Scott H Young

Archive for May, 2012

Why Should Life Be Simple?

Sunday, May 27th, 2012

Readers have often pointed out to me that my site has no consistent message. When you read other blogs, their philosophies are sharpened down to a point: simplicity, live consciously or nonconformity. In comparison, mine seems rather haphazard.

Part of this is focus—writers have opinions about many subjects, but they are only interesting when writing on a few of them. But I suspect a bigger part is that people are naturally drawn to simplified life philosophies. We want to be able to reduce life to a few generalized principles from which all truth can follow.

But why should life be simple? The complexity of even the simplest things is staggering, so why should life boil down to following only a few basic tenets?

Chaos Theory

Physicists often have a dislike for the social sciences. Richard Feynman notoriously claimed many of them were cargo cult sciences, trying to give the appearance of being scientific without upholding the same rigor. If they were truly doing science, why had they discovered no universal laws or principles?

While the merit of Feynman’s comments is questionable, it does raise an interesting point. Why have we discovered laws of physics that hold to amazing accuracy, but have failed to uncover similarly strong principles in sociology or economics? After all, truly few principles of economics could approach the confidence intervals and generality we have for Maxwell’s equations.

I suspect the answer is that simple equations don’t exist for economics or sociology. Sociology is built on psychology, which is built on biology, built on chemistry, built on physics. These disciplines rest upon an enormous tower of complexity, so it’s no wonder we haven’t uncovered an equation with the simplicity and breadth of e=mc^2.

Chaos theory tells us that many common systems can have wildly divergent results based on slight tweaks of the initial conditions. Given that life emerges from so many of these chaotic subsystems, is it any wonder we haven’t found the perfect equation telling people how to live?

Managing Complexity

For their part, the social sciences have seemed to adopt two different approaches to managing the complexity. The most common has been to make simple models, but accept the limitations or lack of generality they create.

Models are often necessary. Simplicity has a power to make decisions, even if they are sometimes wrong. A solution that works much of the time is better than giving up because no perfect solution exists.

But models only work if you accept that they are only approximations. Instead, many people like their models enough that they prefer them to an ambiguous reality. Models are only useful if you recognize the boundaries where they break down.

Learning new models for life can be useful too. But only if you are aware, and expect, the model to break at some point. The most useful truths are rarely universal.

Another, more recent, approach is to avoid trying to make a comprehensible model altogether. Machine translation researcher, Peter Norvig, argues here that machine learning models can better predict grammatical regularities than the theories of linguists. The challenge is that such billion-parameter models are opaque in how they generate predictions.

In many ways this is why learning-by-doing works. The complexity of the situations is too vast to be understood consciously, and the accuracy of any comprehensible model is too low. The average person is an encyclopedia of such unarticulated facts, guided by principles they would never be able to explain consistently.

Searching for Local Truth

Instead of trying to find the perfect set of universal principles, we’re better off trying to find local truths. Things which aren’t true for all people or cases, but which are true enough of the time to be worth believing in.

Exploring local truths means we also need to work harder to find where they break. We’re endowed with thousands of these local truths over the years of our life, but few people ever try to see where they stop working. Experience without experimenting is just reaffirming those little lies.

More than anything it means that, perhaps, universal principles don’t exist. If they do, they are probably more complex than anything religion or philosophy has yet to offer.

Posted by Scott Young on May 27th, 2012 in Personal Development | 31 Comments »