Jordan Crandall on Mon, 21 Jul 2008 04:51:17 +0200 (CEST)

[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]

<nettime> Between Tracking and Formulating

Between Tracking and Formulating
Jordan Crandall
Text of presentation given at the conference "The Everyday Life of
Surveillance Studies Network Seminar Series:  Prediction, Anticipation,
Department of Geography, Durham University, UK, 18 July 2008
Stephen Graham, conference chair

1. The Apparatus of Tracking (as it takes an anticipatory or predictive

We are in a culture that everywhere advances toward real time.  We aim to
close the gap between detection and engagement, or desire and its
attainment.  We combine human and machinic functions through ever-faster
computational systems, in order to generate a realtime perceptual agency,
a concert of forces.

And yet, in an increasingly competitive consumer-security culture, this
realtime perception  is too slow.  We want to gain advantage in our
competitive theaters -- the battlefield, the marketplace, or the social
arena -- and in this competitive, accelerating culture of shrinking space
and time intervals, we want to reach beyond the immediacy of the present. 
We want an increased capacity to see the future.

As this realtime, combinatory, networked seeing becomes anticipatory --
oriented toward the future -- then vision itself changes:  we are no
longer "seeing" but "tracking."

Tracking leapfrogs the expanding present.  It offers up a predictive
knowledge-power: a competitive edge.  It promises to equip us with new
abilities:  the skill of outmaneuvering our foes, intercepting our objects
of suspicion and desire.  (Desire -- because we also track the various
objects of our libidinous interest -- products, markets, sexual partners.)

This anticipatory, predictive orientation has taken root within the
cartographic tradition.  It has traditionally relied upon observational
expertise.  Think of the vigilant scanning of radar screens by human
operators during WWII.  The vigilant observer, harnessed to the screen,
watching movements, extrapolating patterns.  The observational expert,
interpreting movements on schematic maps.

Today however tracking relies, more and more, on data analysis.  This
shifts the role of this vigilant observer.  Statistical procedures inform
tracking, as tracking informs the conditions of statistical analysis.

Even though we can isolate tracking as an activity, what we really need to
talk about is something larger:  a tracking apparatus -- a technical,
discursive, institutional formation, at once an activity, a form, and a
mindset.  An apparatus that takes a predictive orientation --
understanding an object's propensity, the propensity of a situation.

Let's think about the tracking apparatus in terms of its "strata".   We
can think of 4 layers.
1. Tracked phenomena, in the form of datasets
2. Statistical procedures (analytics)
3. A pattern is discovered, which suggests a continuity, and a propensity.
(--you use more analytics to refine the pattern: you use random back-tests
to test the pattern's accuracy, stability.  you test its ability to
forecast-- then you arrive at:)
4. A model, a formula (algorithmic).  It is stable, but you can
continually modify it.  You can always can introduce new factors.

Increasingly, the tracking apparatus is able to reach far back into the
past, further back than was humanly possible, through the use of
regressions.  Regressions are statistical procedures that take raw
historical data and estimate how various causal factors influence a single
variable of interest (for example, the quality of wine, or an enemy's
movement).  A pattern is revealed, derived from the past, and this
demonstrates a likelihood, a propensity, for what could happen today.

This pattern might be stabilized, made operational, in a formula.  You
just plug in the specified attributes into a regression formula, and out
comes your prediction.  A moving phenomenon -- a stock price, a biological
function, an enemy, a product or part -- is codified and understood in a
historical trajectory, in order to extrapolate its subsequent position. 
This formula is not a fixed thing but something subject to continual
modification.  New factors can be introduced.  It might be stable at one
point, unstable the next.  It is  modulated as it is modulating.

Consider the use of personal data-mining in Google.  Its web accelerator,
for example, continually pre-picks web pages for you, predicting what you
are going to want to view next, based on your tracked personal history,
and holds these pages in wait for you.  Even before you fire up your
browser, Google predicts what sites you are going to look at.

Google's PageRank system uses a form of data-mining called social network
analysis.  One might think of this, along with collaborative filtering, as
a way to harness the "wisdom of crowds."  As, also, with the prediction
engines of, which know the books that I am likely next to buy.

Through a technologically--enhanced seeing, a mathematical seeing,
relationships are uncovered among widely disparate kinds of information --
patterns that could not be previously "seen" by the naked eye.  Such
data-mining has only increased with the rise of computing capacity
--processing capacity, but even more important, storage.  In the so-called
"Petabyte" age, data-mining becomes increasingly precise.  I does not
necessarily render observational expertise obsolete, but it does change
its role, its orientation.  Likewise with the so-called "scientific
method" itself.

Chris Anderson, in his recent article for Wired [16.07], put it this way. 
In the new Petabyte age, "The new availability of huge amounts of data,
along with the statistical tools to crunch these numbers, offers a whole
new way of understanding the world? science can [now] advance even without
coherent models, unified theories, or really any mechanistic explanation
at all."  We can stop looking for causal models -- correlation is enough.
"We can analyze the data without hypotheses about what it might show.  We
can throw the numbers into the biggest computing clusters the world has
ever seen and let statistical algorithms find patterns where science

"This is a world where massive amounts of data and applied mathematics
replace every other tool that might be brought to bear."  In this new
world, according to Anderson, theories of human behavior, from linguistics
to sociology, are no longer necessary.  Out with taxonomy, ontology, and
psychology -- all of us!

"Who knows why people do what they do? The point is they do it, and we can
track and measure it with unprecedented fidelity.  With enough data, the
numbers speak for themselves."  This is, according to Anderson, the "new
laboratory of the human condition."  For him it supercedes everything that
has come before.

This dream of a universe of abundant data, increasingly precise in its
predictions, instantaneously transmitted and shared -- this has always
been tracking's utopia.  Think of the ultimate panoptic dream of the
military:  a wireless, unified computing grid that can link weapons,
systems, and personnel in real time, making volumes of information
instantly available to all military and intelligence actors.  A system
that allows, in the military's words, each actor to have a "God's eye
view? of the battlefield.  A system through which the military predicts
that it will be capable of "finding, tracking, and targeting virtually in
real time any significant element moving on the face of the earth." 
Tracking as the ultimate panoptic ideal, propelled by a sense of divine
right, could not be more explicitly stated.

But now let's also connect back to our understanding of tracking as a
visual activity that is situated within the cartographic tradition.  From
the observational expert at the monitor, analyzing tracked spatial data,
to mobile GIS and location-aware technologies:  the user at the handheld
device, accessing traffic information that help to determine the best path
to take in a strange city.  Here tracking is conducted through
sophisticated graphic information systems that are formatted according to
geographic paradigms, oriented for the humans who must interpret it and
transform it into actionable intelligence.  It is an emerging landscape
where actors are tagged with geospatial coordinates: a world of
information overlays that is wedded to objects and physical sites. A world
where communication is tagged with position.

And yet, even as it endures, this cartographic tradition has become simply
one kind of interface to data (albeit an important and enduring one, as we
are, at least for now, spatial creatures).  Maps have not gone away; they
have simply been transformed to interfaces.

Think of tracking technologies like RFID.  It can keep track of the exact
item you buy, how long you hold onto it, how far you transport it, what
other products you buy in conjunction with it, etc.   While it is possible
to map its tracked objects in space, such spatialization is not primary. 
The map is secondary;  the numbers are what speak.

2. The Generativity of Tracking: Materializing Inclinations

Again, we are talking about a tracking apparatus, as this takes a
predictive orientation -- or rather, an understanding of an object's
propensity, the propensity of a situation.

Of course, privacy issues haunt tracking.  Here is a curious twist.  Ian
Ayres points out in his book Super Crunchers that "Traditionally, the
right to privacy has been about preserving past and present information. 
There was no need to worry about keeping future information private? Yet
data-mining predictions raise just this concern? [It] puts future privacy
at risk because it can probabilistically predict what we will do."  For
Ayres, this new world of data-mined predictions "moves us toward a kind of
statistical predeterminism."

But we do not only need to understand it in this way, to glimpse the
generativity of tracking.  Because we willingly step into tracks that are

Here is a poetic way to see it. There exists a probable construct -- a
kind of ideal scenario -- that stands in relation to reality as its
tendency. It configures as a statistical inclination.   It hovers like an
ideal form awaiting a reality that will fill it.  It becomes a silhouette
that models future positions, a ghostly forebear into which reality flows.

Something like that book that will likely appear me on,
recommended through its prediction engine.  A bit of a world that stands
in wait for me, beckoning me.  It subtly shapes my foray, one click away.

When we track -- when we study how something or someone is moving in order
to predict its future location or orientation -- we subject everything to
the classifying schemes available to us.  We fasten our objects (and
subjects) onto a classifying grid or database-driven identity assessment. 
But if we follow Heidegger, and then Foucault and Kittler, the
technically-driven classifying schemes are there right from the start,
informing the conditions of our approach.   What we see is defined within
the discursive paradigm of such technologized seeing.  Subsequently, we
begin to see ourselves in these terms.  We internalize the classification

This mechanism of anticipating the future then is somewhat of a
self-reinforcing one.  Reality is molded to fit the demands of tracking --
in part by establishing the terms through which we understand and engage
it.  Paul Edwards, tracing the history of computing in the military, would
describe it this way:  As computers form the basis for strategic thought,
the world is modeled as a formal machine, subject to its determining
logics. From mid-century onward, the systematic, logical rules of
computing helped produce the sense that everything -- ground realities,
warfare, markets -- could be formalized, modeled, and managed.  Reality
was figured as mathematical and ?capturable? through a formal programming
logic.  The world became predictable, pliable; the future controllable. We
anticipate the future as if we could shape it. This is what Edwards calls
the "closed world" mindset."

There is an emphasis on data patterns over essence: an ever-greater
abstraction of persons, bodies, and things, and an emphasis on statistical
patterns of behavior, where the populace is pictured as a calculus of
probability distributions and manageable functions. Quantitative analysis
impacts real-life decisions, as "real life" becomes an analytical
construct.  Life is remade around the capacity of the database.

3. The Limits of Tracking, at the Macro and the Micro

But now let's talk about the limits of tracking as such.  One is at the
macro, the institutional dimensions of the tracking apparatus, and the
other is at the micro, its affective dimension.  So let us zoom out, and
zoom in, and in both cases, touch the limits.

First, the macro.  Continuing to follow Paul Edwards we can see that
tracking is part of a closed-world mindset that is fast reaching its
limits. In Edwards' terms, we are seeing a shift from Cold War grand
systems theories and their closed world orientations, to network models,
which move us "from hierarchy to intricate mesh, from topography to
topology, from closed to fractally open, from determinism to chaos and
complexity."  A new modality of power: the decline of the closed
disciplinary institution and the rise of distributed organizations -- the
shift from strong discipline to weak discipline.  Where the strong
discipline of carceral institutions relies on the regular exercise of
intensive, direct encounters between authorities and individual subjects
(the confessional, the examinations, the boot camp, etc.), in weak
discipline, the balance between the two has shifted -- the productive
power of subjects has increased, as new forms of networked agency arise
without carcerality.

We could also see this, following Agamben, as the rise of an open-ended
culture of security -- in contrast to the disciplinary and control
societies to which tracking more readily belongs.  While discipline
isolates and closes off territories, he writes, security leads to an
opening and to globalization. While the former wants to prevent and
prescribe, the latter wants to intervene and direct ongoing processes.
While the former wants to produce order, the latter wants to guide
disorder. Following from this line of reasoning, this new apparatus of
security is not, as many would have it, "preventive." It is not preventive
since it "can only function within a context of freedom of traffic, trade,
and individual initiative."

In this context, we could say that the new security ecologies compel a
kind of "open-sourcing" of the apparatus of tracking, on the one hand, and
its further ability to be operationalized, on the other.  This would mean
that it "opens out" onto new ecologies, new assemblages of network-enabled
actors, in a chaotic world which is no longer predictable in the linear,
closed-world sense.

There are those who would disagree with this, by simply saying: the data
just needs to get increasingly granular, the tools for analyzing it

Analytics are only as good as the datasets they rely upon.  And datasets
are only as good as they can become increasingly granular.  No doubt, they
are increasingly precise.  But the question arises:  can data capture

Now, the micro.  If tracking moves toward instantaneity -- eliminating
time and space intervals and connecting and conjoining multiple actors --
then in the extreme case, as Paul Virilio would have it, this realtime
arena is one in which "coincidence" takes the place of communication, and
the emphasis shifts from the "standardization of public opinion" to the
"synchronization of public emotion."  In a real time world where there is
less and less time to act, or where action plays out in barely-measurable
fractions of seconds, there is also less time for thought.  There is no
time for interpretive activity: there is a zooming into the body, into the
pre-interpretative level.  A turning away from exterior movements and
positions and instead a turning inward, toward "interior" states.  We
might call these dispositions -- disposition to act  that accumulate just
at the horizon of the visible.  We have here moved from position to

Dispositions are affectively-grounded behavioral tendencies, attitudes,
moods, or general temperaments that adhere over time.  We are talking
about bodily intensities, in addition to linguistic mediation; affect, in
addition to representation. This is a domain that is occupied with
intangible, corporeal dynamics of movement, sensation, and rhythm, rather
than solely with calculi of symbolic positioning. It is unpredictable,
does not resolve to meanings.

We have here been understanding the tracking apparatus as having a
predictive orientation, or rather, a way of seeing the world in terms of
propensities.  Yet affects and desires, as they might adhere in
dispositions, are not, of course, predictable.   Affects are
pre-individualized, ambiguous sensations that do not necessarily resolve
to emotions.

Of course, one can say that the apparatus of tracking, as we are defining
it here, can generate affects and desires -- produce dispositions.  As,
again, in the prediction engines of, which seem to know what we
want before we know what we want, and therefore help to create a want. 
Tracking derived predictions, put into play as arrays of offerings, can
generate desires and emotions.  We can see tracking as part of a
desire-producing, affect-triggering mechanism.  In the commercial realm
this has always been the case: viewership or readership patterns are
tracked in order to generate more effective advertising -- advertising
that can more effectively push one's buttons.

But again, in a world that moves toward instantaneity, it is already too
late.  And as Leo Bersani would say, "knowledge misses being."  Tracking
is, at its core, semiotically-based, understanding both movement and
position as language. It cannot harness the realm of affect.

One might say that it deals with time (a divisible mobility) but not
duration (an indivisible mobility).  Disposition stands in relation to
position as duration stands in relation to time.

Here an ontological threshold is approached:  how to engage these
pre-individual, pre-symbolic affective capacities?  On the one hand, the
tracking apparatus does attempt to reach deeper into the body through new
micro-technologies and techniques of physiological measurement.
Technologies of bioanalysis are probing more deeply into the intimate,
micro-states of the body.

These technologies have revealed, for example, that a particular action is
already set in motion by the body about 0.8 seconds before we consciously
experience the performance of it. The body readies itself for action
before it has a conscious experience of the action. Compared to our sense
perception, our thought processes are too slow, and so, especially when it
comes to quick events, nature has routed around them: the parts of the
brain that activate movements are linked directly to the centers for sense
perception. In working with the phenomenon of "readiness potential,"
Benjamin Libet, an American neurologist, long ago showed that
consciousness lags hopelessly behind action: thought follows action,
however we do not consciously experience it in this order.

Nigel Thrift says it well: what we experience as the immediate presentness
of the body is, in a sense, already past.  Thrift says that we can expand
the time-space of embodiment accordingly, such that it incorporates a
"constantly moving preconscious frontier." To incorporate this
preconscious frontier in our understanding of embodiment is to widen the
durational expanse of the present moment, opening up a space between
affect and contemplation.

And yet, even as the apparatus of tracking is increasingly able to delve
into them, these new forms of pre-conscious bodily-knowingness counter
determining logics, close-world mindsets.  (There are those like Chris
Anderson who would disagree, of course, saying that the technology just
needs to get better, more granular, its processing and storage capacity

4. What has become of Tracking?  The need to look at the institutional

If we say, then, that the limits of the tracking apparatus are reached
both in terms of the micro and the macro, then what can we suggest is
developing?  At the micro level, we have a "turning inward" of tracking
(into the realm of intension rather than only extension) -- a "turning
inward" of predictive orientations, which are no longer solely aimed at
the external world, but at the body substrate itself.  At the macro level,
we have its "opening out" onto new ecologies, new assemblages of
network-enabled actors, in a chaotic world which is no longer mappable or
"predictable" in the linear, closed-world sense.

Again, we are talking about the tracking apparatus as a technical,
discursive, and institutional formation -- at once an activity, a form,
and a mindset.  We are talking about this apparatus as something that
takes a predictive orientation -- an understanding of an object's
propensity, the propensity of a situation.  So far, in outlining the
different "strata" of the tracking apparatus, we have focused on the
technical and the discursive:

Here, again, the different "strata" of the tracking apparatus:
1. Tracked phenomena, in the form of datasets
2. Statistical procedures (analytics)
3. A pattern is discovered, which suggests a continuity, and a propensity.
4. A model, a formula (algorithmic).  It is stable, but you can
continually modify it.  You can always can introduce new factors.

What is called for at this point, in order to understand what the tracking
apparatus has become, is a more careful look at the institutional paradigm
within which it exists.  We can't understand this new form of tracking
unless we can understand the changed institutional landscape within which
it exists, and consequently, the new modality of power to which it
belongs.  So let us now focus on the institutional strata.

We have already talked about this in terms of open-world (security), in
contrast to closed-world (control) -- a form that involves weak
discipline, in contrast to strong discipline; non-carcerality, in contrast
to carcerality.  The institutional paradigm we need to look at is one that
traffics between these various levels of embodiment, from the grand to the

5. The shift from Institutions to Organizations

In contrast to the institutional paradigm, we can call it the
organizational paradigm.  Let's see what happens to tracking when we
locate it within this new paradigm.

As a quick way of visualizing the difference, let's consider the
difference between the iPhone and the gPhone.

The iPhone is a self-contained, proprietary device, encased in a beautiful
shell, which can only to be used, in the US, with one carrier: AT&T. 
There has been a lot of speculation about what the Google phone will look
like.  Will it be even sexier?

But-- viola! -- The Google Phone is, at its core, not really a phone. 
It's an open-source operating system called "Android" that can be loaded
into any mobile device. Any phone can become a gPhone, simply by loading
the Android DNA.  It is not a device but a means, a seed.  Unlike
Microsoft's or Apple's OS, Google's software is fully customizable, built
on Linux.  People who write for current mobile OSs have to obtain security
keys and cross gateways -- not so here: programming for this phone is like
programming for the Web, building your own customized blog or social
networking site.  Any application can be installed and run, and all
devices are compatible -- they can share information.

Right from the start, Google wanted to harness the creativity and input of
users.  It  announced the "Developers Challenge," which offered prize
money to users who could develop the best applications for the new system.
 1788 entries were received from around the world, and 50 winners were
recently announced.

Now I want to dispel any easy sense that tracking has disappeared, or has
been replaced with something else.  Here is an important component:  All
phones will be locatable through GPS or the cross-referencing of cell
towers, generating an unprecedented degree of locational specificity. Of
course Google will track its users, in order to understand their habits
and sell highly individualized, targeted timespace to advertisers. So this
is a higher-order tracking:  a significant extension of it, however
couched, as it is, in a participatory ecology.  But with a constantly
evolving cosmos of applications, how do you understand what people are
doing, or will be doing?

Two areas emerge here.  First, Google might not know exactly what they're
doing, but, because the phones are locationally-aware, it will know where
they are -- where they are currently and where they are going.  Second,
there is the fact of user's investment in developing their own
applications, and making the device their own, integrating it into their
life in a highly intimate and personal way, and sharing material with
others in an unprecedented way.   One could understand Android in
bi-directional terms:  it learns from its users, is adapted by them and
reincorporated in a constantly evolving cosmos.

There is an intimacy here: a kind of affective formulation that gets
closer to the body, closer to home, and it learns by way of practice.  If
tracking imposes on you, this affective formulation is something that
moves in with you.  It interweaves itself into your habits.  What is
involved is not a "top down" approach but an emergent one: not an imposed
presence, but a form of emergent presencing.

But, again, we have to dispel any notion that tracking has disappeared. 
There is still surveillance here.  But rather than simply a surveilling,
there is also a seeding.  We are talking about something that constitutes
a higher-level order of tracking, a kind of extension of the apparatus of
tracking -- a more generative arm of the apparatus.  An extension of its
predictive orientation.  What is "beyond" prediction, or an understanding
of the world in terms of propensities?  Something that involves emergent
modes of bodily knowingness, which operate at the level of the affective
and the pre-individual.

Of course we are not really talking about two separate realms -- the
institution, the organization -- but of a historical development where the
latter builds on the former.  Let's follow Chris Anderson again for a
simple progression.

"Kilobytes were stored on floppy disks. Megabytes were stored on hard
disks. Terabytes were stored in disk arrays.  Petabytes are stored in the
cloud. As we moved along that progression, we went from the folder analogy
to the file cabinet analogy to the library analogy to ? well, at petabytes
we ran out of organizational analogies."  We're out of analogies!  (We may
have plenty of storage, but we no longer have analogies!) There is no time
for analogies!  So we'll just call it the organization!

Again, following his argument -- (sorry, I should at least be summoning
Kittler here?) -- the organization brings the death of the scientific
method. "The new availability of huge amounts of data, along with the
statistical tools to crunch these numbers, offers a whole new way of
understanding the world. Correlation supersedes causation, and science can
advance even without coherent models, unified theories, or really any
mechanistic explanation at all."  But his argument privileges
statisticalization above all.  And knowledge misses being.

6. The Improvisational Organization

So let's switch to Nigel Thrift, who has a different way of seeing the
organization.  For him, staticalization is augmented with improvisation. 
"From flip charts to divisional structures to inventories to commercial
statistics to various software packages," particular "modes of knowing"
are produced "which are acted out in various ways."  In opposition to
formalist approaches, organizations are composites.  What holds them
together and pushes them on is "performance" -- "an ability to act
convincingly into the situation that presents itself by taking whatever
propensity for dynamism may be offered that is also a practical ethic of
discovery and invention.

Organizations are rarely made up of practices that are so mechanical that
they simply reproduce themselves.  Usually, they consist of a set of root
practices which can very often go wrong or, at the very least, require
radical adjustment to keep the same.  In these circumstances,
improvisation is often called for, improvisation which sometimes produces
solutions that become the base of new practices.

This process of almost continuous improvisation is forced by the exact
configuration of forces that presents itself to actors at any point in
time which in turn requires a more or less skilled response to the
arrangement of things, a sense of the propensity of the situation that the
Chinese call 'shi,' the potential born out of disposition."

"A critical element of 'shi' is space. For much of what counts as
configuration is exactly that: a continuous re-arrangement of things in
response to events.  So what counts as shi requires all manner of spatial
operations:  linking, contrast, separation, combination, tension,
movement, alternation, oscillation worked out in a series of different
registers: bodily movement as exemplified by gesture, the different
combinations of sound, touch, vision, and smell that typify a situation,
the lie of the hand which pushes back on the body in all kinds of ways,
for example, through balance, through the tools that are at hand, and so
on."  [Thrift, "Space", TCS]

This involves a way of seeing technology -- tools -- in terms of
human/object assemblages, couplings, contourings, which also serve as
sensory and proprioceptive conduits.  It is a far cry from the
organizational landscape outlined by Chris Anderson:  a world where
"massive amounts of data and applied mathematics replace every other tool
that might be brought to bear."

7. A model for the Improvisational Organization:  the Assemblage

Now there are many ways of course to look at this new paradigm of the
"improvisional" organization, which has extended that of the institution.
The improvisational organization -- whose actors engage in a configuration
of forces that "requires a more or less skilled response to the
arrangement of things, a sense of the propensity of the situation that the
Chinese call 'shi,' the potential born out of disposition."  The model I
have found extremely useful is that of the assemblage.

My own model is based on the one outlined by Manuel DeLanda, who has
brilliantly expanded Deleuze's original work.  And yet I am incorporating
other references -- for example Brian Massumi's theories of affect; the
actor-network orientations of Vilem Flusser, Matthew Fuller, and Bruno
Latour; Lacan's notion of the "sinthome" by way of Zizek; Keller
Easterling's concept of "spatial formulas."  Assemblage theory is a theory
that is very much in development, and I think it is important to take a
working approach, continuing to try to develop the model, rather than
trying to adhere to some sort of orthodoxy.

Rather than trying to explain the model of the assemblage in its entirety,
I want to focus on this activity of formulating, and use it as a port of

Again, the different "strata" of the tracking apparatus, that we now want
to situate within this new organizational model, the assemblage:
1. Tracked phenomena, in the form of datasets
2. Statistical procedures (analytics)
3. A pattern, which suggests a continuity, and a propensity.
4. A model, a formula (algorithmic).  It is stable, but you can
continually modify it.  You can always can introduce new factors.

The first thing that I want to say is that formulating is not some sort of
antidote to the statistical mindset that is part and parcel of tracking. 
It is an activity that is no stranger to applied mathematics.  Formulation
is a functional modeling of data-mined analytics, which are themselves the
result of tracking.  It is still a modality of surveillance.

What people do -- what things do too -- are tracked and measured.  This
generates the data to be mined.  Statistical algorithms find patterns. 
These patterns could be stabilized in a working model or program -- a
formula -- which allows a generativity.  So the formula is a kind of
statistical program, a kind of stabilization of data-mined analytics.  A
site where statistics are hardened in a productive, working form.

In my model of the assemblage, formulation is one kind of process, which
works in conjunction with two others that I will call resonation and
coding.  As I see it, these are the three recurrent processes that serve
to stabilize and destabilize the assemblage as a whole.   [At its basis,
the assemblage is a moving population of symbiotic, co-functioning actors 
-- actors that can be anything.  These actors synthesize (or network
together) through shared, recurrent processes of stabilization and

As it works in conjunction with coding, we can define a formula as a set
of active-organizational principles and procedures (cues or rules), which
is able to harness the timing, movement, and composition of actors, to the
extent that it can generate spatial, social, and somatic effects.

Yet it as it works in conjunction with coding, it also works in
conjunction with resonation, or affect.  In this way it addresses the
limits of coding itself, trafficking in disposition as well as position. 
In this sense the formula is something like a sensorial motif, a
propagating pattern that generates excitations and structures disposition,
yet at its core is meaningless.

Borrowing a term of Brian Massumi's, we might say that the formula, as it
works in conjunction with resonation and coding, is a kind of "activation
contour."  A contouring a flows:  a contouring that shapes, takes shape,
within the realm of routines, habits, repetitive behaviors.

Companies engage in the production of formulas -- carefully designed sets
of daily cues -- to introduce new routines and generate habits among
consumers.  The holy grail:  when things become so ingrained that we don't
think about them -- eat snacks, apply lotions, wipe counters. "If you look
hard enough, you'll find that many of the products we use every day ?
chewing gums, skin moisturizers, disinfecting wipes, air fresheners, water
purifiers, health snacks, antiperspirants, colognes, teeth whiteners,
fabric softeners, vitamins ? are results of manufactured habits." says
Charles Duhigg in the NYTimes.  A product is understood to be successful
when it becomes part of daily or weekly patterns.
["Warning: Habits May be Good For You," 13 Jul 08.]  We are not talking
about just a company of course but an assemblage of companies:  the one
that makes the product, but also those companies with whom it works in
conjunction, to track user habits and generate the datasets upon which it

A formula is an agent of modulation, a kind of calibrating infrastructure
through which actors move.  But this modulating formula is, itself,
continually shaped.  It is a contouring of flow, a stabilization of flow,
which, while partly planned, can emerge in an unplanned way.  It is an
unpredictable, emergent formation -- a two-way street.  It involves a kind
of improvisational, generative approach, which adapts by way of its uses.

It does not enclose, contain, determine, but rather, sets forth a
compositional dynamic, which interweaves various kinds of actors.  It is
not a compositional imperative, but a kind of teasing.  It sets forth an
arrangement of things, infused with a programmatic impulse, a tendency to
action.  A dispositional trigger.  A form of weak discipline, as Edwards
would call it.   Its outcomes cannot be predicted with certainty.  A
formula can only anticipate -- but it does so by snuggling up to the body,
moving in closer.

I defined a formula as a set of active-organizational principles and
procedures (cues or rules).  As it works in conjunction with coding, it
also works in conjunction with resonation, and in this way addresses the
limits of coding itself.  Here it functions like a sensorial motif -- a
propagating pattern that generates excitations and structures disposition,
yet at its core is meaningless.

Again, the three recurrent processes -- formulation, resonation, coding --
are those that serve to stabilize and destabilize the assemblage as a
whole.  As it stabilizes, the assemblage can accrue an identity, or play a
role.  We can refer to this stabilization as a form of "presencing." 
Again what is involved is not simply a "top down" dynamic but also an
emergent one -- not an imposed presence, but a form of emergent

As an example of such a sensorial motif, let's consider the Fail Whale.

The Fail Whale is basically, a visual error message that appears on
Twitter, the micro-blogging site.  It is an actor playing a role on the
assemblage that is Twitter.

Assemblages are formed through a dynamic of stabilization and
destabilization.  It is a difficult line to navigate: they can't be too
rigid, but they can't be too chaotic either.  We can say that Twitter has
serious stability problems.  At net critic Mez Breeze describes it,
Twitter is a site that experiences continual outages, much to the
frustration of its users, who are very vocal about their complaints.  In
order to "soften" the negative feelings provoked by this constant
technical failure, Twitter created the Fail Whale.

As Mez Breeze refers to it, this is an "affective redirection" designed to
"cushion error evidence" -- to transform an irritating user experience
into a more positive one.  A cushion, and an affective redirection, which
soothes users and allows them to feel connected even when experiencing
site dysfunction.

One could understand this as simply a good corporate PR strategy.  But
according to Mez, an interesting thing has occurred, which she calls
"synthetic presencing."  Fail Whale has been embraced by users in an
unanticipated way:  it has become an "emergent persona."

It now has its own fanclub, Facebook page, and theme song.  It has its own
branded merchandise.  And further, it has inspired creative appropriations
-- "community-generated Fail Whale art."  What began as an imposed
presence, became a form of emergent presencing, which then stabilized into
a brand.  So it's not a question of whether it is an imposed or an
emergent presence, a brand or a community-form, but rather, a dynamic
between stabilization and destabilization -- a seeding and a hardening.

If the Fail Whale cosmos is an assemblage -- a collection of heterogeneous
actors both human and not, who are enaged in a symbiotic co-functioning --
it is one that continually stabilizes and destabilizes itself, through
processes of resonance, coding, and formulation.

[It is easy to see formulating and coding in terms of recurrent processes
of stabilization and destabilization.  But resonation, too, works this
way.  There is a kind of pleasure in displeasure -- a dynamic embodied in
notions of the sublime or the sacred, and the very arc of the tragedy.  In
terms of the stabilization and destabilization processes of resonation
(which works in conjunction with coding and formulation), pleasure can be
understood as a kind of stabilization, but it relies upon, indeed thrives
upon, the destabilization of displeasure.  As such with fear:  we can see
it as part of productive, interlocking mechanism:  to dictate to fear God;
the affective allure of terror.  (At the level of affect, we are not
talking about specific emotions but of a deeper bodily knowingness that
does not resolve easily to emotional distinctions.  Emotions are in a
sense culturalized affects, sensations categorized.)]

The "program" that motivates the Fail Whale is an affective formula.  The
graphic is its user interface.   (Again: the formula is a kind of
statistical program, a kind of stabilization of data-mined analytics.  A
site where statistics are hardened in a productive, working form.)

We can see here other possible graphic manifestations of the formula. 
These are actors that are brewing, actors in a state of emergence.  They
are outputs of the assemblage that are waiting to catch on -- fishing
lines that extend out to you.  In this way the assemblage solicits actors,
seeks to stabilize itself and gain effectivity.  Its strength is
determined by the actors it is able to solicit, and its balance between
stabilization and destabilization.

An assemblage waiting to catch on, to catch you.  Extending its fishing
lines out to you.  A fishing -- but also a "nudging."  You can set Twitter
to "nudge" you.  You can set an Auto-Nudge -- a self-reminding update
feature.  "We know you have busy lives and may not always remember to
Twitter.  By popular demand, auto-nudge has come into being.  You can
schedule a Twitter nudge if you haven't updated in 24 hours."  A hungry
assemblage, this.

Interesting, also, to think about the dynamic of desire.  The
stabilization and destabilization processes of the assemblage -- as a
dynamic of advance and retreat, revealing and concealing -- are the very
dynamic of desire.  Think of the French courtship rituals that so
fascinated Lacan.  Desire is used up if it is too stable -- if its object
is too stable or "knowable."  Desire requires a labyrinth: endless
twistings and turnings.
Think of Caravaggio's Fortune Teller -- if we think of this as tracking,
who is tracking who?  A two-way street, and in whatever direction, it is
infused with desire.  If we are tempted to rely too much on coding, we
have to remember there is also resonance, whether libidinally understood
or not.  Psychoanalysis can be helpful here, but only if we move away from
its emphasis on a fundamental lack, as Leo Bersani would have us do.

What happens when we think of the Fail Whale as an agent of surveillance? 
People willingly divulge what they are doing, where they are going, at
various points throughout the day.  It is a form of surveillance that is
adopted and embraced by its object, and thus its functionality is
increased.  And yet it is transformed in the process.  It is open-sourced,
and yet something of a monitoring function endures -- it can "phone home."
 Seen from this perspective, the perspective of sites like Twitter as
engines of tracking, the goal would be to maintain that thread, those
tracking threads, however tenuous.

The goal of the assemblage's processes is, in Thrift's words:  a "skilled
response to the arrangement of things, a sense of the propensity of the
situation" as a "potential born out of disposition."  We are well from the
domain of the simple prediction.

8. From Geography to Topology

Time prohibits an explanation of the entire working model of the
assemblage.  So at this point I want to simply suggest one step, first,
that can be taken, which connects to our understanding of the history of
tracking:  a shift from geography to topology.  We first understood
tracking as a visual activity, situated within the cartographic tradition.
 From the observational expert at the monitor, analyzing tracked spatial
data, to mobile GIS and location-aware technologies:  the user at the
handheld device, accessing information that can help to determine the best
path of movement in space.  And yet, even as it endures, this cartographic
tradition has become simply one kind of interface to data.  Again, with
tracking technologies like RFID, such spatialization is not primary.

If the cartographical becomes just another interface, what then can we
think about as more primary?  The topological.  To understand RFID, for
example --  which can keep track of the exact item you buy, how long you
hold onto it, how far you transport it, what other products you buy in
conjunction with it, etc.  We don't have spatial relations but assemblages
over time and space, what we might call consumer-object assemblages, or
hand-tool assemblages.  We can also see these as actor-networks.

Topographies can describe the physical layout of the components that make
up such networks.  They can be used to abstract the inherent connectivity
of objects while ignoring their detailed form.  Topography is the
mathematical study of the properties that are preserved in objects in
spite of whatever deformations they might undergo.

The definition of topology leads to the following mathematical joke:  Q:
What is a topologist? A: Someone who cannot distinguish between a doughnut
and a coffee cup.  Ha ha -- but wait, how wonderful if you can see the
world otherwise, in terms of shared properties!  If you see everything in
terms of shared properties and capacities, you can look beyond illusory
form and begin to see the world in terms of connectivities and alignments.
 This is how you begin to see the world in terms of assemblages.

Open-world (security), in contrast to closed-world (control).  Weak
discipline, in contrast to strong discipline.  Non-carcerality, in
contrast to carcerality.  Organization, in contrast to institution. 
Emergent, as opposed to top-down.  As Wes Moore says:  "A model for all
contemporary expressions of power, the topological field is nomadic and
always-eluding the apprehending grasp. Consider the jihad: warfare
directed at nothing, 'total war,' radiating everywhere, beyond space,
existing in the form of fluid, mobile cells and nodes... Jihad is war in
topological, not geographical, space..."

One has to understand this also as stratified.  (In Deleuzian terms,
stratified as well as smooth.)  And here one can talk about it in terms of
a new modality of power.  It is not "control" as such:  When it a degree
of stabilization is reached (as enacted in practice), or a certain
materialization threshold crossed, effects can be produced.  It this way
actors can become functional, operational.  But not necessarily.  To act
as an agent of control, there is a balance to be achieved between
stabilization and destabilization.   A certain degree of reliability, but
not rigidity.  It is a fine line, a precarious zone.

In contrast to the geographical/cartographic, the topological allows a
relation to the body and its affects.  It already exists as part of the
phenomenological tradition [and here we can connect to Merleau-Ponty].  
Here we have to see not only signs but intensive points, valves of
excitation.  They do not refer to meanings, concepts, or precise locations
on the body so much as points of reverberation.

As Alphonso Lingis would have it, these excitations form a layer of
libidinal life, which is not a "depth" but a topography, a surfaceness: 
"the slippery effervescence at the conjuncture of mouth with breast, anus
and exterior, urethra at the point where the urine surfaces, thumb with
lips, finger with nostril.  Couplings, for the sake of the surface
effects..."(Excesses, 27).  "A mouth, it is adjustable.  It can couple
onto a nipple -- or a bottle, or a thumb.  A hand can curl around a
breast, or an arm, or another hand, or a penis?"(36)  For Lingis, this is
something on the order of a bricolage rather than a blueprint.  It is an
erotogenic surface that is extendable:  the points of reverberation act as
demands, appeals -- "they form the gaping cavities of demand, want,
desire, hunger."

These points have yet to become signs, "for what they refer to is not
something ideal, transcendent meaning, but another intensive point?"  They
generate appeals; -- they are like "raised hardnesses on the pliable
flesh" that "call for another's eye, another's touch, finger, nipple,
tongue, penis."  For Lingis, this is not a semiotic system, yet "it is out
of this kind of distributive movement of inscription that the
differentiated material for a semiotic system will be taken, and on this
purely lateral and libidinal function of craving and want that the
intentional reference of signs will be developed."

Which is why we see that the assemblage's processes of resonation, coding,
and formulation as separate (even though they work in tandem) --
reverberation and signification are different things, operating at
different levels.

What would this topology look like?   We have bendings, twistings,
meetings, foldings, which warp any representations of the body as we would
picture it.  Orientated toward couplings -- thumb to lips, finger to
nostril -- the surfacing of fluids, and conjunctions of body parts
appealed to, it is an erotogenic surface that is ever extendable.   What
we have is a very different kind of topographical map.

The ability of the surveillance apparatus to grasp processes and
anticipate outcomes -- the effectiveness of the anticipatory apparatus as
a whole -- is now becoming based on its ability to generate such mappings.
 The new cartographies of tracking -- of anticipatory surveillance --
escape the frame, and operate on the order of a kind of pre-symbolic
graffiti-making, spreading over every surface and warping Cartesian space.

How does one visualize it?  Datasets, diagrams, doodlings.   However one
wants to try to visualize it, one thing is certain:  we have here reached
the end of the panoptic paradigm, the paradigm in which tracking has
played an integral role.

9. Inadequate Confrontations

How does one visualize it -- but also, how does one critique it?  This is
a landscape that doesn't lend itself to the traditional critical approach.
 It doesn't promote a deductive orientation; rather, it would seem to
promote the opposite of that.  Its goal, as Bruno Latour would say, is not
to reduce the dynamic, but to extend it further.  And here Nigel Thrift,
as well as Latour, would agree:  "becoming able to add or assemble is more
important than subtraction."  It calls for a need to move beyond the
disciplinary, enclosing and managing contained arguments, and instead find
ways of incorporating extension.  For with the anticipatory assemblage,
there is always more not less.

As Moten and Harney would put it, this entails a stepping-out of the
"skeptical of the known" and into "an inadequate confrontation with what
exceeds it and oneself." (cited in Clough, 33)  One must ride the surface,
engaging in a kind of politics that Stephen Duncombe would suggest can
speak to the entire range of people's fantasies, passions, and desires --
what he calls a kind of "ethical spectacle."  Stevphen Shukaitis, too,
would say that politics needs to engage intensities; its task, according
to him, is "the composition of common space through processes of intensive
engagement" -- processes that are "not bound by the closure of already
understood identities and positions."

In many ways we are talking about a critical or political practice that is
less "oppositional" than compositional.  This necessitates the development
of unwieldy, awkward, combinatory approaches that are not about dealing
with lack but with excess.

How does this realization change what I'm doing now?  As a theorist, where
to go from here?

Bruno Latour says that criticism is no longer necessary.  Chris Anderson
has said that theory is no longer necessary.  Out with us all!  But now
I've just said it too.  Or have I?

To answer this question, I offer a final note.  There is one other step
that seeing the world in terms of assemblages also necessitates.  The
actors that comprise the assemblage are always playing roles.

In the world of anticipatory assemblages, *everything masquerades.*


With thanks to Stephen Graham and David Murakami Wood

#  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:
#  archive: contact: