In 1545, Girolamo Cardano (d.1576) published his Great Art, or On the Rules of Algebra. It begins: ‘in this book, learned reader, you have rules of algebra. It is so replete with new discoveries and demonstrations by the author – more than seventy of them – that its forerunners are of little account’ (trans. from Witmer, Reference Richard Witmer1993, p. 1). It’s ironic that a work that begins with boasts of its own originality should be notorious for having included an unoriginal result. One of the ‘discoveries’ Cardano revealed to his ‘learned readers’ was the method for solving cubic equations, that is, equations of the form x3 + ax = b. It was not Cardano who discovered this method, nor indeed did he claim to have done so. As he openly explained in the Great Art, he had been shown it by his colleague Niccolò Tartaglia (d.1557), who wanted to publish it himself. But Cardano learned that Tartaglia’s discovery was actually a re-discovery, since the method had already been established decades before by the now dead Scipione dal Ferro (d.1526). Cardano felt this gave him license to include the solution to the problem in his own book, much to the outrage of Tartaglia.
To our eyes this scholarly furore looks pretty familiar. In the modern world, scientists, mathematicians, and even philosophers are frequently anxious about publishing their results before they are ‘scooped’ by someone else. We know too that this phenomenon is not all that recent; just think of the famous dispute over who could take credit for inventing calculus. But in the sixteenth century, the phenomenon was recent. Very few medieval authors trumpeted their own originality the way Cardano does here, never mind bothering to give elaborate explanations about who should (and shouldn't) get credit for discovering what. In the midst of his explanation of the double discovery of the solution to cubic equations, Cardano comes out with a phrase that could appear in a twenty-first century footnote: methods presented in the book discovered by other scholars will receive explicit attribution, but ‘those to which no name is attached are mine’ (Witmer, Reference Richard Witmer1993, p. 9).
Here we have a central paradox of the Italian Renaissance. This was an age where scholars were unearthing long ignored texts from antiquity, and extolling the unparalleled achievements of classical culture in such fields as of philosophy, history, literature, science, and mathematics (where the leading heroes were Pythagoras and Archimedes: see Rose, Reference Rose1975; Joost-Gaugier, Reference Joost-Gaugier2009). Yet it was also an age where originality was prized as a sign of genius, with that prize sometimes being bestowed by others (Vasari writing on Michelangelo) and sometimes awarded to oneself (Cardano). On both counts, the Renaissance seems to depart from the medieval mindset. Of course the whole point of speaking about ‘renaissance’ is that the study of antiquity was being ‘reborn’ after the long, moribund medieval age; and we standardly think of the medievals as being desperate to avoid giving the impression of originality. For the medievals, we are told, originality was not an asset to be claimed with pride but an accusation to be hurled at others. This was especially so in religious contexts, where novelty was at most only a step removed from heresy.
That, at least, is the received wisdom. In fact things were more complicated, and on both scores. The first point, about the recovery of texts unknown to the medievals, need not detain us here. But it is worth saying that while some ancient works were indeed rediscovered and had a major impact, a notable example being Lucretius On the Nature of Things (see Greenblatt, Reference Greenblatt2011), Renaissance philosophy also involved deep attention, and new approaches, to the Aristotelian corpus that had already been at the core of the medieval university curriculum (Schmitt, Reference Schmitt1983). Furthermore, the Greek texts that inspired Renaissance philosophers and scientists were usually not so much rediscovered, as imported from Byzantium, where they had been preserved and known the whole time. For example Greek Byzantine commentaries on Aristotle’s Ethics were an important influence on engagements with this text in Latin (Lines, Reference Lines and Miller2012; Trizio, Reference Trizio and Ziolkowski2014). Finally, the fifteenth century was not the first time that ancient texts had been ‘recovered’ and given new attention. Scholars also speak of a ‘twelfth century Renaissance’ when discussing the contribution of philosophers like John of Salisbury (d.1180) and the Platonists associated with the city of Chartres, and even a ‘Carolingian Renaissance’ animated by scholars like Alcuin (d.804) and Eriugena (d. after 870).
But in what follows my focus will be on the second point. Were medieval thinkers really so authority-bound, so reluctant to think for themselves? Latin medieval culture offers many grounds for a negative answer. For starters, there were occasional thinkers who were happy to proclaim their own originality. The most flamboyantly unusual was Ramon Llull (d.1316), who developed a universal, combinatoric ‘Art’ for answering all possible questions in philosophy and theology. He unsuccessfully tried to get the scholastics in Paris to embrace this new method (Fidora, Reference Fidora2008), which he thought could even lead to world peace by establishing the truth of Christianity on rational grounds. Admittedly Llull is nobody’s idea of a typical medieval philosopher. But about a generation earlier Roger Bacon was stressing the possibility, and urgency, of making new discoveries in philosophy and the sciences by taking recourse to empirical science (scientia expermentalis; see Hackett, Reference Hackett and McEvoy1995). Stubborn readers might say that Bacon too was more exception than rule. His empirical approach is often held up as an anticipation of trends in early modern European science, rather than being at home within medieval thought.
This assumption might be challenged, since it’s not as if Bacon was the only medieval thinker with empirical leanings. But I instead want to consider a phenomenon that no one could deny to be typical of medieval thought, namely the scholastic ‘disputed question’. This form of writing was based on actual teaching practices at the medieval universities. It is most likely to be familiar to you from Thomas Aquinas, but was used pervasively in scholastic philosophy from the thirteenth century onwards (extending, by the way, well into the ‘Renaissance’ and beyond thanks to the longevity of the scholastic tradition, but that’s another story). In a disputed question, the term auctoritas refers not so much to a canonical author whose word is taken to be true without question, but to a given passage from a canonical author, whose word is worth at least taking seriously. (For example Aquinas regularly cites Averroes as an authoritative interpreter of Aristotle, but he also wrote a whole treatise critiquing Averroes’ reading of the Aristotelian theory of mind.) Typically, in a disputed question such passages are arrayed to support alternative approaches to a problematic issue. The goal of the author is to show how the ‘authorities’ can be reconciled, interpreted, massaged, and occasionally set aside, in order to arrive at a position on the issue at hand that both makes intrinsic sense and fits the textual tradition.
In this context, an appeal to authority was not a discussion-ending move that precluded independent thought or creativity. To the contrary, it was an instrument for generating and exploring problems. A particular pure example of this tendency of authorities to create, rather than resolve, difficulties is the work Sic et Non (a title I like to translate On the One Hand and On the Other Hand) by the early scholastic Peter Abelard (d.1142). It simply sets out contradictory passages from the tradition without attempting to reconcile them. And we have other evidence that Abelard already considered the invocation of authority to be inadequate, by itself, for resolving intellectual problems. He tells a story in his own autobiography about a dispute with a colleague who accused him of having said something heretical (Radice and Clanchy Reference Radice and Clanchy2003, pp. 21–22). When Abelard tried to explain himself, the colleague waved this away and demanded that he produce support in the form of an authoritative quotation. After Abelard did so, citing Augustine, the colleague insisted that Abelard was understanding the passage incorrectly. Abelard retorted that this was irrelevant, since the colleague was ‘looking only for words, not interpretation’. Here we are not far from the refreshingly skeptical comment of Alan of Lille, that authorities are like a wax nose: they can be twisted in whichever direction you like (Stiefel, Reference Stiefel1977, p. 359).
All this makes clear that the intellectuals of Latin Christendom did not learn to be wary of authority only in the Renaissance. But it is in medieval Islam that we see an even more thoroughgoing rejection of slavishness to authority. For in that culture, there was a specific word used to denigrate such slavishness: it was called taqlīd. Etymologically this term has to do with being ‘bound’ or ‘gird’, like a collar being hung around the neck of an animal. In English it is often translated as ‘imitation’ or, less pithily, ‘uncritical acceptance of authority’. However it is translated, the important point is that it was a pejorative word. To accuse someone of taqlīd was to suggest that they were not thinking for themselves when they should be. That caveat (‘when they should be’) is important, because according to these same intellectuals, some people should engage in taqlīd. In fact, most people should. If one is not a trained scholar, one should not simply ‘think for oneself’ and make one’s own determinations. Rather one should seek the advice of someone who is a trained scholar. But the scholars, called ‘those who know (ʿulamāʾ)’, were not supposed to be satisfied with merely quoting or imitating reputable authorities. In fact, it was arguably the avoidance of taqlīd that defined membership of the intellectual elite.
The original context of this idea was legal theory (El Shamsy, Reference El Shamsy2008; Ibrahim, Reference Ibrahim2016). In Islamic jurisprudence, the sources of legal injunctions are the Qurʾān itself and ḥadīth, meaning reports of things the Prophet Muḥammad said and did during his lifetime. Where an explicit injunction can be found in these sources, making a legal determination is relatively clear-cut. It says explicitly at Qurʾān 2:150, for instance, that one should face Mecca when praying, so no legal theory or independent reflection is required to determine that this is the correct direction to face in prayer. I say ‘relatively’ clear-cut because even here, of course, questions might arise. What should believers do if the time for prayer has come, and they are not sure which way Mecca lies? But the more acute problem is that many questions are not directly addressed in the Qurʾān and ḥadīth at all. One legal tradition, the ẓāhirī school, took the view that only the explicit injunctions are binding and everything else is permitted; the name of this school expresses their legal position, since they followed the ẓāhir or ‘evident’ meaning of the religious source texts. But the four main branches of Sunni jurisprudence allowed legal scholars to hand down rulings by extrapolating from what they could find in these source texts (Hallaq, Reference Hallaq1997; Hallaq, Reference Hallaq2005). To give the most famous example, wine is explicitly forbidden, so other alcoholic beverages are taken to be forbidden also, by ‘analogy (qiyās)’.
This kind of reasoning was often called ijtihād, literally ‘exerting oneself’ to arrive at a considered judgment. In cases where a question is not straightforwardly settled by the religious source texts, the only alternative to ijtihād is taqlīd. Rather than working out such a judgment for oneself, one follows the reasoning offered by someone else. This might be, for example, one of the authoritative jurists who founded the aforementioned legal schools, or later jurisprudents working within one’s school tradition. For this reason we can actually think about the use of taqlīd as being analogous to the legal concept of precedent (Jackson, Reference Jackson1996). A court may for instance hand down ten years punishment for a given crime simply because this is the amount that was given in the past, without giving much if any thought to the reasons why ten years are neither too few nor too many. But actually qualified jurists were not supposed to just follow these school founders by blindly repeating their judgments. That would be potentially dangerous, after all: if one does not understand the rationale for a previous ruling one might re-apply it in a situation where it is not relevant. Instead, jurists were encouraged to understand the ‘rationale’ or ‘proof’ (ḥujja) for the ruling.
There were degrees of independence here. The extremes would be pure ijtihād, which is figuring something out completely on your own, and pure taqlīd, which is taking the word of a legal authority without knowing why that authority decided as they did. A simple case might be someone consulting a judge about whether it is all right to snort cocaine, and simply accepting the answer ‘no’ without getting into the rationale for this answer, namely that it is intoxicating, and thus analogous to wine. This is qubūl qawl bi-lā ḥujja, ‘accepting a position without any rationale’. Between the two extremes, one might make sure to understand the rationale by which a jurist made their decision. Or, one might make an effort to reach one’s own decisions, but building on the precedent of one’s chosen legal tradition. This was called ‘ijtihād within a school (ijtihād fī l-madhhab)’ or ‘affiliated ijtihād (ijtihād muntasib)’. It should also be noted that, while these points all come from the Sunni legal tradition, things were not that different in Shiite Islam, with the exception that the religious source texts are considerably expanded, with the Qurʾān and ḥadīth being supplemented by the teachings of the Shiite Imams.
There is some historical debate over the question of how much independence was allowed to jurists, and whether, as it is sometimes put, the ‘gate of ijtihād’ was ultimately closed, with legal theory giving way to strict following of precedent (Hallaq, Reference Hallaq1984). But for our purposes it is sufficient to know that in the medieval period, legal scholars often reserved to themselves the right to engage in ijtihād. For the common run of people, those the Greeks called hoi polloi, the recommendation was to stick to taqlīd in order to avoid unwittingly violating God’s law. As one legal theorist put it, ‘it is incumbent on the ordinary person (ʿāmmī) to act by taqlīd if he is incapable of ijtihād’ (Calder, Reference Calder1989, p. 61).
Now, one thing that the medieval Christian and Islamic worlds had in common is that there was quite a bit of overlap between expertise in law and expertise in theology. As a result, the contrast between ijtihād and taqlīd also came to be applied to theological doctrines and religious belief in general. One might, for instance, believe in God simply because one was raised in a Muslim family; but one might also believe in Him because one was in a position to offer a proof (again, ḥujja) of His existence. In theology as in law, the scholarly elite prided themselves on avoiding taqlīd. As members of the ʿulamāʾ, the scholars could explain all their religious beliefs by offering their own rational argumentation and competent interpretation of the Scripture. Common folk were assumed to be incapable of this, whether because they lacked opportunity or lacked talent. A telling remark on this point comes from the theologian al-Isfarāʾīnī (d.1027), sorting such people into two categories:
One [category] consists of people who are not wholly lacking in a kind of reasoning, even if it is imperfect in its expression and its grounding. Such people are truly believers and in the proper sense, know (ʿārif). The other [category] consists of people who are completely unenlightened in this respect and have no real knowledge, rather, since they believe through taqlīd, their belief lacks integrity and not one of them is free of uncertainty and doubt. (trans. Frank, Reference Frank1989, p. 45)
Some theologians went even further, adopting not only a condescending tone towards people of pure taqlīd, but going so far as to say that for core religious beliefs every Muslim had a responsibility to gain at least a modest level of independent understanding. It is reported that in one West African community, for instance, ordinary Muslims were questioned philosophically by the scholars about the basis of their belief in the oneness of God (Van Dalen, Reference van Dalen2016, p. 142). Usually though, taqlīd was felt to be appropriate for the simple believer, and inappropriate for the scholars.
Which raises the question, what is so bad about taqlīd? After all, if I am a Muslim wondering whether I can drink whiskey, it shouldn’t matter whether I figure out that it is forbidden on my own, or am told it is forbidden by a legal scholar. The main thing is that I get the right answer, so that I may fulfill my religious obligations. Does ijitihād have anything to recommend it, apart from being the privilege of an elite class? One concern might be that figuring out one’s religious obligations for oneself is itself a religious obligation, at least for those who are capable of it. Some Muslim theologians taught the following: ‘whoever is capable of knowledge of God becomes an unbeliever if he does not apply knowledge to know God, regardless of whether he abandons knowledge to pursue taqlīd, doubt, conjecture, or ignorance’ (trans. El Omari Reference El Omari2016, pp. 152–53). The thinking seems to be that if one has the capacity to reflect for oneself, then one would be failing to use this God-given gift by depending on the advice of others. To engage in taqlīd would be a kind of culpable laziness.
This view was echoed by figures working within the tradition of Aristotelian philosophy. Which is in a sense unsurprising, since Aristotle’s depiction of the best or ‘happy’ life for humans revolves around the notion of making full use of one’s inborn capacities. In particular, the power of reason is distinctive to human beings, and so being an excellent human demands making excellent use of reason. In the practical sphere this would consist in exercising reason to make virtuous choices. In the theoretical sphere, it would involve achieving full-blown scientific understanding. In neither case would it be sufficient to blindly follow the lead of other people. In the Islamic world, the Aristotelian philosopher who made this point most emphatically was Ibn Rushd, usually known under his Latinized name ‘Averroes’ (d.1198). He wrote that ‘the name “human” is predicated equivocally of a human being who is perfected by a speculative science and of one who is not perfected by it or who does not have aptitude to be perfected by it’ (Adamson, Reference Adamson, Adamson and Fay Edwards2018, p. 93). A human who does not achieve scientific understanding is not a ‘human’ in the full and proper sense, just as an eye that cannot see is not an ‘eye’ in the full and proper sense, because it fails to carry out its natural function.
Averroes thought that this same message could be taken from the Qurʾān. His most widely-read work, The Decisive Treatise (Faṣl al-maqāl), begins with a series of scriptural quotations intended to establish that there is a religious obligation to engage in philosophical inquiry (trans. in Hourani, Reference Hourani1976). For example, ‘reflect, you who have vision’, and ‘have they not reflected upon the kingdoms of heaven and earth, and whatever God has created?’ (Qurʾān 59:2, 7:185). Now, even Averroes did not think that every Muslim has the responsibility to become versed in philosophy. But he did think that those who are capable of it are obligated to do so. As a faithful Aristotelian, he thought the guidelines to be followed here were laid out in Aristotle’s logical writings, especially the Posterior Analytics, which offers an account of scientific, demonstrative understanding. Among the criteria such understanding must fulfill are the requirement to grasp the reason or explanation why things are true. (More technically: a demonstrative argument involves connecting two terms through an explanatory ‘middle’ term.) Having a demonstrative argument is the philosopher’s analogue – or Averroes would have said, the philosopher’s superior version – of the ‘rationale’ invoked in Islamic law and ‘proof’ offered in Islamic theology.
Just like the jurist and the theologian, then, the Averroist philosopher is a member of a self-appointed elite that is distinguished by the use of independent reasoning. And Averroes was certainly no less elitist than the jurists and theologians. He routinely contrasted ‘the scholars (al-ʿulamāʾ)’ to ‘the many (al-jumhūr)’ and in the Decisive Treatise explained that demonstrative arguments are for the former, since only they can understand such arguments. In a clear hint that he took this conception of elite reflection from the legal tradition, he even compared independent philosophical inquiry to the use of ijtihād in jurisprudence. ‘The many’, by contrast, should not even attempt to pursue such inquiry. It will only confuse them to be confronted with demonstrative discourse, which may even lead them into unbelief. An example might be that exposing common folk to proofs of God’s incorporeality could convince them that God does not exist at all, if they cannot imagine a being with no body.
So when approaching these people – who will of course be the vast majority – one should instead use non-demonstrative discourse. This might be merely ‘dialectical’ argumentation, which draws on commonly accepted premises rather than establishing a chain of argument on the grounds of absolutely certain first principles, or ‘rhetorical’ discourse, which aims at mere persuasion. To believe that something is true because of dialectical or rhetorical discourse would be taqlīd, which from Averroes’ point of view is entirely appropriate for the common believer. He was for the most part rather disdainful towards the theologians, whose vaunted proofs he thought were at best dialectical arguments, at worst just bogus sophistries. But he would have agreed with the Muʿtazilite theologian al-Kaʿbī (d. 931), who said that some people are ‘obligated to apply taqlīd and conjecture: these are the laypeople (al-ʿawāmm), the slaves, and many women’ (El Omari Reference El Omari2016, p. 159).
So this is one reason to avoid taqlīd: because if one has the talent and opportunity to do better by engaging in independent reasoning, then one also has the obligation to do so. But there is another reason. If I believe something merely on the basis of authority, then my belief will only be as reliable as the authority I have chosen. But how do I know that this authority is worth following? It could be that they are teaching me falsehoods. After all, I have not reflected for myself as to whether their claims are true, so how can I rule this out? In fact, even if what they have told me to believe is true, then it looks like I have wound up with a true belief only by luck. For a concrete example, we can consider a famous saying ascribed to the Prophet: ‘every child is born in the innate condition (fiṭra) but his parents make him a Jew, Christian, or Magian’. From the Muslim point of view, Christian children are brought up believing falsehoods, such as that Christ was the incarnation of God or that God is a Trinity. If they had been more fortunate, they would have been born into Muslim families and thus would have had true instead of false beliefs. From the Christian point of view, of course, things are the other way around: Muslim children are unfortunate enough to be brought up thinking that Christ was not God and that Muḥammad was a genuine Prophet. But either way, the upshot is disconcerting. In matters of such importance, it seems inadequate simply to hope that we’ve gotten lucky.
The problem is still with us, and discussed by philosophers of religion under the heading of ‘religious pluralism’. Once you learn that there are believers in other religions (and people with no religious belief at all), how can you be satisfied with simply adopting the faith into which you happen to have been born? Actually, this worry about taqlīd would still arise even if there were no diversity of opinion. Even if every human on earth believes something, that doesn’t prove it is true. What would prove it true would be, well, proof: and to have a proof is precisely not to believe by taqlīd. Still, we are more likely to worry about luck and unluck in belief where we are confronted by pluralism, that is, where there is observable difference of opinion. And the problem of specifically religious pluralism was certainly noticed by medieval philosophers. Here we might once again make mention of Ramón Llull, a Christian who grew up among Muslims on his native island of Majorca. He wrote a dialogue showing how his ‘Art’ could resolve disagreement between a Jew, a Muslim, and a Christian (trans. in Bonner, Reference Bonner1993). And Llull was, of course, only one of many medieval philosophers who believed that rational argument could show the viability of their favored Abrahamic religion and establish the incoherence of the rival faiths.
But I’d like to follow this line of thought in a work by someone who, famously, did not consider himself a ‘philosopher (faylasūf)’, at least in the sense this word carried in medieval Arabic. This was al-Ghazālī (d. 1111), who remains famous as a theologian and critic of Aristotelian ‘philosophy (falsafa)’ in the version put forward by Ibn Sīnā (Avicenna, d. 1037). He has this reputation on the basis of his Incoherence of the Philosophers (Tahāfut al-falāsifa, trans. Marmura, Reference Marmura1997), where he aimed many criticisms at Avicenna and other philosophers. Among those criticisms was the charge that they engaged in taqlīd, because they blindly followed authorities like Aristotle (Griffel, Reference Griffel and Günther2005). If you know anything about Avicenna you may find this accusation strange, or even risible. For in fact, Avicenna was distinguished by his innovative and occasionally even irreverent handling of Aristotelian philosophy, which he quite deliberately and explicitly reworked so as to produce a new and distinctive system (Gutas, Reference Gutas1988). But what we have seen so far in this paper may help explain why al-Ghazālī attacked Avicenna in these terms. The label of taqlīd was chosen to sting, since philosophers no less than jurists and theologians prided themselves on their independence of mind. It’s what all members of the intellectual elite in the Islamic world had in common (so much for the notion that all medieval thinkers were happy to present themselves as being bound by authority and unoriginal).
For al-Ghazālī’s own approach to authority and its use, we’ll be better served by turning to another text, his intellectual autobiography the Deliverer from Error (Munqidh min al-dalāl, trans. in Watt Reference Montgomery Watt2000; I quote in my own translations from the edition in Jabre, Reference Jabre1969). In the opening section of this work, al-Ghazālī explains the background to a skeptical crisis he endured, when he came to think that none of his beliefs were absolutely certain. Beliefs grounded in sensation need sometimes to be corrected by reason. So how can we guarantee that reason itself stands in no need of correction, or validation, by a higher epistemic court of appeal (Jabre, Reference Jabre1969, p. 13)? The reason why this realization hit al-Ghazālī so hard is that he had made it his life’s goal to achieve certainty, never contenting himself with beliefs formed through taqlīd:
Thirst for grasping the true natures of things was a habit and practice of mine from early on in my life, an inborn and innate tendency (gharīza wa-fiṭra) given by God in my very nature, not chosen or contrived. So as I neared maturity the bonds of taqlīd weakened for me and I was emancipated from inherited beliefs. For I saw that young Christians always grew up to accept Christianity and young Jews to accept Judaism, while young Muslims always grew up to accept Islam. And I heard the ḥadīth related of the prophet, ‘every child is born in the innate condition (fiṭra) but his parents make him a Jew, Christian, or Magian’ (Jabre, Reference Jabre1969, pp. 10–11).
‘Inherited beliefs’ were not enough for the young al-Ghazālī, because he realized that religious communities who disagree strongly with one another are all following such inherited beliefs. In forming one’s religious convictions, it is surely not enough to hope that one was lucky enough to be born into the right family. Thus he quotes exactly the aforementioned ḥadīth, drawing from it the lesson that taqlīd exposes one to the vagaries of epistemic luck.
Al-Ghazālī’s struggle to overcome skepticism comes in several stages, which we do not need to discuss in detail here. In part his escape is thanks to a ‘light cast into his breast’ by God, which gives him a powerful subjective sense of certainty as concerns certain self-evident principles, like basic truths of logic and mathematics. But he does not try to ground his belief in Islam in such a feeling of certainty, perhaps because a ‘Jew, Christian, or Magian’ could do the same, again leaving al-Ghazālī on an epistemic par with those whose beliefs he considers false. Actually, it seems that al-Ghazālī could have made that move, simply saying that he feels certain he is right in his religious belief, and actually is right. Members of other religions might feel equally certain, but there would be a crucial difference: their beliefs happen to be wrong. This would satisfy some philosophers. For example Alvin Plantinga’s ‘reformed epistemology’ allows the Christian to take belief in God as ‘properly basic’, which means the Christian can rationally commit to it without further argument or justification, even while recognizing that the belief is not basic for atheists. But such expedients do not seem to attract al-Ghazālī. He wants his beliefs to attain a standard of certainty such that all reasonable people would adopt them, if they were to perform due epistemic dilgence.
What al-Ghazālī needs, then, is a means of establishing with certainty that Muḥammad’s prophecy was genuine, since the truth of Islam turns on this claim. A common way to do this would have been to appeal to miraculous feats performed by the prophet. In Islam, emphasis is placed on the Qurʾān itself, which is said to be ‘inimitable’ by normal human means (for the origins of this idea in Islamic theology see Martin, Reference Martin1980). Indeed this is traditionally recognized as Muḥammad’s only miracle. But again, this is a game that can be played by representatives of other Abhramic faiths. For instance Judah Hallevi’s dialogue Kuzari shows a king being converted to Judaism on the strength of the well-attested miracles in Jewish history (for comparison between Hallevi and al-Ghazālī see Baneth, Reference Baneth and Jospe1981; Kogan, Reference Kogan and Inglis2002). Perhaps in part for this reason, al-Ghazālī rejects apparent miracles as insufficient grounds for accepting the veracity of a prophet. What seems to be a God-given miracle could be a trick worked by human means.
Instead, he proposes the following:
If you are in doubt about whether a certain person is a prophet or not, certainty can be had only through knowledge of what he is like, either by personal observation or reports and testimony. If you have an understanding of medicine and jurisprudence, you can recognize jurists and doctors by observing what they are like, and listening to what they had to say, even if you haven’t observed them. So you have no difficulty recognizing that al-Shāfiʿī was a jurist or Galen a doctor, this being knowledge of what is in fact the case and not a matter of taqlīd shown to another person. Rather, since you know something of jurisprudence and medicine, and you have perused their books and treatises, you have arrived at necessary knowledge about what they are like. Likewise, once you grasp the meaning of prophecy and then investigate the Quran and [ḥadīth] reports extensively, you arrive at necessary knowledge that [Muḥammad] is at the highest degree of prophecy. (Jabre, Reference Jabre1969, p. 43)
This passage offers a middle path between the blind obedience of taqlīd and the do-it-yourself ambitions of ijtihād and philosophical proof. Al-Ghazālī recommends obedience that is not blind, the acceptance of authority on good grounds. Just as one could verify that someone is a doctor by considering ‘what he is like (aḥwāluhu)’, so too we should convince ourselves that someone is a prophet by seeing the way he conducts himself, or reading about him. What we’re looking for is not miracles, but wisdom, virtue, and holiness. This then would be al-Ghazālī's answer to the problem of religious pluralism: any fair-minded person who reads deeply about the life of Muḥammad will come away from this with the conviction that this man, surely, was a genuine Prophet, and therefore that Islam is the true faith.
While al-Ghazālī does not say quite as much about this process as we might wish, it is clear that the process is a fairly demanding one. As he says here regarding jurisprudence and medicine, you need to have ‘perused the books and treatises’ of these specialists. His idea is not, though, that you need to have transformed yourself into one of the specialists. Rather, the goal is to be in a position to verify the expertise of the genuine specialist without having such expertise oneself. We will only be in such a position once we have ‘necessary knowledge’ concerning the reliability of the specialist, or prophet. This is a term of art in Islamic theology: knowledge is ‘necessary (ḍarūrī)’ or ‘immediate (badīhī)’ if it leaves no room for doubt, as happens with first principles of reasoning and the immediate deliverances of sense-perception (Abrahamov, Reference Abrahamov1993). Some theologians say that necessary knowledge comes in part through the ‘innate condition (fiṭra)’, mentioned in that same Prophetic report about people born into different faiths.
Whatever its source, ‘necessary knowledge’ can provide the basis for further, ‘acquired (muktasab)’ knowledge, which stands in need of proof and is always less certain than necessary knowledge. For example speculative argumentation, in theology or any other field of science, would yield conclusions that have the status of acquired, not necessary, knowledge. Here we might return to the case of mathematics. When al-Ghazālī says that he had subjective certainty concerning mathematical truths, he means basic ones that are immediately obvious and thus constitute ‘necessary knowledge’, like ‘ten is greater than three’. The conclusion of a complex mathematical proof would, by contrast, qualify as ‘acquired’ knowledge even if it takes its starting point from premises that are known necessarily. Al-Ghazālī’s account of the verification of prophecy and expert knowledge has precisely this structure. We begin by familiarizing ourselves enough with the field of expertise that we are sure (that is, have ‘necessary knowledge’) who is an expert and who is not. Now we find an expert, and take their advice. This is not taqlīd, since that would be following an authority uncritically. But neither is it full-blown ijtihād, the pure use of independent judgment. Here, we do follow an authority figure, but do so after adopting a critical frame of mind and testing the credentials of that authority.
In the legal context that spawned the concept of taqlīd, and serves as one of al-Ghazālī’s main examples, this approach might be contrasted to the aforementioned practice of ‘affiliated ijtihād’. As we saw, that involved accepting the principles of a given legal school and then working out one’s own legal judgments within those parameters. What al-Ghazālī describes has the reverse structure. Instead of taking one’s starting points from an authority and working independently from there, he recommends using independent thought at the stage of choosing one’s authorities, but thereafter depending on those authorities. In a juridical context this could mean, for instance, working out which of the founding legal scholars to follow, then following him. (It is no coincidence that in our key passage, al-Ghazālī mentions verifying al-Shāfiʿī’s (d.820) expertise in law, since al-Ghazālī was himself a jurist of the Shāfiʿite school.)
This, I think, is an attractive answer to the question of how we should use authority. We can recognize the basic policy as one we use all the time, as when – to take al-Ghazālī’s other example – we find a doctor we trust and then follow their advice. But remember that al-Ghazālī sets a fairly high bar when it comes to bestowing one’s trust upon an expert. It’s not good enough to choose a lawyer because they are wearing a sharp suit and carrying a briefcase, or a doctor because of their reassuring aura. Rather, you need to know something of law and medicine yourself in order to make an informed choice. In a modern context, al-Ghazālī might have in mind the extent of expertise you would build up over the course of taking an undergraduate degree. Your BA in literature doesn’t qualify you to write books about Shakespeare, but it will help you verify that a certain Shakespeare scholar you meet at a party really knows their stuff. Of course, we will all still need to engage in taqlīd in many spheres of knowledge. You might be pretty good on Shakespeare, but when it comes to particle physics you have to surrender to epistemic luck, because you aren’t even in a position to tell expert physicists apart from charlatans, never mind do physics yourself.
Even though al-Ghazālī’s proposal comes from the rather special, and for us perhaps unfamiliar, context of a discussion of prophetic veracity, it applies with striking relevance to the problems we face in modern society. We are beset not just by religious pluralism but also disagreement over highly complex problems like climate change, the benefits of international trade agreements, and of course the best way to deal with a global pandemic. The first lesson we should take from al-Ghazālī is to be fairly modest in our pretensions of certainty. If you haven’t devoted years of specialized study to climatology, economics, and epidemiology, then with respect to the issues just mentioned you are better off following the authoritative pronouncements of others than coming to your own, necessarily ill-informed, view. After all, none of us have time to become experts in all these fields. But this does not mean we are doomed to follow authority uncritically. Our goal should be to ‘weaken the bonds of taqlīd’, as al-Ghazālī put it. In other words, we should work at improving our ability to recognize the expertise of other people, so that we may responsibly give those people our credence. This policy, with its combination of humility and demandingness, might be medieval in inspiration, but it remains fit for purpose in today’s world of widespread uncertainty.
