jump to navigation

Revisiting the Classics: Foundation May 12, 2011

Posted by futurewired in Isaac Asimov, Revisiting the Classics.

OK, I’d been planning to do a post on Foundation for a little while, but on Monday morning I happened to wander over to io9 only to discover that as part of their Blogging the Hugos section they’ve made this week Foundation week and have basically said most of the things I was planning to on the first day. Here’s a link:


So for that reason I’m just going to say a few things about on one aspect of the first Foundation book, namely the concept of psychohistory.

(PS if you’re wondering why that particular cover, it just happens to be the one I own and I seem to have made that an unofficial rule so far.)

When I first read Foundation, on holiday in Ireland circa 2003, I was fascinated by the idea of psychohistory; applying to laws of statistics to large groups and societies in order to make predictions about the future. Basically a mash up of historical analysis, sociology and mathematics that, in the Foundation universe, can be used to model large complex groupings and thus predict their behaviour over the course of thousands of years. Its pioneer, Hari Seldon uses it to predict the decline and fall of the galactic empire and to devise a plan to reduce the resulting dark age down to a mere thousand years. A crucial aspect of psychohistory is that its predictions can only be valid of the participants are unaware of them, meaning that Seldon’s plan necessarily involves an element of deception.

Something about this really caught my imagination at the time, but after five years of studying history and political science I’m not so sure. For starters, its a fairly woolly subject area that has resisted mathematical modelling fairly successfully – there are just too many variables that can’t be reduced down enough to be expressed as examples of a common type, either symbolically of mathematically.

Secondly it can be surprisingly difficult to define what the variables are, how they impact on specific events, even what those events are. A couple of examples I remember coming up from my MA were the hypothesis that the first world war was caused by the railways – an only slightly facetious argument that that plans for troop mobilisation were organised via the railway system meant that once the order was given by one side it was almost impossible to bring it to a halt -, when the second world war began – more complicated than you might think given the Sino-Japanese war began in 1937 but ended with the Japanese surrender in 1945 – as well as the question of what caused the Russian and Chinese revolutions.

This last one I think is particularly interesting because its an example of comparative macrohistorical analysis that should allow potentially for future predictions, but in actual fact its extremely limited in that respect.  You can identify common factors in the two and posit them as being key causal factors, but you can also identify vital features of each that did not occur in the other. You also have the issue of validating them through other examples in which their absence lead to there not being a revolution – whats called the double non-event. Further difficulties appear when you consider how limited the examples for comparison can be, particularly when they are dependent on certain historical structural factors, that can limit them to particular periods of history, thus rendering them useless in terms of predictive qualities.

I may seem to have wandered off topic here so I’ll try to bring it back to sf. In Foundation Seldon claims to have predicted the fall of the galactic empire through creating a mathematical model and extrapolating from current trends. Since this is explicitly the first galactice empire there can be no possibility of having a like for like example to compare with, so its possible he modelled up from smaller societies in order to predict the results of these trends eg centralisation of power, aristocratic class – but don’t get me started on the minefield of using class as a tool for comparative analysis. Now I know you can say there are lots of forms of modelling that don’t require comparative analysis, but they tend to deal with factors that are either static or evolve in accordance with certain rules. In addition the method of modelling itself is testable in a way. Can the same be said of psychohistory? While the prediction of the decline of a gigantic entity like the galactice empire can only be proven by its actual occurance, psychohistory is shown to work, predicting when Seldon will be arrested and what planets his Foundation might be exiled to. Somehow though, I just don’t buy it. Mathematics is great and can be used in all sorts of brilliant and weird ways but when it comes to predicting the dynamics of future history the problem isn’t finding the equation to input factors and trends into, it’s defining what those inputs actually are, what kind of impact they have, and finding any that a room full of historians and political scientists wouldn’t get into a fight over.

It would be interesting to imagine an alternative Foundation in which Seldon, unbeknownst even to himself, is actually just really good at reading current events in an intuitive fashion; all the number stuff is just smoke and mirrors. The galactic empire trundles on, or falls three months later to be replaced by a giant ad-hocracy, or a new technology appears on the scene to transform the social structure of the galaxy, ala the internet. Seldon dies never knowing he was essentially the Clever Hans of science fiction.

It would have made a pretty rubbish book though. And definitely one without any sequels.

That was a surprisingly curmudgeonly post I think and probably an unfair one. I greatly enjoyed re-reading Foundation and I’ll be doing the same for the sequels when I get around to it. A bit of suspension of disbelief would probably do me good, as well as some understanding for what the possibilities of psychology looked like in 1951 compared to today. Oh well!



1. Redhead - May 12, 2011

I read Foundation for the first time when I was in college. Enjoyed it, but didn’t really understand much of it. fifteen years later, I suppose it’s time for a re read.

Your post wasn’t really that curmudgeonly!

futurewired - May 13, 2011

It’s worth a second look but the dialogue’s a lot more over the top than I remember…

Oh and on the curmudgeon point I’m just paranoid about turning into a grumpy old man a couple of decades early!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: