Neuroscience and Biobehavioral Reviews 26 (2002) 769776

www.elsevier.com/locate/neubiorev

Aging, expertise and fine motor movement

Ralf Th. Krampe*
Education Center for Lifespan Psychology, Max-Planck Institute for Human Development and Education, Lentzeallee 94,
D-14195 Berlin, Germany
Received 23 March 2002; accepted 22 August 2002

Abstract
Age-graded decrements in accuracies and maximum speed of fine motor movements observed in numerous experimental studies have
nurtured general factor explanations like the assumption of general age-related slowing of central cognitive processes. This review focuses on
two domains of investigation that yielded challenges to general factor models. First, experimental approaches aiming at the decomposition of
fine motor skills provide evidence for the dissociability of timing, sequencing, and executive control components that show differential rather
than general age-related changes. Second, studies on cognitive-motor expertise demonstrate that age-related changes in critical skill
components depend on individuals time investments into specific practice activities. It is argued that the process dissociations observed at
the behavioral level in developmental (i.e. age and expertise) studies reflect individuals long-term adaptations to internal and external
performance constraints. The outcomes of these adaptation processes are stable interindividual differences in component processes.
q 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Age-related slowing; Movement variability; Musicians; Timing; sequencing; Executive control

Numerous experimental studies in the literature point to a

deterioration of older individuals capacities to control fine
motor movements [57]. Based on the ubiquity of empirical
evidence for negative age-related changes in the domains of
both cognitive and motor functioning extant models of
cognitive aging [7,46] posit that general slowing of central,
cognitive processes is the main cause for these developmental trends. Despite its apparent simplicity the general
slowing account must be considered as a strong baseline
model against which any claim related to differential agerelated changes in processing mechanisms must be evaluated. The specific assumption in slowing models is that
slowing affects basic processing steps, the duration of which
is slowed proportionally in older relative to younger adults.
Thus, the typical finding that negative age-effects increase in
size in more difficult task conditions (age difficulty
interactions) can be accounted for by general slowing
models by assuming that more difficult tasks require a larger
number of steps relative to simpler tasks [7]. At the level of
neural processes, proponents of general slowing models
refer to the overall losses of neural connectivity or decreased
levels of neurotransmitters in the aging brain [50] as
* Tel.: 49-30-82406-423; fax: 49-30-8249939.
E-mail address: krampe@mpib-berlin.mpg.de (R.T. Krampe).

potential causes for proportional slowing of basic processing

steps. To demonstrate process specific age-related changes
over and above general slowing experimental designs
require that young and older adults capacities are assessed
across a large range of performance conditions [27,38], or
that cross-over interactions are demonstrated. Most experimental aging studies fail these criteria.
General factor accounts have also been proposed as
explanations for age-related changes in expert performance.
For example, Salthouse et al. [48] argued that the superior
level of visuo-spatial performance in older architects relative
to normal adults in the same age-group reflected preserved
differences in basic cognitive abilities. Specifically, the
argument is that related basic dispositions existed prior to the
process of expertise acquisition and they probably influenced
professional choices. Subsequently these dispositions
undergo the same age-related changes (i.e. general slowing
of related processes) as other cognitive functions with the
result that differences between experts and normal individuals at any age are merely preserved across the life span.
The theoretical perspective taken in this review is to
conceive of interindividual differences in fine motor
performance as outcomes of individuals long-term adaptations to external (i.e. task and situational demands) and

0149-7634/02/$ - see front matter q 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 1 4 9 - 7 6 3 4 ( 0 2 ) 0 0 0 6 4 - 7

770

R.Th. Krampe / Neuroscience and Biobehavioral Reviews 26 (2002) 769776

internal (i.e. mental and physical capacities) constraints.

The notion of constraints equally portrays positive opportunities (e.g. additional degrees of freedom due to acquired
coordination skills, increases in muscular strength from
childhood to adulthood) as well as negative limitations (e.g.
adverse effects of biological aging, bimanual interference)
both types of which require adaptation on the part of the
individual to maintain some level of performance. The
individual difference perspective comprises developmental
changes at an ontogenetic level as they relate to both,
normal aging and the acquisition and maintenance of
expertise [13,14,28]. The evidence compiled in this paper
demonstrates that general factor models fail to account for
differential patterns of developmental change in components of fine motor skills. Instead, it is argued that
individuals adapt to performance constraints by differentially relying on those processes that are less age-sensitive
or on those mechanisms that they managed to maintain
through deliberate efforts.

1. Age-related changes in maximum speed and accuracy

of fine motor movements
Age-related changes in the performance of fine motor
movements enjoyed much interest in occupational psychology and ergonomics already during the first decades of the
20th century. Later on, Welfords [56] seminal contribution
marked the shift in empirical research towards psychological theories of aging and motor control. The general picture
emerging from Welfords work and numerous studies since
is that older adults motor performance is markedly slower
and less accurate than that of young individuals.
Several cross-sectional studies demonstrated a decline
for the maximum repetition rate for finger tapping in
older adults, findings cited as evidence for general agerelated slowing [47]. Other frequently applied experimental paradigms in age-comparative work on motor
control were variants of Fitts [15] aiming task: in related
experiments participants performed discrete or reciprocal
arm movement to targets varying in size or distances.
Discrete aiming tasks permit a clear separation of
movement planning stages and execution phases for a
single movement; in reciprocal aiming tasks participants
perform series of successive movements between (typically) central and peripheral targets. In the latter
paradigm the focus of investigation is on online-control
and the variability of movements. The overall pattern of
results indicates that older adults take longer to plan (i.e.
initiate after a starting signal) or restructure arm movements than young adults do [2] and that their movements
tend to be less smooth, that is, more segmented [35].
Bimanual arm- and hand-movements produce larger ageeffects than unimanual tasks [52], especially if anti-phase
movements or the coordination of non-homologous limbs
are required [51].

A specific form of age-related slowing was evident in

studies of pursuit manual tracking [23] tasks. In these
studies, older adults time delays between their pursuit
movements and the input stimulus were generally larger and
their compensatory or correction movements were also
slower and more variable, especially at higher movement
tempos or frequencies. In their modeling attempts, Jagacinski and his colleagues assume a causal link between agerelated slowing and variability in aiming movements [35]
and manual pursuit tracking [23]. At the same time, these
and related studies mark a more recent shift in empirical
focus in the cognitive aging literature away from maximum
speed or accuracy towards the variability of movement
production.

2. Age-related changes in the variability of fine motor

movements
Originally, the proposal to conceptually link age-related
performance declines in fine motor movements to variability, once again, dates back to Welford [58]. He
hypothesized that the normal biological aging process
yields increasing levels of noise in the human motor system
making the resulting deterioration in the signal to noiseratio a major source of age-related performance decrements.
The most prominent theoretical concept related to the
variability of movement production is the notion of an
internal clock or central timer, a multi-purpose device that is
called upon for any time-critical information in motor
control or perception. Wing and Kristofferson [61,62]
proposed a two-level model of movement timing that
distinguishes between the levels of central timekeeping or
the generation of clock pulses and the level of peripheral
motor implementation. The central clock delineates delays
after which the peripheral motor system implements the
triggered movement with a specific effector after another
(motor) delay. According to the model, observed time
intervals (durations) between overt responses in a timed
motor task are the summed delays caused by these two
hypothetical processes. The cerebellum has been proposed
as a neural substrate of the central timing component [21].
While neuropsychological models differ with respect to the
precise role attributed to different brain areas in controlling
fine motor movements [20,39,60], there is agreement that
the central clock mechanism must be recruited by and
interact with higher-level structures during the production
of complex movements.
In typical experiments focusing on the variability of
movement timing, participants produce repetitive taps with
prespecified target intervals. Wing and Kristofferson [61,62]
developed a method to obtain separate estimates for the
variability of central timing and motor implementation
component processes based on the covariances of observable intervals that participants produce when trying to
generate series of isochronous (i.e. identical target intervals)

R.Th. Krampe / Neuroscience and Biobehavioral Reviews 26 (2002) 769776

taps. The repetitive tapping paradigm and the two-level

timing model are interesting from an age-related slowing
aspect, because mathematical explorations show that (a) any
stochastic clock mechanism produces variabilities that
increase systematically as a function of the delineated
duration, (b) slower clocks produce larger variabilities for a
given target duration than faster clocks delineating the same
interval, and (c) slower clocks produce steeper increases in
variabilities when multiple durations are compared [17,29].
Comparisons of young and older adults movement
variabilities at the same, convenient tapping tempos can
thus be informative regarding the efficiencies of central
timing processes without picking up on biomechanical or
peripheral factors that constrain maximum tapping rates.
Age-comparative studies of repetitive, timed finger
tapping that applied the Wing Kristofferson method
obtained similar measures of efficiencies for peripheral
motor implementation in young and old adults. The same
studies differed with respect to age-differences in central
clock variability: While Woodruff-Pak and Jaeger [63]
found significant age-related increases in clock variability,
other studies [10,19] found no differences between young
adults and healthy older adults who were up to 80 years old.
Interestingly, all three studies listed here used the same,
single tempo (550 ms intervals) in their tapping experiments. One reason for the inconsistency of these findings
could be the selection of older participants. Duchek et al.
carefully screened their large-scale sample for mental
health. They found that only those older participants who
had distinctive symptoms of early dementia showed
increased clock variabilities while healthy older adults
performed at the same levels as young adults. This raises the
possibility that lower-level timing processes remain intact in
normal aging but are sensitive to pathological aging and its
precursors. Furthermore, participants in the three studies
described had very limited amounts of task-related experience and their timing capacities were only assessed at a
single tempo. In the domains of memory functioning and
mental figural transformation tasks, two studies [27,38]
demonstrated that age-graded dissociations of component
processes require systematic assessments of participants
full ranges of performance after intensive practice. Experiments that further scrutinized age-related changes in lowerlevel timing processes and aimed at their dissociation from
higher-level control processes underlying fine motor movements are described in Section 3.

3. Dissociating component processes in fine motor

performance
In most musical performance, maximum speed is rarely
the dominant goal. Rather, movements are typically
generated as rhythmic patterns, which are realized according to the constraints of an overall tempo and a metrum.
From a theoretical perspective of action or motor control

771

[24] rhythm production is an instance of timing specific

intervals and controlling the serial order (sequencing) of
different intervals (target durations) within recurring
patterns. Complex forms of sequencing fine motor movements have for some time been linked to hierarchical
models of control [34] in which distinct cognitive processes
are assumed to underlie timing and sequencing capacities
[36,55].
In a recent series of experiments Krampe et al. [33]
attempted to dissociate low-level timing, higher-level
sequencing components, and executive control processes
in the production of repetitive movements. In their study,
the same young (ages 18 21) and older individuals (ages
67 73) with little musical experience participated in four
different experiments in the course of altogether 14 sessions.
Tasks comprised isochronous tapping (i.e. repetitions of the
same target interval) as well as rhythmic patterns of varying
sequence complexities (i.e. rhythms consisting of different
numbers of target intervals). The methodological approach
in this study was to systematically vary the overall
performance tempos within each task condition, such that
participants timing variabilities for identical target durations could be compared across different task contexts.
Krampe et al. [33] found that variabilities in the
isochronous tapping task increased systematically with
target duration, however, variabilities and the slope of
their increase with target duration were very similar for
young and older adults. Given that variability in isochronous taping tasks can be considered the most direct measure
for the efficiency of the internal clock, these findings
implicate a relative age-graded stability of a low-level
timing mechanism until later adulthood, at least in healthy
adults. In contrast, when the same participants produced
rhythmic patterns made up from multiple target intervals,
variabilities increased systematically with sequence complexity relative to the variabilities observed for the identical
target intervals in the isochronous tapping task. This pattern
demonstrates that sequencing of multiple target durations
relies on additional mechanisms over and above lower-level
timing. The effect of task complexity was pronounced in
older adults, revealing an age-specific deterioration in
sequencing capacity. Older adults increases in variabilities
in complex sequence tasks corresponded with higher
numbers of sequence errors that is, lack of distinctions in
relative target durations within the same rhythmic cycles.
The experiment in the Krampe et al. [33] study that was
directly aimed at executive control functions, used a motor
variant of the task-set switching paradigm [1,43]. In this
rhythm-switching task, participants had to alternate between
a simpler, dominant rhythmic pattern and a difficult pattern
within the same trial. Both patterns were made up from the
same target intervals, which were, however, arranged in
different sequential orders. Older, but not young adults were
found to frequently perseverate the simpler pattern, pointing
to process-specific, age-related decrements in executive
control functions. Krampe et al. attributed their findings to

772

R.Th. Krampe / Neuroscience and Biobehavioral Reviews 26 (2002) 769776

the involvement of executive control functions [37,43] in

the endogenous provision of sequence information.
The term endogenous provision denotes the assumption
that action control relies on mental representations (i.e.
action schema or mental sets) that contain information about
task-relevant cues, appropriate motor programs, and
performance demands. Related representations (task sets)
must be maintained and activated (i.e. endogenously
provided for) at critical points in time during performance.
In the context of rhythm production, action schema or
mental sets correspond to the notion of rhythm programs
[55], abstract representations of the serial order of intervals
and their relative durations. Rhythm programs must be
actively maintained or revised during performance, for
example, during overall tempo changes. The proposition of
abstract mental representations as one aspect of action
control is especially plausible in rhythm production: if, like
Vorberg and Wing assume, participants have a representation of relative durations of single intervals within
rhythmic cycles, overall changes in performance tempo
become possible through parameter tuning [55] without the
rhythm program becoming obsolete. Endogenous provision
of abstract mental representations is critical even in musical
sight-reading, because the information contained in musical
notation is underspecified with respect to expressive
variation and optimal implementation (e.g. the fingering in
piano playing).
The age-graded dissociation of low-level timing processes on the one hand, and sequencing and executive
control mechanisms on the other, as observed in the Krampe
et al. study [33], is in line with models of cerebellar timing
[21] and theories attributing executive control to prefrontal
cortex areas [40,59]. From a developmental perspective, the
described findings lend themselves to an adaptation
interpretation with respect to differential age-related
changes in component processes underlying fine motor
movements: by relying on (or defaulting to) their relatively
intact lower-level timing mechanisms older adults simplify
complex rhythmic structures by sacrificing complex
sequencing and executive control operations that overtax
their processing capacities. In considering the adaptive
value of this behavior it is important that such simplification
comes together with the maintenance of overall structure
and performance tempo; altogether an option preferable to
the breakdown of performance in fine motor tasks like
rhythm production or music.
Another, more obvious form of adaptive potential relates
to the benefits of task-related experience. Both young and
older adults in the Krampe et al. study [33] improved, that is
reduced, their timing variability considerably in the course
of the multi-session experiment. Performances even in
simple fine motor tasks like repetitive finger tapping have
been shown to benefit considerably from practice and this is
equally true for older adults when their performance is
assessed across multiple sessions [31]. Neuropsychological
and neurophysiological studies have also revealed

considerable brain plasticity for motor functions until later

adulthood [6,49]. At the level of daily fine motor activities,
Dixon et al. [9] found that older adults handwriting speed is
only reduced relative to young adults when they have to
write unfamiliar text. In addition, these age-effects were
considerably reduced even through the limited training
administered in the lab. In the context of cognitive-motor
expertises individuals accumulate task-related experience or
invest time in goal-directed practice of fine motor movements over years or decades. Studies focusing on expertiserelated process adaptations and their moderating effects on
age-related changes are discussed in Section 4.

4. Age and cognitive-motor expertise

Expertise research can be characterized by its focus on
stable interindividual performance differences in specific
domains and areas of functioning [13,14]. The apparent ease
and efficiency at which experts perform their skills even at
advanced ages is especially intriguing in the light of the
ubiquitous age-related slowing demonstrated for cognitive
and motor functioning. Two alternative types of explanations for this apparent discrepancy [28,31,45] are
discussed in the literature: The first position maintains,
that experts of any age have always been superior in general
(e.g. cognitive speed, intelligence) and/or expertise-relevant
abilities (e.g. musical talent, finger dexterity) such that their
advantages could be attributed to interindividual differences
with long-term stability that already existed prior to
expertise acquisition. From this account we would expect
that high-levels of proficiency in a certain type of expertise
coincide with superior levels of performance in more
general abilities. An alternative account is that experts have
acquired mechanisms that permit them to circumvent the
specific limitations in general processing only in those tasks
or activities relevant to their domains [8,45]. According to
the latter view, expertise development amounts to a longterm process resulting in maximum adaptation to specific
task-constraints [13,14]. Certain more general functions
might well benefit from long-term training in a specific
expertise (like single-finger-tapping rate could benefit from
practicing the piano or typing), however, the critical sources
of interindividual differences should be found in more
specific mechanisms.
In those areas of expertise that heavily rely on domainspecific knowledge, like chess or medical diagnosis, an
explanation based on general intelligence factors is unlikely
to turn out a primary account. Age-comparative studies on
knowledge-based types of expertise have been conducted
for chess, GO-playing, mastermind, crossword puzzle
solving, and management skills [28]. The general picture
emerging from these studies is that older experts showed
normal (i.e. similar to non-expert controls) age-graded
decline in general measures of processing speed and
performance on unfamiliar materials. At the same time,

R.Th. Krampe / Neuroscience and Biobehavioral Reviews 26 (2002) 769776

older experts showed reduced, if any, age-related declines in

the efficiencies or the speed at which they perform skillrelated tasks and level of expertise was only weakly
correlated to performance in psychometric IQ-tests.
Arguably, accounts based on dispositions existing prior
to the skill acquisition process have more plausibility in
domains of expertise that have strong sensorimotor or
physiological components. The assumption that interindividual differences in basic cognitive-motor functions are
natural precursors to or contribute to interindividual
differences in expert performance can indeed be found in
the literature [25]. Correlational findings from several
studies appear in line with related assumptions. Overall
typing speed is correlated with the maximum rate of
repetitive single-finger tapping [3,44]. Professional pianists
show higher maximum finger tapping rates than musically
untrained individuals or amateur musicians [26,31,53].
Studies using simple interval production or duration
judgment tasks illustrated that the timing accuracy of
central clock mechanisms and the variability of movement
implementation are more efficient in trained musicians
compared to novices [26,30].
Experimental studies that tried to evaluate the effects of
aging in forms of expertise that involve fine motor
movements covered such diverse domains as typewriting
[5,44], piloting [42,54], piano playing [33], and golfing [22,
41]. The general outcome of these studies mirrored the
pattern of findings emerging from investigations of knowledge-based forms of expertise, namely that, at least until
their sixth or even seventh decade of life, older experts
showed little, if any, declines in speed and accuracy of
performance in skill-related tasks while their performance
in general measures of cognitive speed was similar to that of
normal older adults.

5. Age- and expertise-related changes in component

processes of typing and piano playing
Donald Gentners [16] work on typing can be viewed as a
classic example of decomposing a complex fine motor skill
into basic motor functions and more complex specific
mechanisms. His studies showed that beginning and expert
typists show very distinct patterns of transition times
(interkeystroke intervals) for successive keystrokes. The
fastest transitions in novice typing are repetitions of
identical letters typed with the same, single finger, while
the slowest (and most error prone) transitions emerge for
successive letters typed with different fingers in the hands.
For expert typists, the pattern is reversed, with repetitive
keystrokes being the slowest and hand alternations yielding
the fastest transitions. From analyses of high-speed video
films of typists performances Gentner was able to identify
the critical processes underlying this change in performance
characteristics. Skilled typists launch keystrokes with
different fingers or hands almost simultaneously or with

773

very short time delays thereby minimizing the resulting

interkeystroke intervals. Gentner termed this phenomenon
advance preparation. Advance preparation is also a typical
characteristic of expert performance in fine motor skills like
tennis or table tennis [4,18].
Gentner [16] conceived of the process of skill acquisition
in typing as a gradual optimization of those cognitive
processes that support advance preparation such that the
ultimate level of performance is maximally adapted to
peripheral factors (i.e. keyboard layout). Evidence for agegraded adaptation processes in typing came from studies by
Salthouse [44,45], who investigated individuals from a large
age-range. He found that, across agegroups, basic components of movement proficiency like the maximum finger
repetition rate showed a relatively modest relation to overall
typing speed accounting for 42% of the variance. In
contrast, measures reflecting complex expertise-related
mechanisms like the speed of typing letters with alternate
hands or the eye-hand span accounted for more than 70% of
the interindividual differences in overall typing speed.
Salthouse argued that the successful maintenance of typing
skills in his older expert typists relies on cognitively
complex mechanisms, namely extensive anticipation or
advance preparation, as illustrated by older skilled typists
longer eye-hand spans (i.e. the number of letters they looked
ahead prior to executing the actual keystrokes).
Salthouses [44,45] findings suggest that older experts
attain the same level of performance as young experts either
by means of different mechanisms, or at least by
differentially relying on different component processes [5].
However, it is not possible from these studies to determine
whether older typists performances at younger ages were
already superior and associated mechanisms were better
preserved due to a slower age-related decline or whether
they had invested deliberate activities to maintain these
critical capacities during later adulthood.
A study designed to address these issues was conducted
by Krampe and Ericsson [31]. Expert and amateur pianists
from different agegroups in their study performed tests of
maximum repetitive tapping rate, speeded multi-finger and
bimanual sequencing tasks, and non-speeded tasks like
memorization of sequences and expressive musical
interpretation. In the older amateur group, age-effects in
speeded multiple-finger and bimanual coordination tasks
were similar to those observed for general processing speed
as measured by psychometric tests or choice-reaction time
tasks. In contrast, older professional pianists showed normal
age-related decline in measures of general processing speed
while negative age-effects were reduced or fully absent in
expertise-related tasks. Taken together, these findings
yielded a crossover interaction pattern of age, expertise,
and capacity (skill-related vs. general speed). In line with
the arguments made for advance preparation in complex
finger movements in the typing literature, Krampe and
Ericsson found that the expertise-advantage in complex
finger sequencing tasks remained substantial, even if

774

R.Th. Krampe / Neuroscience and Biobehavioral Reviews 26 (2002) 769776

the sizeable expertise-effects in maximum repetition rate

were controlled for.
Krampe and Ericsson [31] argued that their findings
reflect older experts selective maintenance of acquired,
expertise-specific, mechanisms. Their selective maintenance interpretation rests on self-report and diary data
measuring participants time-investment into deliberate
practice and other activities. The degree to which levels of
performance in expertise-related tasks was maintained
depended on the amounts of deliberate practice invested at
the later stages of expertise development, namely in the fifth
and sixth decade of life. Furthermore, skill components
differed with respect to the degree to which deliberate
practice moderated age-related performance changes. While
experts had strong advantages in components without
explicit speed aspects like the memorization of sequences
or the control of timing and force variability during
expressive performance, negative age-effects were absent
in these tasks. The same was true for the maximum rate of
repetitive single-finger tapping where expert pianists were
much faster, but older amateurs performed at levels
comparable to young amateurs. Only when alternate tapping
was compared, older amateurs (but not older experts)
showed a disadvantage relative to their younger counterparts and the degree of skill maintenance was strongly
correlated to amounts of practice. Thus, the relatively small
amounts of practice and the lifelong task-related experience
coming with amateur levels of engagement are sufficient for
age-graded stability in several skill-components that show
age-related decline in the normal population. In contrast,
those mechanisms that portray the strongest expertise
advantage require the largest amounts of practice to acquire
and to maintain in later adulthood.
The correlational evidence for the role of deliberate
practice activities in acquiring and maintaining expertise in
fine motor skills is in line with neuropsychological
evidence: Playing a musical instrument increases the
cortical representation of the hand and single fingers and
the size of the effect correlates to the years of training [11].
Plasticity of related brain areas has also been demonstrated
for older adults who took up playing the violin or who
intensified their practice regime later in life [12].
Further evidence for expertise-specific adaptations in the
processes underlying fine motor movements came from
studies of bimanual coordination and timing. In one study
[32] professional pianists performed two bimanual rhythms
at a large range of tempos. The simpler rhythm had an
isochronous structure that is, all intervals between successive keystrokes regardless of which hand initiated or
terminated them, had the same target durations. The other
task was a 3 (left) against 4 (right) polyrhythm in which
intervals produced within either hand had an isochronous
structure. Bimanual polyrhythms have been studied quite
extensively in the motor literature, because their temporal
structure lends itself to hand-independent, parallel timing in
which cognitive control structures delineating events

produced by one hand are independent of those for the

other at some level. Earlier studies, however, failed to
provide evidence for parallel timing; instead they found that
even trained musicians relied on integrated timing control
when producing polyrhythms. Integrated timing control
means that keystrokes for both hands are strictly delineated
within a common temporal reference frame that is, as
immediately successive events between hands. Similar
findings were obtained by Krampe et al. [32] when slow
and moderate tempos were considered; however, these
authors found that expert pianists switched to parallel, handindependent timing control when performing polyrhythms
at rapid tempos. Considerations of the performance
constraints arising at different tempos reveal that both
integrated and parallel control modes are adaptive processing mechanisms tuned towards the goal of minimizing the
variability of produced intervals [30]. At slow and moderate
tempos, the variability of observed intervals is mostly
determined by the clock mechanisms such that integrated
timing of the shortest intervals (those among immediately
successive keystrokes) is optimally adaptive. At fast
tempos, clock variabilities are very small such that experts
use their superior sequencing and coordination mechanisms
(i.e. parallel timing) to exploit the benefits of within-hand
isochrony. Even within the highly select group of
professional pianists individuals differed considerably with
respect to their capabilities of implementing parallel timing.
In general, those pianists with the largest amounts of
accumulated practice were more likely to show parallel
timing and thereby realize performance tempos up to
500 ms/cycle (i.e. 125 ms for right-hand intervals and
167 ms for left-hand intervals).
A more recent study with young and older, accomplished
amateur pianists [29] confirmed that parallel timing is
outside the performance range of non-expert individuals.
Participants in both agegroups relied on integrated timing
and the maximum performance tempo they achieved was
clearly below that of expert pianists. Older amateur pianists
did as well as their young counterparts in the simpler task
with the isochronous interval structure. When producing the
polyrhythm at convenient or slower tempos, older amateurs
produced higher variabilities than young individuals did,
and their pattern of relative interval durations deteriorated
towards the simpler, isochronous pattern. The latter finding
is in line with older adults selective reliance on their
preserved lower-level timing skills and the age-graded
decline of higher-level sequence mechanisms discussed
earlier [33].

6. Conclusions
Studies aiming at the experimental decomposition of
complex movement control and the investigation of expert
motor performance portray a differentiated picture that
challenges the apparent parsimony of the general slowing

R.Th. Krampe / Neuroscience and Biobehavioral Reviews 26 (2002) 769776

account of age-related changes. In this context, a recent shift

in emphasis of experimental investigation from maximum
speed towards the variability in fine motor movements
proved useful to demonstrate age-graded dissociations of
component processes underlying fine motor movements.
Processes like timing, sequencing, and executive control
show differential patterns of age-related declines that cannot
be reduced to proportional reductions in general processing
speed. Likewise, age by expertise designs reveal dissociable
trajectories of age-related change for general components of
processing speed and expert-mechanisms and demonstrate
that skill components differ with respect to their modifiability through deliberate practice. Studies in expertise and
cognitive aging research converge on findings that show that
older adults selectively rely on those processing mechanisms that are less sensitive to age-related decline or that
individuals managed to selectively maintain until later
adulthood. None of these findings can be accommodated to
general factor models of aging or expertise.
Expertise and cognitive aging research build on a
combination of individual difference approaches and
experimental methods aimed at dissociating the underlying
processes. Process dissociations observed at a behavioral
level in developmental (i.e. age- and expertise-related)
contexts reflect individuals long-term adaptations to
internal and external performance constraints. The stable
outcomes (individual differences) of these long-term
adaptation processes are promising starting points for the
search for underlying neural substrates. In this context, agegraded process dissociations can be cornerstones among the
mutual constraints between experimental and neurophysiological research.
References
[1] Allport DA, Styles EA, Hsieh S. Shifting intentional set: exploring the
dynamic control of tasks. In: Umilta C, Moscovitch M, editors.
Attention and performance, vol. XV. Hillsdale, NJ: Erlbaum; 1994. p.
42152.
[2] Amrhein PC, Stelmach GE, Goggin NL. Age differences in the
maintenance and restructuring of movement preparation. Psychol
Aging 1991;6:45166.
[3] Book WF. Voluntary motor ability of the worlds champion typists.
J Appl Psychol 1924;8:283308.
[4] Bootsma RJ, Van Wieringen PC. Timing an attacking forehand drive
in table tennis. J Exp Psychol: Hum Percept Perform 1990;16:219.
[5] Bosman EA. Age-related differences in the motoric aspects of
transcription typing skill. Psychol Aging 1993;8:87102.
[6] Buonomano DV, Merzenich MM. Cortical plasticity: from synapses
to maps. Ann Rev Neurosci 1998;21:14986.
[7] Cerella J. Aging and information processing rates in the elderly. In:
Birren JE, Schaie KW, editors. Handbook of the psychology of aging.
San Diego, CA: Academic Press; 1990. p. 20121.
[8] Chase WG, Ericsson KA. Skill and working memory. In: Bower GH,
editor. The psychology of learning and motivation, vol. 16. San
Diego, CA: Academic Press; 1982. p. 1 58.
[9] Dixon RA, Kurzman D, Friesen IC. Handwriting performance in
younger and older adults: age, familiarity, and practice effects.
Psychol Aging 1993;8:36070.

775

[10] Duchek JM, Balota DA, Ferraro FR. Component analysis of a

rhythmic finger tapping task in individuals with senile dementia of the
Alzheimers type and in individuals with Parkinsons disease.
Neuropsychology 1994;8:21826.
[11] Elbert T, Pantev C, Wienbruch C, Rockstroh B, Taub E. Increased
cortical representation of the fingers of the left hand in string players.
Science 1995;270:3057.
[12] Elbert T, Sterr A, Rockstroh B. Untersuchungen zu corticaler
Plastizitat beim erwachsenen Menschen: was lernt das Gehirn beim
Geige spielen? [Studies of cortical plasticity in adults: what does the
brain learn from playing the violin?]. Musikphysiol Musikmedizin
1996;3:5765.
[13] Ericsson KA, Krampe RT, Tesch-Romer C. The role of deliberate
practice in the acquisition of expert performance. Psychol Rev 1993;
100:363406.
[14] Ericsson KA, Lehmann AC. Expert and exceptional performance:
evidence on maximal adaptations on task constraints. Ann Rev
Psychol 1996;47:273305.
[15] Fitts PM. The information capacity of the human motor system in
controlling the amplitude of movement. J Exp Psychol 1954;47:
38191.
[16] Gentner DR. Expertise in typewriting. In: Chi MTH, Glaser R, Farr MJ,
editors. The nature of expertise. Hillsdale, NJ: Erlbaum; 1988. p. 1 21.
[17] Gibbon J, Church RM, Meck WH. Scalar timing in memory. In:
Gibbon J, Allan L, editors. Timing and time perception. Annals of the
New York Academy of Sciences, 423. New York: New York
Academy of Sciences; 1984. p. 5277.
[18] Goulet C, Bard C, Fleury M. Expertise differences in preparing to
return a tennis serve: a visual information processing approach. J Sport
Exercise Psychol 1989;11:38298.
[19] Greene LS, Williams HG. Age-related differences in timing control of
repetitive movement: application of the Wing Kristofferson model.
Res Quart Exercise Sport 1993;64:328.
[20] Hazeltine E, Helmuth LL, Ivry RB. Neural mechanisms of timing.
Trends Cogn Sci 1997;1:1639.
[21] Ivry R. Cerebellar timing systems. Int Rev Neurobiol 1997;11:
55573.
[22] Jagacinski RJ, Greenberg N, Liao M-J. Tempo, rhythm, and aging in
golf. J Motor Behav 1997;29:159 73.
[23] Jagacinski RJ, Liao M-J, Fayyad EA. Generalized slowing in
sinusoidal tracking by older adults. Psychol Aging 1995;10:819.
[24] Jordan MI, Rosenbaum DA. Action. In: Posner MI, editor.
Foundations of cognitive science. Cambridge, MA: MIT Press;
1989. p. 72767.
[25] Keele SW, Hawkins HL. Explorations of individual differences
relevant to high level skill. J Motor Behav 1982;14:323.
[26] Keele SW, Pokorny RA, Corcos DM, Ivry R. Do perception and motor
production share common timing mechanisms: a correlational
analysis. Acta Psychol 1985;60:173 91.
[27] Kliegl R, Mayr U, Krampe RT. Time-accuracy functions for the
determination of person and process differences: an application to
cognitive aging. Cogn Psychol 1994;26:13464.
[28] Krampe RT, Baltes PB. Intelligence as adaptive resource development
and resource allocation: a new look through the lenses of SOC and
expertise. In: Sternberg RJ, Grigorenko EL, editors. Perspectives on
the psychology of abilities, competencies, and expertise. New York:
Cambridge University Press; 2002. in press.
[29] Krampe RT, Engbert R, Kliegl R. Age-specific problems in rhythmic
timing. Psychol Aging 2001;16:12 30.
[30] Krampe RT, Engbert R, Kliegl R. The effects of expertise and age on
rhythm production: adaptations to timing and sequencing constraints.
Brain Cogn 2002;48:17994.
[31] Krampe RT, Ericsson KA. Maintaining excellence: deliberate practice
and elite performance in young and older pianists. J Exp Psychol: Gen
1996;125:331 59.
[32] Krampe RT, Kliegl R, Mayr U, Engbert R, Vorberg D. The fast and
the slow of skilled bimanual rhythm production: parallel versus

776

[33]

[34]

[35]

[36]
[37]
[38]

[39]

[40]
[41]
[42]

[43]
[44]
[45]

[46]
[47]

R.Th. Krampe / Neuroscience and Biobehavioral Reviews 26 (2002) 769776

integrated timing. J Exp Psychol: Hum Percept Perform 2000;26:
20633.
Krampe RT, Mayr U, Kliegl R. Timing, sequencing, and executive
control in repetitive movement production. Potsdam: Department of
Psychology, University of Potsdam; 2002.
Lashley KS. The problem of serial order in behavior. In: Jeffress LA,
editor. Cerebral mechanisms in behavior. New York: Wiley; 1951. p.
11236.
Liao M-J, Jagacinski RJ, Greenberg N. Quantifying the performance
limitations of older and young adults in a target acquisition task. J Exp
Psychol: Hum Percept Perform 1997;23:164464.
MacKay DG. The organization of perception and action: a theory for
language and other cognitive skills. New York: Springer; 1987.
Mayr U, Kliegl R. Task-set switching and long-term memory
retrieval. J Exp Psychol: Learn Mem, Cogn 2000;26:1124 40.
Mayr U, Kliegl R, Krampe RT. Sequential and coordinative
processing dynamics in figural transformations across the life span.
Cognition 1996;59:61 90.
Meck WH, Benson AM. Dissecting the brains internal clock: how
frontal-striatal circuitry keeps time and shifts attention. Brain Cogn
2002;48:195 211.
Miller EK. The prefrontal cortex and cognitive control. Nature Rev:
Neurosci 2000;1:5965.
Molander B, Baeckman L. Attention and performance in miniature
golf across the life span. J Gerontol 1994;49:3541.
Morrow D, Leirer V. Aging, pilot performance and expertise. In: Fisk
AD, Rogers WA, editors. Handbook of human factors and the older
adult. New York: Academic Press; 1977. p. 199 230.
Rogers RD, Monsell S. The cost of a predictable switch between
simple cognitive tasks. J Exp Psychol: Gen 1995;124:20731.
Salthouse TA. Effects of age and skill in typing. J Exp Psychol: Gen
1984;113:345 71.
Salthouse TA. Expertise as the circumvention of human processing
limitations. In: Ericsson KA, Smith J, editors. Toward a general theory
of expertise: prospects and limits. New York: Cambridge University
Press; 1991. p. 286 300.
Salthouse TA. The processing-speed theory of adult age differences in
cognition. Psychol Rev 1996;103:40328.
Salthouse TA. A theory of cognitive aging. Amsterdam: North
Holland Press; 1985.

[48] Salthouse TA, Babcock RL, Skovronek E, Mitchell DRD, Palmon R.

Age and experience effects in spatial visualization. Dev Psychol 1990;
26:12836.
[49] Sanes JN, Donoghue JP. Plasticity and primary motor cortex. Ann Rev
Neurosci 2000;23:393415.
[50] Scheibel AB. Structural and functional changes in the aging brain. In:
Birren JE, Schaie KW, editors. Handbook of the psychology of aging.
San Diego, CA: Academic Press; 1996. p. 105 28.
[51] Serrien DJ, Swinnen SP, Stelmach GE. Age-related deterioration of
coordinated interlimb behavior. J Gerontol: Psychol Sci Social Sci
2000;55B:295 303.
[52] Stelmach GE, Goggin NL, Amrhein PC. Aging and the restructuring
of precued movements. Psychol Aging 1988;2:151 7.
[53] Telford CW, Spangler H. Training effects in motor skills. J Exp
Psychol 1935;18:1417.
[54] Tsang PS, Shaner TL. Age, attention, expertise, and time-sharing
performance. Psychol Aging 1998;13:32347.
[55] Vorberg D, Wing AM. Modelling variability and dependence in
timing. In: Heuer H, Keele SW, editors. Handbook of perception and
action. Motor skills, vol. 3. London: Academic Press; 1996. p.
181 261.
[56] Welford AT. Aging and human skill. Oxford: Oxford University Press
for the Nuffield Foundation; 1958.
[57] Welford AT. Motor performance. In: Birren JE, Schaie KW, editors.
Handbook of the psychology of aging. New York: Van Nostrand
Reinhold; 1977. p. 450 96.
[58] Welford AT. Signal, noise, performance, and age. Hum Factors 1981;
23:97109.
[59] West RL. An application of prefrontal cortex function theory to
cognitive aging. Psychol Bull 1996;120:272 92.
[60] Willingham DB. The neural basis of motor-skill learning. Am J
Psychol 1999;8:17882.
[61] Wing AM, Kristofferson AB. Response delays and the timing of
discrete motor responses. Percept Psychophys 1973;14:512.
[62] Wing AM, Kristofferson AB. The timing of interresponse intervals.
Percept Psychophys 1973;13:45560.
[63] Woodruff-Pak DS, Jaeger ME. Predictors of eyeblink classical
conditioning over the adult age span. Psychol Aging 1998;13:
193205.