func-
tions as a sort of responseto the labored definitionforadab that we
findin the Mifldh. In his effortto achieve precision and totality
al-Sakkaki had removed the notion of privilege from dis-
course. He had turned poetry into expression and reduced elo-
quence to competence. The clear message of the is that
such a colorless conception of rhetoricwas incompatible with the
Talkhs.
honored role that artful turns of phrase traditionally played in
Islamic society.

Conclusion.
This "re-focusing"of the rhetoricalarts marks the finalphase in
the scholastic process that we have been considering
here. Nihdyal al-Ijdz, the Mifldh.and the may be seen to
constitute three stages in this process insomuch
Talkhs. each work re-
as
presents some advance (or development) over its predecessor. Al-
Sakkaki, for instance, adopts various elements fromNihdyal al-
Ijdz (e.g. the definitionof eloquence and the presentation of the
badf) but he completely rebuilds al-RJzil's organizational struc-
ture. Al-Qazwini, forhis part, uses the same (albeit reduced) for-
mat as the MifIdh, but places 'ilm al-baldghahinto a new pack-
age. Accordingly,all threeauthors are in some way dependent on
their predecessors,but still manage to make some contributionto
the understandingof their subject.
A sense of development is not, however, the most important
aspect of this literature. It is probably more significantthat each

(30) Although plagiarism was a much considered topic among litterateurs,it is

clear that al-Qazwini draws his material fromal-Jurjini. Both presentationsopen,
forexample, by consideringsimilar descriptionsof bravery(AB p. 313; TM p. 409).

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions
 THE MAKING OF A TEXTBOOK 115

author offersan alternate use for the study of eloquence. For

al-Razi and al-Sakkiki 'ilm al-baldghahis not really a goal in itself,
but an important tool in the academic debates of their
time. Nihdyal al-Ijdz forms a sort of "user-friendly"reference
work for the increasinglysophisticated discussion of the Koran,
best represented by al-Zamakhsharl's Kashshdf. Al-Mifidh, on
the other hand, incorporates rhetoricin a comprehensivesystem
with which the author explores the possibilities of semantics in
language. Neither al-RFzi nor al-Sakkkki presents eloquence for
eloquence's sake, as al-Qazwini attempts to do in the Talkhsy.
And yet it is the baldghahof the that survives in the text
centered environment of the Islamic Middle
Talkhs. Ages. We can see
this in the distribution of commentaries that were writtencontin-
ually between the death of al-Sakkfki and the present
day. Nihdyal al-Ijdz did not attract a single work, and while the
Mifldh attracted a 25
very healthy commentaries, the was
the basis for30 commentaries,16 abridgmentsand over 100 super
Talkhs.
commentaries. Indeed, after the sixteenth century al-Qazwini's
work became the almost sole conduit forbringingal-Jurjani's ideas
out of eleventh century Iran and into all parts and ages of the
Islamic world.(31) Accordingly, the medieval reception of the
very rich Islamic hermeneuticstradition is based on al-Qazwini's
packaging of the material in the Talkhis. On this basis alone the
work - and the process that produced it - should merit our
attention today.
William SMYTH
(Arlington,V.A., U.S.A.)

(31) These numbersare based on the bibliographical discussions in Hijji Khali-

fah's Keshf al-Zundn (Istanbul, 1941, 2 volumes), W. Ahlwardt's Verzeichnisder
arabischenHandschriflender KdniglichenBibliothekzu Berlin (Berlin, 1887 ff,v2)
and R. Sellheim's Arabische Handschrifien-Materialien zur arabischen Literatur-
geschichie,Teil 1, vol. XVII, A. 1 of Verzeichnisder orientalischenHandschriftenin
Deutschland(Wiesbaden, 1976, pp. 299-317, passim).

This content downloaded from 139.184.14.159 on Tue, 22 Dec 2015 18:00:43 UTC
All use subject to JSTOR Terms and Conditions