I picked this up because folks over on the Philosophy in a Time of Software kicked things off by discussing this book by Latour. So, I'm really not terribly knowledgeable about sociology, but I did a fair bit of reading in the social sciences while
getting my library union card studying library/information science. So I wasn't completely underwater, but I definitely felt like I was swimming in the deep end. I didn't get the connection to computer programming until quite late in the book, but it was definitely a bit of a lightbulb moment when I did. Latour's style (at least that of the unmentioned translator) is refreshingly direct, personal, and unabashedly opinionated. He spends much of the book describing just how complicated social science is, and how far it has gone off the tracks...which is quite entertaining at times.
A few things I will take with me from this book and its portrayal of Actor Network Theory:
I will never be able to say or write the word "social" without feeling like I'm glossing over a whole lot of stuff, and that this stuff is what I should actually be researching, talking and writing about. Latour stresses that it's important not to dumb things down by appealing to established social forces (class, gender, imperialism, etc) but by tracing the actors, their controversies, and their relations. This work requires discipline because it's tempting to reduce the complexity by using these familiar abstractions instead of expending energy/effort in documenting the scenarios as faithfully as possible. By letting the actors have a voice, and say what they think they are doing, rather than the researcher telling the actor what they are actually doing. I work in libraries/archives, so I particularly liked Latour's insistence on the importance notebooks, writing, and documentation:
The best way to proceed at this point ... is simply to keep track of all our moves, even those that deal with the very production of the account. This is neither for the sake of epistemic reflexivity nor for some narcissist indulgence into one’s own work, but because from now on everything is data: everything from the first telephone call to a prospective interviewee, the first appointment with the advisor, the first corrections made by a client on a grant proposal, the first launching of a search engine, the first list of boxes to tick in a questionnaire. In keeping with the logic of our interest in textual reports and accounting, it might be useful to list the different notebooks one should keep—manual or digital, it no longer matters much. p. 286.
... and that this is the work of "slowciology" -- it requires you to slow down, and really describe/dig into things.
The other really interesting thing about this book for me was the insistence that social actors do not need to be human. It is fairly typical for social science research to focus on face-to-face interaction between people as the primary focus. Latour doesn't dispute the importance of studying human actors, but emphasizes that it's useful to increase the number of actors under study by studying objects (mediators) as actors. Typically we think of actors as having agency, free will, etc ... but objects are typically complex things, with particular affordances, and extensive relations with other things in the field. You get only a very limited view of what is going on if you don't trace these relations.
Things, quasi-objects, and attachments are the real center of the social world, not the agent, person, member, or participant—nor is it society or its avatars. (p. 237)
As a software developer, I really identified with Latour's insistence on the role that objects play in our understanding of activities around us; how this view necessarily complicates things a great deal, and requires us to slow down to really understand/describe what is going on. It is hard work. And it's only when we understand the various actors and their relations, the actual ones, not the abstract ones in the architecture diagram, or in the theory about the software, that we will be in a position to effectively change things or build anew.