The same principle works for certain types of computer program, for certain types of animal behaviour, and – bringing the two together – for computer programs that are designed to simulate animal behaviour. Suppose we wanted to understand the flocking behaviour of starlings. There are some stunning films available on YouTube, from which the stills on colour page 16 are taken. These balletic manœuvres were photographed over Otmoor, near Oxford, by Dylan Winter. What is remarkable about the starlings’ behaviour is that, despite all appearances, there is no choreographer and, as far as we know, no leader. Each individual bird is just following local rules.
The numbers of individual birds in these flocks can run into thousands, yet they almost literally never collide. That is just as well for, given the speed at which they fly, any such impact would severely injure them. Often the whole flock seems to behave as a single individual, wheeling and turning as one. It can look as though the separate flocks are moving through each other in opposite directions, maintaining their coherence intact as separate flocks. This makes it seem almost miraculous, but actually the flocks are at different distances from the camera and do not literally move through each other. It adds to the aesthetic pleasure that the edges of the flocks are so sharply defined. They don’t peter off gradually, but come to an abrupt boundary. The density of the birds just inside the boundary is no less than in the middle of the flock, while it is zero outside the boundary. As soon as you think about it in that way, isn’t it wondrously surprising?
The whole performance would make a more than usually elegant screensaver on a computer. You wouldn’t want a real film of starlings because your screensaver would repeat the same identical balletic moves over and over, and therefore wouldn’t exercise all the pixels equally. What you would want is a computer simulation of starling flocks; and, as any programmer will tell you, there’s a right way and a wrong way to do it. Don’t try to choreograph the whole ballet – that would be terribly bad programming style for this kind of task. I need to talk about the better way to do it because something like it is almost certainly how the birds themselves are programmed, in their brains. More to the point, it is a great analogy for how embryology works.
Here’s how to program flocking behaviour in starlings. Devote almost all your effort to programming the behaviour of a single individual bird. Build into your robo-starling detailed rules for how to fly, and how to react to the presence of neighbouring starlings, depending on their distance and relative position. Build in rules for how much weight to give to the behaviour of neighbours, and how much weight to give to individual initiative in changing direction. These model rules would be informed by careful measurements of real birds in action. Endow your cyberbird with a certain tendency to vary its rules at random. Having written a complicated program to specify the behavioural rules of a single starling, now comes the definitive step that I am emphasizing in this chapter. Don’t try to program the behaviour of a whole flock, as an earlier generation of computer programmers might have done. Instead, clone the single computer starling you have programmed. Make a thousand copies of your robo-bird, maybe all the same as each other, or maybe with some slight random variation among them in their rules. And now ‘release’ thousands of model starlings in your computer, so they are free to interact with each other, all obeying the same rules.
If you’ve got the behavioural rules right for a single starling, a thousand computer starlings, each one a dot on the screen, will behave like real starlings flocking in winter. If the flocking behaviour isn’t quite right, you can go back and adjust the behaviour of the individual starling, perhaps in the light of further measurements of the behaviour of real starlings. Now clone up the new version a thousand times, in place of the thousand that didn’t quite work. Keep iterating your reprogramming of the cloned-up single starling, until the flocking behaviour of thousands of them on the screen is a satisfyingly realistic screensaver. Calling it ‘Boids’, Craig Reynolds wrote a program along these lines (not specifically for starlings) in 1986.
The key point is that there is no choreographer and no leader. Order, organization, structure – these all emerge as by-products of rules which are obeyed locally and many times over, not globally. And that is how embryology works. It is all done by local rules, at various levels but especially the level of the single cell. No choreographer. No conductor of the orchestra. No central planning. No architect. In the field of development, or manufacture, the equivalent of this kind of programming is self-assembly.
The body of a human, an eagle, a mole, a dolphin, a cheetah, a leopard frog, a swallow: these are so beautifully put together, it seems impossible to believe that the genes that program their development don’t function as a blueprint, a design, a master plan. But no: as with the computer starlings, it is all done by individual cells obeying local rules. The beautifully ‘designed’ body emerges as a consequence of rules being locally obeyed by individual cells, with no reference to anything that could be called an overall global plan. The cells of a developing embryo wheel and dance around each other like starlings in gigantic flocks. There are differences, and they are important. Unlike starlings, cells are physically attached to each other in sheets and blocks: their ‘flocks’ are called ‘tissues’. When they wheel and dance like miniature starlings, the consequence is that three-dimensional shapes are formed, as tissues invaginate in response to the movements of cells;* or swell or shrink due to local patterns of growth and cell death. The analogy I like for this is the paper-folding art of origami, suggested by the distinguished embryologist Lewis Wolpert in his book The Triumph of the Embryo; but before coming to that I need to clear out of the way some alternative analogies that might come to mind – analogies from among human crafts and manufacturing processes.
ANALOGIES FOR DEVELOPMENT
It is surprisingly hard to find a good analogy for the development of living tissue, but you can find partial similarities to particular aspects of the process. A recipe captures something of the truth, and it is an analogy that I sometimes use, to explain why ‘blueprint’ is not appropriate. Unlike a blueprint, a recipe is irreversible. If you follow a cake recipe step by step, you’ll end up with a cake. But you can’t take a cake and reconstruct the recipe – certainly not the exact words of the recipe – whereas, as we have seen, you could take a house and reconstruct something close to the original blueprint. This is because of the one-to-one mapping between bits of house and bits of blueprint. With conspicuous exceptions such as the cherry on top, there is no one-to-one mapping between bits of cake and the words, say, or sentences of its recipe.
What other analogies to human manufacturing might there be? Sculpture is mostly way off the mark. A sculptor starts with a chunk of stone or wood and fashions it by subtraction, chipping away until the desired shape is all that remains. There is, admittedly, a somewhat sharp resemblance to one particular process in embryology called apoptosis. Apoptosis is programmed cell death, and it is involved, for example, in the development of fingers and toes. In the human embryo, the fingers and toes are all joined. In the womb, you and I had webbed feet and hands. The webbing disappeared (in most people: there are occasional exceptions) through programmed cell death. That is a bit reminiscent of the way a sculptor carves out a shape, but it is not common enough or important enough to capture how embryology normally works. Embryologists may briefly think ‘sculptor’s chisel’, but they don’t let the thought linger for long.
Some sculptors work not by subtractive carving but by taking a lump of clay, or soft wax, and kneading it into shape (which may subsequently be cast, in bronze for example). That again is not a good analogy for embryology. Nor is the craft of tailoring or dressmaking. Pre-existing cloth is cut, to shapes set out in a pre-planned pattern, then sewn together with other cut-out shapes. They are often then turned inside out to disguise the seams – and that bit, at least, is a good analogy to certain parts of embryology. But in general, embryology is no more like tailoring than it is like sculpture. Knitt
ing might be better, in that the whole shape of a sweater, say, is built up from numerous individual stitches, like individual cells. But there are better analogies, as we shall see.
How about the assembly of a car, or other complicated machine, on a factory assembly line: is that a good analogy? Like sculpture and tailoring, assembly of pre-fabricated parts is an efficient way to make something. In a car factory, parts are pre-made, often by casting in moulds in a foundry (and there is, I think, nothing remotely like casting in embryology). Then the pre-made parts are brought together on an assembly line and screwed, riveted, welded or glued together, step by step according to a precisely drawn plan. Once again, embryology has nothing resembling a previously drawn plan. But there are resemblances to the ordered sticking together of pre-assembled parts, as when, in a car assembly plant, previously manufactured carburettors and distributor heads and fan belts and cylinder heads are brought together and joined in correct apposition.
Below are three kinds of virus. On the left is the tobacco mosaic virus (TMV), which parasitizes tobacco plants and other members of the family Solanaceae, such as tomatoes. In the middle is an adenovirus, which infects the respiratory system in many animals, including us. On the right is the T4 bacteriophage, which parasitizes bacteria. It looks like a lunar lander, and it behaves rather like one, ‘landing’ on the surface of a bacterium (which is very much larger) then lowering itself on its spidery ‘legs’, then thrusting a probe down the middle, through the bacterium’s cell wall, and injecting its DNA inside. The viral DNA then hijacks the protein-making machinery of the bacterium, which is subverted into making new viruses. The other two viruses in the picture do something similar, although they don’t look or behave like lunar landers. In all cases their genetic material hijacks the protein-making apparatus of the host cell and diverts its molecular production line to churning out viruses instead of its normal products.
Three kinds of virus
Most of what you see in the pictures is the protein container for the genetic material, and in (‘lunar lander’) T4’s case the machinery for infecting the host. What is interesting is the way in which this protein apparatus is put together. It really is self-assembled. Each virus is assembled from several previously made protein molecules. Each protein molecule, in a way that we shall see, has previously self-assembled into a characteristic ‘tertiary structure’ under the laws of chemistry given its particular sequence of amino acids. And then, in the virus, the protein molecules join up with each other to form a so-called ‘quaternary structure’, again by following local rules. There is no global plan, no blueprint.
The protein sub-units, which link up like Lego bricks to form the quaternary structure, are called capsomeres. Notice how geometrically perfect these little constructions are. The adenovirus in the middle has exactly 252 capsomeres, drawn here as little balls, arranged in an icosahedron. The icosahedron is that Platonic perfect solid that has 20 triangular faces. The capsomeres are arranged into an icosahedron not by any kind of master plan or blueprint but simply by each one of them obeying the laws of chemical attraction locally when it bumps into others like itself. This is how crystals are formed, and, indeed, the adenovirus could be described as a very small hollow crystal. The ‘crystallization’ of viruses is an especially beautiful example of the ‘self-assembly’ that I am touting as a major principle by which living creatures are put together. The T4 ‘lunar lander’ phage also has an icosahedron for its main DNA receptacle, but its self-assembled quaternary structure is more complex, incorporating additional protein units, assembled according to different local rules, in the injection apparatus and the ‘legs’ that are attached to the icosahedron.
Returning from viruses to the embryology of larger creatures, I come to my favourite analogy among human construction techniques: origami. Origami is the art of constructive paper-folding, developed to its most advanced level in Japan. The only origami creation I know how to make is the ‘Chinese Junk’. I was taught it by my father, who learned it in a craze that swept through his boarding school during the 1920s.* One biologically realistic feature is that the ‘embryology’ of the Chinese junk passes through several intermediate ‘larval’ stages, which are in themselves pleasing creations, just as a caterpillar is a beautiful, working intermediate on the way to a butterfly, which it scarcely resembles at all. Starting with a simple square piece of paper, and simply folding it – never cutting it, never glueing it and never importing any other pieces – the procedure takes us through three recognizable ‘larval stages’: a ‘catamaran’, a ‘box with two lids’ and a ‘picture in a frame’, before culminating in the ‘adult’ Chinese junk itself. In favour of the origami analogy, when you first are taught how to make a Chinese junk, not only the junk itself but each of the three ‘larval’ stages – catamaran, cupboard, picture frame – comes as a surprise. Your hands may do the folding, but you are emphatically not following a blueprint for a Chinese junk, or for any of the larval stages. You are following a set of folding rules that seem to have no connection with the end product, until it finally emerges like a butterfly from its chrysalis. So the origami analogy captures something of the importance of ‘local rules’ as opposed to a global plan.
Also in favour of the origami analogy, folding, invagination and turning inside out are some of the favourite tricks used by embryonic tissues when making a body. The analogy works especially well for the early embryonic stages. But it has its shortcomings, and here are two obvious ones. First, human hands are needed to do the folding. Second, the developing paper ‘embryo’ doesn’t grow larger. It ends up weighing exactly as much as when it started. To acknowledge the difference, I shall sometimes refer to biological embryology as ‘inflating origami’, rather than just ‘origami’.
Chinese junk by origami, with three ‘larval stages’: ‘catamaran’, ‘box with two lids’ and ‘picture in a frame’
Actually, these two shortcomings kind of cancel each other out. The sheets of tissue that fold, invaginate and turn inside out in a developing embryo do indeed grow, and it is that very growth that provides part of the motive force which, in origami, is supplied by the human hand. If you wanted to make an origami model with a sheet of living tissue instead of dead paper, there is at least a sporting chance that, if the sheet were to grow in just the right way, not uniformly but faster in some parts of the sheet than in others, this might automatically cause the sheet to assume a certain shape – and even fold or invaginate or turn inside out in a certain way – without the need for hands to do the stretching and folding, and without the need for any global plan, but only local rules. And actually it’s more than just a sporting chance, because it really happens. Let’s call it ‘auto-origami’. How does auto-origami work in practice, in embryology? It works because what happens in the real embryo, when a sheet of tissue grows, is that cells divide. And differential growth of the different parts of the sheet of tissue is achieved by the cells, in each part of the sheet, dividing at a rate determined by local rules. So, by a roundabout route, we return to the fundamental importance of bottom-up local rules as opposed to top-down global rules. It is a whole series of (far more complicated) versions of this simple principle that actually go on in the early stages of embryonic development.
Here’s how the origami goes in the early stages of vertebrate development. The single fertilized egg cell divides to make two cells. Then the two divide to make four. And so on, with the number of cells rapidly doubling and redoubling. At this stage there is no growth, no inflation. The original volume of the fertilized egg is literally divided, as in slicing a cake, and we end up with a spherical ball of cells which is the same size as the original egg. It’s not a solid ball but a hollow one, and it is called the blastula. The next stage, gastrulation, is the subject of a famous bon mot by Lewis Wolpert: ‘It is not birth, marriage, or death, but gastrulation, which is truly the most important time in your life.’
Gastrulation is a kind of microcosmic earthquake which sweeps over the
blastula’s surface and revolutionizes its entire form. The tissues of the embryo become massively reorganized. Gastrulation typically involves a denting of the hollow ball that is the blastula, so that it becomes two-layered with an opening to the outside world (see the computer simulation on p. 231). The outer layer of this ‘gastrula’ is called the ectoderm, the inner layer is the endoderm, and there are also some cells thrown into the space between the ectoderm and endoderm, which are called mesoderm. Each of these three primordial layers is destined to make major parts of the body. For example, the outer skin and nervous system come from the ectoderm; the guts and other internal organs come from the endoderm; and the mesoderm furnishes muscle and bone.
Neurulation
The next stage in the embryo’s origami is called neurulation. The diagram on the right shows a cross-section through the middle of the back of a neurulating amphibian embryo (it could be either a frog or a salamander). The black circle is the ‘notochord’, a stiffening rod that acts as a precursor of the backbone. The notochord is diagnostic of the phylum Chordata, to which we and all vertebrates belong (although we, like most modern vertebrates, have it only when we are embryos). In neurulation, as in gastrulation, invagination is much in evidence. You remember I said that the nervous system comes from ectoderm. Well, here’s how. A section of ectoderm invaginates (progressively backwards along the body like a zip fastener), rolls itself up into a tube, and is pinched off where the sides of the tube ‘zip up’ so that it ends up running the length of the body between the outer layer and the notochord. That tube is destined to become the spinal cord, the main nerve trunk of the body. The front end of it swells up and becomes the brain. And all the rest of the nerves are derived, by subsequent cell divisions, from this primordial tube.*