Avicenna and Galen, Philosophy and Medicine: Contextualising Discussions of Medical Experience in Medieval Islamic Physicians and Philosophers

In: Oriens
View More View Less
  • 1 The University of Manchester

In this article I discuss Greek and Arabic philosophical and medical debates about experience (taǧriba, empeiria). I consider the Greek and classical Arabic background for debates about experience among Arabic commentators on the Hippocratic Aphorisms. I argue that these authors are influenced by Galen’s ideas about experience in his pharmacological and dietetic writings, and Aristotle’s ideas about experience, expressed mainly in Posterior Analytics, Book Two. I argue, however, that the Aristotelian viewpoint of experience reaches the Arabic Aphorisms commentators through the intermediaries of Aristotle’s Platonist commentators and Avicenna. I show that most of the Arabic Aphorisms commentators understand experience to have the various meanings Galen assigns it in his medical writings. Ibn al-Quff is the lone, but no less intriguing, exception. In his Aphorisms commentary, Ibn al-Quff uses Avicenna’s definition of experience in the book On Demonstration (Kitāb al-Burhān) from Avicenna’s summa The Healing (Kitāb al-Šifāʾ) to explain Hippocrates’ words. Closely examining Avicenna’s On Demonstration, Book One, Chapter 9, reveals that Avicenna continues late antique trends, which meld medical and philosophical debates. Avicenna uses Galen’s idea of qualifed experience to resolve interpretive challenges in Aristotle’s Prior Analytics, Book Two, Chapter 23 and Posterior Analytics Book two, Chapter 19, where Aristotle speaks about experience’s role in the inductive process of knowledge acquisition. I argue that the fluid way in which Ibn al-Quff deploys Avicenna’s On Demonstration to explicate the Hippocratic Aphorisms marks a shift in which Avicenna’s philosophical thought becomes increasingly influential in post-classical Islamic medical discourse.

Abstract

In this article I discuss Greek and Arabic philosophical and medical debates about experience (taǧriba, empeiria). I consider the Greek and classical Arabic background for debates about experience among Arabic commentators on the Hippocratic Aphorisms. I argue that these authors are influenced by Galen’s ideas about experience in his pharmacological and dietetic writings, and Aristotle’s ideas about experience, expressed mainly in Posterior Analytics, Book Two. I argue, however, that the Aristotelian viewpoint of experience reaches the Arabic Aphorisms commentators through the intermediaries of Aristotle’s Platonist commentators and Avicenna. I show that most of the Arabic Aphorisms commentators understand experience to have the various meanings Galen assigns it in his medical writings. Ibn al-Quff is the lone, but no less intriguing, exception. In his Aphorisms commentary, Ibn al-Quff uses Avicenna’s definition of experience in the book On Demonstration (Kitāb al-Burhān) from Avicenna’s summa The Healing (Kitāb al-Šifāʾ) to explain Hippocrates’ words. Closely examining Avicenna’s On Demonstration, Book One, Chapter 9, reveals that Avicenna continues late antique trends, which meld medical and philosophical debates. Avicenna uses Galen’s idea of qualifed experience to resolve interpretive challenges in Aristotle’s Prior Analytics, Book Two, Chapter 23 and Posterior Analytics Book two, Chapter 19, where Aristotle speaks about experience’s role in the inductive process of knowledge acquisition. I argue that the fluid way in which Ibn al-Quff deploys Avicenna’s On Demonstration to explicate the Hippocratic Aphorisms marks a shift in which Avicenna’s philosophical thought becomes increasingly influential in post-classical Islamic medical discourse.

1. Introduction

In this article, I provide historical context for debates about medical experience (taǧriba) in Arabic commentaries on the Hippocratic Aphorisms. I highlight two distinct accounts of experience that influenced classical authors such as Alfarabi (d. 950 or 951) and Avicenna (d. 1037) as well as the post-classical commentators on the Hippocratic Aphorisms.1 The bases of these accounts can be traced to Aristotle’s (d. 322 BC) Prior Analytics, Book Two, Chapter 23 (hereafter APr. 2.23), Posterior Analytics Book Two, Chapter 19 (hereafter APo. 2.19) and Galen’s (d. ca. 216 AD) mainly pharmacological writings on “qualified experience (dihōrismemē peira).” This article shows that classical and post-classical Arabic authors synthesised these two traditions and transformed the role experience plays in medicine and philosophy in ways that Aristotle and Galen never intended.

In Section 2, I outline what Galen meant by “qualified experience” in his pharmacalogical writings. I show, however, that Galen does not reserve this idea for pharmacology, since the intuitions behind qualified experience are also present in his commentary on the first aphorism in the Hippocratic Aphorisms.

I argue in Section 3 that Alfarabi and Avicenna were keenly aware of Galen’s doctrine of qualified experience. Both authors, however, express deep reservations about how effective qualified experience really is in pharmacology and anatomy. Of them, Alfarabi is by far the most critical of this aspect of Galen’s medical methodology. As we shall see, Alfarabi attacks Galen’s doctrines relating to medical experience by trying to show that the method is logically incoherent. We shall see that in his attack Alfarabi draws on logical concepts in Aristotle’s Categories and on Platonist commentaries on the Categories.

In Section 4, I enumerate the different senses that Arabic commentators on the Hippocratic Aphorisms assigned to experience in their commentaries on the first aphorism. We shall see that most of the commentators assign to taǧriba the same sense that Galen assigns to peira in his Aphorisms commentary and his pharmacology. One important exception to this rule is a commentary attributed to Ibn al-Nafīs (d. 1288) preserved in the unique manuscript Oxford, Bodleian, MS Pococke 294. In this commentary, Ibn al-Nafīs draws on Galen’s commentary on the Aphorisms, but also supplements Galen’s discussion with material inspired by the opening remarks of Avicenna’s Canon, Book Two (On Simple Drugs).

I turn to the Aphorisms commentary by Ibn al-Quff (d. 1286) in Section 5. Like Ibn al-Nafīs, Ibn al-Quff also draws on Avicenna to explain the meaning of experience in the first aphorism. However, Ibn al-Quff paraphrases passages from Book One, Chapter Nine of Avicenna’s Book On Demonstration (Kitāb al-Burhān) in the logic of The Healing (Hereafter, Dem. 1.9). In it, Avicenna addresses a number of objections to Aristotle’s account of the relationship between induction (Gr. epagōgē, Ar. istiqrāʾ) and experience in APo. 2.19 and the indemonstrable first principles of demonstrative science. In this chapter, Avicenna appropriates Galen’s ideas about finding causes by using qualified experience in order to respond to objections that the first principles acquired by induction are not suitable for use in demonstrative science. In Section 6, I examine late antique precedents for using Galenic medical concepts and texts to solve interpretive challenges in Aristotelian philosophy. I conclude that part of Avicenna’s medical legacy is that post-classical medical discourse comes to be infused with doctrines and debates from Avicenna’s philosophical works.

2. Qualified Experience in Galen’s Aphorisms Commentary

Philip van der Eijk has shown that Galen recognized that the conditions under which observations of phenomena are made impact the observations’ evidentiary value with regard to making a judgment about where responsiblity for the observed phenomena lies.2 Van der Eijk focuses mainly on Galen’s pharmacological works, in which Galen is interested in assigning responsibility for certain pharmacological effects on the body to properties of drugs. According to van der Eijk, Galen shows a “keen awareness” that in “an empirical test of a substance’s dietetic or pharmocological power” certain conditions must be fulfilled “in order to make the test have an evidential value and provide sufficiently specific information.”3 Indeed, it seems that Galen developed the concept of qualified experience based on a clear recognition that the conditions under which observations are carried out must allow us to distinguish the drug’s essential (kath’ hauto) effects from its accidental (kata sumbebēkos) consequences.4

Yet, there is evidence from Galen’s commentary on the Hippocratic Aphorisms that Galen recognized that manipulating the conditions under which observations are made to lend them greater evidentiary value is not a task whose importance is limited to pharmacology. In his commentary on the first aphorism, Galen makes the following comment on Hippocrates’ cryptic statement that “the determination is difficult (hē de krisis chalepē).” Typically, Galen’s comments serve as a thinly veiled dig at how his empiricist opponents understand reason’s role in medicine. Nevertheless, Galen alludes to the fact that the problem of how to secure the evidentiary value of empirical observations confronts any empirically-minded physician, not just the empiricist.5

τοῖς δ’ ἐμπειρικοῖς οὐχ ὁ λόγος εἰρῆσθαι δοκεῖ κρίσις, ἀλλ’ ἡ τῶν διὰ τῆς πείρας εὑρισκομένων βοηθημάτων ἐπίκρισις. ὄντως γὰρ καὶ τοῦτο χαλεπὸν καὶ δυσθήρατον, ὅταν πολλῶν ἰαμάτων τῷ κάμνοντι γεγενημένων ἕν τι ἐξ αὐτῶν αἰτιᾶται τῆς ὠφελείας ἢ βλάβης. εἰ γὰρ οὕτως ἔτυχε κοιμηθέντος καλῶς αὐτοῦ καὶ καταιωνηθέντος ἐφεξῆς καὶ καταπλασθέντος ἐπὶ τούτῳ, κἄπειτα κλυσθέντος ἢ αὐτομάτως τῆς γαστρὸς ἐνδούσης, εἶτα καὶ τραφέντος τοιάσδε τινὰς τροφὰς, κἀπὶ τούτοις ἅπασιν ὠφεληθέντος ἢ βλαβέντος οὐ ῥᾴδιον εἰπεῖν διὰ τί τῶν γεγενημένων συνέβη τὸν ἄρρωστον ὠφεληθῆναι ἢ βλαβῆναι.

‮وأمّا أصحاب التجارب فيقولون إنّه لم يرد بالقضاء القياس، وإنّما أراد به الحكم على الأشياء التي توجد بالتجربة وما يعقب. وبالحقيقة فإنّ هذا أيضاً عسر والوقوف عليه صعب، أعني أن يكون المريض قد عولج بأصناف شتّى من العلاج فتظهر منفعته أو مضرّته فيحكم أن تكون المنفعة أو المضرّة إنّما كان سببها واحد من أصناف ذلك العلاج. وذلك أنّك إن جعلت في المثل أنّ مريضاً نام نوماً صالحاً ثمّ انتبه فمسح بدهن ثمّ ضمّد ثمّ حقن أو انطلق بطنه من تلقاء نفسه ثمّ أكل طعاماً بحال ما ثمّ حدثت له بعد هذه الأشياء كلّها منفعة أو مضرّة فليس يسهل أن يقال من قبل أيّ هذه الأشياء كلّها التي كانت حدثت هذه المنفعة للمريض أو المضرّة.‬‎

[L1] The empiricists do not seem to mean by “reasoning” determining (krisis), but rather the verification of remedies discovered through experience. For, in reality, this too is difficult and ellusive whenever one of the many remedies given to the patient is the cause of benefit or harm. For example, if, after the patient slept well, had a massage, was rubbed with oil, was purged or his stomach emptied on its own, and then he ate this or that food, and after all these things he received some benefit or was harmed, it is not easy to say which of these [therapies] is responsible for the benefit or harm to the patient.

In L1 Galen draws attention to the doctrines of the empiricist sect, whose notion of reasoning in medicine does not involve reasoning deductively from universal principles to particular instances. To the contrary, Galen represents the empiricists’ notion of reasoning as a process of verifying by repeated applications that a certain correlation does in fact exist between drugs or foodstuffs and some therapeutic effect based purely on experience. According to Galen’s representation of the empiricist view, medical reasoning is a strictly non-deductive process of inferring conclusions about how to treat particular illnesses based not on deductions from universal principles of pharmacology or humoral pathology but on a finite number of analogous cases histories.

Galen may dismiss the empiricist notion of reason, but in L1 he does concede—given the importance of qualified experience in his pharmacology he must concede—that establishing a causal connection between a drug and its therapeutic effect on the body is complicated by the myriad conditions under which the putative cause (e.g. scammony) and its therapeutic effect (purging yellow bile) is observed. Indeed, the purpose of qualified experience is to lend observations of the consequences of administering scammony to the patient evidentiary value by systematically eliminating those conditions that obscure the drug’s essential effects on the body. In Galen’s view, qualified experience is a method for distinguishing the essential effects of a drug from its accidental effects. In his pharmacological writings, Galen identitifies a handful of conditions that must be controlled in order for these types of observation to have evidentiary value. Philip van der Eijk has identified several of them. These include ensuring that the body the drug is tested on is as balanced in temperament as possible, that the environment in which the drug is tested is balanced, that the drug is simple and pure and not administered to the patient in combination with other types of therapy, that the drug is administered in the season that is most suitable, and that the drug is in line with the natural temperament and age of the patient to whom it is given.6

Galen’s example in L1, however, describes a cocktail of different therapies, which are administered to the patient in such a way that ultimate responsiblity for the benefit or harm to the patient is hopelessly obscured. Nor is this example uniquely pharmacological in scope. This observation underscores the fact that for Galen the problem of improving the evidentiary value of empirical observations by controlling the conditions under which the observations take place is not just a problem in pharmacology. This problem confronts anyone who is interested in providing empirically-based evidence for a claim such as “there is a causal connection between two phenomena X and Y,” or “X causes Y,” or “X is responsible for the existence of Y.”

3. Logic of Experience in Alfarabi’s On Topical Analysis (Kitāb al-Taḥlīl)

There is strong evidence that classical and post-classical Arabic physicians and philosophers thought deeply and critically about medical experience as a problem of scientific method. These authors recognized the fact that controlling the conditions under which observations are made forms a crucial methodological element in any argument in which the conclusion that X is responsible for or causes Y is based on a finite number of premises that are derived from empirical observations.

Evidence from the classical period comes from Alfarabi’s criticism of Galen’s anatomical method in On Topical Analysis, and from Book Two of Avicenna’s Canon of Medicine (al-Qānūn fī l-ṭibb, hereafter Canon 2) on simple drugs (adwiya mufrada).7 Avicenna does not employ an equivalent of Galen’s term “qualified experience” to characterize the seven conditions that he says must be observed to ensure the value of observations of a drug’s pharmacological action. These conditions and the fact that he uses them to distinguish between a drug’s essential and accidental effects, however, show his debt to Galen’s pharmacological thought. At the beginning of this chapter, Avicenna says that the powers in drugs can be discovered by experience (ṭarīq al-taǧriba) or by a process of reasoning (ṭarīq al-qiyās).8 For experience to lead to reliable knowledge about a drug’s power, Avicenna says that the physician must observe seven conditions (šarāʾiṭ).9 It is clear that Avicenna’s thought in this section owes a great deal to Galen’s notion of qualified experience in pharmacology. Nevertheless, according to Pormann, the conditions that Avicenna sets down “are much more detailed than those found in Galen’s extant works.”10 (1) The drug should be free from any “acquired quality (kayfīya muktasaba),” which works to obscure the drugs essential activity with its accidental effect; (2) the experience should be conducted on a “simple illness (ʿilla mufrada)”; (3) the drug’s effects should be observed on a pair of illnesses whose humoral aetiologies are opposite to each other; (4) the power in the drug should correspond to the power in the illness; (5) the time in which the drug’s effects appear should be observed closely; (6) the effect should be constant and occur in most cases; and (7) the human body rather than bodies of animals should be used for discovering the drug’s effects.

In On Topical Analysis, Alfarabi takes up a polemic against Galen’s polemic in the Doctrines of Hippocrates and Plato against Aristotle, Chrysippus and Praxagoras. These authors claimed that the heart rather than the brain is the origin of the nerves. Alfarabi’s comments are intended to highlight Galen’s incompetence as a logician rather than serve as a thorough refutation of Galen’s doctrine that the brain is the origin of the nerves and of the governing part of the soul. Neverthless, the fact that Alfarabi’s comments are directed at Galen’s anatomical method rather than Galen’s pharmacology highlights the fact that Arabic physicians and philosophers believed that the method of qualified experience extends beyond Galen’s pharmacology, ramifying into Galen’s anatomy and therapeutic method. In addition, Alfarabi’s comments occassion the conclusion that Arabic philosophers and physicians believed that in Galen’s writings there is a generic inductive method for finding causes, which is on display equally in Galen’s pharmacology, anatomy, and therapeutics.

In his discussion of Galen’s arguments against Aristotle and Chrypsippus in Doctrines of Hippocrates and Plato, Books One and Two, Alfarabi reduces the premises of Galen’s argument to two “if, then” sentences:

  1. (A)if the nerves are left intact, voice, sensation and voluntary motion remain.
  2. (B)if the nerves are severed, voice, sensation and voluntary motion are impaired.

According to Alfarabi, Galen believes (A) and (B) alone are sufficient to prove that the nerves originate in the brain and that the brain is where the ruling part of the soul resides. Alfarabi analyses (A)’s and (B)’s formal features in terms of existence (wuǧūd) and removal (irtifāʿ), where the existence of property X in Y means that Y possesses the attribute X; the removal of property X from Y means that attribute X is absent from Y.11

Here, removal is not quite a kind of metathetic or privative negation.12 It is not treated as an operation on terms that transforms the term’s extension, as in the case of metathetic terms in Aristotle’s On Interpretation 10.13 Rather, saying that animal is “removed from” human means that the reasoner hypothetically considers what items would fall under the extension of the term “human” if all the items falling under the term “animal” did not exist. In this case, since animal is part of the definition of human, nothing would fall under “human” if there were nothing falling under “animal.”

In such a situation, Alfarabi and Aristotle’s Platonist commentators say that “animal is removed when human is removed, but animal is not removed when human is removed.” For example, in the Categories, Aristotle says that “prior (proteron)” has several senses. One of the senses Aristotle speaks (Cat. 14a30–1) about is “what does not reciprocate in respect of what follows in existence (τὸ μὴ ἀντιστρέφον κατὰ τὴν τοῦ εἶναι ἀκολούθησιν = mā lā yarǧiʿu bi-t-takāfuʾi fī luzūmi l-wuǧūdi).”14 In his commentary on this passage, Simplicius uses the later term “imply (sunepipherein)” to gloss Aristotle’s term “following (akolouthēsis).” It is not clear to me that Simplicius uses these terms with distinct senses. Simplicius explains the phrase as follows:

Δεύτερον δὲ τρόπον τοῦ προτέρου φησὶν τὸ μὴ ἀντιστρέφον κατὰ τὴν τοῦ εἶναι ἀκολούθησιν, ὅταν ἄλλῳ μὲν τεθέντι αὐτὸ ἕπηται, αὐτῷ δὲ τεθέντι τὸ ἄλλο μὴ ἕπηται, ὡς δυοεῖν15 μὲν ὄντων ἓν πάντως ἔστιν, ἑνὸς δὲ ὄντος οὐκ ἀνάγκη δύο εἶναι. ἀφ’ οὗ οὖν οὐχ ἕπεται ἡ ἀκολούθησις, πρότερον ἐκεῖνο, καὶ ᾧ τεθέντι ἕπεται τὸ ἕτερον, ὕστερον ἐκεῖνο. καλεῖν δὲ εἰώθασιν οἱ νεώτεροι τὸ τοιοῦτον πρότερον “συνεπιφερόμενον μὲν μὴ συνεπιφέρον δὲ καὶ συναναιροῦν μὲν μὴ συναναιρούμενον δέ.”

[…]

οὕτως δὲ καὶ τὸ γένος πρότερον δείκνυται τῇ φύσει τοῦ εἴδους, ὡς συναναιροῦν μὲν μὴ συναναιρούμενον δὲ καὶ συνεπιφερόμενον μὲν μὴ συνεπιφέρον δέ· ζῴου μὲν γὰρ ἀναιρεθέντος συναναιρεῖται ἄνθρωπος, ἀνθρώπου δὲ ἀναιρεθέντος οὐ συναναιρεῖται ζῷον, καὶ κατὰ τὴν τοῦ εἶναι δὲ ἀκολουθίαν οὔτε συνυπάρχει οὔτε συνεπιφέρεται τῷ ζῴῳ ὁ ἄνθρωπος ὥσπερ τῷ ἀνθρώπῳ τὸ ζῷον.16

[L2] The second sense of being prior speaks about the what does not reciprocate with respect to following in being, when something follows [call it Y] from something else being posited [call it X], but positing X does not follow from Y. Thus, there being two, there is, in all events, one. There being one, however, it is not necessary that two is. Therefore, the prior is that from which the consequent does not follow [so, one is prior because two does not follow from it]; the posterior is that from the positing of which the other follows [so, two is posterior because one follows from positing that it exists]. The recent [philosophers,17 hoi neōteroi] tend to call what is prior in this way “what is implied but does not imply, and removes but is not removed.”

In this way it is shown that genus is, by nature, prior to species because it [genus] is what removes [species] but is not removed [by species], and it is what is implied (sunepipheromenon) [by species] but does not imply (sunepipheron) [species]. For animal being removed removes human, but human being removed does not remove animal, and regarding following in existence, human does not coexist with or coimply animal, nor does animal coexist with or coimply human.

In Alfarabi’s discussion of Galen’s view, it seems that the Greek sunanhaireitai corresponds to Alfarabi’s irtifāʿ or “removal,” and the Greek sunepipheretai corresponds to Alfarabi’s luzūm al-wuǧūd. Alfarabi’s objection to Galen, then, is based on the logical concepts of removal and following in existence. Though these terms appear in Aristotle’s Categories, the somewhat formulaic manner in which we find them being used by Alfarabi and Simplicius in the form of conditional sentences likely represents a later stage of development, a result of a synthesis carried out by Aristotle’s commentators and adopted by Alfarabi in his critique of Galen. For Alfarabi the logical form of Galen’s premise (A) is therefore:

(I) if X, Y;

and the logical form of (B) is:

(II) if ¬X, ¬Y.

According to Alfarabi, it was common to use the fact that (I) and (II) are true to conclude that X causes (sabab li-) Y. And Galen is one of the thinkers who uses this form of argument.18

‮على مثال ما يرى جالينوس الطبيب يستعمل ذلك كثيرا في ما يشاهده في أعضاء الإنسان بالتشريح فيجعله أسبابا لأشياء أخر لم يشاهدها بأن يستعمل هذا الموضع مثل قوله إذا قطعنا العصب الفلاني، بطل الصوت أو الحركة أو الحسّ. فإذا وجود ذلك العصب هو سبب لوجود الصوت أو الحركة أو الحس.‬‎

[L3] For instance, Galen appears to have used [this kind of argument] frequently in what he observes about the human body when he is dissecting. He makes something a cause for other things he has not observed by employing this topos. For example, he says “when we sever a certain nerve, voice, motion and sensation are impaired. Therefore, the nerves are the cause of speech, motion and sensation.”

Later in the passage, Alfarabi expains why he believes that Galen’s claim that (A) and (B) together demonstrate that voice, sensation and voluntary motion originate in the brain is false.19

‮فنقول نحن الآن أمّا أنّ السبب الذي هو بالفعل ودائمًا سبب لشيء ما يلحقه ضرورة أن يكون إذا ارتفع، ارتفع الشيء وإذا وجد، وجد الشيء فذلك بيّن. وأمّا أن يكون كلّ ما إذا ارتفع، رفع الشيء وإذا وجد، وجد الشيء سببًا لذلك الشيء فليس يصحّ من قبل أنّه ليس يجب عن هذا شيء أكثر من أنّهما يتكافآن في لزوم الوجود. وذلك يتبيّن بمقابل ارتفاع الشيء وهو وجوده لزم من ذلك وجود الأمر. وقد كنّا وضعنا أنّ الأمر إذا وجد، وجد الشيء فيكون الأمر والشيء أي واحد منهما وجد وجد الآخر، فيكونان متكافئين في لزوم الوجود وليس يلزم ضرورة أن يكون أحدهما سببًا لوجود الآخر.‬‎

[L4] What we say now is that it is a necessary feature of the cause that it is actually and perpetually causing E [namely, its effect; literally, “the thing (aš-šayʾ)”], and that when it is removed, E is removed, and if it exists, E exists. This is obvious. However, for every X such that when X is removed Y is removed and when X exists Y exists, it is not the case that X is the cause of Y. For the only fact that is necessary about them is that they follow each other in existence reciprocally (yatakāfaʾāni fī luzūmi l-wuǧūdi). This is made evident when the removal of X is in the antecedent and the removal of Y is in the consequent. If we assert the opposite of the removal of Y, this yields the opposite of the removal of X, which is that X exists, whereas we had already hypothesized that if X exists, Y exists. So whichever of X or Y exists, the other exists too. So they follow each other in existence reciprocally. But it is not necessarily the case that one is the cause of the other.

According to Alfarabi, Galen’s argument in On the Doctrines of Hippocrates and Plato, Books One and Two only allows us to identify phenomena that always coexist in time. However, mere perpetual coexistence does not warrant the conclusion that one of the phenomena is responsible for or causes the other. Once again, Alfarabi is drawing in part on Aristotle’s Categories and in part on some commentary source. In his discussion of the category “simultaneous (hama),” Aristotle identifies things that are “simultaenous in nature,” which are “what reciprocate in relation to following in existence but never does one cause the other to exist (ὅσα ἀντιστρέφει μὲν κατὰ τὴν τοῦ εἶναι ἀκολούθησιν, μηδαμῶς δὲ αἴτιον θάτερον θατέρῳ τοῦ εἶναι ἐστιν).” Aristotle cites double and half as examples. The existence of each one entails the existence of the other, but neither causes the other. And Simplicius adds: “[For] the double and half always imply (suneisagei) and cancel (sunhanairei) one another, and neither is the cause of the other (γὰρ συνεισάγει καὶ συναναιρεῖ ἀεὶ ἄλληλα τὸ διπλάσιον καὶ τὸ ἥμισυ, καὶ οὐδέτερον θατέρου αἴτιόν ἐστι).”20

For Alfarabi, then, Galen’s argument is not a demonstration because it is not impossible that (I) and (II) are true and the conclusion (C) “X causes Y” is false: it is possible that that Y causes X, or that neither X nor Y causes the other.

Thus, aside from his opinions about Galen’s medical method, it is important to realise that Alfarabi believes that Galen has a generic method for inferring that X causes Y based on connections that are observed to always exist between X and Y, where X and Y may be anatomical parts or activities in the body, drugs and foodstuffs, or elements of medical treatment. Based on Aristotle and Platonist commentators, Alfarabi’s objection is that the observations that Galen makes during a vivisection, for example, only yield knowledge about the concurrence of events. We might observe that phenomenon B always follows on phenomenon A, or that B is absent whenever A is. But when we are trying to come to a conclusion about whether A causes B, or B causes A, or neither, the value of this data about mere coincidence in existence is small.

The problem in Galen’s anatomy, it seems, is the same as in the case of qualified experience in Galen’s pharmacology. In relation to the question of whether X causes Y or Y causes X, the value of the observations of connections that perpetually and reliably exist between X and Y are lessened by the presence of ambiguating conditions in the observation. The necessary conditions for sensation (S) to exist in the body are a sound brain (B), nerves (N) connecting the brain to the rest of the body, and other factors f1…, f n . All together N, B, and f1,…, f n constitute the necessary and sufficient conditions for S. This observation may be given the form of the conditional:

(III) if N & B & f1,…, f n , S.

(III) asserts that under all conditions if the nerves, brain and other background factors are present in the body, sensation will exist in the animal. In addition, since N, B and f1,…, fn are necessary conditions, the elimination of any one of them will eliminate S. For example, say that the nerves are damaged or severed. Sensation, then will be eliminated. This fact may also be expressed in a conditional:

(IV) if (¬N) & B & f1,…, f n ,(¬S).

(IV) means that under all conditions if the nerves are severed or damaged but the brain and other factors are left intact, sensation will be eliminated. Alfarabi asks us to consider whether it is really safe to conclude from (III) and (IV) that B is responsible for the exitence of S. In other words, does the fact that S and B coincide as two phenomena that follow each other in existence reciprocally authorize the assignment of responsibility to the brain, the nerves or to some other factors? What is more, Galen cites (IV) if N) & B & f1,…,fn, (¬S) as proof that B is responsible for S, whereas it is obvious that N and f1,…,fn are equally responsible for S. In addition, it is only through the combination of N, B and f1,…,fn that the connection with S holds. So in Alfarabi’s way of thinking (IV) does not establish that B is responsible for or the origin of S, but only that N, B and f1,…,fn together are responsible for S, without implying that B is any more responsible for S than N or f1,…,fn. Galen’s argument according to Alfarabi only identifies certain conditions that always coexist. However, the observation that certain phenomena always coexist is not enough to identify which of the phenomema is the origin of or bears ultimate responsibility for these activities in the body.

4. “Qualified Experience” in the Arabic Commentators on the Hippocratic Aphorisms

Alfarabi criticizes a method Galen employs in his anatomical works to discover where ultimate responsibility for activities in the body lies. Alfarabi says that Galen’s method does not yield conclusions that follow necessarily from the premises derived from experience, which in On the Doctrines of Hippocrates and Plato, Books One and Two is derived from dissection. The basis of Alfarabi’s criticism lies in the fact that certain ambiguating conditions in which the observations are made will inevitably undermine the evidentiary value of these observations in relation to the conclusion that is being sought. Despite the fact that Alfarabi directs his criticism at Galen’s anatomy, it is noteworthy that Alfarabi sees this criticism as valid in any situation in which we seek to identify that X causes Y based on repeated observations about predictable connections that exist (or do not exist) between X and Y.

Galen seems to have formulated the method of qualified experience in order to address this very problem. However, from his extant writings, it is not clear how Galen thought qualified experience should be used to ensure that the observations of particular instances do, in fact, prove the sought after conclusion about the power of a drug. Nor is it clear how Galen’s method of qualified experience might be taken outside pharmacology and introduced into other fields of medicine in which making statements about causes is based on empirical evidence. For hints about how to do both of these, we may turn to some of the Arabic commentators on the Hippocratic Aphorisms.

In their comments on the first aphorism, it is clear that post-classical Arabic commentators were aware of the different senses of “experience” that Galen used in his writings. In addition to qualified experience, Philip van der Eijk has also identified a generic or unqualified type of experience, which in Galen is much more common.21 Unlike qualified experience, which seeks to increase the value of observations by modifying experimental conditions, this more common class of experience is used to test or determine (dihorizein or dihorisasthai, but also dokimazein = imtaḥana) the validity of a “statement, or claim, an issue, idea or notion.”22 For example, the statement that all olive oil irritates the eyes must be tested or validated by means of experience. In their comments on the statement that “determination is difficult and experience dangerous,” Arabic commentators on the Aphorisms invariably mention this type of experience, which they say is used as an instrument for testing universal statements. Thus, Ibn Abī Ṣādiq (d. after 1067), one of the earliest and most influential commentators on the Hippocratic Aphorisms, says:23

‮وأمّا الخطر في التجربة، فإنّ التجربة على وجهين. أحدهما اعتبار ما علم من القوانين الكلّيّة في المشاهدات الجزئيّة. وهذا يحتاج إليه كل واحد، وليس فيه خطر. والآخر امتحان الشيء من غير قياس يؤدّي إليه ولا صادر عن أصل وقانون. ومن البيّن أنّ هذا النوع من التجربة غير موثوق به، ولذلك هو خطر، كما يدلّ عليه الاستقراء. فإنّ من لا يعرف قوانين الكلام فهو إذا تعاطى علم التوحيد، فهو إلى أن يضلّ، فيضلّ أقرب منه إلى أن يصيب ويرشد. ومن جهل قوانين النحو واستعمل الإعراب في كلامه، كان الذي يخطئ فيه أكثر ممّا يصيب. ومن غني عن القوانين الطبّيّة ثمّ شرع في تدبير المرضى، كان ما يفسده أكثر ممّا يصلحه. وهذا النوع هو الذي عناه لا غير.‬‎

[L5] In relation to the danger that is in experience, “experience” has two senses. The first is learning through experience what is known from universal principles in particular observations. Everyone needs this experience, and there is no danger in it. The other [sense of experience] is testing something [for example, a compound drug concoction] without a process of reasoning (qiyās) that leads to it, nor does it [the compound drug concoction] proceed from a foundation or rule. It is clear that this type of experience is unreliable, for which reason it is dangerous, as induction (al-istiqrāʾ) indicates. For whoever does not know the rules of Islamic scholastic theology (qawānīna l-kalām) yet busies himself with the science of oneness [namely, theology] is more likely to go astray than to hit the mark and be guided. Whoever is ignorant of the principles of grammar but uses case-endings in his speech is more likely to make errors than to speak correctly. Whoever dispenses with medical rules and then begins to treat patients does more harm than good. This is the kind [of experience] that [Hippocrates] is speaking about.

Ibn Abī Ṣādiq’s comments are indebted to Galen. He affirms that the Galenic synthesis of experience (al-taǧriba) and reason (al-qiyās) is reliable and safe for use on patients, and even that it is the kind of experience that Hippocrates himself would have favoured. Ibn Abī Ṣādiq also singles out how Galen’s empiricist opponents used experience in a way that Hippocrates regards as a “danger (ḫaṭar).” In the first kind of experience, there is knowledge about a universal statement, which is verified by experience derived from observation of particular instances. Thus, knowledge that the universal “all olive oil irritates the eyes” is verified by repeatedly observing that olive oil irritates people’s eyes. The second type of experience discusssed by Ibn Abī Ṣādiq is not connected with the task of verifying universal statements or principles. Experience in this sense is testing out a drug on a patient in order to bring about some effect. In this case, the (empiricist) physician acts on previous experience alone without recourse to the principles of medical theory. Since the empiricist physician does not make use of reason when he treats patients, his knowledge of a limited number of cases does not adequately equip him to deal with cases the particulars of which he has never come across before. According to Ibn Abī Ṣādiq, ignorance of the principles of medicine is what makes the empiricist physician’s treatment dangerous, since he tests a drug without knowing what the effects of it on the patient will be.

It is possible to read Ibn Abī Ṣādiq’s and Galen’s comments on the meaning of “experience” in their commentaries on the Hippocratic Aphorisms as being directed exclusively against empiricist physicians. On this interpretation, experience plays the negative role of falsifying hypotheses about the properties of drugs, foodstuffs, and therapies more generally. Galen’s comments in L1 about the difficulty of experience in relation to gaining knowledge about the causes and effects of drugs on the body, however, point to another sense of experience, one that is closely tied to Galen’s idea of qualified experience. In this case, when experience is disconnected from the task of falsifying universal statements, it may play a positive or productive role in the inductive process of generating universal principles. Ibn Abī Ṣādiq appears not to assign experience this role. Yet, several important Arabic commentators do. For example, ʿAbd al-Laṭīf al-Baġdādī (d. 1231), who was a prodigious writer in both medicine and philosophy and a strident critic of Avicenna,24 says:25

‮وقوله ”والتجربة خطر،“ أي غرر وإقدام بالفعل على غير بصيرة ولا ثقة. وإنّما يستعمل هذه اللفظة، أعني الخطر، في شيء شريف خطير يقدم على مزاولته من غير ثقة بسلامته. وبدن الإنسان ممّا هو شريف. فالإقدام على معالجته بالتجربة من غير ثقة بسلامة العاقبة خطر. والطبيب يعالج المريض ليبرئ لا ليجرب. فإنّ التجربة تكرير الإحساس في شيء لتثق النفس أنّ فعله وانفعاله منسوب إليه لذاته، لا واقع على طريق الاتّفاق. والمجرّب متعلم، فإذا كنّا أبداً مجرّبين، فيا ليت شعري، متى نكون مداوين.‬‎

[L6] “There is danger in experience,” that is to say, uncertainty (ġarar) and rushing into an action without discernment or confidence. He [Hippocrates] employs this word, that is “danger,” for something that is noble and important which one rushes to do without having any confidence that it will be safe. The human body is something noble. Therefore, rushing to treat it based [only] on experience without confidence that the outcome will be safe is dangerous. The physician treats the patient in order to cure him, not to experiment. Experience is making repeated observations of X [lit. “the thing (aš-šayʾ)”] so that the soul is confident that the activity and passivity (infiʿāl) that is associated with X is essential to X, and is not associated with X by chance. The person who is trying [therapies] out is learning, so if we are always trying [therapies] out, then I would like to know when we shall use them.

Al-Baġdādī also associates experience with uncertainty and doubt about the effects of a particular therapy on the patient. Recalling Galen’s use of qualified experience in order to distinguish a drug’s essential from its accidental effects, al-Baġdādī says that experience is the act of making repeated observations of something (tikrāru l-iḥsāsi fī šayʾin) so that the person acquires reliable knowledge that the characteristics or properties that we observe in the drug belong to the drug essentially (li-ḏātihī) rather than by chance ( ʿan ṭarīqi l-ittifāqi).

The most explicit reference to the Galenic notion of experience as the conscious manipulation of the observational conditions under which a drug or foodstuff is administered to the patient appears in a commentary attributed to Ibn al-Nafīs (d. 1288).

‮والتجربة هي امتحان ما يؤثّره الشيء في البدن بإيراده عليه. وذلك على قسمين. أحدها امتحان ما اقتضاه القياس كما إذا دلّ القياس على برودة الكافور، فأردنا امتحان ذلك. وثانيهما امتحان الشيء من غير قياس. ولا شكّ أنّ الخطر في القسم الثاني أعظم.‬‎

‮وأمّا أنّ ”القضاء عسر،“ والقضاء هو الحكم، فقيل: ”أراد الحكم على المريض بما يؤول إليه حاله من صحّة أو عطب.“ ولا شكّ في عسر ذلك. وقيل: ”أراد الحكم بموجب التجربة.“ وذلك عسر أيضًا لأنّ التجربة إنّما يوثق بها إذا كانت على بدن إنسان وكان الوارد خالياً عن كلّ كيفيّة غريبة، واستعمل في علل متضادّة وبسيطة بما قوّته مقاربة لقوّة العلّة. وكان فعله أولاً ودائماً أو أكثريًّا. ولا شكّ إنّما توقف على ذلك، فهو عسر، وخصوصًا إذا كنّا اتسعملنا صنوفاً من المعالجة كالفصد والاستفراغ وسقي أدوية، فحصل النفع فإنّا لا ندري عن أيّها حصل.‭26‬‬‎

[L7] Experience is testing what effect something has on the body by applying it to [the body]. This is of two types. One is [A] testing what syllogism entails, such as when syllogism indicates that camphor is cold, and we want to test that. The second is testing something without syllogism. Obviously, the danger is greater in the second type.

As for “the determination is difficult”—“determination” means making a judgment—some say that [Hippocrates] meant making a judgment about whether the patient’s condition will end in health or destruction. The difficulty of this is obvious. Others say that he meant making a judgment [B] in accordance with what experience dictates. This is difficult, too, because experience is reliable when it is about the human body only, the drug (al-wārid) is free of any extrinstic quality, and when it [the drug] is used in simple diseases that are opposite to it [the drug in quality] such that [the drug’s] strength is close to the disease’s strength, and when [the drug’s] action is primary and persistent, or for the most part. Doubtless, anything that requires such conditions is difficult, especially when we use different kinds of therapy [at once], for example blood-letting, purging, and drugs. If, after applying [this combination of therapies] a benefit results, we do not know from which treatment the benefit was derived.

In this passage, Ibn al-Nafīs lists several different senses that “experience (taǧriba)” has, the most relevant of which here are senses A and B. Sense A corresponds to the unqualified or simple experience that Galen prescribes for falsifying universal statements that have been inferred from first principles by a process of syllogistic logic. Sense B on the other hand appears in a passage that seems to expand on Galen’s comments in L1. Similar to Galen, Ibn al-Nafīs points out that the difficulty in B-type experience involves the reliability (ṯiqa) of the experience. Like Galen too, he offers an example of compound therapies that makes it impossible to tell which therapy was principally responsible for the benefit to the patient. He lists several of what we can call “ambiguating” conditions the presence or absence of which affects the evidentiary value of the observations. He mentions four conditions, all of which focus on the drug’s properties, and many of which are based on Avicenna’s comments on observing the powers of drugs from Canon 2 (see the Appendix). He says that in order for observations to give insight into the real pharmacological effects, the drug should be “free of any impure quality (ḫāliyan ʿan kulli kayfīyatin ġarībatin).”27 This suggests that the drug should be a simple drug and as pure as possible without compounding it with another drug or foodstuff. He recommends that the nature and strength of the drug accords with the nature and strength of the disease that it is being used to treat.28 He also recommends that the pharmacological activity of the drug whose effects are being observed be associated with it in a primary way (awwalan). The idea here seems to be that observations of a drug’s effects that are secondary or that cannot be directly attributed to the power of the drug are of little use in discovering the drug’s essential therapeutic effects.29 Lastly, he recommends making observations with drugs who effects always come about (ʾiman), that is under all conditions, or at least under most conditions (akṯarīyan30).31 Ibn al-Nafīs seems to have in mind that observations of a drug’s effects that rely on times of the year, or certain conditions in the body, or certain age-groups do not have evidentiary value if these conditions are not taken into account (for example, the true effects of a hot drug administered in cold weather or even in a cold climate will affect the effect of the drug on the patient, and thus the observations’ value to the physician).

The second important observation has to do with how in L7 I translate the term “qiyās” with the technical, Aristotelian term “syllogism” rather than the more generic term “reason” or “reasoning process.” In the early Arabic medical tradition, qiyās was often used to render the Greek logos and its cognates.32 In the Arabic philosophical tradition, qiyās is used to render the Greek sullogismos and its cognates.33 Early Arabic medical authors and philosophers were careful to distinguish between these distinct uses of qiyās,34 one to pick out a generic mental activity of reasoning, the other the specific sense of qiyās as an Aristotelian syllogism. By the end of the thirteenth century at the latest, it appears that in medical discourse qiyās frequently came to refer to the syllogism rather than reasoning in an unqualified sense. The evidence for this is in L10 by Ibn al-Quff (d. 1286) below. In addition, in the same commentary on the Aphorisms preserved in Oxford, Bodleian, MS Pococke 294 from which L7 is taken, Ibn al-Nafīs (d. 1288) follows Galen-Ḥunayn in using qiyās as a gloss for al-qaḍāʾ. However, in the following passage, Ibn al-Nafīs speaks about qiyās in the context of the “science of logic (ʿilm al-manṭiq)” as logic was famously referred to after Avicenna in the Arabic tradition:35

‮وقيل‭:‬ أراد بالقضاء القياس لأنَّ القياس يلزمه الحكم بموجبه فأطلق اسم اللازم على الملزوم. ولا خفاء بعسر معرفة صحيح القياس من فاسده. ولو سهل التمييز بينهما لما خالف العلماء ولا ناقض أحدهم نفسه. ونقول‭:‬العلم الذي يميز بين صحيح القياس وفاسده هو المنطق، فلابد للطبيب القائس من معرفة المنطق.‬‎

[L8] It has been said: “By “determination (al-qaḍāʾ)” he [Hippocrates] meant the syllogism (al-qiyās).” For the syllogism entails the judgment by necessitatating it. Hence, the name of what is entailed (al-lāzim, the judgment that is inferred; the conclusion of the syllogism) is given to what entails (al-malzūm) it. There is nothing concealing the fact that knowing what syllogism is sound and unsound is difficult. For if distinguishing between them were easy, scholars would never disagree with each other, nor would one of them contradict himself. We [Ibn al-Nafīs] say: “The science that distinguishes the sound syllogism from the unsound syllogism is logic. Thus, the physician who uses reason must know logic (fa-lā budda li-ṭ-ṭabībi l-qāʾisi min maʿrifati l-manṭiq).”

Thus, where Ibn Abī Ṣādiq, Avicenna and other physicians before the 12th-century use qiyās to speak about a generic reasoning activity or process of inferring conclusions from principles, by the 13th-century qiyās is being regularly used in medical discourse with the meaning that it had had in Arabic philosophical discourse since 9th-century.

We may conclude, then, that Ibn al-Nafīs and ʿAbd al-Laṭīf al-Baġdādī focus on experience as a task or activity undertaken by the observer in order to establish a causal and essential connection between a drug or type of therapy and its therapeutic effect. According to these thinkers, this task involves (1) repeated observations in which (2) the conditions under which the observations takes place are carefully adjusted to endow them with (3) evidentiary value; meaning that, taken together, the observations are a demonstration or proof that the drug’s or therapy’s effect is essential to it rather than accidental. As a matter of fact, the underlying logic of this procedure is nearly identical to the anatomical method Alfarabi attributes to Galen for finding the parts of the body that are responsible for activities in the body. For example, under conditions c1,…,c3 we are testing whether scammony (call it S) (the presence of scammony is also considered a condition) purges yellow bile (call this P). The initial observations may be put into the form of a conditional proposition:

O1: if S and c1 and c2 and c3, P.

O1 means something like: when it is always the case that if scammony is given to a patient under these three conditions, yellow bile is purged. There is little evidentiary value in this first observation since it is not clear whether the therapeutic effect of purging is due to scammony alone, to one of the conditions, or to scammony in combination with one or more of the conditions. Thus, in the second observation O2 we make sure to eliminate the ambiguating condition c1. If P still comes about we know that c1 is irrelevant to generating P in the patient. The observation will thus have the form:

O2: if S & (¬c1) & c2 & c3, P.

O2 means something like: when it is always the case that if scammony is given to a patient under possibly ambiguating conditions c2 and c3, but condition c1 is eliminated, purging in the patient results. The observations will be repeated two more times, where in each an ambiguating condition is eliminated. Observation O4 looks like this:

O4: if S & (¬c1) & (¬c2) & (¬c3), P.

O4 means something like: when it is always the case that if scammony is given to a patient and all ambiguating conditions c1,…,c3 have been eliminated, purging in the patient results. O4 has strong evidentiary value. The final observation, O5, acts as a control for observations O1,…,O4:

O5: if (¬S) & (¬c1) & (¬c2) & (¬c3), (¬P).

O5 means something like: when it is always the case that if the patient is not administered any drug and all other conditions are in balance, the patient does not experience purging. For Galen as well as the Arabic commentators on the Aphorisms who speak about a generic empirical method for discovering causes, O1O5 count as a demonstration that responsibility for the purging effect experienced by the patient lies in the particular pharmacological properties of scammony alone, and is not a consequence of environmental conditions, time of the year, the temperament of the patient, his age, the strength of the concoction, impurities in the scammony, or any other ambiguating factors. O1O5 is a proof of the pharmacological principle: “Scammony causes purging of yellow bile.”

5. Senses of Experience in Later Aphorisms Commentators from Avicenna’s Philosophical Writings: Synthesising Galen and Aristotle

Alfarabi’s charge against Galen is that his anatomical method and his pharmacological method are not equipped to yield reliable knowledge about causes. He would argue, in effect, that O1O5 do not yield necessary knowledge that scammony is the reason for purging, but only knowledge that purging always follows taking scammony and that, all other conditions being in balance, purging does not occur if the patient does not take scammony. According to this line of reasoning, these are simply temporal conjunctions that do not give insight into which of P or S is responsible for the other.

I believe that Galen’s response to Alfarabi’s criticism would have been: “So what?” In his critique, Alfarabi assumes that Galen’s method will only yield reliable or necessary knowledge about causal relations or relations of responsibility between X and Y if the conclusion that (C) “X causes Y” follows with logical necessity from the premises that (i) “if X, Y” and (ii) “if (¬X), (¬Y).” I mean the following. He reduces the form of Galen’s argument in On the Doctrines of Hippocrates and Plato, Books One and Two into the argument that C follows necessarily from i and ii. Alfarabi assumes that Galen’s position can be refuted if it can be shown that the truth of C does not follow necessarily from the assumption that i and ii are true. He then produces a counter-example of Galen’s argument, in which the premises are true but the conclusion is false, for example, “it is day” could be substituted for X and “the sun is up” for Y so that that the premises of the argument are: (i’) “if it is day, the sun is up,” and (ii’) “if it is not day, the sun is not up.” i’ and ii’ are clearly true, but (C) “the fact that it is day is the cause of (or an explanation for) the fact that the sun is up” is clearly false. Alfarabi takes the fact that Galen’s argument is logically invalid (i.e. there are substitution-instances for X and Y such that the premises i and ii are true but C is false) as a genuine critique of Galen’s argument against Aristotle. According to Alfarabi, Galen’s empirical method does not yield necessary knowledge about the cause because the statement “X causes Y” does not follow with logical necessity from the premises i and ii. Therefore, on this account, Galen’s method should be dispensed with. To my mind, however, Galen never thought qualified experience should conform to the canons of deductive logic in the first place.

In the discussion of experience and reasoning as the two methods for discovering the pharmacological properties of simple drugs in Canon 2, Avicenna also seems sceptical about the value of experience (and reasoning) as a guide to gaining necessary and certain knowledge about the effects that drugs bring about in the body. His reasons, however, are not Alfarabi’s. Avicenna does not fault experience for its failure to conform to the conditions of deductive logic. Instead, he appears to believe that there are ineliminable limitations on the ability of qualified experience to eliminate every ambiguating factor and bring to light a drug’s essential pharmacological effects. We might recall that Avicenna says that in the process of determining a drug’s pharmacological effects, the physician must take care to observe the time (zamān) in which the drug’s effect appears to him (Condition 5 in Appendix).36

‮والخامس أن يراعى الزمان الذي يظهر فيه أثره وفعله. فإن كان مع أول استعماله أقنع أنه يفعل ذلك بالذات. وإن كان أول ما يظهر منه فعل مضاد لما يظهر أخيرا أو يكون في أول الأمر لا يظهر منه فعل ثم في آخر الأمر يظهر منه فعل فهو موضع اشتباه وإشكال عسى أن يكون قد فعل بالعرض كأنه فعل أولا فعلا خفيا تبعه بالعرض هذا الفعل الأخير الظاهر. وهذا الإشكال والاشتباه في قوة الدواء والحدس أن فعله إنما كان بالعرض لقد يقوى إذا كان الفعل إنما ظهر منه بعد مفارقة العضو. فإنه لو كان يفعل بذاته فعل وهو ملاق للعضو، ولاستحال أن يقصر وهو ملاق ويفعل وهو مفارق. وهذا هو حكم أكثري مقنع. وربما اتفق أن يكون بعض الأجسام يفعل فعله الذي بالذات بعد فعله الذي بالعرض. وذلك إذا كان اكتسب قوة غريبة تغلب الطبيعية مثل الماء الحار، فإنه في الحال يسخن وأما من اليوم الثاني أو الوقت الثاني الذي يزول فيه تاثيره العرضي فانه يحدث في البدن بردا لا محالة لاستحالة الأجزاء المستعقبة منه الى الحالة الطبيعية من البرد الذي فيه.‬‎

[L9] The fifth [condition, sc. šarṭ] is that one observe the time in which the drug’s effect and activity appear. If they appear directly after the drug is given, it is convincing that this [effect and actvity] occurs essentially (bi-ḏ-ḏāt). If what appears first is an activity that is opposite to what appears last, or no activity appears at first but then the activity appears last, this then will be a source of ambiguity and suspicion that it may be that the activity the drug has is accidental (bi-l-ʿaraḍ), as if the drug first had a subtle [essential] effect, which is then followed accidentally by this later, obvious effect. This suspicion and ambiguity about drug’s power and the surmise (al-ḥads37) that the drug’s activity is accidental is strong when the [drug’s] activity appears after departing from the body part (al-ʿuḍw). For if the drug were to act essentially, it would act whilst it was in contact the body part, and it would be impossible for it to remain ineffective whilst it was in contact with the body part but to act whilst it departs from the body part. This is a judgment that holds for the most part and is convincing.

However, it sometimes happens that some bodies act essentially after they act accidentally. This happens when they acquire a power that is extrinsic to them that overwhelms the drug’s natural power, as is the case of hot water. For in that state, [water] heats, but the next day or the next moment in which the accidental effect has dissipated, water ineluctably brings about a cold [quality] in the body owing to the fact that parts from which the water is constituted transform back to the cold that is in water in its natural state.

In this passage, Avicenna has two objectives. His chief purpose is to draw the physician’s attention to the fact that the order in which the drug’s various effects appear in the patient’s body may be a reliable indicator of the drug’s essential and accidental activity. To this end, Avicenna sets out the principle that the effect that first appears in the patient’s body is associated with the drug essentially. In L9, however, Avicenna hastens to add that this “principle” is really no more than a rule of thumb. He points to situations in which different kinds of drug and foodstuff naturally behave in a way that obscures the drug’s essential and accidental properties by violating this rule of thumb. For example, hot water, which has the direct effect of heating the patient’s body, produces its heating effect accidentally. After several minutes, the heat dissipates and the cold that is in water naturally returns. Avicenna seems skeptical about this method’s real utility, especially in the face of drugs that naturally produce their essential effect in a delayed fashion, even after the drug has left contact with the affected part of the body. In such cases, Avicenna does not believe that a series of observations of the drug’s effect on the patient’s body will allow the physician to distinguish the drug’s essential and accidental effects, even if we carefully control the conditions underwhich the observations are made.

Avicenna believes that experience, even if carefully monitored in this way, still does not yield reliable, certain and necessary knowledge about causal relations between drugs and their effects. This belief appears, however, to be based on the practical conditions holding for medical treatment. Outside the idealised world of medical textbooks, conditions are never ideal, and thus experience, even suitably qualified, will not necessarily yield reliable results about causal relations between drugs and their effects owing to the nature of particular drugs and foodstuffs.

Yet, as we shall see, Avicenna’s medical views in Canon 2 do not reflect the complexity of his philosophical doctrines regarding the role of experience in the process of knowledge acquisition. An important discussion by Abū l-Faraǧ Ibn al-Quff al-Karakī (d. 1286) draws our attention to the fact that by the end of the thirteenth century at the latest, Avicenna’s medical and philosophical doctrines were treated as relevant to understanding Hippocratic and Galenic medical texts. It also highlights the manner in which Avicenna synthesised Galenic medical thought and Aristotelian philosophy.38

In his commentary on the first aphorism, Ibn al-Quff takes the opportunity to speak about the role experience plays in acquiring first principles of medical science. Ibn al-Quff’s discussion resembles a similar discussion by Aristotle at the end of Posterior Analytics, Book Two. In APo. 2.19 Aristotle discusses the role of experience in the process of acquiring first principles in demonstrative science. In his commentary, however, Ibn al-Quff is not referring to Aristotle, but is closely paraphrasing Avicenna’s discussion of aporias involving experience, demonstration, and cause in Dem. 1.9.39 In this passage from his Aphorisms commentary, Ibn al-Quff discusses what experience means. His choices are pragmatic, however. He copies the definition that Avicenna supplies for experience in Dem. 1.9, uses the same example to illustrate his point, and comes to the same conclusions about how experience has the capacity to yield reliable information about the natures and causes of observed phenomena. He omits, however, the more fundamental philosophical dilemma, which had originally inspired Aristotle and Avicenna, about the incongruity between the fact that experience yields first principles of demonstrations about causes, and the fact that these principles are not themselves demonstrated. In fact they are, like definitions, indemonstrable.40

‮البحث الثاني عشر في معنى التجربة. التجربة هي القضيّة التي يصدق العقل بها بواسطة الحسّ وشركة من القياس. فإنّ الحسّ إذا تكرّر عليه اقتران شيء بشيء مراراً غير محصورة، حصل في الذهن من هذا التكرار شيء، وهو أنّ هذا القدر ليس باتّفاقيّ. فإنّ الاتّفاقيّ ليس بدائم ولا أكثريّ، بل هو أمر طبيعيّ. مثاله السقمونيا، فإنّها تسهل الصفراء. وهذا القدر، لمّا رأيناه قد تكرّر مع شرب السقمونيا، علمنا قطعاً أنّ التكرار هو الموجب لذلك الاعتقاد. إذ لا بدّ لكلّ حادث من سبب، وليس ذلك السبب أمر إراديّ من السقمونيا، فإنّ النبات عادم للإرادة. بل ذلك لأمر طبيعيّ في السقمونيا. وليس هو الجسميّة، وإلا تساوت الأجسام في ذلك. بل هو لقوّة حاصلة فيه. فحينئذ يقضي العقل بأنّ السقمونيا التي في بلادنا تسهل الصفراء إذا كان المنفعل مستعدّاً للإسهال.‬‎

[L10] The twelfth investigation is about the meaning of “experience.” Experience is the proposition to which the intellect gives assent by means of the senses in cooperation with syllogism (al-qiyās). For if the senses repeatedly perceive the conjunction (iqtirān) of X [lit. “the thing (aš-šayʾ)”] and Y [lit. “the thing (aš-šayʾ)”] innumerable times, something is generated in the mind from this recurrent perception, namely that this power [in X, qadr] is not merely by chance (bi-ttifāq), since what is by chance does not happen perpetually or for the most part. Rather, [what occurs to the mind] is that this is something natural [to X].

Take scammony, for example, which purges yellow bile. When we see this power (qadr) [i.e to purge yellow bile] associated with drinking scammony, we know with certainty that the repetition is what necessitates this belief. Since every event must have a cause, in scammony this cause is not something volitional since plants do not have volition. Rather, the cause is something natural in scammony. Nor is it [the power, sc. qadr] on account of the fact that scammony is a body. If it were, all bodies would be alike [in their ability to purge yellow bile]. Rather, [the power] is owing to a capacity that resides in scammony. Then the intellect judges that the scammony in our lands (fī bilādinā) purges yellow bile on condition that the patient is fit for purging.

In L10, Ibn al-Quff takes “experience” to refer to a proposition (qaḍīya) that is a result of the cognitive process of inducting (pharmacological or therapeutic) principles such as “scammony purges yellow bile” from the fact that we repeatedly perceive (ḥiss) the temporal conjunction (iqtirān) of a pair of phenomena. The first step in this process is that the reasoner repeatedly observes that two phenomena always occur together, for example the patient taking scammony and the purging of yellow bile that follows. The second step is that after a sufficient number of observations, a syllogistic process of reasoning (al-qiyās) allows the intellect to move from this collection of perceptions to form the universal judgment that all scammony causes purging of yellow bile. Ibn al-Quff says that we know that this experience gives reliable insight into the true nature of scammony because the perceptions are repeated. Our conviction (iʿtiqād) that scammony causes purging is not necessitated because it has been deduced from first principles, but from repeatedly observing particular instances of scammony purging yellow bile. The repetition of perceptions necessitates certainty in the judgment that scammony causes purging, and that the purging effect is an essential consequence of scammony’s nature, meaning that the purging is not a result of accidental conditions under which the observations were made, or secondary or non-essential pharmacological properties of scammony or extraneous conditions affecting the patient to whom the drug is given.

Ibn al-Quff is paraphrasing Avicenna’s discussion of problems relating to induction’s role in acquiring first principles in Dem.1.9. Yet there are important elements in this text that Ibn al-Quff omits. Though he mentions the syllogism at the beginning of L10, the importance of syllogism’s role in generating universal propositions of the form “(all) scammony purges yellow bile” from experience is not emphasised in Ibn al-Quff’s paraphrase. For his part, Avicenna synthesises Aristotle’s somewhat incompatible accounts of induction (epagōgē) in the Prior and Posterior Analytics. In Deliverance (Kitāb al-Naǧāt), Avicenna briefly summarises Aristotelian induction as it is presented in APr. 2.23. In this chapter, Aristotle holds that induction has an implicit syllogistic structure. Avicenna’s summary account in the Delieverance rightly emphasises this fact as well.

‮الاستقراء هو حكم على كلي لوجود ذلك الحكم في جزئيات ذلك الكلي إما كلها وهو الاستقراء التام وإما أكثرها وهو الاستقراء المشهور فكأنه يحكم بالأكبر على الواسطة لوجود الأكبر في الأصغر. ومثاله أن كل حيوان طويل العمر فهو قليل المرارة لأن كل حيوان طويل العمر فهو مثل إنسان أو فرس أو ثور والإنسان والفرس والثور قليل المرارة ومن عادتهم أن لا يذكروه على هذا النظم بل يقتصرون على ما هو كالصغرى أو ما هو كالكبرى.‭41‬‬‎

[L11] Induction is making a judgment (ḥukm) about a universal (kullī) owing to the fact that this judgment exists in the particulars in the extension of this universal. Either all of them [fall in the extension of the universal], in which case it is complete induction, or most of them [fall in the extension of the universal], in which case it is common induction (al-istiqrāʾ al-mašhūr). It is as if the major term is judged to belong to the middle term because the major term is in the minor term. An example of it is that [Conclusion] “every animal that is long-lived has a small amount of bile” because [major premise] “every animal that is long-lived is like humans, horses and bulls,” and [minor premise] “humans, horses and bulls have a small amount of bile.” Normally, however, they do no mention [the syllogism] in this order. They restrict themselves [to mentioning only] what is like the minor or what is like the major.

Similar to Aristotle in APr. 2.23,42 in L11 what Avicenna calls the “judgment (ḥukm)” is the major term “long-lived,” and what Avicenna calls the “universal (kullī)” is the minor term “having a small amount of bile.” The humans, horses and bulls are the particulars (ǧuzʾīyāt) that, prior to the induction, are recognised to fall in the extension of the universal “having a small amount of bile.” Avicenna says that we inductively form the conclusion of the universal affirmative proposition “everything having a small amount of bile is long-lived” because we recognise that humans, horses, bulls and any other bileless animal have long lives. Given that Avicenna recognises the fact that Aristotle gives induction the propositional structure of a syllogism in APr. 2.23, it is no surprise that in Dem. 1.9 Avicenna speaks about experience as a judgment.

‮والتجربة مثل حكمنا أن السقمونيا مسهل للصفراء فإنه لما تكرر هذا مرارا كثيرة زال عن أن يكون مما يقع بالاتفاق فحكم الذهن أن من شأن السقمونيا إسهال الصفراء وأذعن له وإسهال الصفراء عرض لازم للسقمونيا.‬‎

‮ولسائل أن يسأل فيقول: ”هذا مما لم يعرف سببه. فكيف يقع اليقين الذي عندنا من أن السقمونيا لا يمكن أن يكون صحيح الطبع فلا لكون مسهل للصفراء؟“ أقول‭:‬ إنه لما تحقق أن السقمونيا يعرض له إسهال الصفراء وتبين ذلك على سبيل التكرار الكثير، علم أن ليس ذلك اتفاقا. فإن الاتفاقي لا يكون دائما أو أكثريا. فعلم أن ذلك شيء يوجبه السقمونيا طبعا إذ لا يصح أن يكون عنه اختيارا إذ علم أن الجسم بما هو جسم لا يوجب هذا المعنى. فيوجبه بقوة قريبة فيه أو خاصة له أو نسبة مقرونة به. فصح بهذا النوع من البيان أن في السقمونيا بالطبع أو معه علة مسهلة للصفراء. والقوة المسهلة للصفراء إذا كانت صحيحة وكان المنفعل مستعدا حصل الفعل والانفعال. فصح أن السقمونيا في بلادنا تسهل دائما الصفراء إذا كانت صحيحة. فإذن عرفنا الأعظم للأصغر بواسطة الأوسط الذي هو القوة المسهلة وهو السبب. وإذا حللت باقي القياس وجدت كل بيان إنما هو بيان بواسطة هو علة لوجود الأكبر في الأوسط، وإن لم يكن علة للعلم بالأكبر. فإذن بالسبب حصل لنا هذا النوع من اليقين أيضا.‬‎

[L12] Experience is like our judgment that scammony purges yellow bile, for when this repeats many times, it is no longer something that occurs by chance. So the mind judges that it is in the nature of scammony to purge yellow bile, conviction is placed in this [judgment], and purging yellow bile is a concomitant property (ʿaraḍ lāzim) that scammony possesses.43

Someone may ask, saying “This is the kind of jugment in which the cause is not known. So whence this certainty we have that it is not possible for scammony to have a sound nature and not purge yellow bile?”

I say: “When the fact that purging yellow bile follows scammony is realised (taḥaqqaqa), and it is made evident on numerous occassions, it is known that this [facility for purging yellow bile] is not by chance, for what is by chance is not always the case or for the most part. Thus, it is known that scammony is something that naturally purges yellow bile, for it is not fitting that it happen by way of volition, since it is known that body, insofar as it is body, does not entail this phenomenon (al-maʿ). Thus, it [purging] is entailed by a proximate capacity (qūwa qarība) in it, a property (ḫāṣṣa) that it possesses, or a relation that is connected to it.”

With this kind of account, therefore, it is true that scammony possesses a cause that purges yellow bile, either by its nature (bi-ṭ-ṭabʿ) or by virtue of some property that accompanies it. Furthermore, as long as the capacity for purging yellow bile is sound and the patient is fit for the drug, its [scammony’s] activity and passivity will come about. Therefore, it is true that the scammony in our lands (fī bilādinā) always purges yellow bile as long as it is sound. Thus, we recognise that the major term belongs to the minor term by means of the middle term, which is “the purgative capaciy (al-qūwa al-mushila),” which is the cause. Then, if you analyse the rest of the syllogism, you will find that every demonstration is by means of the thing that causes the major term to be in the minor term, even if it is not what causes knowledge of the major term.44

As McGinnis notes, it is not Avicenna’s view, as Avicenna says explicitly in another passage in Dem. 1.9, that experience alone, that is repeated observations of the temporal conjunction of phenomenon X and phenomenon Y, yields knowledge of the cause for the conjunction of the phenomena. Just as Ibn al-Quff alludes to in the beginning of L10, Avicenna stipulates that the certainty that the principle “X is Y” (“scammony purges yellow bile”) is true additionally requires a syllogism, in which the fact that Y (“purges yellow bile”) belongs to X (“scammony”) is the conclusion, Y (“purges yellow bile”) is the major term, X (“scammony”) is the minor term, and the cause (sabab) of the conjuction of phenomena is the middle term (“capacity to purge”).45 Avicenna does not explicitly set out the syllogism he has in mind, but Gutas reconstructs it thus:46

[Minor] Scammony has by nature the power to purge;
[Major] whatever has by nature the power to purge causes purging when ingested;
[Conclusion] therefore scammony by nature causes purging when ingested.

Dimitri Gutas says that for Avicenna the part of experience that involves “repeated observation of the fact” is a process that is independent of the part of the inductive process that involves the syllogism. It is only in combination that knowledge of the cause is arrived at inductively by experience.

Together, LL11–12 represent Avicenna’s synthesis of Aristotle’s account of induction in APr. 2.23 and APo. 2.19. In this account, experience appears in the stage of induction (mentioned in both APr. 2.23 and APo. 2.19) in which data from particulars are required. According to Avicenna’s synopsis of APr. 2.23 in L11, experience supplies the premises for the induction, which involves particular observations of horses, humans and bulls. The premises based on these observations allow us to form a universal affirmative proposition as a conclusion through induction, in which the middle term is a list of particulars, not a universal term that acts as a cause for the fact that the predicate belongs to the subject. In other words, Avicenna recognised that the inductive process outlined by Aristotle in APr. 2.23 was the ideal form of reasoning for acquiring first principles for a demonstrative science based on perception and experience. Yet, in the final paragraph of L12 Avicenna alludes again to the syllogistic form of induction in which experience participates. In this text, however, the minor premise is a cause (sabab) rather than a set of particulars. Avicenna appears to believe that if the experience upon which the inducted first principles are based is carried out under the conditions prescribed by Galen in his medical works, the middle term for the inductive syllogism need not be merely a list of particulars but can be a cause. This is because qualified experience, unlike common, unqualifed experience, allows the reasoner to see that the connection between the predicate “purges yellow bile” and the subject “scammony” is by nature rather than by coincidence.

LL11–12 are also where Avicenna addresses the question of the conviction (iʿtiqād) or certainty (yaqīn) that we have in first principles that are acquired inductively. Like many commentators, Avicenna recognised that on this particular question Aristotle was silent. It is in order to address the question of how we can be certain that the universal judgments we make are true that Avicenna introduces Galen’s ideas about qualified experience. In his reading of APo. 2.19, Avicenna takes “experience” to have, in effect, the characteristics of Galen’s qualified experience. Not just any sort of experience will yield certain and necessary first principles. Repeated experience of predictable connections between the predicate and subject whilst also remaining sensitive to the conditions in which the connections are observed yields knowledge that the connection between subject and predicate term arises not from mere chance, but from a natural connection or causal relation that exists between them. In short, Avicenna’s discussion in Dem. 1.9 is not a “critique” of Aristotle.47 Rather, Avicenna enlists the aid of what was a well-known element in Galen’s medical thought to counter an obvious objection to Aristotle’s account of inductively acquiring first principles in the Prior and Posterior Analytics.

6. Attitudes about Philosophy in Medicine from Late-Antiquity to the Thirteenth Century

Unlike other Arabic physicians reviewed above, Ibn al-Quff uses Avicenna’s Dem. 1.9 to explain what experience means in the Hippocratic Aphorisms. Why? What historical-textual conditions between Hippocrates and Ibn al-Quff made it possible for the latter to use Avicenna’s On Demonstration to explain the lemma of the Hippocratic Aphorisms? Several explanations may be given.48 Here is one that strikes me as plausible.

Aristotle’s Platonist commentators realised that several elements of Aristotle’s account of the acquisition of first principles of demonstrative science required clarification if not modification, Aristotle’s account of experience not the least. At APr. 2.23, Aristotle outlines the underlying syllogistic structure that contributes to how inductive reasoning (epagōgē) is able to yield conviction (Gr. pistis, Ar. yaqīn) about the universal affirmative judgment “A belongs to every B” from a limited number of observations of particulars in the extension of both A and B.49 However, at APr. 2.23 Aristotle does not explicitly say that this is how the reasoner acquires first principles. At APo. 2.19 99b20, Aristotle begins to describe a cognitive process by which the soul acquires first principles of a science from induction. First principles present something of challenge for Aristotle’s axiomatic hierarchy of the sciences. Whilst knowledge of the theses of a science result from syllogistic demonstrations from first principles, the first principles themselves are “direct” or “immediate (eirētai proteron).”50 The challenge for Aristotle in APo. 2.19 is two-fold. On the one hand, he wants to affirm that we have certain knowledge of first principles that is at least as secure as knowledge from demonstration. On the other hand, Aristotle needs to describe a non-demonstrative, non-syllogistic process by which we come to know first principles that is based on sense perception.51 The process detailed in APo. 2.19 100a2–12 Aristotle calls “induction (epagōgē = al-istiqrāʾ),” an activity that he associates with an “innate faculty that discriminates (dunamis sumphatos kritkē = qūwa ġarīzīya muḫtabira).” This faculty Aristotle calls “perception (aisthēsis = al-ḥiss),” and says that it is shared by all animals, but only in animals such as humans do these perceptions persist to form a memory. Repeated memories constitute a single, unitary experience.52

ἐν οἷς δ’ ἔνεστιν αἰσθομένοις ἔχειν ἔτι ἐν τῇ ψυχῇ. πολλῶν δὲ τοιούτων γινομένων ἤδη διαφορά τις γίνεται, ὥστε τοῖς μὲν γίνεσθαι λόγον ἐκ τῆς τῶν τοιούτων μονῆς, τοῖς δὲ μή.

Ἐκ μὲν οὖν αἰσθήσεως γίνεται μνήμη, ὥσπερ λέγομεν, ἐκ δὲ μνήμης πολλάκις τοῦ αὐτοῦ γινομένης ἐμπειρία· αἱ γὰρ πολλαὶ μνῆμαι τῷ ἀριθμῷ ἐμπειρία μία ἐστίν. ἐκ δ’ ἐμπειρίας ἢ ἐκ παντὸς ἠρεμήσαντος τοῦ καθόλου ἐν τῇ ψυχῇ, τοῦ ἑνὸς παρὰ τὰ πολλά, ὃ ἂν ἐν ἅπασιν ἓν ἐνῇ ἐκείνοις τὸ αὐτό, τέχνης ἀρχὴ καὶ ἐπιστήμης.

‮وأما ما كان من الحيوان يثبت فيه [الحس] فقد يبقى عندما يحس شيء ما في أنفسها. وعندما يكون كثير من أمثال هذه فقد يحدث حينئذ تمييز ما وتفصيل، حتى أن لبعضها قد يكون من ثبات أمثال هذه قول، وبعضها لا. فمن الحس يكون حفظ كما قلنا. ومن تكرير الذكر مرات كثيرة تكون تجربة، وذلك أن الأحفاظ الكثيرة في العدد هي تجرة واحدة. ومن التجربة عندما يثبت ويستقر الكلي في النفس الذي هو واحد في الكثير، ذلك الذي هو في جميعها واحد بعينه هو مبدأ الصناعة والعلم.‬‎

[L13] In those animals in which [the perception] persists, when they perceive, something remains in their souls. When there are many such things, a discrimination and distinction comes about (diaphora tis = tamyīzun mā wa-tafṣīlun) so that there is an account for those of them that persist, but for others not.

As we said, from perception comes memory, and from memory that is repeated many times comes experience. For memories that are many in number are a single experience. When the universal that is one in many, that which is identically one in every one of them [memories]—[when such a universal] persists and is fix in the soul, this is the principle of the art and science.

Aristotle describes a three step non-demonstrative, non-deductive process—this fact is made explicit in Isḥāq’s interpretive translation of this phrase—in which the innate faculty of perception has many particular instances of perception, which together constitute a memory. Many memories all together constitute a single experience, the outcome of which is a universal first principle of the practical arts or the science. Aristotle later in the text calls this process “induction.” According to the scheme that Aristotle has set down, we will not understand the first principle “X is Y” in the sense that we will not derive them by means of a demonstrative syllogism in which the middle term is the cause for the fact that Y belongs to X. Nevertheless, there is a sense in which we grasp the first principles and it is our grasping these principles that serves as the basis for understanding. Yet, even if induction does not yield understanding (epistimē) of the first principles, as we have seen, according to APr. 2.23, induction yields (1) conviction (pistis) about universal statements (2) based on a finite number of observations of particulars. Moreover, the induction process does admit a (3) syllogistic structure, but one in which the middle term does not serve as a reason or the account for the fact that the predicate belongs to all (or none) of the subject.

First principles derived from experience that do not supply information about the cause for the connetion between the subject and predicate are hardly suited to serving as the axioms of a demonstrative science, which Aristotle held to be a discourse yielding necessary knowledge derived from necessary premises linked by middle terms that serve as cause. Yet, in spite of this belief, in APr. 2.23 68b30–32, Aristotle clearly holds that induction is the ideal means for acquiring first principles, since induction “is the sort of syllogism of a primary and unmiddled premise. For the middled syllogism is through the middle term. But the syllogism that is not is by means of induction.”53 Yet, how can the certainty and necessity Aristotle requires of demonstrative science be acquired by means of an inductive process based on perception?54

Consistent with the importance of Aristotle’s and Plato’s philosophy and Galenic medicine in the Alexandrian curriculum, in order to overcome this difficulty it appears that some authors referred to the medical discourse on experience in Hippocratic and Galenic writings. In particular, it seems that medical debates relating to the problem of how to acquire reliable pharmacological data were recognised as being relevant to explaining Aristotle’s text. For example, in his commentary on Posterior Analytics, Book Two, John Philoponus55 (d. 570) says:

ἐκ δὲ μνήμης πολλάκις τοῦ αὐτοῦ γινομένης, ἤγουν πολλῶν δὲ μνημῶν συναθροισθεισῶν, γίνεται ἐμπειρία, ἤγουν γνῶσίς τις δυνάμεως πράγματός τινος. οἷον εἶδον μυριάκις τὸν ἐλλέβορον κενοῦντα χολήν, καὶ πολλὰ αἰσθήματα τοιαῦτα ἐν τῇ φαντασίᾳ μου ἐνετυπώθησαν, ἐξ ὧν ἐπισυνηθροίσθησαν μνῆμαι πολλαί· ἐκ δὲ τῶν πολλῶν μνημῶν γέγονέ μοι ἐμπειρία καὶ γνῶσις, ὡς ὁ ἐλλέβορος δύναμιν ἔχει κενωτικὴν χολῆς. ἡ δὲ γνῶσις αὕτη ἠρεμήσασα καὶ ἐμπαγεῖσα καὶ ἑδραιωθεῖσα ἐν τῇ ψυχῇ μου, ὡς οὕτως ἔχει ὁ ἐλλέβορος καὶ οὐκ ἄλλως, τὸ καθόλου συνῆξεν, ἤγουν τὸ “πᾶς ἐλλέβορος καθαίρει,” ὃ καθόλου ἐστὶν ἀρχὴ τῶν ἀποδείξεων.

[L14] From frequent memory of something coming about—that is, from many memories that have assembled together—experience (empeiria) comes about; that is to say, a certain recognition (gnōsis tis) of the capacity that something has. Such as when on numerous occassions I saw hellebore purge, many of these perceptions are imprinted in my imagination (phantasia), by which many memories are gathered together. And from the many memories comes my experience and recognition (gnōsis) that hellebore has the capacity to purge bile. Likewise, the recognition that hellebore has such [a capacity] and nothing else came to rest, became fixed and stable in my soul, that is to say the universal “all hellebore purges” was made to take shape [in the soul], the universal which is is the principle of demonstrations.56

Of course, this text makes no pretense to supply a solution to the dilemmas in Aristotle’s text. However, it is at least clear that (1) Philoponus perceived that the challenges posed by experience derived from observing pharmacological phenomena were relevant to interpreting Aristotle’s text; and more importantly (2) Philoponus speaks about how numerous observations of hellebore purging yields recognition (gnōsis not epistimē) that there is a capacity (dunamis) in hellebore that causes purging. And though Philoponus does not speak about causes, middle terms or syllogisms, it is important to recognise the fact that Philoponus speaks about “capacity to purge,” which is what Avicenna eventually identifies as the middle term in the syllogism that supplements raw experience, transforming it into knowledge of the cause.

We should see, then, Avicenna’s discussion in Dem. 1.9 as a continuation of this trend in late antique commentaries on APo. 2.19, in which medical discourse on experience was brought to bear on this text. Indeed, when McGinnis observes that in Dem. 1.9 Avicenna urges the reader to “note the variables or various antecedent and background conditions surrounding the observations,” we quickly realise that Avicenna has adapted Galen’s medical theory of qualified experience into the philosophical discourse surrounding Aristotle’s account of experience in APo. 2.19. For example, in Dem. 1.9, Avicenna uses the example of how using experience naively leads a person to the false conclusion “all people are dark-skinned” because he drew his experience exclusively from populations in countries where most or all people are dark-skinned.57 With this example, Avicenna demonstrates that if proper attention is not given to the characteristic of the sample on which observations are made, “accidental or chance relations” between predicate and subject can be mistaken for universal and necessary relations.58 In fact, by enlisting the aid of Galenic qualified experience Avicenna is addressing precisely the problem with Aristotle’s account of inductively acquiring first principles.

Thus, by the time that the debate about experience reached Ibn al-Quff at the end of the thirteenth century, the ancient debates about experience have been synthesised and given rigorous formulation not only in Avicenna’s medical works such as the Canon, but in Avicenna’s philosophical writings as well. Avicenna continues a trend that was apparently commonplace in late Alexandrian medicine and philosophy, in which Galen’s medical thought was brought to bear on solving philosophical issues in Aristotle’s philosophy. In the case before us, Galen’s idea of unqualified experience is used to clarify Aristotle’s discussion of how experience leads to acquring first principles of demonstration. Thus, by the thirteenth century, it would have been natural for medical authors such as Ibn al-Quff to refer to Avicenna’s philosophical works for insights into Hippocratic works. It appears that aš-šayḫ ar-raʾīs himself did the same.

7. Conclusion

In this article, I have tried to place the debates about medical experience in Arabic commentaries on the Hippocratic Aphorisms in their historical context. There were two influential accounts of experience that influenced these authors: Galen’s qualified experience in his pharmacological works and Aristotle’s discussion of experience in the final sections of the Prior and Posterior Analytics. I have shown how Aristotle’s Greek commentators began to transform or even synthesise these senses of experience. However, it seems to me that Alfarabi and Avicenna took the synthesis of these philosophical and medical accounts of experience to new levels of rigour and complexity. Avicenna’s contributions to the form taken by debates about experience among the Arabic Aphorisms commentators is especially noteworthy. Alfarabi’s discussion of Galen’s anatomical method shows that classical Arabic thinkers recognized that Galen had a generic method for discovering causal relations between phenomena that he used in pharmacology, anatomy and therapeutics. Alfarabi’s criticism of Galen’s method relies on the fact that ambiguating conditions reduce the evidential value of these observations for proving statements about causes. On the other hand, in the opening pages of Canon 2 Avicenna distills Galen’s scattered comments on qualified experience into a rule-based method for discovering pharmacological effects.

Yet, I have also tried to show how Avicenna’s philosophical works come to influence post-classical medical authors such as Ibn al-Quff. Unique among the commentators on the first aphorism, Ibn al-Quff quotes from Avicenna’s definition of experience in On Demonstration, which is his philosophical work based on Aristotle’s Posterior Analytics. By drawing on Avicenna’s philosophical works to explain Hippocrates’ words, Ibn al-Quff is following Avicenna’s example. For in Dem. 1.9, Avicenna uses Galen’s notion of qualified experience to address problems he saw in Aristotle’s account of what role experience plays in the acquisition of first principles of demonstration. Ibn al-Quff combines Avicenna’s medical and philosophical thought to produce an entirely new understanding of the Hippocratic text that often has little to do with how Galen understood the same text.

Appendix: Translation of Avicenna on Medical Experience in the Canon of Medicine

This translation is based on a 1971 reprint of the 1877 Bulaq edition of the Canon: Avicenna, al-Qānūn fī al-ṭibb, ed. Qāsim Muḥammad al-Raǧab (Baghdad: Dār al-Muṯannā, 1971), 1:224, l. 30–6, l. 4. Peter E. Pormann has translated the text of Condition 2 (225, ll. 3–9) in Pormann, “Avicenna on Medical Practice, Epistemology, and the Physiology of the Inner Senses,” 99. A partial translation is in Nasser, Tibi and Savage-Smith, “Rules for Assessing the Effects of Drugs.” Finally, there is an adapted translation of Canon 2 by Laleh Bakhtiar, The Canon of Medicine (al-Qānūn fī l-ṭibb) (The Law of Natural Healing): Volume 2 Natural Pharmaceuticals (Chicago: Kazi Publications, 2012).

The Second Discourse: On Coming to Know the Powers of Drug Mixtures by Experience

The powers of drugs come to be known in two ways. One of them is by a process of reasoning (ṭarīq al-qiyās). The other is by experience (ṭarīq al-taǧriba). Let us take up the discussion of experience first. We say that experience leads to reliable knowledge of the drug’s power when certain conditions are observed.

[Condition 1] The first is that the drug is free of any acquired quality (kayfīya muktasaba), whether it is heat that is accidental or cold that is accidental, or that it acquires a quality that comes from a transformation in its substance, or another [substance] combines with it. For even if water is cold by nature, when it is heated, it continues to heat for as long as it is hot. In spite of the fact that euphorbium (al-furbiyūn) is naturally hot, when it is made cold, it makes [other things] cold as long as it is cold. In spite of the fact that almonds (al-lawz) tend to be balanced and fine, when they spoil they heat strongly. And in spite of the fact that the flesh of fish is cold, when it is salted it heats strongly.

[Condition 2] The second is that what experience of the drug is conducted on (al-muǧarrab ʿalayhi) is a simple illness (ʿilla mufrada). For if it is a compound illness (ʿilla murakkaba) in which there are two elements that require therapies that are opposite to each other, and the experience of the drug is conducted on both of them, after which there is a benefit that comes about, the real reason for this [benefit] is not perceived. For example, if a patient has a phlegmatic fever, and we give him agaric to drink and the fever recedes, it is not necessary to judge that agaric is cold because it brings benefit in the case of a hot illness, which is the fever. On the contrary, it may happen that the agaric brought benefit because it dissolved the phlegmatic disease matter, or because it evacuated it from the body so that when the matter was eliminated, the fever receded. In reality, this is an essential benefit that is mixed accidentally. It is essential in relation to the disease matter, and it is accidental in relation to the fever.

[Condition 3] The third is that experience of the drug is conducted on a disease that is the opposite of the disease in question, so that if the drug brings benefit in the case of both diseases, the [drug’s] temperament is not judged to be the opposite of the temperament of one of them. For it may be that the [the drug] benefits in one of the illnesses essentially but in the other accidentally. If experience of scammony is conducted on a cold illness, it is likely that it will bring benefit and heat. If it is conducted on a hot illness such as tertian fever, it is likely that it will bring benefit because it evacuates the yellow bile. But if this is the case, the experience does not yield reliable [information] about whether scammony is hot or cold until it is known that the activity in the first case is essential and the activity in the other case is accidental.

[Condition 4] The fourth is that the power in the drug equals the power in the illness that [the drug] opposes. For the heat in some drugs falls short of the cold that is in the illness, so that [the drug] has no influence on the illness at all, but there might have been a heating action if the drug were used in the condition that the cold was less intense (aḫaff). Thus, it is necessary that experience of the drug be conducted on the weakest form of the illness (al-aḍʿaf), and then [the strength of the disease that experience of the drug is conducted on] is gradually increased so that the strength of the drug is known and there is not any doubt about it.

[Condition 5] The fifth is that one observe the time in which the drug’s effect and activity appear. If they appear directly after the drug is given, it is convincing that this [effect and actvity] occurs essentially. If what appears first is an activity that is opposite to what appears last, or no activity appears at first but then the activity appears last, this then will be a source of ambiguity and suspicion that it may be that the activity the drug has is accidental, as if the drug first had a subtle [essential] effect, which is then followed accidentally by this later, obvious effect. This suspicion and ambiguity about the drug’s power and the surmise that the drug’s activity is accidental is strong when the [drug’s] activity appears after departing from the body part. For if the drug were to act essentially, it would act whilst it was in contact the body part, and it would be impossible for it to remain ineffective whilst it was in contact with the body part but to act whilst it departs from the body part. This is a judgment that holds for the most part and is convincing.

However, it sometimes happens that some bodies act essentially after they act accidentally. This happens when they acquire a power that is extrinsic to them that overwhelms the drug’s natural power, as is the case of hot water. For in that state, [water] heats, but the next day or the next moment in which the accidental effect has dissipated, water ineluctably brings about a cold [quality] in the body owing to the fact that parts from which the water is constituted transform back to the cold that is in water in its natural state.

[Condition 6] The sixth is that one ensures that the drug’s activity progresses continually or for the most part. For if it does not, its activity proceeds from it accidentally, since natural states of affairs proceed from their origins continually or for the most part.

[Condition 7] The seventh is that the experience is on the human body. For if it is conducted on other than the human body, it may be inadequate on two counts. The first is that it is possible that in relation to the human body the drug is hot, but cold in relation to a lion if [the drug] is hotter (aḥarr) in relation to the human but colder than the lion and the horse. It seems to me that rhubarb (al-rāwānd) is very cold for a horse but is hot for a human. The second is that it is possible that in relation to one of the bodies [the drug] has a property (ḫāṣṣīya) that it does not possess in relation to the other body. For with respect to the human body monk’s-hood (al-bīš) [Aconitum napellus] has a poisonous property, but in relation to starlings (al-zarāzīr) it does not.

These, then, are the rules that must be observed when the powers of drugs are derived by way of experience. So understand that.

Bibliography

Manuscript Sources

Ibn Abī Ṣādiq al-Nīsābūrī, Abū al-Qāsim (d. after 1068). Šarḥ fūṣūl Abuqrāṭ. MS, Istanbul, Beyazid Devlet Kütüphanesi, Veliyeddin Efendi 2508.

Ibn al-Nafīs, ʿAlāʾ al-Dīn ʿAlī ibn Abī al-Ḥazm al-Qurašī (d. 1288). Šarḥ al-fuṣūl li-Abuqrāṭ. MS, Oxford, Bodleian Library, Pococke 294.

Ibn al-Quff, Abū al-Faraǧ ibn Yaʿqūb ibn Isḥāq (d. 1286). Kitāb Taysīr al-wuṣūl ilā tafsīr al-fuṣūl li-Abuqrāt. MS, Gotha, Landesbibliothek, 1894.

Edited and Published Sources

Ahmed, Asad Q. Avicenna’s Deliverance: Logic. Karachi: Oxford University Press, 2011.

Alfarabi. Al-Manṭiq ʿinda l-Fārābī. Edited by Rafīq al-ʿAǧam and Majid Fakhry. 4 vols. Beirut: Dār al-Mašriq, 1985–1986.

Aristotle. Prior Analytics. Translated by Robin Smith. Indianapolis: Hackett, 1989.

Avicenna. Al-Šifā, al-manṭiq al-burhān. Edited by Abū al-ʿAlā al-ʿAfīfī. Cairo: al-Maṭbaʿa al-amīrīya bi-l-Qāhira, 1956.

Avicenna. Al-Qānūn fī al-ṭibb. Edited by Qāsim Muḥammad al-Raǧab. 3 vols. Baghdad: Dār al-muṯannā, 1971.

Avicenna. Kitāb al-Naǧāt fī al-ḥikma al-manṭiqīya wa-ṭ-ṭabīʿīya wa-l-ilāhīya. Edited by Majid Fakhry. Beirut: Dār al-āfāq al-ǧadīda, 1982.

Avicenna. The Canon of Medicine (al-Qānūn fī l-ṭibb) (The Law of Natural Healing): Volume 2 Natural Pharmaceuticals. Translated by Laleh Bakhtiar. Chicago: Kazi Publications, 2012.

Badawī, ʿAbd al-Raḥmān. Manṭiq Arisṭū. 3 vols. Cairo: Dār al-kutub al-miṣrīya, 1948.

al-Baġdādī, ʿAbd al-Laṭīf. Šarḥ al-fuṣūl. Edited by Sherif Masry. The University of Manchester, 2012–2017. http://dx.doi.org/10.3927/51688866.

Barnes, Jonathan. Aristotle’s Posterior Analytics. Oxford: Clarendon Press, 1976.

Bayer, Greg. “Coming to Know Principles in Posterior Analytics II.19.” Apeiron 30 (1997): 109–42.

van der Eijk, Philip. “Galen’s Use of the Concept of ‘Qualified Experience’ in his Dietetic and Pharmacological Works.” In Medicine and Philosophy in Classical Antiquity: Doctors and Philosophers on Nature, Soul, Health and Disease, edited by Philip van der Eijk, 279–98. Cambridge: Cambridge University Press, 2005.

Endress, Gerhard. “Reading Avicenna in the Madrasa: Intellectual Genealogies and Chains of Transmission of Philosophy and the Sciences in the Islamic East.” In Arabic Theology, Arabic Philosophy from the Many to the One: Essays in Celebration of Richard M. Frank, edited by James. E. Montgomery, 371–422. Leuven: Peeters, 2006.

Frede, Michael. Die Stoische Logik. Göttingen: Vanderhoek and Ruprecht, 1974.

Forcada, Miquel. “Ibn Bājja on Medicine and Medical Experience.” Arabic Sciences and Philosophy 21 (2011): 111–48.

Galen. Galeni Opera Omnia. Edited by Karl G. Kühn. 20 vols. Leipzig: Car. Cnoblochii, 1821–1833.

Galen. Tafsīr Ǧālīnūs li-Kitāb al-fuṣūl. Edited by Taro Mimura. The University of Manchester, 2012–2017. http://dx.doi.org/10.3927/51689293.

Goldin, Owen. Philoponus(?): On Aristotle’s Posterior Analytics 2. London: Duckworth, 2009.

Gutas, Dimitri. “Medical Theory and Scientific Method in the Age of Avicenna.” In Before and After Avicenna, edited by David C. Reisman, 145–62. Leiden: Brill, 2003.

Gutas, Dimitri. “The Empiricism of Avicenna,” Oriens 40 (2012): 392–436.

Hintikka, Jaakko. Time and Necessity: Studies in Aristotle’s Theory of Modality. Oxford: Clarendon Press, 1973.

Ibn Abī Ṣādiq. Šarḥ fuṣūl Abuqrāṭ. Edited by ARABCOMMAPH. The University of Manchester, 2012–2017. http://dx.doi.org/10.3927/51931955.

Ibn al-Nafīs. Šarḥ al-fuṣūl li-Abuqrāṭ. Edited by Sherif Masry. The University of Manchester, 2012–2017. http://dx.doi.org/10.3927/52844770.

Ibn al-Quff. Kitāb Taysīr al-wuṣūl ilā tafsīr al-fuṣūl li-Abuqrāt. Edited by ARABCOMMAPH. The University of Manchester, 2012–2017. http://dx.doi.org/10.3927/52131919.

Janssens, Jules. “ ‘Experience’ (tajriba) in Classical Arabic Philosophy (al-Fārābī–Avicenna).” Quaestio 4 (2004): 45–62.

Lennox, James. Aristotle’s Philosophy of Biology. Cambridge: Cambridge University Press, 2000.

Martin, John N. “Proclus and the Neoplatonic Syllogistic.” Journal of Philosophical Logic 30 (2001): 187–240.

McCaskey, John P. “Freeing Aristotelian Epagōgē from Prior Analytics II 23.” Apeiron 40 (2007): 345–74.

McGinnis, Jon. “Scientific Methodologies in Medieval Islam.” Journal of the History of Philosophy 41 (2003): 307–27.

McGinnis, Jon. “Avicenna’s Naturalized Epistemology and Scientific Methods.” In The Unity of Science in the Arabic Tradition: Science, Logic, Epistemology and Their Interactions, edited by Shahid Rahman, Hassan Tahiri and Tony Street, 129–52. Dordrecht: Kluwer Academic Publishers, 2008.

Nasser, Mona Aida Tibi and Emilie Savage-Smith. “Ibn Sina’s Canon of Medicine: 11th Century Rules for Assessing the Effects of Drugs.” Journal of the Royal Society of Medicine 102 (2009): 78–80.

Philoponus. In Aristotelis Analytica Posteriora commentaria cum anonymo in librum II (CAG 13.3). Edited by Maximillian Wallies. Berlin: Akademie der Wissenschaften, 1882.

Pormann, Peter E. “Avicenna on Medical Practice, Epistemology, and the Physiology of the Inner Senses.” In Interpreting Avicenna: Critical Essays, edited by Peter Adamson, 91–108. Cambridge: Cambridge University Press, 2013.

Rosenthal, Franz. “ ‘Life is Short, the Art is Long’: Arabic Commentaries on the First Hippocratic Aphorism.” Bulletin of the History of Medicine 40 (1966): 226–45.

Ross, William D. Aristotle’s Prior and Posterior Analytics. Oxford: Clarendon Press, 1949.

Savage-Smith, Emilie. A New Catalogue of Arabic Manuscripts in the Bodleian Library, University of Oxford Volume 1: Medicine. Oxford: Oxford University Press, 2011.

Simplicius. Simplicii in Aristotelis Categorias Commentarium (CAG 8). Edited by Karl Kalbfleisch. Berlin: Akademie der Wissenschaften, 1882.

Simplicius. On Aristotle’s “Categories, 9–15.” Translated by Richard Gaskin. Ithaca: Cornell University Press, 2000.

von Staden, Heinrich. “Experiment and Experience in Hellenistic Medicine.” Bulletin of the Institute of Classical Studies 22 (1975): 178–99.

Strobino, Riccardo. “Avicenna on the Indemonstrability of Definition.” Documenti e Studi Sulla Tradizione Filosofica Medievale 21 (2010): 113–63.

Thom, Paul. “Al-Fārābī on Indefinite and Privative Names.” Arabic Sciences and Philosophy 18 (2008): 193–209.

Ullmann, Manfred. Wörterbuch zu den griechisch-arabischen Übersetzungen des 9. Jahrhunderts. Wiesbaden: Harrassowitz Verlag, 2002.

Winter, Michael. “Aristotle, hōs epi to polu Relations, and a Demonstrative Science of Ethics.” Phronesis 42 (1997): 163–89.

Zimmermann, Fritz W. Al-Farabi’s Commentary and Short Treatise on Aristotle’s De Interpretatione. Oxford: British Academy and Oxford University Press, 1981.

1

Franz Rosenthal, “ ‘Life is Short, the Art is Long’: Arabic Commentaries on the First Hippocratic Aphorism,” Bulletin of the History of Medicine 40 (1966): 226–45; Jules Janssens, “ ‘Experience’ (tajriba) in Classical Arabic Philosophy (al-Fārābī–Avicenna),” Quaestio 4 (2004): 45–62; Jon McGinnis, “Avicenna’s Naturalized Epistemology and Scientific Methods,” in The Unity of Science in the Arabic Tradition: Science, Logic, Epistemology and Their Interactions, eds. Shahid Rahman, Hassan Tahiri and Tony Street (Dordrecht: Kluwer Academic Publishers, 2008), 129–52; Miquel Forcada, “Ibn Bājja on Medicine and Medical Experience,” Arabic Sciences and Philosophy 21 (2011): 111–48.

2

Philip van der Eijk, “Galen’s Use of the Concept of ‘Qualified Experience’ in his Dietetic and Pharmacological Works,” in Medicine and Philosophy in Classical Antiquity: Doctors and Philosophers on Nature, Soul, Health and Disease, ed. Philip van der Eijk (Cambridge: Cambridge University Press, 2005), 279–98, 289.

3

Ibid.

4

Ibid., 290.

5

Greek text: Galen, Galeni οpera οmnia, ed. Karl G. Kühn (Leipzig: Car. Cnoblochii, 1829), 17b:354, l. 10–5, l. 3. Arabic text: Galen, Tafsīr Ǧālīnūs li-fuṣūl Abuqrāṭ, trans. Ḥunayn ibn Isḥāq, ed. Taro Mimura (The University of Manchester, 2012–2017), http://dx.doi.org/10.3927/51689293. The English translation of L1 is based on the Greek. Unless otherwise advertised, this and all other translated texts in this paper are mine.

6

van der Eijk, “Galen’s Use of Qualified Experience,” 287.

7

Peter E. Pormann, “Avicenna on Medical Practice, Epistemology, and the Physiology of the Inner Senses,” in Interpreting Avicenna: Critical Essays, ed. Peter Adamson (Cambridge: Cambridge University Press, 2013), 91–108, 98–9.

8

Avicenna, al-Qānūn fī al-ṭibb, ed. Qāsim Muḥammad al-Raǧab (Baghdad: Dār al-muthannā, 1971), 1:224. I recognise that in his philosophical works, Avicenna invariably uses qiyās to speak about the syllogism in its technical, Aristotelian sense. However, I do not assume that he uses it an identical way in his medical writings. This is why I have elected to use the more generic translation “process of reasoning.” On the further complications associated with translating qiyās in classical and post-classical Arabic medical discourse, see Section 4.

9

See Pormann, “Avicenna on Medical Practice, Epistemology, and Physiology,” 98.

10

Part of this passage is translated in Mona Nasser, Aida Tibi, and Emilie Savage-Smith, “Ibn Sina’s Canon of Medicine: 11th Century Rules for Assessing the Effects of Drugs,” Journal of the Royal Society of Medicine 102 (2009): 78–80. See the Appendix for a complete translation.

11

Fritz W. Zimmermann uses the words “eliminate” and “cancel” to translate rafaʿa and irtafaʿa. For example, see Fritz Zimmermann, Al-Farabi’s Commentary and Short Treatise on Aristotle De Interpretatione (Oxford: Oxford University Press, 1981), cxxx, n. 2.

12

See John N. Martin, “Proclus and the Neoplatonic Syllogistic,” Journal of Philosophical Logic 30 (2001): 187–240.

13

This is the topic of what Alfarabi calls the “Third Section” of Aristotle’s On Interpretation, 19b19–21a33. For Alfarabi’s “long commentary,” see Zimmermann, Al-Farabi’s Commentary and Short Treatise on Aristotle De Interpretatione, 100–57; for this topic in the “short commentary,” see ibid., 234–40. See also Paul Thom, “Al-Fārābī on Indefinite and Privative Names,” Arabic Sciences and Philosophy 18 (2008): 193–209.

14

ʿAbd al-Raḥmān Badawī, Manṭiq Arisṭū (Cairo: Maṭbaʿat dār al-kutub al-miṣrīya, 1948), 1:70. For Aristotle’s use of akolouthein to express various senses of existing together, see Jaakko Hintikka, Time and Necessity: Studies in Aristotle’s Theory of Modality (Oxford Clarendon Press, 1973), 41–61 [Chapter II], especially 43–5 and 51–4. Hintikka notes (ibid., 42) that “usually translators and commmentators [modern, ancient?] do not make any clear-cut distinctions between the different verbs Aristotle uses to express following, compatibility, logical equivalence, and logical entailment.” I am grateful to the peer-reviewer for directing me to this section of the book, and suggesting that I avoid translating akolouthein as “entailment.” I have elected to use the generic notion of “following” because following from is not necessarily a symmetrical relationship in the way that compatible with is. It thus conveys the sense that Aristotle and Alfarabi are trying to impress on the reader about priority in existence. This translation also makes it clear why Simplicius would have felt comfortable glossing it with the Stoic term for logical entailment “sunepipheirei.”

15

Adopteding “δυοεῖν” with other manuscripts against Kalbfleisch’s “δυεῖν.”

16

Simplicius, Simplicii in Aristotelis Categorias Commentarium (CAG 8), ed. Karl Kalbfleisch (Berlin: Akademie der Wissenschaften, 1907) 419, l. 20–20, l. 5.

17

Gaskin says that οἱ νεώτεροι “probably” refers to the Stoics; Simplicius, On Aristotle’s “Categories, 9–15,” trans. Richard Gaskin (Ithaca: Cornell University Press, 2000), 258. This is plausible. Michael Frede notes that ἐπιφορά “is the Stoic term, which corresponds to the Peripatetic συμπέρασμα.” Michael Frede, Die Stoische Logik (Göttingen: Vanderhoek & Ruprecht, 1974), 118, n. 2 (from original German). It may be that sunepipheron is a terminological innovation of Aristotle’s commentators when they began to adapt and mix Stoic and Peripatetic terminology. Neverthless, the idea of implication and cancellation used in the sense they have in L2 recalls another passage in Aristotle’s Categories. The idea of cancellation as it is used here appears to derive from an earlier passage in Aristotle’s Categories, on “relatives (τὰ πρός τι).” At Cat. 7b15–22, Aristotle observes that “it is clear that by nature relatives come to be together (Δοκεῖ δὲ τὰ πρός τι ἅμα τῇ φύσει εἶναι),” using the relation between master (δεσπότης) and slave (δοῦλος) to illustrate his point. Aristotle says: “there being the master, the slave is, and there being the slave, the master is. Other cases are like this. And this one cancels (συναναιρεῖ) the other, for there being no double, there is no half, and there being no half, there is no double.” Thus, what seems likely is that οἱ νεώτεροι refers to some of Aristotle’s Platonist commentators who synthesised these different parts of the Categories into an account of a type of following (ἀκολουθία) that exists between terms.

18

Alfarabi, al-Manṭiq ʿinda l-Fārābī, ed. Rafīq al-ʿAǧam (Beirut: Dār al-mašriq, 1985), 1:103–4. Hereafter, citations from this work appear as follows: Alfarabi, On Topical Analysis.

19

Alfarabi, On Topical Analysis, 106.

20

Simplicius, On Aristotle’s “Categories, 9–15,” 424, ll. 23–5 [p. 172]. Translation is Gaskin’s.

21

van der Eijk, “Galen’s Use of Qualified Experience,” 282–3.

22

Ibid.

23

Ibn Abī Sādiq, Šarḥ fuṣūl Abuqrāṭ, ed. ARABCOMMAPH (The University of Manchester, 2012–2017), doi: http://dx.doi.org/10.3927/51931955; Istanbul, Beyazid Devlet Kütüphanesi, ms. Veliyeddin Efendi 2508, fol. 2b, ll. 2–11.

24

Ullmann, Medizin, 171.

25

ʿAbd al-Laṭīf al-Baġdādī, Šarḥ al-fuṣūl, ed. Sherif Masry (The University of Manchester, 2012–2017), http://dx.doi.org/10.3927/51688866.

26

Ibn al-Nafīs, Šarḥ al-fuṣūl li-Abuqrāṭ, ed. Sherif Masry (The University of Manchester, 2012–2017), http://dx.doi.org/10.3927/52844770; MS, Oxford, Bodleian Library, Pococke 294, fols. 4b, l. 12–5a, l. 10. For questions about authorship, see Emilie Savage-Smith, A New Catalogue of Arabic Manuscripts in the Bodleian Library, University of Oxford Volume 1: Medicine (Oxford: Oxford University Press, 2011), 18 [Entry No. 6].

27

See Condition 1 and 2 in the Appendix.

28

See Condition 4 in the Appendix.

29

See Condition 3 in the Appendix.

30

This phrase is probably another importation of Aristotelian scientific jargon into medical discourse. It is used prominently in the Metaphysics and the Posterior Analytics to charactersise what kind of object scientific knowledge is about. The Greek expression is ὡς ἐπὶ τὸ πολύ, which is translated into Arabic variously as, for example, fī akṯari l-amri or ʿalā l-amri l-akṯari. See, for example, Michael Winter, “Aristotle, hōs epi to polu Relations, and a Demonstrative Science of Ethics,” Phronesis 42 (1997): 163–89. Thus, Posterior Analytics 2.12 (96a17–9): “There will be immediate first principles also in the case of what holds for the most part (fa-takūnu iḏani l-mabādiʾu ġayra ḏawāti awsāṭin li-l-ašyāʾi llatī fī akṯari l-amri = ἔσονται τοίνυν καὶ τῶν ὡς ἐπὶ τὸ πολὺ ἀρχαὶ ἄμεσοι, ὅσα ὡς ἐπὶ τὸ πολὺ οὕτως ἔστιν ἢ γίνεται).” Greek text: William D. Ross, Aristotle’s Prior and Posterior Analytics (Oxford: Clarendon Press, 1949); Arabic text: ʿAbd al-Raḥmān Badawī, Manṭiq Arisṭū (Cairo: Maṭbaʿat Dār al-kutub al-miṣrīya, 1948), 2:461. Likewise, Posterior Analytics 1.30 (87b19–22): “There is no knowledge (epistēmē, ʿilm) through demonstration of what holds by chance (tuchēs, bi-l-ittifāq). For what holds by chance is neither necessary nor does it hold for the most part (fa-ammā š-šayʾu llaḏī ʿani l-ittifāqi fa-lā ʿilma bihi bi-l-burhāni iḏ kāna l-amru llaḏī bi-l-ittifāqi laysa huwa ḍarūrīyan wa-lā ʿalā akṯari l-amri = Τοῦ δ’ ἀπὸ τύχης οὐκ ἔστιν ἐπιστήμη δι’ ἀποδείξεως. οὔτε γὰρ ὡς ἀναγκαῖον οὔθ’ ὡς ἐπὶ τὸ πολὺ τὸ ἀπὸ τύχης ἐστίν).” Greek text: Ross, Aristotle’s Analytics; Arabic text: Badawī, Manṭiq Arisṭū, 2:417.

31

See Condition 6 in the Appendix.

32

See L1 above; Manfred Ullmann, Wörterbuch zu den griechisch-arabischen Übersetzungen des 9. Jahrhunderts [WGAÜ] (Wiesbaden: Harrassowitz Verlag, 2002), 397: for example, baʿḍuhu mina t-taǧāribi wa-baʿḍuhu mina l-qiyāsi = ta men ek tēs peiras, ta d’ ek tou logou.

33

See Ullmann, WGAÜ, 650–1.

34

See Avicenna in Appendix; Ibn Abī Ṣādiq in L5 above.

35

Ibn al-Nafīs, Šarḥ fuṣūl Abuqrāṭ, http://dx.doi.org/10.3927/52844770; MS, Oxford, Bodleian Library, Pococke 294, fol. 6a, ll. 10–13.

36

Avicenna, al-Qānūn, 1:225, ll. 17–27.

37

Avicenna is using ḥads with its lexical meaning, not in the technical sense he assigns it in his psychology and epistemology, in which the philosopher’s powerful intellect spontaneously hits on the middle term of demonstrative syllogism. See Dimitri Gutas, “The Empiricism of Avicenna,” Oriens 40 (2012): 392–436, 400. Nevertheless, Gutas notes that the one of the types of proposition in Avicenna’s logic is what Gutas calls “data provided by finding the middle term of a syllogism” which corresponds to Avicenna’s term “ḥadsīyāt.” Gutas says that these are propositions that are “based on experience.”

38

Gutas has written on this topic in Dimitri Gutas, “Medical Theory and Scientific Method in the Age of Avicenna,” in Before and After Avicenna, ed. David C. Reisman (Leiden: Brill, 2003), 145–62. In this article, Gutas claimed that “medicine, given the status accorded to it in the classification of the sciences, never became part of the mainstream theoretical academic curriculum; it was only a practical craft, learned and transmitted mostly by way of apprenticeship in the hospitals (ibid., 161).” Gerhard Endress’ study on Faḫr al-Dīn al-Rāzī (d. 1210) and the influence of his commentary on Book One of Avicenna’s Canon of Medicine shows that Gutas’ claim is somewhat inaccurate. Gerhard Endress, “Reading Avicenna in the Madrasa: Intellectual Genealogies and Chains of Transmission of Philosophy and the Sciences in the Islamic East,” in Arabic Theology, Arabic Philosophy from the Many to the One: Essays in Celebration of Richard M. Frank, ed. James. E. Montgomery (Leuven: Peeters, 2006), 371–422.

39

See Jon McGinnis, “Scientific Methodologies in Medieval Islam,” Journal of the History of Philosophy 41 (2003): 307–327; Riccardo Strobino, “Avicenna on the Indemonstrability of Definition,” Documenti e Studi Sulla Tradizione Filosofica Medievale 21 (2010): 113–63. Strobino notes (Strobino, “Avicenna on the Indemonstrability of Definition,” 145) that Dem. 1.9 corresponds to APo. 2.19. This is true, but there is important material near the end of the chapter (Avicenna, Burhān, 97, ll. 1–15) that relates to the topic of accidental predication in APo. 1.5.

40

Abū al-Faraǧ Ibn al-Quff, Kitāb Taysīr al-wuṣūl ilā tafsīr al-fuṣūl li-Abuqrāt, ed. ARABCOMMAPH (The University of Manchester, 2012–2017), http://dx.doi.org/10.3927/52131919; MS, Gotha, Landesbibliothek, 1894, fols. 9a, l. 10–10a, l. 6.

41

Avicenna, Kitāb al-Naǧāt fī al-ḥikma al-manṭiqīya wa-l-ṭabīʿīya wa-l-ilāhīya, ed. Majid Fakhry (Beirut: Dār al-āfāq al-ǧadīda, 1982), 93. For other translations of this passage, see McGinnis, “Scientific Methodologies,” 309; and Asad Q. Ahmed, Avicenna Deliverance: Logic (Karachi: Oxford University Press, 2011), 83–4.

42

John P. McCaskey, “Freeing Aristotelian Epagōgē from Prior Analytics II 23,” Apeiron 40 (2007): 345–74, 358.

43

This portion of the text is translated in McGinnis, “Scientific Methodologies,” 317.

44

Avicenna, al-Šifā al-manṭiq al-burhān, ed. Abū al-ʿAlāʾ al-ʿAfīfī (Cairo: al-Maṭbaʿa al-amīrīya bi-l-Qāhira, 1956), 95.

45

McGinnis, “Scientific Methodologies,” 321 reconstructs the syllogism in a similar way.

46

Gutas, “Avicenna’s Empiricism,” 400. I think Gutas is right to make much more of syllogism’s role in Avicenna’s theory of experience. It is certainly not a “throwaway line,” brief as it is (ibid., 322).

47

“Ibn Sînâ’s attack on Aristotelian induction offers perhaps the most rigorous and technical critique of induction, as later ancient and medieval natural philosophers understood induction, until the modern period.” McGinnis, “Scientific Methodologies,” 308.

48

It is likely that Faḫr al-Dīn al-Rāzī played an important role in the process of bringing Avicenna’s philosophical works to bear on his medicine by means of the lemmatic commentary in the twelfth and thirteenth centuries. For example, see Endress, “Reading Avicenna in the Madrasa,” 383–92. I make this argument in an article in preparation: “The Emergence of Verification (taḥqīq) in Islamic Medicine: The Exegetical Legacy of Faḫr al-Dīn ar-Rāzī’s (d. 1210) Commentary on Avicenna’s (d. 1037) Canon of Medicine.”

49

This is the interpretation of this chapter offered by McCaskey. McCaskey, “Freeing Aristotelian Epagōgē.”

50

Isḥāq b. Ḥunayn supplies the explicative translation “ġayr ḏawāt awsāṭ,” lit. “they do not possess middle terms.” See Badawī, Manṭiq Arisṭū, 2:482, l. 21.

51

For a clear statement of the many challenges posed by what Bayer calls this “tortuous last chapter” of Aristotle Posterior Analytics, see Greg Bayer, “Coming to Know Principles in Posterior Analytics II.19,” Apeiron 30 (1997): 109–42.

52

Greek text: Ross, Aristotle’s Analytics, 99b40–100a8. Arabic text: Badawī, Manṭiq Arisṭū, 2:483–4. Translation is based on the Arabic.

53

Compare with Aristotle, Prior Analytics, trans. Robin Smith (Indianapolis: Hackett, 1989), 99–100.

54

Bayer states the difficulty as follows. “Aristotle the empiricist seems to be insisting on sense experience as the instrument for acquiring knowledge of the principles, while Aristotle the rationalist is positing a mental state even ‘truer’ and ‘more accurate’ than scientific understanding, let alone sense experience, for actually possessing them.” Bayer, “Coming to Know Principles,” 109.

55

Regarding the authorship of this text, Owen Goldin judges that it is “largely a paraphrastic condensation of either a lost commentary on An. Post. 2 by Philoponus, or another commentary on this book that derives from the lectures of Ammonius.” Owen Goldin, Philoponus(?): On Aristotle’s Posterior Analytics 2 (London: Duckworth, 2009), 4. Richard Sorabji argues that there is reason to believe that the text is by Philoponus; see Sorabji’s preface to Goldin, Philoponus(?): On Aristotle’s Posterior Analytics 2, vii–ix.

56

Philoponus, In Aristotelis Analytica Posteriora commentaria cum anonymo in librum II (CAG 13.3), ed. Maximillian Wallies (Berlin: Akademie der Wissenschaften, 1882), 435, ll. 18–27.

57

Avicenna, Burhān, 95–6.

58

McGinnis, “Scientific Methodologies,” 322–4. Based on James Lennox’s work on the links between Aristotle’s Posterior Analytics and Aristotle’s biology, it is clear too that in Dem. 1.9, Avicenna sees qualified experience as relevant to Aristotle’s discussion of essential and accidental predication in APo. 1.5; see in particular, Avicenna, Burhān, 97. The links between Avicenna’s Posterior Analytics and the biology of the Healing and its relationship to Aristotle’s philosophy of biology require a separate study; see James Lennox, Aristotle’s Philosophy of Biology (Cambridge: Cambridge University Press, 2000).

  • 1

    Franz Rosenthal, “ ‘Life is Short, the Art is Long’: Arabic Commentaries on the First Hippocratic Aphorism,” Bulletin of the History of Medicine 40 (1966): 226–45; Jules Janssens, “ ‘Experience’ (tajriba) in Classical Arabic Philosophy (al-Fārābī–Avicenna),” Quaestio 4 (2004): 45–62; Jon McGinnis, “Avicenna’s Naturalized Epistemology and Scientific Methods,” in The Unity of Science in the Arabic Tradition: Science, Logic, Epistemology and Their Interactions, eds. Shahid Rahman, Hassan Tahiri and Tony Street (Dordrecht: Kluwer Academic Publishers, 2008), 129–52; Miquel Forcada, “Ibn Bājja on Medicine and Medical Experience,” Arabic Sciences and Philosophy 21 (2011): 111–48.

    • Search Google Scholar
    • Export Citation
  • 2

    Philip van der Eijk, “Galen’s Use of the Concept of ‘Qualified Experience’ in his Dietetic and Pharmacological Works,” in Medicine and Philosophy in Classical Antiquity: Doctors and Philosophers on Nature, Soul, Health and Disease, ed. Philip van der Eijk (Cambridge: Cambridge University Press, 2005), 279–98, 289.

    • Search Google Scholar
    • Export Citation
  • 4

    Ibid., 290.

  • 6

    van der Eijk, “Galen’s Use of Qualified Experience,” 287.

  • 7

    Peter E. Pormann, “Avicenna on Medical Practice, Epistemology, and the Physiology of the Inner Senses,” in Interpreting Avicenna: Critical Essays, ed. Peter Adamson (Cambridge: Cambridge University Press, 2013), 91–108, 98–9.

    • Search Google Scholar
    • Export Citation
  • 9

    See Pormann, “Avicenna on Medical Practice, Epistemology, and Physiology,” 98.

  • 12

    See John N. Martin, “Proclus and the Neoplatonic Syllogistic,” Journal of Philosophical Logic 30 (2001): 187–240.

  • 19

    Alfarabi, On Topical Analysis, 106.

  • 21

    van der Eijk, “Galen’s Use of Qualified Experience,” 282–3.

  • 24

    Ullmann, Medizin, 171.

  • 39

    See Jon McGinnis, “Scientific Methodologies in Medieval Islam,” Journal of the History of Philosophy 41 (2003): 307–327; Riccardo Strobino, “Avicenna on the Indemonstrability of Definition,” Documenti e Studi Sulla Tradizione Filosofica Medievale 21 (2010): 113–63. Strobino notes (Strobino, “Avicenna on the Indemonstrability of Definition,” 145) that Dem. 1.9 corresponds to APo. 2.19. This is true, but there is important material near the end of the chapter (Avicenna, Burhān, 97, ll. 1–15) that relates to the topic of accidental predication in APo. 1.5.

    • Search Google Scholar
    • Export Citation
  • 46

    Gutas, “Avicenna’s Empiricism,” 400. I think Gutas is right to make much more of syllogism’s role in Avicenna’s theory of experience. It is certainly not a “throwaway line,” brief as it is (ibid., 322).

    • Search Google Scholar
    • Export Citation
  • 57

    Avicenna, Burhān, 95–6.

  • 58

    McGinnis, “Scientific Methodologies,” 322–4. Based on James Lennox’s work on the links between Aristotle’s Posterior Analytics and Aristotle’s biology, it is clear too that in Dem. 1.9, Avicenna sees qualified experience as relevant to Aristotle’s discussion of essential and accidental predication in APo. 1.5; see in particular, Avicenna, Burhān, 97. The links between Avicenna’s Posterior Analytics and the biology of the Healing and its relationship to Aristotle’s philosophy of biology require a separate study; see James Lennox, Aristotle’s Philosophy of Biology (Cambridge: Cambridge University Press, 2000).

    • Search Google Scholar
    • Export Citation
  • 1

    Franz Rosenthal, “ ‘Life is Short, the Art is Long’: Arabic Commentaries on the First Hippocratic Aphorism,” Bulletin of the History of Medicine 40 (1966): 226–45; Jules Janssens, “ ‘Experience’ (tajriba) in Classical Arabic Philosophy (al-Fārābī–Avicenna),” Quaestio 4 (2004): 45–62; Jon McGinnis, “Avicenna’s Naturalized Epistemology and Scientific Methods,” in The Unity of Science in the Arabic Tradition: Science, Logic, Epistemology and Their Interactions, eds. Shahid Rahman, Hassan Tahiri and Tony Street (Dordrecht: Kluwer Academic Publishers, 2008), 129–52; Miquel Forcada, “Ibn Bājja on Medicine and Medical Experience,” Arabic Sciences and Philosophy 21 (2011): 111–48.

    • Search Google Scholar
    • Export Citation
  • 2

    Philip van der Eijk, “Galen’s Use of the Concept of ‘Qualified Experience’ in his Dietetic and Pharmacological Works,” in Medicine and Philosophy in Classical Antiquity: Doctors and Philosophers on Nature, Soul, Health and Disease, ed. Philip van der Eijk (Cambridge: Cambridge University Press, 2005), 279–98, 289.

    • Search Google Scholar
    • Export Citation
  • 4

    Ibid., 290.

  • 6

    van der Eijk, “Galen’s Use of Qualified Experience,” 287.

  • 7

    Peter E. Pormann, “Avicenna on Medical Practice, Epistemology, and the Physiology of the Inner Senses,” in Interpreting Avicenna: Critical Essays, ed. Peter Adamson (Cambridge: Cambridge University Press, 2013), 91–108, 98–9.

    • Search Google Scholar
    • Export Citation
  • 9

    See Pormann, “Avicenna on Medical Practice, Epistemology, and Physiology,” 98.

  • 12

    See John N. Martin, “Proclus and the Neoplatonic Syllogistic,” Journal of Philosophical Logic 30 (2001): 187–240.

  • 19

    Alfarabi, On Topical Analysis, 106.

  • 21

    van der Eijk, “Galen’s Use of Qualified Experience,” 282–3.

  • 24

    Ullmann, Medizin, 171.

  • 39

    See Jon McGinnis, “Scientific Methodologies in Medieval Islam,” Journal of the History of Philosophy 41 (2003): 307–327; Riccardo Strobino, “Avicenna on the Indemonstrability of Definition,” Documenti e Studi Sulla Tradizione Filosofica Medievale 21 (2010): 113–63. Strobino notes (Strobino, “Avicenna on the Indemonstrability of Definition,” 145) that Dem. 1.9 corresponds to APo. 2.19. This is true, but there is important material near the end of the chapter (Avicenna, Burhān, 97, ll. 1–15) that relates to the topic of accidental predication in APo. 1.5.

    • Search Google Scholar
    • Export Citation
  • 46

    Gutas, “Avicenna’s Empiricism,” 400. I think Gutas is right to make much more of syllogism’s role in Avicenna’s theory of experience. It is certainly not a “throwaway line,” brief as it is (ibid., 322).

    • Search Google Scholar
    • Export Citation
  • 57

    Avicenna, Burhān, 95–6.

  • 58

    McGinnis, “Scientific Methodologies,” 322–4. Based on James Lennox’s work on the links between Aristotle’s Posterior Analytics and Aristotle’s biology, it is clear too that in Dem. 1.9, Avicenna sees qualified experience as relevant to Aristotle’s discussion of essential and accidental predication in APo. 1.5; see in particular, Avicenna, Burhān, 97. The links between Avicenna’s Posterior Analytics and the biology of the Healing and its relationship to Aristotle’s philosophy of biology require a separate study; see James Lennox, Aristotle’s Philosophy of Biology (Cambridge: Cambridge University Press, 2000).

    • Search Google Scholar
    • Export Citation

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 526 0 0
Full Text Views 758 294 19
PDF Views & Downloads 286 206 19