Wednesday, October 24, 2007

Watch Out for the Razor!

In my previous adages column (see "Obey Murphy - It's the Law!") I started down the road of discovering and discussing those "named laws".

While there are many more adages out there, here are a few I found worth mentioning:

  1. Name: Ockham's Razor (sometimes called "Occam").
    The adage: "Entia non sunt multiplicanda praeter necessitatem" (Explanations should never multiply assumptions without necessity).
    The person: William of Ockham
    Background: William was a Franciscan friar from the village of Ockham in England, and one of the major medieval philosophers.
    General interpretation: When two explanations are offered for a phenomenon, the simplest explanation is preferable.
    The term "razor" is used to "shave off" the unwanted explanations.
    Reality: Yes, it's possible that aliens stole your car. It's more probable that you forgot where you parked it.
    In software development: you can write this amazing recursive algorithm to sort through your amazingly well designed new data structure in O(nlogn), or you can use a standard collection and call the Sort() method...
    Recommended reading: What is Ockham's Razor?
  2. Name: Clark's Law.
    The adage: Any sufficiently advanced technology is indistinguishable from magic.
    The Person: Arthur C. Clark.
    Background: a leading Sci-Fi writer (2001: A Space Odyssey) and an inventor/scientist.
    General interpretation: when no easy explanation is available, people will attribute any phenomenon to a higher power. Think of the light bulb. Now think what would be the reaction to it had you introduced it a 1000 years ago to a superstitious crowd.
    Reality: I'm not stepping into this trap :).
    Recommended Reading: Benford's Law.
  3. Name: Peter's Principle.
    The adage: In a hierarchy every employee tends to rise to his level of incompetence.
    The Person: Laurence J. Peter.
    Background: Peter was an educator specializing in hierarchies and bureaucracies. He first introduced the principle in his 1969 book titled... "The Peter Principle".
    General interpretation: people in an organization are always at one level above where they should be. Some team leads would be excellent developers. Some managers were excellent consultants. Sadly, they do not fit their current position.
    Reality: I'm not stepping into this one either :).
    Recommended reading: The Peter Principle.
  4. Name: Brook's Law.
    The adage: adding manpower to a late software project makes it later.
    The person: Fred Brook.
    Background: Fred Brook was one of IBM's lead developers and in charge of its greatest undertaking in the 60's: the IBM 360 project. He documented his many experiences of that complex project in "The Mythical Man-month" - a book that became the cornerstone of project management books.
    Reality: from my experience, this is a true observation. New people, dropped into a late project, lack the history and knowledge of the project's requirements and origins. They also tend to think of themselves as "the solution" and the existing group as "the problem". Even if everyone play nice together, expensive time will be wasted on inter-team knowledge transfer.
    Recommended reading: The Mythical Man Month.
  5. Name: Stigler's law of eponymy.
    The adage: No scientific discovery is named after its original discoverer.
    The person: Stephen Stigler, a University of Chicago statistics professor.
    The background: Stigler published his law in a 1980 paper titled... "Stigler's law of eponymy".
    General interpretation: throughout history, people published results based on others' research, plagiarized papers, or just plainly stole the works of others.
    Stigler's law just states the obvious: the person whose name is associated with a theory is either the last in line to work on in, or the first in line to steal it.
    Reality: funny enough, the full law should read "No scientific discovery, not even Stigler's law, is named after its original discoverer" - since Stigler attributed the original law to Robert K. Merton - a renowned sociologist.
    Recommended reading: What is eponymy?

Well, I had to finish with Stigler's law - as it may render the rest moot :)
That's it for now. More in the future (?)

No comments: