Nettime mailing list archives

<nettime> Data Retention in the European Union: When a Call Returns
oli {AT} zeromail.org on Sat, 11 Oct 2008 20:15:05 +0200 (CEST)

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

<nettime> Data Retention in the European Union: When a Call Returns

Data Retention in the European Union:  When a Call Returns
Article by Oliver Leistert
Published in the International Journal of Communication
(IJoC): http://ijoc.org/ojs/index.php/ijoc/article/view/302

I have never been satisfied with the discussions around Data
Retention. This text tries to discuss the technology while not
being solely a technical text.


Title: Data Retention in the European Union:  When a Call Returns


Retained (via digital storage) metadata of telecommunications acts
change and transform into the content of something else: a
surveillance program. Originating from telecommunications as a
protocol necessity, the metadata is fed into a data space that freezes
and manipulates the time axis. This measurement of post-9/11 governing
is only one item in an assemblage of surveillance technologies that
are not watching, in the manner of traditional CCTV, but processing
the population under observation. Since data processing is a more
recent, counterintuitive, and still relatively opaque principle, its
capacities are not adequately understood, nor are they established in
the general understanding. The concept of a data space that provides
movement within and between data described here illustrates the powers
of data retention in an imaginable way.


This article's subject is a proposal for how to attain a more
intuitive point of view of the seldom recognised, immense power of
data retention in a supranational information space like the EU. A
problem of massive data collection in huge databases with highly
sophisticated information technologies, such as data mining, lies in
its imperceptibility. While an analog archive is impressive simply by
its sensual perceptibility, digital data collections do not matter at
all on the phenological side.
By providing more intuitive concepts around the metaphor of data
space, this article's main intention is to demonstrate the powers of
data retention from within the technologies' logic as an immanent
approach. Further, it suggests that the so-called metadata of
communication acts cannot be regarded as such until they are being
treated as content data on a different level — the level of data
retention. The article’s cause is the Data Retention Directive of the
European Union, which forces the production of an unequalled
decentralized accumulation of all metadata of communication acts
(e.g., call detail records of telephony and Internet traffic) within
the EU.

The Directive in a Nutshell

The European Union Directive on data retention [1], though less than
10 pages long, is invested with considerable authority. It directs the
member states to pass a law compelling each provider of
telecommunications services to retain traffic and location data for at
least the past six, and at most, the last 24 months. As stated in the
first sentence of Article 1:

This Directive aims to harmonise Member States’ provisions concerning
the obligations of the providers of publicly available electronic
communications services or of public communications networks with
respect to the retention of certain data which are generated or
processed by them, in order to ensure that the data are available for
the purpose of the investigation, detection, and prosecution of
serious crime, as defined by each Member State in its national law.
(2006/24/EC: 56)

Certain data here means traffic data and location data, and as defined
here, is data generated by or during an act of telecommunications with
a mobile phone, a landline, or via the Internet, minus the “content.”
These inquiries around data ask who, when, where, with whom, how long,
and so forth — but do not ask about the nature of the communication.
The data generated during unsuccessful acts of telecommunication is
also similarly analysed.

To “harmonise” means to implement technical standards of retention,
and to do so for data access from anywhere in the EU. [2] The data
that one profiler gets from a member country shall technically be
compatible with the data s/he obtains from any other member state.

Unquestionably, the Directive rests on a differentiation between
traffic/location data and content data. The retention of all data
generated during an act of telecommunication might not fit with
Directive 2002/58/EC on data processing and privacy of July 12, 2002,
and other fundamental human rights. [3]

Traffic Data and Content Data

This differentiation between data that contains the structural
components of communication and data that relates to content is, first
of all, technically inspired, and thus indicates the possibility that
the Directive is produced with this technical differentiation in mind.
Traffic data consists only of the information needed to technically
initiate, sustain, and terminate an act of communication. However, as
the Directive aims at “the investigation, detection, and prosecution
of serious crime,” the relevant data is divided into the following
symbolic subcategories:

- data necessary to trace and identify the source of a communication;
- data necessary to identify the destination of a communication;
- data necessary to identify the date, time, and duration of a
- data necessary to identify the type of communication;
- data necessary to identify users’ communication equipment or what
purports to be their equipment; and
- data necessary to identify the location of mobile communication

Obviously, this subcategorisation of traffic data is not inherent to
the data (as it is produced through the cooperation of user, device,
and infrastructure), but this subcategorisation represents a scheme of
the profiler's questions. In addition, the term communication itself
is problematic here. Article 6, Sentence 2, of the Directive states:
“No data revealing the content of the communication may be retained
pursuant to this Directive.” But the person calling the crisis line is
not ordering pizza. There are undeniably semantic elements in the mass
of so-called traffic data. Data identifying the person calling the
crisis line unavoidably reveals content (by virtue of the nature of
the call.) This is one of the reasons why such data is protected under
privacy legislation in the European Union.

Two Symbolic Worlds

The symbol-processing machines called computers that enable
telecommunications services, and the symbol-“processing” (or more
accurately, “recognising”) humans trying to communicate via these
media are not equivalent to each other in terms of hermeneutic and
cognitive function. The realm of the symbolic mediates the
trajectories of the real and the imaginary for and between humans. The
symbolic — constituted by signs and by modes of signification — is the
basis of any human communication. Machines also process symbols, but
neutrally (i.e., neither in reference to the outside nor to the
inside). [4] They don't rely on a world representation in their minds.
[5] Computers are radically autistic (Krämer, 1992,  p. 339) in terms
of human characteristics. Their only concern is whether “it” (the
encoded datum) is computable or not. [6]

Computer-processed data does not relate to anything outside the
machine. The caller's ID is a device ID. The profiler's assumption
that the device is equal to, or identical with, or identifiable with
its user is an obvious pragmatic reduction. It serves here as a
metaphor for the appropriation of technical data for profiling issues,
because it may be argued that this is not the same kind of data
anymore: Traffic data is transformed into something else in order to
lend itself to specific uses, including those of surveillance, by the

Interfaces as Gates between Symbolic Worlds

Interfaces connect computers with the outside world, and vice versa.
It is via interfaces that input and output can be processed. A
computer without any interface is both a paradox and an impossibility.

Data entered into computers for a telecommunication act are to be
divided into “outer machinic” and “inner machinic.” The ubiquitous
Internet Protocol (IP) is an apt example: Domain names are outer
machinic data as they relate to the outer world, while IPs are
inner-machinic addresses of the Internet, without reference to its
outside. A domain name server does the necessary translation between
the two modes.

In this analogy, IPs are traffic data, while domain names are content
data. This neatly describes the conflict in a nutshell. IPs do not
represent anything but a numerical address of a specific machine.
Domain names do not represent a specific machine, but signify codes
created by and comprehensible to humans, such as “fbi.gov”.

Retaining traffic data shifts the address space of meanings from
machinic to human. This is precisely why ethical problems may occur.
Computers process symbols regardless of their meaning, as long as they
are operational. Human beings process symbols regardless of their
technical viability, as long as they are meaningful. Two ontologically
distinct worlds collide here [7] as data from one world is fed into
the circuits of the other.

While the traffic data is generated during a telecommunication act,
its signification for humans and outside of the processing computers
belonging to the telecommunications infrastructure itself is produced
by its retention, via standardised access and, later, its reference to
people's names.

Taking this shift seriously, one may speak here of “new” data and
reject the notion of the common identity of traffic data and retained
data. The retained data has no operational meaning anymore, but it is
transformed into the symbolic: It now represents the movements and
telecommunication acts of people.

As the traffic data is produced “automatically” by the communications
technology itself as a working necessity, the production of the
profiler's data is very economic. It is not even necessary to
introduce new hardware or sensors into the existent infrastructure.

By retaining the data for the purpose of crime investigation, a
significant shift transforms [8] the prior technically necessary data
into data that now has a meaning for humans, and so is strictly
speaking some sort of content data. The process of retention itself
inherently supports the transformation of data: Now, it is archived on
some dedicated storage media in a dedicated storage form, easily
accessible and searchable.

This transformation of traffic and location data into content data
cannot be described within the logic of the act of communication
itself, as this act is outperformed by machinic operations. The
database containing metadata is a newly generated object, produced by
specific algorithmic operations and strategic settings.

A New World

A database is a "collection of data or information organised for rapid
search and retrieval." [9] In this case, the database is needed to
ensure the persistence of the ephemeral by generating durable data
sets. These data sets consist of successful or unsuccessful acts of
telecommunication registered and executed in the telecommunication
infrastructure within the geographical space of the European Union.
Data thus transformed does not constitute an object that belongs to
the telecommunication infrastructure itself; rather, it is a
materialisation that rests on a massive time-axis manipulation.
Essentially parasitic, it negates existent chronology and transforms
expired data logs into valid current information. [10] It generates an
ahistorical time window of 6 to 24 months and consists of symbolic
representations of a compressed space-time manifold through data
doubles — or “dividuals,” as Deleuze (1992) puts it — preserved for
profiling and crime prosecution. As a form of decentralised
surveillance architecture, the data retention database infrastructure
is very robust. This is because metadata is always processed from two
sides, that of the sender as well as the receiver.

Since the time axis is thus deconstructed and then reorganised by the
storage operation, profilers enter a hyperreal/surreal world of
replicated identity. This strategic copy is navigable through space
and time. User locations and communications now can be reconfigured
into feedback loops and transmitted back and forth arbitrarily.

Twofold is a Ticket

The convention to retain data on both sides of the communication
renders possible data-travel along these electronic meridians by
following communication patterns of constructed groups. Starting with
database d1 and data double a1' of communication act a1 that connects
device A and B, the travel continues to database d2, [11] containing
the data double a1'', which represents the connection of A and B once
again, but from the reversed perspective. Any connection of B with C
is now within reach, followed by C with D, and so forth.

This doubling of representation enables the easy construction of
groups whose members have nothing in common in real life. Where
communication acts of one single person were analysed, the topology
was simple and compact: a star with the person's device in the middle
as the connecting point with the rest. What can be called a “crawling”
topology enabled via the storage of the two perspectives of
communication act is potentially endless and
self-generating/recursive. This arbitrariness of possible searches
makes the decentralised time-space representation susceptible to
mining, and as such to the production of new knowledge. With the
persistent doubled articulation of such data-doubles, all
communication acts in the European Union are continuously in
circulation and within profiler reach — beginning with any act — even
while the devices are presented solely as nodes of a static network
laid down in a database. But the profilers’ intent is to connect the
dots — the data doubles of telecommunication acts.  The system can
connect any data with any data (simply because they have been
generated) in a fully automated scanning mode through predefined
algorithmic search mechanisms.

Lingua Franca in Data Space: Location Data

Location data of mobile phones might be the most important trace for
the profiler, as location becomes the lingua franca in the
surveillance and profiling community (Curry, 2004). This data is of
specific value and shows most precisely the difference between
metadata and the actual data used in data retention: “. . . as long as
the phone is turned on, it serves as a passport into a monitored
electromagnetic enclosure” (Andrejevic, 2007, p. 100).
Location data is a necessity for the operation of the GSM standard
that is solely used in the EU. Any mobile phone switched “on” produces
locality-based data via the closest cells of the phone. This is needed
to identify the phone in the provider's network and to offer the best
connectivity available. However, retention of this data is not
essential for the efficient functioning of the communication
structure. [12]

To retain this data, to process it, and to make it accessible via
searchable databases changes its status from ephemeral to temporally
enduring. But additionally, retained location data remodels a
four-dimensional world. Device locations are laid down as a space-time
continuum of at least six months. Through algorithmic operations,
these data can easily be visualised and brought into navigable and
replicable form.

It is important to emphasise that location data derives not only
through the explicit mobile phone connections; any landline call or
communication service is locatable too, and the customer's data (such
as address and billing information) are retained as well. The mobile
phone's data offers profilers the luxury of the detailed history of a
person's movement itself with remarkable precision as long as the
phone is switched “on.” This density of data results from the real
world, from the material movements of bodies on earth. A one-to-one
representation of the loci of any mobile device is the most accurate
representation possible of people's movements and places (that can be
achieved without invading the body itself). Therefore, participating
in modern life in the 21st century is becoming increasingly dependent
on a device that might also be called location tracker.

Tomorrow, Not Today, and Not in the Future

Once data retention is fully operational throughout the European
Union, location data will be the ubiquitous source for the assessment
of the population's movement.
By the principle of double (twofold) retention of all
telecommunication acts, an operation on this data space is possible in
any logical direction of the database's sets. Double retention in
combination with location data may even serve to generate a more than
four- dimensional world, as any topological construction is
computable. The genesis of new technologically-oriented perspectives
on the past (which will barely then qualify as the past) will then be
cast in the looping circuitry of algorithmic operations.

But it is not only the past that is being re-mastered. With
statistical methods, future predictions are also possible. As the base
of these statistical calculations is a one-to-one representation and
not a sample, the outcome of the prediction will be as accurate as
statistics can possibly be. In other words: with the retention of
location and traffic data, even the future of Europe's population is
under (re-) construction. Groups and the movements of groups will be
predictable at the pace of the latest CPU.

Data Retention as an Element of an Assemblage

Seen from a broader scope, data retention fits well into the post-9/11
war on terror measurements. The shift toward an omniscient
surveillance-state has generally often been compared to scenarios
familiar from the prophetic novel 1984 by George Orwell (1949). But
there are critical differences. The analogy had been underscored by
the historical concept of the panopticon, introduced by Bentham (1785)
and popularised by Foucault (1977). But where the panopticon draws its
power from the fact that the surveilled never know if they are
surveilled, and therefore internalise habits as if they were
surveilled, the present situation, fostered by ongoing modalities such
as data retention, should more accurately be referred to as
panspectron or surveillant assemblage.

The term panspectron, as introduced by Sandra Braman (2006), refers to
a state of things where no surveillance subject is specifically
invoked in order to trigger an information collection process. Rather,
information is collected about everything and everyone all the time.
An individual subject appears only when a particular question needs to
be answered, triggering data mining for particular information within
the mass already gathered, in order to precisely answer that question.
And while populations remain generally aware of the unmoved and
intimidating presence of the panopticon, they tend to be unaware of
the aggressive efficiency of specific modes of information collection.
Data retention exactly fits into this conceptual frame, just as with
Passenger Name Records and SWIFT financial data. These are sustained
and augmented without any specified trigger, and therefore,
potentially infinite.

The concept of the surveillant assemblage, as introduced by Kevin
Haggerty and Richard Ericson (2000), refers to a multiplicity of
heterogeneous objects, whose unity is solely functional. As an
assemblage is always a “potentiality,” this concept can be connected
to the developing panspectron, which also resides in the background as
a formidable and ambivalent latency.

Paradoxically, the weakness of the assemblage is also its power: “As
it is multiple, unstable, and lacks discernible boundaries or
responsible governmental departments, the surveillant assemblage
cannot be dismantled by prohibiting a particularly unpalatable
technology” (Haggerty & Ericson, 2000, p. 609).

The main directive of the assemblage is to transform the body into
virtual bytes of information — data doubles. “And while such doubles
ostensibly refer back to particular individuals, they transcend a
purely representational idiom” (Haggerty & Ericson, 2000, p. 614).
Discrimination and social sorting is amongst the socio-political
consequences that subjects may experience as concrete back-references.
Data retention is an example of a functional element of the assemblage
that might refer back to subjects. It is unpredictable, and possibly
holds radical consequences for the subject.

Conlusion: Communication is Doubled

The Royal Academy of Engineering speaks of necessary conditions for
the application of surveillance technologies:

"Reciprocity between subject and controller is essential to ensure
that data collection and surveillance technologies are used in a fair
way. Reciprocity is the establishment of two-way communication and
genuine dialogue, and is key to making surveillance acceptable to
citizens." (Raeng, 2007, p. 8)
How can this be done? It is currently an open question as to whether a
significant portion of European mobile phone users know about data
retention at all or are sufficiently aware of its repercussions.

This is not only true of data retention, but also of most data
collecting, processing, and mining practices. This significant lack of
knowledge, understanding, and consciousness about new paradigms of
technologically-enabled surveillance and related governmental
practices and commercial business is comparable to a situation of
betrayal: While the people communicating assume that their privacy is
strongly protected, a permanent but virtual eavesdropping operation is
in the making that might become all too real later.

The fact is that data retention of telecommunications data belongs to
an assemblage of new emerging forms of control endemic to a networking
society. Its power results from the wide acceptance and usage that
electronic communications media have gained in the last years. Yet
each link within this mega- or giga-weave of omnipotent
cyber-connectivity is a single, tenuous, finite, vulnerable strand:
the voice of the person at the other end of the line.To take part in
this connected world is coupled with a number of drastic consequences
that have been described above. Calling someone produces data far
beyond the call. And while the call might be forgotten to have taken
place within a couple of days by the people who have spoken to one
another, the technical infrastructure implemented for data retention
ensures for up to 24 months that it has happened, regardless of
whether or not any human being remembers the call.


[1] European Union directive 2006/24/EC of March 2006 on the retention
of data generated or processed in connection with the provision of
publicly available electronic communications services or of public
communication networks and amending directive 2002/58/EC, here after:

[2] It is worth mentioning that this directive has passed as a so
called first pillar directive using the single market power of the
European Union and not as a third pillar directive pursuent to the
Union's power fighting crimes. “Once the choice was made to go ahead
with the Directive as a First Pillar initiative, the Commission and
the Council took the position that, legally speaking, the Directive
could not regulate police access to communications data. Anything
having to do with the police was strictly Third Pillar“ (Bignami 2007:

[3] If the actual implementation does so and if it complies with
national laws on privacy is not of concern here, neither is the
discourse on data retention and citizen rights. For a historical
discussion of data retention in Europe and Canada, see Warner 2005.

[4] For a precise discussion of the common misunderstanding of
computers as enhancements  of humans and the anthropocentric (mental)
case, see Tholen 1994.

[5] A lession any “artificial intelligence” research had to learn.

[6] The radical difference of computers and human beings is constantly
being blurred by anthropocentric thinking of technology. Viability as
the leading paradigm in science and technology as well pretends
connections between machines and the outside world that do not exist
(Winkler 2004: 226-30).

[7] It is the blindness of the machines that in the first place
renders possible the signifying human work. Data collection,
processing and mining can only be done if the machines do not
interfere with opinions about the meaning of the data. Generally,
there are two ways out of this dilemma: enter ethics into machines or
disallow certain machinic operations.

[8] As Bruno Latour puts it, “There is only transformation.
Information as something that will be carried through space and time,
without deformation, is a complete myth [. . .]. From the same bytes,
in terms of “abstract encoding,” the output you get is entirely
different, depending on the medium you use.” (Latour, 2004, p. 154).

[9] http://www.britannica.com/ebc/article-9362288

[10] Winkler calls such as state a co-presence of past and present
(Winkler, 1997, p. 175).

[11] D1 and d2 might be same database if both devices use
communication services from one and the same provider. The bigger the
provider, the more often this is the case. In a monopoly situation,
every communication act is laid down in one single database

[12] It is not a necessity for the member states to force the
retention of this specific data. Location data in the strict EU
directive's sense has to be retained only in combination with a
successful or unsuccessful act of communication. Nonetheless, to
retain all location data is a common practice in law enforcement.


2006/24/EC. (European Union Directive 2006/24/EC of March 2006 on the
retention of data generated or processed in connection with the
provision of publicly available electronic communications services or
of public communication networks and amending Directive 2002/58/EC).
Official Journal of the European Union, L 105, 56-63.

Andrejevic, M. (2007). iSpy: Surveillance and Power in the Interactive
Era. Kansas: University of Kansas Press.

Bentham, J. (1995). The Panopticon Writings. London: Verso.

Bignami, F. (2007). Protecting Privacy against the Police in the
European Union: The Data Retention Directive. Duke Law School Science,
Technology and Innovation Research Paper Series, Research Paper No. 13.

Braman, S. (2006). Tactical Memory: the Politics of Openness in the
Construction of Memory. First Monday, 11(7). Online-Publication.
URL: http://firstmonday.org/issues/issue11_7/braman/index.html

Curry, M.R. (2004). The Profiler's Question and the Treacherous
Traveler: Narratives of Belonging in Commercial Aviation. Surveillance
& Society 1(4): 475-499.

Deleuze, G. (1992). Postscript on the Societies of Control. October 59
(Winter): 3-7.

Foucault, M. (1977). Discipline and Punish: The Birth of the Prison.
New York: Vintage.

Haggerty, K., & R. Ericson. (2000). The surveillant assemblage.
British Journal of Sociology, (51)4: 605-622.

Krämer, S. (1992). Symbolische Maschinen, Computer und der Verlust des
Ethischen im geistigen Tun. In W. Coy, W. et al., (Eds.), Sichtweisen
der Informatik (pp. 335-342). Braunschweig/Wiesbaden: Viehweg.

Latour, B. (2004). There is no Information, only Transformation. In
Geert Lovink, Uncanny Networks: Dialogues with the Virtual
Intelligentsia (pp. 154-160). Cambridge, MA.: MIT Press.

Orwell, G. (1949). Nineteen Eighty-Four. New York: Penguin.

Raeng (The Royal Academy of Engineering). (2007). Dilemmas of Privacy
and Surveillance. Challenges of Technological Change.
Online-Publication. URL:

Tholen, G.F. (1994). Platzverweis. Unmögliche Zwischenspiele zwischen
Mensch und Maschine. In N. Bolz, F. Kittler, G.F. Tholen (Eds.),
Computer als Medium (pp. 111-135). München: Fink.

Warner, J. (2005). The Right to Oblivion: Data Retention from Canada
to Europe in Three Backward Steps. University of Ottawa Law &
Technology Journal. 2: 77-104.

Winkler, H. (1997). Docuverse - zur Medientheorie der Computer.
München: Boer.

Winkler, H. (2004). Diskursökonomie. Versuch über die innere Ökonomie
der Medien. Frankfurt am Main: Suhrkamp.

#  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://mail.kein.org/mailman/listinfo/nettime-l
#  archive: http://www.nettime.org contact: nettime {AT} kein.org