Yet these citations are highly imprecise, and all appear to come from Book I of the Politics. Another author from the same period, Miskawayh, states that either one or two books of the Politics are listed in the catalogue of Aristotle’s writings with which he was familiar (Pines, 153, 155). Pines suggests that only the first two books were available, and perhaps not in the form known to us, but rather in a Hellenistic or Roman paraphrase (155, 160). Pines also raises the possibility that Alfarabi could have altered Aristotle for his own purposes, especially in the Political Regime (156–59).
In an article written in 1993, Rémi Brague comes to a conclusion similar to that of Pines, although he adopts a somewhat more strident tone. Brague also calls our attention to some new pieces of evidence, most notably a couple of quotations from Averroes (Brague 1993, 428–30). According to the first, Aristotle’s Politics “did not fall into our hands” (Averroes 1974, 22.4–6). Whomever “our” may refer to, it evidently does not include Alfarabi, since a second passage, cited by Brague in Latin, suggests that in Averroes’s view Alfarabi did possess this work (Brague 1993, 429). But how could Averroes in twelfth-century Andalūs know what books Alfarabi would have possessed in tenth-century Baghdad? Averroes’s claim appears especially problematic given that Alfarabi has left us no explicit references to the Politics, even in his treatise titled the Philosophy of Aristotle.40 Brague shows, with painstaking philological analysis, that the passage in which Alfarabi invokes an Aristotelian work on political science in fact refers to the Ethics (430–32). He concludes by echoing and even citing Pines: the medieval Muslims had at best unreliable quotations and summaries of the Politics (432).
None of this circumstantial evidence ever convinced Muhsin Mahdi that Alfarabi was unfamiliar with the Politics (Mahdi 2001, 34–36). Mahdi observes that Alfarabi does not explicitly declare his unfamiliarity with the work, and silence does not prove ignorance (35). Yet doesn’t Averroes declare his unfamiliarity? True, but he never declares Alfarabi’s (cf. Brague 1993, 429). Furthermore, the context of Averroes’s claim is suspicious. He states that the presence of Plato’s Republic compensates for the absence of the Politics, since both contain the practical part of political science. But if Averroes was truly unfamiliar with the contents of the Politics, on what grounds could he claim that it contains the practical part of political science, or that Plato’s Republic represents an adequate substitute for it (Mahdi 2001, 35)? Finally, Mahdi doubts that a book which had simply been absent from the Muslim world could resurface so quickly and easily in medieval Europe (34). Mahdi accounts for the absence of the Politics by supposing that the philosophers concealed it deliberately, because they regarded the political philosophy of Plato as more suitable for their project of reviving philosophy in the Islamic world. In particular, Aristotle’s arguments for the self-sufficiency of practical wisdom would have seemed preposterous in a world dominated by divine law (35–36). Mahdi’s conjecture about Alfarabi pretending not to know the Politics resembles my earlier conjecture about Alfarabi pretending not to know Greek: he wished to conceal from his contemporaries his familiarity with something many of them would have viewed with suspicion. Such guesses are plausible and intriguing in both cases, but they hardly amount to certainty.
I regard Alfarabi’s knowledge of Aristotle’s Politics as more doubtful than his knowledge of Plato. As far as Plato is concerned, we should be inclined to take Alfarabi at this word, but in the case of Aristotle’s Politics, there is no such word, since Alfarabi neither betrays nor claims thorough knowledge of the work. Alfarabi certainly could have appreciated Aristotle’s stature as a philosopher, and even as a political philosopher, on the basis of his other works, including the Ethics (cf. Mahdi 2001, 35). I therefore remain somewhat skeptical about Alfarabi’s knowledge of the Politics, and do not attempt to show that he interprets its teachings in any detail. Nevertheless, the passages of Alfarabi that seem most reminiscent of Aristotle, leading both Pines and Brague to think that he must have had access to some summary of Book I (Brague 1993, 432; Pines, 157), can still be profitably compared and contrasted with Aristotle. This will at the very least shed light on the substantial difference between the politics and political philosophy of ancient Greece and medieval Islam, thus helping to explain why the Politics never gained much currency in the medieval Islamic world in the first place.
One of the passages adduced by Pines as having left some echoes in medieval Islam occurs in the very first chapter of the Politics. Pines makes a useful and sensible comparison between this passage and some parallel passages in Alfarabi (Pines 156–59), and I wish to build on his example. Aristotle famously claims that “man is a political animal” (Politics 1253b7–8), and a version of this statement recurs in Alfarabi, in both the Virtuous City and Political Regime (VC 228.1–8; PR 60.64, Ar. 69.16–17). Yet the terms used by each philosopher diverge already at this point. Aristotle speaks of humans as politikon, an adjective whose root is linked to the Greek polis, and takes care to distinguish the ethnē unfavorably from the developed polis (Politics 1252b19–20). Alfarabi speaks of humans as attaining perfection only in an ijtimā‘, or association, which in Arabic signifies a cooperative political community of undefined size: it could describe a household, city, nation, many nations, or even the entire inhabited world (PR 60.64, Ar. 69.17–19, VC 228.10–230.2). This points to a possible difference between the two philosophers: while Aristotle privileges the polis, in this chapter in particular and throughout the Politics in general, it is by no means clear that Alfarabi follows him on this point. Scholarly discussion of this question has been largely inconclusive. Let us take a closer look.
The passages most often cited in favor of the city occur in the Political Regime and the Virtuous City. In the former, Alfarabi says that “The city is the first in the rankings of perfections” (PR 60.64, Ar. 69.20). In the latter, Alfarabi says that “The noblest good and the furthest perfection are obtained first of all in a city, not in any association that is more deficient than it” (VC 230.3–4).41 Fauzi Najjar links this statement to the recently cited passage in Aristotle’s Politics that praises the city (Najjar 1954, 108–9). But Najjar’s assumption seems somewhat hasty. Given how unsure we are about Alfarabi’s knowledge of the Politics, we can hardly suppose that he agreed with it on all points. Besides, Alfarabi distinguishes between the city, Umma, and the multinational association in two major ways: size and capacity for perfection (cf. Naṣṣār 1978, 36–37). The meaning of Alfarabi’s statements therefore depends on the interpretation of “first.” Does it mean first in perfection, as Najjar assumes, or simply smallest in size, among the communities that are equally capable of perfection? The context seems to suggest the latter, since in both passages the city, Umma, and multinational association are listed according to size rather than virtue. These three types of association are all defined as “perfect,” in contrast to households, streets, neighborhoods, and villages, deficient associations that ought to be subordinate to the city (PR 60.64, Ar. 69.19–70.1; VC 228.10–230.1). Thus “first” is more plausibly taken to signify the city’s superiority to these deficient, subpolitical associations, rather than its relation to the Umma. Alfarabi defines the city as the smallest of the associations that are capable of human perfection: on this point he agrees with Aristotle (Politics 1252a27–30). But Alfarabi does not follow Aristotle in proclaiming its superiority to larger associations, as some scholars have already noted (Mahdi 2001, 140; Cité vertueuse, trans. Cherni, 218 n. 4). The smallest might actually mean the weakest, as Naṣṣār assumes, arguing that “the perfection realized in each of these associations differs from the other in degree: the association arising in a city is beneath the association arising in an Umma, and the latter is beneath the association comprising all of humanity” (Naṣṣār 1978, 37). Alfarabi also states that “the unqualifiedly perfect human community is divided into Ummas” (PR 60.64, Ar. 70.5), which Pines plausibly interprets to mean that “bigger communities are more perfect” (Pines, 156).
A further proof that Alfarabi does not regard small size as inherently good is that he makes no attempt, even in his commentaries on Greek works, to limit either the territory or population of the political community to any numerical size. Plato’s restriction of the city of the Laws to 5,040 households (Laws 737e ff.) is still deemed too large by Aristotle (Politics