Not even wrong / by Digvijay Singh

Few weeks ago, I was watching this long interview of Prof. Sydney Brenner. At one point, he used a phrase that Physicist Wolfgang Pauli used a lot. The phrase was 'Not even wrong'.  

What does it mean ? As the phrase suggests, it is used to describe a model that has an intrinsic logical fallacy or in some cases a string of loosely connected ideas which are individually not wrong but make no coherent sense in combination.  The definition I described is negative, but there are important lessons from this:

1. A wrong model is the second most useful model i.e. after a correct model. 

2. Never propose a model where you are playing too safe and conservative due to the fear of proposing a wrong model. A model too afraid to make a definitive bold statement does not advance our understanding. Ofcourse you want to be very rigorous about testing the bold statement but a 'hesistant' model is not too far from 'not even wrong' model. Or is it ?

Fundamentally all proposed modes are not completely right but some are useful and a 'hesistant' model is not even useful.

Here is a fun example of gibberish biological statements that together make no sense, but are not particularly wrong individually. It's a scene from one of my favorite movies Blade Runner.