How Do I Know?

1crowA month ago, I hurt my Achilles tendon.  This week the downer of not being able to run drove me to admit I was injured, and seek internet advice.  The process was time-consuming, but straight forward.  First, I searched for a remedy to my problem.  Then I used the top few results to provide new key words, and iterated.  Along the way, to cull applicants, I took liberty in judging the plausibility of each author’s recommendation based on: personal experience, the reasonability of any proposed mechanisms, the author’s credentials, how often I saw the same advice, and whether or not the author bothered to proofread.  In the end, I had a course of action I could fake myself into thinking would work.  But how could I know?

I has often been said, “correlation is not causation”.  That makes sense.  Pizza burning the roof of your mouth, just before the internet goes out, doesn’t mean the eating disaster disconnected you from the world.  But, it doesn’t mean the pizza burn didn’t cause the internet to fail.  To resolve the true nature of causation we often look for more correlation, or attempt to derive the observation from some accepted theory, or we simply defer to authority.

Looking for more evidence is good science.  Before making an observation, in the absence of any theory, the odds of any one event happening are 1:2.  Once a thing has happened, the odds of it happening again are still 1:2, but the odds of that observations being ubiquitous are 3:4.  This continues with every successful observation; three observations have 5:6 odds of ubiquity, four observations have 7:8.  Thing is, the ratio never really gets to 1:1, and any failure to repeat the correlation must be discredited, neglected or assimilated to prevent killing the causation theory.

Discovering a theory, from data, is solving a system of equations with infinite variables and as many equations as data points.  Having finite data points will always leave the theorist with infinite possible theories.  Being mostly practical, and having a limited capacity for random crap, humans are inclined to narrow the theory pool to a handful.  The likelihood one idea will be accepted seems to rely on a cocktail of familiarity, proliferation, and pedigree.  If a theory shares characters and actions with one we already believe, then it can easily become canon without significant effort.  In the absence of convenience, repetition helps to cement a new idea (much like multiplication tables).  Barring all that, if your daddy says its true, then it is.

A causal theory seeks to eliminate the possibility that an effect can be observed without its cause.  Objective observation can only report what was observed, perpetually leaving open the possibility that correlation won’t be observed.  With no means of scientific proof, and a limited capacity for theories, we are left to rationalize simple hypotheses, repeat them, and agree upon what is true.  Correlation is not causation…consensus is causation.


P.S. Jan’s gone all Pythagorean with her folding.


Leave a Reply

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

You are commenting using your 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