1. About the Authorities of the Ancients

117. About the second principal article, Aristotle seems to say certain things in the book [On Memory 1.449b18-21] from which it follows that memory is in the intellective part. For he says that we remember certain intelligibilities, as that a triangle has three angles equal to two right angles etc. “because we have learnt and considered them.”

118. A response is given [Henry of Ghent, Quodlibet 6 q.8] according to Aristotle’s own remark, for later in the same place [ibid. 1.449a12-13] he says “memory of intelligibilities is not without a phantasm.”

119. On the contrary: not for this reason must memory be denied to be in the intellect, just as understanding is not denied to be in the intellect and yet, according to his opinion there, we do not understand without a phantasm.

120. Another response [Henry, ibid. q.8] is that we remember intelligibilities per accidens; hence Aristotle says there [On Memory 1.450a12-13], when speaking of intelligibilities, that intellection will be per accidens. And Damascene (as cited before, n.76), “we remember intelligibilities just as we learn them, but we do not have memory of the substance of them.”

121. On the contrary: any power that knows an act as the act is of an object, in some way knows the object; but this object ‘a triangle has three angles equal to two right angles’ as it is a demonstrated and known truth can only be known by the intellect, such that no sense is similarly able to know this act ‘I have considered the fact that a triangle has etc.’

122. Again the Philosopher concedes there [On Memory 1.450a16-18, 2.453a8-10] that recollection is present only in man (and Avicenna maintains this above [nn.8, 10; On the Soul p.4 ch.3;]), because there is a sort of syllogizing in it. From this there is a twofold argument. First, that the knowledge proper to man himself seems to belong to the intellect itself; second, more efficaciously, that knowledge through syllogistic discourse pertains to the intellect alone; of this sort is recollection, for recollection proceeds discursively from certain known things to what has in some way fallen away, which it wants to recover the memory of. And although, because syllogism is always from premises to conclusion, there is no syllogism there (for recollection proceeds from contraries or similars, or from something that has, in its being sensed, an ordering toward what we are looking for), nevertheless neither can such conferring belong to the sense power, as it seems; rather the discursive process and the cognition that terminates it belong to the same power, and recollection terminates this discursive process; therefore etc.

123. An objection against this reasoning [Henry of Ghent, Summa a.4 q.4] is as follows:

The Philosopher [On Memory 1.449b6-8, 450b5-12; Book of Six Principles 4 nn.46-47] says that some are good at recollection and others bad, because of diversity in the organ; therefore he attributes recollection to the sense part.

124. Again [Henry, ibid. a.1 q.10] an argument that recollection is impossible is taken from Themistius [On Posterior Analytics 1 ch.1] about a fugitive slave, whereby it is proved that it is impossible to learn anything, because either it was something already known and so it is not learnt, or it was not and so, if it occurs to the intellect, the intellect does not know it to be what it is looking for.

125. This argument there indeed [n.124] lacks evidence, because whatever is necessarily inferred from necessary premises is known by this very fact; nor is it necessary for me to know [sc. first] what I [sc. later] acquired knowledge of, or not necessary for me to know it save in general, because I sought to know whatever I could infer from things I knew.

126. But in the issue at hand there is a difficulty. For the argument goes as follows: has he [the one recollecting] completely forgotten the thing that a is [sc. the thing he is looking for] or not? If he has then, if he could through recollection get back to the memory of it, he does not know it to be what he sought the memory of, and consequently he does not recollect it; because in recollecting he remembers it anew, as a thing having been remembered before and forgotten in the meantime. If he has not completely forgotten a, then he cannot recollect a.

127. The first member of this argument is confirmed by Avicenna from before [n.122]: the desire to remember in particular belongs to no brute, “for if brutes do not remember, neither do they desire to remember.” Likewise the Philosopher [On Memory 1.450a27-30] seems to posit that memory belongs to the imaginative part, “the habit of which,” he says, “we assert to be memory.”a And Damascene, as above [n.76] says, “Memory is imagination left behind by actualized sense.”

a.a [Interpolation] namely, it is manifest that memory is a part of the soul: when and of what there is imagination, of that there is also memory.

128. For the understanding of these authorities [n.127] I say (as was said before [n.94]) that the act of memory has a double object, namely proximate and remote. Now past-ness is sometimes required in each object as it is object, and sometimes in one of them only.

129. Because the senses do not know their object according to any condition save the one they have when they are sensing, according to Metaphysics 7.10.1036a6-7, “when sensible objects are away from the senses, it is not clear whether these objects are or are not,” and so they cannot have memory of their past act as past without also having memory of the sensed object as past, because they have memory of it only in the way it was as sensed when the act of sensing remained.

130. Now the intellect does not require past-ness in each object but only in the proximate one. For because its act can be of something as that something is necessary, as considering ‘a triangle has three angles equal to two right angles’, so its act of remembering this act of considering can have a remote object, not as past, but as always being the same way.

131. As concerns the condition, then, of ‘regarding a past object as past’ [n.88], that is, both remote and proximate object, such that both are known precisely and necessarily as past - memory as concerns this condition belongs necessarily to the sense part and not necessarily to the intellective part, although it could belong to the intellective part, as will immediately be said in the solution [n.136ff.].

132. Another thing [sc. to understand,[n.128] is that memory in the sense part is enough for operation without the intellect, as is plain in brutes; but, conversely, the intellective memory is not enough for operation without the sense memory, just as we cannot understand without a phantasm. And therefore Aristotle [Metaphysics 7.10.1036a6-7] would not say that a man is good or bad at remembering because his intellect is good or bad at conserving the species of something previously understood, but because his sense memory (which goes along also with the intellective memory for intellective remembering) is good or bad at retention. For perhaps any intellective memory always conserves the species, but it has not the power for act because the species has been destroyed from sense memory, without which the intellective memory is not enough for operation.

133. Proof of this:

First because what is received seems to be in the receiver according to the manner of the receiver, and consequently, since the intellect is an immaterial power and not changeable by these bodily undergoings [sc. of the senses], it does not seem that its species remains indelibly. For this reason, therefore, Aristotle [n.129] assigns a falling away of species in the sense memory only, because the sense organ is affected or moved in this way or that.

134. The same is proved secondly because, when someone remembers, he must have something remaining in himself through which he knows that thing to be what he first remembered and later forgot (in the way the argument about the fugitive slave proceeded [n.124]). But this something that remains cannot be placed in the sense part, because it has been destroyed, at least it does not remain perfectly or sufficiently for an act of remembering; therefore it is probable that it is the species remaining perfectly in the intellect. And thus when the species that somehow fell away has been recovered in the sense memory then, by collation of it with the intelligible species that remains, this ‘remembered object’ is known to be that which was known in memory before.

135. So therefore, as concerns primacy or radicality or sufficiency in itself for acting, memory is not in the intellective part but the sense part, even in our case.