Cognitive Neuroscience 61memory are intact in amnesic patients and intact in monkeys
with medial temporal lobe lesions. For example, classical
delay conditioning of skeletal musculature depends on the
cerebellum (Thompson & Krupa, 1994), conditioning of
emotional responses depends on the amygdala (Davis, 1992;
LeDoux, 2000), and habit learning (win-stay, lose-shift re-
sponding) depends on the neostriatum (Packard, Hirsh, &
White, 1989; Salmon & Butters, 1995). Nondeclarative
memory thus refers to a variety of ways in which experience
can lead to altered dispositions, preferences, and judgments
without providing any conscious memory content.
Further work with monkeys has demonstrated that the
severity of memory impairment depends on the locus and
extent of damage within the medial temporal lobe memory
system. Damage limited to the hippocampal region causes
significant memory impairment, but damage to the adjacent
cortex increases the severity of memory impairment. It is im-
portant to note that the discovery that larger medial temporal
lobe lesions produce more severe amnesia than smaller le-
sions is compatible with the idea that structures within the
medial temporal lobe might make qualitatively different con-
tributions to memory function. This is because anatomical
projections carrying information from different parts of the
neocortex enter the medial temporal lobe memory system at
different points (Suzuki & Amaral, 1994).
Another important brain area for memory is the dien-
cephalon. However, the critical regions in the diencephalon
that when damaged produce amnesia have not at the time of
writing been identified with certainty. The important struc-
tures appear to include the mediodorsal thalamic nucleus,
the anterior nucleus, the internal medullary lamina, the
mammillo-thalamic tract, and the mammillary nuclei. Be-
cause diencephalic amnesia resembles medial temporal lobe
amnesia in many ways, these two regions together probably
form an anatomically linked, functional system.
These findings in monkeys are fully consistent with the
findings from human amnesia. Damage limited to the hip-
pocampal region is associated with moderately severe amne-
sia (Rempel-Clower, Zola, & Squire, 1996; Zola-Morgan,
Squire, Rempel, Clower, & Amarel, 1992), and more exten-
sive damage that includes the hippocampal region as well as
adjacent cortical regions is associated with more severe
memory impairment (Corkin, 1984; Mishkin, 1978; Rempel-
Clower et al., 1996; Scoville & Milner, 1957).
The same principle, that more extensive damage produces
more severe impairment, has also been established for the
hippocampus proper in the case of the rat (E. Moser, Moser,
& Andersen, 1993; M. Moser, Moser, & Forrest, 1995). The
dorsal hippocampus of the rat is essential for spatial learning
in the water maze, and progressively larger lesions of this
region produce a correspondingly larger impairment. Thus, in
all three species it has turned out that the brain is organized
such that memory is a distinct and separate cognitive func-
tion, which can be studied in isolation from perception and
other intellectual abilities. Information is still accumulating
about how memory is organized, what structures and connec-
tions are involved, and what functions they support. The dis-
ciplines of both psychology and neuroscience continue to
contribute to this enterprise.
Roger Sperry was another key player in the origins of cog-
nitive neuroscience. He received his doctorate in zoology at
the University of Chicago and then joined Lashley for a year
at Harvard and moved with Lashley to the Yerkes Primate
Laboratory at Orange Park, where he stayed for some years.
Sperry did his pioneering studies on the selective growth
of brain connections during this time (see Sperry, 1951).
Lashley was fascinated by the mind–brain issue—the brain
substrates of consciousness (although he never wrote about
it)—and often discussed this problem with his younger col-
leagues at Orange Park (Sperry, personal communication). In
more recent years, Sperry and his associates at the California
Institute of Technology tackled the issue with a series of com-
missurotomy patients—the human “split-brain” studies. This
work proved to be extraordinary, perhaps the most important
advance in the study of consciousness since the word itself
was developed many thousands of years ago (Sperry, 1968).
Another key origin of the modern field of cognitive neuro-
science is the study of humans with brain damage, as in
Milner’s work on H. M. noted earlier. Other influential scien-
tists in the development of this field were Hans-Lukas Teuber
and Brenda Milner. Karl Pribram also played a critical role.
Teuber received his early training at the University of Basel,
obtained his doctorate at Harvard, and studied with Karl
Lashley. He became chairman of the psychology department
at MIT in 1961. In the 1940s, he published an important se-
ries of papers in collaboration with Bender and others on per-
ceptual deficits following penetrating gunshot wounds of the
brain. Later he also investigated the effects of frontal lesions
on complex performance in humans.
Brenda Milner received her undergraduate training at
Cambridge; then after the war she came to Canada and stud-
ied for her PhD at McGill University under Hebb’s supervi-
sion. Hebb arranged for her to work with Wilder Penfield’s
neurosurgical patients at the Montreal Neurological Institute.
Her work on temporal lobe removal in humans, including
H. M., really began modern study of the memorial functions
of the hippocampus (see earlier). She also collaborated on
studies with Roger Sperry and with Karl Pribram.
Another very important influence in modern cognitive
neuroscience comes from the Soviet scientist Alexander