The simplest of autonomous DNA-copying systems on our planet are bacterial cells, and they need at least a couple of hundred genes to make the components they need. Cells that are not bacteria are called eukaryotic cells. Our own cells, and those of all animals, plants, fungi and protozoa, are eukaryotic cells. They typically have tens or hundreds of thousands of genes, all working as a team. As we saw in chapter 2, it now seems probable that the eukaryotic cell itself began as a team of half a dozen or so bacterial cells that clubbed together. But this is a higher-order form of teamwork and is not what I am {155} talking about here. I am talking about the fact that all genes do their work in a chemical environment put together by a consortium of genes in the cell.

Once we have grasped the point about genes working in teams, it is obviously tempting to leap to the assumption that Darwinian selection nowadays chooses among rival teams of genes – to assume that selection has moved up to higher levels of organization. Tempting, but in my view wrong at a profound level. It is much more illuminating to say that Darwinian selection still chooses among rival genes, but the genes that are favored are those that prosper in the presence of the other genes that are simultaneously being favored in one another's presence. This is the point we met in chapter 1, where we saw that genes sharing the same branch of the digital river tend to become “good companions.”

Perhaps the next major threshold to be crossed as a replication bomb gathers momentum on a planet is the Many-Cells Threshold, and I'll call this Threshold 4. Any one cell in our life form, as we saw, is a little local sea of chemicals in which a team of genes bathe. Although it contains the whole team, it is made by a subset of the team. Now, cells themselves multiply by splitting in half, with each one growing to full size again. When this happens, all the members of the team of genes are duplicated. If the two cells do not separate fully but remain attached to one another, large edifices can form, with cells playing the role of bricks. The ability to make many-celled edifices may well be important on other worlds as well as our own. When the Many-Cells Threshold has been crossed, phenotypes can arise whose shapes and functions are {156} appreciated only on a scale hugely greater than the scale of the single cell. An antler or a leaf, an eye's lens or a snail's shell – all these shapes are put together by cells, but the cells are not miniature versions of the large shape. Many-celled organs, in other words, do not grow the way crystals do. On our planet, at least, they grow more like buildings, which are not, after all, the shape of overgrown bricks. A hand has a characteristic shape, but it is not made of hand-shaped cells, as it would be if phenotypes grew like crystals. Again like buildings, many-celled organs acquire their characteristic shapes and sizes because layers of cells (bricks) follow rules about when to stop growing. Cells must also, in some sense, know where they sit in relation to other cells. Liver cells behave as if they know that they are liver cells and know, moreover, whether they are on the edge of a lobe or in the middle. How they do this is a difficult question and a much studied one. The answers are probably local to our planet and I shall not consider them further here. I have already touched upon them in chapter 1. Whatever their details, the methods have been perfected by exactly the same general process as all other improvements in life: the nonrandom survival of successful genes judged by their effects – in this case, effects on cell behavior in relation to neighboring cells.

The next major threshold I want to consider, because I suspect that it, too, is probably of more than local planetary significance, is the High-Speed Information-Processing Threshold. On our planet this Threshold 5 is achieved by a special class of cells called neurons, or nerve cells, and we might locally call it the Nervous System Threshold. However it may be achieved on a planet, it is important, because {157} now action can be taken on a timescale much faster than the one that genes, with their chemical levers of power, can achieve directly. Predators can leap at their dinner and prey can dodge for their lives, using muscular and nervous apparatus that acts and reacts at speeds hugely greater than the embryological origami speeds with which the genes put the apparatus together in the first place. Absolute speeds and reaction times may be very different on other planets. But on any planet an important threshold is crossed when the devices built by replicators start to have reaction times orders of magnitude faster than the embryological machinations of the replicators themselves. Whether the instruments will necessarily resemble the objects that we, on this planet, call neurons and muscle cells is less certain. But on those planets where something equivalent to the Nervous System Threshold is passed, important further consequences are likely to follow and the replication bomb will proceed on its outward journey.

Among these consequences may be large aggregations of data-handling units – “brains” – capable of processing complex patterns of data apprehended by “sense organs” and capable of storing records of them in “memory.” A more elaborate and mysterious consequence of crossing the neuron threshold is conscious awareness, and I shall call Threshold 6 the Consciousness Threshold. We don't know how often this has been achieved on our planet. Some philosophers believe that it is crucially bound up with language, which seems to have been achieved once only, by the bipedal ape species Homo sapiens. Whether or not consciousness requires language, let us anyway recognize the Language Threshold as a major one, {158} Threshold 7, which may or may not be crossed on a planet. The details of language, such as whether it is transmitted by sound or some other physical medium, must be relegated to local significance.

Language, from this point of view, is the networking system by which brains (as they are called on this planet) exchange information with sufficient intimacy to allow the development of a cooperative technology. Cooperative technology, beginning with the imitative development of stone tools and proceeding through the ages of metal-smelting, wheeled vehicles, steam power and now electronics, has many of the attributes of an explosion in its own right, and its initiation therefore deserves a title, the Cooperative Technology Threshold, or Threshold 8. Indeed, it is possible that human culture has fostered a genuinely new replication bomb, with a new kind of self-replicating entity – the meme, as I have called it in The Selfish Gene – proliferating and Darwinizing in a river of culture. There may be a meme bomb now taking off, in parallel to the gene bomb that earlier set up the brain/culture conditions that made the take-off possible. But that, again, is too big a subject for this chapter. I must return to the main theme of the planetary explosion and note that, once the stage of cooperative technology has been reached, it is quite likely that somewhere along the way the power to make an impact outside the home planet will be achieved. Threshold 9, the Radio Threshold, is passed, and it now becomes possible for external observers to notice that a star system has newly exploded as a replication bomb.

The first inkling external observers will get, as we have seen, is likely to be radio waves spilling outward as a {159} by-product of communications within the home planet. Later, the technological heirs of the replication bomb may themselves turn their deliberate attention outward to the stars. Our own halting steps in that direction have included the beaming out into space of messages specifically tailored for alien intelligences. How can you tailor messages for intelligences of whose nature you have no conception? Obviously it is difficult, and quite possibly our efforts have been misconceived.