So, one discourse, as a package of narrative stuff, labels guerrilla fighters as terrorists. A different discourse labels them as freedom fighters. You can absorb and use components of either or both of these discourses into your running narrative, or not.
Some discourses are widely accepted, and form a hefty part of the foundations of society. For example, there are a number of discourses that describe specific labels and roles for men and women. There's a discourse for "breadwinner and housewife", and all the stuff that piles along with those labels, which has plenty of near relatives. These have been accepted long enough that even while they're being attacked and denied, the repercussions of them (wage differences, glass ceilings) live on. Getting them out of the foundations is hard; they're poured right into the bedrock of many institutions, ingrained into the conversation, assumed by the rules, and so on.
That example is especially handy, because any acceptance of it (including a harsh rejection of it) changes who you'll listen to, who has and doesn't have credibility. In much of society, it means that women's ideas about work, money, and many other topics are given less credibility. In groups that are engaged in a deep rejection of it, it can mean that the ideas of men are looked on with suspicion (but only "can", not "does" - there are as almost as many forms of rejection as there are of acceptance). This means that it affects the way that further discourses are handled and absorbed, changing the rules for what can come after.
Biased discourses create unequal access to the shaping and formation of further discourse. This likely isn't a shocking idea to most readers, but it absolutely needs mentioning. And it does reach further than sexism and racism; it also applies to expertise. The "Ivory tower intellectual" and the "Ignorant nobody" are both discourse-based attempts to change the landscape of credibility.
Each branch (or sub-branch) of science, then, is a large and difficult discourse that has subjected itself to regulation. They're difficult because they don't narrate all that well; the easily-narrated versions of each branch are the "folk science" version. They're subject to regulation in that there's a broadly-accepted scientific method, and specific tools in each branch - all in the service of updating the discourse and attempting to make that discourse, when employed in your narrative, a better tool for simulating objective reality.
And, no, this post isn't subject to those methods. Which makes it folk science. Specifically, this is folk sociology.