Quoting Ken :
"In general, I have been sceptical toward ANT for the reasons that
bother Klaus -- the apparent attribution of agency to objects and
artefacts.
There are other theoretical issues that have bothered me about it as
well. I recall reading Latour's description of a kitchen tool in
which he seemed to describe the tool not merely as an "actor" but
rather using verbs that implied that the "actor" was an "agent."
If this is the case, it raises problems for design process, design
thinking, and design theory -- I know that Klaus finds the term
"theory" troublesome."
I think that Jeremy has provided a lot of good insights on ANT. I am
not a specialist of ANT myself, but a reasonable reader of Latour,
Callon etc. in their own language.
I am not surprised at the slight irritation that anglo-saxon
pragmatists tend to show towards the notion of agency (and the
connection to discourse, to which I shall come back). Agency is
neither intentionality (which could be attributed to humans
only...hum... though I think that Don's post is raising a substantial
point), nor performance (I mean here : the transformation that occurs
after an artefact has been used...hum... but Austin could also says
something about that !). Agency is the actualisation of a potential.
To use Klaus example about the rifle : the rifle is not an agent (and
I think that you would not find Latour claiming this at all).
Artefacts carry agency when they contribute to the semiosis of the
situation. The rifle, out of its holster, at the end of the arm of a
policeman might mean for me that I am protected (and it "represents"
law), while at the end of the arm of a villain, it might mean danger
(for me). The rifle hanging over the chimney might mean that I am a
collector and would allow me to meet other collectors.
Therefore, agency is not independent from semiotics (in the large
sense): it is first and foremost differential and cannot be reduced
to an "essentialist" perspective (that is why it is "so easy" to turn
it into something not rigourous). But I think that it has not been
designed to make an ontology of the world, but rather —as Jeremy
stresses rightfully— as a way to identify layers of concepts that
create, at a given moment in time, a coherent world (or, as
essentialists would say : a world that appears coherent; one simple
measure of it being a "private language").
The (most) important thing is that the coherence is strongly
associated to the power layer that is embedded in discourses. And
here, I do include artefacts as contributing to the formation of
discourse. We all know what "powershow" means : and it is not the
president of the university in short boxing each protagonist. Agency
does not care whether you have been threatened (verbally,
physically...), or whether you are shy, feel inferior, don't care...
agency will notice that one person always applies the interpretation
of another.
Whether this is important to design, I would say yes... absolutely.
If design was only about doing "things", that would be easy. But
design is also creating, negotiating, prioritising properties of a
potential before turning this into an "artefact". And I am afraid
that, in my view, this has not been much covered. Yet, you just have
to sit in meetings (in the industry) to clearly see that discourse
formation is an essential asset for having a successful project.
Finally, I agree with Jeremy: use ANT (or in fact any other theory)
if it helps you understanding things better and fits your purpose.
But, like many other "french theories", use them carefully —go back
to the original. Too many "intellectuals" think that discourse
analysis and deconstruction is an easy playground for general
relativism, write incompetent papers and don't stand behind what they
say. That has never ever been the underlying project of the authors.
Jean
|