A Human Buglist

When I was about eight years old, two crosses on a page in a "Boy's Book of Wonders" showed me that my eyes were blind, right in the centre of my visual field, AND that my vision was lying to me about what was in the spot. (You can see - that is, fail to see - the blind spot yourself with the demonstration on the Wikipedia.)

This is a fault. A bug. It's not the only one and together they tell us something about what we are.

Bugs have different causes and the blind spot falls into the category of bad engineering design. The eye is a camera made out of jelly and paint, so you wouldn't expect it to be good, but it's still a surprise to find that  something we experience so richly is even worse than it needs to be. The blind spot happens because that's where the optic nerves carrying signal from all the retina sensors bunch together to head off to the brain. The centre seems like a logical place for that to happen, except that because the light sensors are installed on the wrong side of the retina, the nerves have to be in front, blocking the view. It really is a bit of a bodge.

Want some more bad engineering? How about our structure? There's obviously something rather provisional and desperate about our spines. Something that hurts that much can't be right. And what possible purpose can flat feet have?

Of course the classic example of specifically human bad design is childbirth. How many women have been killed - or will die today - through breech delivery, head too big or infection? Other mammals don't get this: When humans got huge heads to fit all the brains in, there should have been a design review of parturition, but that obviously never happened.

So there's plenty of bad engineering, and I would call it buggy, but really that's not what most people mean by a bug. A bug is a software fault: information processing by a program that's written wrong. Software bugs are sometimes caused by poor communication between the designers, imprecision in the specification, lack of skill or inadequate testing, but the classic cause is just wrong assumptions, and it seems to me that all our software bugs are this type.

Optical illusions show this nicely. Starting with the raw images, visual processing detects edges, orientations, movement and higher-level interesting constructs like faces or nudity. But to work under under harsh time pressure it needs to assume that the images on the retina are caused by looking at a physically consistent world. Artful graphics exploit that to confuse one or more levels of processing: The lines are parallel, the image is flat, the circles are stationary, there is no face on Mars and the models are decently dressed (sort of.) Any perception to the contrary is the result of your visual cortex struggling with perverse cues.

OK, but we just identified that our eyes are badly made. Perhaps other senses are better?

It doesn't seem that way. Unlike our eyes, ears are really clever engineering. We have direct and precise frequency analysers in our auditory canals, but we can still be deceived by a Shepard tone -- a mixture of sounds that would never arise in the natural world. And your nose contains a fabulous chemical analyser -- possibly not as good as a gas chromatograph/mass spectrometer combo, but a lot faster, so it's a shock that a plant hormone can make sour seem sweet.

The same basic problem - failed assumptions - throws out more trouble in our behavioural choices. The assumptions we make are the those self- and mutually-aware people living in the wild.

• We assume that there is no law. On that basis, honour-based choices - irrational, boundless escalation in the face of any slight or conflict - are a smart move, especially for young males. You get hurt sometimes, but the respect protects you against casual try-ons. But in our society of laws and state-monopoly violence, defence of honour is harmful or horrifying.
• We assume that we're living in an extended group of one or two hundred people. So when we hear of something bad happening to someone, we think it might affect us. That makes us too easily cowed or amazed in a culture that has access to all the interesting or noteworthy events that happen to a global population of billions. We react as if something that has a one-in-a-billion chance has happened to a member of our family.
• We assume that a woman who has sex with a man will bear his children. So, if she's "easy" she's competing - blacklegging - against women who demand a price in fatherly attention for their offspring, or she's sponging his unknowing care on other men's children. And that's why men and women hate slags, even when they benefit from their attentions, or are fond of sex themselves. (I find this quite the most depressing entry on this list.)
• And our appetite assumes that food is scarce, we will do hard physical work most days and you can never get too much sugar, lard or meat. It's no wonder we get fat.

It's worth saying that these are behavioural explanations, not justifications. It's immoral to defend your honour in violation of the law. Natural, thoroughly human, and grossly wrong.

You'll see that I'm going "up the stack" here, from the raw neuronic signals up to more sophisticated processing. There are assumptions all the way, but I'm going to finish with some judgements, with things that as slightly rational thinkers we really might expect to get right, but are still undone by our design assumptions:

• We don't believe in small effects accumulating over a long time, or even small things at all. We have to learn it: "Many a mickle makes a muckle" is wisdom, not intuition. That's why we have to explicitly learn about regular savings, or geology, or washing hands to remove invisible microbes. (And it's why I struggle with computer memory.) We're making judgements in the face of a high assumed probability that we'll die before next spring, so small things really don't matter.
• In the other direction, we care about things that we are aware of, regardless of their actual impact. So when vanishingly small concentrations of pollutants show up on a mass spectroscope, we worry - foolishly.
• We're terrible at statistics. We simply can't do random - we always assume there's a meaning and do what it takes to extract it, even where there is none, with the same avidity that we identify faces on Mars. That's how we end up with webs of belief about magpies or turning three times before looking at the moon.
• Our intuitive metaphysics: cause and effect, things and relations, is an assumption. It works well enough to get through the day, but it has almost nothing to do with the integrated mesh of resonance which is the the way the universe is actually made.
• We think natural things are good and right. But they're not. Which I suppose is my point with this list. We go the wrong way a lot.

I could go on. But I hope this is enough to show that we're a bit of a hack. Stay with me and I'll try and explain why I think that's interesting.

As humans, we cannot really understand - intuitively - what human beings really are. We're too close. But study allows us to see that we are mechanisms (perhaps mechanisms among other things) and what's more, be certain that the mechanism is assembled by natural selection of inherited variations.

Be certain? Certainty is a strong claim. Surely there are other possible explanations, no matter how implausible, for  the excellence of our bodies and the power of our perceptions? There have been many offers: divine design, a striving life force, even alien intervention or ancestor simulators.

I don't think that we can pick between these explanations based on how well we work. Some are more plausible than others, but any of them might be the one. There are lots of reasons to be designed right. What I like to see is where we're working badly. That's interesting.

Tools and limits shape designs. Compare the buildings (or cars) of thirty years ago with those of today to see the effect of full 3D computer design tools. Or study a cheap toy to understand the limits of something designed to be assembled by ill-equipped workers in Chinese sweatshops. (Clue: too many screws.)  It's the flaws and the limits that reveal the capability of the designer. They're side effects, and the only reason for them to be there is because the designer didn't perceive them, didn't care about them, or couldn't get rid of them. Defects and limits speak of necessity, and necessity is hard evidence of identity.

So that's what interests me.

Natural selection never gets to redo from scratch because each change has to be an improvement, no matter how tiny, in one generation. It can make big changes, but only if each step on the way works better in the current circumstances. It has no foresight at all. All the other candidates have a chance to fix what's wrong; God doesn't do mistakes. Simulation authors want to save processor cycles. The aliens can spot the faults just like we can. The life force is defined as striving for perfection. Natural selection is the only really foolish designer.

The mammalian eye is an unfortunate fluke -- mollusc eyes have it the right way round without difficulties. Natural selection can't move the data connections to the back of the retina, because the first intermediate stages to correct it will be blinder than the current design. In the same way, my back hurts because my body plan is a quadruped that has been bent hard wrong to go on two legs all the time. Nothing said "OK, another ape, but this one walking upright." A redesign for two-legged life could have addressed the "big brain delivered through the pelvic girdle" issue too, but it'll never happen because natural selection doesn't do redesigns.

The wrong assumptions come down to the same thing -- a designer utterly lacking in foresight. If God had meant us to see, he would have given us a dedicated GPU, rather than using a brain that can't keep up. Would God have given us an inbuilt honour code when the real moral challenges facing the human race are the global tragedies of the commons, and easy nuclear weapons? (Mind you, I can see simulation authors trying that -- one of the uglier ways to look at the world is that it is a psychology experiment by someone pushing the most vicious and perverse circumstances he can.)

All the stuff we get wrong -- dietary choice, physics and metaphysics, racial distaste, even exotic things like economic fallacies -- all make sense, albeit a bitter and immoral sense in the context of an evolved social species living on the plains. That is who, that is what, we are.

Is that depressing? I think it's rather fine to be a great ape that got lucky; if you don't, that's OK too. But if we can jointly say that there is a mechanism here (whatever else is here too) and that the mechanical works themselves have no special status and many obvious screw-ups, then  there is a way forward. We can try, perhaps not to fix them exactly, but make ourselves the masters, managers and controllers of their harm.