Skimming Lindsey & Pluckrose for A Thing, I run into the claim that postmodernists think science is just a metanarrative, supported by reference to Lyotard, whose argument is that science lacks a metanarrative (& thus finds itself in a crisis of legitimation, which it propagates)

Like, that's Lyotard's whole point about capitalist technoscience: absent any other form of legitimation, it ends up relying on criteria of pure performativity or effectiveness, whose goal is buffing the efficiency of the social/profit-making mechanisms in which it is embedded

L & P want to hook science back up with a liberal account of how it functions as a truth-seeking process equipped with self-correcting mechanisms - which is exactly the sort of thing Lyotard meant by a metanarrative

The "cynicism" of "Cynical Theories" just is the refusal to take this account of how things work on faith, which L & P construe as a moral collapse, rather like religious fundamentalists who argue that loss of belief in God leads inexorably to total ethical abandonment

This explains the relentless hyperbole of their account of "postmodernism" as an unstoppably corrosive radical skepticism which evacuates all consideration for truth and seeks only polemical leverage (they should read Foucault on how polemics are boring and pointless...)

Their mortal terror that science might be treated as "just another narrative" is a displaced fear that their liberal metanarrative ("science, democracy and markets are all self-correcting processes") might be recognised as such

The problem with that metanarrative is that it gets caught up in a circularity of legitimation: it tries to have it that it is *scientifically true* that its account of how science finds things to be true is the correct one

Failing that, the fallback position is indeed one of legitimation through performativity: planes fly, computers compute, ergo science works (but we have sort of given up, at this point, on the criterion of truth, which is what was at issue)

fwiw I do think there are virtuous ratchets which can be set up in systems which subject themselves to criteria of external validation: is this theory empirically defeasible? can this political project pass the test of securing popular consent? does anyone want to buy this thing?

and bad things do tend to happen if a system closes itself off against contestation, stops reality-checking, huffs its own emissions until it passes out

but the actual purpose of any "critical" project is to see around the next corner, to take note of when and how systemic closure sets in: to see how consent is manufactured, how research programs fall to confirmation bias and dodgy bookkeeping, how incentives turn perverse

To take a vivid example: homosexuality was classified as a mental illness in the DSM until 1973. To the extent that psychiatry is or seeks to be a science, the liberal narrative would want to identify this shift as one of those self-corrections: theory adjusting to reality -

but you cannot give a remotely adequate account of how that change occurred without considering the politics of gay liberation, a social movement and body of truth-seeking and theory-making quite apart from the internal discourse of psychiatry. The correction came from "outside".

This is very much, and not coincidentally, the context of Foucault's enquiry into the relationships between different ways in which truth is sought, expressed, contested, made socially cogent or intelligible, set to work in or against institutions, and so on.

Is there a fact of the matter about whether or not homosexuality is a mental illness? Was there a shift in 1973 from stating the facts of the matter incorrectly to stating them correctly? This seems a laughably inapposite way of posing the question.

But the liberal metanarrative about how science works as a self-correcting process doesn't really equip you to pose the question in more appropriate or perspicacious terms - which is part of what "critical theory" in its various forms is there to enable us to do.

postscript: Here's L & P saying postmodernism construes scientific reasoning as a metanarrative

Here's them attributing that position to Lyotard

Lyotard's actual position on "small knowledge"/minor narratives is not exactly one of wistful mourning - he observes that, as a consequence of "incredulity towards metanarratives", people tend to reach for the minor and the local in compensation -

but also remarks that proverbial wisdom and ancient folkways hardly satisfy the criteria of consistency and pertinacity required of a theory of knowledge: "the people's prose...says one thing and its opposite" - "like father, like son" / "to the miserly father, a prodigal son"

At one moment in Lyotard's thinking he does try to articulate a "paganism" of multiple language games without an overarching metalanguage that can render them ultimately commensurable, but it's important to note that this isn't a straightforward relativism:

i.e. it isn't "Western science and traditional magical practices are both equally valid ways of knowing", but more "science, art, politics etc are distinct pursuits with different internal norms and criteria; there's no obvious way to make science just, or justice scientific"

(I'm slightly conflating what Lyotard says about different phrase regimes with what Badiou says about different truth procedures here, because I think the comparison's illuminating)