GHAZALI ON MIRACL ES AND NECESSARY CONNECTIONGEORGE GIACAM AN AND RAJA BAHLUL
Ghazali on Miracles and Necessary
Connection
GEORGE GIACAMAN AND RAJA BAHLUL
Birzeit University
I. INTRODUCTION
Ever since Muslim thinkers came into contact with Greek Philosophy and
science in the seventh century, the relation between Islamicphilosophyand
theology has been an uneasy one. Muslim philosophers often felt com-
pelled to defend their philosophical activities against the suspicions and
attacks of the theologians, and some developed considerable energy and
effort to the harmonization of philosophy and religion on some fundamen-
tal points.
Ghazali (d. 1111) was perhaps the most important Muslim theologian
to attack the activities of Muslim philosophers. His Tahafut al-Falasifah
(“Incoherence of the Philosophers”) constitutes the most systematic and
thorough attack on Neo-Platonism by a Muslim thinker, and has had a
considerable influence on the course of philosophical activity in Islam.
So considerable its impact was that Averroes (d. 1198), the most
prominent medieval Muslim Aristotelian, felt compelled to write a para-
graph-by-paragraph rebuttal of Ghazali’s book. As cogent and persuasive
as Averroes’ Tahafut al-Tahafut was, however, still it failed to counter the
influence which Ghazali’s work was to have on the subsequent course of
Islamic philosophy.
1
Ghazali’s Tahafut is also of clear philosophical interest sincein it he was
not merely content to adduce religious considerations for rejecting the
views of Muslim Neo-Platonists, as represented primarily by Ibn Sina (Avin-
cenna, d. 1037) and al-Farabi (d. 950). He sought to meet the philosophers
on their own ground. As he himself observes in al-Munqith mina’l Dhalal
(“Deliverance from Error”):
1. For a brief critical assessment of the basic line of defense used by Averroes
against Ghazali, see George Giacaman, “Tradition and Innovation: Two Muslim
Views of Causal Relations,” in Philosophie et Culture: Proceedings of the Seventeenth World
Congress of Philosophy (Montréal: Éditions Montmorency, 1986), pp. 247–49.
39
Medieval Philosophy and Theology 9 (2000), 39–50. Printed in the United States of America.
Copyright © 2001 Cambridge University Press 1057-0608
I was convinced that a man cannot grasp what is defective in any of the
sciences unless he has so complete a grasp of the science in question
that he equals its most learned exponents in the appreciation of its
fundamental principles, and even goes beyond and surpasses them,
probing into some of the tangles and profundities which the very
professors of the science have neglected. Then and only then is it
possible that what he has to assert about its defects is true.
2
The Tahafut itself consists of twenty questions or problems that have
relevance to religion and for which Ghazali takes the philosophers to task.
These can be divided into two broad categories: (1) questions that conflict
with some of the fundamental principles of religion and for which the
philosophers are to be charged with irreligion (kufr), and (2) questionsthat
do not come into conflict with a basic religious tenet, yet on account of
which, the philosophers are nevertheless to be considered to have commit-
ted heretical innovation (bid’ah). The problem of causation belongs to the
second category and is dealt with as “problem number seventeen” of the
Tahafut.
Ghazali’s examination of causality occurs in the context of his discus-
sion of the “physical sciences” (al-tabi’iyyat). His overall concern is to affirm
the omnipotence of God and to safeguard the possibility of miracles.
3
Specifically, Ghazali argues against the conclusion that “[the] connection
observed between causes and effects isof logicalnecessity.”
4
This heregards
as relevant in view of the fact that the possibility of miracles, which consti-
tute a “departure from the usual course of events” (khariqah li’l’adah, liter-
ally “violates habit”), appeared to him to conflict with the attribution of
necessary causal efficacy to inanimate particulars. As he states:
As to the first point, it is necessary to contest it, for on its negation
depends the possibility of affirming the existence of miracles which
interrupt the usual course of nature like the changing of the rod into
a serpent or the resurrection of the dead or the cleavage of the moon,
and those who consider the ordinary course of nature a logical neces-
sity regard all this as impossible . . .
5
In what follows, we present a brief critical examination of Ghazali’s
main arguments against the views of the philosophers on causation. We
2. Ghazali, al-Munqith mina’l Dhalal, in The Faith and Practice of Al-Ghazali, trans.
W. Montgomery Watt (Lahore: Sh. Muhammad Ashraf, 1963), p. 29.
3. For theArabic text ofGhazali’sTahafut we have usedTahafut al-Falasifah, ed.
M. Bouyges (Beirut: The Catholic Press, 1962) Henceforth: TAF. Ghazali’s text is
quoted by Averroes in his Tahafut al-Tahafut. Henceforth: TAT. We have used Simon
van den Bergh’s translation of Averroes’ Tahafut (London: Luzac & Co., 1954) as
the English translation of Ghazali’s text. TAF, p. 194; TAT, p. 1:313
4. TAF, p. 191; TAT, p. 1:312.
5. TAF, p. 192; TAT, p. 1:313.
40 GEORGE GIACAMAN AND RAJA BAHLUL
suggest that the phrase “departure from the usual course of events” carries
at least two meanings, only one of which is in conflict with the belief in the
idea of a causal order where events follow one another in an intelligible
manner. Furthermore, we argue that Ghazali’s desire to uphold the possi-
bility of miracles need not constrain him to repudiate the idea of necessary
connection, since he is able to explain miracles in ways that are compatible
with belief in causality and necessary connection. We conclude by examin-
ing some arguments to the effect that Ghazali’s attempt to hold onto both
miracles and necessary connection is inherently unstable; furthermore, we
explore directions which Ghazalians may take in order to counter these
arguments.
II. GHAZALI’S EMPIRICIST EPISTEMOLOGY
Ghazali begins by stating his view and then goes on to challenge the basis
of the argument of his opponents. Simply stated, his view is that the con-
nection between what is believed to be the cause and the effect is not
necessary.
6
It is possible to have a cause without what we normally view as
its attendant effect, and it is possible to have an effect without what we
regard as its cause. Ghazali offers several examples to clarify his meaning:
eating and the satisfaction of hunger, contact with fire and burning, decapi-
tation and death. That hitherto each of these “causes” was followed by what
we understand to have been its effect is not due to any inherent powers or
capacities in the “causes” themselves, but is solely due to the power of God
through whose agency their concomitance has been maintained. The
connection between them is itself neither necessary norindissoluble. More-
over, God has the power to produce satiation without eating and decap-
itation without death.
Formulated in the foregoing manner, Ghazali’s initial position does
not amount to an argument. It is a statement of a position and a point of
view. And while the text itself does contain philosophical argumentation,
Ghazali does not always separate clearlybetween thatand ahost ofreligious
considerations that he marshals to buttress his position.
The main thrust of Ghazali’s attack on the belief in the existence of
necessary causal connections in nature involves challenging the epistemo-
logical basis of such a belief. First of all, he argues, it can be said that the
existence of one thing with the existence of another does not show that one
exists by the other. When we ordinarily observe the constant conjunction of
some objects with others, we begin to assume that they are inextricably
connected. But what right have we to regard them as being causally con-
nected and in a necessary manner, Ghazali asks, since all we observe is the
6. TAF, p. 195; TAT, p. 1:316.
GHAZALI ON MIRACLES AND NECESSARY CONNECTION 41
presence of one object followed by the presence of another, and observe no
connection between them? The only argument that philosophers can pro-
duce to show thatfire,forinstance, has thecapacity to necessarily incinerate
cotton is from observation. Observation, however, affords us no reason to
believe in the existence of anything other than conjunction. It “proves a
simultaneity, not a causation.”
7
Fire, in any case, is an inanimate object and
can have no action. God is the only agent of the creation of blackness in
cotton and the disintegration of its parts, accomplished either directly or
through the intermediary of angels.
8
Ghazali’s argument presupposes an empiricist epistemology. But the
foundation of the argument is not developed as it is, for instance, by Hume,
whose views on causation bear a striking resemblance to those of Ghazali.
Hume’s account of causation, it should be noted, was based, in part, on his
epistemology—in particular, his general thesis that meaningful ideas are
reducible to the sensory impressions from which they are derived. On the
basis of this, Hume proceeds to isolate the three empirical relations of
contiguity, succession, and constant conjunction as the essential elements
in our idea of causation.
9
According to Hume, the idea of necessary connection is a product of
those three empirical relations. It is subjective in origin, in the sense that it
is an idea that is impressed on our minds by the constant conjunction of
certain objects and events. It has no foundation in reality, but is rather
projected by the mind upon nature.
Unlike Hume, Ghazali’s intent is not primarily the analysis of the
meaning of causation. Ghazali’s concern is to guard against compromising
divine omnipotence. This becomes a distinct possibility when miracles are
denied on the strength of a belief in a causal order which is not subject to
divine power. To achieve his end, Ghazali finds it essential to deny that
necessity and causality can be attributed to the ontological order.
Thus far,two main featuresof Ghazali’s position have been mentioned:
(1) his denial of the existence of a necessary causal connection between
matters of fact, which leaves him free to attribute causal agency solely to
God; and (2) the argument that observation does not support the conclu-
sion that necessary relations exist between causes and effects, but only their
constant conjunction. The first point is clearly not a philosophical argu-
ment. Rather, it is a statement of a point of view to be argued for. But
Ghazali’s argument in support of his position falls short of this mark for at
least two reasons.
Firstly, if one grants the validity of the argument, all it demonstrates is
that to the extent that our knowledge of the essential features of causal
relations between particulars is based on observation, we cannot legiti-
7. TAF, p. 196; TAT, p. 1:317.
8. TAF, p. 196; TAT, p. 1:317.
9. David Hume, A Treatise of Human Nature, ed. L. A. Selby-Bigge (Oxford:
Clarendon Press, 1888), bk I, pt. 3, sec. 2,6.
42 GEORGE GIACAMAN AND RAJA BAHLUL
mately suppose that a necessary relation exists between what we regard as
cause and effect, since weobserve nosuch relation.It may wellbe that there
are necessary connections between causes and effects, but if there are any,
we are unaware of them. The evidence we possess does not afford us the
information to warrant the assertion that necessary relations obtain be-
tween matters of fact. Ghazali’s argument thus challenges the empirical
grounds for believing that effects necessarily follow their causes. Of itself,
however, the argument does not suffice to show that there are no necessary
connections between causes and effects. As Madden puts it, one needs to
be a “rigid positivist” in order to believe that failure to experience some-
thing is a good reason for believing that it does not exist.
10
But secondly, and perhaps more importantly, what we have called
Ghazali’s “empiricist epistemology” can be challenged. In particular, one
can question his claim that observation shows only that something happens
with something rather than by it. The most straightforward line of question-
ing that comes to mind here is the one that G. E. M. Anscombe directs at
Hume’s claim that we are not able to perceive causality or necessary con-
nection. Anscombe claims that our use of the concept “cause” presupposes
our application of a host of causal concepts that we use in “reporting what
is observed.” As examples of causal concepts that we regularlyuse inreport-
ing “what is observed,” she offers the following examples: scrape, push, wet,
carry, eat, burn, knock over, keep off, squash, make, and hurt. To Hume’s
challenge to produce an example of a (causal) “efficacy” that is obvious to
our consciousness or sensation,
11
she replies: “Nothing easier: is cutting, is
drinking, is purring not ‘efficacy’?”
12
Of course, die-hard Humeans or Ghazalians might reply that when we
observe a floor being scrubbed, or a clump of cotton going upin flames, all
our senses register is a seriesof appearances: first, the floor looks dirty, then
it is shiny, and one might even see bits of dirt being “detached” from the
floor, and so on. But scrubbing as such, one might say, is not really “ob-
served.” Perhaps one “infers” itfrom whatone does see. Butin sodoing one
goes beyond observable evidence (as is the case with all scientific infer-
ences).
According to this line of reasoning no one really observes anything
acting on anything, or, for that matter, being acted upon. But is this really
a tenable position? Not only does it make one wonder what function all
these verbs of “action” are doing in language, but it makes one also wonder
what observation itself is, if not some kind of activity (“efficacy,” in An-
scombe’s words). When one observes, is not one doing something, namely
10. E. H. Madden, “Averroes and the Case of the Fiery Furnace,” in Islamic
Philosophy and Mysticism, ed. T. Beauchamp (Encino, CA: Dickenson Publishing
Company, 1974), p. 139.
11. Hume, A Treatise of Human Nature, bk I, pt. 3, sec. 14.
12. G. E. M. Anscombe, “Causality and Determinism,” in Causation and Coun-
terfactuals, ed. E. Sosa (Oxford: Oxford University Press, 1975), p. 69.
GHAZALI ON MIRACLES AND NECESSARY CONNECTION 43