Some things that I found interesting while reading the summary of Kosso's book were:
The idea that we are sort of taught how to observe, theorize, and talk about theories--or that inquiry itself is socially constructed. And that this being the case, we may be a good idea to distance ourselves as much as possible from commonly held views of what scientific inquiry really is--in order to attempt to be more objective.
More on science as socially constructed:
Kosso seems to say that existing theories affect and form our perceptions about what the fundamentals of science are: much like LaTour's assertion that once we move upstream, we have to contend with more and more "blackboxes"--this is congruent with Kosso's thinking regarding how we find ourselves "downstream" and find our choices affected by previous works.
Another thing I found interesting, which I mentioned in class, was the complexity of Kosso's argument as it pertains to self reference. He's theorizing about theories--which I think, opens him up to either self criticism, or self-reinforcement, not quite sure what yet--but I think that since he goes on to talk about credibility of experience (and since that is also socially constructed) he may be able to toot his own horn.
Kosso makes a similar point to Latour's in that he describes theories as being "cohesive", where Latour/Flower ideology is that of a "hang-togetherness"--a term repeated many times in class. The idea that similar or congruent theories will tend to reinforce one another, and because of this, tend to be referenced in the same breath. (See also, Kosso's "theories travel in packs")
On another tangent:
I'm not sure if I am taking this correctly, but I also think that Kosso may have been proposing that while it is good for a theory to be falsifiable, a single falsification does not a theory-buster make. Exact conditions cannot always be replicated, such as the situation we saw in LaTour's writing with the guinea-pig gut.
Of note, also, was the idea of the nature of observability--and the thought that as our ability to "SEE" things for ourselves with our actual eyes, or 5 senses decreases, the amount that we rely on "instrumentation" increases, and that this, should in turn foster questions about the reliability or accuracy of the tools we use in lieu of eyes or ears etc.