Do software-generated "connections" really generate inspiration?

Posted by Manfred
Oct 25, 2007 at 08:54 PM

Steve wrote: “Making software connections—in my opinion and experience—amplfies memory, not thinking. Is that a distinction worth making? I think obviously so, although exactly what the implications are may not be immediately obvious.If like me you are fond of analogies, the difference between the effect of enhancing thinking (intelligence) vs. memory is sort of like the difference between a faster computer processor on the one hand and more RAM and a larger hard disk on the other (assuming of course the hard disk space is actually used).”

We are obviously getting into some deep and controversial issues here. Philosophers and cognitive scientists hold various views on these matters. But without getting deeply into any of these, I would say the following.

Yes, the distinction between memory and thinking is worth making. However, and this is important, this does not mean that memory and thinking are independent of one another in the same way that one piece of hardware is independent of one another. They depend on another. One might also say that there are interconnections between the two at various levels. To vary a theme of one of my favorite philosophers: “thinking without memory is empty, while memory without thinking is blind.”

I would go further and say that with regard to any given problem, even the most brilliant thinker needs to know (or remember) the relevant facts. Indeed, to determine what are the relevant facts is often a first step to a solution of the problem. Though there are also instances where a solution shows that certain facts that one already knew are relevant, etc. etc.

Secondly, there are some philosophers and many cognitive scientists (called “connectionists”) who would argue that thinking is nothing but connecting. This goes back to philosophers like Hume, who thought that thinking is just a process of of association, and thinking about thinking meant identifying the principles of association. Some thinkers in this tradition have gone so far as to say that the human brain should be modeled as a neural net, and that you could program a computer to mirror a human brain by establishing associations between different bits of information. There is an interesting novel by Richard Powers, called Galatea 2.2 that explores this issue.

I have some sympathy for this kind of view—and that’s why ConnectedText appeals to me—-while I also think that connectionism is very problematic. Still, I would hold that making connections is a kind of thinking, and therefore I respectfully disagree with Steve on this issue.
Manfred