Published online by Cambridge University Press: 17 August 2017
The twelfth century philosopher-theologian Fakhr al-Dīn al-Rāzī is well known for his critique of Avicennan metaphysics. In this paper, I examine his critique of Avicenna's physics, and in particular his rejection of the Avicennan and Aristotelian theory of place as the inner boundary of a containing body. Instead, Fakhr al-Dīn defends a definition of place as self-subsisting extension, an idea explicitly rejected by Aristotle and Avicenna after him. Especially in his late work, the Maṭālib, Fakhr al-Dīn explores a number of important philosophical issues with reference to this theory of place, including the principle that two indiscernible things (in this case two overlapping extensions) must be identical and the idea that motion and rest are always relative.
Le philosophe-théologien du XIIe siècle Fakhr al-Dīn al-Rāzī est bien connu pour sa critique de la métaphysique avicennienne. Nous examinons, dans cet article, sa critique de la physique d'Avicenne, et en particulier son rejet de la théorie avicennienne et aristotélicienne du lieu – lieu comme limite interne du corps englobant. Alternativement, Fakhr al-Dīn défend une théorie du lieu comme extension auto-subsistante, doctrine explicitement révoquée par Aristote et Avicenne à sa suite. En particulier dans son œuvre tardive, les Maṭālib, Fakhr al-Dīn explore plusieurs questions philosophiques importantes en rapport avec cette théorie du lieu, dont: le principe voulant que deux choses indiscernables (en l'occurrence deux extensions superposables) soient identiques, et l'idée selon laquelle mouvement et repos sont toujours relatifs.
1 al-Rāzī, Fakhr al-Dīn, al-Tafsīr al-Kabīr, ed. ʿImrān, S., 16 vols (Cairo, 2012), vol. 1, p. 205 Google Scholar. My thanks to Andreas Lammer for this reference and extensive discussion of the issues discussed in this paper, and also to Marwan Rashed for very useful comments.
2 On Aristotle's conception of place see Algra, K., Concepts of Space in Greek Thought (Leiden, 1994)Google Scholar; Hussey, E., Aristotle's Physics Books III and IV (Oxford, 1983)Google Scholar; Lang, H., The Order of Nature in Aristotle's Physics: Place and the Elements (Cambridge, 1998)CrossRefGoogle Scholar; Morison, B., On Location: Aristotle's Conception of Place (Oxford, 2002)CrossRefGoogle Scholar; Sorabji, R., Matter, Space and Motion (London, 1988)Google Scholar; Zekl, H. G., Topos: die aristotelische Lehre vom Raum: eine Interpretation von Physik, D 1–5 (Hamburg, 1990)Google Scholar.
3 Al-Tafsīr al-Kabīr, pp. 255–6.
4 al-Rāzī, Fakhr al-Dīn, al-Mabāḥith al-mashriqiyya fī ʿilm al-ilāhiyyāt wa-al-ṭabīʿiyyāt, ed. al-Baghdādī, M. M., 2 vols (Beirut, 1990)Google Scholar.
5 See P. Adamson and A. Lammer, “Fakhr al-Dīn al-Rāzī’s Platonist account of the essence of time,” forthcoming.
6 al-Rāzī, Fakhr al-Dīn, al-Maṭālib al-ʿāliya min al-ʿilm al-ilāhī, ed. al-Saqqā, A. Ḥ., 9 vols (Beirut, 1987)Google Scholar. All references to the Maṭālib will be to vol. 5.
7 For this contrast regarding place, see Aristotle, Physics, 209a29–30; Avicenna, The Physics of The Healing, trans. McGinnis, J. (Provo, 2009)Google Scholar, §2.5.1.
8 See Maṭālib 47, with endorsement of the theory at 49. For discussion see P. Adamson, “The existence of time in Fakhr al-Dīn al-Rāzī’s Maṭālib al-ʿāliya,” forthcoming.
9 For this see Adamson and Lammer, “Fakhr al-Dīn al-Rāzī’s Platonist account.”
10 A briefer discussion along the same lines, but with a more aporetic conclusion, can be found in his Mulakhkhaṣ fī al-ḥikma. For this text I have used MS Berlin Staatsbibliothek, or. oct. 623; the discussion of place is at fols 89v–92v.
11 For his views on the related issue of atomism, see Setia, ʿA., “Atomism versus hylomorphism in the Kalām of al-Fakhr [sic] al-Dīn al-Rāzī: a preliminary survey of the Maṭālib al-ʿāliyyah ,” Islam and Science, 4 (2006): 113–40Google Scholar.
12 For a similar list of criteria posed in atomic terms, see the quotation from Ibn Mattawayh given by Dhanani, A., The Physical Theory of Kalām: Atoms, Space, and Void in Basrian Muʿtazilī Cosmology (Leiden, 1994), p. 62 Google Scholar. Avicenna's list is also echoed by Suhrawardī; see Walbridge, J., “Illuminationists, place, and the void,” in Biard, J. and Rommevaux, S. (eds), La nature et le vide dans la physique médiévale: Études dédiées à Edward Grant (Turnhout, 2012), pp. 119–35, at 130CrossRefGoogle Scholar.
13 As already mentioned there is a further disagreement as to whether space must always be occupied, that is, as to whether void is possible.
14 This underestimates the extent of disagreement over place, space and void within kalām. See Dhanani, The Physical Theory, pp. 62–71.
15 See Mabāḥith 327–62. For instance faṣl 16 which establishes the existence of place is based on Healing: Physics §2.5.1–7, faṣl 17 offers five possible views and endorses the “inner surface” theory, following closely Healing: Physics, §2.6.2, etc.
16 Healing: Physics, §2.5.8, used by Fakhr al-Dīn at Mabāḥith 332 and Maṭālib 116–17, §2.1.6.
17 Fakhr al-Dīn's Mulakhkhaṣ also offers a skeptical argument to the effect that place can be neither substance nor accident (89v–90r). He replies that “the existence of place is known necessarily, for we know necessarily that the moving thing goes from one spot and location to another” (90r, lines 17–18).
18 See Adamson, “The existence of time.”
19 For an argument that Aristotle has Plato largely right here, see Morison, On Location, pp. 116–18. He points out that it may not be helpful to understand Plato's theory in terms of, or translate khôra as, “space” (p. 121), but doing so here is justified in order to bring out the link to Fakhr al-Dīn's position.
20 Translated in Adamson, P. and Pormann, P. E. (trans.), The Philosophical Works of al-Kindī (Karachi, 2012), §22Google Scholar.
21 This is somewhat different in the Mulakhkhaṣ though he there too seeks to defend Plato's view (90v, lines 2–5). He reports that for Plato, “the place of a body is its matter,” but by this Plato meant only that place “has in common with [matter] that placed things successively arrive (tawārid) in them.”
22 I here pass over the contribution of the Illuminationists, since this has been discussed by J. Walbridge, “Illuminationists, place, and the void.” Especially the treatment of Quṭb al-Dīn al-Shirāzī should be compared to that of Fakhr al-Dīn; for instance Walbridge mentions a similar classification of views on place at p. 128. One small caveat to this useful study: at p. 123, Walbridge suggests that faḍāʾ (the modern Arabic word for “space”) was apparently not used as a technical term in this period, but in fact Fakhr al-Dīn frequently uses this word to refer to place when understood as space.
23 al-Rāzī, Abū Bakr, Rasāʾil falsafiyya (Philosophical Works), ed. Kraus, P. (Cairo, 1939), p. 307 Google Scholar.
24 On this feature of Abū Bakr's view see also Adamson, P., “Galen and Abū Bakr al-Rāzī on time,” in Hansberger, R. and Burnett, C. (eds), Medieval Arabic Thought: Essays in Honour of Fritz Zimmermann (London, 2012), pp. 1–14 Google Scholar.
25 al-al-Baghdādī, Abū al-Barakāt, al-Kitāb al-Muʿtabar, ed. Musāwī, Z., 3 vols (Jabīl, 2012), vol. 2, pp. 40–1Google Scholar. See further Pines, S., Studies in Abu'l-Barakāt al-Baghdādī: Physics and Metaphysics (Jerusalem, 1979), pp. 105–6Google Scholar. See Muʿtabar vol. 2, p. 39 for a similar claim of time's epistemic immediacy.
26 Muʿtabar, vol. 2, p. 44.
27 Muʿtabar, vol. 2, p. 43. For a discussion of immersive vs. non-immersive place in Latin medieval philosophy at about the same time, see Trifogli, C., Oxford Physics in the Thirteenth Century (ca. 1250–70): Motion, Infinity, Place and Time (Leiden, 2000), pp. 141–64Google Scholar.
28 Indeed Fakhr al-Dīn invokes replacement as an argument in favor of (A2) at Maṭālib 145, §4.1.5.
29 For the example see Aristotle, Physics, 4.2, 209a35 (en têi gêi), Abū al-Barakāt, Muʿtabar, vol. 2, p. 1, and for the pre-philosophical intuition that place is the surface something sits on, Avicenna, Healing: Physics, §2.6.1 and 16.
30 As Morison, On Location, p. 97 points out, this is how Aristotle could meet the challenge of Zeno's regress argument of places in places. A limit is somewhere, that is, in the body whose limit it is, but a limit has no place. Avicenna also denies that surfaces have place: Healing: Physics, §2.9.6.
31 The exception is §4.2.2, which tries to establish that the extension inside a vessel must be distinct from the extension of the body contained by the vessel; but this reprises arguments already familiar from earlier sections.
32 Here one might compare the approach propounded by Alvin Plantinga in philosophy of religion. Despite asserting that belief in God's existence is “properly basic” and so can be rationally accepted with no further warrant, Plantinga still takes himself to be obligated to defuse considerations offered against God's existence, such as the problem of evil.
33 The only candidate for Y considered by Fakhr al-Dīn is the contained body, so having ruled this out he infers that place is self-subsistent. One could block this move by introducing a further candidate Z. For instance if, as he said in his Koran commentary, God may have created place one might suppose that place continues to subsist through God's power rather than itself. However, Fakhr al-Dīn seems happy to use the phrase “self-subsistent” to refer to other created things like physical substances. So whatever caveats need to be offered here concerning God's continued relation to the world, we can at least say that space has the same degree of metaphysical independence as the bodies that occupy space.
34 This thought experiment is similar to one proposed by Abū Bakr al-Rāzī, in which we should suppose that there are no bodies at all, and conclude that there would still be (empty) place. See Abū Bakr al-Rāzī, Rasāʾil, p. 199.
35 See e.g. Marmura, M. E., “Avicenna's ‘Flying Man’ in context,” Monist, 69 (1986): 383–95Google Scholar; Black, D. L., “Avicenna on self-awareness and knowing that one knows,” in Rahman, S. et al. (eds), The Unity of Science in the Arabic Tradition (Dordrecht, 2008), pp. 63–87 CrossRefGoogle Scholar; Hasse, D. N., Avicenna's De Anima in the Latin West (London, 2000), pp. 80–92 Google Scholar; Kaukua, J., Self-Awareness in Islamic Philosophy: Avicenna and Beyond (Cambridge, 2015)CrossRefGoogle Scholar.
36 Fakhr al-Dīn reaches a different, and more ambiguous, conclusion in the Mulakhkhaṣ. There, after surveying the arguments for the extension and surface theories, he asserts that place must be really, and not just mentally, existent. However he leaves it open whether the Platonic or Aristotelian theory is to be preferred: “it is necessarily known that everything located has a spot and location. That this location which we posit for it is something mental and postulated (amran dhihniyyan taqdīriyyan) is impossible. For how can body be present in something that someone merely posits or supposes – such that if the positing didn't exist, then that which is posited would have no reality – given that the body is [in fact] present in the location? But if we do make it something real, then it is either something particular (mushār ʿalayhi) or not. The latter is problematic too, because we do indicate (nashīru) this location, and how could we do this with something that is not particular? The former, then, must be either dimension or surface, or neither of these. We have laid out everything one can say about these two possibilities, so it is for you to make up your own mind” (fol. 92v, lines 12–19).
37 For more on this argument and its antecedents see McGinnis, J., “A penetrating question in the history of ideas: space, dimensionality and interpenetration in the thought of Avicenna,” Arabic Sciences and Philosophy, 16 (2006): 47–69 CrossRefGoogle Scholar. As McGinnis notes (at p. 51) the argument has its roots in Aristotle, Phys. 4.8, 216b13–15.
38 A surprising later echo of this question comes in the context of Christian debates over the Eucharist in the Latin sphere. We find Giles of Rome for instance insisting that extensions as such are capable of overlap or interpenetration. See Adams, M. McCord, Some Later Medieval Theories of the Eucharist (Oxford, 2010), p. 108 CrossRefGoogle Scholar.
39 The principle is also stated by Avicenna himself; see McGinnis, “A penetrating question,” p. 58.
40 Note that this case assumes a univocal account of the existence of God and creatures. For this issue see R. Wisnovsky, “Essence and existence in the eleventh- and twelfth-century Islamic East (Mašriq): a sketch,” and Eichner, H., “Essence and existence: thirteenth century perspectives in Arabic-Islamic philosophy and theology,” both in Hasse, D. N. and Bertolacci, A. (eds), The Arabic, Hebrew and Latin Reception of Avicenna's Metaphysics (Berlin, 2012), at pp. 27–50 and 123–51Google Scholar.
42 Reading nufūdh.
43 See also Abū al-Barākāt, al-Muʿtabar, vol. 2, p. 43.
44 In fact, Avicenna is not proposing that we define rest and motion relativistically, since he sees these as accidents that inherently belong to a given body. But as we will see Fakhr al-Dīn latches onto this suggestion as Avicenna's primary response.
45 This assumes that the rotation of the sphere doesn't disqualify it as an unmoving place.
46 Avicenna doesn't use this word when discussing the bird and stone problem, but does use it elsewhere, e.g. at Healing: Physics, §3.13.6.
47 My thanks to Andreas Lammer for pointing out this likely response.
48 As Morison, On Location points out, ultimately we can guarantee that everything has an unmoving place, because the universe as a whole is a resting “reference point” (159, 169). Incidentally, one should probably see Avicenna's bird and stone cases as improved variants of Aristotle's boat example: the bird is completely surrounded by a moving medium, whereas Aristotle's boat is in contact with moving water only where the hull is immersed.
49 One might worry about the case of a rotating sphere, which moves but does not change in respect of space. Fakhr al-Dīn doesn't address this, but the answer would presumably be that its parts do change in the requisite manner, even if the whole sphere does not. Another, more problematic issue is whether it is really meaningful to speak of moving within infinite space. I take Fakhr al-Dīn's point to be that regions of space are primitively located – as we already saw, they are not individuated by anything further – and it is relative to these that bodies move or fail to move (of course, spatial regions themselves never depart from their location).
50 For an edition and French translation see Rashed, R., Les mathématiques infinitésimales du IXe au XIe siècle, vol. IV: Ibn al-Haytham. Méthodes géométriques, transformations ponctuelles et philosophie des mathématiques (London, 2002), pp. 666–85Google Scholar. My thanks to Marwan Rashed for drawing my attention to this important source.
51 Or the geometrical thought experiments in Ibn al-Haytham see Rashed, Les mathématiques infinitésimales, pp. 671–3, and 955–6 for Fakhr al-Dīn's quotation in the Mulakhkhaṣ.
52 Rashed, Les mathématiques infinitésimales, p. 669.
53 Rashed, Les mathématiques infinitésimales, p. 677, see also 683, 685.
54 Rashed, Les mathématiques infinitésimales, p. 658, goes so far as to say that on his account, “le lieu devient ontologiquement neutre.”
55 As reported by Simplicius, In Aristotelis physicorum libros commentaria, ed. Diels, H. (Berlin, 1882), p. 604.9Google Scholar. See on this Sorabji, Matter, Space and Motion, chap. 11.
56 These recall, but do not include, the idea of sticking out one's arm from the edge of the cosmos into nothing at all – already used by Archytas to show that the universe must be surrounded by void. See Simplicius in Phys. 467.26–32, and Sorabji, Matter, Space and Motion, p. 125. The argument also features in the earlier al-Rāzī’s On Metaphysics. See Abū Bakr al-Rāzī, Rasāʾil falsafiyya, pp. 133–4.
57 This sort of “removal” argument circulated from early on in Arabic, as it could be found in Philoponus (in Phys. 574–5); on this see McGinnis, “A penetrating question,” 53, and further Sedley, D., “Philoponus’ conception of space,” in Sorabji, R. (ed.), Philoponus and the Rejection of Aristotelian Science (London, 1987), pp. 140–53Google Scholar.
58 As noted by McGinnis, “A penetrating question,” p. 67.
59 al-Rāzī, Fakhr al-Dīn, Muḥaṣṣal afkār al-mutaqaddimīn wa-al-mutaʾakhkhirīn min al-ʿulamāʾ wa-al-ḥukamāʾ wa-al-mutakallimīn, ed. Atay, H. (Cairo, 1991), pp. 311–12Google Scholar.
60 Compare his similar remarks about the use of intuition to establish the existence of time, at Maṭālib 49.
61 Cf. Mulakhkhaṣ, fol. 90v, line 9 and following for an alternative line of argument against the extension view: independently existing extension would have to be capable of motion or not, but both are impossible (if extension could move this would lead to a regress, but extension as such cannot be incapable of motion, since bodily extension can move).
62 This mustard seed example is also used at Mulakhkhaṣ, fols 89v, 91v.
63 This argument also appears almost verbatim at Mulakhkhaṣ, fol. 91r, lines 12 and following.
64 This rationale is also given at Mulakhkhaṣ, fol. 91v, lines 7–8.