My research investigates the processing mechanisms underlying cognitive
development, using converging evidence from behavior, computational modelling,
and cognitive neuroscience. Our lab focuses on understanding the ubiquity
of task-dependent behaviors (success in some, but not all, tasks meant
to measure the same cognitive construct) from infancy through early childhood,
and their implications for the psychological, neural, and computational
mechanisms underlying cognitive development.
One main line of research explores these issues in the context of infants'
memory for hidden objects, which looks very different depending on how
the infants are tested. For example, infants can appear wildly precocious
in the first few months of life in visual habituation paradigms, but seem
to possess an out-of-sight, out-of-mind mentality for several months longer
when tested in searching for hidden objects. Many researchers have tried
to resolve this puzzle by arguing that infants fail such tasks due to
deficits in means-ends analysis abilities rather than memory problems.
However, we have demonstrated that infants use the identical means-ends
skills to retrieve visible objects that they fail to use to retrieve hidden
objects, suggesting that their failures to retrieve hidden objects cannot
be attributed solely to means-ends deficits. Our research explores the
potential role of two other factors motivated by psychological, neural,
and computational considerations: the gradedness of representations of
hidden objects, and distinct types of representations of hidden objects.
We use multiple methodologies in this work, including 1) testing neurologically-intact
infants on marker tasks adapted from behavioral and single-cell recording
studies with non-human primates, 2) developing neural network models of
relevant brain areas and conducting lesion and recording experiments on
the models, 3) testing brain-damaged patients (in collaboration with Akira
Miyake, Mike Mozer, and Randy O'Reilly at CU Boulder), and 4) testing
rhesus macaques (in collaboration with Marc Hauser at Harvard University,
and Liz Spelke and Earl Miller at MIT).
Another line of research explores the issues of task-dependent behavior
in the context of toddlers' spatial cognition. Toddlers (and rats) appear
to reorient after becoming disoriented using geometric information about
the shape of the room, but not featural information such as the color
of the walls, even though they can use featural information for other
purposes. We explore potential mechanisms underlying such non-intuitive
behaviors. Again, we build on what is known about the relevant neural
mechanisms, and conduct our work using multiple methodologies including
behavioral studies with toddlers and neural network simulations.
Our overarching goal is to use children's task-dependent behaviors as
a window onto the mechanisms underlying cognitive development, and the
nature of the origins of our knowledge.
Munakata, Y., & Yerys, B.E. (in press). All together
now: When dissociations between knowledge and action disappear. Psychological
Munakata, Y. and Stedron, J.M. (in press). Neural
network models of cognitive development. In C.A. Nelson & M. Luciana
(Eds.), Handbook of Developmental Cognitive Neuroscience, Cambridge, MA:
Munakata, Y., Santos, L.R., Spelke, E.S., Hauser,
M.D., & O'Reilly, R.C. (2001). Visual representation in the wild:
How rhesus monkeys parse objects. Journal of Cognitive Neuroscience, 13,
O'Reilly, R.C., & Munakata, Y. (2000). Computational
Explorations in Cognitive Neuroscience: Understanding the Mind by Simulating
the Brain. Cambridge: MIT Press.
Munakata, Y. (1998). Infant perseveration and implications
for object permanence theories: A PDP Model of the A-not-B task. Developmental
Science, 1(2), 161-184.
Munakata, Y., McClelland, J.L., Johnson, M.H., &
Siegler. R.S. (1997). Rethinking infant knowledge: Toward an adaptive
process account of successes and failures in object permanence tasks.
Psychological Review, 104(4), 686-713.
Carnegie Mellon University