Point/Counterpoint

The need for empirical research into the practicality and efficacy of software development methods is obvious but most published papers have inadequate experimental design. The assumption that what programmers do is "natural," and somehow right or practical, needs to be questioned seriously. Human beings haven't evolved by natural selection to be good programmers. There are people alive today who worked on the first electronic computers. Further, almost all of today's programmers learned from earlier programmers; either they were explicitly taught or they observed how the programmers that preceded them had done their work. If those pioneers were wrong, the methods that we now perceive as natural or intuitive will also be wrong. We can't simply conclude that what we observe in projects today is the best way to do something.