Copyright 1994 by the American Psychological Association, Inc.

Journal of Experimental Psychology: H

0278-7393/94/53.00
Learning, Memory, and Cognition
1994, Vol. 20, No. 3, 739-749

Role of Elaborative and Perceptual Integrative Processes

in Perceptual-Motor Performance
Katherine Verdolini-Marston and David A. Balota

Three experiments address the dependence of both explicit and implicit memory performance on
elaborative processes for a perceptual-motor task, pursuit rotor. Explicit memory performance was
reflected by recognition of previously encountered pursuit rotor stimuli. Implicit memory
performance (priming) was identified in Experiment 1 as an advantage in pursuit rotor
performance for old stimuli that Ss failed to explicitly recognize. In Experiments 2 and 3, the types
of strategies that Ss engaged in during training and test phases were manipulated. Results
indicated that explicit memory performance depended on elaborative processes that emphasized
which specific stimuli were encountered, whereas reliable implicit memory performance appeared
only under a control no-instruction condition. Discussion focuses on attention to perceptual-
integrative processes for priming.

Within the past decade, considerable research has focused Squire, & Mishkin, 1982). More germane to our research is the
on a distinction between explicit and implicit memory perfor- processing distinction. Evidence from both amnesic and healthy,
mance (see review by Roediger, 1990). Explicit-memory perfor- normal subjects suggests that implicit and explicit memory
mance is revealed on traditional memory tests such as free performance appear to depend on different types of encoding
recall, cued recall, and recognition tests, in which subjects are processes. In particular, there is evidence indicating that
required to consciously recollect an earlier episode. Implicit explicit memory performance depends on elaborative pro-
memory performance is reflected by facilitations in perfor- cesses that relate target events to other contents of memory
mance through previous exposure to a stimulus without spe- (e.g., semantic processes), whereas implicit memory perfor-
cific recollection of earlier episodes (e.g., Graf & Schacter, mance depends more on integrative processes that reflect the
1985). Hence, implicit memory performance is reflected by influence of previous exposures on more unitized perceptual
tasks such as word-fragment completion (e.g., Tulving, Schac- representations (e.g., Graf, Mandler, & Haden, 1982; Jacoby
ter, & Stark, 1982; Warrington & Weiskrantz, 1974), percep- & Dallas, 1981). This differential sensitivity to different types
tual identification (Jacoby & Dallas, 1981), and homophone of encoding operations laid the groundwork for the dual-
spelling (e.g., Jacoby & Witherspoon, 1982). In each of these process model for implicit and explicit memory performance
tasks, subjects are not required to consciously recollect an (see, e.g., Graf et al., 1982; Mandler, 1979,1980).
earlier episode to perform the tasks, and yet there are clear Although there has been some compelling support for the
benefits from such earlier episodes. processing approach to explicit and implicit memory perfor-
The interest in the implicit-explicit distinction has been mance, some exceptions have been reported. These exceptions
nurtured by the possibility that fundamentally different memory raise questions about the theoretical utility of distinguishing
systems, processing modes, or both underlie these two types of between these different memory manifestations on the basis of
memory manifestation. Regarding the "systems" viewpoint, processing characteristics. One of the best examples of such an
data from amnesic subjects have provided evidence for distinct exception is a study reported by Blaxton (1989, Experiment 1).
and different neuroanatomical substrates underlying implicit In Blaxton's study, both explicit and implicit memory perfor-
and explicit memory performance (cf. Damasio, 1989; Mar- mance benefited from elaborative processing, in some cases,
tone, Butters, Payne, Becker, & Sax, 1984; Zola-Morgan, and from perceptual processing, in other cases, depending on
the retrieval demands of the memory test. That is, perfor-
Katherine Verdolini-Marston and David A. Balota, Department of mance on two explicit memory tests (free recall and semanti-
Psychology, Washington University. cally cued recall) and on one implicit memory test (a test of
The experiments included in this article were conducted as part of a general knowledge) was best following the elaborative process-
doctoral dissertation by Katherine Verdolini-Marston, under the ing of target stimuli, and performance on one explicit memory
direction of David Balota. We wish to thank other dissertation test (graphemically cued recall) and on one implicit memory
committee members, Richard Abrams and Ira Hirsh. Thanks are also test (word-fragment completion) was best following the simple
extended to Peter Graf, Dan Schacter, Kavitha Srinivas, and an perceptual processing of the targets. The importance of these
anonymous reviewer for their valuable comments on drafts of this results is that it would appear that explicit and implicit memory
article. We appreciate the contributions of numerous others, in
particular Dale Marston and Kevin Spratt. performance may not be fully distinguished by distinct process-
Correspondence concerning this article should be addressed to ing modalities but rather might be better understood in terms
Katherine Verdolini-Marston, who is now at the Department of of the match between encoding operations and the retrieval
Speech Pathology and Audiology, University of Iowa, Iowa City, Iowa demands of the memory test. The emphasis on the match
52242. Electronic mail may be sent to verdolini@wjshc-po.shc.uio- between encoding and retrieval operations has been most
wa.edu. clearly detailed in the transfer-appropriate processing frame-

739
740 KATHERINE VERDOLINI-MARSTON AND DAVID A. BALOTA

work (for further discussion, see Roediger, Srinivas, & Wel- episodes. For example, Fleshman (1984) emphasized the use
don, 1989; Roediger, Weldon, & Challis, 1989; also see Morris, of the metaphors of "the sea within" and "the sea without" to
Bransford, & Franks, 1977).1 facilitate acquisition of graceful dance movements. Lessac
One of the goals of our research was to provide further (1967) described speech and voice as an "orchestra" to
evidence regarding the notion that explicit and implicit memory facilitate acquisition of resonant voice and intelligible speech
performance are regulated by qualitatively different types of for theater. If in fact such metaphors facilitate implicit
encoding processes (e.g., elaborative and integrative pro- perceptual-motor memory performance following training, it
cesses, respectively). In pursuit of this goal, we attempted to might be possible to show a dependence of implicit memory
assess the sensitivity of both implicit and explicit memory performance on elaborative processes by incorporating (elabo-
performance to two different types of elaborative processes. rative) metaphors during training.
According to the dual-process (dissociation) model, implicit The second reason for using a perceptual-motor task was
memory should be relatively unaffected by distinct types of related to the anticipated dependence of implicit memory
elaborative processes because this type of memory perfor- performance on elaborative processes that emphasize how to
mance is relatively insensitive to elaborative encoding opera- do a task. Although implicit verbal memory, in general,
tions in general. However, different types of elaborative supposedly depends on procedures and operations or on
processes should modulate explicit memory performance. On knowing how (Cohen & Squire, 1980; Squire, 1986; Squire &
the other hand, according to the transfer-appropriate process- Cohen, 1984), the relevance of knowing how would appear, at
ing framework, one may find an influence of type of elaborative least at a surface level, to be stronger for a perceptual-motor
processing on both explicit and implicit memory performance task, as compared with verbal tasks that require subjects to
because the strongest predictor of performance within this complete word fragments or to identify rapidly presented words.
approach is the degree to which the encoding operations The third and final reason we used a perceptual-motor task
match the retrieval operations that are demanded by the was that only a limited number of studies have assessed the
memory tests. Hence, the crucial dimension is not the type of mental processes that mediate explicit and implicit memory
processing but rather the match between the encoding and performance beyond the verbal domain (e.g., Nissen & Bulle-
retrieval operations. mer, 1987; Schacter, Cooper, & Delaney, 1990; Schacter,
A critical aspect of our research is the identification of the Cooper, Delaney, Peterson, & Tharan, 1991). Although the
appropriate two classes of elaborative encoding operations. influence of elaborative processing on implicit memory perfor-
The two types of elaborative processes selected were based on mance is relatively rare in the verbal domain, such a relation-
the views expressed by a number of memory researchers ship may be more readily obtained in other domains. As
regarding explicit and implicit memory performance. It has pointed out previously, the perceptual-motor domain is a
been pointed out that explicit memory performance involves prime candidate for such a demonstration.
the unique specification of events (see, e.g., Damasio, 1989; In summary, our purpose was to provide further evidence
Tulving, 1972), and according to some speculations, implicit regarding the dissociation between explicit and implicit memory
memory performance involves procedures and operations, or performance as a function of processing level in a perceptual-
"knowing how" to perform a task (Cohen & Squire, 1980; motor task. We assessed the influence of two types of elabora-
Squire, 1986; Squire & Cohen, 1984). Thus, in our research, tive processes on implicit and explicit memory performance.
one class of elaborative encoding conditions emphasized which On the basis of theoretical descriptions regarding the opera-
specific stimuli subjects encountered, and the other class tions that underlie implicit and explicit memory performance,
emphasized how to do the experimental task. According to a one might expect that elaborative processes that emphasize
transfer-appropriate processing view, one might expect that which unique stimuli were encountered would benefit explicit
elaborative processes that individuate stimuli would benefit memory performance, whereas elaborative processes that
explicit memory performance, whereas elaborative processes emphasize knowing how to do a task would benefit implicit
that emphasize how to do a task would benefit implicit memory memory performance, especially on a perceptual-motor task.
performance. In contrast, on the basis of an unembellished On the other hand, according to an unembellished dual-
dual-process model, one might expect that the different types process (dissociation) model, one might expect that explicit
of elaborative processes would be relatively ineffective in memory performance would be especially sensitive to the type
modulating implicit memory performance and would be primar- of elaborative processing, whereas implicit memory perfor-
ily effective in modulating explicit memory performance. mance would be relatively uninfluenced.
In our study, a perceptual-motor task (pursuit rotor) was
used to address the influence of these two types of elaborative Overview of the Experiments
processes on explicit and implicit memory performance. The The subjects' task in each experiment was to track a rotating
selection of a perceptual-motor task is important for the visual target with a wand (pursuit rotor task). In each experi-
following three reasons: First, there is anecdotal evidence that
suggests that elaborative processes can enhance implicit type
1
memory manifestations in the perceptual-motor domain. Spe- Note that at a general level, the dual-process model can also be
cifically, some professional perceptual-motor trainers claim considered under the rubric of a transfer-appropriate processing
that metaphoric images, which relate target productions to approach (Graf & Ryan, 1990). However, in this article, transfer-
other contents of memory, promote benefits in perceptual- appropriate processing refers to the specific version described by
motor performance without reference to previous training Roediger and colleagues.
 PERCEPTUAL-MOTOR PERFORMANCE 741

ment, subjects tracked several different targets that were

created by varying target radii and target speeds. Each specific
speed-radius combination is referred to as a stimulus. The
measure of general perceptual-motor skill was time on target
(TOT), in seconds, for each minute of practice.
During an initial phase, subjects first warmed up on the
pursuit rotor task by practicing on relatively easy stimuli
(stimuli with comparatively low tangential velocities). Subjects
then received several more difficult, critical stimuli during the
training phase. After a delay, subjects returned for a test
phase. During the test phase, subjects first practiced on the
same warm-up stimuli as they had done previously and then
received a set of critical stimuli. Some of these critical stimuli
were old, and some were new. After each critical test trial,
subjects responded (yes or no) as to whether they recognized
the preceding stimulus from the training phase. This was the Figure 1. Pursuit rotor equipment used in Experiment 1. It includes
first indication that there would be an explicit memory test. modified pursuit rotor, rotary timer, and wand. Replacement disks are
Proportion of correct responses on the recognition test consti- not shown.
tuted the measure of explicit-memory performance. Chance
results on this test were reflected by a recognition score of .50.
The purpose of Experiment 1 was to provide evidence of when the goal is to quantitatively compare changes in different
implicit perceptual-motor memory performance, defined as a measures as a function of the experimental manipulations.
facilitation in perceptual-motor performance (in TOT) follow- Neither of these problems occurred in this study. To anticipate
ing practice, without explicit reference to previous learning the results we modulated by processing condition both explicit
episodes. To be consistent with most previous studies in the and implicit memory performance across all the experiments.
verbal domain, an additional requirement for the measure of Thus, neither explicit nor implicit measures of memory perfor-
implicit memory performance was that facilitations be shown mance were insensitive. Furthermore, our purpose was not to
for specific stimuli that subjects did not reliably recognize as quantitatively compare variations in explicit and implicit
old (priming without awareness). memory performance as a function of processing condition but
The purpose of Experiments 2 and 3 was to address the rather to detect the dependence of these performance types on
primary theoretical question of interest, that is, the role of various encoding operations. Thus, although scaling differ-
distinct types of elaborative processes for explicit and implicit ences across implicit and explicit-memory tests are a potential
perceptual-motor memory performance. In these experi- concern in this area of research, these concerns did not
ments, the strategies that subjects engaged in were manipu- compromise our conclusions.
lated, and the effects on explicit and implicit memory perfor-
mance were observed. Specifically, subjects in different groups Experiment 1
received (a) elaborative processing instructions that empha-
As noted, the goal of the first experiment was to identify a
sized which specific stimuli were encountered, (b) elaborative
measure of implicit perceptual-motor memory performance,
processing instructions that emphasized how to do the pursuit
which is defined as a performance facilitation (priming) for
rotor task, (c) perceptual-integrative instructions that empha-
previously encountered stimuli that subjects do not reliably
sized attention to the surface characteristics of the stimuli
recognize.
(Experiments 2 and 3), or (d) no instructions about mental
strategies (Experiment 3).
A final introductory comment is worth noting here. The Method
measures of explicit and implicit memory performance involve Subjects. Twenty-two healthy adults, 12 men and 10 women rang-
different scales. That is, the measure of explicit-memory ing from age 18 to 42 years (M = 24.2 years) participated in the
performance was the proportion of correct identifications on a experiment as volunteers. With the exception of one subject, all were
recognition test, whereas the measure of implicit memory undergraduate or graduate students at Washington University. All but
performance was facilitation in pursuit rotor performance (in two subjects were right-handed, according to self-reports.
TOT) for previously encountered stimuli. One might ask Apparatus. The modified pursuit rotor used in Experiment 1 is
whether the results for explicit and implicit memory perfor- shown in Figure 1. This device consisted of a 20.7 cm x 10.1 cm x 30.1
mance could be compared because of these scaling differences. cm wooden frame that housed one of five disks that rotated in a
clockwise direction at 30, 60, or 90 rpm. A copper target with a
Although it is clearly desirable to use similar scales across
diameter of 0.95 cm was embedded in each of the disks. For each disk,
measures within an investigation, there was no obvious way to the center of the target was located at a different distance from the
create similar scales for explicit and implicit memory perfor- center of the disk: 3.45 cm, 5.39 cm, 7.27 cm, 9.15 cm, and 11.04 cm,
mance in the present experiments. More important, it should and 12.94 cm (Disks 1, 2, 3, 4,5, and 6, respectively). When the metal
also be noted that scaling issues in general pose major stylus (Lafayette Instrument Company Pursuit Rotor Stylus) made
interpretative problems when one of the measures under contact with the copper target, a timer (Lafayette Instrument Com-
examination is insensitive to experimental manipulations and pany rotary timer) was driven and indicated TOT in 10-ms increments.
742 KATHERINE VERDOLINI-MARSTON AND I WID A. BALOTA

Table 1 Results aru i Discussion

Performance in Training and Test Phases in
Experiment 1 (N = 22) As displayed in Table 1, subjects' performance (in TOT) on
the pursuit rotor task improved from training to test phases.
Test Implicit The average TOT for critical training stimuli was 22.81 s,
memory:
TOT, TOT, TOT, Explicit whereas the average TOT for the same stimuli at test was
Training: old new old-new memory: 32.50 s, and the average TOT for new stimuli of the same
Measure TOT stimuli stimuli stimuli PCR approximate difficulty was 30.37 s. Thus, the average TOT for
M 22.81 s 32.50 s 30.37 s 2.13 s* .54 old stimuli was higher than the average TOT for new stimuli
SD 5.82 s 6.53 s 7.74 s 3.45 s .19 during the test phase by a margin of 2.13 s. A one-way,
within-subject analysis of variance (ANOVA) confirmed that
Note. TOT = time-on-target; TOT advantage of old over new test
stimuli in implicit-memory performance; PCR = proportion correct on this performance advantage, or priming, was reliable, F(l,
recognition test. 21) = 8.36,/> < .01,AfSc = 5.95. Thus, there was clear evidence
*p <.01. of performance facilitation for previously encountered stimuli.
The mean proportion of correct identifications on the
recognition test was .54. This value, also shown in Table 1, was
not reliably different from chance, z = .37.
A cassette recorder was used to deliver instructions about when to
start and stop pursuit-rotor practice for all phases of the experiment. The poor explicit recognition performance might appear a
Stimulus materials. Warm-up stimuli involved Disk 1 that was bit surprising, given that each subject received only four
presented at 30, 60, and 90 rpm. Critical training stimuli were drawn different critical stimuli during the training phase and was later
from a pool of 8 stimuli, a subset of 15 possible stimuli, involving Disks required to distinguish this limited number of stimuli among a
2, 3, 4, 5, and 6, presented at 30, 60, and 90 rpm. Each stimulus pool of only eight stimuli. Poor recognition performance was
involved a unique combination of target radius and target speed. For clearly related to the difficulty of remembering not only which
each subject, 4 different, critical training stimuli were selected from target radius was encountered but also the specific radius-
the pool of 8 stimuli, and each of the 4 stimuli was presented three speed combination. Thus, even if a subject correctly recog-
times during training. Each training set of 4 critical stimuli exhibited nized a disk from the training phase, this would not necessarily
the following characteristics: (a) The average tangential velocity across lead to correct recognition unless the subject was able to
the 4 different stimuli was approximately 60 cm/s and (b) for each remember the association between target radius and target
critical training stimulus that was presented to a given subject, another speed.
subject received a different critical stimulus with approximately the In summary, item-specific priming occurred for stimuli that
same tangential velocity. Critical test stimuli included all 8 stimuli from subjects did not reliably recognize. Therefore, the results of
the selected stimulus pool, 4 old stimuli and 4 new stimuli. For each
Experiment 1 provide evidence of implicit memory perfor-
subject, test stimuli were ordered according to the following con-
mance in a perceptual-motor task.
straints: (a) The average tangential velocity of stimuli in the first and
second halves of the test sequence was similar, ranging from approxi-
mately 59 cm/s to 62 cm/s, and (b) 2 old and 2 new stimuli appeared in
each half of the test sequence. All stimuli were counterbalanced across Experiment 2
subjects so that each stimulus appeared equally often as an old and as a
The purpose of Experiment 2 was to address the influence of
new stimulus.
elaborative processes on both explicit and implicit perceptual-
Procedure and design. For all pursuit rotor trials, subjects held the motor performance. During both training and test phases, the
tracking wand in the dominant hand. Throughout the experiment,
mental strategies that subjects engaged in were manipulated,
subjects received 1-min pursuit rotor trials that were separated by
1-min rest periods. Subjects first received the warm-up stimuli.
and the effect on explicit and implicit memory performance
Including 3 min of practice and 2 min of rest, the first warm-up phase was observed. Two of the mental strategies promoted elabora-
lasted 5 min. Following the warm-up phase, subjects received the tive processing because they required subjects to relate pursuit
critical training stimuli. Including 12 min of practice (4 stimuli by 3 rotor stimuli to other contents of memory. One of the
trials each) and 11 min of rest, the critical training phase lasted 23 min. elaborative strategies individuated the specific stimuli that
Approximately 24 hr later (22.4 to 27.8 hr), subjects returned for the subjects received (album condition), and the other elaborative
test phase. For this phase, subjects first warmed up on the same strategy emphasized how to perform the pursuit rotor task in
warm-up stimuli they had received the previous day. Following the general (stir condition). A third processing condition was
warm-up phase, subjects received the critical test stimuli and also the intended to promote perceptual processing by directing
recognition test. Including 8 min of pursuit rotor performance and 7 subjects' attention to the surface characteristics of the stimuli
min of rest, the total time for the critical test phase was 15 min. (concentrate condition).
Throughout the experiment, TOT information was provided imme- The predictions are relatively straightforward. According to
diately after each trial, except during the critical test phase. During the the dual-process model, one should expect that both types of
test phase, subjects first provided recognition responses and then were elaborative processing would facilitate explicit memory perfor-
informed about TOT. Additional information about performance was mance, as compared with the perceptual-concentrate condi-
provided by audible clicks in the pursuit rotor apparatus each time the
tion. Moreover, depending on the depth of elaborative process-
tracking wand went on or off target.
For both explicit memory (recognition) performance and implicit
ing across these two tasks, there may be reliable differences
memory (priming) performance, we used a one-way (old vs. new between the two types of elaborative processing in explicit
stimulus) within-subject design. memory performance. Turning to implicit memory perfor-
 PERCEPTUAL-MOTOR PERFORMANCE 743

mance, according to the dual-processing framework, one Procedure and design. The principal difference in the procedure, as
should expect performance to be either equivalent across all compared with Experiment 1, was the introduction of mental strate-
three conditions or possibly highest in the perceptual- gies during training and test phases. Following warm-up trials, which
lasted 5 min including 3 practice min and 2 rest min, subjects were
concentrate condition. On the other hand, according to the
given one of three different instructions: album, stir, or concentrate.
transfer-appropriate processing framework, one should expect (The full instructions are described elsewhere, see Verdolini-Marston,
that elaborative processing that individuated the different 1991.) Briefly, in the concentrate group, subjects were instructed to use
stimuli should result in superior explicit memory performance the pursuit rotor task as an exercise in concentration and to attend to
and that an elaborative strategy that emphasized how to do the the rotating target and to the target's path. These instructions were
pursuit rotor task should result in superior implicit memory assumed to promote perceptual processing, that is, attention to the
performance. surface characteristics of the stimuli. In the stir group, subjects were
instructed to think of the pursuit rotor task as stirring in a bowl (the
target path) with a wooden spoon (the tracking wand). The assump-
Method tion was that these instructions promoted elaborative processing
Subjects. Ninety-six healthy adults, 72 women and 24 men, volun- because they emphasized the relation between the pursuit rotor
teered for the experiment. Subjects were recruited from undergradu- stimuli and other contents of memory. It was further assumed that
ate and graduate classes at Washington University, as well as from the these instructions emphasized how to do the pursuit rotor task, on the
local community. Ages ranged from 17 to 46 years (Af = 31.2 years). basis of the results of a two-stage pilot study.2
Twenty-four women and 8 men were assigned to each of the three Finally, in the album group, subjects were instructed to view the
experimental groups. On the basis of subject reports, 87 subjects were pursuit rotor machine as a record player and the four different target
right-handed, 7 were left-handed, and 2 were ambidextrous. paths as different songs on an album. Subjects were asked to mentally
Apparatus. The pursuit rotor equipment was updated with a more hear a specific song during critical trials, depending on which target
recent apparatus, a Lafayette Instrument Company Photoelectric path was presented, and to hear each song at a speed that depended on
Rotary Pursuit (Model 30014), connected to a Lafayette digital the speed of target rotation. The assumption was that these instruc-
clock/counter (Model 54035). The pursuit rotor machine was 36 cm tions also promoted elaborative processing because they encouraged
wide x 36 cm deep x 21 cm high. Five removable glass plates that were the relating of pursuit rotor stimuli to other contents of memory. It was
placed on the top of the pursuit rotor machine formed the superior further assumed that these instructions emphasized which specific
surface. These plates were painted black, except for a transparent stimuli subjects received for practice because each pursuit rotor
circular path 2.0 cm wide. For each of the five plates, the radius of the stimulus corresponded to a unique combination of mental song and
circular path was different: 3.4 cm, 5.0 cm, 6.6 cm, 8.2 cm, and 9.8 cm mental song speed.
for Plates 1,2,3,4, and 5, respectively. After receiving instructions about mental strategies, subjects were
When the machine was activated, a light spun in a Clockwise given the opportunity to practice their respective mental strategies
direction beneath the surface of whichever plate was installed and briefly before proceeding with the training trials. After each critical
appeared as a target of 2.0 cm x 1.7 cm within the circular path of the trial, subjects indicated (yes or no) whether they had used the intended
plate. The speed of light rotation could be varied continuously from 1 mental strategy during the preceding trial, and they rated both the
to 100 rpm. A photoelectric sensor embedded in a hand-held wand clarity and persistence of the strategy during the preceding trial on a
activated a digital counter when the wand was directly above the light. 5-point scale. This enforced the processing strategies by introducing
The counter measured TOT to the nearest millisecond. The sensitivity some form of accountability.
of the light receptor was set at the maximum level throughout the In this experiment, 1-min critical trials during the training phase
experiment. (and also during the test phase) were separated by rest intervals of 1
Stimulus materials. Both training and test phases began with three min 15 s, as opposed to rest intervals of 1 min as in Experiment 1. In
warm-up stimuli, which involved Plate 1, and target rotations of 30 Experiment 2, subjects were reminded of the mental strategies before
rpm, 60 rpm, and 100 rpm. each trial, and extra time was provided for the reminders. Thus,
For each subject, critical stimuli for training and test phases including 12 min of practice and 13 min 45 s of rest, the critical training
included four different stimuli from a pool of eight stimuli involving phase lasted 25 min 45 s.
four plates (Plates 2,3,4, and 5) and two rotational speeds (60 rpm and Following a rest of 12 min 30 s, subjects returned for the test phase.
100 rpm). As in Experiment 1, each different critical stimulus involved The delay interval was 12 min 30 s in this experiment, as opposed to 24
a unique combination of target radius and target speed. For a given hr in Experiment 1. The interval was shortened Experiment 2 because
subject, each of the four different critical training stimuli was repeated effects of mental strategies on both explicit and implicit memory
three times, and together, the set of four stimuli reflected the following performance were the focus of interest in Experiment 2, and it is quite
constraints: (a) The average target tangential velocities across the possible that explicit memory performance might become insensitive
stimuli ranged from approximately 60 cm/s to 64 cm/s; (b) the four to the manipulations of interest over a 24-hr period. The test phase
different stimuli included one exemplar each of Plates 2,3,4, and 5; (c) began with the same warm-up stimuli that preceded the training
a given plate was always presented with the same rpm (60 rpm or 100 phase; for these warm-up trials, subjects were told they could use
rpm); (d) two of the four stimuli had rotations of 60 rpm, and two had whatever mental strategy they wished. Following the second warm-up,
rotations of 100 rpm; and (e) the same stimulus was not repeated on
successive trials, and each of the four different stimuli appeared at
least once within the first five trials. 2
In the first stage of the pilot study, 8 of 10 subjects generated
Critical stimuli for the test phase included all exemplars from the stirring as an image that they thought would help them to perform the
pool of eight stimuli. Two old stimuli and two new stimuli appeared in pursuit rotor task. In a second stage, 10 of 20 new subjects preferred
each half of each test sequence, and average tangential velocities for the stirring image over a circle-drawing image in terms of perceived
stimuli in the first and second halves of each test sequence were usefulness of the images for how to do the pursuit rotor task. Only 5 of
similar, ranging from 60 cm/s to 64 cm/s. As in Experiment 1, each these subjects preferred the circle image. Further details about the
stimulus was counterbalanced so that it occurred on an equal number pilot study are available from Katherine Verdolini-Marston and David
of trials in the old and new conditions across subjects. A. Balota on request.
744 KATHERINE VERDOLINI-MARSTON AND DAVID A. BALOTA

Table 2 aspects of pursuit rotor skill acquisition is secondary to the

Performance in Training and Test Phases in main focus of our investigation (i.e., item-specific implicit-
Experiment 2 (N = 96) memory performance), this pattern is noteworthy because it
Implicit does not support anecdotal claims of general benefits for a
Test
memory perceptual-motor task with metaphoric (elaborative) process-
TOT, TOT, TOT, Explicit ing conditions, even though subjects in our pilot study gener-
Group/ Training: old new old-new memory: ally perceived such instructions as facilitatory (stir condition,
measure TOT stimuli stimuli stimuli PCR see Footnote 2).
Concentrate Explicit memory. As shown in Table 2, explicit recognition
M 27.17 s 32.55 s 32.10 s 0.45 s .51s performance was clearly influenced by mental strategy. Recog-
SD 6.08 s 6.01 6.15 s 2.83 s ,15 s nition performance was poorest for the concentrate group
Stir
M 28.37 32.92 s 32.70 s 0.22 .55 s (.51), slightly better for the stir group (.55), and considerably
SD 6.90 s 7.75 s 7.91s 2.75 s .18 s better for the album group (.65). A one-way between-subjects
Album ANOVA yielded a main effect of mental strategy, F(2, 93) =
M 26.50 s 32.33 s 31.85 s 0.48 s .65 s* 5.59, p < .006, AfSe = 0.029. Post hoc Tukey comparisons
SD 7.67 s 7.70 s 8.25 s 2.10 s .19 s
indicated that explicit recognition performance for the album
Grand means 27.35 s 32.60 s 32.22 s 0.38 s .57 s " group was reliably higher than performance for the concen-
Note. TOT = time-on-target; TOT advantage of old over new test trate group, but none of the other comparisons produced
stimuli in implicit-memory performance; PCR = proportion correct on reliable differences. In fact, only recognition performance for
recognition test. the album group was reliably different from chance, z = 1.75,
*p < .05, one-tailed. * *p < .01 for main effect of group.
p < .05,one-tailed.34
To summarize, elaborative processing instructions resulted
which lasted a total of 5 min, subjects received the critical test. in relatively good explicit memory performance but only when
Including 8 min of pursuit rotor performance and 8 min 45 s of rest, the these instructions emphasized which unique stimuli subjects
final critical phase lasted 16 min 45 s. received (album instructions). Conversely, none of the instruc-
For explicit memory (recognition) performance, the design was a tions, including elaborative processing instructions (stir and
one-way (3 mental strategy groups), between-subjects design. For album instructions) and perceptual processing instructions
implicit memory (priming) performance, the design was a two-way (2 (concentrate instructions) resulted in implicit memory perfor-
[old-new status] x 3 [group]), mixed-factor design, with old-new mance (priming) that was better than chance.
status as a within-subject variable and group as a between-subjects The failure to obtain any priming in Experiment 2 was
variable. puzzling, especially in light of the fact that priming was clearly
demonstrated in the first experiment, in which mental strate-
Results and Discussion gies were not imposed. Assuming that the results from Experi-
Implicit memory. Table 2 shows that although all groups ment 1 were not spurious, one possible explanation is that the
improved in pursuit rotor performance from training to test strategies developed in Experiment 2 were simply inefficient
phases (M = 27.35 s for critical stimuli during the training for implicit memory performance. Thus, if one were to develop
phase, M = 32.60 s for old stimuli during the test phase, and
3
M = 32.22 s for new stimuli during the test phase), there was Hits and false alarms were evaluated to determine whether group
little evidence of an item-specific performance benefit for old differences in recognition performance were related to differences in
over new stimuli in the test phase, for any of the groups. Across sensitivity, response bias, or both. All groups showed a positive
groups, the average performance advantage for old stimuli was response bias of the same approximate magnitude: For all groups,
approximately 65% of recognition responses were yes (66.4% for the
0.38 s, and within-group averages ranged from 0.22 s (stir concentrate group, 64.5% for the stir group, and 65.6% for the album
group) to 0.48 s (album group). A 2 (old-new status) x 3 group). Therefore, changes in recognition as a function of group were
(mental strategy group) mixed-factor ANOVA failed to reveal not associated with changes in response bias. Rather, recognition
a significant main effect of priming (old vs. new status), F(l, differences were attributable to differential sensitivity.
93) = 2.03,p < .16,M5e = 3.32. In addition, the interaction of 4
In Experiment 2 and also in Experiment 3, post hoc questioning
Group x Old-New status, which indicated whether priming revealed that subjects were insightful about the relative usefulness of
differed as a function of group, did not approach significance, their respective instructions for explicit memory performance. Sub-
F(2, 93) = 0.10. Thus, on average, there was little or no jects in the album group rated the instructions as most helpful,
evidence of perceptual-motor priming in any of the mental whereas most other subjects rated the instructions as not helpful (stir
strategy groups. and concentrate groups, Experiment 2) or as helping very little
(locomotive and concentrate groups, Experiment 3); recognition data
It is interesting to note that not only were item-specific corroborated these impressions. However, subjects were not insightful
priming effects similar (null) across strategy groups, but a about the relative usefulness of their instructions for pursuit rotor
measure of generalized learning for the pursuit rotor task also performance. Subjects in the concentrate and stir groups in particular
did not change across strategy groups. Specifically, if one tended to rate their instructions as quite helpful, whereas subjects in
considers improvements from training items compared with the album group tended to think that their instructions did not help
new items in the test phase, there was little evidence of any pursuit rotor performance much, or helped less. However, these
group differences, F(2, 93) = 0.89, p < .42, MSe = 4.81. impressions were unrelated to pursuit rotor performance data. This
Although the main effect of instructional group on general finding is discussed in detail elsewhere (Verdolini-Marston, 1991).
 PERCEPTUAL-MOTOR PERFORMANCE 745

better elaborative strategies, one might find implicit priming in Table 3

the pursuit rotor task under such instructions. This possibility Performance in Training and Test Phases in
was addressed in Experiment 3. In addition, Experiment 3 Experiment 3 (N = 128)
addressed another simpler hypothesis that we had not consid- Implicit
Test
ered at the outset. Specifically, it is possible that implicit memory:
perceptual-motor memory manifestations, reflected by item- TOT, TOT, TOT, Explicit
specific priming effects, fail to develop under conditions in Group/ Training: old new old-new memory:
measure TOT stimuli stimuli stimuli PCR
which subjects are required to attend to mental strategies.
Mental-strategy groups
Concentrate
Experiment 3 M 29.02 s 33.97 s 33.57 s 0.40 s .60s
SD 6.38 s 7.35 s 7.72 s 2.69 s .15 s
The purpose of Experiment 3 was to address the failure to Locomotive
obtain reliable priming in any of the mental strategy groups in M 28.82 s 33.95 s 33.97 s -0.02 s .59 s
Experiment 2. Experiment 3 addressed the hypothesis that SD 6.80 s 7.74 s 7.62 s 1.86 s ,13 s
Album
some mental strategies (e.g., elaborative strategies that empha- M 27.19 s 32.62 s 32.66 s -0.04 s .79 s**
size how to do a task, or even perceptual strategies that SD 6.48 s 8.21s 8.31s 2.92 s .20 s
emphasize attention to surface characteristics of a stimulus)
No-instruction group
can result in reliable and even superior priming, but the M 28.31 s 35.43 s 34.18 s 1.25 s .54 s
particular strategies used in Experiment 2 were simply poor SD 5.62 s 6.52 s 6.35 s 2.10 s .17 s
ones. Thus, in Experiment 3, an attempt was made to improve Grand means
the effectiveness of the elaborative strategy that emphasized M 28.34 s 33.99 s** 33.60 s 0.39 s* 0.63 s**
how to do the pursuit rotor task (stir strategy) and of the
Note. TOT = time-on-target; TOT advantage of old over new test
perceptual strategy that emphasized attention to surface stimuli in the implicit-memory performance; PCR = proportion
characteristics of the stimuli (concentrate strategy). In addi- correct on recognition test.
tion, an attempt was made to increase the sensitivity of the *p < .05 For preplanned comparison between mental-strategy groups
priming measure in general by increasing the number of and no-instruction group. **p <-01. ***p <.01 For preplanned
repetitions for each critical training stimulus from three to comparison between mental-strategy groups and no-instruction group.
four. Finally, Experiment 3 also addressed the possibility that
perceptual-motor priming fails to develop when mental strate- The stir instructions used in Experiment 2 were substituted with
gies are imposed. This hypothesis was tested by reintroducing a "locomotive" instructions. This change in elaborative processing
no-instruction group in Experiment 3. If the results of Experi- instructions that emphasized how to do the pursuit rotor task was
ment 2 were due to the possibility that perceptual-motor based on a pilot study.5 The locomotive instructions required subjects
priming does not develop with imposed mental strategies, to view the circular target path as a wheel on a locomotive and to view
then the results of the third experiment should yield no the tracking wand as a rod attached to the target light on the wheel as it
implicit priming for any of the mental strategy groups and turned. The suggestion was that as the wheel turned, it would pull the
rod and thus the arm around with it.
reliable priming for the no-instruction group, as found in
Rest periods between critical trials were 1 min in Experiment 3, as in
Experiment 1.
Experiment 1. Rest periods were increased to 1 min 15 s in Experiment
2 to allow for reminders about mental strategies before each practice
Method minute. Reminders were also used in Experiment 3, but the rest
intervals were decreased in duration to offset the increase in the
Subjects. Subjects were 128 healthy adults who volunteered for the overall time required for the experiment as a result of the additional
experiment. Subjects were again recruited from Washington Univer- training trial for each stimulus during the training phase.
sity and the local community. Ages ranged from 17-44 years (M = 27.5 Including 16 min of training trials (4 stimuli x 4 repetitions) and 15
years). Sixty-three women and 65 men participated, and approximately min of rest, the initial critical phase lasted 31 min. The critical test
the same number of women and men were assigned to each of four phase lasted 15 min, including 8 min of practice and 7 min of rest.
experimental groups. According to self-reports, 110 subjects were All other aspects of the procedure and design were similar to
right-handed, 14 were left-handed, and 4 were ambidextrous. Experiment 2.
Apparatus and stimulus materials. The same apparatus and stimu-
lus materials used in Experiment 2 were used in Experiment 3. Results and Discussion
Procedure and design. The fundamental procedure and design
were the same as for Experiment 2, with the exception of the following Implicit memory. As shown in Table 3, subjects in all
changes (see Verdolini-Marston, 1991, for further details). groups improved in pursuit rotor performance from training to
Subjects in the concentrate group received instructions that were test phases (M = 28.34 s for critical stimuli during the training
similar to those in Experiment 2, with two exceptions that were
phase, M = 33.99 s for old stimuli during the test phase, and
expected to increase attention to perceptual aspects of the task by
reducing competing operations. First, in addition to concentrating on
5
the pursuit rotor circle—light ensemble, subjects were specifically Specifically, in this pilot study, 12 of 20 subjects preferred the
instructed to exclude extraneous thoughts and emotions. Second, locomotive instructions to stir instructions in terms of perceived
subjects were also instructed to use "free and easy breathing" as an aid usefulness for how to do the pursuit rotor task, as compared with 6
to limiting performance apprehensions that might distract attention subjects who preferred the stir instructions. Further details are
from the perceptual-motor task. available from K. Verdolini-Marston and David A. Balota.
746 KATHERINE VERDOLINI-MARSTON AND DAVID A. BALOTA

M = 33.60 s for new stimuli during the test phase). More dotal claims about benefits for a perceptual-motor task with
important, however, subjects in the mental strategy groups appropriate elaborative processing (metaphoric) strategies.
showed little evidence of a performance advantage for old Explicit memory. The results from the explicit recognition
stimuli over new stimuli during the test phase (i.e., item- memory test were similar to those of Experiment 2. That is, as
specific priming). For these groups, average TOT advantages shown in Table 3, subjects in the album group (.79) performed
for old stimuli during the test phase ranged from —0.04 s better than did subjects in the other groups on the recognition
(album group) to 0.40 s (concentrate group). Individual t tests task (for locomotive, .59; for concentrate, .60; and for no-
confirmed that priming was unreliable for all of the mental instruction, .54). The main effect of group was significant in a
strategy groups. However, as in Experiment 1, subjects who did one-way, berween-subjects ANOVA, F(3, 124) = 13.42, p <
not receive any instructions about mental strategies did pro- .0001, MSC - 0.028. Post hoc Tukey comparisons confirmed
duce reliable priming; the TOT advantage for old stimuli over that average recognition performance for the album group was
new stimuli during the test phase was 1.25 s for the no- reliably higher than the average performance for each of the
instruction group, t(31) - 3.37, p < .01. Furthermore, al- remaining three groups. However, none of the other groups
though a 2 x 4 (Old-New Status x Group) mixed-factor differed from each other. Individual z tests indicated that only
ANOVA indicated that the main effects of old-new status the average recognition score for the album group was reliably
(priming), F(l, 124) = 3.40, p < .07, MSe = 2.95, and the above chance, z - 3.96, p < .01. Thus, elaborative processes
interaction of old-new status by group, F(3, 124) = 1.97, p < that emphasized which stimuli were presented (album strat-
.13, MSe = 2.95, failed to reach significance, a preplanned egy) again resulted in superior explicit memory performance,
comparison comparing priming for the no-instruction group and elaborative processes that emphasized how to do the
with the mean of the priming effects across the mental strategy pursuit rotor task (locomotive strategy) and perceptual pro-
groups did yield a reliable difference, f(124) = 2.29, p < .03. cesses (concentrate strategy) again resulted in performance
Thus, despite attempts to improve the effectiveness of mental that did not reliably differ from chance.
strategies and to increase the sensitivity of priming measures in
general, item-specific priming failed to develop in mental
strategy groups, as in Experiment 2, but such priming was General Discussion
again evident for the no-instruction group, as in Experiment 1. The primary goal of our investigation was to provide
It should be noted that the performance advantage for old information regarding the role of distinct types of elaborative
stimuli as shown by the no-instruction group, but not the processing in both implicit and explicit perceptual-motor
mental strategy groups, is attributable to superior item-specific memory performance. On the basis of arguments from a
benefits (improvements from training stimuli to old stimuli transfer-appropriate processing framework, we assumed that
during the test phase) in the no-instruction group, rather than elaborative processes that emphasized which specific stimuli
to superior generalized benefits (improvements from training were encountered would primarily facilitate explicit memory
stimuli to new test stimuli) in the mental strategy groups. That performance, whereas elaborative processes that emphasized
is, subjects in the no-instruction group improved more from how to perform the perceptual-motor task would primarily
training stimuli to old test stimuli (M = 7.12 s for the no- influence implicit-memory performance. On the other hand,
instruction group, as compared with M = 4.95 s, 5.13 s, and according to the dual-process (dissociation) model, implicit
5.43 s, for concentrate, locomotive, and album groups, respec- and explicit memory performance are best distinguished by
tively). A preplanned comparison indicated that the superior different processing levels. Hence, one should expect explicit
item-specific improvement for the no-instruction group, as memory performance to be primarily modulated by elaborative
compared with the mental strategy groups, was reliable, processes, whereas implicit memory performance should be
f(124) = 2.76, p < .01. However, all groups improved in a relatively uninfluenced by distinct levels of elaborative process-
similar fashion from training stimuli to stimuli that were new in ing and possibly be best under conditions that emphasize
the test phase (M = 5.87 s for the no-instruction group and perceptual-integrative processes.
M = 4.55 s, 5.15 s, and 5.47 s for concentrate, locomotive, and Despite our repeated attempts, the results fail to show a
album groups, respectively). A preplanned comparison compar- common reliance of explicit and implicit memory on elabora-
ing the no-instruction group with the mental strategy groups tive processes. Elaborative processing conditions that empha-
supported this observation, t(\24) = 1.08, p < .29. Thus, as sized which stimulus was stored (i.e., the album conditions in
noted, the evidence of priming in the no-instruction group, but Experiments 2 and 3) consistently produced the highest level
not in the mental strategy groups, was related to superior of explicit memory performance. Conversely, despite anec-
item-specific benefits in the no-instruction group, as opposed dotal claims about the benefits of elaborative (metaphoric)
to superior generalized benefits in the mental strategy groups. processing strategies for perceptual-motor tasks (Fleshman,
Finally, it is noteworthy that there was no reliable difference 1984; Lessac, 1967), there was no evidence that any elaborative
across any of the groups in a measure of generalized skill processing condition, including two that emphasized how to do
development. Specifically, the improvement between perfor- the pursuit rotor task (stir condition in Experiment 2 and
mance in the training phase and the new items in the test phase locomotive condition in Experiment 3), produced higher
did not reliably vary as a function of group, F(3, 124) = 0.73, implicit memory performance than did any other condition
p < .53, MSe = 6.76. As in Experiment 2, the failure tofindany (e.g., the concentrate or album conditions in Experiments 2
and 3). In fact, implicit memory, as reflected by item-specific
benefit of mental strategy group on either generalized or
priming (without corresponding above-chance explicit memory),
item-specific memory performance is inconsistent with anec-
 PERCEPTUAL-MOTOR PERFORMANCE 747

was evident only when no instructions of any type were stimuli. Thus, implicit memory performance should be similar
imposed (no-instruction groups in Experiments 1 and 3). across instructional conditions, or possibly superior, following
Thus, our results appear most consistent with the view that conditions emphasizing perceptual processes (e.g., concen-
explicit and implicit memory performance in the perceptual- trate conditions). In the current experiments, neither of these
motor domain are regulated by qualitatively different types of predictions was confirmed. Implicit perceptual-motor memory
processing modalities. (See also Heindel, Butters, & Salmon, manifestations (item-specific priming) developed only when
1988, for some evidence of a neuroanatomical dissociation mental strategies were not imposed. Therefore, the most
between explicit and implicit memory performance, including straightforward explanation of our results for implicit memory
information about a pursuit rotor task.) Specifically, the results is that it depends on spontaneous or nonstrategic processing,
indicate that explicit memory (recognition) performance for a or both.
perceptual-motor task appears to depend primarily on elabo- Of course, one might ask whether one can reconcile these
rative processes that individuate target events. Conversely, results with the dual-process model. Perhaps this model can
item-specific implicit perceptual-motor memory as reflected account for ourfindings,with a single qualifier. The qualifier is
by priming appears to depend primarily on spontaneous or not related to domain (perceptual-motor vs. verbal), as might
nonstrategic processes or both. Arguments for a dissociation seem obvious, but is related to our use of novel stimuli
approach are particularly strong on the basis of the present compared with the use of familiar stimuli in previous studies
results because a condition that benefited explicit memory supporting the dual-process model. That is, as suggested by
performance (album condition in Experiments 2 and 3), as this model, implicit memory phenomena (item-specific prim-
compared with the no-instruction control (Experiment 3), ing) depend on the integrative processing of perceptual frag-
resulted in depressed implicit memory performance. Although ments. Possibly, integrative processing requires full attentional
rarely reported in previous studies, similarfindingswere noted allocation when novel stimuli, such as those we used, are
by Schacter and colleagues (Schacter et al., 1990, Experiment presented, whereas full attentional allocation may not be as
2) in a study on memory for novel three-dimensional objects. crucial for the familiar verbal stimuli typically used in the
In that study, elaborative encoding of the objects benefited previous studies. The final piece of the argument is that
explicit memory (recognition) performance for the objects but imposed strategies essentially function as secondary tasks.
resulted in impaired implicit memory performance (priming According to this logic, priming should in fact occur when
on a task requiring possible-impossible object decisions). mental strategies are imposed and the stimuli are familiar to
Schacter et al. (1990) concluded that implicit memory perfor- subjects, because the processes that support priming do not
mance was impaired with the imposition of an (elaborative) require full attentional allocation that might be disrupted by
encoding strategy because the strategy interfered with percep- imposed strategies. In contrast, priming should not occur when
tual processing. Thus, the finding of a strategy that benefits mental strategies are imposed along with novel stimuli be-
explicit memory performance while disrupting implicit memory cause, in this case, the perceptual processes that support
performance is not new to this literature. priming require full attentional allocation and are disrupted
Now consider our results within the specific dual-process when strategies are imposed.
model that motivated the present research (e.g., as originally Of course, this interpretation is post hoc. However, there
proposed by Mandler and his colleagues; Graf et al., 1982; are a series offindingsin the literature that appear to support
Mandler, 1979, 1980). For explicit memory, the findings are this interpretation. For example, there is clear evidence that
consistent with the suggestion that explicit memory perfor- attention plays a diminishing role as a function of practice for
mance depends on elaborative processes. However, on the both verbal and nonverbal tasks. Performance decrements are
basis of the present results, a caveat is necessary. Elaborative anticipated with secondary loads in early phases of acquisition,
processes benefited explicit memory performance, but only whereas increasingly diminished decrements are anticipated in
under conditions in which these processes uniquely specified later phases (see, e.g., LaBerge & Samuels, 1974; Schneider &
target events. Specifically, elaborative processing conditions Fisk, 1982). More relevant to the present discussion are
that individuated target events (album conditions) consistently reports about the role of attention for implicit memory
resulted in superior explicit memory performance, as com- performance across varying acquisition levels. For example,
pared with elaborative processing conditions that emphasized according to a study by Nissen and Bullemer (1987), item-
how to do the pursuit rotor task (stir and locomotive condi- specific performance benefits (essentially priming) for a novel,
tions). In fact, the elaborative conditions that emphasized how repeated, 10-movement sequence were disrupted under second-
to do the task did not reliably yield above-chance explicit ary load conditions. Along the same lines, Ferraro, Balota, and
memory performance. Of course, this pattern is quite consis- Connor (1993) reported that individuals with Alzheimer's
tent with theoretical approaches emphasizing the critical role disease (subjects with well-documented attentional break-
of individuating operations for episodic memory (Damasio, downs; cf. Nestor, Parasuraman, & Haxby, 1991) also produce
1989; Tulving, 1972) and extends the importance of this factor decreased item-specific priming in the Nissen and Bullemer
beyond the verbal domain to the perceptual-motor domain. Task. In contrast, priming for familiar verbal stimuli and for
Regarding implicit memory performance, our results appear pictures of familiar objects is apparently unimpaired under
inconsistent with the most straightforward predictions from secondary load conditions (Parkin, Reid, & Russo, 1990;
the dual-process model. Specifically, according to this model, Parkin & Russo, 1990). Moreover, priming of familiar represen-
implicit memory depends on integrative or perceptual pro- tations also appears to be unimpaired in individuals with
cesses that occur whenever subjects are exposed to target Alzheimer's disease (Balota & Duchek, 1991; Moscovitch,
748 KATHERINE VERDOLINI-MARSTON AND DAVID A. BALOTA

Winocur, & McLachlan, 1986).6 The results across these Butters, N., Heindel, W. C , & Salmon, D. P. (1990). Dissociation of
studies imply that attention becomes increasingly less impor- implicit memory in dementia: Neurological implications. Bulletin of
tant for critical integrative processes to occur as acquisition the Psychonomic Society, 28, 359-366.
level advances. Hence, at this level, our results may be viewed Cohen, N. J., & Squire, L. R. (1980). Preserved learning and retention
of pattern-analyzing skill in amnesia: Dissociation of knowing how
as still consistent with the dual-process model. For novel
and knowing that. Science, 210, 207-210.
stimuli, perceptual processes that support priming require
Damasio, A. R. (1989). Time-locked multiregional retroactivation: A
attentional allocation and are interfered with when strategies systems-level proposal for the neural substrates of recall and
are imposed. For familiar stimuli, perceptual processes can recognition. Cognition, 33, 25-62.
occur even under secondary load conditions (e.g., imposed Ferraro, F. R., Balota, D. A., & Connor, L. T. (1993). Implicit memory
encoding instructions, wherein attentional allocation is some- and the formation of new associations in non-demented Parkinson's
what disrupted). Of course, this explanation requires consider- disease individuals and individuals with senile dementia of the
able and further investigation by using tasks beyond the Alzheimer type: A serial reaction time (SRT) investigation. Brain
pursuit rotor and others described here. and Cognition, 21, 163-180.
To summarize, it is possible to view our results as consistent Fleshman, B. (1984). Mime and movement training with mental
with the original dual-process model. That is, it may be the imagery. Journal of Mental Imagery, 8, 91-98.
Graf, P., Mandler, G., & Haden, P. E. (1982). Simulating amnesic
case that perceptual-integrative processes are the crucial
symptoms in normal subjects. Science, 218, 1243-1244.
factor for implicit memory performance. For familiar stimuli,
Graf, P., & Ryan, L. (1990). Transfer-appropriate processing for
such as those used in most previous studies, such processes implicit and explicit memory. Journal of Experimental Psychology:
may occur with minimal attentional allocation and thus are not Learning, Memory, and Cognition, 16, 978-992.
disrupted by the imposition of mental strategies. For novel Graf, P., & Schacter, D. L. (1985). Implicit and explicit memory for
perceptual-motor stimuli, such as those used in our study, the new associations in normal and amnesic subjects. Journal of Experi-
critical perceptual processes may require fuller allocation and mental Psychology: Learning, Memory, and Cognition, 11, 501-518.
are thus disrupted when mental strategies are imposed, result- Gross, D. A., Wilson, R. S., & Fox, J. H. (1990). Preserved word-stem-
ing in impaired priming. completion priming of semantically encoded information in
In conclusion, our results are consistent with a general Alzheimer's disease. Psychology and Aging, 5, 304-306.
Heindel, W. C , Butters, N., & Salmon, D. P. (1988). Behavioral
dissociation approach to explicit and implicit memory and
Neuroscience, 102, 141-147.
extend the dissociation approach beyond the verbal domain. Jacoby, L. L., & Dallas, M. (1981). On the relationship between
Explicit perceptual-motor memory performance appears to autobiographical memory and perceptual learning. Journal of Experi-
depend on elaborative processes that individuate target stimuli. mental Psychology: General, 110, 306-340.
Implicit perceptual-motor memory performance for novel Jacoby, L. L., & Witherspoon, D. (1982). Remembering without
stimuli appears to depend on spontaneous or nonstrategic awareness. Canadian Journal of Psychology, 36, 300-324.
processes, or both. One goal of future research in this area LaBerge, D., & Samuels, S. J. (1974). Toward a theory of automatic
should be a clearer understanding of the role of attentional information processing in reading. Cognitive Psychology, 6, 293-323.
allocation for priming with familiar and unfamiliar patterns. Lessac, A. (1967). The use and training of the human voice: A practical
approach to speech and voice dynamics. New York: Drama Book
Publishers.
6
There is evidence that individuals with Alzheimer's disease (AD) Mandler, G. (1979). Organization and repetition: Organizational
do produce some deficit in priming as measured by stem-completion principles with special reference to rote learning. In L. G. Nilsson
measures (see review by, Butters, Heindel, & Salmon, 1990). This (Ed.), Perspectives on memory research (pp. 293-397). Hillsdale, NJ:
indicates that AD individuals do produce breakdowns even in the Erlbaum.
priming of familiar representations such as words and pictures. There Mandler, G. (1980). Recognizing: The judgment of previous occur-
are two points to note here: First, it is possible that AD individuals rence. Psychological Review, 87, 252-271.
produce the deficits in stem completion-type tasks because this task in Martone, M., Butters, N., Payne, M , Becker, J. T., & Sax, D. S. (1984).
general produces a high load on attentional resources. Second, Gross, Dissociations between skill learning and verbal recognition in
Wilson, and Fox (1990) have recently reported preserved word-stem amnesia and dementia. Archives of Neurology, 41, 965-970.
completion of semantically encoded information in AD individuals. Morris, C. D., Bransford, J. D., & Franks, J. J. (1977). Levels of
When a breakdown in the priming of familiar representations occurs in processing versus transfer appropriate processing. Journal of Verbal
a stem-completion task, it may be due to the possibility that AD Learning and Verbal Behavior, 16, 519-533.
individuals are less likely to fully encode the primes. Moscovitch, M., Winocur, G., & McLachlan, D. (1986). Memory as
assessed by recognition and reading time in normal and memory-
impaired people with Alzheimer's disease and other neurological
disorders. Journal of Experimental Psychology: General, 115, 331-347.
References Nestor, P. G., Parasuraman, R., & Haxby, J. V. (1991). Speed of
information processing and attention in early Alzheimer's disease.
Balota, D. A., & Duchek, J. M. (1991). Semantic priming effects, Developmental Neuropsychology, 7, 243-256.
lexical repetition effects, and contextual disambiguation effects in Nissen, M. J., & Bullemer, P. (1987). Attentional requirements of
healthy aged individuals and individuals with senile dementia of the learning: Evidence from performance measures. Cognitive Psychol-
Alzheimer type. Brain and Language, 40, 181-201. ogy, 19, 1-32.
Blaxton, T. A. (1989). Investigating dissociations among memory Parkin, A. J., Reid, T. K., & Russo, R. (1990). On the differential
measures: Support for a transfer-appropriate processing frame- nature of implicit and explicit memory. Memory and Cognition, 18,
work. Journal of Experimental Psychology: Learning, Memory, and 507-514.
Cognition, 15, 657-668. Parkin, A. J., & Russo, R. (1990). Implicit and explicit memory and the
 PERCEPTUAL-MOTOR PERFORMANCE 749

automatic/effortful distinction. European Journal of Cognitive Psychol- Squire, L. R., & Cohen, N. J. (1984). Human memory and amnesia. In
ogy, 2, 71-80. G. Lynch, J. L. McGaugh, & N. M. Weinberger (Eds.), Neurobiology
Roediger, H. L., HI. (1990). Implicit memory: Retention without of learning and memory (pp. 3-64). New York: Guilford Press.
remembering. American Psychologist, 45, 1043—1056. Tulving, E. (1972). Episodic and semantic memory. In E. Tulving & W.
Roediger, H. L., Ill, Srinivas, K., & Weldon, M. S. (1989). Dissocia- Donaldson (Eds.), Organization of memory (pp. 381—403). San
tions between implicit measures of retention. In S. Lewandowski, J. Diego, CA: Academic Press.
C. Dunn, & K. Kirsner (Eds.), Implicit memory: Theoretical issues Tulving, E., Schacter, D. L., & Stark, H. A. (1982). Priming effects in
(pp. 67-84). Hillsdale, NJ: Erlbaum. word-fragment completion are independent of recognition memory.
Roediger, H. L., Ill, Weldon, M. S., & Challis, B. H. (1989). Journal of Experimental Psychology: Learning, Memory, and Cognition,
Explaining dissociations between implicit and explicit measures of 8, 336-342.
retention: A processing account. In H. L. Roediger III & F. I. M. Verdolini-Marston, K. (1991). Processing characteristics of perceptual-
Craik (Eds.), Varieties of memory and consciousness: Essays in honour motor memories with and without awareness. Unpublished doctoral
ofEndel Tulving (pp. 3—41). Hillsdale, NJ: Erlbaum. dissertation, Washington University, St. Louis.
Schacter, D. L., Cooper, L. A., & Delaney, S. M. (1990). Implicit Warrington, E. K., & Weiskrantz, L. (1974). The effect of prior
memory for visual objects and the structural description system. learning on subsequent retention in amnesic patients. Neuropsycho-
Bulletin of the Psychonomic Society, 28, 367-372. logia, 12, 419-428.
Schacter, D. L., Cooper, L. A., Delaney, S. M., Peterson, M. A., & Zola-Morgan, S., Squire, L. R., & Mishkin, M. (1982). The neuro-
Tharan, M. (1991). Implicit memory for possible and impossible anatomy of amnesia: Amygdala-hippocampus versus temporal stem.
objects: Constraints on the construction of structural descriptions. Science, 218, 1337-1339.
Journal of Experimental Psychology: Learning, Memory and Cognition,
17, 3-19.
Schneider, W., & Fisk, A. D. (1982). Degree of consistent training:
Improvements in search performance and automatic process devel- Received April 14,1992
opment. Perception and Psychophysics, 31, 160-168. Revision received July 21,1993
Squire, L. R. (1986). Mechanisms of memory. Science, 232, 1612-1619. Accepted July 22,1993