This short article is to try to put discussion about surveillance into
theoretical framework. It is far from rigorous and is more a guide to
a certain way of thinking about the topic.
The word `sousveillance', coined in the late 90s by Steve Mann in
analogy with `surveillance' was meant to invert the prevailing power
dynamic. The canonical example being the hypothesis that giving
cameras to homeless people would make it less likely that they would
be beaten by police. The `veillance' or watching was in the direction
from less to greater power.
Locally, at any point in space, we can write the veillance equation:
E = V·∇P
This is a compact way of saying that supposing we can assign a `social
power' to every point in space, we can then find, with certain
technical assumptions, the direction and magnitude of greatest change,
∇P. This is the power gradient. It points up the hill from the less
powerful to the more powerful.
Equally, following Mann [1] we can form the `sight' or `veillance
intensity' vector V at that point. This is the direction and intensity
of watching emanating from there. It is the sum of all the gazes that
pass through that point. The · symbol means inner product which is a
way of multiplying vectors such that the result is a single number, a
scalar, that says how well aligned V and ∇P are.
The result, E, is the amount of `veillance' happening at that point
scaled according to the disparity in power at that place. We can
further take the integral over a region in space, ∫Edv, to find out
the aggregate amount, and nature of veillance in that region.
The sign of E is instructive. If it is positive, the total gaze and
the power gradient are aligned. In this situation we have
`sousveillance', the weak watching the powerful. If it is negative,
the direction is reversed, with the powerful watching the weak, which
is `surveillance' as the word is normally used. If it is zero, then
the gaze is among peers with no disparity in power. Making a short
film of a birthday party perhaps. This is `isoveillance'.
The magnitude of E has an interesting interpretation. It can be small
or zero in three circumstances: when the power gradient is zero, when
there is nobody watching, or when watching is only directed among
equals. In other words this might happen in either or both of an
egalitarian society or a society without either sur- or
sous-veillance.
On the other hand if E is some non-zero value, the larger the power
gradient, the smaller the amount of watching necessary to get that
particular value and vice-versa. If E is interpreted as some sort of
measure of the effect on society of veillance, then taken together
with its sign, if there are great power disparities, a small amount of
surveillance has a large negative effect, whereas a small amount of
sousveillance has a large positive effect. In less unequal societies
it takes more veillance activity to achieve the same thing.
This theory of veillance does not take into account the following,
important, phenomenon. Suppose Alex takes a video of Larry at dinner
one night to remember a pleasant evening by. On the surface we could
imagine that this is simply isoveillance. A harmless activity. However
Alex is in the habit of using a server owned by Gerald to store these
video-memories. Gerald is in a position of privilege and power and if
he looks at the video, he is committing surveillance on Larry and
using Alex as an unwitting accomplice.
Similar indirect or hidden surveillance -- implying that E should be a
large negative number -- is possible in a variety of other
circumstances as well. For example even if Alex did not entrust
video-memories to Gerald for safe-keeping, a state could covertly
steal them or force Alex to hand them over.
This indirection, veillance happening through several hops, means that
the in calculating V it is necessary to sum up the indirect gazes as
well. Indeed it is necessary to know all possible paths for
information to pass from Larry to Gerald, together with their
bandwidth, in order to find out the amount of veillance being
committed by Gerald on Larry. This issn't so obvious at first glance.
It is also not obvious that it is well-defined to speak of a `power
field' with a value at every point in space. Certainly it is plausible
that we could associate a number representing some notion of power for
every person, and for every pair of people a difference between these
numbers, but to arrive at something like a gradient we need a notion
of distance between them. Two candidates are physical distance which
has the advantage of being continuous, or distance across a social
graph which would take more work. People, of course are discrete, not
continuous entities, so we might speak of their `sphere of influence'
when adding up their contribution to the `power field'.
Making this rigorous would take some work, but in the continuous limit
we should arrive at something like the veillance
equation. Nevertheless it is helpful to have this framework in mind
when thinking about the social and economic dynamics of information
flow on the Internet and elsewhere.
Edinburgh.
December, 2015
[1] http://www.eyetap.org/docs/Veillametrics_JanzenMann2014.pdf
# distributed via <nettime>: no commercial use without permission
# <nettime> is a moderated mailing list for net criticism,
# collaborative text filtering and cultural politics of the nets
# more info: http://mx.kein.org/mailman/listinfo/nettime-l
# archive: http://www.nettime.org contact: [email protected]
