Friday, 4 July 2008

Machines and metaphors

A passing reference to Paul Ormerod in this post on Stumbling and Mumbling made me go and have a quick flick through my copy of Why Most Things Fail. I found it quite a challenging read, as he is very sceptical about the record of social democracy. But to be honest I have quite a bit of sympathy with the general pitch that it's very difficult to plot a successful course of action given an unknowable future. There's a great bit early on where he translates a statement from a director of GM about a new launch as basically stating "Our new product model might do well, or it might not. We don't know."

But the bit I was reminded of when flicking through the book last night was this nice little para late on where Ormerod is describing Hayek's views:

The visions of the world articulated by orthodox economics and by Hayek are fundamentally different. Conventional theory describes a highly structured mechanical system. Both the economy and society are in essence giant machines, whose behaviour can be controlled and predicted. Hayek's view is much more rooted in biology. Individual behaviour is not fixed, like a screw or cog in a machine is, but evolves in response to the behaviour of others. Control and prediction of the system as a whole is simply not possible.

The bit I really like is the reference seeing systems as machines. I think this is a widespread and fundamental misconception. In my bit of the world I think it plagues some attempts at systemic reform. I think that there is an implicit assumption, for example, that the investor-company relationship would function radically differently if only different information was being fed into the machine.

(Incidentally this is another reason why I continue to think that the unions could be a serious force here. There has always been an element of scepticism on the unions' part about SRI because of its failure to address labour issues effectively. As a result they have often gone off and done their own thing, often quite successfully. Union investor activism can be successful (in my view) because it has immediate goals, rather than a systemic approach. Unions still need to have a view on the systemic issues, but seem to get less bogged down in this area than SRI proponents.)

More broadly I think this conception of systems as machines is further evidence of the way we attracted to certain ideas and ways of viewing the world. Seeing systems as machines is comfortable because it implies that control is possible, and that you can have an idea of what the system working well looks like (and therefore how to achieve good results). I think these little short-cuts we use to describe bits of the world must have a fairly significant influence on how we understand it too.

And with that in mind, my next book purchase is going to be this.


Nick Drew said...

i may be back to you on this one Tom, but it's the w/end now ...

Tom P said...

I'm all ears!

Nick Drew said...

nothing earth-shattering !

you are IMO right that machine-metaphors abound, and lots of folk have inveighed against them (Ryle, Taleb ...)

and are (sometimes) too simplistic, and worse, may be fundamentally misleading

although probably more through incompleteness than utter uselessness - just as in physics when a wave-metaphor runs out of steam and a particle-metaphor (/model) is needed: but both are illuminating the same phenomena, and both have their uses

but for practical purposes we do need models. One of the wisest things ever said to me in a practical context was by a chap with no formal education beyond the age of 16, whose work depended citically (potentially life'n'death) on his being very much on the ball with understanding + predicting people's behaviour, & he said

the key is to divide people into enough categories to do them justice, but not so many that you lose sight of the similarities

(& his goals were 100% functional - no ethical overtones in the use of 'justice' here)

Tom P said...

I've never read any Ryle, maybe I should give it a bash.

I think you're right that metaphors actually play a useful role too. Otherwise how would we ever be able to transfer insights from one scenario to another that differs only in appearance.

The machine metaphor bothers me particularly when it is applied to scenarios involving human interaction. There is a tendency to characterise people as objects who will inevitably follow a certain path if a certain force/incentive is applied to them (or not).

btw I realise that a lot of the time the things I get most annoyed about are those that I suffer from myself. I'm sometimes think that blogging for me is a bit like arguing with a younger version of myself!