Objects, Networks and Sociality
I’ve been catching up on some reading lately, especially around activity theory and much of what I was reading seemed to resonate with some of my other interests, especially that of Actor Network Theory. What initially got me thinking was a really interesting post
from Jyri about about scalable and effective social software. His argument is basically that it is object
orientated systems that work – like flickr – where there are ‘things’ to cohere
around. Without objects it is harder to orient to anything as LinkedIn have
found [though they are now moving toward ‘jobs’, services and job ‘uri’]. The focus on objects is noting new – look at
object centred database models etc. but explicitly acknowledging their
centrality in “social” systems is, if not original, then very welcome. Actor-Network Theory [or SST or post-ANT] is a collection of ideas that posits the importance of ‘things’, be they human
or non-human, in their coming together, in their effects. This ‘heterogeneous network’ of humans and non-humans is agnostic toward the materiality of what is being studied, rather it looks at how things are facilitated and engendered. The wikipedia entry for ANT uses the example of the supermarket to elucidate on this approach – where the cashier is only as important as the cash register in processing the good being bought, and these ‘things’ in turn depend upon other machinations some of which are fixed [numerical system] and some of which are more fluid [language]. This ‘machinic arrangement’ does away with traditional sociological positioning of subject – object. As Anne Galloway states, the “object is
already in the subject and vice versa”.
However in practice there is often little
difference with existing methodologies of ‘thick description’, where you
describe ‘practices’ and trace patterns of activity. And this is where I think there are interesting parallels with activity theory [AT]. AT focuses on
process and practice rather than the user centred design [UCD] approach of
objectifying people within certain segments, tribes, or typologies and applying traits or characteristics to the design process. This could be the design of social software
systems or interface design, whatever. The latter tends to be a snapshot, an objectified moment, whereas the former can be seen as a dynamic descriptive process not unlike that used in ANT*. So what is the reticence around using activity theory and seeing object-centred sociality? I think in part it is to do with our ideas of ‘reality’.
Baudrillard and other post-modern writers argued that we were moving away from
‘real’, authentic experience toward ‘hyperreality’, Bruno Latour, the most
eloquent and nuanced of all SST writers [and, in my opinion one of the greatest thinkers of the last thirty years], described the same move toward a more
technological society as actually more ‘real’, not less. I like this. His explanation was that in the modern [rather than the post-modern]
world it is actually harder to create strong, effective networks – and the effort to create associations among different elements makes for more reality. Whilst
Baudrillard wrote that “The Gulf War Did Not Happen”, that it was hyperreal
‘simulacra‘ – that we experienced as the Gulf War was not what actually happened –
Latour would argue that it was a reality, and just as ‘real’ and ‘authentic’ an
experience exactly because of the huge machinations that had to be
created and put in place to make it happen for us, the viewers and consumers at
‘home’. Rather than there being an authentic experience ‘beneath’ the one we experienced through the media, Latour’s reality operates on the same epistemological plane. As a result the ‘subject’ becomes just another actor in a drama with no pre-defined role and that, fundamentally, is difficult for people to grapple with not only because ‘Humanism’ has been a central plank of Western though for centuries but on a more practical level because we see the world ‘subjectively’, through our own experiences, and consequently believe in our ability to effect change.
If you apply ANT
to developing scalable ongoing experiences [I would call them social software
but they’re not just social now are they?], it puts the focus back into relationships rather than ‘things’. Looking
at what makes relationships ‘tick’ and re-evaluating the importance of objects
[heterogeneous networks of humans and non-humans] in creating ‘society’. This reinforces Jyri’s point about object oriented systems as crucial to building sociality in the digital world. It also serves as a useful counterpoint to popular perceptions of contemporary society, as seen in phenomena like internet dating, as somehow ‘less real’ than ‘traditional’ ways of doing things, ie where the subject was made the all seeing, all doing action-hero and the objects merely invisible actors.
* But as
seductive as this move to process, activity, practices and ‘heterogeneous networks’
is, it is not necessarily and inherently any ‘better’ than other
approaches. Indeed, the ‘science’ behind
the approach is itself a result of objectification and network effects [in
other words the theory has within itself the seeds of its own deconstruction]. Ultimately, you use whatever ‘works’ and by
constantly trying different methods, techniques and approaches you find
something that, for a while will ‘gel’. And by describing the ‘network’ and looking for strengths and weaknesses other
people can look to learn from it.