Numbers, Damned Lies & Statistics Again


Time to get onto what Tarde said about figures, or numbers.

We think about numbers and figures in general all inside out, back-to-front and upside down. If you want to find some of the most committed mysticism in the world today, look no further than the way we all grovel before facts and figures. There's a pretty widespread belief or assumption that when you 'quantify' something you get closer to the reality or truth of it. So numbers and figures aren't just practical things, they have this veneer of mysticism added to them - they somehow access reality better than everyday words can, for example.

Added to that is our usual idea that you can only really say accurate things with numbers when you're applying them to large numbers of things. If you use numbers for small numbers of things, you get what statisticians call a small 'sample size', which creates (apparently) inaccurate results. Social scientists do this all the time, they feel that they can't possibly say anything quantitaively valid about 'social' phenomena unless they use some huge sample size, of thousands of people. They emulate what they think 'natural' scientists do (and what natural scientists think they do as well), namely collect the data, churn the numbers, and come up with laws. They truly feel they can't say quantitative things unless they work at the 'top', aggregate level of things.

It's all much more simple, Tarde realised. Natural scientists use huge numbers of things because they don't have a choice. When you study gases, for example, you have no way to access all of the little movements of the billions of gas molecules that go together to create a phenomenon. So you study the aggregate of these huge numbers, and come up with some averaged result that describes the behaviour of this aggregate. But you don't know much at all about how the interactions of all of those billions of molecules came to produce that aggregate behaviour. However this is where the mysticism creeps in, because natural scientists and everyone who emulates and admires them then take this aggregate and say it's somehow the law 'behind' the movements of those billions of individual molecules. Completely back-to-front, they mistake what is merely the effect of all of those billions of interactions for the cause of these interactions.

What social scientists try to emulate then is actually the real and unavoidable weakness of natural science, not its strength. Social scientists are dealing with much smaller numbers of entities (usually people), and can therefore quite practically trace the way the aggregate derives from the interactions of all of the individuals. The internet makes this easier every day, you can increasingly trace in minute detail how huge, colective phenomena derive from millions of individual transactions, such as via Google Analytics (which tells me for example the spectacularly low numbers of people who bother coming here). But social scientists haven't registered that yet, and nor have most of the rest of us. They still like to think that the aggregate or pattern/law is what is somehow 'behind' the individual interactions.

At the end of the day numbers and figures are usually just measurements. And measurement never gets more complicated than putting a ruler on something and reading the number off the ruler. You can use much more complicated devices, but it's the same process every time. You're doing a completely practical thing, for practical reasons. Never does measurement take you behind the scenes of reality into some transcendental realm of truth. And you don't need to gather thousands or millions of things to start measuring, you can measure anything. And you won't get closer to reality the more things you measure, reality is actually always right there, in each individual. Aggregates of things are always much less complex than each of the individual things within the aggregate, just like governments are in fact massive simplifcations of the lives of everyday people, as I talked about last time.

What Tarde brilliantly realised is that things get much less complex the further away from the individual you go, not more. Our atomic way of looking at things always looks for simple building blocks (like atoms) at the foundation of things, which are then added together to create more complex things. But that's backwards. The closer to individual things you get, the more complex things get, as anybody who actually works with atoms knows. Who would ever really say that the government department dealing with social problems is more complex than any of the individuals and their problems they're dealing with? And yet we continue to think that 'big' things actually exist and are more complex than 'little' things or individuals. And we think that the patterns and numbers we derive from studying these aggregates of individuals are somehow the law that explains the individuals.

It's our mysticism. God isn't dead, we just gave him a calculator.

Comments

Popular posts from this blog

The Morality of a Speed Bump. Latour.

Reductio Ad Hitlerum, or what's wrong with Godwin's Law

Counterpoint (P.S.). Queen.