[Bullet Magnet]

You're not attempting to prove or establishing a link there, the one is a necessary result of the other. It's like dividing the population into people in prison and people not in prison and noticing that those in prison are more likely to have been convicted of a crime at some point. It's like, wow! Great work there.

But then you might notice that, in America at least, the prison population is disproportionately black compared to the non-prison population, and thereby conclude that black people are genetically predisposed toward criminal activity, and if I have to explain what's wrong with that conclusion then I have severely misjudged the character and intellect of this forum.


Your coma example has another, much more subtle difference, in that unlike the man/woman categories, by describing the two options as x and not x you have by the law of identity, outlined logically exclusive and exhaustive categories. You might instead have listed "people in comas" and "people who are healthy", which are not at all exhaustive: you don't have to be in a coma to be unhealthy.

The point here is that "man/woman" are not exhaustive in the same way as "coma/not coma". The category of "woman" is not equivalent to the category of "not man", and you make numerous errors both technical and social when you divide the people of the world that way at the start of your investigation, and it will get you wrong answers. It might work most of the time, but when reality produces exception you will struggle to deal with it, and we have enough real-world tragedies to look to to know how that can go. I'm hoping that this is not as mind-blowing to some of you here as I fear it might be.