Content uploaded by Michael Decker
Author content
All content in this area was uploaded by Michael Decker on May 20, 2015
Content may be subject to copyright.
FOCUS
The role of ethics in interdisciplinary technology assessment
Published online: 27 January 2004
ÓSpringer-Verlag 2004
Abstract Technology Assessment (TA) is a problem oriented endeavour
dealing with political, societal, ecological, etc. problems. Only in rare cases is
one individual scientific discipline sufficient to assess these problems. Usually
the perspectives of different scientific disciplines have to be combined in order
to develop interdisciplinary based recommendations to act. In this paper a
quality controlled interdisciplinary discussion process is described which
encourages an expert group to generate argumentation chains cross-cutting the
disciplinary boundaries. The role of ethical reflection in this procedure depends
on the problem situation. Whenever a technical application is on the agenda
which cannot be allocated to a so-called ‘‘business-as-usual’’ case, one would
ask for ethical reflection. This contribution argues that this ethical reflection
has to take place together with the interdisciplinary discussion due to two
reasons. Firstly, the technical, economical, legal and social aspects are deeply
cross-correlated with the ethical reflection. And secondly, participating in such
interdisciplinary discussions enables an ethical reflection which keeps in touch
with the real world. Two case studies dealing with robotics applications in
health care are mentioned as examples for problem settings, in which inter-
disciplinary TA succeed in developing discipline-crossing argumentation
chains.
Zusammenfassung Technikfolgenabscha
¨tzung (TA) ist ein problemorientiertes
Unterfangen, das sich unter anderem mit politischen, gesellschaftlichen und
o
¨kologischen Problemen befasst. In den seltensten Fa
¨llen reicht eine einzelne
wissenschaftliche Disziplin aus, diese Probleme zu beurteilen. Gewo
¨hnlich sind
die Perspektiven verschiedener wissenschaftlicher Disziplinen zusammenzubr
Poiesis Prax (2004) 2: 139–156
DOI 10.1007/s10202-003-0047-0
M. Decker
Institut fu
¨r Technikfolgenabscha
¨tzung und Systemanalyse,
Forschungszentrum Karlsruhe,
PF 3640, 76021 Karlsruhe, Germany
E-mail: Michael.Decker@itas.fzk.de
Michael Decker
ingen, um interdisziplina
¨r fundierte Handlungsempfehlungen zu entwickeln. In
diesem Beitrag wird ein interdisziplina
¨rer Diskussionsprozess beschrieben, der,
unter Aufrechterhaltung einer Qualita
¨tskontrolle, eine Expertengruppe dazu
ermutigt, Argumentationsketten aufzubauen, die die Grenzen ihrer Disziplinen
durchbrechen. Welche Rolle ethische Reflektion in dieser Prozedur spielt, ha
¨ngt
von der Problemsituation ab. Wann immer eine technische Anwendung auf der
Agenda steht, die keinem so genannten ‘‘Business-as-usual’’-Fall zugeordnet
werden kann, ist ethische Reflektion gefragt. Das Argument dieses Beitrags ist,
dass ethische Reflektion zusammen mit der interdisziplina
¨ren Diskussion stat-
tzufinden hat, und zwar aus zwei Gru
¨nden: Erstens korrelieren die technischen,
o
¨konomischen, rechtlichen und sozialen Aspekte stark mit der ethischen
Reflektion. Zweitens ermo
¨glicht die Teilnahme an solchen interdisziplina
¨ren
Diskussionen eine ethische Reflektion, die die reale Welt nicht aus den Augen
verliert. Als Beispiele fu
¨r Problemstellungen dienen zwei Fallstudien u
¨ber Ro-
botikanwendungen im Gesundheitswesen, in denen es gelingt, im Rahmen einer
interdisziplina
¨ren TA Argumentationsketten u
¨ber die Grenzen von Disziplinen
hinweg zu entwickelt.
Re
´sume
´L’e
´valuation technologique (ET) est une entreprise axe
´e sur les
proble
`mes, a
`savoir les proble
`mes politiques, socie
´taux, e
´cologiques, etc. Les
cas sont rares dans lesquels une seule discipline scientifique suffit a
`e
´valuer ces
proble
`mes. En ge
´ne
´ral, les perspectives provenant de diffe
´rentes disciplines
scientifiques doivent eˆ tre combine
´es afin de de
´velopper pour l’action des
recommandations a
`base interdisciplinaire. Cet article de
´crit un processus de
discussion interdisciplinaire soumis a
`un controˆ le de qualite
´, qui encourage un
groupe d’experts a
`constituer des chaıˆ nes d’arguments franchisant les limites
des disciplines. Le roˆ le de la re
´flexion e
´thique dans cette de
´marche de
´pend de
la situation proble
´matique. De
`s qu’une application technique ne pouvant eˆ tre
affecte
´ea
`la cate
´gorie dite des « cas courants » est a
`l’ordre du jour, une
re
´flexion e
´thique est demande
´e. La pre
´sente contribution pose que cette
re
´flexion e
´thique doit avoir lieu en meˆ me temps que la discussion interdisci-
plinaire, et ceci pour deux raisons : d’une part, les aspects techniques, e
´co-
nomiques, juridiques et sociaux sont e
´troitement lie
´sa
`la re
´flexion e
´thique.
D’autre part, la participation a
`de telles discussions interdisciplinaires permet
une re
´flexion e
´thique demeurant en prise avec le monde re
´el. Les deux e
´tudes
de cas portant sur des applications robotiques dans le domaine de la sante
´
illustrent des proble
´matiques dans lesquelles une ET interdisciplinaire parvient
a
`ge
´ne
´rer des chaıˆ nes d’argumentation au-dela
`des frontie
`res de diffe
´rentes
disciplines.
1 Introduction
New technologies influence our every day life. Once introduced to the real
world, one can identify those people who profit from them, those who are not at
140 M. Decker
all or less affected, and those who have disadvantages.1Therefore decisions
about new technologies contain decisions about who should win and who
should loose. Conflicts are foreseeable. Political decision makers are asking for
solutions acceptable to the voters. Technology Assessment (TA)2contributes to
problem solving by evaluating different options, sorting out various arguments,
etc. Thereby, TA is answering questions like ‘‘in which future society do we want
to live?’’ or more normatively ‘‘in which future society should we want to live?’’.
Societal problems in connection with new technologies need, in general, an
interdisciplinary TA-approach, because only in rare cases is one individual
scientific discipline able to develop the problem solutions alone. Mostly tech-
nical, economical, ecological, legal, ethical and other aspects have to be taken
into account. Moreover, the development of problem solutions needs intense
cross-correlations between the scientific disciplines, because the argumentation
chains3leading to acceptable solutions are interdisciplinary themselves, i.e., they
are a combination of many disciplinary (part-)arguments. In this contribution,
this will be demonstrated on the case study of autonomous robots in health
care.4
Robots have to possess a number of basic skills in order to be capable of
performing actions in the world. One of these skills is the capability of loco-
motion, which is often achieved by means of wheels. Alternatively, ‘‘natural’’
ways of movement are copied through the construction of legs, wings, scales,
fins and the like. The possibility of perception is implemented with cameras and
other sensors, which provide data for modelling the environment. The third
important aspect is the capability of learning. The learning of movements,
interpreting the model of the world and of reflection, e.g., with regards to the
relation of the robot to its environment, are thereby considered to be the most
important areas. Robots which possess these capabilities of acting are called
‘‘autonomous robots’’.5
Especially in the field of health care many robot applications are developed
and some have already reached maturity of market. A problem-oriented health
technology assessment must first deal with technical aspects. Are the robots able
to do what they are supposed to do? Usually the developer of a robot defines the
technical criteria the robot has to achieve along with a health care expert; the
latter will also be involved in checking the capabilities of the prototype. In
addition, economical aspects have to be taken into account. Is the overall per-
formance of the robot cheaper than comparable methods used up to now? This
is typically done by an extensively cost-benefit analysis. Moreover legal aspects
have to be considered due to liability aspects. Who is responsible for malfunc-
tions of the robot? This becomes relevant especially if the robot is equipped with
1A new city highway for example is connected with advantages for those who live in the
countryside and work in the city. People who live next to the new highway have disadvantages
because they are exposed to the pollutants and to the noise and their plot of land looses value.
Citizens who live in the opposite part of the city are less affected.
2A definition of TA in this sense, i.e., focusing on social relevance, can be found in Grunwald
(2002).
3These argumentation chains are mainly if-then clauses.
4This example refers to the research project ‘‘Robotics. Options for the Substitutability of
Humans’’ organised from The Europa
¨ische Akademie GmbH (Christaller et al. 2001).
5Within the robotics community (cf. the problematic with the notion ‘‘autonomy’’ in the
‘‘introductory step’’).
The role of ethics in interdisciplinary technology assessment 141
a learning algorithm. Finally, ethical aspects have to be taken into account,
because treatment and nursing of patients, elderly people, handicapped etc. is
deeply rooted in our social behaviour. Therefore one has to ask in which areas
we, as a modern society, should replace6previous human completed tasks by
robots and in which areas we just do not want a robotic act instead of a human
being.
In the following, the above sketched need for interdisciplinary technology
assessment will be underlined by referring to two detailed examples from health
care. Afterwards a proposal is made on how to organise a quality controlled
interdisciplinary expert discussion in order to reach one common (interdisci-
plinary) statement of science. Finally, the role of ethics within the game of the
scientific disciplines will be described.
2 Two examples7
The first case study is ROBODOC, developed from ISS (Integrated Surgical
System, Inc.) (Fig. 1). The ROBODOC Surgical Assistant System is intended for
Fig. 1 Robodoc from ISS (source:http://www.robodoc.com)
6By using the notion ‘‘replacement’’ it is assumed that the task to be fulfilled by the robot has
up to now been taken over by humans.
7The aim of these examples is to demonstrate that there are questions to be answered from
different scientific disciplines. The questioning is not meant to be complete in the sense of the
below mentioned work programme.
142 M. Decker
use on patients requiring primary cementless total hip replacement surgery. The
ROBODOC System requires the use of the ORTHODOC Preoperative Planning
Workstation. A computed tomography (CT) scan of the patient’s femur is
loaded from a tape into a workstation allowing the surgeon to position a 3D
representation of a selected prosthesis within a femoral image created from the
CT data. After the surgeon has selected a prosthesis and designated its location,
the coordinates of the prosthesis relative to landmarks on the femur are calcu-
lated. Three titanium pins, implanted prior to the CT scan, serve as the land-
marks for spatial orientation. Prior to hip replacement surgery, data from the
preoperative planning session are transferred to the ROBODOC Surgical
Assistant. The ROBODOC Surgical Assistant consists of a robotic arm with a
distal high-speed milling bur and is controlled by a computer containing
ROBODOC control software. The robotic device features a base with wheels,
which is positioned next to the operating table. It has a bone clamp fixation
device that prevents movement of the femur. After incision, exposure and
removal of the femoral head and fixation of the femur, the device mills the
femoral canal to create a cavity of the appropriate size and shape for the selected
femoral prosthesis. The device offers the ability to complete preoperative
planning of the femoral implant procedure including surgeon-controlled selec-
tion of prosthesis size, shape, and placement and controlled milling of the femur
in preparation for a prosthesis, replacing (compare footnote 6) the use of a
manual surgical broach and hand-held milling tools.8
Concerning technical aspects, one would check if the above mentioned skills
are achievable with ROBODOC and also compare the overall capability with
the so-called standard treatment, i.e., with a surgeon who mills the hole by hand.
The criteria to scrutinise the technical performance may include the following
questions. How good is the contact between the surfaces of the bone and the
prosthesis? How exactly was it possible to reach the planned optimal position of
the prosthesis? Is the patient treated with care during the operation process? Is
the new treatment more risky than the standard treatment? An economical
perspective includes a cost-benefit analysis. Are hip replacement surgeries with
ROBODOC cheaper than those without robot assistance? How does the surgery
duration differ? Is the same amount of human assistance (assistant doctors,
operating room nurse, etc.) required? Is a longer durability of the artificial hip
achievable? From a legal point of view, primarily liability aspects are relevant.
The surgery is planned by the surgeon but the execution is done by the robot
equipped with the control system of the robot producer. Is there a problem with
shared responsibilities? Legal aspects could also refer to emergency perfor-
mances of the robot. How is the robot managed in the case of a power failure,
disruption of the data line, etc? From an ethical perspective, the question has to
be answered whether the introduction of a semiautonomous robot into an
operation room opens a new category of technology or if it is still within the
frame of ‘‘business-as-usual’’9or ‘‘is the relation between physician and patient
varied in a virulent way by the introduction of the robot’’?
8This description can be found at http://www.orthopaedics.northwestern.edu/orthopaedics/
research/robodoc.htm.
9For the description of the ‘‘business-as-usual’’ case, see the paper of Armin Grunwald in this
volume.
The role of ethics in interdisciplinary technology assessment 143
The second case study is fictive but based on Care-O-bot (Hans et al. 2002), a
prototype of a multi-functional robot assistant for housekeeping and home care
to be used by elderly or handicapped people in order to live independently in
their homes for a longer time. Therefore, an easy, intuitive, and dependable
operation of the home care system was required. Care-O-bot is able to cope with
many different situations and fulfils complex tasks even in dynamic environ-
ments. Furthermore, the robotic assistant is able to execute not only single tasks
at a time but several tasks concurrently. The further developed Care-O-bot II is
equipped with a manipulator to perform household tasks such as fetch-and-
carry duties, setting the table or basic cleaning (Fig. 2). Moreover it is a mobility
aid which enables the patient to move behind the robot, thus keeping the patient
grounded. In addition, Care-O-bot is a communication tool which includes a
camera system and a video phone. In order to create our case study, we imag-
ined a combination with a patient supervisory tool, e.g., a ring to wear on a
finger which detects medical data like blood pressure, pulse frequency, oxygen
concentration in the blood etc. These data can be transferred automatically to
the next hospital or the family doctor. The robot can be equipped with a
learning algorithm, which enables him to manoeuvre in an unknown environ-
ment. This learning algorithm could also include an adaptive human-machine
Fig. 2 Care-O-bot II (source: http://www.care-o-bot.de)
144 M. Decker
interface in order to observe and learn about the individual behaviour of ‘‘its’’
patient.
In addition to these technical criteria, economical aspects must be taken into
account. Due to the fact that nursing in recuperation houses is extremely
expensive, even an expensive robot borrowed or leased via health insurance
would have the potential to save money for the health care system.10 Taking into
account the wish of many elderly people to stay as long as possible within their
social setting, one could expect a win-win situation. The legal perspective of this
scenario refers mainly to liability aspects. Who is reliable should malfunction of
the robot occur? Due to the learning algorithm there might be a conflict between
robot producer and robot owner. The robot producer builds a learning algo-
rithm which defines several corridors of behaviour of the robot while the robot
owner, on the other hand, trains the robot according to its own needs thus
making the robot’s actions unpredictable to the robot producer. He would reject
liability for malfunction due to wrong training. On the other hand, the robot
owner is not an expert in robotics. He also does not want to be responsible for
malfunctions of the robot. Ethical reflection would focus on the nursing scenario
in general. Up until now, it is unusual to transfer nursing tasks to robots at all.
Nursing is a task typically fulfilled by human beings. It remains to be seen if a
modern society should and wants to hand over some nursing tasks to robots.
3 The project group concept
These case studies underlined that we are dealing with an interdisciplinary
problem which has to be solved by taking into account all relevant scientific
disciplines. The so-called project group concept is a well established tool that
was developed at the Europa
¨ische Akademie Bad Neuenahr-Ahrweiler GmbH
to organise interdisciplinary research (Decker and Grunwald 2001). The orga-
nisation and completion of a new project follows a procedure which is divided
into several phases, within each of which several steps can be distinguished. The
first phase is called the pre-project phase.11 Here, the first step is the preparation
of the respective problem field, which began with some initial questions, and
which, typically, is rooted outside science. Each new problem or question has to
be prepared in such a way that it can be approached by an interdisciplinary
expert group. In general, the preparation of the problem is accompanied by the
selection of the relevant scientific disciplines, followed by the selection of indi-
vidual scientists who can represent these disciplines in the discussion of the
putative project group.
Now that the groundwork is prepared, it is ready to enter the project phase,
which is also structured in three steps. Starting with a disciplinary stock-taking
(introductory step), the participants develop a common basis for discussion,
which serves to prepare an interdisciplinary perspective on the problem in
question (analytical step). Finally, at the end of the project, the project group
10 Personal communication with Professor Dr. Karl Lauterbach, Institute of Health Economics
and Clinical Epidemiology, University of Cologne.
11 The pre-project of the robotics project is published in Decker (1997).
The role of ethics in interdisciplinary technology assessment 145
should come up with concrete, joint recommendations for achieving goals
(recommendation step.)12
This procedure is accompanied by several evaluation processes. In the fol-
lowing, the particular steps of the pre-project and the project phase are pre-
sented and commented on in more detail.
3.1 Specification of questioning and problem definition
Typically TA-questionings start in situations which can be described as ‘‘tech-
nology push’’ or ‘‘technology demand’’ type. Technology push means that a new
technical development has reached the prototype level and now the different
contexts of use are discussed. ROBODOC can be described in this sense where
achieving precise positioning without ‘‘getting tired’’ was the basis for the suc-
cess of industrial robots. ROBODOC transfers the precision of a CNC-milling
machine into the new context of an operation room. Technology demand for-
mulates a societal problem which might be solved by new technologies. The
increasing average age of the population of developed countries is such a
problem. A huge demand of geriatric nursing will appear accompanied by the
question of who covers the costs for these services13 (Loerzer 2003). Therefore,
Care-O-bot can be seen in this context.
In both cases the problem definition as a starting point for interdisciplinary
research structures the field around the new technique. Decisions regarding
relevance must be made referring to well founded justification, e.g., as a basis
upon which to decide which scientific disciplines are viewed as relevant and
necessary for solving the respective problem. By means of this specification, a
kind of transformation to a scientific formulation of the initial question takes
place. In the context of expert discussions, this can be seen as developing an
interdisciplinary work program. Practically, these first steps are developed and
worked out by one person who preferably should have some experience with
interdisciplinary research and with problems of technology assessment.14
3.2 Selection of the scientific disciplines
Organizers of interdisciplinary projects are, in general, confronted with the
following dilemma. On one hand, all scientific disciplines found to be relevant
should participate so as to enable a more or less fully comprehensive research
project. This is seen as crucial to the legitimating of the results. On the other
hand, interdisciplinary research needs an intensive debate between the partici-
pating disciplines if it is to go beyond the level of pure multidisciplinary re-
search. Such an intensive debate requires that the size of the respective group of
experts be limited. A successful procedure must take this predicament into
account.
12 A chart of these phases and their accompanying evaluation loops can be found in (Decker
and Grunwald 2001).
13 See the contribution of Leo Hennen in this volume.
14 At the Europa
¨ische Akademie this person is typically a member of the scientific staff who
would also follow through with the entire project phase as the project manager.
146 M. Decker
The following guidelines are suggested to resolve the dilemma. In general, a
problem does not require the same expenditure of energy from all scientific
disciplines found to be relevant. This becomes obvious when also sub-disciplines
of the particular ‘‘university faculties’’ are considered. One can make use of this
observation by introducing different degrees of participation in the procedure
according to the different degrees of relevance of the scientific disciplines. These
degrees of participation can be, e.g. ‘‘full membership of the project group’’ or
‘‘written contributions based on additional study’’ or ‘‘attendance of a selected
and limited number of meetings in order to give evaluative remarks’’.
Full members constitute the actual project group.In addition to the project
manager they participate in all meetings and are authors of the final report.15
Furthermore, the project group can call for additional studies from competence
fields which are not represented in the group, and it can invite guest experts if
there is a need for discussion with particular sub-disciplines.
Which disciplines participate to which degree has to be justified in relation to
the initial question and work program. This justification is an ex ante perspec-
tive on the state of knowledge at the end of the pre-project. However, during the
project itself it is possible that the focus of the discussion changes.16 This can be
taken into account by planning the project in a flexible way. On one hand, this
may result in the necessity to call an additional expert (or an additional
sub-discipline) into the group as a full member even after the start of the actual
project phase. On the other hand, it could become evident that a discipline
evaluated as relevant in the beginning is losing significance for the problem. This
would have to be reflected in the course of the project and in the final report. All
changes of the initial work program must be thoroughly justified.
On the basis of these considerations it is possible to list all the scientific
disciplines found to be relevant along with their putative degree of participation
in the project group. Now the question is: which criteria should be taken into
account in selecting particular experts from these disciplines?
3.3 Selection of the experts
The quality of interdisciplinary research depends first of all on the quality of the
disciplinary contributions. Thus, it makes sense to refer to the quality standards
of the respective scientific disciplines, which are usually well established and
accepted. Searching within the different scientific communities, experts must be
called upon who are able to represent their scientific discipline as a whole in an
interdisciplinary discussion. This means in particular that one has to make sure
that the respective scientists are competent in regard to the initial problem and,
moreover, that they are able to explain the different streams of their disciplines
in an objective way to the representatives of the other scientific disciplines. In the
course of such exchanges, each expert will, hopefully, also reveal the underlying
assumptions of her or his discipline. These assumptions are typically not dis-
cussed inside disciplinary boundaries, and usually they do not have to be. In
interdisciplinary contexts, however, these assumptions can gain crucial impor-
15 Striving for the goal ‘‘common authorship of all full members of the project group’’ is a
fundamental aspect of the project group concept.
16 Such a change of focus may result from one of the evaluation processes mentioned below.
The role of ethics in interdisciplinary technology assessment 147
tance. Moreover, the experts should be able to reflect on the preferences of
stake-holders connected with their respective disciplines. For example, a medical
expert should report about patients’ interests17 and an expert for jurisprudence
should reflect on necessary safety regulations for users of newly developed
products.
In addition, and also of primary importance, the respective experts must be
generally prepared to participate in an interdisciplinary discussion in the first
place. Typically, so-called ‘‘prominent’’ experts of scientific disciplines already
feel responsible for answering questions from ‘‘outside’’. But the intensive
procedural discussion intended here means that one has to agree to a pre-
discursive consent18 concerning the rules of discussion (Gethmann 1979). Every
disciplinary statement must be justified in an acceptable manner, i.e., acceptable
for all other disciplines. Moreover, every participant has to be prepared to
identify and give feedback on those deficits of his or her ‘‘own’’ discipline, which
may have only appeared due to the new interdisciplinary perspective of the
problem. In summary, all participating scientists must be able and willing to
take part in a constructive, critical discussion with the goal to develop, in a
concerted undertaking, tangible, practical recommendations for action. This
requires an intensive collaboration over a longer period of time, and not every
scientist is willing and able to devote so much effort in such a time consuming
and sometimes strenuous discussion process.
Once having identified some candidates among the respective scientific com-
munity as potential project members, it is advisable to select two to three experts
from each sub-discipline by the process described above. Out of these only one19
person will be finally invited to participate as a full member of the expert group.
This is not only for organisational reasons but also to avoid dominance of a
particular scientific discipline.
The project phase begins by convening the experts. Initially, each expert is
required to make his or her arguments clear to the experts of the other scientific
disciplines. Thereby, it will be clarified with a view to ‘‘interdisciplinary plau-
sibility’’ how much weight individual disciplinary perspectives should be given in
regard to the respective problem at hand. This procedure usually prevents a
disciplinary bias or narrowing.20
3.4 The project phase in rational TA
Each scientist is an expert only in relation to her or his own scientific discipline.
Therefore, an interdisciplinary process has to start with the identification of
basic notions found to be relevant for the project (introductory step), followed
by a description of disciplinary definitions of these notions. Thereby, underlying
17 This advocacy is not claimed to be complete, but one can assume that there are at least some
ideas about the stakes of clients, and a discussion about that could give rise to invite a repre-
sentative from a particular stake-holder group to participate in a meeting of the project group.
18 Grunwald introduces the notion ‘predeliberative’ agreement as an advancement (Grunwald
2000a; and his contribution in this volume).
19 The remaining experts could be candidates invited to the evaluation meetings.
20 Of course this depends on the personality traits of the participating individuals. Unfortu-
nately, it is almost impossible to include this in the selection process of potential participants.
148 M. Decker
disciplinary assumptions must be revealed since, as explained above, it is only
within each particular scientific discipline that taking certain assumptions for
granted might be considered unproblematic in the sense that there is ‘‘no need
for justified arguments’’. The indispensability of this process becomes particu-
larly obvious if identical words are used to denote different concepts in different
sub-disciplines. Likewise, a careful process for introducing a notion is required if
the group decides to use or to re-define a respective notion in a particular way
for the purpose of the interdisciplinary project.21
During the next step of the project phase the disciplinary perspectives of the
initial problem will be scrutinised. Due to the fact that this criticism is coming
from outside the particular scientific discipline, the disciplinary perspective will
be adapted to the interdisciplinary requirements, whereby the criteria for
‘‘successful adaptation’’ are derived from the initial question or problem,
respectively.
This analytical step can be best characterised as mainly descriptive stock
taking. Nevertheless, it is already at this point that the development of trans-
parent argumentation chains begins to emerge, i.e., chains of argumentation,
upon which all participants of different disciplinary backgrounds can agree.
During the recommendation step of the project phase, these argumentation
chains have to be extended into the area of pragmatic consequences. The goal is
the formulation of concrete recommendations for action that always include
normative aspects. This means that, in practice, on the basis of the argumen-
tation chains developed, those chosen have to be of particular relevance for the
problem at hand. This selection should be founded on a consensus between the
members of the expert group. In this phase again the transparent presentation of
the arguments is of crucial importance.
3.5 Evaluation loops
Quality control in interdisciplinary research is of crucial importance. The criteria
for distinguishing between good and ‘less good’ interdisciplinary research must be
developed in reference to the initial problem and the evaluation of whether the
recommended solution to the problem is practicable. To enhance the chances for a,
in this sense, successful project, the interdisciplinary procedure includes several
evaluation loops. All evaluation processes are organised in a multidisciplinary
manner, which should enable an evaluation of both disciplinary research and
‘‘additional’’ interdisciplinary aspects. In the following, the evaluation loops
during the distinct steps of the procedure are described in detail.
3.5.1 The project proposal
In the project proposal, the initial problem is described to those whose problem
the project wants to solve. Here, it has to be justified why it seems to be a
21 ‘‘Autonomy’’ was, for example, a notion within the robotics project which is a well defined
concept in robotics, philosophy and jurisprudence, but in completely different senses. There-
fore, a common definition of the notion ‘‘autonomy’’ was one of the first tasks on the project
group’s agenda.
The role of ethics in interdisciplinary technology assessment 149
promising endeavour to solve this problem by an interdisciplinary expert group
discussion.22 During this step it becomes possible to identify the scientific dis-
ciplines that appear to be essential for solving the problem. In technology
assessment, for example, it is typically possible to identify a scientific discipline
as being technically competent. Referring to a project about the replaceability of
human beings by robots would be robotics and artificial intelligence research.23
These experts of the respective essential scientific disciplines are called into a
‘‘core group’’. Now, they begin their work by evaluating the proposal itself,
checking it for feasibility, and transforming it into a work program. This work
program will identify specifically all relevant scientific disciplines and determine
their degree of participation.
3.5.2 Work program
The work program is the basis of the interdisciplinary project. Therefore an
interdisciplinary expert commission24 evaluates its chances of successfully
solving the initial problem. A central feature of the evaluation process focuses
on scrutinising the decisions concerning the relevance of the scientific disci-
plines to participate and the selection of the particular experts. Feasibility
criteria are also taken into account, e.g., is the transition from initial multidis-
ciplinary perspectives to an interdisciplinary whole feasible during the planned
project phase? Is it possible to identify milestones which can be expressed in
terms of a progression of interim results which may influence future decisions
concerning the continuation of the project? Are there any assumptions con-
nected with the initial social problem, which have to be examined concerning
their validity?
3.5.3 The start of the project
In this phase the experts participating in the project group are asked to present
their initial statements, whereby they should include the relevant major para-
digms of their sub-disciplines. Since, however, every expert has her or his per-
sonal perspective on his or her scientific discipline, it is advisable to invite
additional experts from the same discipline to give statements from their indi-
vidual points of view. If it were possible during the pre-project phase to identify
conflicting perspectives inside disciplinary boundaries, one would make sure that
these conflicts are presented at the beginning of the project. In this way, the
project group member is confronted with an alternative perspective, whereas the
other participants of the group have the opportunity to get a second opinion.
Instead of inviting experts out of relevant disciplines one after another, it is
suggested that such a hearing of external experts, which broadens the basis of
the discussion right from the beginning, should be completed in a so-called kick-
off meeting. Such a meeting gives the additional opportunity for the disciplinary
22 In some cases it might be more promising to go for a so-called participatory TA-method,
which includes stake-holders, citizens, etc.
23 Concerning a project on ‘‘xenotransplantation’’, for example, one would definitely need a
surgeon, a virologist and an immunologist.
24 At the Europa
¨ische Akademie this evaluation step is done by its scientific council.
150 M. Decker
statements to not only be presented to, and evaluated by, the members of the
project group but also to and by external experts of other disciplines.25
Therefore, the kick-off meeting is an evaluation process in itself in the sense
that external experts put their statements on record and these statements have to
be considered by the project group — either by taking them into their own
argumentation or by rejecting them explicitly on valid grounds.26
3.5.4 Work in progress: internal aspects
The next step is that every member of the project group prepares a paper which,
although originating from the respective disciplinary perspective, aims at the
inclusion of the relevant inputs from other experts. In other words, it is the guiding
principle for each author to consider the results of the ongoing interdisciplinary
discussion in the formulations of their own papers. In this way an interdisciplinary
approach is ensured. Ideally, this approach culminates in a report, which is agreed
upon by all members of the group, and which, therefore, is published under joint
authorship. The process described in this section can very well be viewed as an
internal evaluation process of the disciplinary perspectives presented.
3.5.5 Work in progress: external aspects
The work in progress has to be evaluated for two different purposes. Firstly, the
results reached so far must be scrutinised. Has the goal been reached to combine
interdisciplinary argumentation chains? If so, are they presented in a transparent
manner? Secondly, evaluation steps might draw attention to new aspects which
have so far not been considered. For the reasons already given in favour of the
kick-off meeting, we suggest that, after the first versions of a report have been
presented, a second meeting with external experts be convened. This meeting is
called a midterm meeting. Two goals should be pursued through such a meeting.
On the one hand, the midterm meeting should allow an evaluation of disci-
plinary results by inviting the relevant external experts. On the other hand, these
external experts should assess the interdisciplinary results as well. Especially for
the latter purpose, a briefing of the experts will be necessary, which means that
the written materials (papers for the report) are made available to them in
advance. Furthermore, they have to be informed about the initial societal
problem, the transformation processes during the pre-project, the means chosen,
and the ends to be reached. Therefore, the preparation of an interdisciplinary
evaluation takes significantly more time than for pure disciplinary evaluations.
It is expected from members of the project group that they use the results of
the midterm meeting for further elaboration of their reports, that is, they should
consider all inputs and criticism from the external experts, either by completing
their own argument or by an explicit rejection.
25 Examples of statements presented during the kick-off meeting of the robotics project can be
found in Decker (1999).
26 At the end of the kick-off meeting the work program of the project and thus the initial
problem definition has been scrutinised three times. Firstly by the so-called ‘‘core-group’’,
secondly and thirdly by external experts. This is due to the high importance of this phase for the
whole project; wrong decisions during this ‘‘setting the course’’ process are difficult to correct
during the project phase.
The role of ethics in interdisciplinary technology assessment 151
3.6 Final report
After the midterm meeting, the next goal of the project group was to work out
feasible recommendations for action, based on the argumentation chains
developed so far. The final report is compiled along these lines. To emphasise
once again, it is of crucial importance that, besides the recommendations, the
complete chains of arguments are presented in a transparent manner and
without gaps. Thereby, the arguments of the different scientific disciplines have
to be woven into an interdisciplinary whole.27 Only then is the goal of the
project group reached. Whether this undertaking has been successful will be
evaluated again by an external multidisciplinary expert group.28
In summary, the project group concept can be described as a step-by-step
interdisciplinary discussion process in which each step is accompanied by
external evaluation processes. The crucial importance of the problem defini-
tion for the whole process is considered in a threefold evaluation. The eval-
uation processes focus on both the disciplinary foundations and the
interdisciplinary cross-correlations. They are therefore promising tools to
overcome the so-called expert dilemmas (Nennen and Garbe 1996), which
refer to the fact that expert statements sometimes consider a too narrow
perspective leading to counter expert opinions, or the experts express their
personal (political, economical) preferences within their scientific statements.
This quality controlled interdisciplinary discussion process results in final
recommendations which can be taken as ‘‘science’s voice’’ within the societal
discussion process.
4 The role of ethics in the concert of the scientific
disciplines
In order to describe what a professional ethicist is doing in such an interdisci-
plinary discussion process, it is sensible to distinguish between morals and
ethics.29 Morals, or more precisely moral rules, can be found in all social
groups30 and are therefore factual, existing and accessible by empirical research.
Ethics can be described as the professional reflection of these different morals,
especially if they are conflicting, because as long as different morals exist in
parallel without any potential conflict, an ethical reflection is not at all neces-
sary. But conflicting morals lead, in the above mentioned sense, to different ends
to be reached by new technologies. Ethicists start a fictive discourse (Gethmann
1991), thereby taking into account the arguments of the factual existing morals.
Ethical reflection is a comparison and scrutinisation of the moral arguments and
their underlying assumptions. This is what an ethicist is trained to do. It might
27 Christaller et al. (2001) have an example for such a final report in common authorship.
28 At the Europa
¨ische Akademie, this multidisciplinary expert group is the scientific council of
the Academy.
29 cf. Armin Grunwalds contribution in this volume.
30 The Catholic or Protestant churches, trade unions or patient associations are examples of
such groups.
152 M. Decker
be necessary that the factual moral arguments of social subgroups be identified
by a moral inquiry.31 This is the case when there is some doubt about the moral
arguments in the fictive discourse. But in most cases it is not necessary to inquire
about the moral arguments of all sub-groups for each ethical decision making
because moral arguments do not change very quickly.32
The role of ethical reflection in TA in general can be justified because technical
actions nowadays are actions under uncertainties and inequality33 (Gethmann
1999a, 1999b). Therefore, decision making is confronted with problems con-
cerning the fair distribution of chances and risks and concerning actions under
uncertainties. This is because usually the ‘‘winners’’ (beneficiaries) of a new
technology are different from the potential ‘‘losers’’ (affected from unintended
consequences) (Grunwald 2002).
Technology Assessment is dealing with new technologies in a means-end
relation, i.e., a new technology is developed or should be developed to reach a
particular end. During the discussion, one compares different means referring to
their intended and unintended consequences. In addition, the scrutinising of the
end itself is of crucial importance. Concerning their role in the interdisciplinary
discussion described above, all scientific disciplines are equal. However, taking a
second look, it becomes obvious that there are some differences concerning their
inputs, e.g., ethical reflection deals mainly with the ends of technology devel-
opment. Contributions to answer questions like ‘‘which technologies could be
developed, which should be developed and which should not be developed?’’ are
made by professional ethicists.
Usually the discussion about the results should take place before the debate
on different means to reach these ends because the potential to reach the end is a
strong argument in the discussion on means. Therefore one could argue that
ethical reflection should take place at the very beginning of the TA project and it
sets the unchangeable boundaries for the interdisciplinary discussion. Later, one
‘‘only’’ discusses about the means to reach these ends. However, during the
discussion about ends it becomes obvious very soon that there are conflicting
ends in the debate. The ends to be reached by new technologies are in the focus
of criticism themselves (Grunwald 2000c). In order to come to rational deci-
sions, ethical reflection has to justify the ends to be reached by new technologies.
This endeavour needs the input of, and is deeply cross-correlated with, the other
scientific disciplines as is shown in the following examples.
Having autonomous robots assisting in the operation room seems, upon first
look, not that extraordinary. An operating room is already equipped with high-
tech inventory like a heart-lung machine. What makes a semi-autonomous robot
which mills a hole in a bone special? Questions like these have to be answered
referring to technical and legal aspects. Where are the differences between a
hand-held mill and a robot controlled mill? Technically viewed, one compares
criteria like ‘‘push foreword pressure’’, ‘‘accuracy’’, ‘‘average probability of
bone damaging’’, and ‘‘quality of the transfer of the 3D operation plan to the
surgery itself’’. The legal aspects refer to these technical aspects. They focus on
31 cf. the contribution of Rob Reuzel in this volume. The deaf community can be described as
such a social group.
32 Which is the reason why they can be used as a basis for rational decisions at all (Grunwald
2000b).
33 cf. Pedro Gallo’s contribution referring to ‘‘equity’’ in this volume.
The role of ethics in interdisciplinary technology assessment 153
the liability in case of malfunction. The high resolution 3D data of the CT are
used by the surgeon in order to plan the surgery. This plan has to be fulfilled by
the robot during the surgery. To make sure that there are no discrepancies
between the operation plan and the actual milling of the robot, both the plan
and the robot control system must be recorded. One can imagine a kind of black
box within the robot.34 If malfunction should occur, it can be identified whether
the failure originated in the planning or by the robot, i.e., by the robot producer.
In addition, a kind of emergency automatism might be necessary. This allows,
for example, an emergency stop during operation and the manual pulling back
of the robot from the patient without causing additional damage. If these as-
pects are taken into account, the use of ROBODOC can be seen as a ‘‘business-
as-usual’’ case from an ethical point of view (Grunwald 2000d). Another point
touches on the economical perspective where up to now an overall cost-benefit
analysis is not possible due to the fact that a manually milled hip prosthesis lasts
around 15 years and the overall duration of use is a crucial factor in the benefit
analysis. The costs of the surgery with robot assistance are higher and the
operation takes around 15 min longer. The hope of longer use is mainly based
on better contact between prostheses and bone, but one can also find some less
optimistic statements.35
Autonomous robots in geriatric nursing are not in our ‘‘business-as-usual’’
scenario concerning the nursing of elderly members of society. Until now this
task has been carried out by human caregivers and the introduction of robots
into the relationship between patients or elder people and the nursing staff needs
ethical reflection. It would be the best solution to provide caretaking by humans
which takes into account individual needs, the autarky and the dignity of the
aged people. However, it appeared in inquiries, that the nursing by humans is
sometimes not very ‘‘human’’ (Loerzer 2003). Moreover there is a decrease of
the secondary growth in geriatric nursing.36 But, however, this is very expensive
and there is the additional danger that relations of dependency can be misused.
Within the market of health care, the cost aspect plays an important role due to
the overall limited budget, e.g., in Germany. Within an exhausted budget, the
justification to consider a more expensive treatment than what is standard ar-
gues not only against this present standard treatment (concerning the same
disease) but against all the treatments of all other diseases. This is a typical
question of fairness of allocation. Due to which reasons should it be possible
then to reject a technical solution with an enormous health care economic
potential? This is a typical conflict on the level of ends.
Another aspect the ethicist has to consider, for example, is the weighing
between the prohibition of instrumentalisation of human beings against the gain
of increased autarky. The latter can be reached by the cooperation of a human
being and a robot which, since it is not a human being, is available all the time
and for unlimited duration. Especially younger patients and handicapped people
welcome the increased independence from other people or nursing staff. The
former takes into account that the human being might be instrumentalised
34 Establishing a black box in robots equipped with learning algorithms was one recommen-
dation given in Christaller et al. (2001).
35 ZEIT Nr.20/99, p 42 und Nr.32/99, p 29; Frankfurter Rundschau Nr. 133 12/6/99, p 6.
36 Su
¨ddeutsche Zeitung 03.01.2002.
154 M. Decker
within this cooperation. Nursing is a very intense relationship between human
beings (caregiver and patient), especially in long term nursing. Within such a
relationship, the nurse may have to guide the patient. Transferred to a human-
robot relationship, this would mean that the robot reaches a kind of control over
the human.37 This contradicts the universal agreement on the prohibition of
instrumentalisation of human beings.38
These case studies demonstrate that several scientific disciplines contribute to a
common chain of argumentation, which requires the intense interdisciplinary
discussion.39 Through this process, the ethical reflections ‘‘remain in contact’’ with
real world problems. The other scientific disciplines consider the relevance of the
ethical reflection (Grunwald 2000c). Applied ethics in TA deals with ethics of
technical action (Gethmann 1991). The detailed understanding and handling of
the technical, economical and legal implications of this technical action is thus a
necessary prerequisite.
5 Conclusion
Technology Assessment constitutes dealing with new technologies in a means-
end relation. It scrutinises both the means to reach a certain end and the end
itself. Both assessments are correlated as shown in the above mentioned
examples. Therefore they need to take place within the same discussion. It has
been demonstrated that Technology Assessment of autonomous robots in health
care is a problem which needs to be solved by an interdisciplinary discussion
including ethical reflection. The development of common recommendations for
action is based on that interdisciplinary discussion. The role of ethics within this
interdisciplinary discussion is included despite the fact that ethical reflection
takes place mainly at the end-level because the discussion about ends accom-
panies the whole interdisciplinary endeavour. Quality control of interdisciplin-
ary expert discussions has been identified as fundamental to reach
intersubjective valid results. Relevance decisions made by the project group
members contain normative aspects and must therefore be securitized referring
to both disciplinary quality and problem solving potential in the interdisci-
plinary context.
References
Christaller T, Decker M, Gilsbach J-M, Hirzinger G, Lauterbach K, Schweighofer E, Schweitzer
G, Sturma D (2001) Robotik. Perspektiven fu
¨r menschliches Handeln in der zuku
¨nftigen
Gesellschaft. Springer, Berlin Heidelberg New York
37 It is not self-evident that the human is at the top within the control hierarchy of the coop-
eration system between human and robot. Especially in the case of autopilots where many
accidents are caused by problems of handling the pilot-machine interface, one then thinks about
leaving the last decision to the computer or robot. (Bild der Wissenschaft 5/97, p 59).
38 Christaller et al. (2001) has recommended the use of robots in the context of nursing only as
tools for the human nurse and as instruments to raise patient’s autarky. One has to avoid the
reduction of the human nursing staff.
39Again, the description of the case studies is focusing on the cross-correlations among the
different scientific disciplines. For complete argument see Christaller et al. (2001)).
The role of ethics in interdisciplinary technology assessment 155
Decker M (1997) Perspektiven der Robotik. U
¨berlegungen zur Ersetzbarkeit des Menschen,
Graue Reihe Nr. 8, Europa
¨ische Akademie GmbH, Bad Neuenahr-Ahrweiler
Decker M (ed) (1999) Robotik. Einfu
¨hrung in eine interdisziplina
¨re Diskussion. Graue Reihe
Nr. 16, Europa
¨pische Akademie GmbH, Bad Neuenahr-Ahrweiler
Decker M, Grunwald A (2001) Rational technology assessment as interdisciplinary research. In:
Decker M (ed) Interdisciplinarity in technology assessment. Implementation and its chances
and limits. Springer, Berlin Heidelberg New York, p 32–60
Gethmann CF (1979) Proto-Logik. Untersuchungen zur formalen Pragmatik von Beg-
ru
¨ndungsdiskursen. Suhrkamp, Frankfurt
Gethmann CF (1991) Ethische Aspekte des Handelns unter Risiko. In: Lutz-Bachmann M (ed)
Freiheit und Verantwortung. Morus-Verlag Berlin, p 152–169
Gethmann CF (1999a) Die Rolle der Ethik in der Technikfolgenabscha
¨tzung. In: Petermann T,
Coenen R (eds) Technikfolgen-Abscha
¨tzung in Deutschland. Bilanz und Perspektiven.
Campus Frankfurt am Main
Gethmann CF (1999b) Rationale Technikfolgenbeurteilung. In: Grunwald A (ed) Rationale
Technikfolgenbeurteilung. Konzepte und methodische Grundlagen. Springer, Berlin Hei-
delberg New York, p 1–10
Grunwald A (2000a) Technik fu
¨r die Gesellschaft von morgen. Mo
¨glichkeiten und Grenzen
gesellschaftlicher Technikgestaltung. Campus, Frankfurt, New York
Grunwald A (2000b) Technology policy between long-term planning requirements and short-
ranged acceptance problems. New challenges for technology assessment. In: Grin J, Grun-
wald A (ed) Vision assessment: shaping technology in 21st century society. Springer, Berlin
Heidelberg New York
Grunwald A (2000c) Technik fu
¨r die Gesellschaft von morgen. Mo
¨glichkeiten und Grenzen
gesellschaftlicher Technikgestaltung. Campus, Frankfurt
Grunwald A (2000d) Against overestimating the role of ethics in technology. Sci Eng Ethics
6:181–196
Grunwald A (2002) Technikfolgenabscha
¨tzung – eine Einfu
¨hrung. Edition Sigma, Berlin
Hans M, Graf B, Schraft RD (2002) Robotic home assistant Care-O-bot: past—present—future.
In: Proceedings of the 11th IEEE international workshop on robot and human interactive
communication, ROMAN2002, Berlin, pp 380–385
Loerzer S (2003) Immer mehr Beschwerden. Verbesserungen in der Altenpflege scheitern am
Geld. In: http://www.Sueddeutsche.de, 16th January 2003
Nennen H-U, Garbe D (1996) Das Expertendilemma: zur Rolle wissenschaftlicher Gutachter in
der o
¨ffentlichen Meinungsbildung. Springer, Berlin Heidelberg New York
156 M. Decker