Skip to content


"In general, what passes for reflexivity in most social sciences is the sheer irrelevancy of questions raised by the analyst about some actors' serious concerns." (Latour 33)


The above from: Latour, Bruno. Reassembling the Social: An Introduction to Actor-Network-Theory. Oxford: Oxford University Press, 2005.

I have just started reading it today, and so far (touch wood), I am finding it refreshingly lucid and useful. Something that is not always a feature of social theory :-)

Update 20/1

Well, I finished the book and am find myself to be somewhat inspired by it; the explanations are mostly clear, and the basic message is simple: don’t invent agencies and overarching ‘frames’ for whatever subjects you are studying – listen to them and follow them where they take you.

Laying the whole mass of actors and connections on a ‘flat’ social world is a useful way of visualising ‘social’ relations: no actor/actant is ‘above’ another, and all potentially influence each other. More importantly perhaps, all that connect are ‘part’ of each other, they have contributed to building the ‘site’ as it is. Thus we are all sites, and understanding how sites are ‘stabilised’ is an important way of effectively describing the situation.

Actor-Network Theory is an essentially descriptive process, and to counter accusations of ‘mere’ description he argues that if something is properly described, in all its minute detail, then no extra explanation (i.e. frame, context, meaning, …) should be required. And he has a point there, though a difficulty is to decide when to stop tracing the connections between different actors, or actants.

In spite of what I said above about lucidity and all that, the last two chapters before the conclusion (“Second Move: Redistributing the Local”, and “Third Move: Connecting Sites”) get pretty convoluted. He comes across sometimes as somewhat structuralist/determinist: he mentions “structuring templates” that circulate, and uses that analogy of ‘plug-ins’ to explain how we cannot do anything (e.g. rationally choosing something in a supermarket) without having these ‘plug-ins’ ‘downloaded’ from other sites. For me this ignores our basic sociality, and ability to negotiate with the world based on some basic instincts (i.e. survival, social, and reproductive). In a way, it’s like he’s being an ‘atomised structuralist’, or ‘powerless structuralist’ – he’s mapping out the social world in a ‘flat’ manner, arguing that all element affect others etc., but just saying there’s no central motor force to it all, which ain’t a bad point.

In the second-to-last chapter, he asks what will be between the lines that connect the sites in the network, and starts to speculate about ‘plasma’ and “vast outside” – all very strange… My feeling is that there doesn’t have to be anything ‘between’ the connections – for the connections are what make our world, mediators affect it, things change, traces are left and actions initiated; but as he said, if it leaves no trace it is not there. I would also say, perhaps tautologically though, that what is visible is visible because it’s part of ‘society’; though he’s arguing that with his method we can discover more, etc.

His conclusion (for anyone looking for a shortcut to reading the whole book ;-)) isn’t actually a summary of the book, but an argument for the political validity of ANT. He argues that ANT is political; proper data collection needs to let subjects have all their agency, and include all relevant entities; secondly, it is political because it has as a task to make the ‘composition’ of the data in a manner accessible to all, and in a manner that loops between the elements and the composer (so to speak).


No Trackbacks


Display comments as Linear | Threaded

Olumide Abimbola on :

I like the book really much... I guess the thing that made the most impression was the issue that a good description would beat any attempt to explain.

The main problem people have had with ANT has been that it grants agency to non-humans. But I think the problem comes because people often take agency for intentionality, and ANT does not deal with intentionality but with actions.....

julian on :

"a good description would beat any attempt to explain" - some people would argue that that's one of the weaknesses, because it's not 'theorised' enough. Which is exactly what he's trying to avoid though :-) I'm not completely sure myself, and in some ways his work is more useful as a guide to a rigorous methodology rather than a theory.

Re. intentionality etc, I think you're probably right. Perhaps another way of looking at it would be passive and active agency?

Olumide Abimbola on :

Hmm... yea, some people would argue that the argument against explaining is one of its weaknesses. Actually, one of my friends says that it is an attempt to escape responsibility for the written text of the research, since one can simply say that they have described, and not attempted to theorise on behalf of the field or the subject or the matter of concern. But I think it is a wrong assumption that description is easy... Just ask me as I deal with smuggling here in West Africa.

Active and passive agency? I very much doubt that Latour would like the idea that agency could be active or passive. One can see him trying to deal with that, for instance, in his chapter on the agency of objects.... Remember the section A list of situations where an object's activity is made easily visible? (pg 79). I think passive and active agency is something to think about.

julian on :

It could be used to avoid taking into account various political/power factors, but on the other hand I think that 'theorising' can sometimes be overdone.

Umm... I don't remember, but I'll check it out. Also relevant would perhaps be 'patiency' (I think), Mark Hobart talks about it.

Add Comment

Enclosing asterisks marks text as bold (*word*), underscore are made via _word_.
Standard emoticons like :-) and ;-) are converted to images.
E-Mail addresses will not be displayed and will only be used for E-Mail notifications.
:'( :-) :-| :-O :-( 8-) :-D :-P ;-) 
BBCode format allowed
Form options