Sometimes they do. A single mistake might be neutral or beneficial. Suppose a couple of entrepreneurs plan a weekend community newspaper/site. In the promos they send out to potential advertisers, they call their proposed publication “The Sundry Times.” There is one wrong letter (bit) of information in the title; they had meant to write: The Sunday Times.
But that wrong bit of information may turn out to be neutral. While “The Sundry Times” is hardly a grabber, it conveys at least some meaning as a name for a publication. We could consider it a neutral mutation. If the community their proposed paper is intended to serve happens to be called Sundry Place, the accidental name may be widely accepted, and therefore beneficial.
But what if the flyers asking for advertisers’ cash had gone out introducing the “Sundzy Times”? The “Sunfay Times”? Or the “Sundy Times”? Would you advertise with a publication that makes such errors in its own promotional copy? We could consider that extinction.
As anyone who types knows, there are many more ways to make such a mistake that are harmful than helpful. And as any reporter knows, no long chain of random letters is going to turn out to be a local news story by accident.
It works the same way in genetics. Mutations are far more likely to introduce a problem than a benefit. And no long string of such mutations is going to happen to produce a complex new physiological system.
*Australian biology writer Stephen E. Jones offered an initial similar analogy which I (O’Leary for News) have slightly reworked: Stephen E. Jones, found at ASA List, May 14 2000: http://tinyurl.com/9srssjq