Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

1 

 

Cartesian Critters Can't Remember 

Devin Sanchez Curry 

Studies in History and Philosophy of Science 69 (2018), 72–85 

 

Abstract: Descartes held the following view of declarative memory: to remember is to 

reconstruct an idea that you intellectually recognize as a reconstruction. Descartes 

countenanced two overarching varieties of declarative memory. To have an intellectual memory 

is to intellectually reconstruct a universal idea that you recognize as a reconstruction, and to 

have a sensory memory is to neurophysiologically reconstruct a particular idea that you 

recognize as a reconstruction. Sensory remembering is thus a capacity of neither ghosts nor 

machines, but only of human beings qua mind-body unions. This interpretation unifies 

Descartes’s various remarks (and conspicuous silences) about remembering, from the 1628 Rules 

for the Direction of the Mind through the suppressed-in-1633 Treatise of Man to the 1649 Passions of 

the Soul. It also rebuts a prevailing thesis in the current secondary literature—that Cartesian 

critters can remember—while incorporating the textual evidence for that thesis—Descartes’s 

detailed descriptions of the corporeal mechanisms that construct sensory memories. 

 

Keywords: Descartes; Sensory Memory; Animal Spirits; Machine Psychology; Mind-Body Union 

 

1. Memories and mere imaginings 

 You've lost your keys again. Consider two psychological acts you could perform in 

order to find them. You could remember where you left them. Or, if your memory fails, you 

could imagine where you might have left them. In either case, you would bring an idea to bear, 

such as an idea of keys resting in your coat pocket. In either case, your ability to bring this idea 

to bear would rely on traces left in your brain by previous sense experiences. Indeed, the 

distinct acts of memory and imagination might produce apparently identical ideas of keys 

stowed in coat pockets. 

René Descartes famously argued that, while you happen to be a union of body and soul, 

your “essence consists solely in the fact” that you are a soul: “a thinking thing” (CSM 2:54; AT 

7:78).1 Nevertheless, Descartes did not believe the psychological acts just described to be purely 

intellectual functions of your immaterial soul. Instead, he held that remembering (or imagining) 

where you have (or might have) left your keys is mainly to be chalked up to the operation of 

                                                 
1 Descartes’s writings are cited by the following standard abbreviations. 

AT = Charles Adam and Paul Tannery (Eds.) (1964–76). Oeuvres de Descartes (Vols. 1-11). CNRS/J. Vrin. 

Cited by volume and page number. 

CSM = John Cottingham, Robert Stoothoff, and Dugald Murdoch (Eds. and Trans.) (1984–85). The 

Philosophical Writings of Descartes (Vols. 1-2). Cambridge University Press. Cited by volume and 

page number. 

CSMK = John Cottingham, Robert Stoothoff, Dugald Murdoch, and Anthony Kenny (Eds. and Trans.) 

(1991). The Philosophical Writings of Descartes (Vol. 3): The Correspondence. Cambridge University 

Press. Cited by page number. 

TM = Thomas Steele Hall (Ed. And Trans.) (1996). Treatise of Man. Prometheus Books (Great Minds Series). 

Cited by page number. 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

2 

 

physiological mechanisms in your brain. He dissected the heads of various animals in order to 

unearth these mechanisms. 

Despite the success of these dissections, there remains an interpretive puzzle about 

whether or not Cartesian critters can remember. Although Descartes frequently and 

enthusiastically attributed the faculty of imagination to nonhuman animals, he conspicuously 

shied away from unequivocal attributions of sensory memories—such as your memory of keys 

resting in your coat pocket—to creatures without souls. As Dennis Sepper writes, “the fact is 

that Descartes wrote very little about memory, and that little is more enigmatic than clarifying” 

(1998: 295). The enigmatic and sparse state of the textual evidence has given rise to two 

competing strands of interpretation in the scholarly literature. 

According to ‘corporealist’ scholars, Descartes mechanized all of the central functions of 

the sensitive soul, and thereby explained how soulless animals remember (Landormy 1902; 

Joyce 1997; Sutton 1998, 2016; Gaukroger 2000, 2002; Hatfield 2000, 2006, 2008, 2012, 2016; 

Clarke 2003; Ott 2017).2 Thus, John Sutton writes that “Descartes is consistent in attributing 

memory to animals” (2016: 490) and that “only intellectual memory is unique to humans: the 

celebrated beast-machine doctrine does not deny corporeal memory to animals” (1998: 74). And 

Richard Joyce (1997: 380) argues that Descartes claimed “that a system with no 

phenomenological mental states has the ability to remember … (A plausible view, though not 

Descartes's, is that remembering necessarily involves the conscious.)” Corporealists hold that 

Cartesian sensory remembering is the purview of the body, consisting solely in physiological 

processes. 

 ‘Incorporealist’ scholars agree that Descartes attributed memories to nonhuman animals 

in the 1630s, but argue that Descartes later renounced his youthful view and conflated sensory 

memory with intellectual memory (Morris 1969; Fóti 2000; Des Chene 2001; Scribano 2016). 

Emanuela Scribano argues that “the outcome of Descartes’ mature reflections on memory is that 

brain traces, which he searched for by dissecting animal heads, deserve only metaphorically to 

be called memory” (2016: 146). Animal brains lack a mechanism to render these metaphorical 

‘memories’ genuinely past-oriented. Thus, Véronique Fóti writes that in the 1640s Descartes 

“could no longer recognize a strictly bodily memory that humans would share with animals.” 

Instead, he came to believe “that animals do not possess genuine memory, but that their 

behaviour attests only to the conditioning of their bodily mechanisms” (2000: 598) 

Incorporealists hold that all genuine Cartesian remembering is the purview of the soul, 

consisting in acts of the intellect. 

 I have an intermediate interpretation to offer, according to which Cartesian sensory 

remembering is the purview of the mind-body union. Corporealists are right that Descartes 

mechanized the (unconscious) functions of the sensitive soul, including the physiological 

process of constructing sensory memories. But incorporealists are right that Descartes provided 

no mechanical means of distinguishing memory from mere imagination. If my arguments are 

good, both corporealism and incorporealism are false. Descartes held both that remembering 

                                                 
2 Corporealists disagree amongst themselves about just how richly sensitive Cartesian automata might be. 

For example, Sutton (1998) and Gaukroger (2000) interpret Descartes as countenancing purely 

mechanistic varieties of sentience and intentionality. Hatfield (2008) does not. 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

3 

 

necessarily involves the conscious 3  and that sensory memories consist in physiological 

processes (as opposed to conscious products of intellectual acts)—albeit physiological processes 

that directly engender conscious feelings and bear the right relationship to conscious acts of 

reflection. Descartes's understanding of memory was thus such that he could coherently dissect 

the heads of animals in order to explain what sensory memories are, while simultaneously 

denying animals the ability to remember. Humans, as mind-body unions, remember sensory 

ideas when our souls reflectively recognize that our bodies are reconstructing ideas (rather than 

merely fantasizing). Insofar as they lack the intellectual capacity for reflection, Cartesian critters 

possess the requisite neurophysiological machinery for memory yet cannot remember.4 

 

2. Memory as reconstruction recognized as such 

Both remembering and imagining a past sensation produce a particular idea, such as the 

idea of keys resting in your coat pocket. Neither of these acts produce the relevant idea out of 

thin air. They recruit traces that past events have left in your brain, in order to construct an apt 

image for the situation at hand. The most obvious difference is that imagination sometimes 

cobbles together fresh images, whereas memory reconstructs previously sensed images. (I will 

call the latter phenomenon ‘reconstruction.’) 

Many renaissance scholastics held reconstruction to be sufficient for remembering (Des 

Chene 2000; Edwards 2013), as did some seventeenth-century philosophers who shared 

Descartes’s project of mechanizing the sensitive soul, including Pierre Gassendi (1658)5 and 

Nicolas Malebranche (1674).6 However, most early modern European philosophers stressed that 

remembering necessarily involves an additional element: the reflective awareness that the 

newly constructed idea closely resembles a previously experienced idea. (I will call this 

phenomenon—the reflective awareness of a reconstruction as a reconstruction—‘recognition.’) 

Following Aristotle, Rudolph Goclenius (1613) countenanced recognition as a necessary element 

of remembering.7 So did prominent philosophers writing soon after Descartes, including Henry 

                                                 
3 Per Joyce, this view is plausible; pace Joyce, it is also Descartes’s. 
4 I presume that Cartesian critters lack rational souls. Descartes admitted that it cannot be proved that 

animals lack souls, “since the human mind does not reach into their hearts” (CSMK 365; AT 5:276-

277). Nevertheless, he argued severally that we have (inconclusive) reason to doubt that nonlinguistic 

animals have souls, since we need not posit souls in order to account for their behavior (CSM 1:139-

141; AT 6:55-59; CSMK 302; AT 4:573-576; CSMK 365-366; AT 5:277-279; CSMK 374; AT 5:344-345). 

Descartes presumed “the fact that animals lack a mind” (CSMK 181; AT 3:370) elsewhere as well 

(CSMK 148; AT 3:85). 
5 Gassendi (1659a: 560): “Nothing [corporeal] acts on itself … This is the reason that sight cannot see itself 

or know its vision or apprehend that it sees; nor can any other faculty, which is corporeal, do the like; 

and moreover, neither can phantasy, which is corporeal, perceive its own imagining or apprehend 

that it imagines.” Gassendi took memory to be the unreflective corporeal reproduction of ideas 

(Michael & Michael 1988; see also note 13 below). 
6 Malebranche (1674-75/1997: 106): “our brain fibers, having once received certain impressions through 

the flow of the animal spirits and by the action of objects, retain some facility for receiving these same 

dispositions for some time. Now, memory consists only in this facility, since one thinks of the same 

things when the brain receives the same impressions.” See Sutton (1998: Appendix 2) for commentary.  
7 Goclenius (1613: 680-681): “Memory is of those things which we have previously known, and notice that 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

4 

 

More (1659),8 Louis de La Forge (1666),9 John Locke (1689),10 Gottfried Leibniz (1714),11 and 

perhaps Thomas Hobbes (1651).12 According to these philosophers, memories are ideas that you 

recognize as reconstructions—aha! I felt my keys in my coat pocket earlier! Ideas that you do not 

recognize as reconstructions—could my keys be in my coat pocket?—do not count as memories, 

even when they happen to reconstruct earlier sense experiences. 

In this respect, Descartes’s understanding of memory was typical of his milieu. Late in 

life, Descartes provided a five-part analysis of memory. He wrote to Arnauld in 1648 that 

 

If we are to remember (recordemur) something, it is not sufficient that the thing should 

previously have been before our mind and have left some traces in the brain which give 

occasion for it to occur in our thought again; it is necessary in addition that we should 

recognize (agnoscamus), when it occurs the second time, that this is happening because 

it has already been perceived by us earlier. Thus verses often occur to poets which they 

do not remember ever having read in other authors, but which would not have occurred 

to them unless they had read them elsewhere.  

 

From this it is clear that it is not sufficient for memory (memoriam) that there should be 

traces left in the brain by preceding thoughts. The traces have to be of such a kind that 

the mind recognizes (agnoscat) that they have not always been present to us, but were at 

some time newly impressed. Now for the mind to recognize this, I think that when these 

traces were first made it must have made use of pure intellect to notice that the thing 

which was then presented to it was new and had not been presented before; for there 

                                                                                                                                                             
we have known, the marks of which are evident, and which without delay arouse our power of 

remembering. In other words, [in memory] there is a recognition of the thing previously known, as it 

was known.” (Thanks to Nabeel Hamid for advice on this translation.) 
8 More (1659: §10): “For there is necessarily comprehended in Memory a Sense or Perception that we have 

had a Perception or Sense afore of the thing which we conceive ourselves to remember.” 
9 La Forge (1666/1997: 182): “When some species re-appears on the gland it is always an effect of memory, 

unless the re-appearance depends completely on the object. But it is not always an effect of 

remembering. For in order to remember it is not enough simply to perceive a species which comes 

back again, if one does not also know that this is a re-appearance and that it is not the first time one has 

had this thought. Thus remembering or the power we have of recalling something consists in our 

faculty of recalling the original species on the gland and being aware that this is not the first occasion 

on which it gave us the thought which is present to the mind at the time.” 
10 Locke (1690/1975: 150): to remember is “to revive Perceptions, which it once had, with this additional 

perception annexed to them, that it has had them before.” 
11 Leibniz (1714/1996: 161): “I shall say then that it is sensation when one is aware of an outer object, and 

that recollection is the recurrence of it without the return of the object; but when one knows that one 

has had it before, this is memory.” See Jorgensen (2011) for an interpretation that fits this quotation 

with Leibniz’s various other remarks about memory. 
12 Hobbes (1651: 5): “This Decaying Sense, when wee would express the thing it self, (I mean Fancy it 

selfe,) wee call Imagination, as I said before; But when we would express the Decay, and signifie that 

the Sense is fading, old, and past, it is called Memory. So that Imagination and Memory, are but one 

thing, which for divers considerations hath divers names.” 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

5 

 

cannot be any corporeal trace of this novelty. (CSMK 356; AT 5:220) 

 

For Descartes, you remember something if and only if (1) the thing was previously presented to 

you (2) as a novel idea, (3) that idea made traces in your brain, (4) the traces in your brain led to 

the thing being presented to you again, and (5) you recognize that the thing was previously 

presented to you as a novel idea. On the reading I will advance, Descartes held soulless animals 

incapable of fulfilling conditions (2) and (5), since mechanical brains are equipped to keep track 

of neither novelty nor familiarity as such.  

On Gassendi and Malebranche’s broadly Cartesian views, conditions (1), (3), and (4) are 

jointly sufficient for memory, and poets who accidentally plagiarize do remember the verses 

they write, even though they fail to recognize themselves as reconstructing old ideas rather than 

innovating. If Descartes had adopted this simple analysis of memory, then he would have 

allowed that cats and dogs can remember, even though they cannot recognize their memories 

(or imaginings) as such.13 Descartes attributed a variety of sophisticated psychological capacities 

to the brains of animals, including the capacity to exploit traces of past experiences to 

reconstruct corporeal ideas and initiate action routines.14 Cats physiologically reconstruct ideas 

of the mice of their past, and “if you whipped a dog five or six times to the sound of a violin, it 

would begin to howl or run away as soon as heard that music again” (CSMK 20; AT 1:134), 

reconstructing corporeal associations with (the unconscious aspects of) pain. Descartes thus 

easily could have attributed memories to Cartesian critters, simply by defining memory as 

reconstruction. 

However, Descartes had substantive reason to eschew this simple analysis. Sensory 

memory and imagination are different faculties that play different functional roles in human 

cognition (CSM 2:40; AT 7:57; CSM 1:216; AT 8A:32). Somebody who remembers where they left 

their keys will confidently retrieve them, or reassure their partner that the keys are safely 

stowed away. Somebody who merely imagines where they might have left their keys is more 

likely to waver, even when their fancy has faithfully reconstructed a past sensation. Descartes 

defined memory as reconstruction plus recognition because he saw no other principled way to 

distinguish true memories from mere imaginings. Only the reflective capacity of the human 

soul does the trick. Cartesian critters fail to recognize their occurrent reconstructions as 

reconstructions, and thus fail to transform them into memories.15 

                                                 
13 Gassendi (1649b: 441): “Certainly nothing can be more absurd than to think that a dog, for example, 

says to himself, I imagine that I imagine or something else of this sort.” 
14 Michael & Michael (1989) provide a seminal discussion of the use of the term ‘idea’ in the seventeenth 

century, with foci on Descartes’s early use of ‘idea’ to refer exclusively to corporeal ideas—purely 

physiological configurations of brain matter—as well as his shift to using the term to refer to any 

image that appears before the mind. My gloss of Cartesian corporeal ideas can be found in §3, below. 
15 Similar debates about the nature of memory still occupy philosophers in the twentieth and twenty-first 

centuries. Intellectual heirs of Gassendi and Malebranche support the causal theory that you 

remember insofar as you reconstruct an idea and the reconstructed idea was causally operative in the 

reconstruction (Martin & Deutscher 1966; Bernecker 2010). Intellectual heirs of Descartes and 

company support the sense-of-pastness theory that you remember insofar as you reconstruct an idea 

“and take the relevant experience into account when judging about the past” (Debus 2010: 25; 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

6 

 

In what follows, I square this interpretation with Descartes's sundry claims about 

remembering, from the Rules for the Direction of the Mind (1628) through the Passions of the Soul 

(1649). §3 provides a crash course in Descartes’s “machine psychology” (Hatfield 2007), with an 

emphasis on the neurophysiology of sensory memory. §4 contrasts three competing 

interpretations: corporealism, incorporealism, and my intermediate interpretation. §5 concludes. 

 

3. Mechanizing the sensitive soul 

 Nowadays, philosophers and scientists use the term ‘memory’ variously to invoke 

declarative memory (including episodic memory and semantic memory), nondeclarative 

memory (including procedural memory), and memory traces, among other mental and neural 

phenomena (Squire 2009). On my reading, Descartes taxonomized these cognitive functions in 

roughly the same way, and denied only declarative memory—the recognized reconstruction of 

particular ideas—to animals.16 Descartes used the term ‘habit’ instead of ‘memory’ to describe 

procedural memory—acquired skill in motor function, stimulus-response association, pattern 

detection, and so on—except when the relevant procedure features a declarative memory.17 

Descartes did frequently use ‘memory’ to refer to memory traces—metaphorical stores of ideas 

in the brain. He also wrote of “the retention or imprint of ideas in the Memory (la Memoire)” 

(TM 113; AT 11:202), using ‘the Memory’ (usually capitalized) to refer to the brain region 

featuring memory traces. This much is undisputed: according to the machine psychology 

Descartes developed in the Treatise of Man (~1633), Dioptrics (1637), and Passions of the Soul 

(1649), soulless creatures exploit traces in the Memory to learn procedures and make sensory 

ideas available for physiological processing. The principal interpretive question that divides 

corporealists and incorporealists is whether Descartes identified any of these sensory ideas in 

animals’ brains with declarative memories. 

In 1632, Descartes wrote to Mersenne that “my discussion of man in The World 

                                                                                                                                                             
Fernandez 2008). A third theory rejects the presupposition—made by both causal theorists and their 

opponents—that remembering must be distinguished from mere imagining, and posits that 

remembering just is imagining the past (Michaelian 2016). After all, cognitive neuroscience has 

revealed that Descartes was right that the same brain regions are implicated in memory and 

imagination, though recent work has suggested that different patterns of activity in the hippocampus 

underlie the two processes (Addis et al. 2007; Schacter et al. 2012; Kirwan et al. 2014). 
16 On this usage, declarative memory includes episodic memory—the recollection of episodes or states of 

affairs—and semantic memory—the recollection of facts, concepts, or meanings. For Descartes, on my 

reading, each of these kinds of recollection consists in the reconstruction of ideas (of previously 

cognized episodes, states of affairs, facts, concepts, or meanings), and counts as memory only insofar 

as it is recognized as a reconstruction. Episodic memory is the purview of mind-body unions, 

whereas semantic memory is the purview of pure intellects. 
17 For example, the lute player’s memory of passages (CSMK 146; AT 3:48) might be considered an 

instance of procedural memory, but Descartes only calls it ‘memory’ because it crucially involves the 

recognized reconstruction of particular ideas. When discussing instances of procedural memory that 

do not involve clear-cut declarative memories, such as the dog running from the violin (CSMK 20; AT 

1:134) or the human controlling her passions (CSM 1:348; AT 11: 368-370), Descartes sometimes 

invokes ‘the Memory,’ but never explicitly calls procedural memories ‘memories.’ Instead, he speaks 

of ‘habits’ (CSM 1:344; AT 11: 361-362). 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

7 

 

[including the Treatise] will be a little fuller than I had intended, for I have undertaken to 

explain all the main functions in man.” Having finished writing about digestion, circulation, 

and sensation, Descartes had turned to “dissecting the heads of different animals, so that I can 

explain what memory, imagination, etc. consist in” (CSMK 40; AT 1:263). In the Treatise, 

Descartes presented his neurophysiology as a study of the (apparently) psychological capacities 

of human bodies, considered without reference to rational souls. He urged that, even without 

recourse to the rational faculties of understanding and will, the fabulous humanoid automata 

described in the Treatise would be capable of perfectly “imitat[ing] all the movements that real 

men will make when the soul is present” (TM 96; AT 11:185).18 In so doing, Descartes rejected 

the scholastic notion of a sensitive soul, and located the abilities to sense, imagine, and 

remember squarely in brain matter.19 

 Descartes’s machine psychology was inspired by hydraulic machines found “in the 

grottoes and fountains in the gardens of our kings” in seventeenth-century Europe (TM 21; AT 

11:130). However, Descartes envisioned animal (including human) bodies to be powered by the 

rapid flow of fluids finer than water. Following the Galenic tradition, Descartes held “animal 

spirits” to be produced by siphoning off “the liveliest, strongest, and subtlest parts of … blood” 

(TM 17; AT 11:128).20 Descartes's metaphors for animal spirits—“a certain very subtle wind, or 

rather a very lively or very pure flame” (TM 19; AT 11:128)—emphasize the delicacy, speed, and 

volatility with which they course through nervous systems. Blood and water are too sluggish to 

be the means by which brains communicate nearly instantaneously with peripheral nerves. 

 Animal spirits flow downhill from a single source: the pineal gland (where, not 

coincidentally, Descartes located the mind-body connection). The pineal gland, “situated near 

the middle of the brain's substance just at the entrance to its cavities,” receives only the subtlest 

parts of the blood—animal spirits—from the arteries, coarse parts having been soaked up to 

“nourish [the brain's] substance” (TM 19-20; AT 11:128). When arteries pump blood into the 

brain, they cause the pores surrounding the pineal gland to dilate, and thereby become 

receptive to the backflow of animal spirits. Upon reaching the highest peak of the pineal gland, 

refined animal spirits perform an about-face and flow back through “the fibrous mesh of the 

brain substance,” “whence they spill out into all the nerves.” The constant flow of animal spirits 

from the pineal gland to the nerve endings at the tips of fingers and toes serves to “keep all the 

                                                 
18 Drieux (2016) provides a close reading which highlights the ambiguity of this passage. 
19 Descartes announced that The World, like other texts discussing his machine psychology, should be read 

as a fable: a model presenting how the material world could be, rather than an assertive description of 

actuality. There is reason to believe that this disclaimer was a facetious smoke screen, since Descartes 

told Mersenne in 1633 that he suppressed publication of The World—in which he “decided to explain 

all the phenomena of nature, that is to say, the whole of physics” (CSMK 7; AT 1:70)—after the church 

censured Galileo’s Dialogue Concerning the Two Chief World Systems, since “for all the world I did not 

want to publish a discourse in which a single word could be found that the Church would have 

disapproved of” (CSMK 41; AT 1:271). Even if Descartes sincerely meant the mechanistic details 

presented in the Treatise to be taken only as a model, his overarching metaphysical system ensures 

that the construction of sensations, imaginings, and memories must be mechanized somehow or other. 

After all, soulless animals in the real world have sensory ideas. 
20 Despite their name, animal spirits are modes of material—not mental—substance. 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

8 

 

filaments that compose the nerves and the brain so tense that even those actions that have 

barely force enough to move them are easily communicated from one of their extremities to the 

other” (TM 84; AT 11:174). A pinprick to an animal's toe causes nerves, taut with animal spirits, 

to yank open the relevant pores in the brain, calling the pineal gland to attention. 

 

3.1. Sensation 

 The most developed part of Descartes's hydraulic neurophysiology is his theory of 

sensation, and specifically his optics.21 Descartes famously denied the Aristotelian assumption 

that sensations must directly resemble their objects. He proposed that the task of the 

philosopher of perception “is to know simply how [the images formed in our brain] can enable 

the soul to have sensory perceptions of all the various qualities of the objects to which they 

correspond—not to know how they can resemble these objects” (CSM 1:166; AT 6:113-114). 

How can corporeal ideas represent their particular objects without (necessarily) physically 

resembling them?22 For starters, the physiology of sensory representation approximates the 

physiology of pain responses. Corporeal representation involves the pineal gland—yanked to 

attention by the object of sensation—appropriately adjusting the volume and velocity of animal 

spirits coursing from its surface. Consider Diagram 50 from the Treatise. 

 

 
 

Light rays bounce off points A, B, and C on the arrow and “exert pressure on the back of the 

eye.” The optic nerves which connect the eyes to the pineal gland (H) comprise many filaments, 

including those labeled 1-2, 3-4, and 5-6. Because these filaments are taut with animal spirits, the 

slightest pressures “pull the whole of thread 1-2 [& 3-4 & 5-6] and enlarge the opening of the 

tubule marked 2 [& 4 & 6] (TM 84; AT 11:175),” allowing the filaments to receive animal spirits 

                                                 
21  Descartes distinguished three grades of sensation (CSM 2:294-296; AT 7:437-439). The physical 

sensations I discuss in this section are of the first grade. Second grade sensations are the qualitative 

experiences that arise directly from physical sensations (in mind-body unions). Third grade 

sensations are intellectual judgments based on first/second grade sensations, such as the reflective 

recognition that an idea is a reconstruction. Useful discussions include Garber (1993), Simmons (1999, 

2017), Chignell (2009), and Kouivuniemi & Curley (2015). 
22 Descartes countenanced both particular ideas (generated by the brain) and universal ideas (in the mind). 

Neither kind of idea is guaranteed to resemble its object. 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

9 

 

from the pineal gland “more freely and rapidly,” per lines a-2, b-4, and c-6. Finally, “the 

spirits … draw the gland after themselves a little, and cause it to lean if it is not otherwise 

prevented from so doing; so that, changing the position of its pores, it begins to conduct a much 

greater quantity of spirits through a, b, and c to 2, 4, and 6 than it did before” (TM 97; AT 11:185). 

Thus, rather than transmitting an intact image, the light from the arrow sets off a chain reaction 

culminating in a specific pattern in the pineal gland's distribution of animal spirits, which 

mirrors the retinal pattern in its two-dimensional order. This patterned flow of animal spirits is 

the brain’s corporeal sensation of the arrow. 

Corporeal sensations that physically represent—and enable souls to intellectually 

represent—material objects are four steps removed from, and should not be assumed to 

resemble, the visual properties of their objects. The visual properties of the arrow (A, B, and C) 

are physically represented at one remove by retinal images (1, 3, and 5), at two removes by 

enlarged openings of tubules (2, 4, and 6), at three removes by the lean of the pineal gland, and 

at four removes by the faster and freer pattern of flow from points on the pineal gland (a, b, and 

c). Descartes wrote that “among these figures, it is not those imprinted on the organs of external 

sense [the retinal images], or on the internal surfaces of the brain [the opening of the tubules], 

but only those traced in spirits on the surface of gland H, where the seat of imagination and common 

sense is, that should be taken to be ideas, that is to say, to be the forms or images that the 

rational soul will consider directly when, being united to this machine, it will imagine or will 

sense any object” (TM 86; AT 11:176). Descartes did not consider retinal images or enlarged 

tubules—intermediaries between external world and pineal gland—to be sensations. Patterns in 

the flow of animal spirits from the leaning pineal gland are the only sensations that allow souls 

to represent material objects. While patterns of flow correlate with the shape, “movement, size, 

distance, colors, sounds, odors, and other such qualities” of objects (TM 85; AT 11:176), they 

need not resemble these qualities. Animal spirits do not flow redly, loudly, or stinkily, but from 

different points, with varying gradients of subtlety and agitation of matter, in lesser or greater 

quantities with lesser or greater intensities. Ideas representing different objects vary solely in 

accordance with these four variables: release point, subtlety gradient, volume, and velocity (TM 

75; AT 11:169). 

 

3.2. Imagination 

 Physical sensations are one kind of corporeal idea: animal spirit surges generated in 

response to light impinging on the eyes, ears, nose, tongue, or skin. But animal spirits 

sometimes flow in the absence of outside forces. As Descartes remarked, corporeal ideas “are all 

to be attributed to the common sense when they depend on the presence of objects, but … can 

also proceed from several other causes ... and should then be attributed to imagination” (TM 87; 

AT 11:176). Descartes’s term 'imagination' encompasses all sensory ideas that are not caused by 

physically present external objects. Some imaginings are amalgams of previous ideas, and do 

not represent any previously sensed object. Others are more or less faithful reconstructions of 

previous patterns of flow of animal spirits.23 

 Humans can will themselves to conjure up new ideas or recall old ones. When you 

                                                 
23 Sepper (1998) provides an in-depth treatment of Descartes on imagination. 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

10 

 

concentrate on finding your keys, your willful soul compels your pineal gland to lean 

appropriately, releasing the patterned flow of animal spirits that constitutes your imagined or 

remembered idea. Animals, alongside the humanoid automata of the Treatise, lack souls and 

therefore lack willpower. But soulless machines do not lack the faculty of imagination; they too 

form ideas in the absence of objects. Daydreaming is a variety of unwilled imagination; even 

automata sometimes “let their fancy wander listlessly here and there without external objects 

diverting it and without the fancy's being directed by reason.” When daydreaming, the pineal 

gland releases novel patterns of animal spirits: unprecedented combinations of velocity, volume, 

subtlety gradient, and release points. Often, these new combinations are amalgams of old 

combinations—ideas of lions, snakes, goats, horses, and eagles—and thereby represent 

imaginary monsters such as “chimeras and hypogryphs” (TM 96; AT 11:184). Daydreamers are 

capable of drawing on previous sensory ideas to dream up these imaginary creatures because 

old ideas are traced in their Memory folds. 

 

3.3. Memory 

 Descartes described “the internal part of the brain” (labeled 'B' in Diagram 50) as “the 

seat of the Memory” (TM 88; AT 11:178). 24  Descartes doubted that the pineal gland could 

construct abominations like chimeras and hypogryphs on its own. He wrote to Mersenne in 

1640 that imaginative humans “would not have the great facility which they have in imagining 

an infinity of things which they have never seen, if their souls were not joined to some part of 

the brain that was very well equipped to receive all kinds of new impressions, and consequently 

very ill equipped to preserve them” (CSMK 145; AT 3:47-48). The pineal gland sacrifices the 

ability to retain ideas for greater flexibility in creating ideas. The Memory is the storage space 

required to supply the pineal gland with subject matter beyond immediate sense impressions. 

 As with idea creation, the mechanisms of idea retention and reconstruction revolve 

around differential flows of animal spirits. The pineal gland pumps animal spirits, not only 

through tubules connected to sense organs, but also “into the pores or intervals that occur 

between the filaments composing part B of the brain” (TM 87-88; AT 11:177). Ideas send 

different volumes of animal spirits at different velocities into different pores in the Memory. 

While physically resistant to new ideas, the Memory slowly but surely becomes receptive to oft-

repeated ideas. If a faint new idea is formed on the pineal gland, its flow of animal spirits leaves 

the Memory more or less as it was, because the pores associated with that idea remain narrow 

and rigid. Spirits bore their way into the Memory folds “increasingly effectively in the measure 

that their action is stronger, or lasts longer, or is more often repeated.” When an idea has 

released a strong flow of spirits on many occasions, the traces left in the Memory “are no longer 

                                                 
24 In the Treatise, Descartes consistently used the capitalized (and sometimes italicized) term 'la Memoria' 

to refer to the part of the brain responsible for the storage of memory traces. Elsewhere, he used the 

lowercase term ‘memoria’ to refer to declarative memories—reconstructed corporeal ideas that, like 

all corporeal ideas, are patterns in the flow of animal spirits from the surface of the pineal gland—and 

used the term ‘plis de mémoire’ to refer to memory traces. However, Descartes used the lowercase 

‘memoria’ only twice in the Treatise. The first instance refers to memory traces (TM 83; AT 174). The 

second instance may refer to declarative memory, but does not ascribe memoria to the automata under 

discussion (TM 107; AT 195). 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

11 

 

so easily erased, but are retained there in such a way that by means of them the ideas that 

existed previously on this gland can be formed again long afterward, without requiring the 

presence of the objects to which they correspond” (TM 88; AT 11:178). Persistently recurring 

ideas rearrange the filaments in the Memory, forcing the relevant pores to become wider and 

looser. 

 Descartes concluded this discussion by claiming that “it is in this that the Memory 

consists” (TM 88; AT 11:178). He thereby identified the state of the Memory being amenable to 

receiving certain ideas—patterned flows of animal spirits—with the retention of those ideas. To 

make sense of this identification, Descartes invoked a few analogies. He noted (with regard to 

the following image reproduced from the Treatise) that “if one were to pass several needles or 

engravers' points through a linen cloth as you see in the cloth marked A, the little holes that one 

would make would stay open as at a and at b after the needles had been withdrawn; or if they 

closed again, they would leave traces in this cloth, as at c and at d, which would enable them to 

open quite easily again” (TM 89; AT 11:178-179). 

 

 
 

In a letter to Meysonnier in 1640, meanwhile, he wrote that “the impressions that are preserved 

in the Memory … are simply like the folds that remain in this paper once it has been folded” 

(CSMK 143; AT 3:20; cf. CSMK 233; AT 4:114). Finally, Descartes wrote to Arnauld in 1648 that 

“we say that there are no human tracks in the sand if we cannot find any impressions shaped 

like a human foot, though perhaps there may be many unevennesses made by human feet, 

which can therefore in another sense be called human tracks” (CSMK 356-357; AT 5:220). The 

cloth does not permanently store the holes, nor the paper the fold, nor the sand the tracks. 

Nevertheless, the cloth carries the disposition to reopen old holes, the paper carries the 

disposition to refold along original fault lines, and the sand carries the disposition to receive 

familiar feet. By analogy, Descartes denied that the Memory stores memories, in the sense of 

retaining discrete figures in its folds. Instead, the Memory dispositionally preserves memories, 

in the same sense that the cloth and paper and sand dispositionally preserve their holes and 

folds and tracks. The brain reconstructs ideas in virtue of being disposed to receive well-worn 

ideas more readily than it receives novel ones.25 

                                                 
25 Descartes's treatment of corporeal memories parallels his treatment of innate ideas. In the Comments on 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

12 

 

 The pores of part B of the brain are no different than the other bodily pores through 

which animal spirits travel. Indeed, in 1640 Descartes wrote to Mersenne that “all the nerves 

and muscles can also be so utilized” to retain corporeal ideas; “a lute player, for instance, has a 

part of his memory in his hands: for the ease of bending and positioning his fingers in various 

ways, which he has acquired by practice, helps him to remember the passages which need these 

positions when they are played” (CSMK 146; AT 3:48; cf. CSMK 143-144; AT 3:20). Lute players' 

fingers are more receptive to animal spirits than undexterous people’s fingers, leading lute 

players' pineal glands to be more likely to release spirits in patterns that constitute ideas of 

musical refrains for their fingers to pluck. Descartes even extended the ability to preserve ideas 

beyond the body. He declared that “when we have read a book, not all the impressions which 

can remind us of its contents are in our brain. Many of them are on the paper of the copy which 

we have read” (CSMK 146; AT 3:48). Rereading a book gives rise to the same sensations 

entertained on the first read. Just as the connection between the pineal gland and the 

neurological Memory facilitates repetition, access to the written word disposes us to reconstruct 

certain ideas again and again. 

 There is an apparent tension between Descartes's endorsement of the embodied and 

extended Memory and his later analysis of memory. Conditions (3)—that the original idea leave 

traces in one's brain—and (4)—that the traces in one's brain lead to the recurrence of the idea—

of his analysis seem to require that memory traces be situated in brains. Either Descartes 

rescinded belief in the embodied and extended Memory by 1648, or he intended to require only 

that the original idea leave traces somewhere suitably accessible to the rememberer. 

 There are a couple of reasons to prefer the latter interpretation. First, Descartes 

countenanced intellectual memories which leave incorporeal traces in souls (CSMK 148; AT 

3:84-85; CSMK 233; AT 4: 114), so he was generally open to the possibility of non-neural 

memory traces. More pertinently, Descartes was committed to the view that “nature always acts 

by the simplest and easiest means” (TM 113; AT 11:201). The aim of his neurophysiology was to 

demonstrate that “it is not necessary to conceive of any vegetative or sensitive soul or any other 

principle of movement and life than [the animal machine's] blood and its spirits, agitated by the 

heat of the fire which burns continually in its heart and which is of no other nature than all 

those fires that occur in inanimate bodies” (TM 113; AT 11:202). Descartes set out to prove that 

nutritive and sensitive functions—including sensory memory—could be carried out by the 

same deterministic motions that characterize whirlpools and the flight of arrows and other 

phenomena of the material world. Given this project, Descartes was right to deny that only gray 

matter can retain traces of corporeal ideas. Descartes expected most memory traces to be located 

in part B of the brain, but his mature neurophysiology also allowed neural machines to be 

                                                                                                                                                             
a Certain Broadsheet, Descartes wrote that innate ideas should not be thought of as immortally stored 

in the mind, but rather as innate in “the same sense as that in which we say that generosity is 'innate' 

in certain families, or that certain diseases such as gout or stones are innate in others: it is not so much 

that the babies of such families suffer from these diseases in their mother's womb, but simply that 

they are born with a certain 'faculty' or tendency to contract them” (CSM 1:303-304; AT 8B:358). In 

short, to have innate ideas is to be innately disposed to have those ideas, just as to store memories is 

to become disposed to reconstruct ideas. As I suggest in §§4.2-4.3, Descartes likely understood the 
intellectual capacity to recognize ideas as reconstructions to be analogously dispositional. 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

13 

 

hooked up to (mechanistic) limbs and environments such that peripheral nerves and written 

notes could retain memories as ably, if not as immediately, as the brain’s Memory. Indeed, 

Descartes may have counted fingers, books, and even other organisms as literal extensions of 

the Memory itself, insofar as they feature memory traces. 

 Regardless, Descartes's paradigm case of remembering involves the willful exploitation 

of traces in the folds of the brain’s Memory. Descartes described this process at some length in 

the Passions: 

 

when the soul wants to remember something, this volition makes the [pineal] gland lean 

first to one side and then to another, thus driving the spirits towards different regions of 

the brain until they come upon the one containing traces left by the object we want to 

remember. These traces consist simply in the fact that the pores of the brain through 

which the spirits previously made their way owing to the presence of this object have 

thereby become more apt than the others to be opened in the same way when the spirits 

again flow towards them. And so the spirits enter into these pores more easily when 

they come upon them, thereby producing in the gland that special movement which 

represents the same object to the soul, and makes it recognize the object as the one it 

wanted to remember. (CSM 1:343-344; AT 11:360) 

 

Willful remembering involves trial and error. When humans ransack their brains for a memory, 

their pineal glands lean this way and that, releasing animal spirits willy-nilly until they stumble 

upon the correct pattern of release points, subtlety gradients, volumes, and velocities. Better 

remembered ideas are easier to stumble upon, since the Memory is more receptive to their 

unique flow. 

 Humans also remember idly. A reconstructed image of your keys resting in your coat 

pocket may pop into your head at any moment, unbidden. Descartes held that in such cases the 

pineal gland “is inclined in one direction by the force of the spirits alone, without the aid of the 

rational soul or of the external senses.” If, “at the region of the brain toward which the gland is 

inclined, the shape of one particular object is imprinted more distinctly than any other, the 

spirits tending to that region cannot fail to receive an impression thereof. And it is thus that past 

things sometimes return to thought as if by chance and without the Memory of them being 

excited by [either] any object impinging on the senses” or any act of the will. Our pineal glands 

sometimes happen to reconstruct ideas irrespective of whether we want to remember. More 

frequently, though, when humans let their minds wander “several different figures are traced in 

this same region of the brain almost equally perfectly” (TM 96; AT 11:184), producing 

imaginary mishmashes like chimeras and hippogryphs. 

 

4. Three interpretations of Descartes on sensory memory 

 Cartesian cats and dogs lack the ability to will themselves to remember, but their fancies 

do wander. Although the pineal gland swaying of its own accord usually results in the creation 

of obviously imaginary ideas, it sometimes results in true memories, even in the brains of 

soulless animals. Or so goes the corporealist interpretation of Descartes's theory of sensory 

memory. 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

14 

 

 

4.1. Corporealism 

 Corporealists emphasize Descartes's descriptions of the neurophysiological processes 

that suffice for the reconstruction of ideas. For example, he wrote (well into the 1640s) of “the 

corporeal memory, whose impressions can be explained by these folds in the brain” (CSMK 151; 

AT 3:143), and that “the motion of these brain particles leaves behind the traces on which 

memory depends” (CSMK 190; AT 3:425). Indeed, Descartes boasted that “the effect of the 

Memory that seems to me to be most worthy of consideration here consists in [the fact] that 

without there being any soul in this machine it can be naturally disposed to imitate all the 

movements that real men (or many other, similar machines) will make when the soul is present” 

(TM 96; AT 11:185). He concluded the Treatise by asking his readers to consider that functions 

such as “the retention or imprint of [sensory] ideas in the Memory … imitate those of a real man 

as perfectly as possible and that they follow naturally in this machine entirely from the 

disposition of the organs—no more nor less than do the movements of a clock or other 

automaton, from the arrangement of its counterweights and wheels” (TM 113; AT 11:202). The 

Cartesian brain is equipped to retain memory traces and reconstruct memories like fabulous 

hydraulic clockwork, interventions by the rational soul notwithstanding. Corporealists, citing 

Descartes’s passionate testaments to the sensitivity, flexibility, and creativity of soulless 

automata, argue that “this is not 'memory' in any derivative sense: this just is (corporeal) 

memory” (Sutton 1998: 74). 

 Much of the textual evidence does favor the corporealist view that Descartes attributed 

sensory memories to animals, but for a couple of snags. Here is the first snag: Descartes never 

unequivocally claimed that critters can remember. As illustrated in §3, he maintained that 

memories consist in patterns in the flow of animal spirits from the surface of the pineal gland 

which, facilitated by traces in the folds of the Memory, reconstruct sensory ideas. But he 

conspicuously avoided explicitly identifying reconstructed ideas in either animal or humanoid 

automata brains as memories. 

The second snag is that no mechanism supported by Descartes’s physics could 

distinguish true memories from mere imaginings.26 Corporealists err in inferring attribution of 

memory from attribution of reconstruction. Mere imaginings can consist in the same patterns of 

animal spirits—the exact same images—as memories, and are facilitated by the same traces in 

the folds of the Memory. Crucially, imaginings are sometimes reruns too. The question of 

whether your idea of keys in your coat pocket is a memory or mere imagining is not settled by 

the fact that a physically identical sensory idea previously left traces in your Memory. Instead, 

as discussed in §2, Descartes argued that remembering necessarily involves the reflective 

recognition that a reconstruction is a reconstruction. 

An ecumenical corporealist might allow that Descartes was concerned to distinguish 

                                                 
26 Twentieth and twenty-first century philosophers who define memory as reconstruction tend to rely on 

the causal history of a reconstruction to mark it as a memory rather than mere imagining. It is unclear 

how causal histories could play this kind of metaphysically individuating role in the context of 

Descartes’s austere corpuscularian physics, though an anonymous reviewer has intriguingly 

suggested that Descartes could be seen as anticipating embodied and extended versions of the causal 

theory of memory (as proposed by Sutton & Windhorst 2009). 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

15 

 

reflective remembering from mere reconstruction, but insist that Cartesian critters nevertheless 

remember insofar as they reconstruct ideas. 27  To see the crux of this reply, consider three 

idealized psychological events: (a) utterly novel imagination, (b) reconstruction without 

recognition, and (c) reconstruction with recognition. For example, a poet could (a) devise brand 

new verses, (b) deliver somebody else’s verses without realizing they are plagiarizing, or (c) 

recite verses and recognize having previously heard those verses. The ecumenical corporealist 

and I agree that (a) is a case of mere imagining, and (c) is a case of distinctly human 

remembering. The contentious issue is how to categorize cases like (b). Does reconstruction 

suffice for memory, or is something more needed? 

Ecumenical corporealism is worthy of serious consideration; perhaps Descartes wanted 

to distinguish reflective remembering—a variety of sensory remembering only found in 

humans—from unreflective remembering—which is found in other animals too.28 Descartes's 

goal in the Treatise was to demonstrate that un-ensouled human bodies can perform an 

impressive range of apparently intelligent activities. Many apparently intelligent activities are 

enabled by animal spirits boring traces into the Memory and subsequently exploiting those 

traces in order to reconstruct ideas. Still, the two snags provide evidence that Descartes did not 

classify cases like (b) as memories. Even in his neurophysiological works, Descartes never 

explicitly called cases like (b) ‘memories’; on the contrary, he explicitly insisted that “verses 

often [re]occur to poets which they do not remember” (CSMK 356; AT 5:220). On my reading, 

Descartes refrained from calling mere reconstructions ‘memories’ because he understood 

sensory memory and imagination to be functionally differentiated. Whereas imagination serves 

to (re)construct ideas for physiological (and, in mind-body unions, conscious) processing, 

memory serves to reconstruct ideas for processing as recognized reconstructions. In other words, 

memories function as objects of reminiscence; mere imaginings—even when they are 

reconstructions—do not. I will say more about this functional difference in §4.3. In the 

meantime, the key point is that this functional conception of sensitive capacities draws the line 

dividing imagination from memory between cases (b) and (c), rather than between cases (a) and 

(b). Sensation and imagination are the only functional capacities required to power the 

mechanized sensitive souls that make Cartesian critters tick. 

Descartes never attributed the ability to recognize ideas as reconstructions to creatures 

without rational souls. 29  Indeed, in the 1648 letters to Arnauld, Descartes denied soulless 

                                                 
27 Thanks to an anonymous reviewer for suggesting this middle way between strong corporealism and 

my intermediate interpretation. 
28 Stephen Gaukroger pursues just such an ecumenical corporealism. Gaukroger steadfastly supports the 

claim that Cartesian critters remember (2012: 204-206, 217), but also allows that humans remember 

differently, writing that the key difference “seems to lie in the ability of human beings to stand back 

from, make judgements about, and ultimately control their cognitive and affective states” (214). 
29 One possible exception is in Rule 12, where Descartes equated imagination with “the Memory which is 

corporeal and similar to the one which animals possess” (CSM 1:43; AT 10:416). But given how 

Descartes used the terms 'imagination' and 'memory' in the context of the Rules, it is probable that 

Descartes referred not to memories (qua ideas) but to the Memory here. Throughout Rule 12, 

Descartes used ‘the common sense’ to designate the place in the brain (which he would only later 

identify with the pineal gland) where ideas are formed, and mainly used ‘the imagination’ to 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

16 

 

machines both the ability to recognize reconstructions as reconstructions (condition 5) and the 

prerequisite ability to mark memory traces as novel (condition 2). If these letters stand as the 

definitive statement of Descartes's theory of memory, then the corporealist view that Descartes 

attributed memories to animals is false. Realizing this predicament, the corporealists John 

Sutton (1998) and Desmond Clarke (2003) have independently argued that the 1648 letters are 

not definitive, on the grounds that Descartes later countenanced the corporeal recognition of 

corporeal traces of novelty. Sutton writes that the “claim that there can be no corporeal trace of 

this novelty is contradicted by Descartes's own later discussion of wonder” (1998: 71). This later 

discussion occurs in §§70-75 of Part Two of Descartes's last published work: the Passions of the 

Soul. Clarke offers an opiniated summary: 

 

The definition of wonder that is offered in the Passions and its explanation in terms of 

characteristic flows of animal spirits presuppose the capacity of a subject to recognize 

the novelty of, for example, a perceptual experience because it causes the spirits to flow 

in unusual ways. Presumably, familiar experiences are associated with flows of spirits in 

patterns that are already well established and benefit from the inertial dispositions 

mentioned above, whereas a novel experience will direct the spirits in ways that 

encounter more resistance. This suggests, in a very schematic form, a physical basis for 

recognizing novelty or its absence. (Clarke 2003: 104-105) 

 

When a sense impression fills an animal with wonder (as novel sense impressions are wont to 

do), it prompts the pineal gland to send animal spirits into previously unexplored pores in the 

Memory. On Sutton and Clarke’s reading, this flow of spirits yields a corporeal trace of novelty 

by “conspir[ing] to isolate a memory trace and render[ing] it, temporarily, independent of 

others” (Sutton 1998: 71). Finally, soulless animal brains recognize that memories are 

reconstructions insofar as the repetition of previously wondrous ideas sends animal spirits into 

isolated and relatively unexplored memory traces. 

 Sutton and Clarke's interpretation of Descartes's discussion of wonder contradicts 

Descartes's claim that there can be no corporeal trace of novelty. It also contradicts another 

important aspect of Descartes's neurophysiology of memory. Descartes asserted (in the Passions 

and elsewhere) that when it comes to remembering, practice makes perfect. Though we do not 

remember much from our childhood, “those [things] we do remember are remembered not only 

because of the impressions we received when we were young, but mainly because we have 

done the same things again and renewed the impressions by remembering the events over 

again from time to time” (CSMK 151; AT 3:143-144). Sutton and Clarke's interpretation requires 

Descartes to renounce this view, and hold that we recognize traces of novelty—and thus 

remember—only when the relevant folds of the Memory are relatively uncharted terrain. As we 

remember events over and over again, corporeal traces of novelty must fade. On the 

corporealist interpretation, it therefore must become progressively harder for soulless animals 

to recognize well-worn memories as memories. 

                                                                                                                                                             
designate the place in the brain which he would later refer to mainly as ‘the Memory’ (see CSM 1:41-

43; AT 10:414-417) 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

17 

 

 Descartes's text does not commit him to this implausible consequence, because it does 

not commit him to the corporealist interpretation of his views. The Passions states that wonder 

“has two causes: first, an impression in the brain, which represents the object as something 

unusual and consequently worthy of special consideration; and secondly, a movement of the 

spirits, which the impression disposes ... to flow with great force to the place in the brain where 

it is located so as to strengthen and preserve it there” (CSM 1:353; AT 11:380-381). Contra Sutton 

and Clarke, this passage does not imply lasting corporeal traces of novelty, suitable for the 

recognition of a reconstruction as a reconstruction. Instead, it implies that wondrous 

impressions lead objects to be represented as novel to the soul, and that wondrous impressions 

bore unusually strong memory traces into new areas of the Memory (without leaving corporeal 

traces of novelty).30 These implications are consistent with Descartes's letters to Arnauld. The 

passion of wonder, and associated surge of animal spirits into uncharted Memory folds, spurs 

the soul to reflect on the novelty of a sensory idea.31 Upon reconstructing the wondrous idea, 

the intellect is thereby equipped to reflect on the fact that the corporeal idea is a reconstruction 

(rather than a fantasy). 

Incorporealists, capitalizing on the lack of corporeal traces of novelty, argue that the 

mature Descartes identified this intellectual act of reflection with remembering, and memories 

with the purely conscious products thereof. 

 

4.2. Incorporealism 

 The most radical incorporealist, John Morris (1969), has argued that the 

neurophysiology of the Treatise is untenable, and that Descartes came to realize its flaws in the 

1640s. In particular, Morris claims that “there is really no effective mechanical way to search for 

and retrieve a particular memory pattern from among those stored in the brain. The process of 

shuffling through millions of individual images in order to locate a particular image simply 

cannot be accomplished in the brief time that real human beings take” (1969: 460). Morris’s 

charge is a version of the frame problem which challenges any mechanistic theory of memory 

(Dennett 1978: 125). However, we have no reason to believe the frame problem occurred to 

Descartes, much less that Descartes abandoned his neurophysiology of memory on its basis.32 

 Morris also claims that “if sensory patterns are represented by folds in the cortex, there 

                                                 
30 Descartes also asserted that “it is also certain that objects of the senses that are novel affect the brain in 

certain parts where it is not normally affected; and that since these parts are more tender or less firm 

than those hardened through frequent agitation, the effects of the movements produced in them are 

thereby increased” (CSM 1:354; AT 11:382). Again, this implies only that wondrous impressions lead 

to unusually strong memory traces in new areas of the Memory (but not traces of novelty). 
31 The passion of wonder may be said to involve an ephemeral corporeal mark of novelty (on the basis of 

which the soul reflectively constructs an incorporeal mark of novelty), but not anything like a 

corporeal trace of novelty, since Descartes used the term 'trace' to mean a reasonably long-lasting 

mark. 
32 Sutton (1998) provides an intriguing defense of Descartes, against the frame problem, by arguing that 

Descartes had a proto-connectionist theory of memory. Sutton acknowledges that this proto-

connectionist theory is more inspired by than representative of Descartes’s actual thinking, but still 

musters strong textual support for the thesis that Descartes’s machine psychology was equipped to 

operate a proto-connectionist system of superpositional idea retention and retrieval. 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

18 

 

simply is not enough room in the brain to hold all the information that obviously must be 

stored there” (1969: 460). Unlike the frame problem, we do have evidence that Descartes 

considered this problem of raw storage space. Descartes's contemporary critics beat Morris to 

the criticism (MacIntosh 1983), and Descartes wrote to Mersenne in June 1640 that “there is no 

doubt that the folds of the Memory get in each other's way, and that there cannot be an infinite 

number of such folds in the brain; but there are still quite a number of them there” (CSMK 148; 

AT 3:84).33 However, the storage space problem did not faze Descartes. Two months later, 

Descartes told Mersenne that “I do not think that there has to be a very large number of these 

folds to supply all the things we remember, because a single fold will do for all the things which 

resemble each other” (CSMK 151; AT 3:143). This response is strengthened by Descartes's 

position that ideas are neither literally nor individually stored in the folds of the Memory; 

instead, the Memory is receptive to whole classes of ideas, disposing the pineal gland to 

recreate any and all patterns of flow towards memory traces. 

 Descartes also introduced intellectual memory partly to address the problem of storage 

space, writing that “in addition to the corporeal Memory, whose impressions can be explained 

by these folds in the brain, I believe that there is also in our intellect another sort of memory, 

which is altogether spiritual, and is not found in animals,” and that “it is this [intellectual 

memory] that we mainly use” (CSMK 151; AT 3:143). In addition to superpositional physical 

traces in the Memory folds, “the intellectual memory has its own separate impressions, which 

do not depend in any way on these folds. So I do not believe that the number of folds is 

necessarily very large” (CSMK 148; AT 3:84-85). Whereas corporealists invoke these passages to 

argue that Descartes attributed corporeal memories to animals, Morris invokes them to argue 

that Descartes’s mature view was that “pattern recognition is an intellectual operation” (1969: 

452). According to Morris, Cartesian brain architecture cannot support declarative memories, so 

Descartes renounced the corporeal reconstruction of sensory ideas. 

Véronique Fóti (2000) and Emanuela Scribano (2016), who endorse otherwise milder 

incorporealisms, join Morris in conflating intellectual memory with the intellectual recognition 

of ideas as reconstructions. However, Descartes did not take corporeal memory and intellectual 

memory to provide different routes to the reconstruction of the same ideas. Instead, he 

maintained that intellectual memory is “of a wholly different kind from [memory of material 

things].” In particular, we use intellectual memory to remember “intellectual things” (CSMK 233; 

AT 4:114)—universals—including innate ideas (CSM 2:48; AT 7:70; CSMK 228; AT 3:695) and 

the meanings of words (CSMK 336-337; AT 5:150).34 Descartes's appreciation of intellectual 

                                                 
33 Colin Chamberlain (personal communication) has plausibly remarked that Descartes should allow for 

the possibility of an infinite number of folds in the brain, given the infinite divisibility of Cartesian 

matter. Nevertheless, Descartes’s 1640 letters to Mersenne clearly imply that he held the brain to 

(dispositionally) retain a finite number of memories (as a contingent, empirically discernable fact, 

rather than an a priori rule of physics). 
34 Descartes also told Huygens that, qua immortal souls, we can use intellectual memory to remember the 

past after we have died (CSMK 216; AT 3:598). Many scholars plausibly discredit this passage in light 

of Descartes’s transparent desire to console his grieving friend, who had recently lost a brother. 

However, there is a straightforward way of making this claim consistent with Descartes’s other 

remarks about memory; he might have meant that disembodied souls can remember universal ideas 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

19 

 

memory entails his denial that mechanized sensitive souls are capable of yielding all the 

memories that ensouled humans enjoy. But intellectual memory could not serve to replace 

corporeal memory. Descartes held that “intellectual memory has universals rather than 

particulars as its objects, and so it cannot enable us to recall every single thing we have done” 

(CSMK 337; AT 5:150). We cannot use intellectual memory to reconstruct particular sensory 

ideas retained in our Memory as a result of past experiences. Only corporeal memory functions 

to recall sensory particulars. Cartesian pattern recall is not an intellectual operation; pattern 

recall is a bodily procedure involving the mechanical reconstruction of patterns of animal spirits. 

Incorporealists’ references to the increased role that Descartes granted intellectual 

memory in the 1640s are red herrings in the debate about sensory remembering. Descartes did 

become more willing to countenance intellectual memory, but never at the expense of corporeal 

memory. The intellectual recognition of reconstruction could not be reserved solely for 

intellectual memory, since it often takes particular sensory ideas—such as verses that occur to 

poets—rather than universals as its object. Nor did Descartes dispense with the need for purely 

intellectual memory by countenancing an intellectual component of sensory remembering. 

Intellectual memory enables souls to reconstruct universal ideas.35 Rather than being equivalent 

to intellectual memory, the intellectual recognition of reconstruction is a necessary component 

of each of the distinct faculties of intellectual memory and sensory memory.36 

At the beginning of his metaphysical career, Descartes suggested that the reflective 

power of the intellect plays a key role in the construction of corporeal memories. In the twelfth 

of his 1628 Rules, Descartes asserted that corporeal “memory is no different from imagination” 

(CSM 1:43; AT 10:416). The intellect, he wrote, “is one single power ... when applying itself 

along with imagination to the 'common' sense, it is said to see, touch etc.; when addressing itself 

to the imagination alone, in so far as the latter is invested with various figures, it is said to 

remember; when applying itself to the imagination in order to form new figures, it is said to 

                                                                                                                                                             
acquired and/or reflected upon during their lives. If anybody in the history of thought were to 

consider the immortal ability to recall the experience of having clearly and distinctly perceived 

universal truths genuinely consoling, it would be Descartes. 
35  Scribano is wrong to imply that “Descartes only once speaks of the intellectual memory of past 

experience working … without applying the mind to brain traces” (2016: 146). Descartes plainly 

invoked memories of innate ideas in the Meditations (CSM 2:48; AT 7:70) and 1643 correspondence 

with Elisabeth (CSMK 228; AT 3:695). 
36 Descartes once insinuated that the intellectual memory—being equivalent to the innate disposition to 

have intellectual ideas—is not really memory, writing in a letter to Hyperaspistes that “where purely 

intellectual things are concerned, memory in the strict sense is not involved; they are thought of just 

as readily irrespective of whether it is the first or second time that they come to mind—unless, as 

often happens, they are associated with certain names, in which case, since the latter are corporeal, we 

do indeed remember them” (CSMK 190; AT 425). Some commentators accordingly identify 

intellectual memories with innate ideas (Gaukroger 2002: 221-222; Simmons 2012: fn. 43). However, as 

Lisa Shapiro (2015) has argued, considering the role of memory in the Meditations provides good 

reason to differentiate intellectual memories—qua universal ideas that the thinker intellectually 

recognizes as having actually previously entertained—from mere innate ideas—qua universal ideas 

that (for all the thinker knows) previously existed only potentially and are perceived for the first time. 

I discuss my Shapiroesque take on intellectual memory more in §4.3. 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

20 

 

imagine or conceive; and lastly, when it acts on its own, it is said to understand” (CSM 1:42; AT 

10:415-416). Left to its own devices, the soul possesses only willpower and understanding. Brain 

physiology provides patterns of animal spirits as conduits through which the soul transforms 

the monolithic power of understanding into the diverse powers to sense, imagine and 

remember. The young Descartes concluded that “according to its different functions, then, the 

same power is called either pure intellect, or imagination, or memory, or sense-perception” 

(CSM 1:42; AT 10:415). On the incorporealist reading of this passage, it is the soul that senses, 

imagines, and remembers, even though it makes use of corporeal ideas to do so. 

 In the two 1648 letters to Arnauld, Descartes reaffirmed the essentials of his stance from 

twenty years prior. You remember an idea if and only if (1) the thing was previously presented 

to you (2) as a novel idea, (3) that idea made traces in your brain, (4) the traces in your brain led 

to the thing being presented to you again, and (5) you recognize that the thing was previously 

presented to you as a novel idea. The soul alone fulfills conditions (2) and (5). Again, Descartes 

wrote that “for the mind to recognize [novelty], I think that when these traces were first made it 

must have made use of pure intellect to notice that the thing which was then presented to it was 

new and had not been presented before; for there cannot be any corporeal trace of this novelty” 

(CSMK 356; AT 5:220). Physical changes to the Memory render a host of ideas more likely to 

recur, but cannot indicate that any particular idea has already been constructed on the pineal 

gland. Later, the reflective power of the soul understands certain corporeal images to be 

reconstructions of previously novel ideas. 

 Sutton and Clarke’s reading of Descartes’s treatment of wonder notwithstanding, 

Descartes was consistently clear that the ability to recognize novelty (and the lack thereof) 

belongs to the intellect alone. In the Conversations with Burman, Descartes suggested that “we 

have no memory of the thoughts we had in infancy … because no traces of these thoughts have 

been imprinted on the brain” (CSMK 336; AT 5:150; cf. CSM 2:246-247; AT 7:356-357). But two 

months later, in the first letter to Arnauld, Descartes took this suggestion back, writing that 

while the “confused sensations [of infancy] leave some traces in the brain, which remain there 

for life, that does not suffice to enable us to remember them” (CSMK 354; AT 5:192). Descartes 

did not mean that ideas from our infancy could not recur. He meant that such recurrences 

would not qualify as memories. For ideas formed during (or before) infancy to qualify as 

memories, “we would have to observe that the sensations which come to us as adults are like 

those which we had in our mother's womb; and that in turn would require a certain reflective 

act of the intellect, or intellectual memory, which was not in use in the womb” (CSMK 354-355; 

AT 5:192-193). 

 Fóti and Scribano consider the Rules and letters to Arnauld to support a mild form of 

incorporealism, according to which the metaphysically (rather than physiologically) minded 

Descartes identified remembering with intellectual acts that interpret corporeal reconstructions. 

Scribano writes that “when a mind is connected with the machine, it will not record brain traces 

as conscious recollections, but will interpret the traces … thereby transform[ing] them into 

conscious recollections” (2016: 146). On Scribano’s reading, sensory memories are the conscious 

outputs of these intellectual interpretations of corporeal ideas, as opposed to corporeal ideas 

themselves. Fóti likewise concludes that Descartes's mature verdict was that there is no such 

thing as corporeal memory, although the intellectual memory of sensory particulars has “a 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

21 

 

corporeal basis” (2000: 591). 

 Putting aside the erroneous conflation of intellectual memory and reflective recognition, 

these incorporealist conclusions are on the right track. Nevertheless, they are too quick to 

identify Cartesian remembering with a pure act of the intellect, and Cartesian sensory memories 

with the purely conscious products of that act. Descartes reaffirmed the machine psychology of 

the Treatise in the 1649 Passions—where he also reaffirmed that remembering involves 

reconstruction plus recognition (CSMK 343-344; AT 11:360). The central doctrine of all of his 

neurophysiological writings is that all sensory ideas—sensations, imaginings, and episodic 

memories—are corporeal patterns of animal spirits located in brain matter. Descartes 

consistently claimed that he had successfully mechanized these core states of the sensitive soul, 

and explained in detail how sensory ideas—including memories—represent (without 

resembling) external objects. Corporealists are right to insist that Descartes’s career-long 

commitment to a sensitive, flexible, and creative machine psychology indicates that the 

corporeal ideas produced by physiological mechanisms are components of—rather than mere 

prerequisites for—the process of remembering. 

Descartes’s distinction between direct and reflective thought also tells against Fóti- and 

Scribano-style incorporealism. In the second letter to Arnauld, Descartes clarified his point 

about infant memory: “if ever I wrote that the thoughts of children leave no traces in their brain, 

I meant traces sufficient for memory, that is, traces which at the time of their impression are 

observed by the pure intellect to be new” (CSMK 356; AT 5:220). Infants have the capacity for 

“direct and not reflective” phenomenal ideas, but lack the reflective capacity to distinguish 

between novel ideas and reconstructions. Descartes explained that “when an adult feels 

something, and simultaneously perceives that he has not felt it before, I call this second 

perception reflection, and attribute it to the intellect alone, in spite of its being so linked to 

sensation that the two occur together and appear to be indistinguishable from each other” 

(CSMK 357; AT 5:220-221). Descartes was careful to distinguish active intellectual reflection on 

ideas from the direct occurrence of those ideas in thought. Fóti and Scribano rightly take 

reflection on corporeal ideas to enable sensory remembering, but wrongly identify sensory 

memories with the products of this reflection. 

Throughout his career, Descartes identified sensory memories with patterned flows of 

animal spirits that represent objects to the mind-body union. When he wrote (in Rule 12, the 

Fifth Replies, and elsewhere) that the soul remembers sensory ideas, he must have meant that 

the soul wills the mind-body union to remember, is phenomenally presented with a direct 

representation of a sensory object via the mind-body union, and then reflectively recognizes the 

corporeal idea underlying that phenomenal presentation as a reconstruction. Intellectual acts of 

the soul are components of this process of remembering, but so are physiological processes of 

the body. If this is right, then sensory memories are not the conscious products of interpretation; 

they are the bodily objects of recognition. To be precise, sensory memories are physical patterns 

in the flow of animal spirits. However, they are not just physical patterns: they are physical 

patterns appropriately housed in mind-body unions. When a mind-body union remembers, the 

soul feels its body constructing an idea, and then actively renders that corporeal idea a memory 

by recognizing it as a reconstruction. 

 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

22 

 

4.3. An intermediate interpretation 

On this intermediate reading, sensory remembering is a uniquely human capacity 

comprising both physiological and intellectual components. Animals physically reconstruct 

sensory particulars and angels intellectually remember universals, but neither animals nor 

angels can remember where they left their keys. One pragmatic disadvantage of this 

interpretation is that it entails that a complete account of Cartesian memory requires a viable 

understanding of Descartes’s metaphysics of the union of body and mind. Then again, this 

entailment may indicate why Descartes wrote only enigmatic scraps about the metaphysics of 

memory: he infamously failed to articulate the nature of mind-body unions. 

There are almost as many ways of articulating human nature on Descartes’s behalf as 

there are Descartes scholars. Some scholars take Cartesian bodies and minds to be causally 

connected by either divine fiat or (divinely instituted) psychophysical bridge laws (Wilson 1978; 

Garber 1993; Loeb 2005; Koivuniemi & Curley 2015); other scholars take the mind-body union 

to be a third substance, metaphysically distinct from body and mind (Broughton & Mattern 

1978; Clarke 2003; Hoffman 2009; Cottingham 2008). Most recently, Alison Simmons has argued 

that Descartes considered the mind-body union to “fall outside the domain of any humanly 

possible metaphysics.” On Simmons’s reading, “Descartes’ position is that we have an internal-

sense access to ourselves as mind-body unions that provides a rich phenomenology of 

embodiment that is of both theoretical and practical interest to a Cartesian study of human 

nature,” but does not furnish clear and distinct ideas useable in the search for metaphysical 

truths (2017: 2). 

Fortunately, my intermediate interpretation of Descartes’s view of sensory memory does 

not rely on any particular metaphysics of the mind-body union—or even on the notion that a 

metaphysics of the mind-body union is humanly possible. On the contrary, understanding 

Cartesian sensory remembering as corporeal reconstruction paired with intellectual recognition 

sheds light on whichever conception of mind-body relations is latent in Descartes’s philosophy. 

If Descartes is a dualist, then corporeal ideas cause direct phenomenal ideas, which prompt the 

intellect to reflect on their status as sensations, memories, or mere imaginings. If Descartes is a 

trialist, then corporeal ideas appear directly in consciousness, and thereby prompt the intellect 

to reflect on their status. If Descartes is a metaphysical quietist about the mind-body union, then 

human beings mysteriously feel corporeal ideas (God only knows how), which prompts the 

intellect to reflect on the status of those ideas. (I will use the term ‘phenomenal face’ to refer to 

the second grade conscious idea that directly accompanies a first grade corporeal idea housed in 

a mind-body union, whether as effect, flip side, or mysterious feeling.) 

In any case, Descartes identified sensory memories with corporeal ideas reconstructed 

thanks to memory traces dispositionally retained in the folds of the Memory. Nevertheless, he 

denied animals the capacity to remember. Cartesian critters can’t remember because sensory 

remembering is an essentially psychophysiological act. Sensory remembering is essentially 

psychophysiological because intellectual reflection is a necessary companion to sensory 

memories, which are themselves patterns—different release points, subtlety gradients, volumes, 

and velocities—of animal spirits in human bodies. Whether these corporeal ideas are causally 

connected or token-identical to their phenomenal faces, their felt effect on (or appearance in) 

consciousness is crucial for securing their status as memories. There is no physiological mark 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

23 

 

that distinguishes memories from mere imaginings. Only the intellectual power of the soul—

reflecting on feelings present in consciousness—is capable of distinguishing the case in which 

you remember that your keys are in your coat pocket from the case in which you merely 

imagine that your keys are in your coat pocket. Recognition bestows a nonphysical property to 

a reconstruction: the relational property of being appreciated as a non-novel idea by the 

rememberer. (If Descartes was a trialist, and physiological ideas housed in mind-body unions 

are identical to their phenomenal faces, then memories have phenomenal properties as well.37) 

Indeed, on my reading, Descartes held that brain processes become memories only upon 

achieving this relational distinction. Nonhuman animals are incapable of appreciating the 

(non)novelty of their ideas. It follows that animals do not qualify as true rememberers, even 

though their brains sometimes execute physiological processes that physically mirror processes 

in human brains that constitute memories when intellectually recognized as such. 

Descartes sensibly stressed that humans need not grasp the details of our own 

neuroanatomies in order to recognize our corporeal ideas as reconstructions.38 When willing 

ourselves to remember, “we are not conscious of the manner in which our mind sends the 

animal spirits into particular nerves; for that depends not on the mind alone but on the union of 

the mind with the body.” The pineal gland automatically obeys the conscious will “because of 

the appropriate way the body is constructed, of which the mind may not be aware, and because 

of the union of the mind with the body, of which the mind is certainly conscious” (CSMK 357; 

AT 5:222). When the pineal gland reconstructs a corporeal idea at the will’s bequest (or via idle 

swaying), the phenomenal face of that idea occurs directly in thought (whether via causation or 

token-identity). The soul then acts a second time, and reflectively recognizes the idea as a 

reconstruction—something the mind-body union has sensed before—without understanding 

the idea’s precise physical mode. This confused recognition of reconstruction transforms the 

                                                 
37  My intermediate reading may fit best with trialism about the mind-body union. Intuitively, 

phenomenal faces—second grade images formed in the mind’s eye—are components of (rather than 

caused by) sensory memories. Descartes seems to support this idea in the correspondence with 

Arnauld, reporting that memories “occur in our thought” and are “presented to … the pure intellect” 

(CSMK 356; AT 5:220), as well as the Sixth Replies, where he writes that both first and second (but not 

third) grade sensations “should be referred to the sensory faculty, if we wish to distinguish it clearly 

from the intellect” (CSM 2:295; AT 7:437). Indeed, this identity is so intuitive—surely my memories 

intrinsically feature (not merely efficiently cause ideas of) phenomenal colors, sounds, and odors—

that some readers may think trialism and my intermediate reading will stand or fall together. In this 

connection, Gary Hatfield (personal communication) has suggested that corporeal ideas provide the 

specific contents of sensory ideas, phenomenal faces bring those contents under a conscious aspect, 

and then reflective recognition transforms that whole psychophysiological package into a memory. (I 

myself interpret Descartes as either this kind of trialist or a quietist, depending on the day of the 

week.) However, I am not sure Descartes’s texts definitively settle this issue on the side of intuition. It 

may be that, contrary to intuition, Cartesian people consciously experience the direct phenomenal 

effects of their sensations, imaginings, or sensory memories—which, strictly speaking, are to be 

identified with first grade ideas alone. Regardless, sensory remembering—the process of crafting 

sensory memories—always comprises first grade, second grade, and third grade sensations. 
38 I am hereby putting more of my metaphysical cards on the table, insofar as I reject a strong version of 

the ‘direct inspection’ model of Descartes’s theory of sensation (cf. Chignell 2009). 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

24 

 

corporeal idea into a memory. 

Admittedly, Descartes failed to provide a wealth of insight into precisely how this 

intellectual recognition works. As discussed at the end of §4.1, when a new idea is constructed, 

the attentive soul will reflect on its novelty, and thereby become disposed to recognize a 

reconstruction of that idea as a reconstruction. In a long 1641 letter to Hyperaspistes, which 

both Fóti and Scribano cite in support of the conflation of intellectual memory and recognition, 

Descartes remarked in passing that “where purely intellectual things are concerned, memory in 

the strict sense is not involved,” since “purely intellectual things … are thought of just as 

readily irrespective of whether it is the first or second time that they come to mind” (CSMK 190; 

AT 3:425). In 1640 correspondence with Mersenne (CSMK 148; AT 3:84; CSMK 151; AT 3:143) 

and 1644 correspondence with Mesland, however, Descartes explicitly avowed memory “traces 

which remain in the mind itself” (CSMK 233; AT 4:114), enabling intellectual memory. Both 

incorporealists and corporealists usually take these letters to be incompatible, but a 

straightforward reading unifies them. Strictly speaking, memory involves reconstruction. But 

the memory of “purely intellectual things”—that is, innate ideas—does not (strictly speaking) 

involve construction at all, since innate ideas are always equally accessible within consciousness. 

Hence Descartes’s remark to Hyperaspistes. So what are the traces of intellectual ideas which 

Descartes told Mersenne and Mesland exist in the mind? They are incorporeal traces of 

intellectual recognition, which come in a couple of varieties. Some are traces of recognition of 

the meanings of words (CSMK 336-337; AT 5:150). Others are traces of recognition of novelty: 

traces of having previously consciously experienced an idea. 

 In the same letter to Mesland, Descartes delineated the sources from which the intellect 

passively receives ideas: “partly from objects which come into contact with the senses, partly 

from impressions in the brain, and partly from prior dispositions in the soul and from 

movements of the will” (CSMK 232; AT 4:113-114). Given the impossibility of corporeal traces of 

novelty, conscious ideas of novelty must have their source in mental dispositions. I thus suspect 

that Descartes believed incorporeal traces of novelty, like innate ideas, to “exist within us 

potentially” (CSM 1:361; AT 8B:361). On an influential interpretation of Cartesian ‘potential 

existence,’ innate ideas are latent in our conscious thoughts (McRae 1972; Simmons 2012). 

Because they are innately potentially existent, innate ideas are always latent in our conscious 

thoughts. Incorporeal traces of novelty, on the other hand, become potentially existent through 

our interaction with the world. Recognizing an innate, adventitious, or factitious idea as novel 

changes our thinking. When we recognize an idea as novel, the intellectual ability to recognize a 

reconstruction of that idea becomes latent in our thoughts, available to be dredged up by 

reflection. The reflective recognition of novelty thereby dispositionally enables both intellectual 

memory and sensory memory. 

Descartes attributed both willful recall and reflective recognition to the soul, but he 

attributed transformed sensory memories themselves to the body, despite the three “being so 

linked … that [they] occur together and appear to be indistinguishable from each other” (CSMK 

357; AT 5:221). This psychophysical separation of the components of remembering bears a 

striking similarity to Descartes’s denial of sentience to animals. He infamously wrote that “I do 

not explain the feeling of pain without reference to the soul. For in my view [the feeling of] pain 

exists only in the understanding. What I do explain is all the external movements which 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

25 

 

accompany this feeling in us; in animals it is these movements alone which occur, and not pain 

in the strict sense” (CSMK 148; AT 3:85). My contention is that Descartes propounded a similar 

understanding of memory. In the Treatise, Descartes explained all of the physiological 

movements that reconstruct ideas and accompany the direct feeling of remembering in humans. 

In soulless animals (and the fabled, soulless humans of the Treatise) it is these movements alone 

which occur, and not memory in the strict sense. Just as only ensouled and embodied humans 

can feel their sensations as pains—and thereby experience painful sensations—only ensouled 

and embodied humans can entertain the phenomenal faces of corporeal ideas, recognize those 

ideas as reconstructions, and thereby remember. Painful sensations and memories are patterns 

of animal spirits leaving the pineal gland. But they are also more than that: they are corporeal 

ideas that bear an intimate relationship to particular passivities and activities of the soul. 

Physiological ideas become memories only when accompanied by the soul's conscious 

recognition that the mind-body union is remembering, just as physical sensations become 

painful only when accompanied by human suffering.39 

Corporealists and incorporealists alike might reasonably wonder what kind of 

transformation—what kind of becoming a memory—is posited by this intermediate interpretation. 

Does recognition actually transform reconstructed corporeal ideas into memories in some 

genuine way? Or is the difference merely a matter of how an independent theorist might 

designate each case?40 The short answer is that the recognition of reconstruction functionally 

transforms ideas, and that this functional transformation is a variety of genuine transformation. 

The necessary linkage of sensory memories with the reflective recognition of them as 

reconstructions causes memories to function genuinely differently than mere reconstructions, 

even though their formal physiological qualities (and direct phenomenal effects or qualities) 

remain unaltered. 

When a human being feels pain, their corporeal sensation is functionally transformed 

into a state that the human can will themselves to avoid—or grit their teeth and bear. In the 

same way, when a human being recognizes a reconstruction, their corporeal idea is functionally 

transformed into a memory: the sort of sensory idea that the thinker can treat as old hat—or 

attempt to forget. In Descartes’s view, animals cannot do anything for which sensation and 

imagination fail to suffice. Humans, when they remember, do an extra thing with their sensory 

ideas: they use them to reminisce about the past. A poet who actually remembers verses can 

recite them with a distinctly human humility, and either attribute them to the true artist or 

intentionally plagiarize. Indeed, both sensory memories and intellectual memories play 

important roles in Descartes’s metaphysical method that mere imaginings could not possibly 

play. Skepticism can give birth to certainty only insofar as doubting enables the Cartesian 

Meditator to ascertain the provenances of her ideas (Curry 2016). Memory serves the Meditator 

as a way of reviewing prior meditations, as an epistemological scapegoat—received ideas from 

childhood are epistemologically problematic not merely because they are reconstructed, but 

                                                 
39 There is a disanalogy here, in that the feeling of pain is direct, whereas the recognition of a memory as 

such is reflective. The feeling of a sensory memory, qua direct phenomenal face of a corporeal idea, is 

indistinguishable from the feeling of a sensation or imagining. What makes a memory feel like a 

memory is the second-order recognition that it is not a novel idea. 
40 Thanks to an anonymous reviewer for posing these (paraphrased) questions. 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

26 

 

because the Meditator recognizes them as reconstructed from her dogmatic Aristotelian 

schooling (CSM 2:12; AT 7:17)—and as a pragmatic savior in the search for truth—since the 

Meditator cannot spend all of her time clearly and distinctly perceiving truths, she must 

sometimes instead recognize memories as reconstructions of ideas that she has previously 

clearly and distinctly perceived (CSM 2:48; AT 7:70; CSMK 228; AT 3:695). In order to play each 

of these functional roles, it is crucial that remembering both reconstruct previous ideas and 

occasion a distinctive phenomenology: the felt act of realization that a memory is a recollection, 

rather than a mere figment of the imagination. 

On the reading I have advanced, a Cartesian idea is a memory only if it functions as a 

memory, and functioning as a memory necessarily involves being recognized as a 

reconstruction. Imagining (the past, present, or future) constructs ideas for processing. 

Remembering constructs ideas for reminiscent processing as recognized reconstructions. The 

fabulous machines Descartes envisaged in the Treatise function in intricate—sensitive and 

imaginative—ways, but never as though they (not just reconstruct but) truly remember. 

 

5. Conclusion 

I have argued that Descartes held the following account of declarative memory: to 

remember is to reconstruct an idea that you intellectually recognize as a reconstruction. To have 

an intellectual memory is to intellectually reconstruct a universal idea that you recognize as a 

reconstruction, and to have a sensory memory is to neurophysiologically reconstruct a 

particular idea that you recognize as a reconstruction. 

More specifically, I have taken up the question of whether Descartes thought animals 

remember. Corporealists say 'yes’; incorporealists say ‘no’ (or, to be precise, ‘not at first, then he 

did for a while, but ultimately no’). I have proposed that Descartes never thought animals 

remember. He held that Cartesian critter brains feature the Memory. Moreover, animal brains 

physiologically reconstruct ideas that parallel the sensory memories of humans in all physical 

respects. Nevertheless, the corporeal ideas that flow from the pineal glands of soulless animals 

are not memories, because memories necessarily feature a nonphysical property: the relational 

property of being recognized as a reconstruction by the rememberer. Descartes endorsed a 

neurophysiology that could not differentiate the psychological faculties of memory and 

imagination. Only intellectual reflection on the fact that ideas are reconstructions transforms 

those ideas into memories. Cartesian sensory memory, like Cartesian suffering, is a uniquely 

human capacity that arises “from the union and, as it were, intermingling of the mind with the 

body” (CSM 2:56; AT 7:81). It takes a chimera like you—a mishmash of body and soul—to really 

remember where you’ve left your keys. 

 

Acknowledgements 

This article results from a graduate career whiled away arguing about Descartes with Gary 

Hatfield. (I should have been writing my dissertation.) Thanks also to Patrick Ball, Dave Barack, 

Justin Bernstein, Colin Chamberlain, Andrew Chignell, Amelia Curry, David Curry, Karen 

Detlefsen, Martha Farah, Andree Hahmann, Nabeel Hamid, Jorge Secada, Jordan Taylor, and 

two extraordinarily thorough and thoughtful anonymous reviewers, as well as audiences at the 

2015 Penn Workshop in the History and Philosophy of Science and 2017 Representing Reality 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

27 

 

conference at SUNY Potsdam.  

 

This research was partially supported by a Mellon/ACLS Dissertation Completion Fellowship 

from the American Council of Learned Societies. 

 

References 

Addis, Donna Rose, Alana T. Wong, and Daniel L. Schacter (2007) “Remembering the past and 

imagining the future: common and distinct neural substrates during event construction 

and elaboration”, Neuropsychologica. 45(7): 1363–1377. 

Bernecker, Sven (2010) Memory: A Philosophical Study. OUP. 

Broughton, J. & R. Mattern (1978) “Reinterpreting Descartes on the Notion of the Union of Mind 

and Body”, Journal of the History of Philosophy 16: 23-32. 

Chignell, Andrew (2009) “Descartes on Sensation: A Defense of the Semantic-Causal Model”, 

Philosophers’ Imprint. Vol 9, No 5, 1-22. 

Clarke, Desmond (2003) Descartes's Theory of Mind. OUP 

Curry, Devin Sanchez (2016) “Cabbage à la Descartes”, Ergo, Volume 3, No. 24, 609-637 

Cottingham, John (2008) Cartesian Reflections: Essays on Descartes’ Philosophy. OUP 

Debus, Dorothea (2010) “Accounting for Epistemic Relevance: A New Problem for the Causal 

Theory of Memory”, American Philosophical Quarterly, Vol 47., No. 1, 17-30 

Dennett, Daniel (1978) Brainstorms: Philosophical Essays on Mind and Psychology. MIT Press. 

Descartes, René (1964–1976) Oeuvres de Descartes (Vols. 1-11). Charles Adam and Paul Tannery 

(Eds.). CNRS/J. Vrin. 

Descartes, René (1985-1991) The Philosophical Writings of Descartes, ed. Cottingham, Stoothoff, 

Murdoch, and Kenny. Cambridge University Press. 

Descartes, René (1996) Treatise of Man, Thomas Steele Hall, translator. Prometheus Books. 

Des Chene, Dennis (2000) Life’s Form: Late Aristotelian Theories of the Soul. Cornell UP. 

Des Chene, Dennis (2001) Spirits and Clocks: Machine and Organism in Descartes. Cornell UP. 

Drieux, Philippe (2016) “Machine and Communication of Corporeal Dispositions in 

Descartes and La Forge: The Mysterious ‘Article 83’ of  L’Homme and La Forge’s 

Comments”. In Antoine-Mahut D., Gaukroger S. (eds) Descartes’ Treatise on Man and 

its Reception. Springer, 127-138 

Edwards, Michael (2003) Time and the Science of the Soul in Early Modern Philosophy. Brill 

Fernández, Jordi (2008) “Memory, Past and Self”, Synthese, 160(1): 103–121. 

Fóti, Véronique (2000) “Descartes’ intellectual and corporeal memories.” In S. Gaukroger, J. 

Schuster, and J. Sutton (eds.), Descartes’ Natural Philosophy, 591-603. Routledge. 

Garber, Daniel (1993) “Descartes and Occasionalism” in (Nadler, Steven, ed.) Causation in early 

modern philosophy: Cartesianism, occasionalism, and pre-established harmony. Pennsylvania 

State Press, 9-26. 

Gassendi, Pierre (1659a) "Esse Animos Hominum Immortaleis, Contra Epicurum" 

Animadversiones in Decimum Librum Diogenis Laertii, Vol 1., Lyon: G. Barbier. 

Gassendi, Pierre (1659b) Syntagma philosophiae Epicuri cum refutationibus dogmatum quae contra 

fidem christianam ab eo asserta sunt, Lyon: G. Barbier. 

Gaukroger, Stephen (2000) “The resources of a mechanist physiology and the problem 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

28 

 

 of goal-directed processes” In S. Gaukroger, J. Schuster, and J. Sutton (eds.), Descartes’ 

Natural Philosophy, 630–658. London: Routledge. 

Gaukroger, Stephen (2002) Descartes' System of Natural Philosophy. CUP 

Goclenius, Rudolph (1613) Lexicon Philosophicum. Marburg 

Hatfield, Gary (2000). “Descartes’ naturalism about the mental.” In S. Gaukroger, J. Schuster, 

and J. Sutton (eds.), Descartes’ Natural Philosophy, 630–658. London: Routledge. 

Hatfield, Gary (2007) "The Passions of the Soul and Descartes’ Machine Psychology," Studies in 

History and Philosophy of Science 38 (2007), 1-35. 

Hatfield, Gary (2008) “Animals.” In J. Carriero and J. Broughton (eds.), Companion to Descartes, 

Blackwell, 404-425. 

Hatfield, Gary (2012). “Mechanizing the sensitive soul.” In Gideon Manning (ed.), Matter and 

Form in Early Modern Science and Philosophy, 151–186. Leiden: Brill. 

Hatfield, Gary (2016) L'Homme in Psychology and Neuroscience. In Descartes' Treatise on Man 

and Its Reception, ed. Stephen Gaukroger and Delphine Antoine-Mahut. Springer, 269–85. 

Hobbes, Thomas (1651) Leviathan. Original folio 

Hoffman, Paul (2009) Essays on Descartes. OUP. 

Jorgensen, Larry (2011) “Leibniz on Memory and Consciousness”, British Journal for the History 

of Philosophy, Vol 19, Iss 5, 887-916. 

Joyce, Richard (1997) “Cartesian Memory,” Journal of the History of Philosophy, Volume 35, 

Number 3, 375-393 

Kirwan, C Brock; Ashby, Stefania R; Nash, Michelle I (2014) “Remembering and imagining 

differentially engage the hippocampus: a multivariate fMRI investigation.” Cognitive 

neuroscience vol. 5 (3-4) 177-85 

Koivuniemi, M. & E. Curley (2015) “A Kind of Dualism,” Oxford Studies in Early Modern 

Philosophy 7, OUP, 83-122  

Landormy, Paul (1902) “La Mémoire corporelle et la mémoire intellectuelle dans la philosophie 

de Descartes”, bibliothéque du Congrés Internationale de Philosophie 4, 259-98 

La Forge, Louis de (1997) Treatise on the Human Mind, trans. Desmond M. Clarke, Kluwer. 

Leibniz, G.W. (1996) New Essays on Human Understanding, trans. P. Remnant and J. Bennett. CUP 

Locke John (1690/1975), Essay concerning human understanding, (P. Nidditch, ed.), OUP. 

Loeb, Louis (2005) “The Mind-Body Union, Interaction, and Subsumption” in Early Modern 

Philosophy: Mind, Matter, and Metaphysics (C. Mercer and E. O’Neill, eds.), OUP, 65-85 

MacIntosh, JJ (1983) “Perception and Imagination in Descartes, Boyle, and Hooke”, Canadian 

Journal of Philosophy 13, 327-52. 

Malebranche, N. (1674-1675/1997). The Search after Truth. (T. M. Lennon and P. J. Olscamp, 

Trans.). Columbus: Ohio State University Press, 1980; Cambridge: Cambridge University 

Press. 

Martin, C. B. & Deutscher, M. (1966) “Remembering,” Philosophical Review, 75: 161-96. 

McRae, Robert (1972) “Innate Ideas”, in Cartesian Studies (ed. R.J. Butler), OUP, 32–54 

Michael, E., & Michael, F. S. (1988). “Gassendi on sensation and reflection: A non-cartesian 

dualism,” History of European Ideas, 9(5), 583. 

Michael, E. & Michael, F. S. (1989) “Corporeal Ideas in Seventeenth-Century Psychology,” 

Journal of the History of Ideas, Vol. 50, No. 1, 31-48 



Please cite published version: https://doi.org/10.1016/j.shpsa.2018.03.001 

29 

 

Michaelian, Kourken (2016). Mental Time Travel: Episodic Memory and Our Knowledge of the 

Personal Past. MIT Press 

More, Henry (1659). The immortality of the soul, so farre forth as it is demonstrable from the knowledge 

of nature and the light of reason, Preface section 10, London. 

Morris, John (1969). “Pattern recognition in Descartes’ automata.” Isis 60(4): 451–460. 

Ott, Walter (2017) Descartes, Malebranche, and the Crisis of Perception. OUP 

Schacter, Daniel, Donna Rose Addis, Demis Hassabis, Victoria C. Martin, R. Nathan Spreng, 

and Karl K. Szpunar (2012) “The Future of Memory: Remembering, Imagining, and the 

Brain,” Neuron 76, 677-694 

Scribano, Emanuela (2016) “La Forge on Memory: From the Treatise on Man to the Treatise on 

the Human Mind”, In Descartes' Treatise on Man and Its Reception, ed. Stephen Gaukroger 

and Delphine Antoine-Mahut (New York: Springer), 139-254. 

Sepper, Dennis (1998) Descartes' Imagination. Berkeley: Cal Press. 

Shapiro, Lisa (2015) “Memory in the Meditations”, Res Philosophica, Volume 92, Issue 1, 41-60 

Simmons, Alison (1999) “Are Cartesian sensations representational?” Nous 33(3): 347–369. 

Simmons, Alison (2012) “Cartesian Consciousness Reconsidered”, Philosophers’ Imprint, Vol 12, 

No. 2, 1-21 

Simmons, Alison (2017) “Mind-Body Union and the Limits of Cartesian Metaphysics”, 

Philosophers’ Imprint, Vol 17, No. 14, 1-36 

Squire, Larry R. (2009) “Memory and Brain Systems: 1969–2009”, Journal of Neuroscience, 29(41): 

12711–12716. 

Sutton, John (1998) Philosophy and Memory Traces. CUP. 

Sutton, John (2016) “Memory”. In Larry Nolan (Ed.), The Cambridge Descartes Lexicon, 490-

492. Cambridge: Cambridge University Press. 

Sutton, J. & C. Windhorst (2009) “Extended and Constructive Remembering: Two Notes on 

Martin and Deutscher,” Crossroads, Vol. 9, Iss. 1, 79-91 

Wilson, Margaret (1978) Descartes. Routledge.