To explain how the blind process of natural selection—a process dependent on random events (mutations)—can generate complex adaptations, richard dawkins invites us to imagine that an evolved attribute is like an english sentence—like a line from shakespeare’s hamlet, such as, methinks it is like a weasel (dawkins 1986). the odds that a monkey would produce this line by tapping at a typewriter are vanishingly small, 1 in 10,000 million million million million million million (1 in 1040). these are not good odds. but instead of trying to get a monkey or a computer to get the “right” sentence in one go, let’s change the rules so that we start with a randomly generated letter set, such as swajs meiurnzmmvasjdna ypqzk. now we get a computer to copy this “sentence” over and over, but with a small error rate. from time to time, we ask the computer to scan the list and pick the sequence that is closest to methinks it is like a weasel. whatever “sentence” is closest is used for the next generation of copying, again with a few errors thrown in. the sentence in this group that is most similar to methinks … weasel is selected to be copied, and so on. dawkins found that this approach required only 40 to 70 runs (generations) to reach the target sentence. what was dawkins’s main point in illustrating what he called cumulative selection? in what sense is this example not a perfect analogy for natural selection?