Uncritical acceptance of social-epistemological categories shouldn't fly in political philosophy of science. The undifferentiated "scientist" as technical expert might be good enough for epistemologists, but separating economic positions among scientists is necessary for political analysis

From H. H. Gerth and C. Wright Mills (1942), "A Marx for the Managers".

technicians may be found on all political sides of many social fences. The technical knowledge of managers ... is one thing; their class position, political loyalties, and their stake in the current system is quite another. There is no intrinsic connection between the two."

of trained skill may be a production engineer with a fixed salary and fixed stages in his career within an organization. The possession of a skill may well mean quite heterogeneous interests, class positions, and political loyalties. In a democracy, apart from common technical knowledge, ...

(Btw if you want more details on the empirical and theoretical analyses that Latour and Woolgar conducted, I really would recommend reading the book--it's quite good!)

Perhaps you think knowledge based on models of of human social activity doesn't really count. You'd have to take a lot of other stuff down with you for that to be coherent. Or, if it's the modelling practice in particular you have issue with, lots of natural-scientific knowledge might have to go too

generations of science studies research, which now have an arsenal of anthropological techniques and paradigms for investigating science as a material phenomenon. Does a new discovery about Bach's childhood need to change the way science is done to count as a contribution to knowledge?

social negotiation, etc.). It gives us a better model of how science works that other accounts, which entirely ignore such dynamics.

To address your second Q: why would something need to change the way natural science is done in order to contribute to human knowledge? It's certainly shaped several

The answer to your first Q is already given in my response! It contributed to human knowledge about something that happens in the world – scientific research – by giving an improved model of the dynamics under which that phenomenon operates (e.g., that it operates in part according to processes of

instruments, institutional relationships, and strategic decisions. If you think this is all just sociological mumbo jumbo, perhaps you had already committed wholeheartedly to your beliefs before asking the question--fair enough but not a particularly scientific attitude

object *of* a science), rather than mystifying it as an abstract, almost magical process driven by the innate genius of scientists. Latour and Woolgar’s ethnographic approach, one of the most influential texts in social studies of science, shows that scientific knowledge is the product of labour,

even though its physiological presence in the body remained "unproven". To drive the wider point home: this kind of sociological analysis invites us to investigate science materially, as a real, situated phenomenon governed by identifiable and analysable dynamics (i.e., as something that can be the

through a complex sequence of experimental work, rhetorical negotiation, and inter-actor validation. They demonstrate how a statement like "TRF is Pyro-Glu-His-Pro-NH2" sheds its historical and social context to become a "fact" – a stable, uncontroversial reference point for other researchers,

If I must! Based on their anthropological studies of two research labs operating in the 1960s, Latour and Woolgar (1979) contributed to our knowledge of scientific research processes by showing that scientific facts, such as the structure of TRF(H), are not simply "discovered", but are constructed

I can post you the certificate for my honours degree in natural sciences if you want, it's useless to me since I progressed to the nobler art of historical analysis anyway

I have a limited use link to a non-paywalled version of the article for those who want access (first come, first serve)

will DM you!