The American Journal of Bioethics, 7(9): 3–11, 2007
Copyright c© Taylor & Francis Group, LLC
ISSN: 1526-5161 print / 1536-0075 online
DOI: 10.1080/15265160701518466
Perspective
Rethinking Neuroethics in the Light
of the Extended Mind Thesis
Neil Levy, University of Melbourne and University of Oxford
The extended mind thesis is the claim that mental states extend beyond the skulls of the agents whose states they are. This seemingly obscure and bizarre claim has
far-reaching implications for neuroethics, I argue. In the first half of this article, I sketch the extended mind thesis and defend it against criticisms. In the second half,
I turn to its neuroethical implications. I argue that the extended mind thesis entails the falsity of the claim that interventions into the brain are especially problematic
just because they are internal interventions, but that many objections to such interventions rely, at least in part, on this claim. Further, I argue that the thesis alters the
focus of neuroethics, away from the question of whether we ought to allow interventions into the mind, and toward the question of which interventions we ought to allow
and under what conditions. The extended mind thesis dramatically expands the scope of neuroethics: because interventions into the environment of agents can count as
interventions into their minds, decisions concerning such interventions become questions for neuroethics.
Keywords: neuroethics; philosophy of mind; internalism; science; psychopharmaceuticals; authenticity
Neuroethics exists at the confluence of many disciplines:
neuroscience and ethics, most obviously, but also psychol-
ogy, cognitive science, and philosophy of mind. Doing neu-
roethics justice, therefore, sometimes requires engagement
in relatively obscure debates. In this article, I shall argue
that one such debate, the debate over the mind’s location,
has direct neuroethical relevance. Some thinkers claim that
the mind is not contained within the skull, but extends be-
yond it, into the world. If their claim, the so-called extended
mind thesis is true, then many neuroethical issues must be
rethought. The extended mind thesis dramatically expands
the scope of neuroethics, so that many apparently tangential
questions are seen to be neuroethical questions. To that ex-
tent, the thesis increases the significance of neuroethics. At
the same time, it forces us to drastically rethink our approach
to traditional neuroethical concerns. In the first half of this
article, I shall sketch the extended mind thesis and briefly
sketch a defense of it against some of its critics. I shall then
turn to the far-reaching implications of the extended mind
thesis for neuroethics.
THE EXTENDED MIND THESIS
The name extended mind thesis is the name originally given to
a radical view in the philosophy of mind by its two earliest
defenders, the prominent philosophers of mind Andy Clark
and David Chalmers (Clark & Chalmers 1998). The view has
since been defended under other names: environmentalism
(Rowlands 1999), locational externalism (Wilson 2004) and
extended cognition (Rupert 2004), but Clark and Chalmers’
term is nicely descriptive and I shall use it here. The ex-
tended mind thesis is, roughly, the view that the mind is
Received September 16, 2006; accepted February 1, 2007.
Address correspondence to Neil Levy, Centre for Applied Philosophy and Public Ethics, University of Melbourne, Parkville 3010, Australia.
E-mail: nllevy@unimelb.edu.au
not confined within the skull of individual agents, but ex-
tends into the world. The thesis bears a family resemblance
to, but is far more radical than, the view known in the phi-
losophy of mind as content externalism: the view, associ-
ated with Tyler Burge (1979) and Hilary Putnam (1975), ac-
cording to which the content of mental states is partially
determined by social institutions and by the environment.
Whereas content externalism holds that the content of a
mental state is partially determined by features external
to the head of the person whose state it is, the extended
mind thesis holds that the mental state itself—however it
gets its content—is partially located external to the head of
the person whose state it is. The extended mind thesis there-
fore denies that mental states are constituted solely by brain
states.
At first sight, this is a startling, even absurd, claim. How
can mental states be located external to the brain? The argu-
ment for the claim is essentially functionalist: some states
and processes external to the brain play the same functional
role in cognition as some internal states and processes, and
therefore should be accorded the same status. Consider non-
occurrent mental states, such as a belief that is not currently
being entertained by a subject. Most of our beliefs and mem-
ories are like this, most of the time. Philosophers often call
these dispositional beliefs, because subjects are disposed to en-
tertain them under appropriate circumstances. Now, where
are these mental states stored when they are not being en-
tertained? Neuroscientists have an answer, or at least part
of an answer, to this question. Memories are stored first in
the medial temporal system, then (when they persist suffi-
ciently long) distributed across networks in cortical regions
ajob 3

The American Journal of Bioethics
(Schacter 1996). Now, what makes it the case that a memory
is stored in a particular region? Just this: under the appro-
priate circumstances (for instance, in response to a question
about where she had lunch), the subject recalls the memory,
and that region of the cortex plays an appropriate role in
the process of recall. Spelling out what kind of role is the
appropriate one is a difficult task, but we can confidently
say some things about it. We can say that something counts
as a part of a subject’s memory if the role it plays in re-
call is causal, and if it encodes information that is activated
in recall. The cortical regions identified by neuroscientists
clearly satisfy these conditions. But these same conditions
can be—indeed, are—satisfied by regions of the world ex-
ternal to the subject’s skull. Hence, non-occurrent mental
states like memories and beliefs can be stored external to
the subject.
Consider Clark and Chalmers’ (1998) famous example
of Otto and his notebook. Otto suffers from dementia; he
therefore takes the precaution of writing down everything
he thinks he might need to recall. Otto hears that there is a
new exhibition on at the Museum of Modern Art in New
York. He takes out his notebook and flips to the appro-
priate page. There he reads that the museum is on 53rd
Street. Armed with this knowledge, Otto sets out for the
museum. Clark and Chalmers argue that since Otto’s note-
book plays the same functional role in memory retrieval
as neural states paradigmatically play in ordinary agents’
memory, we ought to count Otto’s notebook as part of his
mind. More generally, proponents of the extended mind the-
sis defend the parity principle: if something plays a role in
cognitive activity, such that, were it internal we would have
no difficulty in concluding that it was part of the mind, it
should be counted as part of the mind whether or not it is
internal.
Of course, Otto is not a typical agent. He has a brain dis-
ease, which degrades his internal memory and leads him to
substitute an external prosthesis. But extending one’s mind
beyond the head is not something that only rare or damaged
individuals do; it is not a second-best substitute for internal
mental states and capacities. Instead, it is something that all
humans do, to a greater or lesser extent. Indeed, there is a
case for saying that the capacity for extending the mind is
distinctive of human beings, not in the sense that no other
animal possesses it (there is evidence, as we shall shortly see,
that at least some other primates can extend their minds out-
side their skulls to some extent), but insofar as extending our
minds is something that we are particularly good at, that we
do by nature, and insofar as our ability to extend our minds
underlies the many cognitive achievements that we rightly
prize.
One way that extending our minds outside our skulls
expands our abilities is by freeing up internal resources
for other tasks. Storing representations outside our heads
takes cognitive load off our brains. Consider two possible
ways of representing the visual scene. One way we might
have performed this task is by constructing and storing rich
and detailed internal representations of the world around
us. We do in fact construct internal representations of the
world, but they are neither rich nor detailed. Why go to
all the trouble and expend valuable resources on construct-
ing such a model when the world is already there to serve
as its own best model (Clark 1997)? It seems to us that we
have a rich and detailed model of the world in our heads,
but this is an illusion, caused by the reliable availability of
facts about the visual scene. It seems to us that our inter-
nal representations are rich because they are constantly up-
dated by our eye movements. The human eye has a very
small area of high-resolution vision; less than 0.01% of the
entire visual field. But our eyes constantly dart about, mov-
ing this window of high resolution across the visual scene.
These movements, called saccades, are intelligent; they are
not random, but instead gather information relevant to the
tasks currently confronting the person. They are also very
fast, averaging about three movements per second. Our fre-
quent and repeated saccades allow us to inspect the world
and update our picture of it, so that it seems to us that we
have a rich representation of it. And so we do, but it is not
an internal representation.
There is a cost to offloading the job of representing the
world onto the world itself. The cost comes in the form
of change blindness, our inability to notice even dramatic
changes in the world under certain conditions (Rayner 1998;
Simons & Levin 1998). However, because these conditions
are rarely encountered outside the psychology laboratory,
these costs are more than compensated for by the bene-
fits that flow from it: the benefits of freeing cognitive re-
sources for other tasks. Great though these benefits are,
however, they are comparatively small when compared
to the benefits that flow from cultural means of extending
minds.
Consider how using external resources can expand the
range of cognitive abilities even of chimpanzees. The un-
adorned chimp brain is able to learn to categorize pairs of
objects on the basis of their similarity or differences from one
another. So, for instance, chimps can be trained to put any
pair of identical objects—two cups, say, or two bananas—
into one box, while placing any pair of dissimilar objects into
another (say one cup and one banana). But without external
aids, they cannot sort pairs of pairs by similarity or differ-
ence. Two pairs of pairs are identical just in case they share
their first-order properties: they are either both identical, or
they are both dissimilar. Otherwise, they are dissimilar. So
the pair (of pairs) apple–banana and cup–shoe is identical,
whereas the pair cup–cup and apple–banana is different.
This higher-order task is difficult even for the human brain,
but chimps can learn to accomplish it by, in effect, turning it
into a first-order task. They do this by learning to associate
tokens with the first-order pairs. For instance, they might
learn to associate a plastic triangle with a pair of objects that
are identical to one another, and a plastic square with a pair
of objects that are dissimilar. Once they have accomplished
that task, the higher-order task is simple. If you want to
know whether a pair of pairs is similar or dissimilar, sim-
ply compare the tokens associated with them: if they are
4 ajob September, Volume 7, Number 9, 2007