Surveillance, Persuasion, and Panopticon

Page 1

Y. de Kort et al. (Eds.): PERSUASIVE 2007, LNCS 4744, pp. 109�120, 2007.

� Springer-Verlag Berlin Heidelberg 2007

Surveillance, Persuasion, and Panopticon

Julie Leth Jespersen, Anders Albrechtslund, Peter �hrstr�m,

Per Hasle and J�rgen Albretsen

Aalborg University, Denmark

{jlj2203,alb,poe,phasle,jalb}@hum.aau.dk

Abstract. The surveillance in public and private places, both physically and

digitally, is increasing for different reasons. In this paper we intend to discuss

surveillance and persuasive technology in an ethical perspective with an eye to

its historical and cultural context. In section 1, we present some different

tendencies of surveillance in society. In section 2, we elaborate on some

important historical ideas on surveillance. In section 3, we consider the use of

persuasive technology for surveillance purposes. In section 4, we discuss the

development towards increasing surveillance in society, at work, in public

places etc. In section 5, we draw up some ethical concerns on surveillance, and

finally. In section 6, we discuss the question of a possible need for a public and

democratic control of the use of surveillance technology.

Keywords:

Captology,

persuasive

technologies,

Panopticon, ethics,

surveillance and democratic control.

1 Introduction

There is clearly an increasing interest in various forms of surveillance and

information gathering activities. This tendency may be understood in terms of the

Greek neologism �Panopticon� (�pan� meaning �everything� and �opticon� meaning

�vision�), which was made famous by the British jurist and philosopher Jeremy

Bentham (1748-1832). Panopticon was originally the name for a specific kind of

prison (see below), but it has come to signify the wish or the will to let everything be

seen and recorded. In modern society this desire seems to be very strong. All sorts of

companies, organizations, and institutions want to store data about members,

customers, patients etc. Such data are sometimes used for other purposes than those

for which they were originally intended. Surveillance systems and IT systems are

used to monitor people for different purposes � for instance the purpose of facilitating

learning or daily work, or for security reasons.

Surveillance Tendencies

Since the introduction of computerized surveillance � also known as �dataveillance� �

in the latter part of the 20th century, we have witnessed an increase in monitoring

practices in everyday life. Today, we are being monitored at home, at work and as

consumers. Prevalent and emerging technologies such as closed-circuit television,

Page 2

110

J.L. Jespersen et al.

electronic tagging and handheld computers contribute to the flow of private

information in public space. These developments have led to what some people

consider an erosion of civil rights and liberties, and a number of studies into this

question have appeared. In the wake of the new surveillance technologies and the

concern about their implications for the individual and society as a whole,

surveillance studies have grown to be a broad field of research. The idea of the

Panopticon has provided a metaphorical framework within which to discuss

surveillance in a wide variety of contexts.

2 Surveillance in a Historical Perspective

The famous Panopticon building, �the all-seeing place�, was designed by Jeremy

Bentham towards the end of the eighteenth century. It is a type of prison, also known

as the �Inspection House�, whose construction enables an observer to watch all the

prisoners without their knowledge.

Bentham is also known as a founding father of the ethical theory of utilitarianism,

and it is in light of this theory that Panopticon should be understood. The basic

principle of utilitarianism, according to which any moral behaviour is evaluated

according to its utility, is realizeopd in the prison building, since the design facilitates

effective surveillance and control with a minimum of human resources. Moreover, the

aim is to punish effectively, but without undue human suffering, the punishment

consisting in systematic confinement rather than physical torture. The development

from physical torture to systematic confinement is by many considered a humanistic

improvement in the history of punishment. This view has later been questioned by

Michel Foucault (1926 - 1984), who in his influential book Surveiller et punir [7]

argued that the change from physical torture to systematic confinement is not

necessarily a humanistic improvement, but rather just a different way to exercise

power.

For his day, Bentham was a controversial figure, and his prison design stirred

attention. Bentham proposed a universal solution to the problems of his day: �Morals

reformed - health preserved - industry invigorated instruction diffused - public

burthens lightened - Economy seated, as it were, upon a rock - the Gordian knot of the

Poor-Laws are not cut, but untied - all by a simple idea in Architecture!� [2 p. 31]

Not only would Panopticon solve the perennial problem of what to do with

prisoners, the utilitarian architecture also provided answers to broader political

questions. The prisons of this time were brutal, unsanitary, and overcrowded. Hence

humanitarian improvements were on the political agenda, and in this context

Bentham's prison building apparently offered a more humane alternative. Panopticon

made it possible that prisoners could remain on British soil, rather than for instance

being deported to Australia, which was not unusual at the time.

However, Bentham was unable to persuade the political establishment that the

Panopticon would suffice. The decisive issue leading to the rejection of Bentham's

plans was neither the prison design itself, nor its� underlying utilitarian ideas, but the

other part of the project, namely the proposal that Panopticon should be open to

private contractors. He imagined the prison as a private, profit-seeking enterprise,

making money from the labour of the prisoners, and Bentham himself wanted to be

Page 3

Surveillance, Persuasion, and Panopticon

111

the first contractor. The financial element coursed the project to be turned down,

while the commercial gain from running prisons was considered inappropriate. The

authority and control of prison labour was best kept in bureaucratic hands, and prisons

should not meet economic criteria. Even though Panopticon was rejected by

Bentham's fellow countrymen, his idea of such a prison has been very influential. This

is among other things due to its effective application to most kinds of institutions

where people are to be watched.

The prison was designed to have an inspection tower in the middle surrounded by a

circular building with cells. The prison guard should be placed in the tower and the

prisoners isolated in the cells, which should all be visible at any time from the tower

in the middle. As much as possible of the cells should be made visible, whereas the

tower is designed to make the guards inside the tower invisible from the cells, using

screens and light as devices to hide the guard from the prisoners. The principle that

the prisoners do not know whether they are being watched or not is essential to

Panopticon. The prisoners have to presume that they are being watched all the time

and must thus behave accordingly. Actually, this renders the guard dispensable, since

the prisoners in a way have internalized the inspection by the guard. Thus the prisoner

carries out his own surveillance. It should therefore be emphasized that Panopticon is

not just the guard watching the prisoners. All personnel and all prisoners as well as

the building, that is, the prison as a whole, make up the panoptic structure. An

example blueprint can be seen in Figure 1.

Bentham was mainly occupied with prisons, but the utilitarian design could in

principle be useful everywhere. Factories, military barracks, schools, and hospitals are

institutions where the panoptic design would have obvious advantages. The specific

prison design was never realized. Bentham had high hopes, but the utilitarian building

was not to make him a wealthy man. Few prisons have been built with a direct

influence from the Panopticon, but there are some, for example Eastern State

Fig. 1. Panopticon blueprint by Jeremy Bentham, 1791[15]

Page 4

112

J.L. Jespersen et al.

Penitentiary in Pennsylvania, USA. However, many prisons, factories, military

barracks, schools, and hospitals all over the world may have been indirectly

influenced by the panoptic principles from Bentham�s prison design. As a proposal

for a specific architecture the Panopticon is an idea of the past, but Panopticism, i.e.

the underlying ideas, constitute a principle still at work today.

Bentham was declaredly a secular thinker. Like many other social thinkers of his

time, he was occupied with providing an alternative to the theologically motivated

ideas which still dominated society. Religion should be replaced with what he

considered to be reason and rationality. Even so, there seem to be some religious

undertones in Bentham�s assumption that the constant gaze of the inspector will

discourage prisoners from doing evil - and perhaps even remove the incentive to think

about evil deeds. Since the constant gaze of the inspector can hardly be realized,

because continuous supervision would be unpractical and expensive, Bentham comes

up with the essential principle that the required supervision only needs to be in the

minds of the observed. God�s omnipresent eye is thus replaced by the internalization

of the observer's potential gaze in the minds of the observed. Bentham rejected

Christianity as a solution for social problems. Nevertheless, some of the useful social

functions performed by Christianity could in his opinion with advantage be kept - just

without its religious core. With the Panopticon, God�s eye was thus transformed into a

secular context [12 p. 599]. It is, however, clear that this kind of panoptic supervision

is in fact qualitatively different from the divine omnipresence included in the

Christian idea of an almighty, omnipresent, all-knowing, and all-loving God.

In his Surveiller et punir [7], Foucault discussed Bentham�s ideas of Panopticon in

the context of the European history of criminal law from medieval physical torture to

modern day imprisonment. This history is a development leading from public,

spectacular, and instantaneous punishment - e.g. the cutting off of limbs - to

systematic confinement, putting away criminals into prisons for durations of months

or years. Foucault described Panopticism as a new political anatomy, in which

discipline replaces the earlier sovereign power (e.g. the king) that was manifested in

pomp and circumstance. The sovereign was replaced by a more subtle and hidden

authority. This new kind of authority exercised its power by objectifying the subjects

which it desired to control, and by creating knowledge about them. Therefore,

Panopticism implies a disciplinary power that aims to train and manipulate the body,

and Panopticism thus has both a negative and a positive function. The negative

function is to set up such limits as are necessary for maintaining discipline, and the

positive one is the production which is the outcome of strict discipline. Disciplinary

power comprises a series of means including drills, constant reports, testing,

regulation, and not least surveillance. Among these means, surveillance plays a

prominent part as a kind of �visibility instrument� that ensures control of the

individual. Disciplinary power thus mainly exercises its power through the gaze, more

specifically the all-seeing eye.

The Panopticon was also a laboratory or a testing ground for social techniques. It

could be used as a �machine� to carry out experiments with the aim of altering

behaviour, or of training or correcting individuals; to experiment with medicines and

monitor their effects; to try out different punishments on prisoners, according to their

crimes and character, and to seek the most effective ones; to teach different

techniques simultaneously to workers in order to decide which is the best; to try out

Page 5

Surveillance, Persuasion, and Panopticon

113

pedagogical experiments � in particular the well-debated problem of secluded

education, by using orphans. [8 p. 3] To sum up the characteristics of Panopticism, at

least five points can be made:

1. The observer is not visible from the position of the observed;

2. The observed subject is kept conscious of being visible (which together with

the principle immediately above in some cases makes it possible to omit the

actual surveillance);

3. Surveillance is made simple and straightforward. This means that most

surveillance functions can be automated;

4. Surveillance is depersonalized, because the observer's identity is

unimportant. The resulting anonymous character of power actually gives

Panopticism a democratic dimension, since anybody can in principle perform

the observation required;

5. Panoptic surveillance can be very useful for research on human behaviour,

since it due to its practice of observing people allows systematic collection of

data on human life.

Bentham�s idea of the prison building Panopticon is really just the prototype of

Panopticism, where discipline, normalization, and surveillance come together. This

observation led Foucault to ask the polemic question, why factories, military barracks,

schools, and hospitals have a striking resemblance to each other [7 p. 264]. Foucault's

own answer was the contention that Panopticism has become general and is found

everywhere. Hence, we live in a prison-like society founded on discipline and

surveillance. The formation of this society stems from many historical processes, but

it is a surveillance society and its purest form is the prison. Conceived in this way the

Panopticon is not only a building. According to Foucault it is rather a schema which

can be used for characterizing many aspects of society. In his own words it is �the

diagram of a mechanism of power reduced to its ideal form� [8 p. 3]. Although

Foucault�s Panoptic interpretation of society has been widely criticised, it remains the

preferred framework for discussing surveillance and an important inspiration within

surveillance studies.

3 The Use of Persuasive Technologies for Surveillance

In his book Persuasive technology � Using Computers to Change What We Think and

Do, B.J. Fogg gives a definition of surveillance technology, which is adapted to the

context of persuasive technology: �� surveillance technology is defined as any

computing technology that allows one party to monitor the behaviour of another to

modify behaviour in a specific way.� [6 p. 46] Fogg�s discussion of surveillance

presupposes that surveillance must be known by the person observed, who gives the

input to the surveillance system in question. Output is sent to the observer, who then

interacts with the person observed by either rewarding or punishing [6 p. 57 (Note

30)]. According to Fogg this feedback makes it reasonable to characterize the system

as interactive.

Fogg explains the widespread use of surveillance with the observation that it

works, that is, that surveillance usually yields the desired results. It has for a long time

Page 6

114

J.L. Jespersen et al.

been a much researched topic within social psychology, and the conclusion from this

research is that observation changes people�s behaviours. Behaviour is changed when

the observer is given the ability to reward or punish. This causes the observed subject

to try to meet the expectations of the observer [6 p. 46]. Furthermore, the surveillance

must be overt, since secret monitoring (covert surveillance) cannot be persuasive

technology. However it must be admitted that the way the Panopticon was to be used

implies a kind of surveillance which can also be viewed as Persuasive Technology,

even though it is not a benign sort. In fact there is an interesting parallel between

Foucault�s discussion of the Panopticon and such a use of Persuasive Technology. In

Foucault�s terms the designer of (Panopticon-like) persuasive technologies may be

viewed as an �inspector� whereas the user can be viewed as a prisoner caught in the

framework of the system. Fogg suggests that companies, in order to motivate

positively rather than punishing, should call their surveillance systems �incentive

systems� or �incentive management technology� [6 p. 48].

In dealing with �Public Compliance without Private Acceptance�, Fogg describes

some further results of surveillance. People might accept surveillance while they are

being observed and behave according to the relevant prescribed standards. When the

monitoring stops, however, they may behave like before, except if they have private

reasons for continuing the new behaviour. Fogg mentions ethical concerns in

connection with the use of surveillance. He sees ethical questions arising in relation to

the preservation of the individual�s privacy and dignity [6 p. 49]. Furthermore, he

thinks it makes a decisive difference how a system works and whether the intention is

��supportive or helpful rather than punitive� [6 p. 226].

This figure is based on Eric Neuenschwander and Daniel Berdichevsky�s model

[3 p. 55]. Eric Neuenschwander and Daniel Berdichevsk�s model differs from

Fig. 2. Fogg�s model of the ethical nature of persuasive technology [6 p. 227]

Outcome

Unintended

Intended

Ethical Unethical

Unethical

Ethical

Not reasonably

predictable

Reasonably

predictable

Ethical Unethical

Praise-

worthy

Responsible

and at fault

Not

responsible

Responsible

and at fault

Not

responsible

Responsible

and at fault

then the

designer

If the

and

Page 7

Surveillance, Persuasion, and Panopticon

115

Fogg�s on the case when the outcome is unintended, not reasonably predictable and

unethical. In this case Eric Neuenschwander and Daniel Berdichevsky do not see the

designer as responsible and at fault.The main criterion of Fogg�s proposal for

evaluating persuasive technology is indicated by the statement that �the ethical nature

of persuasive technology can hinge on whether or not the outcome was intended�

[6 p. 227]. Fogg puts forth a hierarchy of questions which in fact make up a decision

tree or decision algorithm as depicted above.

Thus, for instance, one must ask as the first question whether the outcome of some

use of a piece of persuasive technology was intended or not. Let us try to follow a

path through the decision tree. If for instance the outcome was unintended, the next

question should be whether it was reasonably predictable or not; if reasonably

predictable, then it is time to evaluate whether the outcome was ethical, relative to the

values and assumption underlying the particular analysis. If ethical, there is no special

ethical value to attach to the designer(s) in question � they cannot be praised, since

the positive outcome was not intended, but on the other hand there is no reason for a

negative assessment either. If on the other hand the outcome was unethical, the

designers are �responsible and at fault�, since they ought to have foreseen the

unethical effect (being in this case �reasonably predictable�). It should be clear how

other paths through the decision tree can be followed [6 p. 227].

Fogg presents his version of a stakeholder analysis as a methodology for analyzing

ethics [6 p. 233 f.]. In his analysis Fogg suggests that, like stakeholders, we focus on

consequences, both gains and loses. To judge consequences one must also consider

who gains the most and who loses the most. Having made our analysis we can then

draw ethical conclusions by examining gains and losses in terms of a set of values. In

this connection we must acknowledge the values and assumptions we bring into our

analysis. This stakeholder analysis does not include the intentions behind the use of a

technology, or the methods used to persuade.

Thus Fogg [6] raises various questions � and proposals � concerning the ethical

dimension of persuasion and persuasive technology. It is acknowledged that there

may be negative consequences of the use of surveillance technology. On the other

hand we must keep in mind that surveillance seems to be very effective and useful

when it comes to influencing people�s behaviour.

It should be added that surveillance is in some cases aimed at behavioural patterns

of groups rather than individuals. This kind of surveillance is more complicated, since

it involves handling large collections of data in a specific manner. The kind of

interactivity involved will also tend to be more complex. If for instance a relevant

authority wants to confront a population group with the result of social surveillance

(e.g. the drinking habits of young people) in order to obtain a certain change of the

group�s behaviour, this would presumably call for a sophisticated use of mass media.

4 Modern Surveillance Technology in a Social Perspective

In modern society the possibilities of monitoring people with the purpose of changing

behaviour is growing � and hence the temptation to do so is growing, too. Below are

Page 8

116

J.L. Jespersen et al.

listed a few examples of surveillance systems, all overt in some sense (the intention

may be covert, but the surveillance as such is overt).

Educational systems. Such systems may be designed and implemented with

the purpose of educating staff, but may easily develop into a tool for seeing

who improves his or her skills, and who does not.

Accountancy systems. Basically, these are intended to make everybody�s job

easier. Data are formalized in a way that makes it possible for many systems

to read and condition the data. But such systems are also potentially capable

of registering if somebody makes a mistake, and to count the number of

mistakes.

Security systems. Such systems are installed in order to protect personnel.

For instance, a small card with a chip and a pin-code can be used when

entering and leaving all rooms. This also provides the management with the

possibility of registering where all employees are at all times.

Such systems are all aimed at the behaviour of identifiable individuals. In other cases,

however, the focus of the surveillance is rather on the collective behaviour pattern of

specific groups in society. Here the surveillance will often be covert, i.e., not known

by the persons being observed. This surveillance of various groups involves using

various kinds of search strategies, statistics, and social sorting. In many cases data

mining techniques could be utilized. Some algorithms for performing data mining are

even classified as privacy preserving data mining (PPDM). They veil data in a given

data set, typically extracted from databases, such that privacy is ensured while the

veiled or aggregated data can be used as a reliable source for system administration,

or indeed research. Even if the focus of surveillance is on social groups and not on

individuals, researchers will in many cases be given access to sensitive personal and

individual data. And this may of course be seen as a threat to the individual right to

privacy. Philosophically, use of PPDM can be seen as a solution to the problem that

from a societal point of view the argument �right to know� must be balanced by the

individual�s �right to private life�.

Panoptic Sort and Social Sorting in Research on Personal Data

An increasing number of data describing different aspects of our personal activities

are being stored in databases. Our search activities and various kinds of metadata

related to what we consider to be our private writings and photos are continuously

inspected and gathered for use in databases. In some cases, these databases can be

accessed via the Internet. In fact, most of the databases on personal data may in a not

too distant future be integrated into one comprehensive system. Information

technology may in this way give rise to a new kind of Panopticism. This will

obviously be very attractive from a research point of view � and so it may be for

private purposes, for that matter. The new Panopticon will be the integrated system of

all the personal databases to which the researcher is given access. Given the relevant

political permission, the researcher may of course in principle use this new

Panopticon as a laboratory � as suggested by Foucault in order to carry out social

experiments. However, it is more likely that researchers will generally have to

concentrate on observations rather than experiments, the latter being ethically and

Page 9

Surveillance, Persuasion, and Panopticon

117

politically much more sensitive. But the researcher, like the Panopticon inspector, will

be able to �spy� on � or to put it more neutrally, to observe � all the individuals in the

system in order to analyze and evaluate their behaviour in various ways.

From the perspective of the welfare of the persons being described, research on

personal data may obviously in many cases be beneficial. However, in other cases

people may find it problematic that somebody is carrying out research on their

personal data and thereby providing results which may turn out to affect their lives.

The results of this kind of research will inevitably be used in various kinds of decision

making, notably and not least when employing personnel. One important problem has

to do with the so-called �panoptic sort�, which is a concept Oscar Gandy introduced

and described in The Panoptic Sort: A Political Economy of Personal Information

(1993) [10] as well as in other works:

The panoptic sort is a complex discriminatory technology. It is panoptic in that

it considers all information about individual status and behaviour to be

potentially useful in the production of intelligence about a person�s economic

value. It is discriminatory because it is used to sort people into categories

based upon these estimates. [11 p. 133-34]

Usually, information is gathered in order to profile and categorize, and in this way

people are included and excluded, qualified and disqualified, in all sorts of contexts.

The panoptic sort is thus a pre-emptive means in the hands of governments and

commercial enterprises. Gandy�s primary concern back in 1993 seems to have been

discriminatory practices related to consumer surveillance, using databases to target

�valuable� customers for further advertising, while undesired consumers are

dismissed. In the light of the suicide attacks of September 11, 2001, on New York�s

Twin Towers and The Pentagon in Washington, D.C., the panoptic sort has become

topical in the context of the so-called �War on Terror�. All the suicide attackers were

Muslims, and this fact could induce a panoptic sort profiling and categorizing of this

group of people as potentially dangerous. This could in practice lead to wide

discrimination in airports and at border crossings against the millions of people that fit

this profile.

This leads to another point relating to the discriminatory technology of the

panoptic sort, since profiling and categorizing is not only something that affects

people at the individual level. David Lyon has described how surveillance goes

beyond the individual realm to the social realm:

[I] argue that there are dangers inherent in surveillance systems whose crucial

coding mechanisms involve categories derived from stereotypical or

prejudicial sources. Given that surveillance now touches all of us who live in

technologically �advanced� societies in the routine activities of everyday life,

on the move as well as in fixed locations, the risks presented go well beyond

anything that quests for �privacy� or �data protection� can cope with on their

own. [13 p. 2]

Thus, discussions of surveillance in the context of research on personal data should

not be limited to considerations of privacy � the observation that certain information

due to its sensible and private nature should not be accessible without the consent of

the relevant person(s). We must also be concerned with the super-individual level and

Page 10

118

J.L. Jespersen et al.

include questions relating to the social justice of using information for research

purposes. Similarly, we should be wary of prejudices which could be inherited in the

way information has been gathered.

5 Modern Surveillance Technology in an Ethical Perspective

As mentioned above a distinction should be made between surveillance of the

individual and surveillance of groups in society. However, in both cases there are

important ethical problems to consider.

Individuals may be watched for security reasons. An example of this kind of

surveillance could be digital surveillance, where the system registers the behaviour of

the individual. This kind of surveillance is personalized and may imply, or utilize,

various accusations concerning the individual (whether proved or not). Just for

example, a store owner might video-tape a specific employee whom he suspects of

stealing. This problem has prompted the invention of the cash register [14].

Mass surveillance targets groups or certain profiles, which can lead to

discrimination of certain groups or people with certain traits. But it may also be

justified and in fact necessary to make use of this surveillance technique, for instance

to avoid terror activity.

Both kinds of surveillance raise concerns with respect to individual privacy and

other civil liberties. On the other hand, surveillance can be seen as a form of

persuasive design. If the individual is aware of being monitored he or she will be

motivated to change behaviour to what is considered desirable according to society�s

norms. The same thing applies to group surveillance.

Surveillance of social groups is often carried out using various kinds of data

mining, as mentioned earlier. The technique can popularly be defined as �the

nontrivial extraction of implicit, previously unknown, and potentially useful

information from data� or as �the science of extracting useful information from large

data sets or databases� [16]. The purpose of data mining makes it applicable to

behaviour patterns of selected social groups, especially when wishing to make general

observations without giving researchers access to all individual data. In [4], [5] and

[9] data mining and its privacy preserving potentials are discussed.

There is however one further question that applies to the ethics of persuasion and

surveillance of individuals as well as groups. As argued by B.T.C. Atkinson [1] the

intent of �the persuader� should be made clear to whoever is exposed to the use of

persuasive technology. If this condition is not fulfilled, the process should be

characterized as manipulation. The view is that openness regarding the purpose of the

technology employed could serve as a kind of �ethical safeguard�. In relation to the

use of surveillance technology for persuasion purposes this means that the people

under observation, should be made aware of the intent of those conducting the

surveillance. However, in many cases the purpose of surveillance is rather vague and

general. When it comes to group surveillance, mass media will usually be involved

when the result is communicated. We agree with this basic principle, but it must be

admitted that it is difficult to make a clear distinction between the use of mass

communication for persuasive purposes as opposed to the purpose of informing.

Page 11

Surveillance, Persuasion, and Panopticon

119

6 Conclusions and Further Perspectives

Surveillance makes it possible to change the behaviour of individuals � at least as

long as they know they are being monitored. Surveillance may also influence the

behaviour patterns of social groups, especially if the results of the surveillance studies

are communicated in an adequate manner using mass media.

If one wants to limit surveillance involving private and ethically sensitive data, it

may be required that access to the data is limited. All data may be depersonalized

before they are used for research purposes. But such requirements may also make

many important research projects very difficult or outright impossible to carry out.

For this reason, it may be relevant to look for ethically acceptable alternatives to

depersonalization and other requirements limiting research access to all relevant data.

One obvious solution is to make use of privacy preserving data mining algorithms

which make it possible to run the surveillance systems without human inspection of

sensitive data. The appealing aspect of this approach is the fact that data need not be

traced back to any individual. On the other hand, this makes an assessment of the

correct use of the data with reference to ethical norms more important.

There is no reason to underplay the importance of developing good data mining

algorithms, as long as this kind of computer based privacy preserving tools is not

presented as the solution to all ethical problems of panoptic research and surveillance.

The development of data mining algorithms is important as a possible way to ensure a

high ethical standard in panoptic research and surveillance of personal data. But it is

also clear that the use of such algorithms is not sufficient for that purpose. We also

need some sort of public and democratic awareness and discussion of the values on

which the data mining algorithms are based. Here the only acceptable solution seems

to be some sort of control on the collection as well as the use of data in panoptic

research and surveillance. This control should also include other aspects and problems

such as social sorting by means of which can be disadvantageous for certain people in

an unjust manner. Another potential problem, which calls for democratic awareness is

data dredging, i.e.,��the inappropriate (sometimes deliberately so) search for

�statistically significant� relationships in large quantities of data� [17].

Acknowledgements

We thank Mikkel Leth Jespersen for very useful comments on an earlier version of

this paper. We also acknowledge the stimulus and support of the �European project on

delimiting the research concept and the research activities (EURECA)� sponsored by

the European Commission, DG-Research, as part of the Science and Society research

programme � 6th Framework.

References

1. Atkinson, B.: Captology: A Critical Review, in IJsselsteijn, W. In: IJsselsteijn, W., de

Kort, Y., Midden, C., Eggen, B., van den Hoven, E. (eds.) PERSUASIVE 2006. LNCS,

vol. 3962, pp. 171�182. Springer, Heidelberg (2006)

2. Bentham, J., Bozovic, M.: The Panopticon writings, London, New York, Verso (1995)

Page 12

120

J.L. Jespersen et al.

3. Berdichevsky, D., Neuenschwander, E.: Toward an ethics of persuasive technology.

Communication of the ACM (1999)

4. Bertino, E., Nai Fovino, I., Parasiliti Provenza, L.: A Framework for Evaluating Privacy

Preserving Data Mining Algorithms. Data Mining and Knowledge Discovery 11, 121�154

(2005)

5. Domingo-Ferrer, J., Torra, V.: Privacy in Data Mining. Data Mining and Knowledge

Discovery 11, 117�119 (2005)

6. Fogg, B.J.: Persuasive technology � Using computers to change what we think and do.

Morgan Kaufmann, San Francisco (2003)

7. Foucault, M.: Surveiller et punir: naissance de la prison, Paris, Gallimard (1975)

8. Foucault, M.: Discipline & Punish. Translated from the French by Alan Sheridan , 195�

228 (1977),

http://foucault.info/documents/disciplineAndPunish/foucault.disciplineAndPunish.panOpti

cism.html

9. Fule, P., Roddick, J.: Detecting Privacy and Ethical Sensitivity in Data Mining Results. In:

Estivill-Castro, V. (ed.) Conferences in Research and Practice in Information Technology,

Australian Computer Society, Inc. (2004)

10. Gandy, O.: The Panoptic Sort: A Political Economy of Personal Information. Westview

Press (1993)

11. Gandy, O.: Coming to Terms with the Panoptic Sort. In: Lyon, D., Zureik, E. (eds.)

Computers, Surveillance, and Privacy, Minneapolis University Press (1996)

12. Lyon, D.: Bentham�s Panopticon: From Moral Architecture to Electronic Surveillance.

Queen�s Quarterly 98, 596�617 (1991)

13. Lyon, D.: Surveillance as Social Sorting: Privacy, Risk, and Digital Discrimination,

Routledge (2003)

14. Quorion Data Systems (accessed March 28, 2007),

www.quorion.de/Cash%20Registers/cash_register_history_types.htm

15. Wikipedia 2006a. Image: Panopticon (accessed March 9, 2006), available at the website

http://en.wikipedia.org/wiki/Image:Panopticon.jpg

16. Wikipedia: 2006b. Data mining (accessed February 27, 2006), available at the website

http://en.wikipedia.org/wiki/Data_mining

17. Wikipedia: 2007c. Data dredging (accessed January 30, 2007), available at the

website http://en.wikipedia.org/wiki/Data_dredging