Of all Aristotle’s short biological treatises that make up the collection known as Parva Naturalia, the treatise On Memory and Recollection is the most likely one to grab the attention of modern readers. As Richard Sorabji states in the opening lines of the introduction to the second edition, the treatise contains the first sustained treatment of memory in the West, the earliest description of a powerful ancient mnemonic technique, and something that has been hailed as the ancient formulation of the laws of association of ideas. In addition, the treatise is important for Aristotle’s conception of thinking and representation, his theory of ‘common sense’, his views about the physiological processes that underlie or influence various mental activities, and also for his methodology and proneness to diagrammatic representations.
Before the publication of the first edition of Aristotle On Memory in 1972, English-speaking readers mostly relied on Beare’s somewhat paraphrastic translation (Oxford, 1908) and the commentaries of G.R.T. Ross (Cambridge, 1906) and W.D. Ross (Oxford, 1955). The need for a new, more faithful translation, and a more focused commentary, was satisfied by Sorabji’s book. It contained three introductory essays, a translation with explanatory notes, a selective bibliography, a list of departures from W.D. Ross’ text, and a useful index.
The introductory essays aimed at explaining the main concepts and questions discussed in Aristotle’s treatise. In the first introductory essay (‘Memory’), Sorabji elucidated what memory, according to Aristotle, was and how it worked. Sorabji’s central claim was that memory required mental images, that is quasi-visual mental representations. That is to say, we remember something when we consider a mental image as a copy of the thing experienced in the past on the basis of the image’s likeness to that thing. Also constitutive of memory, according to Aristotle, is what he calls ‘perception of time’, which seems to be some sort of awareness of the temporal relation in which things experienced in the past stand to the subject representing them in the present.
The second essay (‘Mnemonic Techniques’) is devoted to a mnemonic technique which seems to have been widely used in antiquity and is also employed by some modern mnemonists. The technique consists of memorizing in advance a set of places, for example a street of houses, which form background images on which one subsequently superimposes images that represent things to be memorized. When one wants to recall these things, one scans the set of places in one’s mind and, finding the superimposed images, he remembers the things represented by the images. Assuming that Aristotle had this particular technique in mind, which no doubt relied on vivid mental imagery, Sorabji was able to give a definitive elucidation of a difficult passage at 452a17-24.
In the third essay (‘Recollection’), Sorabji clarified Aristotle’s conception of ‘recollection’ ( anamnêsis), mostly by drawing comparisons with Plato’s. Very briefly, for Plato anamnêsis involves being reminded of one thing by another thing, notably being reminded of Forms apprehended by the soul before birth by experiencing particulars that instantiate them. For Aristotle, by contrast, anamnêsis is a deliberate and rational process of searching and recovering a particular piece of information — something experienced or learned before — by going through a chain of associated images. The two conceptions share two underlying ideas, first that anamnêsis is recovery of temporarily lost knowledge from within one’s soul, and second that it proceeds by association, one thing leading to another. The points of divergence are numerous and they are metaphysically and epistemologically motivated. Two glaring differences are that Plato’s anamnêsis seems to require neither a deliberate effort nor going through a series of mental images.
Therefore, Sorabji is right when he observes (p. 41) that the more natural English rendering of the Platonic anamnêsis is ‘being reminded’, and of the Aristotelian ‘recollecting’. One might support this decision by quoting Locke who says that if the same idea is “sought after by the mind, and with pain and endeavour found, and brought again in view, ’tis Recollection” ( Essay, II.19.1). This is worth mentioning because in the second edition, Sorabji seems to have changed his mind and now prefers ‘being reminded’ both for Plato’s and Aristotle’s anamnêsis. This can be inferred from the opening sentence of the introduction to the second edition, where Sorabji translates the title of Aristotle’s text as On Memory and Reminding Oneself and speaks of “mnemonic processes of reminding oneself”.
The fifty pages of notes in the first edition were devoted to elucidating particular phrases and passages, and they complemented the introductory essays. However, the introductory essays occasionally contained passages that complemented the notes too (e.g. an interpretation of the diagram at 452b15-22, explaining how we date past events is found in the first essay, pp. 18-21), so one would have been ill-advised to read one without the other. The text of the translation was conveniently broken down into smaller chunks preceded by helpful interpretative summaries. The translation was generally more faithful to the original as well as more readable than the earlier English translations.
The first edition of Sorabji’s Aristotle On Memory was a nicely produced book which deservedly achieved the status of the standard translation and commentary of Aristotle’s treatise. The second edition, published in 2004 by Duckworth in the UK and now by Chicago in the US, brings a phototype reprint of the first edition, preceded by a new introduction and a list of corrections to the text of the first edition (pp. ix-xxvi).
The introduction to the second edition touches upon some interesting issues, such as animal memory and perception of time, but mostly it is devoted to defending the central claim of Sorabji’s interpretation, namely that memory, according to Aristotle, always involves quasi-visual representations. We should not be impressed, Sorabji argues, by criticisms of mental imagery voiced by some modern philosophers, because they do not take fully into account the experience of vivid eidetic or mnemonic imagery with which Aristotle was much better acquainted than the present age and on which he relied. I would readily concede to Sorabji that quasi-visual representations are indeed paradigmatic phantasmata in Aristotle’s explanations, but I do not see why we should saddle him with the implausible view that the phantasmata operating in memory cannot be of a different, for example quasi-auditory character. I find it hard to believe, for example, that Aristotle would say that we remember a particular melody by picturing, say, an instrument on which we heard it performed. Furthermore, I assume that representations of the congenitally blind persons (mentioned by Aristotle in De Sensu 1, 437a15-17) cannot be of quasi-visual character, yet surely they have memory, and I see no reason to suppose that their memory requires any different explanation than the one given by Aristotle.
Not saddling Aristotle with this view would also remove the need to argue that the word phantasma in Aristotle invariably refers to a pictorial mental image. This is particularly difficult to argue if one wants to claim, as Sorabji does on pp. xiv-xix, that phantasia has to do with all sorts of appearances, not only visual or quasi-visual ones. The difficulty is generated by the key passage for Aristotle’s concept of phantasia ( De Anima III.3, 428a1-2), in which he says that ” phantasia is the faculty by which we say that a phantasma occurs to us.” Sorabji deals with the difficulty by trying to weaken the link between phantasia and phantasma, taking Aristotle to be saying that supplying phantasmata, in the sense of pictorial mental images, is only one of the things that phantasia does. This is not an intuitive reading of the passage, and it does not sit well with Sorabji’s later claim that ” phantasmata are not objects of phantasia, but simply a means by which it apprehends other objects” (p. xx). If phantasia apprehends all objects by means of phantasmata, then supplying phantasmata is not, after all, only one of the things that phantasia does, but rather the essential thing that it does. On the other hand, if phantasia is supposed to apprehend some objects by other means, then it is difficult to see what these means are and how the unity of the faculty of phantasia can be preserved.
Although the introduction makes about a dozen references to recent literature on Aristotle’s treatise and closely related subjects, it is regrettable that the second edition does not include an expanded and updated bibliography. It would have taken little effort to compile a bibliography to supplant the reprinted one which was exceedingly short already at the time of publication of the first edition and lamentably inadequate at the time of publication of the second.
The list of corrections improves the text of the first edition at 14 places, half of them found in the translation. These are largely flaws pointed out by the reviewers of the first edition. However, some of the mistakes pointed out have been unduly ignored. For instance, John Cooper has observed ( Archiv für Geschichte der Philosophie 57, 1975, pp. 63-69) that Sorabji’s divergence from Ross’ text at 453a20 had not been listed on page 115 (‘Textual Changes’) in the first edition, and that Sorabji illegitimately dissociated pote from pôs in his translation of 450a26. Neither of these points has been taken into account in the list of corrections.
I have noted two minor typographic errors in the introduction: an omitted comma between ‘imagination’ and ‘perception’ in the first sentence of the third paragraph on p. xiv, and a missing full-stop after the fourth sentence of the third paragraph on page xviii. Also, on p. xvii Sorabji credits Martha Nussbaum ( Aristotle’s De Motu Animalium, Princeton, 1978, p. 254) with a “clever suggestion” about the metaphorical use of the word phantasia mentioned in De Anima III.3 428a2, whereas Nussbaum herself refers this suggestion to Jacob Freudenthal’s study from 1869. On p. xxiv, note 14 contains a reference to Malcolm Schofield’s important paper ‘Aristotle on the Imagination’, but fails to add that it was reprinted in Nussbaum and Rorty’s collection Essays on Aristotle’s De Anima, which is the sort of thing duly listed in notes 3 and 16.
The first edition of Sorabji’s Aristotle On Memory has long been out of print, and the second edition will make it available again, accompanied with a list of the most important corrections and a stimulating introductory essay. It is disappointing, however, that the second edition does not bring a revised and expanded text of the first, since much work has been done on this treatise and the related issues in Aristotle’s psychology since the early 70s, largely thanks to Sorabji himself.