The Theology of Abū l-Qāsim

al-Balkhī/al-Kaʿbī (d. 319/931)

By

Racha el Omari

|
Cover illustration: Interior detail; decoration on so t of arch resting on south pillar (8th-9th century
Abbasid Mosque at Balkh) Copyright Horst P. Schastok, image courtesy of Fine Arts Library, Harvard
University.

Library of Congress Cataloging-in-Publication Data

Names: Omari, Racha Moujir el, author.

Title: The theology of Abu l-Qasim al-Balkhi/al-Kaʿbi (d. 319/931) /
by Racha el Omari.
Description: Boston ; Leiden : Brill, [2016] | Series: Islamic philosophy,
theology and science ; 99 | Includes bibliographical references and index.
| Description based on print version record and CIP data provided by
publisher; resource not viewed.
Identi ers: LCCN 2015046409 (print) | LCCN 2015044807 (ebook) | ISBN
9789004259683 (E-book) | ISBN 9789004259690 (hardback : alk. paper)
Subjects: LCSH: Kaʿbī, ʿAbd Allāh ibn Aḥmad 931 or 932. |
Muslim scholars--Biography. | Islam--Doctrines--History. | Motazilites.
Classi cation: LCC BP80.K28 (print) | LCC BP80.K28 O43 2016 (ebook) | DDC
297.8/34092--dc23
LC record available at http://lccn.loc.gov/2015046409

Want or need Open Access? Brill Open o fers you the choice to make your research freely accessible
online in exchange for a publication charge. Review your various options on brill.com/brill-open.

Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface.

-
- - - - (hardback)
- - - - (e-book)

Copyright 2016 by Koninklijke Brill , Leiden, The Netherlands.

Koninklijke Brill incorporates the imprints Brill, Brill Hes & De Graaf, Brill Nijho f, Brill Rodopi and
Hotei Publishing.
All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system,
or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise,
without prior written permission from the publisher.
Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill provided
that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite
910, Danvers, 01923, .
Fees are subject to change.

This book is printed on acid-free paper and produced in a sustainable manner.

The Attributes

The Muʿtazilīs’ understanding of divine unity (tawḥīd) was shaped by reason

and to a secondary extent by the imperatives of the scriptural account of God.
Reason dictated to the Muʿtazilīs an austere expression of divine unity and
transcendence that found obstacles in the scriptural depiction of God with
many, often anthropomorphic, attributes. The centrality of these obstacles,
and how the Muʿtazilīs sought to resolve them, partially explains why the
Muʿtazilīs’ treatment of divine unity came under the subject heading of the
attributes (al-ṣifāt). The topic of the attributes was recorded among the rst
subjects of concern for nascent mutakallimūn, and it became one of the
Muʿtazilīs’ key de ning subjects; they ranked divine unity as rst among their
ve self-de ning principles.
To their proto-Sunnī and Sunnī kalām opponents the Muʿtazilīs were guilty
of “divesting God of His reality” (taʿṭīl). Ibn Kullāb (died 241/855?) for example,
conceived of God’s reality as expressed in the attributes as “entitative determi-
nants” (maʿānī) in God. Despite the Muʿtazilīs’ agreement on deeming the
“attribute”’ (ṣifa) as only “descriptive” (waṣf), they were far from uni ed on the
issue of accounting for them; thus they developed varied and con icting strat-
egies for this purpose. For Abū Hāshim, the sources document a systematic
and comprehensive methodology that eclipsed the experiments of earlier
Muʿtazilīs. His methodology came to dominate Muʿtazilī discussions of this
subject by future generations of his supporters and opponents alike.
Abū Hāshim’s theory of “states” (aḥwāl) sought to resolve many of the prob-
lems that early Muʿtazilīs struggled with in carving out an account of the attri-
butes that does not compromise divine unity. His theory drew from the

The discussion of the attributes in kalām surfaced in the early second/eighth century with
Jahm b. Ṣafwān and Ḍirār b. ʿAmr. See Frank, Beings and Their Attributes, 10; and van Ess,
Theologie, 425–439.
See, for example, Sayyid (ed.), Faḍl al-iʿtizāl, 213; al-Jishumī, al-ʿUyūn fī l-radd, fols. 12b–13b;
and Mānkdīm, al-Taʿlīq, 149–291.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 169–170, 546–547; al-Ashʿarī, Kitāb al-Lumaʿ, 10–14; Ibn
Fūrak, Mujarrad maqālāt, 38–42; al-Bāqillānī, Kitāb al-Tamhīd, 23–25; al-Juwaynī, Kitāb
al-Irshād, 30–34; al-Baghdādī, Kitāb Uṣūl al-dīn, 88–89.
Daniel Gimaret, La Doctrine d’al-Ashʿarī (Paris: Les éditions du Cerf, 1990), 26–27.
This distinction was rst noted by al-Jubbāʾī, and also adopted by al-Kaʿbī: al-Ashʿarī, Maqālāt
al-islāmiyyīn, 527–528; Frank, Beings and Their Attributes, 17–19.

© , , | . / _
conceptual repertoire of the grammarians. He gave attributes the ontological
weight of “an accusative of state” (ḥāl), thereby eliminating the multiplicity
that resulted from speaking of God as having many predications. But Abū
Hāshim’s solution to the problem resided in expanding the aims of the theology
of attributes to ve categories, which he crafted to describe not only God but
all other beings, including the components of those beings, namely “bodies”
(ajsām), “accidents” (aʿrāḍ), and “atoms” (jawāhir).
The rst category of attributes is a rmed for every being, including God,
and it was called the “attribute of the essence” (ṣifat al-dhāt). The second cat-
egory was the “essential attribute” and it derived from the rst category (li-mā
huwa ʿalayhi fī dhātihi). When this second category of attributes is applied to
God, it refers to the attributes: “powerful” (qādiran), “knowing” (ʿāliman), “living”
(ḥayyan), and “existing” (mawjūdan). While God was considered knowing
because His knowledge is derived from the attribute of the essence, a human
being is knowing because of an “entitative determinant” (maʿnā) in him that is
knowledge. An example of such a category of attributes applying to another
being can be seen in the “occupation of space” (taḥayyuz) being an attribute of
the “atom” (jawhar). The only di ference is that in the case of God this category
is always a rmed, while for other beings, that category must be added to
them. Abū Hāshim’s third category of attributes was this latter kind of attri-
bute that derived from an “entitative determinant”; these attributes inhered, in
the majority of cases, in a “substrate” (maḥall). One attribute of God—the
attribute of volition—was included in this category, though it was not consid-
ered to inhere in a substrate. The fourth category applied to attributes derived
from an “agent” (bi-l-fāʿil) and included the attributes of beings created by God,
as well as attributes of God’s acts, such as His bene cence and mercy. The last
category of attributes was one derived neither from the essence nor from an

The extent to which Abū Hāshim’s conceptualization of the attributes consisted of a full-
scale system was recognized and explained by Richard M. Frank. Abū Hāshim’s concept
of ḥāl for an attribute was developed and borrowed from the Basran grammariansʼ inter-
pretation of the accusative predicate of kāna (Frank, Beings and Their Attributes, 20–24).
Frank, Beings and Their Attributes, 19–20, 27; Ibn Mattawayh, Kitāb al-Majmūʿ fī l-muḥīṭ
bi-l-taklīf, 1:97–112.
Frank, Beings and Their Attributes, 53–55.
The attributes of essence are recognized as these four in Mānkdīm, al-Taʿlīq, 182.
Richard Frank, “Attribute, Attribution, and Being: Three Islamic Views,” in Philosophies of
Existence, Ancient and Medieval, ed. P. Morewedge (New York: Fordham University Press,
1982), 268.
Frank, Beings and Their Attributes, 93–111.
Ibid., 135–136.
“entitative determinant” (lā li-dhāt wa-lā li-maʿnā), and to it belonged God’s
attribute of hearing and seeing that in turn derived from the attribute of living,
but is separate from the attribute of knowing.
Al-Kaʿbī’s views on the attributes were, as we see, at least in part, a continu-
ation of earlier Muʿtazilī experiments on the attributes, which were later
eclipsed by Abū Hāshim’s theology of the attributes. Later biographical dic-
tionaries of ḥadīth transmitters especially note al-Kaʿbī’s view of divine voli-
tion for its a front to the doctrine of divine omnipotence. He was described as
maintaining the stance that God’s volition is equivalent to His knowledge.
Similarly, al-Kaʿbī’s denial of the reality of the attributes of hearing and volition
led Abū l-Qāsim Ismāʿīl al-Bustī (d. late fth/eleventh century) to single it out
as an example of doctrinal di ferences serious enough to lead earlier Muʿtazilīs
to accuse its supporters of unbelief.
Al-Bustī and the later biographical dictionaries of ḥadīth transmitters espe-
cially condemn al-Kaʿbī for his view of divine volition. There is also additional
evidence for al-Kaʿbī’s contributions on the subject of volition as a cosmologi-
cal (daqīq or laṭīf al-kalām) question in its own right. Tracing the potential
in uence of these contributions on his doctrine of the divine attributes, how-
ever, remains beyond the scope of the present chapter. Thus, among the extant
titles of al-Kaʿbī’s lost works, two deal with the subject of volition. These are
the correspondences he exchanged with al-Jubbāʾī: Naqḍ kitāb Abī ʿAlī l-Jubbāʾī
fī l-irāda (Refutation of Abū ʿAlī’s book on volition [irāda]), and al-Kitāb
al-thānī ʿalā Abī ʿAlī fī l-[irāda] (The second book in refutation of Abū ʿAlī
on the subject matter of volition [irāda]). Moreover, parts of Abū Rashīd

Ibid., 148–154.
Aḥmad b. Muḥammad Ibn Khallikān, Wafayāt al-aʿyān, 3:45. Ibn Shākir al-Kutubī includes
Ibn Abī l-Damm’s report, ʿUyūn al-tawārīkh, Micro lm A 408 (American University in
Beirut: Jafet Library), fol. 28a. Al-Dhahabī explained that al-Kaʿbī limited God’s volition to
nothing other than His knowledge (al-Dhahabī, Taʾrīkh al-Islām wa-wafayāt al-mashāhīr
wa-l-aʿlām, ed. ʿUmar ʿAbd al-Salām al-Tadmurī (Beirut: Dār al-Kitāb al-ʿArabī, 1987)),
23:584–585.
al-Bustī, Kitāb al-Baḥth, 24. Al-Bustī also included the attribute of aversion (kārihan) that
is not noted in other sources (see Table 1). Al-Bustī’s nal verdict was that, if unbelief were
based on the serious doctrines he listed, most Muʿtazilīs would be deemed unbelievers.
Hence, he concluded that accusations of unbelief based on grave disagreements in doc-
trine and the conclusions of these earlier Muʿtazilīs should be abandoned (Madelung and
Schmidtke, “Introduction,” Kitāb al-Baḥth, iv–v). Regardless of al-Bustī’s softer position
regarding accusations of unbelief, he listed al-Kaʿbī’s stance on the attributes of volition,
hearing, and seeing as among the doctrines of earlier Muʿtazilīs that warrant a declara-
tion of unbelief (takfīr).
al-Nīsābūrī’s al-Masāʾil fī l-khilāf document al-Kaʿbī’s views on volition from his
work, ʿUyūn al-masāʾil.

al-Kaʿbī’s Precursors on the Attributes

al-Naẓẓām
Where the sources highlight in uences on al-Kaʿbī’s views of the attributes, it is
mostly of al-Naẓẓām’s in uence that they speak. In some cases, Baghdadis, who
remain unidenti ed, were also linked to al-Naẓẓām’s doctrine of the attributes.
Although the sources do not describe them as in uences, there are also cases
where the in uence of al-Naẓẓām on al-Kaʿbī is apparent through kindred ten-
dencies that are outlined here. Al-Naẓẓām’s understanding of the divine attri-
butes had an apophatic character that was often presented as a response to
Abū l-Hudhayl’s formulation. The latter a rmed the reality of independent
attributes but simultaneously reduced them to only a rmations of God; Abū
l-Hudhayl depicted God as “knowing with a knowledge that is He, powerful
with a power that is He, living with a life that is He.” In response, al-Naẓẓām
acknowledged the eternity of the attributes, but used only participles and not
nouns that would imply their existence as independent and separate entities.
He pronounced that they exist only as a rmations of God’s “essence” (bi-nafsihi).
Al-Naẓẓām described God as “eternally (lam yazal) knowing, living, powerful,
hearing, and seeing by His essence (bi-nafsihi) and not by the existence of the

See for example, Abū Rashīd al-Nīsābūrī, al-Masāʾil fī l-khilāf, 363.

See van Ess, Theologie, 3:399–401, and al-Naẓẓām’s position on the attribute of volition,
ibid., 3:401–403.
Abū l-Hudhayl is the rst Muʿtazilī for whom a cohesive formulation of the attributes is
attested. He sought to simultaneously deny the existence of attributes as “entitative deter-
minants” (maʿānī) and to a rm them as identical with God in his famous formula: “He is
knowing with a knowledge that is He, He is powerful with a power that is He, He is living
with a life that is He” [added emphasis] (al-Ashʿarī, Maqālāt al-islāmiyyīn, 165, 484–486;
Frank, “The Divine Attributes”). In a similar vein, he also a rmed the attributes that por-
tray God in anthropomorphic terms, but only as a rmations of God, he thereby negated
their separate, independent realities: “God has a face that is He, thus His face is He, His
essence (nafsuhu) is He. He [Abū l-Hudhayl] used to interpret the mention of Godʼs hand
to mean His blessing (niʿma),” al-Ashʿarī, Maqālāt al-islāmiyyīn, 165. To emphasize the
absence of entitative determinants in these attributes, Abū l-Hudhayl speci cally spoke
of “God being knowing without knowledge, powerful without power, living without life,
hearing without hearing.” al-Ashʿarī, Maqālāt al-islāmiyyīn, 166.
Ibid., 165.
separate individual attributes knowledge, power, life, hearing, and seeing.”
The attributes, as participles, simultaneously a rmed God’s essence and
negated their opposites. For example, to state that God is knowing is to “a rm
His essence (ithbāt dhātihi) and to deny (nafy) that He is ignorant.” Thus,
prompted by the threat of multiplicity in God, al-Naẓẓām a rmed the attri-
butes, but only designated them as participles. As for the relationship of the
attributes to one another, again, to avoid positing multiplicity in God’s essence,
he formulated the di ferences (ikhtilāf) between the attributes of God as di fer-
ences between their opposites.
Yet, al-Naẓẓām’s reference point for speaking of these attributes, and not of
others, was scripture. It was this dictate of scripture that justi ed, in his eyes,
the special status of the attributes of knowledge and power. He recognized
knowledge and power as attributes that derived from His being knowing and
powerful, but he also recognized that they exist in “an absolute fashion,”
namely as attributes of His essence, precisely because of the “absolute” (iṭlāq)
fashion in which they appear in scripture.

We hold that God has knowledge [because] we go back to the [verbal

meaning of the text of scripture] (narjiʿu ilā) that states His being know-
ing. We hold that He has power [because] we go back to the [verbal
meaning of the text of scripture] of His being powerful. This is because
God expressed (aṭlaqa) knowledge in an absolute manner when He said:
“He revealed with His knowledge.” He also expressed in an absolute man-
ner (aṭlaqa) the attribute of power…and did not apply an absolute expres-
sion [when referring to] any other attribute of essence.

While for al-Naẓẓām scripture dictated that the attributes of knowledge and
power should have a less apophatic exposition than other attributes, al-Naẓẓām
claimed that it was also scripture that informed his treatment of “volition”
(irāda) as not actually an attribute, but only a metaphor. Volition is, thus, not
separable from God’s acts and commands: God’s willing of creation (takwīn) is
His creation, and His willing of human acts is His command of human acts.

Ibid., 486.
Ibid., 166–167, 178.
Ibid., 167. In the second part of the same work (486–487), al-Ashʿarī attacked al-Naẓẓām,
stating: “you only a rm God’s essence” (anta lā tuthbitu illā l-dhāt).
Ibid., 187–188. Again in the second part of that work, al-Naẓẓām was quoted as specifying
that God spoke of the attribute of power and knowledge in an absolute manner (aṭlaq), in
contrast to the attributes of hearing (samʿ) and seeing (baṣar), 487.
But, al-Naẓẓām maintained the distinction between the command and the
object of the command.

God’s volition is only (innamā) His act, or His command (amruh), or His
decree (ḥukm). This is [so] because this is the meaning of volition accord-
ing to Arabic usage (fī l-lugha), either it is a secret thought (ḍamīr), or
shows a thing’s position in relation to something else. This [usage can be
observed] in His saying, He is exalted, “a wall that is on the verge of falling
[literally that wishes to fall] but he [al-Khiḍr] erected it.” Since a secret
thought (ḍamīr) cannot be applicable (yastaḥīl ʿalā) to God, the meaning
of His volition has to be what we mentioned. He [al-Naẓẓām] also noted
that the object of volition (al-murād) is designated as volition (irāda) in
Arabic usage (fī l-lugha). Thus it is said: “Bring me my wish (irādatī),
meaning the object of my volition (murādī).” He [al-Naẓẓām] also says:
“God willed judgment day to take place, meaning that He decreed it.”

Clearly scripture did not compel al-Naẓẓām to take a literalist position along
the lines of the “without how” principle (bi-lā kayfa) that was adopted to vary-
ing degrees of exigency by proto-Sunnī and Sunnī mutakallimūn. Indeed, the
attributes of face and hand are not real (fī l-ḥaqīqa) for al-Naẓẓām. They are,
instead, simply expressions, that the hand of God is His blessing (niʿma), and
His face meant that He perdures (yabqā). They are to be understood as meta-
phorical expressions precisely because Arabic usage in the Qurʾān treats them
as such. To regard these descriptions of God as real would be a breach of that
usage.
Among al-Naẓẓām’s views surveyed here, explicit textual support that
al-Kaʿbī followed al-Naẓẓām exists only for al-Naẓẓām’s stance regarding God’s

Qurʾān 18:77.
ʿAbd al-Jabbār, al-Mughnī: al-Irāda, 6/2:3–4. This doctrine is noted in al-Ashʿarī’s Maqālāt
al-islāmiyyīn (190–191) with the added mention that the Baghdadis (who remained
unidenti ed) agreed with al-Naẓẓām on this. See also van Ess, Theologie, 3:401–403, 6:154.
Richard Frank, “Elements in the Development of the Teaching of al-Ashʿarī,” and “The
Science of kalām,” Arabic Sciences and Philosophy 2 (1992): 7–37.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 167. Regarding how to render the attribute of God’s face
(wajh), the Baghdadis and most of the Basrans followed al-Naẓẓām: “The term wajh is
used in the broadest sense (tawassuʿan). We a rm a face that is He (nuthbit wajhan huwa
huwa). This position is justi ed in the linguistic usage of the term face (wajh) in lieu of the
thing itself. Such as someone saying, ‘Had it not been for your face I would have not done
such and such’.” (Ibid., 189).
Ibid.
volition. In al-Naẓẓām’s scripturalist stance there is a signi cant precedence
for al-Kaʿbī’s choice. But the linguistic arguments that al-Naẓẓām invoked for
his position on the attribute of volition and how he formulated the attributes
of knowledge and power were never attributed to al-Kaʿbī, nor was al-Kaʿbī’s
scripturalist stance explicitly tied to al-Naẓẓām. Furthermore, al-Naẓẓām’s
in uence on al-Kaʿbī may have been even more pervasive than indicated by
the speci c similarities of doctrines just reviewed. The similarity and possibil-
ity of al-Naẓẓāmʼs in uence on the Baghdadis is likely, though individual pro-
ponents remain unidenti able, as we see in what follows.

Baghdadi Predecessors
Like other early Muʿtazilīs, the Baghdadis also distinguished between the
attributes of essence and of act. One group relegated God’s volition of cre-
ation to an attribute of act, while the attributes of knowledge, power, life,
hearing, and seeing were deemed attributes of essence. Moreover, their
disagreement in de ning the attributes of act was documented, speci cally
those related to the attribute of generosity. The challenge in determining
the doctrines of the early Baghdadi precursors of al-Kaʿbī on the attributes is
not only that the information is fragmentary, but also that it is sometimes
only identi ed as being that of the general category of Baghdadis.
A work of al-Khayyāṭ on the subject of the “seen” (shāhid) and “unseen”
(ghāʾib) seems to address the methodological concerns underlying his discus-
sion of the attributes. As for al-Khayyāṭ’s views on the attribute of volition,
the available evidence remains inconclusive.
On the formulation of the attribute of essence, the Baghdadis seemed to
agree in principle with al-Naẓẓām in considering the attributes of essence, in the

Ibid., 505.
There are two positions recorded on the Baghdadisʼ understanding of whether jūd (gener-
ous giving) was an attribute of act. ʿĪsā l-Ṣūfī held that His generosity was an attribute of
act but refused to answer the question whether God could be eternally (lam yazal) non-
generous (ghayr karīm). Al-Iskāfī, however, distinguished two kinds of attributes in gen-
erosity, one an attribute of act, when generosity (karam) is generous giving (jūd) and the
other an attribute of essence designating the Being as elevated above other things (ibid.,
178, 506).
See for example, ibid., 504, 508.
Ibn Mattawayh, Kitāb al-Majmūʿ, 165; Madelung, “ʿAbd al-Raḥīm b. Moḥammad b. ʿOṯmān
al-Ḵayyāṭ,” Encycolopedia Iranica, 1:143–144.
Al-Nasafī reports that al-Khayyāṭ agreed with al-Kaʿbī’s doctrine on divine volition, but
the content of al-Kaʿbī’s doctrine that he reports is not accurate, as God’s volition is
described as the absence of God’s forgetfulness (al-Nasafī, Tabṣirat al-adilla, 1:374–375).
form of participles, as eternal attributes. The de nition available was terse and
focused on the negative theology they upheld: “They deny the existence of any
attribute of essence and [they] state that the Maker (al-bāriʾ) is a thing unlike
other things.” In another context, some Baghdadis agreed with al-Naẓẓāmʼs
special treatment of the attributes of power and knowledge, to the exclusion of
the attributes of hearing, seeing, and living because of the way these attributes
appear in scripture. Again, while it is tempting to see this as the position of
al-Kaʿbī, there is no direct evidence to support this identi cation.
Although the early Baghdadis disagreed among themselves about how to
de ne God’s volition, they shared a common concern for separating God’s voli-
tion from His servant’s act and volition. It seems that all Baghdadi and other
Muʿtazilī de nitions of this attribute were in part anchored to theodicean pre-
occupations. The majority of the Muʿtazilīs deemed the attribute of volition as
an attribute of act. Bishr b. al-Muʿtamir alone held that God’s volition was an
attribute of essence, albeit only when the objects of Godʼs volition did not
include evil deeds: “He is eternally willing the [servants’] acts of obedience to
the exclusion of the [servants’] acts of disobedience” (lam yazal murīdan
li-ṭāʿatihi dūna maʿṣiyatih). Bishr postulated two volitions for God, the one
just noted and another as an attribute of act (hiya ʿl min afʿālihi). His signa-
ture “two volitions” spoke of his preoccupation with preserving the integrity of
God’s agency and His servant’s freedom of action.
Other Baghdadi Muʿtazilīs did not opt for Bishr’s concept of two volitions,
but they were equally concerned with de ning God’s volition without compro-
mising Him by association with evil acts. Abū Mūsā l-Murdār (d. 226/841) under-
stood God’s willing of “the servants’ acts of disobedience” (maʿāṣī l-ʿibād) to

al-Ashʿarī, Maqālāt al-islāmiyyīn, 503.

Ibid., 504.
Ibid., 188 and 508.
Ibid., 509. The term attribute of essence was used by al-Ashʿarī in reference to earlier
Muʿtazilī views in his heresiographical accounts on the attributes. For one interpretation
of al-Ashʿarī’s tracing of this distinction to the Baghdadis, see van Ess, Theologie,
4:436–437.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 509.
Ibid., 190, 513.
Bishr b. al-Muʿtamir also disagreed with the majority of the Muʿtazilīs in holding that
creation (khalq) is distinct from the created (al-makhlūq) and that creation was identical
to God’s willing of a particular thing (huwa al-irāda min allāh li-l-shayʾ) (ibid., 364–365).
The identity of His volition and His creation (khalq) are understood to be one since Bishr
b. al-Muʿtamir is reported to regard them as interchangeable in another passage: Bishr
considered His volition to be His creation (yajʿal al-irāda khalqan lahu) (ibid., 510).
mean that He allows them to occur: “He left no barrier between these acts of
disobedience and the servants” (khallā baynahum wa-baynahā). The ambigu-
ity of al-Murdār’s statement, in its apparent a rmation of divine volition, led
al-Ashʿarī to describe al-Murdār as having deemed God to be willing His ser-
vants’ evil deeds. Jaʿfar b. Ḥarb (d. 236/850) understood God’s willing of unbe-
lief and belief as no more than His awareness of the distinction between the
two. God willed them to be di ferent only in the sense that He declared their
“characteristics” (ḥakama) to be distinct from one another, speci cally the char-
acteristics of evil versus good (qabīḥan ghayr ḥasan).
As for the attributes of hearing and seeing, their meaning was open to great
disagreement, and this was documented for the generations following Bishr b.
al-Muʿtamir and Abū l-Hudhayl. Among the Baghdadis, a negative theologi-
cal stance was documented as early as al-Iskāfī, who regarded these two attri-
butes as equal to God’s knowledge.

God is eternally (lam yazal) hearing (samīʿan), seeing (mubṣiran), and

hearing of objects (sāmiʿan), seeing of objects (mubṣiran), He hears
sounds and speech. The meaning of this is that He knows sounds and
speech and that they are not hidden from Him (lā yakhfā ʿalayhi) because
the meaning of hearing (samīʿ) and seeing (baṣīr) according to him
[al-Iskāfī] and whomever agrees with him is that the objects of hearing
and seeing (masmūʿāt and mubṣarāt) are not hidden from Him (lā takhfā
ʿalayhi).

Al-Iskāfī acknowledged the transitive quality of the attributes of hearing and

seeing as being the same as knowing, thereby including the objects of hearing
and seeing. This position of al-Iskāfī seemed to have dominated the Baghdadi
school, and was followed by al-Kaʿbī. Al-Iskāfī was also responsible for

Ibid., 190.
Ibid., 512.
Ibid., 191, 513–514. As to the identi cation of God’s volition and creation (makhlūq), it is
not found in al-Murdār, ibid., 190.
Ibid., 173–174.
Ibid., 175.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 506.
Certain cosmological considerations were taking center stage in Muʿtazilī theology at the
time, and these underlay the choices of al-Iskāfī and other Muʿtazilīs when they spoke
about the objects of God’s knowledge. The most important detail to note is al-Kaʿbīʼs dis-
agreement with al-Khayyāṭ’s controversial view that the non-existent (al-maʿdūm) is not
only a thing (shayʾ) but also a body (jism). Al-Khayyāṭʼs view was universally opposed by
promoting the view that God’s attribute of life is His power; this was also widely
followed by the Baghdadis who remain unnamed and who seem to have held a
variety of opinions on the exact relationship between the attributes of knowl-
edge and power. Al-Kaʿbī did not, however, follow this position of al-Iskāfī.

al-Kaʿbī’s Doctrine of the Attributes

The Attributes of Essence and of Act

Like his contemporary al-Jubbāʾī, and many of his early Muʿtazilī predecessors,
al-Kaʿbī maintained the distinction between the “attributes of essence” (ṣifāt
al-dhāt) and “attributes of act” (ṣifāt al- ʿl). This distinction had shaped discus-
sions among the Baghdadis at least since the time of al-Murdār, though its details
are little documented. Al-Kaʿbī’s distinction between God’s attributes of
essence and His attributes of act, which remain external to God, is documented
most explicitly and clearly in Māturīdī and Bahshamī sources. In these sources,
al-Kaʿbī’s formulation of the attribute of essence is rendered as “by His essence”
(fī nafsihi), in contrast to al-Jubbaʾī’s “because of His essence” (li-nafsihi). Thus
al-Shahrastānī and al-Mufīd’s description of al-Kaʿbī’s formulation of the attribute

Muʿtazilīs, because it could lead to the conclusion that bodies are co-eternal with God
(al-Ashʿarī, Maqālāt al-islāmiyyīn, 504; al-Shahrastānī, Kitāb al-Milal wa-l-niḥal, 53).
Instead al-Kaʿbī understood the non-existent (al-maʿdūm) to be a thing (shayʾ), but nei-
ther an atom (juzʾ) nor an accident (ʿaraḍ) (Abū Rashīd al-Nīsābūrī, al-Masāʾil fī l-khilāf,
37–38). In this al-Kaʿbī still held a minority position, as the majority of Basrans, starting
with al-Shaḥḥām, thought that things, bodies, and accidents are only known by God
before their existence (al-Ashʿarī, Maqālāt al-islāmiyyīn, 162). These cosmological discus-
sions had an immediate in uence on the doctrine of the attributes. For example, al-Jubbāʾī
accepted hearing and seeing as eternal attributes (lam yazal) distinct from knowledge,
but denied them their transitive quality, it seems, to accommodate al-Shaḥḥāmʼs view,
which he followed. That is, al-Jubbāʾī did not deem God to be samīʿan, mubṣiran, because
this would have required the eternity of the objects of hearing and seeing with God (ibid.,
175–176, also 492–493). On the analysis of the non-existent (maʿdūm) as a discussion of
the ontology of the possible, see Richard M. Frank, “The Non-Existent, the Existent, and
the Possible in the Teaching of Abū Hāshim and his Followers,” Mélanges de l’Institut
Dominicain d’Etudes Orientales du Caire 14 (1980): 185–209.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 168.
Ibid., 176–177. The equation of God’s knowledge with God’s power is in some instances
misattributed to al-Kaʿbī (see below).
al-Māturīdī, Kitāb al-Tawḥīd, 78; al-Nasafῑ, Tabṣirat al-adilla, 1:225.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 178, 506.
al-Jishumī, Sharḥ ʿUyūn al-masāʾil, vol. 1, fol. 171b; al-Nasafī, Tabṣirat al-adilla, 1:225.
of essence as “because of His essence” (li-nafsihi) must be rejected. Al-Kaʿbī
characterized the attributes of essence as those attributes whose opposite is
impossible. This is documented by al-Jishumī, al-Nasafī, and al-Māturīdī; this
is a distinction that had been used by earlier Muʿtazilīs to de ne God’s attributes
of essence. However, in al-Māturīdī we nd an additional argument for the
attribute of essence ascribed to al-Kaʿbī, one that is not attested elsewhere.
Al-Māturīdī cites an unnamed work of al-Kaʿbī directly. But al-Māturīdīʼs state-
ment, as promising as it is indicative of a distinct line of argumentation, cannot
be accepted without reservation because, on occasion, al-Māturīdī cites al-Kaʿbī
as a spokesperson of shared Muʿtazilī stances. Below I analyze al-Māturīdīʼs
statements in tandem with accounts of al-Kaʿbīʼs other arguments proposed by
earlier Muʿtazilīs and given by al-Nasafī and al-Jishumī.

The Scripturalist Basis for Knowing the Attributes

While al-Kaʿbī agreed with other Muʿtazilīs that each of God’s attributes (ṣifa)
is merely an attribution (waṣf), unlike al-Jubbāʾī, al-Kaʿbī saw that it was not
through reason but only from scriptural evidence that God’s attributes are
known. This is attested mainly through the statement of al-Jishumī. In treat-
ing the question of the source of human knowledge of God’s attributes, the
attributes are referred to by the word “names” (asmāʾ). The use of the term
“name” (ism), as opposed to attribute (ṣifa) re ects the linguistic concern of
how a name relates to its meaning. Al-Jishumī’s statement not only allows for
the identi cation of the otherwise unidenti ed proponents of this position in
al-Ashʿarī’s Maqālāt al-islāmiyyīn but also for a perspective on its signi cance
for al-Kaʿbī’s theology. This perspective is otherwise lost in ʿAbd al-Qāhir
al-Baghdādī’s statement, as well as, to a lesser extent, by that of al-Mufīd.
Al-Kaʿbī’s scripturalism continued an earlier Muʿtazilī position, one that was
implied in al-Naẓẓām’s scripturalist interpretation of the attributes.

The Attributes of Hearing, Seeing, and Volition

With one exception, the Baghdadisʼ view that God’s attributes of hearing and
seeing—sometimes just described as perceiving—are only His knowledge is

al-Mufīd, Awāʾil al-maqālāt, 12; al-Shahrastānī, Kitāb Nihāyat al-iqdām, 341.

al-Jishumī, Sharḥ ʿUyūn al-masāʾil, vol. 1, fol. 171b; al-Nasafī, Tabṣirat al-adilla, 1:225; al-Māturīdī,
Kitāb al-Tawḥīd, 78.
See, for example, al-Ashʿarī, Maqālāt al-islāmiyyīn, 166–167, 178.
al-Māturīdī, Kitāb al-Tawḥīd, 79.
al-Jishumī, Sharḥ ʿUyūn al-masāʾil, vol. 1, fols. 176b–177a.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 525.
al-Mufīd, Awāʾil al-maqālāt, 13; al-Baghdādī, Kitāb Uṣūl al-dīn, 115–116.
attested across the four theological traditions. Only late Ashʿarī sources ascribe
this article to al-Kaʿbī—all other sources describe it as a general Baghdadi doc-
trine. Thanks to al-Ashʿarī’s Maqālāt al-islāmiyyīn, we know that it was
al-Iskāfī who started it. Other di ferences exist in the statements as well, but
they are minor. Ibn al-Malāḥimī is alone in relating an argument by al-Kaʿbī in
defense of his equation of the attributes of hearing, seeing—and perception—
with knowledge. The argument states that for al-Kaʿbī to say that God is hear-
ing and seeing implies change in Him, and thus hearing and seeing cannot be
accepted as attributes separate from knowledge. In a few cases, late Ashʿarīs,
speci cally al-Juwaynī, link al-Kaʿbī to al-Naẓẓām, in the statement that they
both held that God does not see anything, including Himself. Furthermore,
al-Shahrastānī relates arguments that he alleges to be al-Kaʿbī’s arguments
in defense of the view that the attributes of hearing and seeing are God’s
knowledge.
In the case of al-Kaʿbī’s stance on divine volition, there are four recensions
with signi cant di ferences. But only the rst one can be accepted uncondi-
tionally. In recension A (under Table 1: 5) al-Kaʿbī is described as having fol-
lowed al-Naẓẓām in deeming divine volition as equivalent to God’s act and
command. Minor details exist in this rst recension. Its details are shortened
sometimes to include only the polemical accusation against al-Kaʿbī. For
example, early Ashʿarī sources only speak of God’s volition as not real. In a
longer version of this recension, Godʼs volition is described as His creation of
His own acts, and of His command and decree of His servants’ acts. Further-
more, Ibn al-Malāḥimī’s statement interprets al-Kaʿbīʼs article along the lines
of Abū l-Ḥusayn al-Basrī’s theology.

al-Jishumī, ʿUyūn al-masāʾil, fols. 21a–21b; Sharḥ ʿUyūn al-masāʾil, vol. 1, fol. 148b; al-Mufīd,
Awāʾil al-maqālāt, 13; Mānkdīm, al-Taʿlῑq, 168; Ibn al-Malāḥimī, Kitāb al-Fāʾiq, 36–37;
al-Shahrastānī, Kitāb Nihāyat al-iqdām, 341; al-Bāqillānī, Kitāb al-Tamḥῑd, 253; al-Ashʿarī
Maqālāt al-islāmiyyīn, 175.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 175.
Ibn al-Malāḥimī, Kitāb al-Fāʾiq, 36–37.
al-Juwaynī, Kitāb al-Irshād, 176.
al-Shahrastānī, Kitāb Nihāyat al-iqdām, 341.
al-Jishumī, Sharḥ ʿUyūn al-masāʾil, vol. 2, fol. 191a; al-Mufīd, Awāʾil al-maqālāt, 13; al-Ashʿarī,
Maqālāt al-islāmiyyīn, 509–510.
al-Bāqillānī, Kitāb al-Tamhīd, 252; Ibn Fūrak, Mujarrad maqālāt, 76; al-Baghdādī, Kitāb
Uṣūl al-dīn, 91.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 509–510.
Ibn al-Malāḥimῑ, Kitāb al-Fāʾiq, 42.
 In the second recension (Table 1: 5B) al-Kaʿbī understands God’s volition
to mean His knowledge—and in one single statement, His power. This
recension must be treated as indicative of later developments in the recep-
tion of al-Kaʿbī’s doctrine because this statement that Godʼs volition is His
knowledge veers from al-Kaʿbī’s article that God’s volition is His act or His
command, an article whose ascription to al-Kaʿbī is validated by the agree-
ment of all the testimonies. It should also be treated as a later development
because the arguments reported in this second recension include signi -
cant divergences among themselves. Indeed the statements of this recen-
sion all derive from late Ashʿarī sources. But the reports of this recension
also include one veri ed component of al-Kaʿbī’s theology, namely one
inherited from al-Naẓẓām which states that the meaning of divine volition
is His act and His command. Because of this component, this recension
can be read as consisting of a late “reception” of al-Kaʿbī’s doctrine on the
attributes.
A third recension includes the most strongly and consistently docu-
mented stance attributed to al-Kaʿbī, namely that God’s volition is His act
and command (Table 1: 5C), but this recension simultaneously mixes this
with other theologiansʼ doctrines that do not agree with al-Kaʿbī. Thus the
statements of Mānkdīm and al-Nasafī allege that al-Kaʿbī held that God’s
volition means “the absence of His forgetfulness (sahw),” and this was a doc-
trine of al-Najjār. Fakhr al-Dīn al-Rāzī cites two mixed reports indepen-
dently from one another. In both reports al-Kaʿbī is linked through hybrid
articles to other gures, namely al-Jāḥiẓ and Abū l-Ḥudhayl, though he has
little in common with them on the doctrine of volition. The fourth and
last recension (Table 1: 5D) is also mixed: al-Ījīʼs statement links al-Kaʿbī
with al-Jāḥiẓ and Abū l-Huhdayl. Clearly these mixed recensions cannot
be used to reconstruct al-Kaʿbī’s theology, rather they re ect Ashʿarī theo-
logical concerns in the course of the sixth/twelfth to eighth/fourteenth
centuries.

al-Shahrastānī, Kitāb Nihāyat al-iqdām, 238; al-Shahrastānī, Kitāb al-Milal wa-l-niḥal, 43–
44; al-Juwaynī, Kitāb al-Irshād, 63–66; al-Zabīdī, Itḥāf al-sāda, 2:141; Ibn al-Malāḥimī,
Kitāb al-Fāʾiq, 42.
For example, compare the arguments in al-Shahrastānī, Kitāb al-Milal wa-l-niḥal, 43–44 to
al-Zabīdī, Itḥāf al-sāda, 2:141.
For example, al-Shahrastānī, Kitāb al-Milal wa-l-niḥal, 43–44.
ʿAbd al-Jabbār, al-Mughnī: al-Irāda, 6/2:5.
al-Rāzī, al-Maṭālib al-ʿāliya, 3:179; al-Rāzī, Kitāb al-Arbaʿīn, 147.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 189–190, 363–364, 510.
 God’s Speech and God’s Relation to Place
While Muʿtazilīs stipulated that God’s speech is created, there were some dis-
agreements on where the accident of speech took place. Al-Kaʿbī followed an
earlier Baghdadi Muʿtazilī stance promulgated by Jaʿfar b. Ḥarb and Jaʿfar b.
Mubashshir (d. 234/849) that stipulates that the Qurʾān as a created accident
could not exist in more than one place. It only exists in the eternal tablet. All
other manifestations of the Qurʾān, meaning its recitation, writing, and hear-
ing, were only imitations (ḥikāya) of the letters created on the tablet, the object
of imitation (maḥkī). In this al-Kaʿbī followed the earlier Muʿtazilī position of
Jaʿfar b. Ḥarb and Jaʿfar b. Mubashshir, and he disagreed with al-Iskāfī’s posi-
tion that was subsequently followed by al-Jubbāʾī. The latter understood the
Qurʾān as an accident that exists in more than one place at the same time.
Thus al-Jubbāʾī, following al-Iskāfī, also thought that each recitation is identi-
cal with the created accident of the Qurʾān. This is documented by al-Nasafī,
whose statement allows for the identi cation of a similar statement in
al-Ashʿarī’s Maqālāt al-islāmiyyīn.
In another recorded position on the attributes, al-Kaʿbī also favored an ear-
lier Muʿtazilī stance. He followed the minority position of Hishām b. ʿAmr
al-Fuwaṭī and his student ʿAbbād b. Sulaymān al-Ṣaymarī (d. c. 250/864) in
believing that God is not constituted spatially. In this al-Kaʿbī disagreed with
the position of earlier Baghdadi Muʿtazilīs, including the two Jaʿfars, and
al-Iskāfī, as well as the Basrans Abū l-Hudhayl and al-Jubbāʾī, who both consid-
ered God to be in every place in the sense that He is “managing” (mudabbir)
every place. It is al-Māturīdī who allows us to identify al-Kaʿbī among the
group of unidenti ed Muʿtazilīs noted in Maqālāt al-islāmiyyīn who, in this
regard, are believed to have followed this minority position. Moreover,
al-Māturīdī’s statement added one further detail to al-Kaʿbī’s endorsement of
this position, al-Kaʿbī explained God’s relation to place as one of “knowing it”
(ʿalā maʿnā annahu ʿālim) and “preserving it” (ḥā ẓ lahu). This relationship was
comparable to the role of a person in relation to the building of a house, in the

al-Nasafī, Ṭabṣirat al-adilla, 1:288; al-Ashʿarī, Maqālāt al-islāmiyyīn, 193.

al-Nasafī, Ṭabṣirat al-adilla, 1:260–261, 286; al-Ashʿarī, Maqālāt al-islāmiyyīn, 192.
van Ess, Theologie, 4:1–4.
Sulaiman Mourad, “ʿAbbād b. Salmān,” Encyclopaedia of Islam, Three, online edition, pub-
lished 2009.
al-Māturīdī, Kitāb al-Tawḥīd, 115.
“God is in every place (bi-kulli makān) in the sense that He is knowing (mudabbir) of every
place and that His knowledge (tadbīr) is in every place” (al-Ashʿarī, Maqālāt al-islāmiyyīn, 157).
sense that he “participates in its creation” (kamā yuqāl fulān fī bināʾ al-dār ay fī
ʿlihi).

The Case of the Basran Muʿtazilī and Māturīdī Testimonies:

Attributes of Essence
Al-Kaʿbī’s formulation of the attribute of essence is recorded in his answer to
the following question: How do we know that an object is the act of a divine
agent? As reported by al-Nasafī, al-Kaʿbī’s answer is evidence of his disagree-
ment with his Basran counterparts’ understanding of the attribute of essence.

A perfect (muḥkam) and sound (mutqan) act proves (yadullu) that its
agent (fāʿiluh) is knowing and powerful only (faḥasb). Then we apply
rational inquiry (naẓar) to [the result of the earlier proof], [if we nd]
that ignorance (jahl) and incapacity (ʿajz) are admissible to [describe]
the agent, [we learn] that He is knowing with a knowledge (ʿālim bi-ʿilm)
and powerful with power (qādir bi-qudra). If [we nd that] ignorance
and incapacity and all the contraries of knowledge and power are inad-
missible for him (istaḥāla ʿalayhi), He is [proven to be] knowing by His
essence (bi-nafsihi), powerful by His essence (bi-nafsihi).

Thus a perfect act is the starting point for the proof that knowledge constitutes
its agent’s attribute of essence. The agent of this perfect act is then subjected to
rational inquiry to examine if the contrary of knowledge and power can char-
acterize him. When the result of this examination is positive, then the agent is
declared as having an essence according to an entitative determinant, be it
knowledge or power. But when the result of the examination is negative—that
is, when the agent cannot be characterized by the opposite of the attributes of
knowledge and power—He is then declared to be knowing and powerful by
His essence (bi-nafsihi). This is how al-Kaʿbī arrived at the conclusion that
God’s attributes of essence are those that cannot tolerate their opposite.
For al-Jubbāʾī, God’s knowledge was distinguished from that of other agents
in the sense that while they know because of an “entitative determinant”
(maʿnā) of knowledge, God knows “because of His essence” (li-nafsihi).

al-Māturīdī, Kitāb al-Tawḥīd, 115.

This framework has long been acknowledged to have dominated the extant material on
the doctrine of the attributes. See Dhanani, Physical Theory of Kalām, 25.
al-Nasafī, Tabṣirat al-adilla, 1:225 [emphasis added].
 If an object of an act is perfect (muḥkam), it provides evidence both in the
seen (shāhid) and unseen (ghāʾib) world that its agent is knowing (ʿāliman)
and powerful (qādiran). In the seen world, the agent is knowing and
powerful with knowledge and power, thus the perfect object of the act is
a proof for both of them [the entities of knowledge and power].
In the unseen world, the knowing and powerful agent is knowing and
powerful because of His essence (li-nafsihi). Thus, [knowing and being
powerful] are proofs for [the existence] of His essence (fa kāna dhālika
dalīlan ʿalā dhātih).

Al-Jishumī also accounts for al-Kaʿbī’s formulation and understanding of the

attribute of essence from this epistemic lens. Just as al-Jishumī disapproved of
the Ashʿarī position that a servantʼs understanding of God’s attribute of knowl-
edge is derived from the “entitative determinant” of knowledge in Him, he also
disapproved of al-Kaʿbī’s understanding of God’s knowledge as by “His essence”
(fī nafsihi). Instead al-Jishumī endorsed Abū Hāshim’s perception of God’s
attribute of knowledge through His state of knowing.

Abū Hāshim, and those who followed him, upheld the view that knowl-
edge of His being knowing, powerful, and living is neither derived from
(yataʿallaq bi) His essence only, nor from an entitative determinant that
is other than Him, but rather it is correlated to His essence in a state (bi
dhātihi ʿalā ḥālatin) … Abū ʿAlī [al-Jubbāʾī] upheld this position in [some]
instances. However, though he upheld the term ḥāl (lafẓ al-ḥāl) in his
[work] Jawāb al-Khurāsāniyya [Response to the Khurāsānīs], he men-
tions a divergent term (bi-khilāf dhālik) in other instances. Abū l-Qāsim
[al-Kaʿbī] held that [knowledge of His being knowing, powerful, and liv-
ing] is knowledge by His essence (ʿilm bi-dhātihi).

Although al-Jishumī did not describe al-Kaʿbī’s understanding of God’s attri-

butes of essence as those that cannot tolerate their opposite, as al-Nasafī does,
he criticized the epistemic consequences of al-Kaʿbīʼs understanding. Holding
that knowledge of God’s attribute of essence is correlated (mutaʿalliq) to His

Ibid.
Daniel Gimaret, “Matériaux pour une bibliographie des Jubbāʾī: note complémentaire,” in
Islamic Theology and Philosophy, Studies in Honor of G.F. Hourani, ed. M.E. Marmura
(Albany: State University of New York Press, 1984), 33.
al-Jishumī, Sharḥ ʿUyūn al-masāʾil, vol. 1, fol. 171b.
Ibid., vol. 1, fols. 171b–172a.
essence amounts to not really knowing anything about God. Al-Jishumī
explained that there are two conditions that remain unmet in such a scheme.
First, knowledge of two separate things cannot be identical. Second, speci c
and general knowledge are distinct from one another. If knowledge of God’s
essence is tied to (mutaʿalliq) knowledge of His attributes, then this is the same
as maintaining that knowledge of God’s existence is equal to knowledge of His
attributes. This was unacceptable to al-Jishumī because these are two separate
types of knowledge, namely speci c versus general knowledge about God.
Al-Jishumī thought that al-Kaʿbī’s formulation of knowledge of the attribute of
the essence did not lead to any speci c knowledge about God, other than that
He is an unknowable essence.
Al-Māturīdī provides a more detailed account of al-Kaʿbī’s understanding of
the attributes of essence, starting with his account of how an attribute of
essence is di ferent from an attribute of act. An attribute of act—as opposed to
an attribute of essence—is characterized by three features. First is its tolerance
of discord, such as the attribute that God is “compassionate” (raḥmān), or
“sustaining” (rāziq). God could be compassionate in one case and not com-
passionate in another; He could sustain one servant but not another. Both
examples of the attributes of act express “discord” (ikhtilāf), the rst spelled out
as “discord in state (ḥāl),” meaning in God’s state, and the second as discord in
relation to the person to whom God’s attribute is directed. Second, an attribute
of act is dominated by God’s power: God could speak in one instance and not in
another. He is not necessarily held to be speaking all the time. The third fea-
ture of al-Kaʿbī’s understanding of the attribute of act is the existence of its
opposite, such as the opposite of the attribute of mercy. The attributes of act are
external to God, they are derived from His power and can be opposed to one
another, discordant with one another.
“Discord” is the criterion that de nes the attributes of act, while its oppo-
site, harmony, which remains unnamed but implied in this passage, de nes
the attributes of essence:

The attribute of act admits (iḥtamala) the possibility of discord with

regard to God’s state (ḥāl) and relation to individuals. As in the state-
ment: “He blesses so and so (fulān)” and “He shows mercy in one state
(ḥāl) and does not show it in another state.” Such is the case of the

al-Māturīdī, Kitāb al-Tawḥīd, 78.

Ibid.
Ibid.
 attribute of speech in its [varying] states and [varying] relations to
individuals.

Conversely, an attribute of essence is identi ed rst by the impossibility of tol-

erating any discord; second, it could not depend on God’s power (qudra); and
third, it is identi ed by the fact that its opposite could not exist. Underlying
al-Kaʿbīʼs characterization of the attribute of essence, namely that it could not
include contradictions, is the assumption that having discordant attributes
amounts to implying that there could be discord in God.

This admission of [discord] is not [valid] for the [attributes] of power,

knowledge, and life, for they are attributes of essence. … [Each attribute]
which occurs by power (kul mā yaqaʿ ʿalayhi al-qudra) is an attribute of
act, such as mercy and speech. The [attribute] which does not occur by
power is an attribute of essence. For example, it is not said: “Is He capable
of knowing or not?” Then when [al-Kaʿbī] is asked regarding the attribute
of essence: “Why is it not possible to describe [God] by the contrary of
the attribute of essence.” He [al-Kaʿbī] would respond: “Because it [an
attribute of essence] derives from His essence and His essence is not dis-
cordant (mukhtalif), for [upholding that an attribute of essence is discor-
dant] implies discord (ikhtilāf) in God.”

Finally, al-Kaʿbī spelled out the principle of the “absence of discord” as a prin-
ciple of “internal necessity” that persists for as long as His essence endures.
“Since His essence (dhāt) is not discordant, discord is not possible as long as
His essence (nafs) lasts, just as a thing (shayʾ) is necessary because of a cause
that endures as long as it [a thing] lasts.” Al-Kaʿbī compared the relationship
of the necessity of a thing (shayʾ) to an internal cause, which persists for as
long as that thing exists, to the relationship of lack of discord with God’s
essence. The comparison implied that lack of discord was due to an internal
cause that continues for as long as He exists. While the function of the “absence
of discord” was explained through this comparison as a necessary internal
cause, it was not identi ed categorically.
This necessary internal cause rst described as “absence of discord” appears
again in al-Māturīdī’s statement of al-Kaʿbī’s explanation of the attribute of

Ibid.
Ibid.
Ibid. [emphasis added].
essence. This time, however, the “absence of discord” is worded in a rmative
language. It is an internal necessary cause, and attributions are derived from it.

He [God] was named (summiya), in reality (fī l-ḥaqīqa), knowing (ʿālim),

creating (khāliq), and powerful (qādir) by way of veri cation (fī l-taḥqīq).
Thus, God is not de ned based on His attributes (fa lā wajh li-taʿrī hi min
ḥaythu wuṣifa), for the reality of His attributes [and names] is derived
from that in which harmony exists (idh ḥaqīqatuhu mā tarjiʿu ilā mā fīhi
al-wifāq).

Of course, like other Muʿtazilīs, for al-Kaʿbī the attributes and names were
words, attribution (waṣf) or (tasmiya), and not attributes (ṣifa). These attri-
butes, speci cally those of essence, derived from “that in which harmony
exists” (mā fīhi al-wifāq). Just as the “absence of discord” was not recognized as
a category, though its function was, the same was true in the case of “that in
which harmony exists.” Indeed, aside from being described as “an internal nec-
essary cause,” the ontological status of “that in which harmony exists” cannot
be uncovered from these passages.
Al-Naẓẓām and other early Muʿtazilīs had already spoken of the attribute of
essence as negating its opposite and a rming the essence of God. But none
of their descriptions attest to the mention of “necessity” when describing the
attribute of essence, and most importantly there was no mention of a principle
of harmony or absence of discord. As for his Basran counterparts, both
al-Jubbāʾī and Abū Hāshim found fault in al-Kaʿbīʼs formulation of the attri-
butes of essence as by His essence (fī nafsihi), and thus could not have favored
his explanations of what it was. With the little information available about the
arguments used by early Muʿtazilīs for understanding the attribute of essence,
al-Māturīdī’s statement cannot be taken as nal evidence for the uniqueness
of al-Kaʿbī’s use of the notion of an “internal necessary cause” to de ne the
attribute of essence.
Notwithstanding that al-Kaʿbī’s sources remain unknown and that his ter-
minology here must be documented further, what is attributed to al-Kaʿbī by

Ibid., 79 [emphasis added].

“Al-Kaʿbī held that God has, in reality, no attribute, rather it [what is called the attribute]
is the attribution (waṣf) of the one making an attribution (wāṣif) or the naming (tasmiya)
of the one who names. These two propositions (amrān) are applicable in the quali cation
of the quali ers whenever they qualify Him with knowledge, power, and action without
any disagreement with regard to attribution (min haythu al-waṣf)” (ibid., 79).
al-Ashʿarī, Maqālāt al-islāmiyyīn, 166–167.
al-Māturīdī as a principle of internal necessary cause that consists of inner
harmony is a promising place to follow up with Robert Wisnovsky’s hypoth-
esis on the origins of Ibn Sīnāʼs (d. 428/1037) concept of the “necessary of exis-
tence.” In his theory, Wisnovsky argues that Ibn Sīnāʼs concept of the “necessary
of existence” (wājib al-wujūd) to describe God was the response of the Sunnī
mutakallimūn to what he calls the speci c problems of the attributes, and the
attribute of eternity in particular. He also notes that their response may have
predated Ibn Sīnā among the Sunnī mutakallimūn and draws attention to
al-Kaʿbī as a possible origin of that response. With this hypothesis, al-Kaʿbī’s
doctrine on the attributes stands a chance of being among the earliest expres-
sions of philosophical material to resolve kalām-speci c problems.
Moreover, the term wifāq (harmony) used in al-Kaʿbī’s de nition of the
attribute of essence also requires further investigation. In his discussion of
causality, the philosopher al-Kindī used synonyms of the notion of harmony;
speci cally he used the term taʾlīf (union), iʿtidāl (equilibrium), or
mutanāsiba (proportion). Of course, Abū Zayd al-Balkhī, in the lineage of
al-Kindī’s school, was closely acquainted with al-Kaʿbī, and may have exposed
him to sources that in turn may explain al-Kaʿbī’s use of the notion of harmony
to account for the Muʿtazilī distinction between the attributes of essence and
act. This textual parallelism, though, as striking as it may be, does not, on its
own, establish historical in uences.

The Case of the Basran Testimony versus ʿAbd al-Qāhir al-Baghdādī:

Scripture and God’s Names
Al-Kaʿbī asserted that scripture is the only source for knowing God’s names.
For al-Kaʿbī, this meant that if a name of God could be expressed in two syn-
onyms, the choice of which one is acceptable would be the one grounded in
scripture. In this, al-Kaʿbī disagreed with al-Jubbāʾī and the Basrans, who
deemed reason alone to be a su cient source for knowing God’s names.
Al-Kaʿbī’s position was not based on a linguistic principle that meanings do

Robert Wisnovsky, “One Aspect of the Avicennian Turn in Sunnī Theology,” Arabic
Sciences and Philosophy 14 (2004), 88.
Peter Adamson, al-Kindī (Oxford: Oxford University Press, 2007), 172.
Ibid., 162.
Ibid., 163.
Frank, Beings and Their Attributes, 15. For a discussion of al-Jubbāʾī’s adoption of reason as
su cient for knowing the attributes, and its in uence on Abū Hāshim’s stance that the
origin of language is convention (muwāḍaʿa) and not revelation (tawqīf), see Sophia
Vasalou, “Their Intention was Shown by Their Bodily Movements,” Journal of the History
of Philosophy 47 (2009): 201–221.
not exist separately from words, or that a particular meaning could only be
expressed in one word. Rather it was based on the view that God’s names must
only be determined by scripture. Indeed al-Kaʿbī upheld the same view of lan-
guage, as known by convention (muwāḍaʿa), that al-Jubbāʾī upheld, but he
stopped short of applying this rule to God’s names speci cally.

The Baghdadis disagreed with him [al-Jubbāʾī]—they claimed that it is

not permissible to name God, He is exalted and praised, by a name on the
basis of reason, unless God named Himself by it. They claimed that the
meaning of [the attribute] knowing (ʿālim) is equal to being cognizant
(ʿārif), but that we name Him knowing (ʿālim) because He named Himself
by it [this name] and we do not name Him cognizant (ʿārif). Similarly the
meaning of [our] saying [He is] perceiving (fāhim) and rational (ʿāqil)
means knowing (ʿālim), but we do not name Him by them [by the names
fāhim and ʿāqil]. Similarly the meaning of “He becomes angry” (yaghḍab)
is equal to “He becomes resentful” (yaghtāẓ), but the term resentful is not
used [by us]. Similarly, the meanings of eternal (qadīm) and ancient
(ʿatīq) are equal [but we do not use the name ancient].

Al-Jishumī attributed to some of al-Kaʿbīʼs followers an explanation more

detailed than the general invocation of usage: These followers contended that
there are names of God that contradict reason, such as His being both the
apparent (al-ẓāhir) and the hidden (al-bāṭin). There are also names by which
He is glori ed, and they too can be known only through scripture. For these
followers of al-Kaʿbī, the resolution to these instances in which God’s names
contradict reason lay, therefore, in proclaiming complete dependence on
scripture as the only reliable source for knowing God’s names. Al-Jishumī sug-
gested an alternative resolution to these two. He reframed the options they
depicted: Not all divine names must be known from scripture just because
some of them can be. Instead, al-Jishumī proposed two categories of divine
names, those that are based on scriptural knowledge and those that are known
by both reason and scripture. Unfortunately, given the polemical nature of

Al-Jubbāʾī held that reason should su ce for knowing a given name of God and that God’s
names could not be known by scriptural postulation (talqīb). This is based on al-Jubbāʾī’s
stance that any given meaning was independent of the word that represented it, with
only human convention tying them together (al-Ashʿarī, Maqālāt al-islāmiyyīn, 525).
See Table 1: 3A1, where this position is identi ed as al-Kaʿbī’s.
al-Ashʿarī, Maqālāt al-islāmiyyīn, 525.
al-Jishumī, Sharḥ ʿUyūn al-masāʾil, vol. 1, fol. 177a.
al-Jishumī’s report, al-Kaʿbī or his followers’ potential retort to this view was
never disclosed.
Moreover, al-Kaʿbī was accused of inconsistency because he deemed reason
su cient to rule out what could not be considered divine attributes. He was
also accused of inconsistency when he accepted the Basran linguistic principle
that meaning could be expressed in di ferent words and that meaning was
apprehended through reason, yet he did not apply these two principles to
scripture. Al-Kaʿbī also shared with the Basrans the principle that language
is known by convention (muwāḍaʿa), but refrained from applying it in the case
of the divine names. What is framed as inconsistency by al-Jishumī only illus-
trates how far al-Kaʿbī’s scripturalism went: he made exceptions to divine
names over other linguistic matters.

[According to al-Jishumī] A name (ism) is an attribution (waṣf) that con-

veys a meaning (fāʾida), and it [the name] is a reality—in so far as its
conveyance of meaning is concerned (huwa ḥaqīqa fī tilka al-fāʾida)—
and this meaning (fāʾida) is valid in its reference to the Eternal, He is
exalted. Thus the use of this term (ʿibāra) to describe God is permissible
(yajūz) without [the permission] of scripture (samʿ). Abū l-Qāsim
[al-Kaʿbī] held that this [i.e., the iṭlāq of the ʿibāra about God without the
permission of scripture] is not permissible.… [Further, al-Kaʿbī held that]
whatever [name] bears (iḥtamala) two meanings (maʿnāyayn), one cor-
rect and the other false, both meanings require a scriptural authorization
(idhn samʿī) [by agreement (bi-l-ittifāq)].

Al-Kaʿbī’s choice of scripture as the source of knowledge of the divine names

was also acknowledged by ʿAbd al-Qāhir al-Baghdādī. But according to
al-Baghdādī, al-Kaʿbī included the sunna and consensus under the Qurʾān
(al-kitāb).

The Basran claimed that God’s names (asmāʾ Allāh) are derived from
usage (iṣṭilāḥ) and analogy (qiyās). The consensus of the ahl al-sunna is
that God’s names are derived from revelation (tawqīf) and that it is not
possible to name God by a name based on analogy (min jihat al-qiyās),
rather He is named by that which appears in the book (al-kitāb),

Al-Jishumī understood that even the names by which God is described in ritual prayers
can also be known by reason.
al-Jishumī, Sharḥ ʿUyūn al-masāʾil, vol. 1, fols. 176b–177a.
 the correct sunna, or the consensus of the community. Al-Kaʿbī
followed them [the ahl al-sunna] in this matter.

But al-Baghdādī’s claim about the meaning of scripture for al-Kaʿbī cannot be
accepted without contradicting what is known about the limits of al-Kaʿbī’s
toleration of consensus and the sunna, and the limits he put on the Qurʾān as
the basis for theological knowledge. Furthermore, al-Mufīd, who associ-
ated the majority of Baghdadis with other groups (such as the Zaydīs, Murjiʾīs,
and the ahl al-ḥadīth), gives no details on what scripture meant for the
Baghdadis when they accepted scripture as a source for knowing the attri-
butes. Al-Kaʿbī’s scripturalism, as attested by al-Jishumī and the Maqālāt
al-islāmiyyīn, invites parallels between him and the ahl al-sunna. But
al-Baghdādī’s claim that al-Kaʿbī’s position on scripture for knowing God’s
names is equal to that of the ahl al-sunna is ultimately misleading, as it entails
many contradictions of al-Kaʿbīʼs theological commitments that are known to
us with certainty.

The Case of the Late Ashʿarī Testimony

As we already established, late Ashʿarī sources document spurious recensions
of al-Kaʿbī’s understanding of the meaning of the attributes of hearing and
seeing on the one hand and the meaning of the attribute of volition on the
other. These same recensions, however, include otherwise valid arguments
from al-Kaʿbīʼs epistemology and cosmology that warrant close examination.

God’s Hearing and Seeing

Al-Shahrastānī claims that al-Kaʿbī argued for the position that the attributes
of hearing and seeing are not separate from God’s knowledge; he argued this,
initially at least, on the basis of the Ashʿarī principle of the applicability of the
seen world (shāhid) on the unseen world (ghāʾib). On the basis of this princi-
ple, al-Kaʿbī was said to have proven that hearing and seeing are not separate
from knowing in the seen world (shāhid), and therefore not separate from one
another in the case of God as well. First al-Kaʿbī had to establish that humans
do not really hear and see the objects of knowledge because the objects of
knowledge can only be known with reason and the heart.

al-Baghdādī, Kitāb Uṣūl al-dīn, 115–116 [emphasis added].

al-Kaʿbī, Qabūl al-akhbār, 1:17; El Omari, “Accommodation and Resistance.”
al-Mufīd, Awāʾil al-maqālāt, 13.
al-Shahrastānī, Kitāb Nihāyat al-iqdām, 343.
 A human being nds in himself (yajiduhu min nafsihi) his perception of
the objects of hearing and seeing with his heart and reason. But a [human
being’s] sight does not sense the object of sight rather he senses (yuḥissu)
the object of hearing, but not the ear.

For al-Kaʿbī, the ears and eyes were only media (wasāʾiṭ) that mediated the
production of objects of knowledge; they did not apprehend the end result of
their toil. Their products—what al-Kaʿbī refers to as the objects of seeing and
hearing—are seated in the heart and in reason. This description of knowledge
implies that both a nal seat and a medium of production are required, and
that these two are distinct from one other.
Once he had established the distinction between the medium and the seat for
the production of knowledge, al-Kaʿbī provided a more speci c account of the
mechanism of human knowledge, in order to preface its distinction from divine
knowledge. A human being is in need of the speci c medium of hearing and see-
ing to attain the object of knowledge, even when that object can only be appre-
hended by reason and the heart. “Because knowledge is not realized (yaḥṣulu
lahu) for the human being except by means of his sight, the latter was labeled a
[separate] sense (ḥāssa). Otherwise, the perceiver (al-mudrik) is the knower
(al-ʿālim) and his perception is not additional to his knowledge.” Just as hearing
and seeing are media (wasāʾiṭ) for attaining knowledge, there are also other
media for attaining knowledge. These media do not, however, a fect the quality of
knowledge, which a knowing person would nd in himself (yajiduhu fī nafsihi).

The proof for this [that di ferent media of knowledge do not a fect the
nal result of knowledge] is that whoever knows something by means of
a report then sees it with the eyes discovers a di ferentiation (tafriqa)
between the two states [yielded by the two media of knowledge]. Except
that this di ferentiation is not one of class (jins) or species (nawʿ) but
rather one between the general (jumla) and the speci c (tafṣīl), univer-
sality (ʿumūm) and particularity (khuṣūṣ), and the absolute (iṭlāq) and
concrete particular (taʿyīn). Otherwise the state of the soul (nafs) is iden-
tical in both cases (ḥālatayn).

Knowledge remains the same even when the medium is di ferent; it consists of
the same class (jins) or species (nawʿ) regardless of its medium. There are,

Ibid.
Ibid.
Ibid.
however, quantitative di ferences in knowledge generated by di fering media.
These non-qualitative di ferences pertain to the order of the general versus the
speci c. An example of this is transmitted knowledge yielded by a report: this
was deemed equal to knowledge yielded by other media in what pertained to
class, but not in what pertained to the general and the speci c.
Al-Shahrastānī noted another of al-Kaʿbīʼs arguments in support of the
independence of the object of knowledge from the medium of hearing and
seeing. Al-Kaʿbī argued that if objects of hearing and seeing result from the
existence of hearing and seeing as independent attributes, an illogical result
would ensue. This illogical result would be that each time a person knows
something—in other words each time knowledge is produced—what was
simply an instrument of this knowledge would have to be produced along with
the knowledge itself. Because the medium of the production of an object of
knowledge was not always present with its object, al-Kaʿbī concluded that
hearing and seeing did not really exist; they were just media of knowledge.

[Al-Kaʿbī added] that should the perceiver perceive with perception [i.e.,
not with knowledge as al-Kaʿbī upheld] a musical instrument being
played, freely moving animals, drummed drums, and blown (tunfakh)
images would have to be present by means of an intact sense (al-ḥāssa
al-salīma) in the perceiver, but [we know that] he [the perceiver] does
not see them or hear them, for God did not create these perceptions for
him. In the same way, [should the perceiver perceive with perception],
it would have been possible [for the perceiver] to see a person in the dis-
tance and not see someone who is close because the perception of
distance was created for him while [the creation of the close person] was
not. We already know by necessity (ʿalimnā ḍarūratan) that the truth of
the matter (amr) is contrary to this [hypothesis].

This passage invokes the idea that knowledge can be attained by di ferent
means, and this supports al-Kaʿbī’s position that Godʼs hearing and seeing are
means of knowledge, and not separate attributes. Moreover, this view that
knowledge can be attained by di ferent means is a central tenet of al-Kaʿbī’s
epistemology, and led him to tolerate imitation (taqlīd) as a means for attain-
ing knowledge.

Ibid.
Ibid., 344.
See Chapter 4.
 Divine Volition
Al-Kaʿbī adopted al-Naẓẓām’s view that God has no volition in reality, that His
volition means His act and His command of His servant’s act. Yet, the premise
that God’s volition is His knowledge (and is sometimes described as from eter-
nity) is a late addition and thus a late Ashʿarī reception and interpretation of
al-Kaʿbī’s original article.
Al-Shahrastānī argued against this alleged position of al-Kaʿbī as follows:
Since al-Kaʿbī and al-Naẓẓām conceded that volition was a class of accidents
(jins min al-aʿrāḍ) based on the speci cation (ikhtiṣāṣ) of one act in distinction
from another in the seen world (fī l-shāhid), the same distinction should follow
for the unseen world (al-ghāʾib). Al-Kaʿbī hypothetically responded by
explaining that volition is necessary for humans because of the limitations of
human knowledge and power. This is how al-Kaʿbī is cited as describing the
function of volition in the human agent.

In the seen world (shāhid) speci cation (ihktiṣāṣ) is proof for [the exis-
tence of] volition because the agent’s (fāʿil) knowledge neither encom-
passes all the aspects (wujūh) of the act, nor the objects that are unseen,
nor the time (waqt) and amount (miqdār), hence he [the agent] is in
need of intent (qaṣd) and determination (ʿazm) in order to specify [the
choice of] one time (waqt) rather than another, and [one] amount
(miqdār) rather than another.

The elements that are necessary for the realization of an act by a human agent
are speci cation, intent, determination, and the choice of one moment in
time. These same elements have no place for an act of God. This is because His
knowledge and power renders these faculties unnecessary.

The Maker (bāriʾ), He is exalted, is knowledgeable of the unseen [worlds]

(ghuyūb), overseeing (muṭṭaliʿ) their secrets, and characteristics (aḥkāmihā)
such that His knowledge of them [the ghuyūb and their secrets] is satisfac-
tory; it [i.e., His knowledge] makes volition and intent (qaṣd) for speci ca-
tion (takhṣīṣ) unnecessary. [This is so] because He knows that every
accident is speci ed with (yakhtaṣṣu) time, form, and power (qudra), and
that only what He knows exists. So what need does He have [i.e., He has no

This is based on the Ashʿarī principle guiding the relationship between the world of the
seen (shāhid) and the world of the unseen (ghāʾib) (al-Shahrastānī, Kitāb Nihāyat
al-iqdām, 239).
Ibid.
 need] for intent (qaṣd) and volition... Thus it is established (taʿayyana)
that the Eternalʼs volition, He is exalted, has no meaning except His
being knowing, powerful, and being an agent ( fāʿilan).

The rst assumption of al-Kaʿbī’s argument in favor of this position is that voli-
tion in the seen world requires intent for speci cation. The second assumption
is that God’s knowledge, by its very nature, makes this intent super uous, since
His knowledge already includes speci cation of time, form, and power. The
premise of the argument rests on the proposition that what God knows exists
by necessity. But how this assumption was justi ed and what its implications
were for al-Kaʿbī’s corollary view that “God’s volition is His knowledge, power,
and act” remain open questions.
Al-Shahrastānī pointed to al-Kaʿbī’s cosmology as the reason for his stance
on the attribute of volition. When al-Shahrastānī repudiated the broadly held
Muʿtazilī position on the attributes, he accused al-Kaʿbī of following the propo-
nents of the doctrine of nature and of denying God’s freedom of choice: “There
is no reason to deny volition as al-Kaʿbī did. He deemed it necessary (li-annahu
yūjib) for the compulsory acts (al-afʿāl ghayr [al-]ikhtiyāriyya) to be similar
(shabīha) to natural acts (al-afʿāl al-ṭabīʿiyya) according to the proponents of
the doctrine of nature (ahl al-ṭabāʾiʿ).”
Al-Shahrastānī does not, however, explain how al-Kaʿbī’s doctrine of nature
in uenced his conception of compulsory actions. This absence of explanation
does not necessarily imply that al-Shahrastānī was merely alluding to the doc-
trine of nature with polemical intent. Indeed, when al-Māturīdī spoke of
al-Kaʿbī’s position on God’s choice in His act as equal to what is naturally deter-
mined (maṭbūʿ) even he alluded to the in uence of the doctrine of nature.
Al-Māturīdī’s discussion, however, remains brief and requires further elucida-
tion before it can be taken as corroborating evidence for al-Shahrastānī’s
statement.

…
Al-Kaʿbī maintained the earlier Muʿtazilī distinction between the attributes of
essence and act. Unlike al-Jubbāʾī and, to a degree like al-Naẓẓām, he saw

The text that records this additional argument attributed to al-Kaʿbī contains many cor-
ruptions and is not fully legible (ibid., 240).
Ibid. [emphasis added].
Ibid., 245.
“God’s acts are by choice because the acts of whatever is naturally disposed (maṭbūʿ) are
one species (nawʿ) only” (al-Māturīdī, Kitāb al-Tawḥīd, 92).
scripture as the only source by which to identify the attributes of God. Following
al-Iskāfī, he understood God’s hearing and seeing, sometimes described as His
perception, to be His knowledge. Following al-Naẓẓām, he deemed God’s voli-
tion to be nothing other than His own act and His command of His servants’
acts. Al-Kaʿbīʼs continuation of earlier views on the attributes, views that his
Basran counterparts rejected, is evidenced in the case of the scriptural basis of
knowing the attributes, and the formulation of the attribute of essence as “by
His essence” (bi- or fī nafsihi) rather than “because of His essence” (li-nafsihi) as
al-Jubbāʾī thought.
One historiographical question, with historical implications, emerges from
these conclusions. Why do the sources single out al-Kaʿbī so emphatically for
doctrines that he did not originate? Do sources associate him with these doc-
trines only because it is convenient: he was the latest representative of these
earlier views? Or did he utilize a new methodology with which he reformulated
them. Based on the evidence examined here I cannot o fer a conclusive answer
to these questions. In the case of his de nition of the attribute of essence, the
likelihood that it was based on his particular ontology is credible but not con-
clusive. The evidence for his de nition of the attribute of essence is found in
al-Māturīdīʼs work. Although there is nothing to contradict this evidence, fur-
ther perspective is needed to contextualize it. Furthermore, al-Shahrastānī
noted the role of al-Kaʿbī’s epistemology in support of his equating the attri-
butes of hearing and seeing with the attribute of knowledge. Al-Shahrastānī
also recounted arguments derived from al-Kaʿbī’s cosmology and epistemology
in support of the latter’s understanding of divine volition. The problem with
al-Shahrastānī’s statement in the case of the attributes of hearing and seeing is
that it is combined with the allegation that al-Kaʿbī initiated this view, when it
was originated by al-Iskāfī. As for al-Shahrastānīʼs statement about al-Kaʿbī’s
reasoning on the attribute of volition, it is harder to accept this without reser-
vation because it is tied to an element that is not corroborated by the majority
of the sources. This uncorroborated element is the claim that al-Kaʿbī under-
stood divine volition to be divine knowledge, a claim that was prevalent in late
Ashʿarī sources and Sunnī biographical dictionaries as well.
Thus, while al-Kaʿbī did not innovate in his formulation of the doctrines of the
attributes, there is preliminary yet signi cant evidence that he sought to re ne
these earlier positions with advances he made in cosmology and epistemology.