[FoRK] Why can't PhDs program?

Stephen D. Williams sdw at lig.net
Sun Jul 6 01:30:26 PDT 2008

I always felt that programming had to be learned by doing a lot of it
and reading a lot of it. You need to get to plateaus of understanding,
or milestones of insight, and get them ingrained so that you think in
those concepts.

This is not to say that the theory, mathematical understanding, logic,
proofs, etc. aren't important, but that they don't naturally provide a
code-producing / code-thinking brain.

My perspective is hack-as-a-kid, get-a-computer-job-as-a-kid, learn to
be professional, accomplish some cool stuff along side
engineers/PhDs/professional developers, then learn the academic
perspective (more as a curiosity / filling out than anything and because
I want to be able to express myself in that realm), so YMMV.


Joseph S. Barrera III wrote:
> Or: how can you be so smart but program so poorly?
> One of the primary skills in programming is to recognize
> patterns and repetition and to encode such in common
> routines so that future changes need be made only once.
> Isn't this the same skill required for finding patterns and
> producing proofs? But then why are so many Math &c
> PhDs so incredibly bad at programming, producing
> line after line of (imperfectly) duplicated code? I don't get it.
> Is it that code doesn't appear to have the same purity as
> mathematics, and thus doesn't deserve the same care and
> respect? It is a question that drives me to distraction.
> - Joe

More information about the FoRK mailing list