Often it's a tangent, or supplementary information. Either way, all of these add something to the discussion; it's hard to make a positive comment that's actually interesting.

Middlebrow dismissal. https://news.ycombinator.com/item?id=4693920

Those who agree with the linked article tend to be less motivated to vote a concurring comment than those who disagree with the article and can't wait to express their disagreement via voting for a contrarian argument.

Things get interesting at the 2nd level; the highest voted reply to the top comment is usually a counter-counter-argument. The roles have been reversed and the yea-sayers are now more motivated to vote than nay-sayers. It's fascinating

I wonder if it's because of ties to academia. The academic origins of the internet, the Google/Stanford thing, the generally high quantity of people with degrees...

Academia is all about criticism. You don't succeed by doing something well, you succeed by situating something new in a context of previous work. That requires you first to understand previous work and its limitations.

Even in the business world, which is quite competitive, you can follow in footsteps. You can see how someone else created a hot dog stand, and just make your own. You don't have to find some flaw in the other hot dog stand, you just have to do it.

Academia demands you to find a flaw in the other hot dog stand and then argue that yours is different.

Coders seem to follow along in that path. We are always needing to find the next thing, so we are constantly finding ways to argue the existing stuff won't do.

On the other hand if one reads from the bottom of the page the calibre of comment leaves a lot to be desired.

Ergo, scroll half way down and start reading. Problem solved.

Maybe you're reading HN submissions about things that people think are bad?