Re: Algorithms and Creativity

From: HARNAD Stevan (
Date: Fri Jun 07 1996 - 22:50:11 BST

> From: "Belcher, Ian" <>
> Date: Tue, 28 May 1996 13:57:09 +0100 (BST)
> An algorithm is a mechanical procedure which manipulates symbols using
> a series of rules/codes that one can follow, in order to obtain a
> result. An example would be building a chair with instructions, like
> "...take the base, and attach all four legs using screws..." The
> algorithm must be mechanical, so that you don't have to think about it,
> or know what it means, (thus no homunculus is needed.)
> Creativity on the other hand reuires a homunculus.

Do you really mean that? Another mind inside our minds? Is creativity
then something, unlike the ordinary things we do, unconsciously,
something that we do superconsciously? But a homunculus can't be
necessary for creativity, otherwise there would be no creativity, as
there is certainly no homunculus!

> When we are
> creative, we may do alot of thinking which requires understanding. For
> instance, the person who invented the computer realised that there was
> a need for some sort of information-storing machine, and set about
> creating one. This person had to understand firstly there was a need
> for a computer, then had to design one, using his creativity. How could
> you "program" something or someone with a series of rules making them
> creative? To design something original, based on no previous knowledge
> of what is being designed is unalgorithmable!

Is ordinary cognition "algorithmable"? Is this a Granny objection?
Explain why a rule couldn't do it: Because part of what we DEFINE as
creative is that we don't yet know a rule for accomplishing something;
if we did, we wouldn't call simply applying it "creative." But we often
don't know the algorithms we use even in noncreative activity. So
couldn't there be an (unconscious) algorithm for creativity after all?

> In conclusion therefore, I believe that one cannot have an algorithm,
> (something that doesn't include a honunculus,) for creativity,
> (something that does require a homunculus.)

I'm afraid you'll have to go back and sort out the homunculus; you don't
seem to have appreciated the homunculus problem: It was used as an
argument AGAINST imagery; why would you want to make creativity open to
the same objection?

This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:47 GMT