Western University Western University
Scholarship@Western Scholarship@Western
Brain and Mind Institute Researchers'
Publications
Brain and Mind Institute
8-2010
Neural reuse: A fundamental organizational principle of the brain Neural reuse: A fundamental organizational principle of the brain
Michael Anderson
Department of Psychology, Franklin & Marshall College, Lancaster, PA 17604, USA
,
michael.anderson@fandm.edu
Follow this and additional works at: https://ir.lib.uwo.ca/brainpub
Part of the Neurosciences Commons, and the Psychology Commons
Citation of this paper: Citation of this paper:
Anderson, Michael, "Neural reuse: A fundamental organizational principle of the brain" (2010).
Brain and
Mind Institute Researchers' Publications
. 42.
https://ir.lib.uwo.ca/brainpub/42
Neural Reuse in the Evolution
and Development of the Brain:
Evidence for Developmental
Homology?
ABSTRACT: This article lays out some of the empirical evidence for the impor-
tance of neural reuse—the reuse of existing (inherited and/or early developing)
neural circuitry for multiple behavioral purposes—in defining the overall
functional structure of the brain. We then discuss in some detail one particular
instance of such reuse: the involvement of a local neural circuit in finger aware-
ness, number representation, and other diverse functions. Finally, we consider
whether and how the notion of a developmental homology can help us understand
the relationships between the cognitive functions that develop out of shared neu-
ral supports. ß 2012 Wiley Periodicals, Inc. Dev Psychobiol 55: 42–51, 2013.
Keywords: brain imaging; plasticity; evolution
INTRODUCTION
How are neural resources deployed to support cognitive
functioning in the adult organism, and how does that
architecture come about? That is, what evolutionary
and developmental pathways does the brain follow in
acquiring its repertoire of capacities? Consider two pos-
sible options, one that has been largely identified with
the embodied/embedded school of cognitive science,
and another associated with evolutionary psychology.
A long-standing guiding principle of both embodied
cognitive science (ECS) and evolutionary psychology
(EvoPsy) is that cognition was built within a system
primarily fitted to situated action. The central nervous
system—the neocortex most definitely included—is
first and foremost a control system for an organism
whose main job is managi ng the myriad challenges
posed by its environment. ‘‘Higher’’ cognitive faculties
like language and abstract reasoning had to find their
neural niche (Dehaene, 2011) within the constraints im-
posed (and the opportunities offered) by the way exist-
ing neural resources were deployed for this purpose, in
a way mediated and guided by whatever continuing se-
lection pressure there is to maintain fast, effective and
efficient solutions to pressing environmental challenges.
Insofar as this is true, then—and this is the other guid-
ing principle shared between EvoPsy and ECS—this
phylogenetic history should have left detectabl e traces
on both brain and behavior. Where EvoPsy and ECS
part company is in their understanding of what those
traces will look like, and where to find them.
In particular, EvoPsy has adopted a methodological
focus on the challenges posed by the environment of
selection (Buss, 2005), which has in turn led many
researchers in this area to spotlight the efficiency of
individual algorithmic and heuristic solutions to those
problems. One result of this focus had been the devel-
opment of the ‘‘adaptive toolbox’’ model of mind
(Gigerenzer & Selten, 2002). Given the presumed pres-
sures on these tools of mind for immediate and efficient
operation, independent, modular neural implementa-
tions of these tools seem a sensible solution.
In contrast, ECS is especially interested in under-
standing the ways in which thinking is both influenced
and partially constituted by emotional and physical
states, bodily activity, and interactions between self,
Developmental Psychobiology
Michael L. Anderson
Marcie Penner-Wilger
Department of Psychology
Franklin & Marshall College
P.O. Box 3003, Lancaster
PA 17604-3003
E-mail: michael.anderson@fandm.edu
Manuscript Received: 26 October 2011
Manuscript Accepted: 11 May 2012
Correspondence to: M. L. Anderson
Contract grant sponsor: NSF
Contract grant number: 0803739
This material is based upon work supported by the National
Science Foundation under Grant No. BCS-1023899.
Article first published online in Wiley Online Library
(wileyonlinelibrary.com): 18 June 2012
DOI 10.1002/dev.21055 ß 2012 Wiley Periodicals, Inc.
others and environment (Ackerman, Nocera, & Bargh,
2010; Cha ndler & Schwarz, 2009; Chemero, 2009;
Kelso, 1995; Lee & Schwarz, 2010; Varela, Thompson,
& Rosch, 1990). When considering the neural supports
for cognition, this perspective naturally places greater
weight on the functional relations and interactions be-
tween neural structures than on the actions of individual
regions. Moreover, this perspective has led ECS to
focus less on the efficiency of individual solutions, and
more on overall efficiency in the use of bodily (and
environmental and social) resources for cognitive ends.
For ECS, resource constraints and efficiency consider-
ations dictate that whenever possible neural, behavioral,
and environmental resources should have been reused
and redeployed in support of any newly emerging cog-
nitive capacities. Functionally isolated and dedicated
neural modules just do not seem to make good design
sense given the importance of efficient use of available
resources, and of ongoing interactions to shaping func-
tion. For ECS, cognition is largely supported by ‘‘old
wheels, springs and pulleys only slightly altered’’ and
reconfigured to serve present purposes.
A logical place to look for evidence of such a histo-
ry is in the distribution of and relationships between
the neural circuits supporting various cognitive func-
tions. ECS predicts that neural circuits originally
evolved or developed for one purpose will be reused in
later emerging functionality. That is, rather than follow-
ing an evolutionary/developmental pathway whe rein we
develop specialized, dedicated neural hardware to meet
each new environmental/behavioral challenge, ECS
suggests that much local neural structure is conserved
but is often combined and recombined by different
organisms in different ways to achieve diverse
purposes.
Imagine a simple brain consisting of six local neural
circuits that could be combined in various ways to
support two cognitive-behavioral tasks. Figure 1 illus-
trates three logical possibilities for how the local neural
circuits could be functionally arranged to support
the tasks in question. In a modular brain, shown in
Figure 1a, local circuits 1, 2, and 3 would combine to
support one task (represented using broken gray lines),
and 2, 4, 5, 6 would work together to support the othe r
(represented with black lines). Although there might be
some neural and functional overlap between the mod-
ules (local circuit 2 active in supporting both tasks),
the neural underpinnings of different behaviors and
abilities would b e largely segregated. In contrast, if
the brain is more holistically organized, all the local
circuits might be engaged in supporting both tasks,
with the behavioral differences possi bly reflected in
such things as different oscillatory dynamics. Finally,
it could be the case that many of the local circuits are
used to support both tasks, but for each task, they coop-
erate in different patterns, with different partners. So
for instance, in Figure 1c, local circuit 1 cooperates
with local circuits 2 and 3 in the black task and with
local circuits 5 and 6 in the gray task.
If such reuse (an especially pure case of which is
illustrated in Fig. 1c) obtains in the brain, then we
should expect at least three things to be true of its func-
tional structure. First, local neural circuits should be
used and reused for diverse purposes in various task
domains. That is, in contrast to what is illustrated by
Figure 1a, local circuits should not be classically selec-
tive in the sense of responding only to a highly restrict-
ed class of stimuli or tasks. Second, we should expect
the functional differences between task domains to be
reflected less in differences in what neural real estate is
implicated in supporting the domains than in the differ-
ent patterns of interaction between many of the same
elements (in contrast to the brain illustrated in Fig. 1b).
And third, we should expect later emerging (evolving
FIGURE 1 Three logical possibilities for the functional structure of the brain.
Developmental Psychobiology Neural Reuse—Evolution and Development of Brain 43
or developing) behaviors/abilities to be supported by a
greater number of local circuits, more broadly scattered
in the brain. Th e reason is simple: the later something
emerges, the more potentially useful existing circuitry
there will be, and little reason to suppose it will be
grouped locally. There is in fact evidence for all three
of these predictions, some of which will be reco unted,
below.
Taking up the first prediction, one recent study
(Anderson & Pessoa, 2011) examined the functional di-
versity of 78 standard anatomical regions of the brain
(based on the Freesurfer atlas) by determining whether
(and how often) each was active in 1,138 experimental
tasks in 11 different BrainMap task domains: action ex-
ecution; action observation; action inhibition; attention;
audition; vision; emotion; language semantics; reason-
ing; explicit (semantic) memory; and working memory
(Fox et al., 2005). Using a diversity scale ranging from
0 (active in only a single cognitive domain) to 1 (equal-
ly active across all 11 cognitive domains), it was deter-
mined that the overall average diversity of the 78 large
anatomical regions was .70 (SD .12). The overall aver-
age diversity of cortical regions was .71 (SD .11) and
of subcortical regions was .63 (SD .17). Put differently,
the regions were active in an average of 95 tasks
spread across 9 cognitive domains. These results are
represented graphically in Figure 2 using a cool-to-hot
scale (gray indicates no information).
The analysis was also performed in a brain divided
into 1,054 neural regions. The overall average diversity
of the 574 small cortical and 21 small subcortical
regions activated by 5 or more experiments was .52
(SD .13). Those 595 regions were activated by an aver-
age of more than 10 experiments across 5 cognitive
domains. The overall average diversity of the cortical
regions was .52 (SD .13) and of the subcortical regions
was .59 (SD .12). The upshot: local neural circuits are
not highly selective, and typically contribute to multi-
ple tasks across domain boundaries.
To examine the second prediction, we performed a
functional coactivation analysis of 1,127 experimental
tasks from the dataset (Anderson, Brumbaugh, &
Suben, 2010), falling into 10 of the BrainMap task
domains (Fox et al., 2005; for this study we excluded
action inhibition, as it contained too few experiments
for this approach). In a functional connectivity analysis,
one looks to see how often regions of the brain co-
activate under various tasks conditions. If the regions
co-activate more often than would be expected given
the activation likelihood of the individual regions—that
is, if the probability of region A and region B being
active in the same task is significantly (p < .01) higher
than would be predicted from the general probability of
A being active and the general probability of B being
active—then this indicates there is a ‘‘functional con-
nection’’ between the regions.
FIGURE 2 Task diversity of brain regions. Image prepared by Josh Kinnison and Srikanth
Padmala, University of Maryland.
44 Anderson and Penner-Wilger Developmental Psychobiology
The results of such a study can be represented as a
graph. A graph is simply a set of ‘‘nodes’’ joined by
‘‘edges,’’ where the nodes and edges can represent vari-
ous aspects of a modeled system. For instance, in an
airline route map nodes are airports and edges represent
flights between them, and in a Facebook-style social
network nodes are people and edges indicate ‘‘friend-
ship.’’ In a brain functional network like that depicted
in Figure 3, the nodes represent individual brain
regions, plotted in a 3D anatomical space, and
the edges represent functional connections between
them—that is, a higher-than-expected likelihood of co-
activation during tasks in a given cognitive domain.
Looking at the data in this format, it is easy to com pare
how often a given region is active in more than one
domain, and how often it has the same neural partners
in more than one domain.
Figure 3 highlights the functional partners of left
precentral gyrus (the functional roles of which will
be discussed further below) during semantics tasks,
FIGURE 3 The fu nctional partners o f left precentral gyrus under three dif ferent task cond itions.
(a) Semantics, (b) emotion, (c) attention. Graphs rendered with Gephi http://www.gephi.org
Developmental Psychobiology Neural Reuse—Evolution and Development of Brain 45
