. 31
( 62 .)


table 8.2 Higher Cognition and Its Realizations

Cognitive Capacities in
Symbol-Using Creatures Realization of the Capacity
purely internal internal cognitive arrangement
of the brain
cerebral + bodily con¬guration
enactive bodily
cerebral arrangement + environmental
world involving
The Embedded Mind and Cognition 189

shall concentrate more heavily on world-involving capacities than enac-
tive bodily capacities in what follows.

4 memory
Memory has been one of the most active areas of research in psychology.
From Ebbinghaus™s initial experiments on the recall of lists of nonsense
syllables to current PET or fMRI research on the localization of particular
memory systems, the bulk of this research has re¬‚ected psychology™s ori-
gin as an experimental discipline. There is, I think, a real question about
the scope and ultimate value of the bulk of the work in this tradition, a
question that has manifested itself in recent debates over the relation-
ship that such research bears to everyday memory, memory in the wild,
memory as it is used in our day-to-day lives. Over two decades ago the
cognitive psychologist Ulric Neisser provocatively stated that

the results of a hundred years of the psychological study of memory are some-
what discouraging. We have established ¬rm empirical generalizations, but most
of them are so obvious that every ten-year-old knows them anyway. We have made
discoveries, but they are only marginally about memory; in many cases we don™t
know what to do with them, and wear them out with endless experimental vari-
ations. We have an intellectually impressive group of theories, but history offers
little con¬dence that they will provide any meaningful insight into natural behav-
ior. Of course, I could be wrong: perhaps this is the exceptional case where the
lessons of history do not apply, and the new theories will stand the test of time
better than the old ones did. . . . But because they say so little about the everyday
uses of memory, they seem ripe for the same fate that overtook learning theory
not long ago.3

For all that has happened in the past twenty-¬ve years, there is little reason
to revise this general judgment.
Part of Neisser™s point about the experimental tradition in memory
research is that the near exclusive concentration on controlled condi-
tions under which one ¬nds a signi¬cant difference between two groups
of subjects has led many investigators to forget what memory is for, what
it does for us as individuals in our day-to-day lives. “Real life” memory “
for example, on memory as it is used by an individual to construct nar-
ratives about what has happened to her, or on memory as it operates in
eyewitness testimony in the legal system “ is the remedy that Neisser him-
self sees as a way of rectifying this de¬cit in traditional memory research.
Reconceptualizing the study of memory as a form of externalist psychol-
ogy can contribute to this general project of reenvisioning memory.4
Thinking Through and Beyond the Body

Consider ¬rst the classic “storehouse” model of memory that derives
from a brief comment that the philosopher John Locke made in An Essay
Concerning Human Understanding. Locke says

. . . This is Memory, which is as it were the Store-house of our Ideas. For the narrow
Mind of Man, not being capable of having many Ideas under View and Consider-
ation at once, it was necessary to have a Repository, to lay up those Ideas, which
at another time it might have use of.

While Locke is often credited with originating the storehouse metaphor
in this passage, what is seldom noted is that having introduced it, Locke
immediately warns us about how it should be understood, with this
passage continuing:

But our Ideas being nothing, but actual Perceptions in the Mind, which cease to
be anything, when there is no perception of them, this laying up of our Ideas in the
Repository of the Memory, signi¬es no more but this, that the Mind has a Power,
in many cases, to revive Perceptions, which it has once had, with this additional
Perception annexed to them, that it has had them before. And in this Sense it is,
that our Ideas are said to be in our Memories, when indeed, they are actually no
where, but only there is an ability in the Mind, when it will, to revive them again.5

Here Locke is cautioning against reifying memory as a storehouse for
ideas, since he holds that ideas, including memories, exist only when they
are perceived consciously by the mind. Locke is also, I think, suggesting a
more dispositional conception of memory. If this is correct, then there is
a question whether this ability is, as Locke says, “in the Mind,” or whether
it is what, in Chapter 6, I called an extrinsic or wide disposition.
The psychologists Asher Koriat and Morris Goldsmith have noted that
while the storehouse metaphor has structured the bulk of memory re-
search over the last hundred years, this metaphor has been challenged
in recent years by research that draws on what they call the “correspon-
dence metaphor.” According to this view, it is of the essence of memory to
correspond to some past state of affairs, rather than simply to act as a store-
house for readily identi¬able chunks of data. The storehouse metaphor
facilitates a conception of memory as a place that can contain a ¬nite
number of individual units, and thus suggests that memory be studied
in terms of the quanti¬cation of those units. By contrast, the correspon-
dence metaphor is conducive to investigating and assessing memory in
terms of the accuracy of the contents of the memory, where this is not
simply a matter of how much is remembered.6
The correspondence metaphor of memory invokes a taxonomically
externalist conception of memory. That is, what individuates memory
The Embedded Mind and Cognition 191

so conceived in general from other types of cognitive processes (for ex-
ample, imagination, fantasy, wishes) is the relationship memory bears to
past, experienced states of affairs in the world. Moreover, there is a second
way in which the correspondence metaphor relies on a taxonomically ex-
ternalist conception of memory. What individuates particular memories
from one another, on this view, is at least in part what they are memories
of or about, that is, their intentionality. On this view, that memories are
not simply self-standing encodings but records of events and episodes in
the world is an important fact about them, not something that should be
factored out or bracketed off in investigating them.
Ulric Neisser has recently claimed that the underlying metaphor struc-
turing real-life memory research is not that of correspondence, but one
of “remembering as doing,” with emphasis given to the activity of remem-
bering, rather than the results of that activity, things called “memories.” In
this respect, Neisser™s view is interestingly like Locke™s dispositional gloss
on his storehouse metaphor. This conception of memory in the experi-
mental literature has its locus classicus in Frederic C. Bartlett™s in¬‚uential
Remembering. Bartlett criticized the view of memory as a storehouse of im-
ages or traces, arguing instead that remembering, as an activity, was essen-
tially a constructive process. With some misgivings, Bartlett adapted the
term “schema” in referring to the “active organization of past reactions.”
In tracing parallels between the role of schemata in individual memory
and conventionalization in societal remembering, Bartlett laid the foun-
dation for the development of the theory of cultural models that has been
in¬‚uential in recent anthropology and ethnology.7
This performative, enactive, or constructive model of memory opens
the way for a locationally externalist conception of memory, where what
is enacted does not simply stop at the skin but involves engaging with
the world through cognitively signi¬cant, embodied action. On this view,
internally stored memories lose the privileged role that they play on the
storehouse conception and become simply one resource that is used in
the act of remembering. What are critical are acts of remembering, and
while these use internal resources they also take place through bodily
activity that is, in turn, engaged with the world. Remembering, on this
view, involves exploiting internal, bodily, and environmental resources in
order to produce some sort of action, often social in nature.
To adapt Neisser™s own examples, to tell a joke or recite an epic tale
is not simply to make certain mouth and body movements, any more
than it is to produce a certain number of items from memory or to re-
call something accurately. Rather, it is to make a suitable impact on an
Thinking Through and Beyond the Body

audience there and then through one™s memory-driven actions. We can
conceive of such memory itself as extending into the world through the
way in which it engages with and appropriates external symbols, treating
them in just the way that internal symbols are treated, and thus giving us
a locationally externalist cognitive system.
Such a conception of memory is operative in the psychologist Merlin
Donald™s view of the evolution of human cognition. On Donald™s view, the
critical transition that makes distinctively human cognition possible is the
formation of what Donald calls the external memory ¬eld. This is constituted
chie¬‚y by visual symbols and the devices that generate them, and these
derive from human cultural achievements, such as the development of
writing systems and pictorial conventions. We have hybrid minds that
combine internal with external symbolic processing. In more recent work,
Donald has developed this view with an eye to highlighting its implications
for consciousness and its evolution.8
Following Donald, the philosopher Mark Rowlands points not only
to visuographic representations but also to the role of bodily grounded
mimesis as providing a new form of semantic memory, one that allowed
for the development of both learned bodily skills and the communica-
tion of sophisticated information about one™s self and its environment.
Rowlands also identi¬es sound, including spoken language, as an external
medium for memory, one whose patterns of repetition, such as rhyme and
rhythm, allow complicated sound patterns to be remembered. Sound and
the patterns that can be generated through it constitute external memory
resources, and provide the basis of the oral traditions that characterize
all human cultures even in the absence of permanent external vehicles
of representation, such as writing systems.9
Thus, this general idea can be developed without restricting oneself
to thinking of memory exclusively in terms of symbols. For example, we
can think of the performative memory system that extends beyond the
head of the individual as incorporating aspects of an individual™s envi-
ronment that are nonsymbolic, including the agent™s bodily orientation
and actual objects in her environment. So conceived, enactive, proce-
dural memory that is locationally wide is an extension of traditionally
conceived procedural memory. The idea that procedural memory may
involve doing things with one™s body, while itself old hat, does suggest
an idea that seems more novel: that one may remember by doing things
with one™s environment. Perhaps even this idea is old hat; after all, we all
know that we can use environmentally cued mnemonics, such as tying a
piece of string around one™s ¬nger, or leaving a note on the refrigerator.
The Embedded Mind and Cognition 193

My suggestion is that these need not simply be prompts to remember
but themselves are ways of remembering “ ways that involve a sustained,
reliable causal interaction between an organism and its environment. The
magnitude of our symbol-laden environments should be taken seriously,
and to do so is to see the mind as extending beyond itself, that is, as being
constituted by such symbols and thus as locationally wide.
To make this graphic, consider the popular problem-solving game for
children, Rush Hour. The game is played on a square board made up
of thirty-six small squares that are snuggly ¬tted by cars (of length two
squares) and trucks (of length three squares). The cars and trucks are of
various colors and can be placed either horizontally or vertically on the
board. The aim of the game is to move the cars and trucks sequentially
so that a designated car (the red car) can proceed to the sole exit on
the board. The game comes with two-sided cards. One side depicts the
way in which the cars should be set up initially (Figure 8.1, with the red
car designated “X”). The other side provides a code for completing a
series of sequential moves suf¬cient to solve that particular “rush hour”
problem (Figure 8.2). The ¬rst move, in the game depicted, is to move
car G one square to the right. And in just forty-six more simple moves
you™re home free!
The way in which most of us go about solving even a relatively sim-
ple Rush Hour problem involves a sustained perceptual and cognitive

¬gure 8.1. Rush Hour: A Problem
Thinking Through and Beyond the Body

(a) (b)
¬gure 8.2. (a) and (b): Rush Hour: A Solution

interaction with a highly structured environment. The board of ¬xed di-
mensions, the rules for the movement of the cars and trucks, and the
objective of the game all structure and constrain what we can do in play-
ing the game. But in playing it we do not simply encode all of this and
then solve the problem. (Go ahead, be my guest!) For most of us, at any
rate, that is not possible. Rather, we solve the problem by continually
looking back to the board and trying to ¬gure out sequences of moves
that will get us closer to our goal, all the time exploiting the structure
of the environment through continual interaction with it. We look, we
think, we move. But the thinking, the cognitive part of solving the prob-
lem, is not squirreled away inside us, wedged between the looking and
the moving, but developed and made possible through these interactions
with the board.
Of course, there is a second way to solve any given Rush Hour prob-
lem (apart from asking your kids to do it). Set up the board, ¬‚ip to the
“answer side” of the card, read off the code, and move each car or truck
accordingly. (This was actually my six-year-old daughter™s preferred solu-
tion, after she got the hang of the game “ and perhaps after she tired
of my fumbling around with standard ways of proceeding.) While we
might be tempted to think of this as involving a problem-solving tech-
nique that is more purely internal, note how much external structure
even it exploits. There is the labeling of the cars and trucks on the board,
the code for each move, the convention that we read from left to right,
The Embedded Mind and Cognition 195

both in each encoded move and from one instruction to the next. We
then have to put all of that together with ¬nding the car correspond-
ing to the symbol, moving it in accord with the instruction, then ¬nding
our place back on the solution card (lose your place there and you™re
The ¬rst point to make is that much of our everyday cognition is more
like the ¬rst way than the second way of solving a Rush Hour problem.
We are geared to interact cognitively with external structures. This is not
simply because of our memory and reasoning limitations. Rather, it is
because in doing so we can take advantage of both the rich environmental
structures that we ¬nd all around us and our ability to control ourselves
and adjust our relationship to those structures. The second point is that
even the second way of solving the problem, the method intuitively that
relies on encoding rather than exploitation, is still not purely internal
and symbolic in how it proceeds. With both methods the mind extends
itself beyond the purely internal capacities of the brain by engaging with,
exploiting, and manipulating parts of its structured environment.
The externalist perspective is most compelling in cases in which sys-
tems of external symbols come to change, in signi¬cant ways, the cog-
nitive capacities of individuals who interact with them. Most striking


. 31
( 62 .)