Moreover, we know what a watch is for, and this colours our thinking. Suppose, instead, that our nineteenth-century heath-walker chanced upon a mobile phone, left there by some careless time traveller from the future. He would probably still infer `design' from its intricate form ... but purpose? What conceivable purpose would a mobile phone have in the nineteenth century, with no supporting network of transmission towers? There is no way to look at a mobile phone and infer some evident purpose. If its battery has run down, it doesn't do anything. And if what was found on the path was a computer chip - say, the engine manager of a car - then even the element of design would be undetectable, and the chip might well be dismissed as some obscure crystalline rock. Chemical 'analysis would confirm the diagnosis by showing that it was mostly silicon. Of course, we know that these things do have a designer; but in the absence of any clear purpose, Paley's heath-walker would not be entitled to make any such inference.
In short, Paley's logic is heavily biased by what a human being would know about a watch and its maker. And his analogy breaks down when we consider other features of watches. If it doesn't even work for watches, which we do understand, there's no reason for it to apply to organisms, which we don't.
He is also rather unfair to stones.
Some of the oldest rocks in the world are found in Greenland, in a 25-mile-long band known as the Isua supracrustal belt. They are the oldest known rocks among those that have been laid down on the surface of the Earth, instead of rising from the mantle below. They are 3.8 billion years old, unless we cannot reliably make inferences from observations, in which case the evidence for cosmic design has to be thrown out along with the evidence of the rocks. We know their age because they contain tiny crystals of zircon. We mention them here because they show that Paley's lack of interest in `stones', and his casual acceptance that they might have `lain there forever', are unjustified. The structure of a stone is nowhere near as simple as Paley assumed. In fact, it can be just as intricate as an organism, though not as obviously `organised'. Every stone has a story to tell.
Zircons are a case in point.
Zirconium is the 40th element in the periodic table, and zircon is zirconium sulphate. It occurs in many rocks, but usually in such tiny amounts that its presence is ignored. It is extremely hard - not as hard as diamond, but harder than the hardest steel. Jewellers sometimes use it as a diamond substitute.
Zircons, then, are found in most rocks, but in this instance the important rock is granite. Granite is an igneous rock, which wells up from the molten layers beneath the Earth's crust, forcing a path through the overlying sedimentary rock that has been deposited by wind or water. Zircons form in granite that solidifies about 12 miles (20 km) down inside the Earth. The crystals are truly tiny: one 10,000th of an inch (2 microns) is typical.
Over the last few decades we have learned that our apparently stable planet is highly dynamic, with continents that wander around over the surface, carried by gigantic `tectonic plates' which are 60 miles (100 km) thick and float on the liquid mantle. Sometimes they even crash into each other. They move less than an inch (about 2 cm) per year, on average, and on a geological timescale that's fast.
The north-west of Scotland was once part of North America, when the North American plate collided with the Eurasian plate; when the plates later split apart, a piece of America was left behind, forming the Moine thrust. When plates collide, they slide over each other, often creating mountains. The highest mountains on Earth today, the Himalayas, formed when India collided with the Asian mainland. They are still rising today by more than half an inch (1.3 cm) a year, though are often weathered away faster, and India is still moving northwards.
At any rate, granite deep within the Earth may be uplifted by the collision of continental plates, to appear at the surface as part of a mountain range. Being a hard rock, it survives when the softer sedimentary rocks that surround it weather away. But eventually, even granite weathers, so the mountain erodes. The zircon crystals are even harder, so they survive weathering; they separate out from the granite, to be washed down to the coast by streams and rivers, deposited on the sandy shore, and incorporated into the next layer of sedimentary rock.
As well as being very hard, zircon is chemically very stable, and it resists most chemical changes. So, as the sediment builds up, and the zircon crystal is buried under accumulating quantities of incipient rock, the crystal is relatively immune to the increasing heat and pressure. Even when the rock is cooked by deep heat, becoming metamorphic - changing its chemical structure - the crystal of zircon survives. Its one concession to the extreme environment around it is that eventually it builds a new layer, like a skin, on its surface. This `rim', as it is called, is roughly the same age as the surrounding rock; the inner core is far older.
Now the process may repeat. The core of zircon, with its new rim, may be pushed up with the surrounding rocks to make a new mountain range. When those mountains weather, the zircon may return to the depths, to acquire a second rim. Then a third, a fourth ... Just as tree rings indicate the growth of a tree, so `zircon rims' reflect a sequence of mountain-building and erosion. The main difference is that each ring on a tree corresponds to a period of one year, whereas the rims on the tiny zircon crystal correspond to geological cycles that typically last hundreds of millions of years. But, just as the widths of tree rings tell us something about the climate in the years that are represented, so the zircon rims tell us something about the conditions that occurred during a given geological cycle.
By one of those neat coincidences that Paley would interpret as the Hand of God but nowadays we recognise as an inevitable consequence of the sheer richness of the universe (yes, we do see that those statements might be the same), the zirconium atom has the same electric charge, and is much the same size, as an atom of uranium. So uranium impurities can easily sneak into that zircon crystal. This is good for science, because uranium is radioactive. Over time, it decays into lead. If we measure the ratio of uranium to lead then we can estimate the time that has elapsed since any given part of the zircon crystal was laid down. Now we have a powerful observational tool, a geological stopwatch. And we also have a simple prediction that gives us confidence in the hypothesis that the zircon crystal forms in successive stages. Namely, the core should be the oldest part of the crystal, and successive rims should become consistently younger, in separate stages.
A typical crystal might have, say, four layers. The core might date to 3.7 billion years ago, the next to 3.6 billion years, the third to 2.6 billion years, and the last one to 2.3 billion years. So here, in a simple `stone', we have evidence for geological cycles that last between 100 million and one billion years. The order of the ages agrees with the order in which the crystal must have been deposited. If the general scenario envisaged by geologists were wrong, then it would take only a single grain of sand to disprove it. Of course that doesn't confirm the huge geological cycles: those are deduced from other evidence. Science is a crossword puzzle.
Zircons can teach us more. It is thought that the ratio of two isotopes of carbon, carbon-12 and carbon-13, may distinguish organic sources of carbon from inorganic ones. There is carbon in the Isua formation, and the ratio there suggests that life may have existed 3.8 billion years ago, surprisingly soon after the Earth's surface solidified. But this conclusion is controversial, and many scientists are not convinced that other explanations can be excluded.