2. Background: Aristotle and the Arabic Peripatetics

In his brief treatise, De Memoria,Footnote ⁷ Aristotle gives an account of memory that heavily emphasizes the idea that memory is of the past. “Memory,” states Aristotle in some of the first lines of the treatise, “is of what has happened” (449b14).Footnote ⁸ And again: “It is always the case that whenever somebody remembers he says in his soul that he has heard or sensed or understood this thing before” (449b22).Footnote ⁹ For this reason, there is no memory of the present or of the future, but only of the past (449b10–20). Thus, pastness — and not only pastness, but the subjective awareness of pastness — is arguably the essential or defining characteristic of memory.

However, it is important to note that in a critical passage Aristotle links this ability to remember something as having occurred in the past closely to the ability to consider a preserved phantasm as a representation (imago) of something previously experienced (450a25–451a3). Without this ability, he notes, we could only “remember” phantasms preserved in the imagination, rather than the absent, previously experienced things of which those phantasms are likenesses. If that were so, there would not really be memory at all, since all we would be “remembering” would be images in our minds.Footnote ¹⁰ Memory, in other words, sees through the imagined phantasm to the previously sensed thing, of which the phantasm is a representation. To illustrate, Aristotle gives the example of a picture of an animal inscribed on a tablet (450b21–451a3).Footnote ¹¹ This picture can be considered in one of two ways: first, secundum se, i.e., as a picture in its own right, and second, as a representation of something else (alterius fantasma), i.e., the real animal used by the artist as a model for the picture. This second mode of considering a phantasm is the mode that is proper to memory. Aristotle, however, does not explain precisely how this shift in perspective — i.e., from considering a phantasm secundum se to viewing it as a representation of some previously experienced thing — comes about.

Although the great Arabic-language Peripatetics Avicenna and Averroes do not necessarily deny that memory involves the past,Footnote ¹² their accounts of remembering are striking in the way they de-emphasize pastness, and instead focus on providing a cognitive account of how it is possible for memory to be “of” a real, previously experienced extra-mental object, rather than merely of a phantasm. As we will see, this account was profoundly influential on Aquinas's theory of memory.

Post-Aristotelian writers famously attempted to flesh out Aristotle's thin account of such things as imagination and cogitation, ultimately postulating a number of what came to be termed “interior senses” to explain various aspects of post-sensationary cognition (see Barker, Reference Barker2012; Black, Reference Black1993, Reference Black1996; Wolfson, Reference Wolfson1935). This effort reached a high degree of sophistication in the writings of Avicenna, who postulated in his De Anima Footnote ¹³ the existence of five different interior senses: the common sense, imagination, compositive imagination, estimative power, and memory (Avicenna, Reference Riet1972, pp. 79–90). Crucially, for our purposes, Avicenna distinguished between two different genera of sensible objects: first, sensible forms, and second, “un-sensed intentiones” (intentiones non sensatas) (Avicenna, Reference Riet1972, p. 89).

In Avicenna's model, four of the five interior senses are split into two receptive/retentive pairs, with each pair responsible for receiving and retaining one of the two kinds of sensible objects. The first pair is constituted of the common sense and the imagination, which are responsible (respectively) for receiving and retaining the sensible forms conveyed by all five of the proper senses. The second pair is constituted of the estimative power and memory, which have for their object the intentiones non sensatas, with the estimative power receiving and the memory retaining these intentiones. To give an account of what, precisely, these intentiones non sensatas are would require more space than is available here. However, in general, we can say that these estimative intentiones represent various sensible aspects of objects that are nevertheless not sensed by any of the exterior senses (Avicenna, Reference Riet1972, p. 86),Footnote ¹⁴ whether because they exceed what the exterior senses are capable of apprehending, or because they are being sensed per accidens by another sense (e.g., when sight “accidentally” senses “sweet” when it apprehends the yellowness of honey (Avicenna, Reference Riet1968, p. 35).Footnote ¹⁵ Importantly, the apprehension of intentiones like these cannot be attributed to intellect, because these intentiones are fundamentally particular aspects of individual sensible objects, rather than universal features conceptualized as universal, which is why non-rational animals are also capable of apprehending them. Thus, according to Avicenna, there must be a power other than the proper senses or intellect that is responsible for apprehending these unsensed sensible features of objects, i.e., the estimative power. For Avicenna, intentiones of this kind include such things as the suitability or unsuitability of objects to the sensing agent (Avicenna's favoured example is the intentio of “hostility,” which is apprehended by the estimative power of a sheep when sensing a wolf (Avicenna, Reference Riet1972, p. 86)). At times, however — and particularly when discussing memory (Avicenna, Reference Riet1968, pp. 9–10) — Avicenna speaks of the estimative intentio as being a cognitive object that in some way represents a sensed object as the very individual object that it is, i.e., as an integral whole with a unique identity (see Black, Reference Black1993, pp. 226–227).

Averroes, in turn, accepted Avicenna's distinction between sensible forms and intentiones non sensatas. Although Averroes eventually rejected Avicenna's account of the estimative power in animals (Averroes, Reference van den Bergh1987, p. 336), he nevertheless retained the concept of unsensed intentiones, postulating that humans cognize these intentiones through an analogous power that he called the “cogitative power.” Averroes, however, significantly pared down the account of these intentiones, de-emphasizing the idea that intentiones represent relational/affective content (e.g., wolf as “hostile”), and instead singling out the idea that the intentiones somehow represent an object qua individual. As he describes it, the intentio of an object is like the kernel (medullam) of a fruit, while all the various sensible features are like the rind (cortices) (Averroes, Reference Shields and Scot1949, p. 42). The upshot, as Deborah L. Black summarizes, is that “the perception of an intention for Averroes would seem to entail the recognition of an individual precisely insofar as it is an individual” (Black, Reference Black1996, p. 169).Footnote ¹⁶

Importantly for our purposes, this idea that the intentio represents an individual qua individual plays a key role in both Avicenna's and Averroes's accounts of memory. Instead of emphasizing that memory is of the past, as Aristotle does, Avicenna and Averroes both single out the presence of the estimative or cogitative intentio as being what differentiates memory from other cognitive states, such as sensation or imagination. In both of their accounts, remembering occurs when the agent combines the correct intentio (which is stored in the memory) with the correct sensible forms (which are stored in the imagination). In his De Anima, Avicenna states that remembering requires at least three things, i.e., the sensible form, the intentio, and the putting of these two things together by the estimative power. It is only when these three things co-occur that the act of remembering occurs (Avicenna, Reference Riet1968, pp. 8–10). In his Epitome of the Parva Naturalia, Averroes states that there are four things involved in an act of remembering, i.e., the sensible image, the intention, causing that intention to be present (facere illam intentionem esse presentem), and the judging that the intention is the same intention that was previously sensed (Averroes, Reference Shields and Scot1949, pp. 51–52). Like Avicenna, then, for Averroes, remembering occurs when the correct intentio (preserved in the memory) is combined with the correct sensible form (preserved in the imagination) by the agent.

While Avicenna's and Averroes's accounts of interior sense and memory differ in some significant ways (see Black, Reference Black, Bernecker and Michaelian2017), in both cases, the intentio is a cognitive object that provides the intentional content that links sensible forms preserved in the imagination to a real, previously experienced object extra mente. In this way, both authors single out and explain Aristotle's idea that memory happens when the sensible phantasm in the imagination is apprehended as a representation of something previously experienced. Without the appropriate intentio, the phantasm in the imagination is merely a collection of sensible forms, i.e., a phantasm considered secundum se. Combined with the appropriate intentio, however, it is a representation of some previously experienced thing, and thus remembering can occur. In this way, they explain Aristotle's description of two radically different modes of considering the same phantasm, i.e., by postulating a cognitive object that represents the individual as such, and a corresponding cognitive power (estimative or cogitative) that, working in concert with memory and imagination, recombines the correct intentio with the correct sensible forms to produce an act of remembering. Although Avicenna and Averroes both typically assume that memory is of something experienced in the past,Footnote ¹⁷ their accounts of memory and remembering do not typically emphasize pastness, instead focusing on the cognitive objects that must cohere and the cognitive processes that must occur for remembering to successfully come about.

3. Aquinas's Hybrid Account

Black rightly suggests that Aquinas's account of the internal senses is essentially a hybridization of Avicenna's and Averroes's theories (Black, Reference Black2000, pp. 66–68). A detailed comparison between Aquinas's theory and that of the two great Arabic-language philosophers would require more space than is available here. However, a brief summary might go like this.Footnote ¹⁸ Aquinas agrees with Avicenna on the existence of a common sense, imagination, estimative power, and memorative power. However, he agrees with Averroes (and disagrees with Avicenna) that there is no separate compositive imagination. Those functions that Avicenna attributes to an ontologically distinct compositive imagination can instead be attributed simply to the imagination. Whereas Avicenna postulates the existence of an estimative power in animals and humans alike, Aquinas argues that the equivalent of the estimative in humans is sufficiently different in mode of operation as to merit being called by a different name, i.e., the cogitative (see De Haan, Reference De Haan2014, Reference De Haan and Băltuță2019) (in accepting the animal estimative, however, Aquinas disagrees with Averroes, who ultimately rejects this power altogether (Averroes, Reference van den Bergh1987, p. 336)). Thus, where Avicenna postulates five interior senses, Aquinas accepts only four, i.e., the common sense, imagination, estimative/cogitative, and memory.

Importantly, following both Avicenna and Averroes, Aquinas accepts the distinction between two kinds of sensible objects, i.e., sensible forms and unsensed intentiones. Like Avicenna, he argues that the interior senses are split into two receptive/retentive pairs, with the common sense and imagination receiving and retaining (respectively) sensible forms, and the estimative/cogitative power and memory receiving and retaining (respectively) the unsensed intentiones. The “hybridization” is particularly evident in Aquinas's balancing of the Averroistic emphasis on the intentio as representing an object qua individual, and the Avicennian notion of the intentio as representing various relational/affective qualities of individual objects. So, for example, in Summa Theologiae I.78.4, Aquinas, in describing the function of the estimative power, repeats Avicenna's example of the sheep sensing the wolf as “enemy,” and adds his own example, i.e., that of a bird apprehending straw as “useful-for-building a nest.” However, in his commentary on the De Anima, Aquinas takes a distinctly Averroistic line in explaining that the estimative power (in animals) or cogitative power (in humans) is responsible for apprehending an object as a this, i.e., as an integral individual. As he writes:

Similarly, when the sheep recognizes its own offspring, this is the estimative power at work.

We can summarize the upshot of Aquinas's account of estimation/cogitation by repeating a principle we have already encountered in the case of Avicenna, i.e., for Aquinas, any kind of apprehension at the level of sense with respect to an individual that cannot be explained by the proper or common senses (which are confined to apprehending only the proper and common sensibles), is an instance of an estimative/cogitative intentio. Hence, Aquinas's definition of the object of the estimative/cogitative power in negative terms, i.e., “intentiones that are not received by the senses,” (Summa, I.78.4) or, as Avicenna put it, “unsensed intentiones.”

4. The Intentio of Pastness

With this foundational background material now in place, it is time to turn our attention to the main topic of this article, i.e., the intentio of pastness. As already noted, although Aquinas is heavily indebted to Avicenna and Averroes for his account of the interior senses and the estimative/cogitative intentio, neither of these thinkers ever speak of the aspect of pastness (i.e., ratio praeteriti) as an estimative/cogitative intentio. Nor does Aquinas's teacher, Albertus Magnus, an avid disciple of the Arabic Peripatetics, especially on the matter of the interior senses. Thus, there is good reason to take note when Aquinas writes:

What Aquinas is clearly saying here is that the apprehension that something was thought or sensed in the past, which is essential to the act of remembering, is the apprehension of a certain kind of estimative/cogitative intentio. That is to say, the apprehension of pastness with respect to a specific phantasm is something that is not apprehended by the proper or common senses, but rather by the estimative/cogitative power.

Clearly, this claim holds important implications for understanding Aquinas's theories of time perception and memory. The difficulty is that Aquinas provides absolutely no further description of the intentio of pastness in the Summa. Nor does he ever speak explicitly of an intentio of pastness in any other work. However, in the rest of this article, I will argue that he does provide significant detail about this intentio of pastness in other works, particularly in a key passage in his commentary on the De Memoria, in which he takes up Aristotle's skeletal description of how the soul cognizes past time. Although Aquinas makes no mention of an intentio here, if indeed the awareness of past time involves a kind of intentio, then it seems that this passage is the primary candidate in Aquinas's corpus to provide further detail as to what sort of thing this intentio of pastness might be. Other clues are to be found in his descriptions of time and time perception in his commentary on the Physics.

Let us begin with a brief look at the commentary on the Physics. Footnote ²¹ In the Physics, Aristotle famously speculates on the nature of time, noting that time is “either motion or something of motion.”Footnote ²² However, he discards the notion that time is motion, arguing that it is “something of motion.” More specifically, he states, time is motion “insofar as it has number” with respect to a before and after (218b21–219a10).Footnote ²³ Thus, while time is not motion, it is the case that we only apprehend time when there is motion.

Aquinas accepts this account, repeatedly emphasizing that time is the measure of motion. Thus, in cases when an agent fails to apprehend motion of any sort, including the motion of the succession of his own interior cogitations or imaginations, he also loses all apprehension of time (In Phys., IV.17, n. 572).Footnote ²⁴ This is what happened in the famous case of the Sardinian sleepers, who fell into so deep and dreamless a sleep that they lost all consciousness, and thus any awareness of motion. Awakening later, they failed to realize that any time had passed (In Phys., IV.16, n. 517). Given the outline of Aquinas's theory of the cogitative power given above, however, it should strike us as significant that Aquinas argues that the apprehension of motion, which is propaedeutic to the apprehension of time, is dependent upon the continuous apprehension of a particular individual (hoc aliquid) (In Phys., V.18, n. 587) across a prior and a posterior. Thus, he notes, we can see Coriscus in the forum and Coriscus in the theatre. In both cases, Coriscus is the same in subject. And it is because we apprehend Coriscus as the same in subject that we are empowered to apprehend motion with respect to Coriscus, and thus to apprehend time, i.e., because “insofar as there is something that is the same in motion, so also it must be the same in time” (In Phys., IV.18, n. 585).Footnote ²⁵ Although, in this passage, Aquinas makes no mention of the cogitative power, we know from other passages that for Aquinas to apprehend Coriscus as Coriscus, is to apprehend an individual intentio by means of the cogitative power. Critically, then, this apprehension of Coriscus (or any object) as a discrete, continuous individual across a prior and a posterior, which would seem to be a prerequisite for the apprehension of motion, and thus for time, is one way that the cogitative (or estimative, in animals) power is involved in time perception (and, in turn, memory).

That time is a measure of motion, which measurement depends upon the agent's capacity to apprehend an individual as maintaining a stable identity across a prior and posterior, is the first key lesson we learn from the commentary on the Physics. The second idea worth highlighting from the commentary, is Aquinas's claim that we measure motion, and in turn the passage of time, by comparing two discrete “nows,” i.e., the “now” of the immediate present, and the “now” of some past moment. He writes:

The present now, in other words, is the anchoring point of all determinations of time, with past time being determined by some sort of measurement or numbering from the present “now” to some past, discrete “now.” Here, however, Aquinas does not explicitly state how or by means of which faculties the agent accomplishes this comparison of nows, and thereby apprehends time. Let us now turn to the De Memoria commentary.

In the key portion of Book VII of the De Memoria commentary that we will spend the rest of the article examining, Aquinas tackles some of the same themes as in the Physics commentary, while providing considerably more detail. Here, he is commenting on a passage in De Memoria in which Aristotle argues that the soul recognizes and judges past time through a sort of proportion or ratio, which takes the form of an image or phantasm (452b8–452b22). Aristotle's description involves a variety of points, labelled with letters of the Greek alphabet. Scholars who have grappled with this obscure passage have graphed the points as forming a triangle, which is sectioned into segments of different sizes that are, however, proportionate to one another. The details of how precisely this image is supposed to function are famously murky.Footnote ²⁷ Nevertheless, the upshot is that Aristotle supposes that the one remembering gauges how far in the past something occurred by means of a kind of phantasm that represents temporal lapses in terms of geometrical proportions. When the image of the thing to be remembered, and this temporal phantasm,Footnote ²⁸ occur together, then memory occurs (452b23).Footnote ²⁹

If the Greek manuscripts of this passage are confusing, we should not be surprised that William of Moerbeke's Latin translation is no better. Nevertheless, Aquinas provides a detailed reading that is, on some points, significantly clearer than the Latin Aristotelian original. In tackling this passage, Aquinas notes that Aristotle says that there is “something in the soul whereby it judges a greater or lesser measure of time.” This “something” involves a ratio with respect to corporeal magnitudes that represent a certain distance. It is to these magnitudes or distances that “the quantity of time is made proportionate, which is grasped with respect to the distance from the present now” (In De Mem., VII, n. 387).Footnote ³⁰

What Aquinas appears to be saying here is that, in order to grasp how far something is in the past, we must apprehend the distance of time by means of a sensible phantasm that represents a certain, proportionate measure of time from the present, which is applied to a certain imaginative form considered as a representation of something experienced in the past. The present “now,” then (as we already saw in the Physics commentary), is the anchoring point for all temporal judgements, while proportionate imagistic magnitudes represent disparate quantities of time from the present moment. Aquinas goes on to note that this is similar to the way that the soul grasps any corporeal magnitude, i.e., by means of a movement in the soul that bears a proportion to the magnitude outside the soul: “for there are in the soul certain forms and motions similar to the things, by which it knows things” (In De Mem., VII, n. 388).Footnote ³¹ What he is saying is that, when comparing the size or magnitude of any number of things (including time), the images we have in our mind are not actually the size of the things of which we are thinking; they are, rather, proportions. For instance, if one thinks about the difference in height between the CN Tower and the Empire State building, one does not have in one's imagination two images that are, in fact, 553 metres and 381 metres, respectively, but rather two images that bear the same proportion as 553 metres to 381 metres, and which represent and map onto the real-world quantities.

Then follows a difficult, but in the end fairly clear description of a triangle made up of points, which is subdivided into smaller, but proportionate triangles by subsecting the triangle at various points horizontal to the base. As Aquinas explains it, the lines formed by the original and subsected sides of the triangle represent a certain motion or magnitude in the soul, while the lines formed by the bases of the original and subsected triangles represent a certain quantity of time. The point is that, as one moves up the side of the triangle, the side and base lengths reduce in size proportionately. And so, whatever magnitude we have as a motion in the soul corresponds to a proportionate magnitude of time in the world.Footnote ³²

The ensuing descriptions of how various magnitudes are compared all riff on the same basic theme, i.e., that temporal distances are known by comparing and contrasting different magnitudes in the soul. The whole passage, however, concludes with this crucial principle, quoted nearly verbatim from Aristotle: “when in the soul there occurs at the same time a motion of the thing to be remembered and of past time, then there is an act of memory” (In De Mem., VII, n. 396).Footnote ³³

If we assume that Aquinas is being consistent across the De Memoria and Physics commentaries and the Summa, then this motion of past time — which is apprehended by means an imaginative magnitude that is proportionate to the distance from the present to a past “now,” and then applied to a particular thing being remembered — must be a case of an intentio, i.e., the intentio of pastness mentioned in Summa I.78.4. This certainly seems to me to be a reasonable hypothesis, and one that coheres perfectly well with Aquinas's overall account in the De Anima commentary of the estimative/cogitative intentio as an instance of a per accidens sensible (In De Anima, II.13, n. 385–396). That something sensed was sensed or experienced in the past is not something that either the proper or common senses can apprehend (i.e., it is not a proper or common sensible). Neither can the imagination, since the imagination only stores sensible species without intentional reference to the things of which they are likenesses. Without that intentional reference to a real, previously experienced thing, the phantasm in the imagination is simply a phantasm considered secundum se. Nor can the intellect apprehend this aspect of pastness, since the intellect abstracts from temporality altogether, in which case the only candidate left is the estimative or cogitative power. By apprehending the intentio of this sensed thing as a particular individual, the estimative/cogitative power is thus empowered to judge via an imagistic magnitude proportionate to a certain quantity of lapsed time that this phantasm stored in the imagination is a representation of something extra mente that was known or experienced a certain quantity of time in the past. This estimative/cogitative judgement that something was experienced in the past both enriches and depends upon the individual intentio (i.e., that intentio which represents the individual thing as the very individual it is), which in turn enriches the sensible forms stored in the imagination, and is an instance of an estimative/cogitative intentio, i.e., the intentio praeteriti.

Note that it would not seem that the temporal phantasm by which past time is apprehended is itself the intentio. Given Aquinas's overall account of the interior senses, the imagistic content of the phantasm here would have to be provided by the imagination. It is the imagination, after all, that preserves and uses the sensible forms that constitute the imagistic content of a sensible phantasm. The estimative or cogitative powers, however, apprehend intentiones that are not sensed by the proper senses, such as the aspect of “hostility,” or an object's very individuality (e.g., Socrates), or the aspect of pastness (ratio praeteriti). Thus, it must be the judgement that this particular thing was experienced a certain length of time in the past that is the intentio prateriti. This is something that is arrived at by means of an imagistic phantasm that represents a certain length of time, but is not reducible to an imagistic phantasm. As is the case with certain other estimative/cogitative intentiones, the intentio praeteriti represents a kind of relation,Footnote ³⁴ i.e., a relation between the now and a certain time in the past, which is then applied to an imaginative form considered as a representation of a previously experienced thing.

5. Conclusion: Uniting the Aristotelian and Arabic Peripatetic Traditions

In brief remarks in her monograph Ratio Particularis, Di Martino notes Aquinas's acceptance of the Arabic Peripatetic distinction between sensible forms and estimative/cogitative intentiones. However, she adds, Aquinas's move to identify the ratio praeteriti as itself being an estimative/cogitative intentio is ultimately a “bold but brilliant exegetical idea” by which he “recovers the Aristotelian definition that links memory to past time by making the temporal determination, the ratio praeteriti, an intentio.” In this way, he “brilliantly resolved the primary problem in the tradition, which proposes two definitions of memory: the faculty of the past, or the faculty of intentiones” (Di Martino, Reference Di Martino2008, p. 137).Footnote ³⁵

This seems basically right to me, although perhaps a touch overstated. As we have seen, it's not quite correct to say that Aristotle only links memory to pastness. He also links memory to the cognition of a phantasm not secundum se, but rather as a representation of something previously experienced. The significant innovation of the Arabic-language philosophers was to develop an account of how the agent is capable of shifting from considering a phantasm secundum se to considering it as a representation of a real, previously experienced thing, i.e., by means of an interior sense power that cognizes a particular cognitive object (i.e., the estimative or cogitative intentio), that represents the very individuality of the sensed object.

In his psychology, Aquinas, fundamentally a synthesizer, combines the best qualities of Avicenna's and Averroes's accounts of the interior senses, and in a single clause in Summa I.78.4, neatly stitches this hybrid account together with the purer Aristotelian doctrine to which he has access. Although Aristotle has no account of an estimative or cogitative power, or of the estimative/cogitative intentio as a different kind of cognitive object, Aquinas takes these Arabic innovations and reconciles them with Aristotle's account of memory in the De Memoria. Yes, memory has to do with pastness. It also, importantly, depends upon the capacity to consider a sensible phantasm as a representation. However, one and the same cognitive power, and the same kind of sensible object, explain both. The estimative/cogitative power, by apprehending the individual intentio, explains how an agent can shift from viewing a sensible phantasm in the imagination secundum se, to considering it as a representation of something previously experienced. And the estimative/cogitative power, by apprehending the intentio of pastness (which is a kind of judgement made by means of a proportionate phantasm of the kind described in the De Memoria commentary), also explains how memory involves the awareness of pastness. The important thing to note is that the latter capacity (i.e., the capacity to cognize pastness) depends and builds upon the prior capacity (i.e., the capacity to cognize an individual as an individual).