Philosophy-RAG-demo/transcriptions/HoP 132 - Eye of the Beholder - Theories of Vision.txt

 Hi, I'm Peter Adamson and you're listening to the History of Philosophy podcast, brought to you with the support of King's College London and the Leverhulme Trust, online at www.historyoffilosophy.net. Today's episode, Eye of the Beholder, Theories of Vision. If the history of science teaches us anything, it is to beware of the obvious. Many things that once seemed obvious have turned out to be false, for instance that the earth is not moving. Conversely, many things that we now take to be self-evidently true are far from obvious. My favorite example of this is the fact that we think with our brains. If this were really so obvious, there would have been no ancient debate about whether the soul's ruling faculty was in the heart or the brain. Another example is eyesight. You probably just take it for granted that when you see something, it is because light is bouncing off what you are seeing and being directed to your eyes. And you're right, that is more or less what is happening. But so far is this from being obvious that no one in the ancient world even proposed such a theory of vision. It wasn't for lack of trying. In ancient Greek and Roman science there were at least three rival theories to explain human eyesight associated with three philosophical sources, Plato, Aristotle, and the atomists. But the first author to set down something like the correct theory of vision wrote neither in Greek nor in Latin, and he did not live in antiquity. His name was Ibn al-Haytham, and he lived in 10th to 11th century Cairo, and he wrote in Arabic. Something else the history of science teaches us is that such innovative leaps nearly always depend on the work of previous thinkers. Ibn al-Haytham was no exception. His treatise Kitāb al-Manādir, or Book of Optics, draws on ancient treatments of eyesight and human anatomy while also exploiting ideas put forward in the earlier Arabic tradition. Of course that detracts nothing from his achievement, but it means that if we are to understand his breakthrough we will need to consider a range of previous optical theories. So, if you'll pardon a pun weaker than the eyesight of an elderly bat, let's first have a look at Plato. He proposed a theory of eyesight in his dialogue Timaeus. Like I said when I covered this dialogue about 100 episodes ago, it is not Plato's most popular work nowadays, but it has historically been one of his most influential, especially in the medieval period. This dialogue deals with the providential design of the universe as a whole, and also of the human body. Within this account, vision is particularly important, because according to Plato, philosophy itself would never come about if we could not see. For it is our observations of the heavens that lead us to discover number and to investigate the nature of the universe. The dialogue tells us that we see, thanks to an invisible stream of very pure fire that is emitted from our eyes. In order for eyesight to take place, this stream must encounter a kindred fire outside, namely light. The resulting connection causes a motion in the soul, namely our seeing. Thus, Plato can answer the most basic questions one might want to pose about sight. Why can we only see what we are directly looking at, and not for instance what is behind us? And why can't we see in the dark, or when our eyes are shut? His answers would be that we see whatever the stream of fire from the eyes can reach, which means that the eyes must be open and the visible objects in front of us, and that without external illumination, the fire has nothing akin to it for it to connect with. Plato also takes up a question that will play a major role in subsequent discussions of eyesight—how do mirrors work? The answer is that the visual stream meets an external illumination on the surface of the reflective body. Plato even tries to explain why mirror images are reversed, and why curved mirrors yield different kinds of reflections. Historians of optics call this sort of theory extra-missionist, because it involves something being sent out of the eyes. We find a later extra-missionist theory in one of Plato's biggest fans, the doctor Galen. He supplements the Timaeus account with his own anatomical ideas. Unlike Plato, he is writing after the discovery of nerves, and in particular he's aware that there are nerves that connect the brain to the eyes. Also unlike Plato, he adopts a Stoic-inspired understanding of the human body, which sees many functions of soul as being carried out by a very fine sort of breath, or pneuma, that pervades the body. The finest kind of pneuma, the sort involved in perception, is distilled in the brain out of the less subtle breath taken into the lungs and then circulated around the body by the heart. Thus, Galen modifies Plato's theory by proposing that the brain is sending pneuma to the eyes. So refined is this pneuma that it has a nature akin to that of light itself. Yet Galen doesn't claim that the pneuma itself is being emitted out of the eyes to whatever we can see, like Plato's fiery visual stream. Instead, the pneuma affects the air in front of the eyes, transforming it into an instrument that brings the visual organ into contact with the visual object. This allows Galen to avoid a standard objection to extra-missionist theories, which is that the human body could never generate sufficient visual stream required to see out over a whole countryside or as far as the heavens. To this, Galen would say that the pneuma causes a chain reaction in which the whole transparent medium, even as far as a distant horizon or the heavens, is transformed into an instrument for seeing. In common with Plato though, he has an intuition that underlies all ancient extra-missionist theories of vision. In order to see a distant object, we need somehow to get into contact with that object. The extra-missionists are effectively saying that seeing is a special way of reaching out to touch other things, even if they are as remote as the vault of the heavens. We use either the air or a stream of visual fire to do this. To make this point, the Stoics compared the visual stream to a walking stick, which the viewer is using to tap whatever is seen. Other ancient philosophers quite literally took a different view. For them, when Muhammad sees a mountain, it is not Muhammad's sight that goes to the mountain, but the mountain that comes to Muhammad. Such a view is intro-missionist, that is, holds that something from the outside world is sent into the eyes. A prominent example is found in atomist authors, notably the Epicureans. They believed that very thin films of atoms are constantly being shed by all visible objects. The atomic sheets are called aidola, or images. When such an atomic image reaches the eyes, it literally collides with atoms of the soul through the portal of the eye. Again, sight is effectively being reduced to touch, but in this case we are touching something that reaches us from a distance, instead of our somehow reaching out to make contact. One advantage of the atomic theory was that it could claim to account for some visual illusions. A famous example is the square tower that looks round from a distance. The explanation would be that the atomic image is buffeted by the air on its way to us, the sharp corners literally being knocked off in the process. The same process might also explain how the images are reduced in size by the time they reach us so that they can fit into the eye. Critics were quick to point out the numerous weaknesses of this theory. To give just one example, if these atomic films are so flimsy, wouldn't they entangle with one another in mid-air, being destroyed or mixing together? Fortunately for the in-crowd though, there was another candidate theory for intro-missionists to adopt, that of Aristotle. For him, as you might remember, we see when the potential of our vision is activated by some external form. In order for this to happen, there must be an illuminated, transparent medium, like a stretch of air filled with sunlight, between the viewer and what is seen. Yet again, we see the need for some kind of contact. The illuminated air fills the gap between seer and scene, and by being in touch with both, transmits the visual form from the object to the eye. The theory solves some puzzles well, for instance by explaining why we can't see in the dark. It is because unilluminated air is incapable of carrying the image. Notice by the way that for Aristotle, we couldn't see through a void, because there would be no medium to carry the image to us. Again, there is room for criticism here. John Philoponus, the late ancient Christian who attacked Aristotle concerning the eternity of the world, also complained about the Aristotelian account of eyesight. He pointed out that Aristotle doesn't solve that most basic of questions, why can we only see what is in front of us? After all, something that is behind me in a well-lit room is touching the illuminated air that touches my eye, so the air should convey the image to me. Yet, with the exception of those of us who are primary school teachers, we are not able to see what is happening behind our backs. Much better positioned to deal with this problem were those authors who applied the tools of geometry to explain vision. The tradition of geometrical optics begins just after Aristotle with the work of Euclid, who I guess I don't need to say was pretty good at geometry. He saw that you could use this branch of mathematics to model what is happening in human vision. Actually, it's been claimed that this technical branch of applied geometry might have originated in Greek theater, when they were figuring out the sight lines for the audience. The dramatic insight here, at any rate, is that we can see only those objects that lie on a straight, unobstructed line drawn to the eye. As a whole, the visual field can be modeled as a cone whose vertex is at the eye and broadens out from there to cover everything we can see, with the edges of the cone corresponding to the edges of our peripheral vision. If something falls inside the cone and is not blocked by an opaque object, then we will see it. The only exception is what we see in a reflective surface like a mirror, which of course does let us see what is behind us. Here, geometry is again useful. If you look into a mirror obliquely from the right, you'll see what is located to the left, and at the corresponding angle. We can make diagrams representing what happens here by drawing a line from the eye to the surface of the mirror to the object seen. Now in Euclid, we are really talking about a mathematical model of vision. There is not much hint as to the physical process being modeled, albeit that it seems to go nicely with the kind of view found in Plato. The visual cone would represent the flow of rays from the eyes to what is seen, and the straight line within the cone would abstractly represent the visual rays. That possibility was exploited by the other great ancient figure in the history of Greek geometrical optics, Ptolemy, whom we already know from his work on astronomy and astrology. He puts some physical meat on the bones of the Euclidean theory, making it clear that the lines of our model do represent visual rays emitted from the eyes. He uses the model to account for a range of otherwise inexplicable phenomena. For instance, how do we tell how far away something is? Within his Platonic and Euclidean model, this is easy to explain. Since we are touching something with rays sent out from our eyes, we can tell how far the rays must travel before they light upon each object. We can also explain refraction, for instance the infamous straight stick that looks bent in water. This happens because the visual rays are being slowed and dragged away from their straight path when they meet a medium that is denser than air. The promise of this geometrical model and the transmission of Euclid and other optical works from Greek into Arabic meant that in the Islamic world, this general approach underlay all serious philosophical theories of vision. Ever ready to reflect on every topic under the sun, Akinde wrote extensively on optics, including numerous works on mirrors. Like Ptolemy, he adopts the extra-missionist theory and makes the lines of the model correspond to visual rays. He repeats a powerful rejection of the kind of intermission we find in Aristotle, with an objection already suggested by the ancient scientist Theon of Alexandria. If Aristotle were right that objects transmit visual forms through the air, then they would look the same from every angle. But consider what happens when we look at a circle from an oblique angle, we don't see a circle, but an oval. One consequence of Akinde's rejection of Aristotle's theory is that he no longer has much use for the idea of a transparent medium. Aristotle thought that illuminated air must be present to serve as a carrier of forms, and that air's transparency consists in its being able to do this job. In fact, for him, air is only potentially transparent when there is no light. Illumination makes it actually transparent, that is, actually able to transmit visual forms. By contrast, Akinde thinks of the transparency of air in negative terms as we would today. Transparency just means that which does not get in the way of our seeing, what is on the far side of it. Akinde, though, would understand this in terms of the visual ray theory. For him, air is transparent because it does not block our vision. When something intercepts the visual ray, then we see it instead of seeing through it. For this reason, Akinde tells us, it is the element earth that gives rise to visibility in objects. Unlike air, fire, and water, it is dense enough to intercept the rays from our eyes. That gives rise to yet another puzzle which should be familiar to anyone who has spent time with a four-year-old child. Why is the sky blue? After all, it is presumably made of air, which should have no color at all. Akinde rises to this challenge too, writing a little treatise specifically on the question. He explains that there are exhalations from the ground which ascend into the air, and that the blue color we see is the result of earthy particles suspended in the atmosphere. More generally, the different colors around us are the result of different elemental proportions, with dark colors belonging to things that have more earth in them. Highly polished surfaces like mirrors do not only intercept the rays, but actually reflect them so that they fall on other objects placed in appropriate positions. Finally, Akinde explains why we can't see in the dark, by saying that even dense objects are seen only when their surfaces are illuminated. Light only occurs when the visual ray and a ray of light fall on the same spot. So it is not the air between me and what I see that needs to be lit up, as Aristotle thought. Rather, the surface of what I am seeing must be illuminated. So in principle, I could see through a void after all, though as it happens Akinde didn't believe that void could exist. So powerful was the geometrical version of the visual ray theory that even staunch Aristotelians like Al-Farabi relied on it. In a work called The Enumeration of the Sciences, he devotes a brief section to optics, and actually says that its main purpose is to account for such phenomena as optical illusions. He also alludes to its use for determining such things as the height of mountains, another topic that had been discussed by Akinde. Still, it's not hard to mount a challenge to the visual ray theory. One of the biggest difficulties is this. If we are sending rays out of our eyes, then all the action seems to be happening at the far end, where the rays make contact with the visible object. But the sensation is happening here, in our brains. If we want to see, it isn't good enough to send something out that makes contact with a distant object. Information also has to return to the eye from the object, so that we can register what the visual ray has touched. In that case, every extra mission theory must also suppose some kind of intro mission, from object to eye, not only eye to object. But that seems pointless. If something comes from the object to the eye anyway, what is the point of supposing that anything at all comes out of the eye? This objection is found in two authors who were contemporaries, the philosopher Avicenna and the hero of our story Ibn al-Haytham. They abandoned the extra mission view entirely, but continued to exploit the advantages of geometrical optics. This means accepting the same visual cone postulated by Euclid, with its vertex at the eye and spreading to cover the whole visual field. But the direction of flow is different. Now, instead of the eye sending a cone of rays to the things it sees, it will be the visible objects that send rays to the eye. And now, finally, both Ibn al-Haytham and Avicenna suggest, these will be rays of light. No special visual emanation or pneuma is needed. Rather, as now seems so obvious, vision occurs when light bounces off objects and travels in straight lines to our eyes. Of course, the illuminated surfaces are in fact sending light in all directions, not just to the eyes of whoever is looking at them. You might think that the result would be nothing but blurred confusion, since every point on our eyes should be getting light from every point on every visible surface. But Ibn al-Haytham, adapting an idea Al-Kindi had used in describing the visual ray, explains that the points on the surface of the eye register only the light rays that fall on them most directly. So, each point on the eye's surface will be affected only by the light that hits it along a perpendicular path. The result is that the effect on the eye is a perfect map of the world, with each point on the eye corresponding to one, and only one, point on the surfaces in the field of vision. Ibn al-Haytham's theory was not only much closer to the truth than those of his predecessors, it also played a crucial role in the later development of optics. When his work on optics was translated into Latin, this inspired thinkers like Kepler. As the leading historian of medieval optics David Lindbergh has remarked, This was only one of the bright ideas to emerge from mathematical thought in the Islamic world and to illuminate the European scientific tradition. Ibn al-Haytham's whole project in the tradition of authors like Euclid and Ptolemy was an application of geometry to the problem of explaining sight. And geometry was only one of the mathematical sciences. Since antiquity, usually four such sciences were recognized, namely arithmetic, geometry, astronomy, and harmonics. Some measure of the impact of work done on these topics in the Arabic-speaking world can be found in our language. You might remember my saying that the first syllable of the word alchemy is simply the Arabic definite article, al. The same goes for our word algebra, which derives from the Arabic al-jabr, pioneered by the mathematician al-khwarizmi. Now this is a podcast about the history of philosophy, not mathematics. But as we've been seeing in this episode, figures we recognize as philosophers also worked in fields like geometrical optics, and the mathematical sciences are always included in overviews of the intellectual disciplines in texts by al-Kindi, al-Farabi, and others. To pay due respect to this feature of the formative period, I want to look at one other mathematical science and its philosophical implications. Next time will be a very special episode in which I discuss the discipline of harmonics. We'll not only be hearing what some philosophers of this period thought about music, but also hearing some examples of traditional music from the Islamic world, including an extended version of the clip I've been using to open and close this current series of episodes. So make sure to tune in next time for Music and Philosophy, here on The History of Philosophy Without Any Gaps.