Imaging or Imagining?
Mann, J. M. 1997. Medicine and public health, ethics and hu-
man rights. Hastings Center Report 27(3): 6–13.
Racine, E., B. Ofek, and J. Illes. 2005. fMRI in the public eye.
Nature neuroscience 6:160.
United Nations. 1998. Universal declaration on the human
genome and human rights, adopted by General Assembly res-
olution 53/152, 9 December 1998, available at http://www.ohcr.
org/english/law/genome.htm
Copyright c© Taylor & Francis, Inc.
DOI: 10.1080/15265160590960276
Neuroethics: A Philosophical Challenge
Kathinka Evers, Uppsala University
Illes and Racine (2005) give a detailed and useful descrip-
tion of how developments in neuroscience and related tech-
nologies have brought about numerous ethical, social, and
legal challenges. These challenges, they argue, are akin to
those posed by modern genetics; describing, for example,
how both domains require scientific as well as socio-cultural
levels of interpretation. Their main message, however, is
that neuroethics goes beyond “genethics” in raising un-
precedented difficulties in interpretation and application.
This message is basically correct and well presented,
but there is a further theme that ought to have been treated
more thoroughly: the need for philosophical analyses of
central notions used by neuroscientists when describing
their theories and data.
Bioethics, of which neuroethics is a part, is indeed a
question of interpreting scientific data and placing these in
ethical, legal and social contexts, but it is also a question of
carrying out broad and thorough conceptual analyses of key
notions. The authors occasionally mention this need but do
not emphasize it sufficiently. The challenges posed by scien-
tific discovery are not two-fold, as Illes and Racine describe
it, but three-fold: to the scientific and the socio-cultural
levels of interpretation, a more general philosophical level
of interpretation must be added, analyzing the meaning
of neuroscientific terms and theories and their relations to
how the same or similar terms are used in other disciplines,
and in ordinary, non-scientific discourse (Illes and Racine
2005). In the case of neuroscience, this level will largely
be constituted by traditional philosophy of mind, and the
more modern neurophilosophy.
In the 1980s, following a long history of relative mutual
ignorance and even hostility (Churchland 1986; Damasio
2000, 7), neuroscience and philosophy of mind finally
merged in an interdisciplinary academic field: neurophi-
losophy (Churchland 1986, 2002). The aim of this new
discipline was to understand the mind and the brain con-
ceptually, using the methods of analytic philosophy, as well
as empirically, using neuroscientific methods, in the aim of
developing a unified theory of mind/brain. Neuroethics is
an area of neurophilosophy and needs to make use of this
framework. The neurophilosophical level of interpretation
cannot be described as an aspect of the scientific level of
interpretation. Scientists, as a rule, are not equipped by
their education to perform these conceptual analyses any
more than philosophers are taught to perform fMRIs, and
neurophilosophy is a subject in its own right.
A clear and explicit emphasis on a philosophical level of
interpretation can help avoid some serious mistakes, such
as introducing terminology with misleading connotations.
For example, the authors use the expression “brain maps” as
an equivalent to the expression “thought-maps,” which are
described as “quantitative profiles of brain function” (Illes
and Racine 2005, 2). Of course, a researcher can choose to
stipulate a technical definition by which a brain activity is
called a thought, but this is not how those words function
in ordinary language and can therefore be quite mislead-
ing. A thought is, in common discourse, a thought about
something, such as “I’ll cook fish for dinner tonight.” The
correlated brain activity is what goes on in the brain when
I think this thought. The expression “brain-map” is rea-
sonable since it sounds like what is at issue: a mapping of
brain activities. In contrast, the expression “thought map”
appears to suggest that we measure what we actually are
thinking about, namely, as if one could “read thoughts,”
which is erroneous. When we map brain activity, we do not
measure the actual contents of that activity (i.e., “thoughts”
as such), but merely traces, or correlates of them in the brain.
This distinction is important, not least in order to have a
realistic discussion of privacy issues.
In their review, Illes and Racine (2005) list vari-
ous applications of modern neurotechnology. Some are
highly interesting and useful, such as using fetal fMRI
March/April 2005, Volume 5, Number 2 ajob 31

The American Journal of Bioethics
to provide “better diagnosis of structural central nervous
system anomalies,” or “using fMRI to determine levels of
consciousness in patients in minimally conscious . . . states”
(Illes and Racine 2005, 5). Other suggested applications
appear overly simplistic and need to be reformulated con-
siderably to be meaningful; for example, the list of alleged
real possibilities for using brain maps to. . . screening passen-
gers in our busy national airports for a propensity to violence,
selecting investment managers to handle our financial portfo-
lios, and even choosing lifetime partners based on compatible
brain profiles for personality, interests and desires. (Illes and
Racine 2005, 2)
Not one of these possibilities qualifies as real. They remain
logical thought experiments until considerable philosoph-
ical and scientific ground has been covered.
Focusing the discussion on one of them, it seems highly
unlikely that brain scans could detect dangerous individuals
by measuring “propensity to violence.” To begin with, such
propensity is not by itself a problem—it can even be an
asset, depending on the context. The urge to be violent need
not be harmful. It only becomes a problem if this urge is not
under control. On the other hand, lack of emotional control
can be a problem even in the absence of “propensity to
violence.” Furthermore, an emotionally non-violent person
can commit horrendous acts in cold blood. It is not the
violence in thoughts or feelings that are socially problematic—
at least not in the context of airport safety—but the level of
violence in actual events or acts that people cause or commit,
and the connection between personality, propensities and
actual dangerousness is not so simple or easily quantifiable
as such examples suggest.
To illustrate, imagine a well-brought-up person with
a passionate temper that includes a propensity for violence
under habitual control, who is accustomed to being
respected, arriving at a U.S. airport where he or she is
requested to go through a brain-scan to prove that he
or she is not dangerous. This request, that many would
consider deeply offensive, is possibly not delivered in
overly polite terms. It requires only a small amount of
imagination to realize that a brain scan of that person
would reveal a person about to explode with anger, simply
longing to punch the person in charge of this demeaning
activity—assuming that such a thing can be revealed at
all. Shall he or she be turned away from the airport?
Compare this situation to an emotionally peaceful hit-
woman or man, entering the airport because that is the place
where her or his next victim works. “Violence” does not
enter the picture of that person’s personality, for this is just a
living, a profession, unconnected to personal feelings either
for or against the person to be killed. (Rather like a soldier
who goes to war prepared to kill to pay the college fees.)
The request for the brain-scan does not disturb her or him,
and it would reveal no notable “propensity to violence.” He
or she will enter the airport cleared by the brain-scan.
Brain scans measuring propensity to violence would
fail to detect the most dangerous people of all: the ones
that commit violence without any emotional involvement
(who may well be the ones to perform the scans). Perhaps
such scans might detect potentially violent people—but
these individuals, if they are able to remain in control, can
be extremely attractive and pleasant. Behavioral sophisti-
cation coupled with the “right” ideology are what social
and personal safety demands, and no brain scan can even
theoretically measure those qualities.
Suggesting unrealistic applications can have harmful
social consequences. In a society permeated by fear, par-
ticularly, they can cause enormous damage by reinforcing
the seeing of threats where there are none whilst the real
dangers go unnoticed (Evers 2003). In the criminal justice
system, where many applications have been suggested, a
very high level of realism, conceptual clarity, and scientific
justification are necessary in order to safeguard the legal
rights of the individual.
As Illes and Racine (2005) point out, neuroscience
poses new challenges to a number of distinct disciplines.
Understanding the neural correlates of complex behavior,
such as moral judgments or rational decision making, is
of great interest to natural and social sciences, philoso-
phy of mind, and ethics. To attain such understanding,
and be able to use it well, knowledge from many aca-
demic domains is required that meet in the interdisci-
plinary field of neuroethics. Researchers in this new and
exciting field have a joint responsibility to focus on real-
istic suggestions of future applications. In order to ensure
that the ethical analyses are based on substantial scientific
ground coupled with realism and conceptual clarity, stud-
ies in neuroethics should be construed as a three-fold task
involving:
1. Scientific interpretation of theories and data
2. Philosophical interpretation of central concepts
3. Ethical interpretation of problems arising in connection
with applications and use
Ethical discussions can precede neither scientific nor philo-
sophical interpretations.

REFERENCES
Churchland, P. S. 1986. Neurophilosophy: Toward a unified sci-
ence of the mind-brain. Cambridge, MA: MIT Press.
Churchland, P. S. 2002. Brain-wise: Studies in neurophilosophy.
Cambridge, MA: MIT Press.
32 ajob March/April 2005, Volume 5, Number 2