An article by Moldbug got me thinking on what the null hypothesis would truly represent. The null hypothesis works on the principle of the Presumption of Innocence, evident in the United States justice system enshrined in the directive “the defendant is innocent until proven guilty”.
This is a messy directive which is fully in contradiction with an average human’s thought process. After all, if it looks like a duck, swims like a duck, and quacks like a duck, then it probably is a duck. Logical reasoning is better represented through the work of Thomas Bayes, a Reverend who proposes a prior conviction rather than beginning with no information at all. Moldbug describes it as such:
A Bayesian court would reason as follows: the defendant is a known ghetto gangbanger (prior conviction) accused of slinging rock (evidence). His attorney asserts that this “rock,” which officers somehow mislaid without entering into evidence, was actually Finish brand dishwasher detergent. However, since defendant has six priors for aggravated narcotics distribution and there is no evidence that he has ever actually done the dishes, we calculate the probability of his story as… etc.
One can plainly see the utility of Bayesian thinking, strictly in daily life. Decisions are made on the basis of intuition more often than not, the ever present “gut feeling” guides our hand. This is an informed opinion on the decision which is completely emprical. An observation updates beliefs about the world, which act as a basis for decisions.
This would make the null hypothesis wholly incompatible with the Hindu notions of Karma. This author has long maintained the view that Kismat (to use the urdu word) is a multi-dimensional Bayesian probability distribution and Karma the manner in which it expresses itself.
What do we mean by Bayesian? Consider Karma to be a vector of a persons attributes, something we could think of as a state vector in the control systems theory sense (but the analogy is somewhat imprecise). A “Karma state” is thus a vector of properties such as experiences and opportunities. Experiences do not change with time but only expand , or get replaced (a good experience at a restaurant replaces one’s earlier bad experience). Opportunities however get updated every time an action takes place, by which we mean they grow, remain the same or shrink.
This action is dependent on present opportunities and once an action occurs it results in post-action opportunities. A trivial example is once I spend money, that money can no longer be used elsewhere i.e a new set of opportunities now exist which previously did not.
This follows quite rationally – an intelligent hardworking individual invariably attains success of some measure while lazy individuals remain static. The former individual uses opportunities to act and produce greater opportunities. The latter stagnates.
There are other aspects to Karma which are not fully addressed here, such as the metaphysical idea of ‘luck’. Luck can be thought of as a random variable thrown into the mix which changes the probability of opportunities.
These conclusions, lead me to fundamentally question the nature of the ad hominem and what qualifies as one. Is it truly unfair to attack the holder of a notion than the notion itself? What if the holder is predisposed to have silly notions? If so what relationship must the personal attack and the notion have to justify such a position?
This requires us to qualify what a silly notion is before discarding both the notion and the person. This is a great tool in dismissing ideas like feminism, secularism and the much lauded Ganga Yamuni Tehzeeb, resulting in the subsequent dismisal of the holders.
Here the crux of the ad hominem strikes us, does the opposite hold? A person may hold incorrect notions but it does not mean all their notions are wrong. How do we take the detour around this minor inconvenience?
Reject the null hypothesis.