Much of the research in programming languages has been to develop languages that allow programmers to write programs that can be easily read and understood by humans. To what extent should a programmer be required to use such capabilities? That is, to what extent is it good enough for the program to perform correctly even though it is not well written from a human perspective?