http://www.jstor.org/stable/4057086

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/action/showPublisher?publisherCode=bap.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

BRILL is collaborating with JSTOR to digitize, preserve and extend access to Arabica.

http://www.jstor.org

ORTHODOXY AND HANBALITE FIDEISM *

BY

AZIZ AL-AZMEH

RTHODOXY is an intangible quality. It is not, from the view- point of the historian, a substance which inheres in different

situations and practices, but is rather an attribute of all religious authority, an attribute which acquires veracity to the extent to which it can be made binding and paradigmatic and to the degree to which it can be consolidated, enforced, and endowed with the capacity for dissemination and continuity. Just as orthodoxy, from the historical point of view, is not merely a dogmatic fact but a situation which has a precise dogmatic expression, heterodoxy in its turn betokens a dispute over religious authority-one which in- volves dispute over social and political territories where religion is practiced and spread, and conflict over the sources for the deter- mination and distribution of religion (dogma, observance, rituals of life-cycle, and other components). It is rare for a religion to achieve the evenness of belief and practice required by an or- thodoxy, and this evenness, when approached, would normally be restricted to limited geographical areas (in islamic history, largely to cities) on which is directly exercised the control of a temporal authority (which could be coterminous with the religious authority, as in certain forms of imamism or in patches in the history of the papacy, among other historical phenomena).

In Islam, as in other religions, there exist a variety of mechanisms for assimilating the unevenness of societies, polities, and cultures made inevitable by conditions prevalent before the homogenizing possibilities of the present era. Some of these mechanisms were of what we might term a Jacobin nature, involv- ing the destruction of tokens of unevenness. But such were usually of a temporary or otherwise limited nature, and constitute the im- plementation of a fundamentalist programme latent in every

* Paper presented at colloquium on 'Official and Unofficial Islam' held at University of Exeter, 26-27 September 1986.

Arabica, tome xxxv. 1988

254 AZIZ AL-AZMEH

religion, a programma which interjects a rigorist interlude in the natural course of things. Islam covered vast geographical areas, civilizations and cultural units, societies, polities, and economies over a millennium and a half. Like Christianity and Buddhism, the other world-religions, it has proved capable of assimilating a bewildering variety of forms, some of which might well appear an- tithetical. It has, as might be expected and like these other equally catholic religions, shown an omnivorous capacity for acquiring and ratifying popular practices, political requirements, social exigen- cies, cultural modes, mythological motifs, and other matters which were not present in the original writ of the religion and in its pristine condition. This is only natural, as the pristine order of a religion is, in religious terms albeit non in those of historical reason, recoverable only in terms of the scriptures that express it and of such commentaries on these scriptures as are deemed legitimate; and such material is less the authentic condition of a religion than the ratifying charter of its various orthodoxies. Ultimately, therefore, the unity of a religion is the unity of a genealogy. Like all genealogies, this one is constructed by a spurious history which has the task of eliminating unevenness in the genealogical relation. In the world of scriptural religions, such a task is undertaken by interpretation whose task it is to establish the concordance of the present (dogma, ritual, political and social order, etc.) with the original event, invariably in the form of show- ing the present as an ineluctable result of this absolute rectitude which is the past in question. In this context, orthodoxy would come to stand for a specific scripturalist genealogy, the general form of all scripturalist orthodoxies. Such fundamentalist myths of origin seem to be written into all religious phenomena1.

The ratification of mundane matters by a religion thus appears to be largely nominal. This does not detract from the momen- tousness of this matter, as the act of naming is a matter of great consequence in culture and ideology, and religions, as components of both ideology and of culture, have always deployed the resources of rhetoric. The hermeneutical and exegetical notions at the

I These matters have been discussed with respect to islamic thought in A. Al- Azmeh, Arabic Thought and Islamic Societies, London, 1986, pp. 166 ff., 228 ff., 260 ff.

HANBALITE FIDEISM 255

disposal of Muslim divines are particularly sophisticated and geared towards this act of naming; but not even this obviates the tangible textual reality of the origin, and at issue in Muslim hermeneutical practice is not a fact of conjuration but, in the rigorous sense, of interpretation. The act of interpretation is per- formed on well-defined material, the sacred text whose integrity is assured by a particular line of transmission (the seven variant readings of the Koran that came to be canonical, and the hadit), and it is this sacred text which provides the minimal contents of dogma and ritual. Thus in addition to genealogical affiliation, orthodoxy posits a number of credal articles and rules of ritual observance as the tokens of religious rectitude. Departure from this basic corpus constitutes a token of difference, and hence of heterodoxy, or of ir- religion.

Of the various tendencies that have constituted Sunnite Islam, Hanbalism has generally been singled out as the most consistently rigorist, and has generally been construed as the prime compo- nent of orientalist demonology2. Hanbalism has indeed striven for doctrinal purity, and has had a fair share, though by no means a monopoly, of rigorist and fundamentalist tendencies, per- sonalities, and social movements. But Hanbalism has also striven systematically to adopt a minimalist view of dogmatics and of ritual observance, and has been historically open to the exercise of wide latitude in matters of legal and political theory and practice that exclude and go beyond the bounds of the strict and minimalist rigorism of its dogmatics. This is the natural counterpart of rigorous and minimalist dogmatics, for it implies a practical divi- sion between the sacred and the profane and strictly delimits areas for the operation of regulations pertaining to both. Hence the pro- found realism of Hanbalite political theory, and their notion of the caliphate from the days of Ibn Hanbal3. This realism is equally

2 On the reputation of Hanbalism and for the elements of a fresh perspective, see G. Makdisi, 'Hanbalite Islam', in Studies on Islam, ed. M. Swartz, New York and Oxford, 1981.

3 See the statements by Ibn Hanbal in Ibn al-Gawzl, Mandqib al-Imam Ahmad b. Hanbal, n.e., Beirut, 1973, pp. 175-176 and by Barbahari in Ibn Abi YaCla, Tabaqdt al-Handbila, ed. M. M. Al-Fiql, Cairo, 1952, vol. 2, pp. 22-23. See also Abiu YaCla, Al-Ahkkm al-sultdnzyya, ed. M. M. Al-Fiql, 2nd ed., Cairo, 1966, pp. 20, 22; Ibn Batta, K. Al-Sarh wal-ibana 'an usuil al-djayna, ed. H. Laoust, Damascus, 1958, pp. 67-68; Ibn Qudama, text in H. Daiber, 'The Creed (CAqfda)

256 AZIZ AL-AZMEH

well highlighted by the liberalism of Hanbalitefiqh and its openness to the wide consideration and application of uncanonical modes of legislation, based on matters of equity, expediency, and custom, a matter shared with other schools of sunnite law4.

The title under which the doctrinal minimalism of the Han- balites was officiated was fideism (tafwid, tawqff, irj), the affirma- tion of dogmatic articles without a qualification that would discur- sively carry them beyond the bounds of their given textuality, on the assumption that the sense of the divine statement is sui generis and comprehensible to us only within the terms of this very state- ment, and that any qualification of the divine statement, in the form of explication or otherwise, is tantamount to imputing to the Divinity and to the Prophet utterances they did not make. This is the sense of the statement attributed to Ibn Hanbal and repeated with many variations by countless followers, that religion as such is the Book of God and veracious narratives about the Prophet and his associates, that religion is imitation, and that the substance of scriptural statements is textually incontrovertible while being in- tellectually ineffable. In antithesis stands ta 'wil, interpretation, execrated by the Hanbalites for its venture beyond the letter of the sacred text, and most often identified by them with the allegorical interpretation of the scriptures which they said added to them what was not in them, and consequently presumed on their behalf in a manner which can only lead to perdition5.

Characteristically, and due to the historical conjunctures out of which this position emerged, it referred to the divine attributes. Sacred textual pronouncements on divine attributes are accepted on faith,

of the Hanbalite Ibn Qudama al-Maqdisi. A Newly Discovered Text', in Studia Arabica et Islamica. Festschriftfor Ihsan CAbbas, ed. W. Al-Qadi, Beirut, 1981, p. 111.

4 On these matters in general, see A. Al-Azmeh, 'Islamic Legal Theory and the Appropriation of Reality', in Islamic Law: Social and Historical Contexts, ed. A. Al- Azmeh (in press). For classical Hanbalite jurisprudence involving these and other elements, see A. W. I. Abu Sulaiman, 'The Role of Ibn Qudama in Hanbali Jurisprudence', unpub. Ph.D. thesis, University of London, 1970.

5 See the classic answers of Ibn Hanbal during the mihna as related by the Mu'tazilite al-Gahiz: Gahiz, Rasa-'il, ed. M. 'A. Harutn, Cairo, 1964, vol. 3, pp. 293-4 and vol. 4, p. 7, and other statement by Ibn Hanbal in Ibn al-Gawzl, Mdnaqib, pp. 156, 165, 170-172, 434. See also Ibn Batta, Iba-na, pp. 47, 56, Ibn Qudama, Tahrz}m an-nazar fJ kutub ahl al-kaldm, ed. G. Makdisi, London, 1962, para. 43-46, 55, 73, and Ibn Taimiyya, 'Al CAqida al-wasitiyya', in idem., Magmuicat al-rasa-dil al-kubra-, n.e., Beirut, 1972, vol. 2, pp. 393 ff.

HANBALITE FIDEISM 257

'... a faith which accords with the letter of the text which is indubitable and whose veracity is incontestable, and whose meaning is known best to Him who uttered it, [a text] in which we believe according to the meaning He in- tended for it' 6.

Why this fideism? and what does it indicate in the context of the presumption of orthodoxy? It is essential to look not only at the content, but also at the form which islamic writers inaccurately praise as 'old women's religion'. Yet it would be more appropriate for us to look beyond this identification of fideism and self- evidence, and to regard this identification of nature and unadulterated scripturalist genealogy less as a mean, just, median and hence natural position than as a limiting case for the expres- sion of doctrinal orthodoxy. The form of dogma is one of a very specific discourse, that of the divine as properly transmitted. The language used therein is not everyone's vernacular, neither are its conceptions those of the people at large; this would have been im- possible anyway in an authoritarian and severely stratified social structure as that of all islamic societies. The language was technical, specific to the discourse itself, deriving its sense from its subjection to the semantic dictates of the divine discourse, and ultimately to the ineffable intent of God himself. The statement is inextricable from the context of the statement, that of divine discourse. This is not an irrationalism, for the opposition is not one of reason and unreason, but one of nomothetic discourse, this be- ing the divine, and that which is not. The nomothetic statement, the discourse of sarc, is significant in itself, and the terms in which this discourse is expressed is not necessarily related to other senses, nor is it permeable with respect to them, although it is possible to use the technical terminology of this discourse (al-ciba-ra al-s'arctyya) to express, and thus assimilate, other senses intended by the speaker 7.

The dogmatic utterance is therefore a technical one, not native to the human understanding nor innate to it; and although Han- balite writers often use the notion of innate natural knowledge (fitra) to express what they see as the self-evidence of their posi- tions, they do so in different discursive contexts to those in which

6 Ibn Qudama, Tahrim, para. 73 (my translation). 7 See the fine analysis of these matters in Ibn Taimiyya, Muwdfaqat sah4h al-

manqiul li-sarfh al-maCqiil, ed. M. M. CAbd al-Hamid and M. M. Al-Fiql, Cairo, 1950-51, vol. 1, pp. 116, 144-146 and passim.

258 AZIZ AL-AZMEH

dogmatic affirmations are made, and as part of a dialectical reper- tory. The technical rectitude of binding credal statements is assured by preserving the integrity of the original text, and this in turn is assured by the scrupulousness and exactitude of genealogical transmission. Hanbalites often referred to themselves and were referred to as ahl al-hadit or ashdb al-hadft and often preferred to view themselves as representing the neutral and thus just and natural attitude towards the textual realities of the faith. The scrupulous regard for pristine textuality is seen to generate a neutral, dispassionate, and therefore correct notion of the faith based on the very textuality whose integrity and veracity is assured by transmission. Such is the meaning of the statement often made by Hanbalite authors that their school represents the authentic sunna, that they are the evidence of Sunnism as such8. They are not a firqa, but merely the continuation of the pristine condition of Islam. And indeed, we find this reflected in the history of the school by Ibn Abi Yacla, who includes in his account personalities such as Bukhari, Muslim, Tirmidhi, Tabaran1, all hadit scholars9. Also included by this author among Hanbalites is Muhammad b. Idris al-SafiClPO; and the illustrious Hanbalite vizier Ibn Hubaira in- sisted that not only should Safici belong to the history of the school, but that Hanbalite predecessors included cAlI b. Abi Talib in a far more legitimate manner than that in which he is claimed by the Sicites ". The genealogical purity of the school and its sources was assured by constant reworking of older works, as were, for in- stance, the $Sjfat as-safwa and Minha-j al-qa-sidin of Ibn al-Gawz1 reworkings of Abu Nucaim al-IsfThan-l's HIilyat al-awl'y a and Gazali's IAyO' culu-m ad-din.

At the heart of the matter is not only the telescoping and reorganization of genealogies, but the consideration of the cen- trality of Hanbalism. It is not entirely untrue that Ahmad b. Han- bal is regarded by all sunnites as a paragon of rectitude 12, although assent to his doctrinal position was by no means universal. Groups

8 For instance, Ibn Rajab, Kasf al-kurba bi-kasf hdl ahl al-gurba, ed. J. M1adi, Alexandria, 1983, passim, and cf. Ibn Taimiyya, CAqida', p. 400.

9 Tabaqdt, no. 387, 388, 488, 594. 10 Ibid., no. 389. 11 Ibn Rajab, Al-Dail (ald Tabaqdt al-Handbila, ed. M. Al-Fiql, Cairo, 1952, vol.

1, p. 273. 12 Ibn Abi YaCla, Tabaqdt, vol. 1, p. 14.

HANBALITE FIDEISM 259

that he anathemized, such as the Mu'tazilites, were generally anathemized by subsequent Sunnism13. The Hanbalite school at- tempted the assimilation of all that is evidently and historically legitimate by its assimilation of SafiCi and others, and by the erec- tion of itself as the arbiter of rectitude. Hence the attempts to por- tray 5afiC as a pupil of Ahmad's in certain respects'4, and to pre- sent the opinion of the Khidr himself as being in agreement with this assessment'5. And though Ahmad is said to have rejected the works of SaficI and Malik b. Anas16, this was in favour of the unadulterated atar'7, of more original and pristine works devoid of commentary or selection, and not in opposition to these per- sonalities.

The tradition which became Hanbalism is therefore seen as directly derived from the fount and origin, and other traditions are legitimate in so far as they form part of this uninterrupted evenness of the surface of history connecting the present of Hanbalism with the Prophet and his companions, and the line leading directly from them to the Hanbalism of the era through the node constituted by Ahmad b. Hanbal. The sacred unalterable sui generis utterance which is the profession of orthodoxy, of Sunnism, is inseparable from the authority which polices its integrity. It is irrelevant without this authority, for without it it will be an unauthorized ut- terance; and as it is not, in the context of fideism, the sense of the utterance which is of relevance but its physical reality, and since such reality is inseparable from the genealogical sanction which guarantees it, it follows that, in a determinate sense, the utterance and the authority are one.

The genealogical sanction of fideism is an inseparable compo- nent of it, and could be indeed said to override in operational ef- ficacy the text itself. It goes without saying that the text regarded with fideist eyes as containing the integrity of the minimalist creed can not be inclusive of the whole textual repertory which might have been so regarded. Of the corpus of sacred texts, certain mat- ters are given centrality, others occluded, certain shades and potentialities of meaning realized and others suppressed. Hanbalite

13 Ibid., p. 16. 14 Ibid., pp. 6, 5-14. 15 Ibn al-Gawzi, Manaqib, p. 144. 16 Ibid., p. 192. 17 Ibn Abi Yacla, Tabaqdt, vol. 1, p. 207.

260 AZIZ AL-AZMEH

creeds, some of which have become fairly well known to modern scholarship-of Ibn Hanbal, Barbahari, Ibn Batta, Abu- YaCla, Ibn Qudama, and Ibn Taimiyya'8-have in common a definite and constant cluster of topics and textual supports which constitute the basic dogmatic repertory of the school.

These matters Hanbalism shares to no inconsiderable degree with As'arism. The relation between the two schools has been complex and occasionally stormy. It is very likely that the original A'sarism as contained in the Ibana of A'Car1 could well have served as the theological arm of the Hanbalites, who at that time were still in the process of constitution. Ascarism then migrated to Hurasan where it bloomed in close proximity and in constant polemical struggles with Hanafite-Muctazilite tendencies and movements. When it returned to Baghdad in the fifth century, it did so in the context of the struggle between the Caliphate, which had the sup- port of the Hanbalites, and the Sultanate, which had sustained both A'Carism and MuCtazilism"9. The latter half of the fifth cen- tury of the Hijra and the first half of the subsequent century witnessed acute conflicts in Baghdad between the divines of both tendencies, supported by riotous local popular constituencies20. But irenical attempts were many, usually in the form of attempts at concordances between Ahmad b. Hanbal and 5afi 21, under the auspices of whose legal school Ascarism was attempting to institute

18 For a general account of the history of Hanbalite creeds, see H. Laoust's In- troduction to Ibn Batta, Ibana.

19 See Al-Azmeh, Arabic Thought, pp. 214 ff. The A'carism of Ascari has long been doubted and discussed. For the latest, see the fresh specifications of D. Gimaret, ' 'Un document majeur pour l'histoire du kalam: Le mugarrad maqdldt al-a&Cari' d'Ibn Furak', in Arabica, 32(1985), pp. 185-218.

20 For Hanbalite accounts: Ibn al-Gawzl, Al-Muntazam ft ta-rih al-muliik wal- umam, ed. Krenkow, Hyderabad AH 1357-58, vol. 8, pp. 163, 305-307, vol. 9, pp. 3-4, vol. 10, pp. 6, 106, 108, 111; Ibn Abi Yalla, Tabaqdt, vol. 2, p. 239; Ibn Rajab, Dail, vol. 1, pp. 19-22. For AMCarite accounts: Ibn CAsakir, Tabyzn kadib al-muftari-t fma nusiba ila al-Imam Abil-Hasan al-As carl-, ed. H. Qudsi, Beirut, 1979, pp. 108, 113-114; Subkl, Tabaqat al-Sdficjyya al-kubra-, ed. cA. Hiluc and M. M. Tanahi, Cairo, 1964 ff., vol. 3, pp. 390 ff. Indications on the social constituency of Baghdadian Hanbalism in this period are given by G. Makdisi, Ibn CAqil et la resurgence de l'Islam traditionaliste au XL siecle, Damascus, 1963, pp. 329-330, who also discussed these conflicts on pp. 340 ff.

21 For instance, the Hanbalite attempt spoken of by Ibn Rajab (Dhail, vol. 1, pp. 33, 35), and the Ascarite attempt of Baihaqi discussed by H. Laoust, 'Le Hanbalisme sous le Califat de Bagdad (241/855-656/1258)', in Rev. d'Etudes Islamiques, 27(1959), p. 96.

HANBALITE FIDEISM 261

itself22. AsCari was also accorded some of the centrality of Ibn Hanbal that was discussed above by large numbers of divines, in- cluding the Hanbalite Ibn CAqil 23. Many of the attributes of dogmatic neutrality, and hence rectitude, that we saw applied to Ibn Hanbal by his followers were claimed for A's'ari by his followers: he is said to have founded no sect, and led no tendency, but simply continued the pristine condition of the original texts24, and his defence of sunnism, of the correct creed, had been adopted as the dogma of all hadit scholars25. His detractors are reminded that As'arite theology was a strong support for Baghdadian Han- balites before the middle of the fifth century, and that this was due to the fact that both schools were adherents of the affirmation of the divine attributes as they occur in the Koran26. Confirmation of this comes from the Hanbalite Ibn Taimiyya, with his usual perspicacity, rigour and impatience with historical differences that accrue and distort original identities27. And in fact As'ari had ex- plicitly praised Ibn Hanbal for his position on divine attributes and stated his intention of proceeding in like manner28, and at least one prominent As'arite said his school was, like Hanbalism, one of hadit the difference being the engagement of the former in polemical activity against detractors, thus requiring the use of more than the sheer repetition of textual material29. In both cases, the position was one of fideism which was presented as the recen- sion of texts whose sense is specific to them. We have in fact a fund of common Hanbalite-classical As'arite credal topoi-this had been expanded and modified by Hurasanian As'arism from Guwaini onwards30-whose lynchpin is the question of divine attributes,

22 See n. 19 above. 23 Subkl, Tabaqdt, vol. 3, pp. 375-376. 24 Ibid., vol. 3, pp. 354 ff. See the general discussion of G. Makdisi, 'Ash'ari

and the Ashcarites in Islamic Religious History', pt. 1, in Studia Islamica, 17(1962), pp. 37-81.

25 Subkl, Tabaqdt, vol. 4, p. 32. 26 Ibn cAsakir, Tabyfn, p. 163. 27 Muwdfaqa, vol. 2, pp. 10-11 and idem., Naqd al-mantiq, ed. M. Hamza, S.

Sanic and M. Fiql, Cairo, 1951, pp. 10-11. 28 Ascari, Al-Iba-na can usu-l al-djydna, n.e., Cairo, n.d., pp. 46-47. 29 Ibn Furak, Kitdb muskil al-hadit, M. cAbd al-Mucid Khan, Hyderabad, 1970,

p. 3. 30 The sources, and therefore the direction, of the other Ascarite tradition,

which produced the apologetic works of Ibn cAsakir and Subkl, are indicated by Subkl, Mucid an-nicam wa mubfd an-niqam, ed. D. W. Myhrman, London, 1908, p. 107.

262 AZIZ AL-AZMEH

which had been made into the touchstone of orthodoxy, a fund to which appeal was made in apologetic literature-of sufism for in- stance31. The difference, as Ibn Fuirak had indicated, resided in the aversion of the Hanbalites towards all acts of interpretation32.

The relationship with As'arism illustrates the manner in which the minimalist fund of Hanbalite dogmatics emerged. Other com- ponents, such as the cardinal principle of the increation of the Koran and it being the physical word of God, have equally deter- minate conjunctures of emergence-in this case, the dispute with the Mu'tazilites during the reign of Ma'mu-n33. The physical vision of God in the afterlife was also the result of the conflict with the Mu'tazilites, the historical assessment of personalities in early islamic history was the result of conflict with the Silites as well as with sections of the Muctazilites, one which became more vehe- ment with the adoption of the Hanbalites of the Abbasid cause in direct contrast to the cAlid legitimism of the Sica. Eschatological articles were similarly the result of determinate disputes with various cAlid and other groups in the early period of Baghdadian Hanbalism. All in all, the minimalist fund which constituted Hanbalite dogmatics represents the institutional concatenation, of uncertain date but certainly by the middle of the fourth century, of a number of doctrinal topol which had the support of a determinate body of sacred textual material. It was not the result of a deliberate construction except by posterity. This dogmatic repertory was moreover constructed topically and aggregatively, with the systematic connection between different parts of the topical cluster almost wanting, and this was in conformity with the structure of islamic dogmatic theology in general, whose schools had, like Hanbalism, grown of the convergence of particular polemical topics and groups engaged in them, and constituted the institu- tional, social and epistemic, result of these convergences34.

31 Kalabadhi, Al-Tacarruf li-madhab ahl al-tasawwuf, ed. M. A. Nawawi, Cairo, 1969, pp. 47-77; Qusairi, Al-Risdla, ed. 'A. Mahmfid and M. Bin al-Sarif, Cairo, 1972-74, vol. 1, pp. 33-35.

32 Ibn Abi Yacla, Tabaqdt, vol. 2, p. 210. 33 This question was enmeshed in complex semantic and linguistic controver-

sies which is very imperfectly understood. On the various positions on the crea- tion and increation of the Koran, see Ascari, Maqdldt al-islhmiyyfn, ed. M. M. cAbd al-Hamid, Cairo, 1950-1954, vol. 2, pp. 231-244.

34 Al-Azmeh, Arabic Thought, pp. 176-178, 211 ff.

HANBALITE FIDEISM 263

The fideist regard towards textual evidence of dogma was the one element that lent coherence to the Hanbalite dogmatic enter- prise. This coherence was therefore really external to dogma, and endowed by the agency which carries this evidence-the epistemic agency of fideism, and the socio-political agency of the Hanbalites. It was difficult always to maintain a strict fideist position, and the history of Hanbalism is by no means devoid of anthropomorphic positions 35 -there is no need to be horrified at the possibility that Hanbalites might have been in their earliest constitution, one of the many anthropomorphic sects that flourished in early islamic Iraq and Iran, as there is indeed no reason for revulsion at the idea that Islam, still indeterminate in the first century, might under different conditions have developed into an anthropomorphism; we should be wary of identifying beginnings and emergences with outcomes. Hanbalism was also not devoid of instances of serious theological study of the problem of divine attributes which, due to the re- quirement of maintaining certain metaphysical positions in- separable from any such speculation in Arab-islamic thought, had to adopt positions different only in name to some As'arite posi- tions. These later positions were normally considered inadmissible, and were indeed considered as such by the same individuals who propounded them. Abu Ya'la, the great Hanbalite theologian and divine, adopted, among other things, the As'arite typology of at- tributes (attributes of essence and attributes of act) under the name essential attributes and conceptual (ma'nawiyya) attributes36. But he did so in a theological treatise which, as such, contained the normal topical repertoire of the theology of his time, including the Five Principles of the Mu'tazilites, which he interpreted after a Han- balite fashion37, and reflected the social and cultural elevation of Hanbalite divines of the time. The same Abu- Ya'll, in his creed and therefore in a different discursive context, repeated the stan- dard position of Hanbalite fideism by requiring belief in the way that God described himself or was described by His Prophet 'without disqualification [of the literal sense of the attributes], or analogy, or exegesis, or interpretation'38.

35 For instance, Ibn Rajab, Dail, vol. 1, pp. 29, 174. 36 Abu YaCla, Kitab al-muCtamad fr usuil al-din, ed. W. Z. Haddad, Beirut, 1974,

para. 66, and cf. Al-Azmeh, Arabic Thought, p. 68. 3 Abu Yacla, Muctamad, para. 385 ff. 38 Text in Ibn Abi Yacla, Tabaqdt, vol. 2, pp. 207 ff.

264 AZIZ AL-AZMEH

The requirement is therefore one of a strict fideism, a rejection of the anthropomorphism brought about by analogy-and an- thropomorphism was a charge constantly levelled against Han- balites-as well as of the denial of the reality of attributes, as did Mu'tazilite interpretation which was aimed at avoiding any analogy which might detract from God's absolute transcendence. Not many divines had the vast learning and the dialectical and rhetorical genius of an Abut Yacla, or were like him capable of answering the requirements of fideism in the folds of the very act of interpretation39. Are we therefore to consider as correct the con- tention of the Hanbalites that their position is one of a perfect equipoise, the just middle, the authentic continuator of authentic beginnings? When asked whether one may transmit a hadft, authenticated by no less than Qatada after cIkrima after Ibn cAbbas, which related of the Prophet his statement that he saw God in the shape of a young man, Ahmad b. Hanbal readily declared that one may, given the authority of this transmission40. Can we agree with the Hanbalite position that such a statement has no an- thropomorphic intension when it is not subjected to an allegorical interpretation?

It is clear that it is possible to maintain that the Prophet's state- ment referred to and Koranic verses speaking of God's hand, face, wrath, and other attributes can be stripped of an anthropomorphic sense only if, in addition to eschewing allegory as the Hanbalites require, one were to eschew the work of the understanding and the imagination. It has been indicated how, in the work of Abui Yacla, dogma exists in two registers, or some might say two levels, one in the fideist narrative mode, and another in a discursive mode, the one in a catechistic form, the other in the form of a dissertation, the one arbitrary, the other reasoned. And it is a primary characteristic of fideism that it should appear arbitrary, as this is the counterpart of the literalness which is yet utterly specific in sense. It is not reasonable to make a statement whose lexical sense is an- thropomorphic, yet interdict both allegory and anthropomor- phism. Very much like the linguistic sign, it is arbitrary but no less binding for its arbitrariness. Dogma in this sense, posited by its disseminators and having been the aggregative result of separate

39 For instance, Abui Yala-, MuCtamad, para. 84 ff., 100, 115, 477. 40 Ibn Abi YaCla, Tabaqdt, vol. 1, p. 218.

HANBALITE FIDEISM 265

historical developments and conjunctures, is quite arbitrary in relation to the understanding and imagination, no less than in relation to popular religion and symbolism.

Fideist dogma is hence quintessentially and thoroughly technical, and therefore constitutes a quarry of authority, that of the Hanbalite leaders in whatever context they had been active. Fideism can only be rigorously maintained by administrative means.

But over and above this, this arbitrary is centred around the sacerdotal word, and is an expression of dogma in the elocutionary as distinct from the discursive form, the oral as distinct from the written. The arbitrariness of fideism indicates it not as a form of knowledge or of apprehension, but as a form for the expression of a particular authority that decrees and disseminates it. It indicates, furthermore, a form of religious knowledge and for the transmis- sion of such knowledge that is appropriate for preachers, and it is a fact that Hanbalites were very active at this level of communica- tion4'. Popular preachers were very active in Baghdad and were of grave consequence for popular movements, and the annals of medieval Baghdad record numerous instances of banning popular preaching in periods of unrest. Hanbalite histories are replete with accounts of Hanbalite preachers, and the Hanbalite Ibn al-Gawzi was possibly the greatest preacher in islamic history42. In his manual of preaching, Ibn al-Gawz1 stated that which was per- missible for popular preachers to say about the divine attributes and the increation of the Koran in terms very similar to those we encounter in Hanbalite creeds, while he specifically interdicts ut- terances of anthropomorphic sense43. It comes as no surprise to us that, given the interdiction of allegory and insistence on literalness, streetcorner religiosity purveyed an anthropomorphic Allah.

Fideism thus appears a form of knowledge appropriate for a religious party with a keen interest in the populace at large, and hence with a decided tendency towards proselytism. Knowledge appropriate for preachers and their audience is also minimalist

41 This has already been indicated by H. Laoust in his introduction to Ibn Batta, Iba-na, pp. LXIII-LXIV n.

42 For a vivid eye-witness account, see Ibn Jubair, Rihla, Beirut, 1964, pp. 197-199, which compares with Ibn al-Gawzl's own account of his preaching methods, on which see his Muntazam, vol. 10, pp. 243, 270, 284 and Kitdb al-qudss wal-mudakkirfn, ed. M. Swartz, Beirut, n.d., pp. 324-325.

43 Quyyas, p. 142.

266 AZIZ AL-AZMEH

therefore on didactic grounds. But its fideist minimalism also in- dicates a further characteristic of preachers' knowledge and a specific form of popular religiosity. It is a form of knowledge which is devoid of epistemic content, and whose utterance is consumed immediately by its audience. Though it instructs by example, it presents itself as symbolic power and as the expression of the authority of the custodian of this knowledge and the genealogy which he presents as his charter. It is located on the interface be- tween dogma properly so called, and devotion, between disserta- tion and incantation. It is a purely affirmative form of expression, and belongs properly to an act of devotion more than to one of in- tellection. In an evolutionist perspective, one might be able to assert that this accords well with the belief in the primacy of devo- tional rituals in all religions, in which case fideism could be described as a very basic, logolatrous form of religion. Ibn al- Gawzi understood what he was doing very well. He only concluded that the audience of his preaching had understood him properly after it had been enthralled and entranced by his words44. And in- deed, he himself wished his body purified after his own death with water heated by the shavings of pens he used for transcribing hadit45.

It was the sheer untenability of fideism that made it correlative with its authority, an authority which also had a plenitude of magical instances illustrated by thaumaturgic and wondrous acts of Hanbalite divines46, a realm of the miraculous to which was sometimes assimilated a thaumaturgic notion of kingship47. This untenability, correlative with authority and the miraculous, freed for this authority-the Hanbalite establishment-a latitudinarious framework of the profane within which the affairs of the world could be managed, one reflected in their political realism and legal liberalism that have already been indicated.

University of Exeter

44 Maqqari, Nafj al-thb, ed. I. CAbbas, Beirut, 1968, vol. 5, p. 162. 45 Ibn Khallikan, Wafayat al-a'ydn, ed. I. 'Abbas, Beirut, 1970, vol. 3, p. 141. 46 For instance: Ibn al-Gawzl, Manaqib, pp. 186 (but cf. p. 267), 295, 296, 297,

483, 513; Ibn Abi Yacla, Tabaqdt, vol. 1, pp. 81-82, vol. 2, pp. 248-249, 252, 256; Ibn Rajab, Dail, vol. 1, pp. 69-70, vol. 2, pp. 57, 60, 138, 164.

47 For an explicit statement, see Ibn al-Gawzl, Al-Misbdh al-mudi'fr iildfat al- Mustadlf, ed. N. 'A. Ibrah-m, vol. 1, Baghdad, 1976, pp. 289-291. For more or- dinary evidence, see, for instance, Ibn Abi Yacla, Tabaqdt, vol. 2, p. 240.