13:04
I made this pilot back in 2018 in collaboration with MinuteEarth and Derek of Veritasium. Originally it was behind a paywall, but the statute of limitations has now run out so here it is on YouTube. Enjoy! There's an old Irish myth of a god-king named Balor the Mighty, who had a poisonous eye of fire in his forehead that he covered with seven cloths to keep the eye cold and everything else cold too. Balor ran his kingdom like a scoundrel and he would take the biggest tributes. When a person didn't pay, Balor came before them himself, and took the cloths off the eye one after the other.
When he took off the first cloth, the bracken began to wither; the second cloth, the grass began to turn red; the third cloth, the woodpiles and forests began to warm; the fourth cloth, there was heat and smoke coming out of the wood and out of the trees; after the fifth cloth everything was getting red; it has been forgotten what happened after the sixth cloth, but after the seventh cloth, the everything was burned, the whole land was on fire. All from the gaze of his eye.
The word paradigm comes from the Ancient Greek for "to see side by side", and that is what we are going to be doing today: seeing, side by side, different ways people have tried to understand seeing. The most popular theory among the Greeks themselves was that vision was basically a long-distance version of touch, and involved something reaching out from the eyes to get an impression of objects in the world. The philosopher Plato suggested that sight was a "visual fire" that melded with daylight, somehow allowing the "forms" of objects to travel long distances and into the observer's eye.
Plato was pretty vague on the details, but the basic idea of "visual fire" rang true for people for a number of reasons: animals' eyes seem to gleam at night, and people often appear to have a twinkle in their eye. Plus, everyone knew that a gaze could have certain mystical powers, from that eerie feeling you get when you know you're being watched, to a lover's enchantment, to the 'evil eye' - the idea that you can harm somebody just by looking at them. Even back then, people did realize there were problems with this "visual rays" understanding
of vision, like: when I open my eyes at night, if beams have to shoot out and travel all the way to the stars and back again, how come I see the stars right away? Or, why can't we see things in the dark? Does darkness somehow stop the rays? But the thing about mental models of the world - about paradigms - is that it's really tough to kill an established theory - especially one that makes a lot of intuitive sense - just by pointing out its weaknesses; you have to have an alternative that can explain at least as much, more compellingly, and with fewer problems. There were other theories in Ancient Greece, like the idea that vision worked like how we now know smell works:
maybe when you're looking at a mountain, little pieces of that mountain - vision particles - carry the essence of the mountain out to be detected by your eyes. But this and other theories of vision had more problems, or explained less, or were less intuitive. So for centuries, people thought "rays shooting out of your eyes" was the truth, and tried to explain even more stuff within this paradigm. Like the mathematician Euclid, who lived in an abstract world of straight lines and the angles and distances between them. He took the idea of visual rays going out in all directions from the eye, and used geometry to explain why closer objects appear
larger and in more detail: the closer the object, the more visual rays it intercepts. And the surgeon Galen used the paradigm too, He sliced up tons of animal eyeballs and drew up an official anatomy of the eye: it's a ball filled with fluid, with an opening in front that we call a pupil, and behind that a hard, somewhat mysterious, crystal-like object that people called the "crystalline humor". But Galen so strongly believed in the "something coming out of the eye" theory of vision that it clouded his own sight: he described the optic nerve as a hollow tube for sight to travel down, even though the optic
nerve is solid. Even though he was getting his hands dirty and collecting data, he was stuck in the paradigm of vision as 'stuff leaving the eyes', and wasn't able to break free of this view. And one thing that everyone had trouble fitting cleanly into that paradigm was light. They knew that light rays emanated from bright objects like candles or the Sun, and that the presence of light is important for seeing (because you can't see without it), but the precise role light rays play in vision remained shrouded in darkness. It took almost another millenia for the flash of brilliance necessary to develop a new theory of sight. It took until Ibn al-Haytham, the first to see the light.
Al-Haytham was an Arab mathematician, astronomer, and physicist who lived during the Islamic Golden age. And he was arguably the first real modern scientist, because while he was happy to read other people's work, he only believed ideas that could be verified with sound reason or experiments. And in al-Haytham's eyes, the Greek theory of eyes emitting rays didn't hold up to scrutiny. For example, Al-Haytham spent a lot of time staring at bright lights, which can be really painful, so he wondered, 'if our eyes are putting out a visual fire, why don't they hurt all the time'? Additionally, he noticed that when you look at a bright source of light for a while and then look away, you see ghostly after-images,
which suggests that light from the outside is directly affecting our eyes. Al-Haytham also used dark rooms to study light. If you poke a hole in one wall of a dark room and sit inside, the light from the outside creates an image on the opposite wall. In fact, the word "camera" has the same root as the word "chamber". So sitting in his camera, al-Haytham noticed that during a solar eclipse, an image of the eclipsed sun would appear on the wall - except it was upside-down and leftside-right. Al-Haytham understood this happened because the light from the top of the crescent would follow straight lines, pass through the pinhole, and end up on the bottom of the wall,
while the light from the bottom of the crescent would end up on the top. And so, al-Haytham knew that dark chambers with a little hole on one side were a kind of optical instrument for creating images from light sources. And he knew from his own observations that light directly affects the eye. And he knew from reading Galen's anatomy that the eye is essentially a dark chamber with a little hole on one side. So al-Haytham put geometry and physics and anatomy together, and concluded that perhaps the eye itself was an optical instrument, just like a pinhole chamber. And if the eye is a pinhole chamber receiving light from the outside,
there was simply no need for a theory of outgoing visual rays any more. Incoming light IS sight! This was such a big deal that it's worth repeating: al-Haytham figured out that light is the source of vision and the eye is an optical instrument. To explain vision, there wasn't any need for separate visual rays. But there were two big problems that al-Haytham struggled with. First, pinhole cameras create images that are upside-down and flipped - it's simple geometry. So if the eye works like a pinhole camera, why don't we see the world upside down?
Secondly, he also wondered why the image in our eye isn't super blurry. A pinhole camera has a really tiny hole (a pinhole!), and this allows it to make a really sharp image, because light rays from each point outside end up at basically a single point inside. But if you make the pinhole too big, then you start letting in a bunch more overlapping light rays, and the image starts to smear out and get really blurry. The pupil of our eyes is like this - it's way too big. So why do we see sharp images? Al-Haytham came up with ad-hoc ways to patch up both of these problems, but he wasn't really able to solve them in a satisfying,
coherent way. And then the Golden Age of Islamic Science came to an end, and the correct resolution of the upside-down blurry image problem would have to wait another 600 years until the European scientific revolution. It would have to wait until Johannes Kepler. Kepler was an astronomer, so he was interested in taking accurate measurements of stuff in the sky. But he needed to know how the eye worked to make sure he could trust his observations, and he couldn't find anyone who had solved al-Haytham's two problems yet. So Kepler decided to do it himself. In thinking about the blurriness problem, Kepler had a fresh perspective. There was a
recent development in pinhole chamber technology. People knew that if you make the pinhole larger, you get a blurry image. But someone noticed the curious fact that if you put a lens in front of the pinhole, it makes the image sharper! This is because a lens bends light rays, and can focus the overlapping, diverging rays back together to a single point, removing the blur. Kepler deduced that the hard "crystalline humor" of the eye had the right shape and properties to be a lens and keep the image sharp. First problem solved. As for the problem of the upside down image - Well, Kepler tried to figure out a way that the eye might flip images rightside up again,
but there was no other place for the light to refract other than the lens (which doesn't flip images rightside up). So eventually he threw up his hands and said, "sorry, but the image IS upside-down." There's no getting around it. Maybe something happens in the brain or the optic nerve to make us think it's right-side up, but that's not my concern - I'm not a doctor!" You could argue that Kepler was just making light of the matter, but in hindsight, this was a key moment. What Kepler did was part of a now time-honored scientific tradition of not biting off more that you can chew. He
cleanly split vision into two categories - the optical stuff that he understood, and the biological stuff that he didn't. This break demystified vision and helped define optics as the study of light and its interaction with mirrors, lenses, and matter; nothing more. And the study of optics opened the doors to much of the science and scientific discoveries of the last 250 years! It led to telescopes, which furthered our understanding of our tiny place in the giant cosmos, and microscopes, which showed us the even tinier world of cells, bacteria, DNA, and beyond.
Kepler's division of the optics of light from the biology of vision also focused us on the remaining mysteries of sight: what happens to those upside-down, backward images the eye captures, AFTER the eye captures them. It turns out that the brain IS hugely involved in interpreting and modifying the signals detected by the eyes. Images arrive at the brain not just upside down and backward, but also covered in a web of shadows from the blood vessels that flow over the retina, and with a gaping hole where the optic nerve plugs into the eye. Your brain fills in the gaps, removes the shadows, and flips the image, so that while your eyes see these things, you don't.
And optical illusions hint at a whole lot more of "seeing" that happens in the brain: here we have two squares that appear to be different colors, but when we remove the rest of the image, suddenly it's clear they're the same color. And it's well known that an object looks larger when drawn closer to the horizon than farther down, even if it's actually the same size on the screen. These kinds of illusions, and many more, suggest that our brains actually make a mental model of what we think the world should be like, and use incoming visual information to constantly update
that model, rather than presenting to us the raw images that come in through the optic nerve. It makes sense that we'd perceive these two squares as being different colors even if the light coming into our eyes from them is the same, because we have a mental model of how squares on checkerboards are different colors, and a mental model of how shadows make things darker, so our brain knows it's just the shadow that makes the light square as dark as the dark one. This is our modern vision paradigm: light makes an image on the back of the eye (which al-Haytham and Kepler figured out),
and then the brain processes information in those images using some combinations of neural pathways that we're only just beginning to understand. The very words we use to describe these processes, like "information" and "processing" suggest that we often think of the brain like a computer or a network, because computing is the dominant paradigm of OUR time. But a lot of the details of how the brain works are still pretty blurry. We have a long way to go, and ultimately, our current picture of the brain processing images like a computer might one day look as vague as Plato's visual fire.
You can now learn about math and coding with step-by-step interactive lessons and puzzles designed to improve your problem solving skills - all powered by this video's sponsor, Brilliant. If you get stuck on a particular problem - Brilliant provides a digital tutor to give you clues and make suggestions to help you understand faster. There are plenty of topics to choose from: Math, data analysis, programming, science, and more, and whatever you pick, Brilliant lets you learn in a more engaging and effective way
than just watching videos; maybe eventually you'll help usher in the next scientific revolution. To sign up for Brilliant for free, go to Brilliant.org/YOUR_URL, scan the QR code onscreen, or click on the link in the description. You'll also get 20% off an annual Premium subscription, which gives you full digital tutoring support and unlimited access to everything on Brilliant. Again, that's Brilliant.org/YOUR_URL - and thanks to Brilliant for supporting MinutePhysics.
