D. THE NEW PHYSIOLOGY AND THE RECONCEPTION OF PHANTASIA

At the end of chapter 13 of Le Monde, Descartes promises to give a further description of how the motions of light cause human sensations, but that is not taken up except in L'Hornme, which begins by extending the fable to include new human beings "composed, as we are, of a soul and a body" (AT XI 119). The account given there of the physiology of perception is in some small but significant ways at odds with the imagination constructed in Le Monde.

In our reconstruction of Descartes's conception of imagination on the threshold of Le Monde we did not pay special attention to the physiological questions involved. As I have pointed out earlier, Descartes began his anatomical researches no later than early 1630, and other evidence points to

a beginning in 1628 or 1629.^[12] To flesh out the hypothetical psychophysiology of Rule 12 it would have been necessary for Descartes to study the medium by which the motions that bear objective information are carried to the brain and also the organ or organs of the brain that receive the impression. One might expect that a fuller account would say more about how and by means of what the impression on the sense organ is passed along.

The physics of Le Monde does not directly explain how patterns or images are transmitted, however. The physics of light is a mechanics of pushing: the action of luminous bodies, which consist of the finest, fiery element, is to push the particles of the second, airy element by virtue of those luminous bodies' very rapid rotatory or vortical motion. Light is thus not itself a particle but a pressure carried through the particles of the second element. Because the universe is a plenum and because Descartes conceives of the particles of the second element as immediately touching one another, this pressure is instantaneously transmitted in all directions from the luminous body. One can imagine a universe of billiard balls packed together, representing the second element, and the luminous body pressing out simultaneously over the entire surface where it contacts the second element. The pressure results in not so much a motion as a tendency to motion. This tendency to motion is said to follow the same laws as motion itself and so proceeds in straight lines. Thus, without being a particle traveling through space or an undulation of a medium, light is propagated in the familiar way. Descartes leaves unexplained the details of how this tendency to motion is refracted and reflected in the third and second elements; however, he does apparently refer to the Dioptrics, noting that the explanation given there, which uses the model of collisions of hard bodies to explain reflection and the analogy of a tennis racket striking a ball for refraction, has to be modified to the circumstances of the plenum and the propagation of light as pressure.^[13] Lacking the mechanisms that would clarify this, we may doubt that a plausible modification could be made, but apparently Descartes saw no problem in principle.^[14] Thus the Dioptrics read in conjunction with Le Monde gives the assurance that the theory of focusing rays on the retina (first presented by Johannes Kepler) can be easily converted into a theory of the focusing of lines of pressure. This means that, at least up to the retina, physics can give a complete account of how

[12] See Seppcr, "Descartes and the Eclipse of Imagination,' and "Ingenium, Memory Art, and the Unity of Imaginative Knowing in the Early Descartes."

[13] The account of light is given in chaps. 13 and 14 of Le Monde (AT XI 84-103). The allusion to the Dioptics is at AT XI 102.

[14] The explanation would have to account for how the light pressure's direction of motion is changed at the surface between one body and another.

information about the configuration of objects in the world can be transmitted to the eye and reconstituted there without any significant loss of precision.

What happens thereafter? In L'Homme the answer depends on anatomy. The retina is dotted with nerve ends, which have corresponding ends at the interior surface of the cavity in which the conarion, or pineal gland, is to be found. Whenever the retinal nerves are moved by light pressure

they pull at the same instant the parts of the brain from which they come, and open by the same means the entrances of certain pores that are on the internal surface of this brain, through which the animal spirits that are in its concavities immediately begin to take their course, and go on to spread through these into the nerves, and into the muscles, which serve to cause, in this machine, movements entirely similar to those to which we are naturally incited, when our senses are touched in the same way. (AT XI 141)

This explanation is predicated on the composite structure of the nerve fiber. It is conceived as an essentially hollow tube with a fiber running down the center. The fiber transmits a pull that originates in the sense organ, while the hollow is filled with animal spirits that are responsible for actions initiating within the body. The pull on the fiber at the retina opens pores in the interior cavity of the brain, which act of opening sets off a flow of spirits from the brain cavity into the nerve hollow; the flow of spirits causes the muscular contractions. But of course not just one nerve is involved in any sensation but a multitude. The retina has many such nerve ends; each operates independently of its neighbors. An "image" falling on the retina is thus broken up into myriad discrete channels of mechanical action by the different nerve endings upon which its various parts fall.^[15]

There are two peculiarities in L'Homme's account of nerve action that deserve special attention: the nerves operate by pulling rather than pushing, and there is a complicated nerve structure that contrasts with the simplicity of the Regulae's account.

Given the pressure mechanism of light, there is something very odd about describing the effect on the nerve as a pull rather than a push. Up to the retinal surface the action of light is pushlike, a pressure. Even in Rule 12 the action between the eye and the phantasia was described as an impression. For some as yet inapparent reason, Descartes decided in L'Homme on a mechanical account that includes a reversal in the direction of tendency to motion. This does not contradict the foundations of Des-

[15] This does not in principle rule out all interaction of the parts of the image, however. For example, whether or not the pull on a given nerve opens a pore in a certain way could depend on amplification or inhibition by the other nerves; or the flows of the spirits in the brain cavity could be programmed to compensate for the decomposition of the image at the nerve ends in the sense organ. But Descartes did not discuss such things.

cartes's physics, since in Le Monde he carefully distinguishes between the quantity and the direction of motion. The former must be preserved in the system, but direction is determined independently, and so in principle it is possible for a quantity of motion to be reoriented in any direction without addition to or subtraction from that quantity. We might simply conclude that at the sensory organ the structure of the nerves is so constituted that a pushing motion is converted into a pulling one.

Yet there is still something counterintuitive about this. Why not simply maintain a pushing force and, if it proves necessary for Descartes's further account of physiological functioning, allow for, say, a motion-reversing, valvelike action at the pores in the brain cavity? Did Descartes lose confidence in the ability of the nerves to transmit a pushing motion?

A possible explanation arises from a comparison with the Rule 12 account. Rule 12 had talked in terms of an impression on the first opaque membrane, the retina, an impression that was instantly carried to the common sense that in turn impressed it on the phantasia. What was impressed was seallike; the impression was the figure or image pressed into the membrane or body by the act of light. But can this happen physically, given the body's anatomy? If Rule 12's scheme is to hold up, we must conceive of something connecting the retina to the common sense. Even though Rule 12 mentions nerves only in connection with motion and not perception, it seems reasonable to believe that in the Regulae Descartes thought of something nervelike as transmitting the impression. But there is no evidence at all that at the time of writing Rule 12 Descartes attributed to the nerves a complex structure. He mentions no animal spirits, no hollows, no tugging central fibers. The character of the explanation in terms of impression strongly suggests that, insofar as he had even raised the question for himself, he thought of the medium as something more or less rigid that is nevertheless capable of preserving the figure of an original impression. What sort of stuff would this have to be? It could not be too solid, like a staff, since a pressure applied to any single point at one end would be transmitted indiscriminately to the other, as though the pressure had been applied equally over the entire cross section. This possibility, which would be readized in the case of any normal pressure applied to a solid, would entail a loss of information: the specific point of pressure application would be irrecoverable at the opposite end. The staff could not be too liquid, either, since then the motion would be immediately dispersed in all directions. The fact is that there is no ordinary material that will achieve the desired effect, if pressure or motion is what is transmitted.^[16]

[16] If the nerves were made out of transparent material or Descartes's second element, they could of course function as optical fibers transmitting light and images. But this would be acceptable only in a contrafactual fable.

Thus the properties of matter defeat the holistic transmission of an impression or image, which was a desideratum of the Rule 12 hypothesis. Undoubtedly on the basis of anatomical study, Descartes came to reject holistic transmission in favor of what one might call analytic transmission. Besides the question of how a uniform nerve could transmit an integral but differentiated image, there was the problem that the optic nerve is composed of many smaller nerves and that the retina has not a single nerve receptor but multitudinous receptors. At the very best, then, an image would have to be broken down into many parts, transmitted along separate channels (nerve strands) through the optic nerve to the brain, and there somehow reassembled into a whole. It is clear even from a schematic consideration that there will be a loss of detail unless the receptors are so packed together that no spaces are left between. But even if they are, one would again run into the problem of the inability of a nerve or nerve strand to transmit images (or parts of images) integrally. The solution is to forget about the image or image impression as what is transmitted and instead to substitute some kind of motion conveyed as a signal rather than as an image. Once again, an instantaneous pushing motion will not do, since the nerves are simply too soft and pliant to fill the bill. The alternative is to have a pulling motion, which can be more or less instantly transmitted in nonrigid materials under appropriate circumstances, for example, a cord already pulled fairly taut. Since the hollow tubular structure that Descartes ascribed to the nerve is not one that any modern anatomist would be able to verify, it seems likely that he interpreted into the anatomy he knew a variant of Stoic pneumatic physiology that would allow the nerves to perform the double duty that the Regulae had left in the dark: the central fiber transmits the pull, and the rest of the chamber serves as a conduit for animal spirits that produce self-initiated body motion.

Surprisingly, despite the primacy of motion, Descartes never fully abandoned the language of images. In both L'Homme (AT XI 175-177) and the Dioptrics (AT VI 128-130) he describes how the opening and closing of the pores at the ends of the optical nerve fibers in the brain concavity trace a figure on the internal surface of the concavity corresponding to the image on the retina; the differential spirit flow from the pineal gland that is induced by the opening and closing of the pores produces the trace of another corresponding figure on the pineal gland. Indeed, he reserves the term 'idea' in the proper sense for the trace on the pineal gland.

Now, among these figures, it is not those that impress themselves on the organs of the exterior senses, or in the interior surface of the brain, but only those that trace themselves in the spirits on the surface of the gland, where the seat of the imagination, and of the common sense, is, that must be taken for the ideas, that is for the forms or images that the rational soul will consider

immediately when, being united to this machine, it will imagine or sense some object. (AT XI 176-177; emphasis in original)

More than a decade later, in the Passions of the Soul (1649), the duality of motion and image recurs; Descartes describes the process of sensation first as a transmission of motion but later as a transmission of images (cf. AT XI 338-339, 346 with 355-356).

What remains hidden from view is the way in which the soul is joined to the pineal gland and whether the interior or only the surface of the gland is relevant to imagination. The quotation in the previous paragraph from L'Homme suggests that in imagining it is only the surface figures that count. But in the Passions, imagining, directing mental attention, and voluntarily moving the body are all associated with the power of will to act in the gland in whatever way is required for driving the spirits toward the brain pores so that the soul might perceive things in the appropriate fashion.

The phenomenological requirements of Le Monde's discussion of pure imaginary space would seem to favor an alternative: that the power of the soul extends throughout the whole gland and that it is within its three dimensions that the fabular replica of external space is produced. One would then also have a simple criterion for discriminating imagining and sensing. In the accounts of L'Homme and the Dioptrics it is not immediately evident how we can distinguish a pineal gland surface trace made by sensation from one made by imagination. How do people know the difference between them? Why do they not confuse them with one another, with imagining mistaken for sensations and vice versa? If imagination were an activity within the gland, however, or the result of an inward action that could be mapped or traced outwardly on the surface, whereas sensation were at the surface alone or the result of an inward-moving process, there would be a physiological correlate of this difference.

At any rate, the furthest Descartes went in his description of the physiology of the imagination was the passage in the Dioptrics that explained how we perceive the distance of objects. In the first place, the eye changes its shape according to the distance of an object; "in the measure that we change it in order to proportion it to the distance of the objects, we change also [a] certain part of our brain, in a fashion that is instituted by Nature for making our soul perceive that distance" (AT VI 137). We ordinarily do this without any conscious activity, just as we adapt the shape of our hands to any material object we grasp without having to think about it. Moreover, a blind man with a stick in each hand is able "as though by a natural Geometry" to determine an object's distance; similarly our two eyes, having slightly different perspectives, or even a single eye that moves so that on the retina there are traced lines proportioned to the size and distance of objects, allow us to perceive the distance: "and this through an

action of thought that, being only a completely simple imagination, does not stop enveloping in itself a reasoning completely similar to what Surveyors do when by means of two different stations they measure inaccessible places" (AT VI 138). Even the relative distinctness or indistinctness of the shape seen or the strength or weakness of light reflected from the object can enable us "not properly to see but to imagine its distance" (AT VI 138-139). Differences in the shapes, colors, and light give us a comparative perspective that may allow us to judge distances as well.

It is hard to know what to make of this. The passage maintains a central role for imagination that is simple yet involves an implicit quasi-trigonometric reasoning (by similar, i.e., proportional, triangles). The imagination works by a kind of proportionalizing function that is beneath consciousness proper, yet that function is called a thinking and reasoning. Was Descartes picturing to himself an actual, real-time process of imaginal calculation—for example, a step-by-step manipulation of geometric lengths of the kind to be found in the Regulae? Or a dynamic production of space and geometry as in Le Monde? Or a continuous generation of more complex curves from less complex ones as in the Geometry? Would this reasoning take place on the surface of the gland, in its interior, or not in the gland but in mind alone? He does at least suggest the relevance of the third dimension of the gland in both sensing and imagining by analogizing the change in shape of the eye in focusing on objects to the hand's adapting its shape when it grasps an object; this change in the eye could correspond to a change in the brain in a manner ordained by God or nature to make our soul perceive distance.

A reasoning that is like trigonometric calculation but is very simple, a logic of surface traces that nevertheless penetrates beneath the surface, an irresolution between images and motions—these are signs of a ferment of thought that regarded imagination as important enough to allow tensions and inconsistencies to emerge in public, in Descartes's published scientific writings. Yet they appeared elsewhere, too, perhaps below the threshold of discrimination for most readers, but nevertheless there: imagination as activity, an activity that might transcend itself to grasp, willingly, the highest, spiritual truths. I am referring not to the writings of his youth but to the work that philosophers take as the keystone of Cartesianism, the Meditations. And so it is to this and other works of the fully matured philosopher that we must finally turn.