2006 • volume 2 • no 2-3 • 113-130 Advances in Cognitive Psychology

Practicing perfection: How concert soloists

prepare for performance

Roger Chaffin and Topher Logan

University of Connecticut

Received 03.04.2006
Accepted 15.16.2006

Keywords
music performance, expert memory, skill learning

ABSTRACT musician attends to these performance cues in

order to ensure that the performance unfolds
Musical performances by concert soloists in as planned. Performance cues are established
the Western classical tradition are normally by thinking about a particular feature of the
memorized. For memory to work reliably under music during practice so that it later comes
the pressures of the concert stage, the perfor- to mind automatically. Performance cues help
mance must be practiced until it is thoroughly the soloist consciously monitor and control the
automatic. At the same time, the performance rapid, automatic actions of playing, while ad-
must be fresh and spontaneous in order to justing to the needs of the moment. During
communicate emotionally with the audience. practice, the musician attends mostly to basic
The resolution of this apparent contradiction is performance cues representing critical techni-
provided by longitudinal case studies of concert cal features (e.g., fingerings), and interpretive
soloists preparing new works for performance. performance cues, representing phrasings, and
Like expert memorists in other domains, expe- changes in dynamics, tempo, and timbre. Dur-
rienced musicians use highly practiced retriev- ing performance, the musician hopes to attend
al schemes to accomplish their extraordinary mainly to expressive performance cues repre-
feats of memory. Performers have a mental map senting the musical feelings to be conveyed to
of the piece in mind as they perform that tells the audience (e.g. excitement). We illustrate
them where they are and what comes next – a this analysis with a typical case study of a con-
series of landmarks, hierarchically organized by cert pianist learning J.S. Bach’s Italian Concer-
the sections and subsections of the music. The to (Presto).

INTRODUCTION second for minutes on end. How does a performer do

this: remembering it all, hitting every note, and at the
We have all experienced the fascination and awe of
same time creating beautiful music? On the one hand,
witnessing a world class performance, whether a mu-
extended practice is needed if memory is to operate
sician in a virtuoso rendition, an ice skater making
reliably (Ericsson & Kintsch, 1995), particularly under
triple axel leaps, or a kayaker hurtling down a class six
the pressure of the concert stage. On the other hand,
rapid. And most of us have marveled at the skill that
audiences value the feeling of freshness and spontane-
makes such feats possible. For example, the perform-
ance of even a moderately complex piano piece places Correspondence concerning this article should be addressed
to Roger Chaffin, Department of Psychology U-1020 Univer-
incredible demands on memory and physical dexterity,
sity of Connecticut Storrs CT 06269-1020 USA email: Roger.
requiring the execution of between 10 and 20 notes a Chaffi[email protected] tel: 860-529-5428

113
http://www.ac-psych.org
 Roger Chaffin and Topher Logan

ity (Repp, 1998; Sundberg, Friberg, & Fryden, 1991). 1995), intuitive problem solving (Gobet & Simon, 1996),
How does a performer resolve these apparently con- and motor skill (Van Orden, Holden, & Turvey, 2003).
tradictory demands? Experts in any domain memorize with a facility that
The answer comes from longitudinal case studies of seems superhuman (Chase & Simon, 1973, Gobet &
concert soloists preparing for performance. Music prac- Simon, 1996). Musicians are no exception; their bi-
tice naturally produces a detailed behavioral record that ographies are full of tales of amazing memory feats,
is not available for the learning of most types of highly such as the story of the young Mozart’s ability to
skilled behavior. During practice, musicians constantly write from memory Allegri’s Miserere after two hear-
start, stop, back up, and repeat, making split-second ings (Cooke, 1999) or Jorge Bolet’s memorization of
decisions as they review every aspect of technique, in- Liszt’s Mephisto Waltz in an hour and a quarter (Noyle,
terpretation, and performance. The behavioral record 1987). The abilities of other expert memorists have
generated in practice provides a window into the cogni- been attributed to the use of highly practiced retrieval
tive processes involved in developing a high level skill. strategies (Ericsson & Kintsch, 1995). In describing
By itself, however, the behavioral record is relatively memory for skilled activity, Ericsson & Kintsch argue
uninformative unless supplemented by an understand- that the end products of processing can be stored in
ing of the musician’s goals (Chaffin & Imreh, 2001). long term memory and accessed through retrieval
Experienced musicians have a detailed understanding cues in short term memory. It is not, however, obvi-
of the factors that go into their decisions during practice ous that principles of expert memory derived from the
(Hallam, 1995a, 1995b) and in the studies reviewed study of memory for chess boards (Chase & Ericsson,
here, this understanding provides an important source 1982; Chase & Simon, 1973), digit strings (Chase &
of information. The musicians were asked to provide Ericsson, 1982; Thompson, Cowan & Friedman, 1993),
detailed, retrospective reports about all of the features and dinner orders (Ericsson & Oliver, 1989) apply to
of the music they thought about during practice. They musical performance. Motor and auditory memory
marked on copies of the score every point where they play crucial roles in musical memory but not in these
had thought about each aspect of the music: structure, other domains. In spite of the large differences, the
basic technique, interpretation, and performance. When case studies of concert soloists reviewed here suggest
these reports were correlated with measures of practice, that the same principles apply.
recall, and performance, those aspects of the music af- The feats of expert memorists have been explained
fecting each activity could be identified. For example, if in terms of three principles: meaningful encoding of
playing started or stopped at places where the musician novel material, use of a well learned retrieval structure,
reported needing to recall information from memory, this and extended practice to decrease the time needed for
suggested that the musician was practicing memory re- retrieval from long-term memory (Ericsson & Kintsch,
trieval. In this way, the detailed behavioral record natu- 1995). The same three principles appear to apply to
rally provided by music practice can be interpreted. expert music performance (Chaffin, Imreh & Crawford,
Case studies are the method of choice for studying 2002; Krampe & Ericsson, 1996). First, experts’ knowl-
highly developed skills like those of a concert soloist. edge of their domain of expertise allows them to en-
The 20+ years of training required for a career on the code new information in terms of ready-made chunks
concert stage increases the normal range of individual already stored in memory (Miller, 1956; Tulving, 1962;
differences so that aggregating observations across Mandler & Pearlstone, 1966). For a musician, these in-
individuals risks obscuring the phenomena of interest clude familiar patterns like chords, scales, and arpeg-
(Ericsson & Oliver, 1988). Here we describe conclusions gios, whose practice forms an important part of every
about expert musical performance based primarily on musician’s training (Halpern & Bower, 1982). Second,
case studies of a concert pianist (Chaffin & Imreh, expert memory in any domain requires a retrieval
2001, 2002, Chaffin, Imreh & Crawford, 2002; Chaffin, scheme to organize the cues that provide access to the
Lemieux & Chen, 2004, 2006; Chaffin, Imreh, Lemieux chunks of information in long-term memory (Ericsson
& Chen, 2003), a jazz pianist (Noice, Chaffin, Jeffrey & Oliver, 1989). For a musician, the formal structure
& Noice, in press), a cello soloist (Lisboa, Chaffin, of the music conveniently provides a ready-made hie-
Schiaroli & Barrera, 2004), and a soprano and con- rarchical organization to serve as a retrieval scheme
ductor (Ginsborg, Chaffin & Nicholson, 2006). We use (Chaffin & Imreh, 1997, 2002; Williamon & Valentine,
these data to test the application to music of theories 2002). For example, Figure 1 shows the structure
developed to account for skilled memory in other do- of one of the pieces whose preparation was the sub-
mains, such as chess and acting, (Ericsson & Kintsch, ject of a case study, the Italian Concerto (Presto)

114
http://www.ac-psych.org
Practicing perfection

by J.S. Bach. The piece is hierarchically organized into ITALIAN CONCERTO

movements, sections, sub-sections, and bars within
Movement
which more specific retrieval cues (performance cues) Allegro Andante Presto

are embedded. The third principle of expert memory is

Section
that prolonged practice is required to bring the speed
of operation of this retrieval scheme up to the point A A1 B B1 A2 C A3 D A4 B2 A5 A6 C1 B3 A A1

where an expert can rely on long-term memory to per- Subsection

Ca1 Ca2 Cb
form tasks for which most people would use working
memory (Ericsson & Kintsch, 1995). For the musician,
Bars
this involves practicing memory retrieval until it is
1 2 3 4
rapid and reliable enough to keep pace with the tempo
required for performance. Expressive Performance Cues
"Light but mysterious" "Surprise"
Rapid memory retrieval is important in music per-
formance to prevent the hands from “running away” as Interpretive Performance Cues

motor performance outstrips the corresponding con- Piano Subito forte

(soft) (suddenly forceful)
ceptual representation in working memory. Practice is Basic Performance Cues
needed to coordinate motor performance with retrieval flourish in feft hand
from declarative long-term memory so that the mem-
ory for what comes next arrives in working memory at Figure 1.
Hypothetical hierarchical retrieval scheme “unpacked” for
just the right moment, before the corresponding mo- Section C of the Presto. Main themes (sections) are rep-
tor sequences, but not so far in advance as to interfere resented by capital letters. Section C is “unpacked” into
subsections (Ca1, Ca2, Cb). Subsection Ca1 is further “un-
with the execution of the preceding notes.1 A goal of packed” into its performance cues (adapted with permis-
the preparation for public performance is to devise a sion, from Chaffin et al., 2002, p. 200)

plan that ensures that memory does not fail while un-
der the pressure of a live performance. In an attempt switching point I would automatically think about where
to explain the mechanisms by which performers plan I was, and the switch would go” (Chaffin et al., 2002,
production sequences, Palmer and Pfordresher (2003) p. 224).
developed an incremental planning model, which We call the kind of memory retrieval cues that the
treats memory for serial order in music performance musician is talking about here, performance cues.
as response preparation rather than response selec- Performance cues are the landmarks of the piece that
tion. In their framework, serial ordering in music is an experienced musician attends to during perform-
determined by contextual associations among pitch ance, carefully selected and rehearsed during practice
events, such as those prescribed by rhythm and meter. so that they come to mind automatically and effort-
They argue that response preparation for a sequence lessly as the piece unfolds, eliciting the highly practiced
of events is in part a function of metrical similarity (the movements of fingers, hands, and arms. Performance
similarity of metrical accent strength between events) cues become an integral part of the performance and
and serial proximity between events. By fitting the provide a means of consciously monitoring and con-
predictions of the model to behavioral data, they de- trolling the rapid, automatic actions of the hands.
termined that errors increased with metrical similarity Performers are faced with a paradox. On the one
and serial proximity and that the range of planning hand, the performance must be completely automatic
increased with experience and decreased with tempo. to cope with the highly charged atmosphere of the
This finding could potentially help explain why even concert stage. Every action must be automatic or it will
experienced performers often find it difficult to keep be forgotten in the adrenaline rush of stepping out in
head and hands together in a fast piece where there front of an audience. On the other hand, performance
are no pauses or long notes which would allow time is a creative activity, not simply the rote repetition of
to think ahead. As the pianist in a study of the Italian overlearned movements. Most performers would agree
Concerto described below explained: with the eminent Russian pianist Emil Gilels that:
“My fingers were playing the notes just fine. The “When I am in top form… the ideas are always
practice I needed was in my head. I had to learn to different. Sometimes I play with greater changes in
keep track of where I was. It was a matter of learn- dynamics, sometimes with less… I must say it is dif-
ing exactly what I needed to be thinking of as I played, ferent each time I play, and it is a process which…
and at exactly what point so that as I approached a includes mastery of the work, knowing the details, be-

115
http://www.ac-psych.org
 Roger Chaffin and Topher Logan

ing comfortable with it, and then adding the fantasy expressive, interpretative, and basic. The sections
(Mach, 1991, vol. 2, p. 123). and subsections of the formal structure provide the
Gilels apparently considered his ability to vary upper levels of the hierarchy, with expressive cues
his performances to be a reflection of his artistry further dividing subsections into expressive phrases.
and skill. Gilels is not talking here about the sort of Interpretive and basic performance cues are embed-
random variability that is present in any psychologi- ded within this hierarchical organization, providing
cal process (Gilden, 2001), but about differences in different types of information about critical points of
the musical gestures of his interpretation. These are interpretation and technique.
the carefully calibrated nuances of timing, dynamics, Different musicians are likely to use the same struc-
and timbre that are responsible for the ritardandos, tural cues for a particular piece (although Ginsborg et
crescendos, and phrasings that bring a piece of music al., 2006 describe a case where a singer and conduc-
to life (Shaffer, 1984; Shaffer, Clarke & Todd, 1985; tor understood the musical structure slightly differ-
see Gabrielsson, 1999, for a review). Each performer’s ently), and to differ more from one another at lower
interpretation is unique, and soloists’ reputations de- levels in this hierarchy. The basic performance cues
pend on their ability to develop interpretations that are provide the most unique information and at this level
both distinctive and convincing. Flexibility is a general the requirements of different performers and different
characteristic of skilled musical performance as it is of instruments are likely to differ widely. For example,
other motor skills (Latash, Scholz & Schoner, 2002; basic performance cues for a cellist include issues of
Todorov & Jordan, 2002; Van Orden et al., 2003). For intonation, bowing, and changing string (Lisboa et al.,
the musician, flexibility is necessary to adapt to the 2004) that are not relevant for a pianist (Chaffin et
idiosyncrasies of each instrument, hall, and audience, al., 2002). For solo works, the only performance cues
to provide a feeling of freshness and spontaneity, and required are those for the individual musician, while
to recover from the mistakes that are an inevitable for ensemble performance the musicians must also
aspect of any live performance. establish shared performance cues to coordinate their
Performance cues provide flexibility by allowing the actions (Ginsborg et al., 2006)
performer to remain mindful of a memorized perform- In learning a new piece, the musician moves up and
ance that has become automatic through extended down the hierarchy, attending to each level of organi-
practice. Performance cues allow the musician to at- zation and each type of cue in turn at different points in
tend to some aspects of the performance while allow- the learning process (Williamon, Valentine & Valentine,
ing others to be executed automatically (Chaffin et al., 2002). Experts in other fields approach a new problem,
2006). Performance cues are the retrieval cues that by looking at the “big picture” (Chi, Feltovich & Glaser,
elicit the knowledge of what comes next from long 1981; Glaser & Chi, 1988; Gobet & Simon, 1996).
term memory, providing the musician with a mental Musicians approach the task of learning a new piece in
map of the piece in working memory that continuously the same way, by getting an overall “artistic image” of
unfolds as the performance progresses. Careful prepa- how the music should sound (Neuhaus, 1973; Chaffin
ration of these cues makes it possible for soloists to et al., 2003). During this initial stage, the musician
reliably perform challenging works from memory on identifies the formal structure of the piece as well as
the concert stage. many of the places that will later become expressive,
We can distinguish four types of performance cue. interpretive, and basic performance cues. In later prac-
Structural cues are critical places in the formal struc- tice sessions, the musician focuses on basic technique
ture of the music, such as section boundaries, where and then on interpretation, revisiting each repeatedly
musical material changes. Expressive cues represent as learning progresses. When the piece is ready, at-
musical feelings to be conveyed to the audience, e.g., tention returns to expressive performance cues during
surprise or excitement. Interpretive cues are places final preparation for a public performance. Ideally, the
where some aspect of interpretation requires atten- musician performs with expressive and structural cues
tion, e.g., a change of tempo or dynamics. Basic cues in the spotlight of attention with basic and interpre-
represent the critical details of technique or musi- tive performance cues in the background, ready to be
cal structure that must be executed exactly for the called upon if needed.
performance to unfold as intended, e.g., the use of In the studies reviewed here, four types of evi-
a particular fingering in order to set up the hand up dence show that musicians attend to different types
for what follows. The four types of performance cue of performance cues as their memory for a piece
are organized hierarchically, as in Figure 1: structural, develops and they continue to attend to them dur-

116
http://www.ac-psych.org
Practicing perfection

ing polished performances at the end of the learning was thinking about as she practiced and the changes
process. First, the musicians’ spontaneous comments in the kind of cue she attended to as her mastery of
to the video camera taping their practice indicate the the piece developed. Third, recall of the score after
issues that are the focus of their problem solving ef- the final performance reveals the organization of the
forts at that moment. Second, the locations of starts, music in the pianist’s memory at the end of the learn-
stops and repetitions during practice reveal which as- ing process. Fourth, a comparison of practice perform-
pects of the music the musician was attending to dur- ances near the end of the learning process with the
ing practice, both those that they were deliberately performance released on the CD demonstrate the kind
thinking about as well as those that affected their of flexibility the pianist was able to introduce into her
activities more automatically. Third, tests of recall of polished performances.
the musical score after the final performance indicate The pianist had played Bach throughout her career
how the piece was organized in a performer’s mem- and had taught the Italian Concerto to a student three
ory when it was last played. Fourth, a comparison of years before, but had never performed the piece her-
bar-to-bar fluctuations in the tempo and dynamics of self before the start of the study. The Presto is of mod-
polished performances indicates which aspects of the erate difficulty (Hinson, 1987), is scored in 210 bars
performance can be spontaneously altered from one in 37 sections, notated in 2/4 time, and lasts for 3-4
occasion to another. minutes at performance tempo. The pianist recorded
her practice from the first time she sat down at the
piano until the piece was performed without the score
A CASE STUDY: THE ITALIAN
at the recording session.
CONCERTO (PRESTO)
We will illustrate what can be learned about how ex-
Description of the learning process
perienced musical soloists memorize by describing one
study of a concert pianist, Gabriela Imreh, learning the Preparation of the piece took 57 practice sessions,
Italian Concert (Presto) by J.S. Bach for professional totaling 33 hours 25 minutes, over 10 months in-
recording as part of an all-Bach CD (Imreh, 1996). cluding two breaks totaling almost 6 months dur-
[The entire Presto can be heard at http://psychlops. ing which the piece was not practiced (see Table 1).
psy.uconn.edu/rchaffin/Italianconcerto(presto).mp3] After an initial sight reading through the whole piece,
First we will describe three occasions on which the practice began with the pianist working her way
pianist talked to the camera immediately after a prac- through the piece a few sections at a time making
tice performance, describing what she was attending fingering decisions and establishing motor memory
to as she played. Second, transcription of the practice through repetition. This continued through session
clearly reveal the performance cues that the pianist 6, at which point there was a three-day break dur-

Table 1.
Distribution of practice sessions across 42 weeks, showing mean session duration and mean length and number of practice
segments (uninterrupted playing of the score) per session.

Session duration Practice Segments

Weeks from Mean length
Session Description Total Mean Mean number
start of practice (bars)
1-6 Section-by-section 1-3 6:20 1:03 15.4 381.7
7-8 First performance from memory 3 2:18 1:09 18.0 362.5
9-16 Developing automaticity 3-20 5:32 0:41 15.5 134.5
BREAK 1 5-19
17 Memorization 20 1:02 1:02 47.6 160.0
18-24 Polishing for first performance 21-22 4:13 0:36 39.3 135.5
BREAK 2 23-30
26-27 Re-learning 30 0:55 0:21 16.3 160.5
28-36 Increasing the tempo 32-36 5:55 0:39 21.0 193.2
37-44 Polishing for second performance 37-41 3:15 0:24 63.0 120.0
45-57 Maintenance 41-42 3:55 0:18 (not recorded) (not recorded)

117
http://www.ac-psych.org
 Roger Chaffin and Topher Logan

ing which the pianist worked on the first movement. to the G ... I have to concentrate on the fingering
When she returned to the Presto in session 7, work in bar 65, the left hand divided between two, four
in each session ranged over the entire piece, rather fingers... I have, oh boy, the scale in the left hand
than being limited to a few sections as before. The at [bar] 124, the two fours in a row... The fingering
goal was now to develop the ability to play through in 186”.
the whole piece fluently and session 8 ended with The third description was at the end of Session 24,
the first practice performance “mostly from mem- the day before the pianist performed the piece in public
ory”. The pianist played the piece through without for the first time. Now the focus was on interpretation;
the score for the first time at the end of session 12 basic cues were hardly mentioned.
to see how well it was memorized before setting the “Then I return to very lightly pianissimo. And again,
piece aside for nearly four months. Developing flu- just the left hand B flat (accented), and then I return
ency continued after the break until, in session 17, to pianissimo... And that gives me again room for a
she began to systematically practice playing from nice crescendo in [bar] 86 and on... ”
memory. Another 4¼ hours of practice were needed There was almost no mention of expression. This
to get ready for the first public performance after does not mean that the public performance the
which she put the piece aside again. next day was not expressive. There was plenty of
When the pianist returned to the piece two expression built into the automatic actions of the
months later, she decided that the tempo needed to performance. But the pianist was not yet focusing on
be faster and put in another seven hours bringing it expression as she played. The first direct mention of
up to the new tempo. By session 36 she reported, expressive performance cues came in a more formal
“The mistakes are starting to fade out a little bit… report made after Session 31. The pianist was asked
It’s coming along” and in session 45, decided that to give some examples of the performance cues she
the piece was ready and stopped recording her was using. Her reply took the form of a diagram
practice. She played through the piece each day un- (see Figure 2) in which she marked the performance
til the recording session two weeks later, however, cues she was using for bars 77-92. Expressive goals
and in sessions 49 and 50 she recorded six practice were mentioned for the first time in this diagram as
performances in order to have a performance on expressive performance cues, along with basic and
videotape to play to audiences during talks about interpretive performance cues. Some bars contained
the research. all three. For example, bar 85 contains an expressive
cue marked, “start building tension”, an interpretive
cue labeled, “start crescendo”, and a basic cue la-
Pianist’s comments after practice
beled “left hand leads, pattern changes, more lega-
performance
to”. The three cues represent three different ways
There were three occasions when the pianist played of thinking about the same point in the music. The
the piece through from memory and described the different ways of thinking would have been salient at
performance cues she was using. The first time she different points in the learning process. In session 17
did this, in Session 12, the only cues mentioned were attention would have been on “Left hand leads…”. In
structural. session 24, the interpretive performance cue at the
“Probably now the seams [between sections] same point was referred to as the “nice crescendo in
are quite obvious... It’s going to take a while to get [bar] 86”. Finally, in the report of performance cues
through this, but it’s good [for me]. Now I have to given after session 31, the pianist identified the ex-
check each transition [between themes] because each pressive effect of the crescendo with the expressive
time it’s something different. That’s the second time, performance cue, “start building tension”.
so ... Oh, I confused them.” The pianist’s descriptions of what she was attend-
The next time was in Session 17, and this time ing to in practice provide the first indication that she
structure was hardly mentioned. Now the focus was attended to different types of performance cues at dif-
on basic cues – technical difficulties, fingerings, and ferent points in the learning progress. The progression
patterns of notes. in her comments suggest that she attended first to
”Eventually at this level you start to have a sort of musical structure, then to basic cues, next to interpre-
map of the piece in your mind and you ... focus on tive cues, and finally to expressive cues. We turn now
certain places in it. I’ll try to tell you... I have a thing to behavioral data for further evidence supporting this
in bar 42 where I have to remember to go all the way conclusion.

118
http://www.ac-psych.org
Practicing perfection

Figure 2.
The performance cues (represented by arrows) that the pianist reported attending to during practice and performance for
bars 77-92

Starts, stops and repeats during performance cues being set up as the pianist worked
practice on this passage for the first time. Starting at a particu-
lar spot establishes a retrieval cue so that later, simply
Another source of information about what the pianist thinking of that spot in the music is sufficient to initiate
was attending to is provided by her practice. Where playing (Chaffin & Imreh, 2002; Chaffin et al., 2003).
did she start and stop; which bars were repeated This conclusion is not based solely on the small sam-
more? Starts, stops, and repetitions provide a reveal- ple of practice in Figure 3. Performance cues affected
ing behavioral record of what the pianist was thinking practice throughout the 10 months of practice with dif-
about during practice. Decisions are made rapidly and ferent types of performance cues affecting practice at
continuously, reflecting the moment-to-moment judg- different points in the learning process.
ments and decisions of the musician. Starting at a par- Effects of each type of performance cue on practice
ticular location requires attention to the music at that were identified by transcribing the practice in each
location, as does deliberate stopping. Stopping, when session and counting the number of starts, stops, and
not deliberate, indicates problems. Repetition shows repetitions for each bar. These three measures served
that a passage was singled out for practice. as dependent variables in regression analyses in which
Figure 3 shows a portion of the practice record for the number of performance cues of each type of per-
session 9 in which the pianist worked on the same formance cue reported for each bar were the predic-
short passage for which we just described the perform- tors. Table 2 summarizes the results of these analy-
ance cues. The record shows that the starts, stops, ses. Most of the effects were positive, indicating that
and repetitions were not random. For example, some the pianist was starting, stopping or repeating bars
bars were used as starting places more than others. containing performance cues significantly more than
What was special about these bars? The performance at other bars not containing cues. Negative effects
cues in Figure 2 provide the answer. The pianist was also indicate that the pianist was paying attention to
starting at performance cues. In Figure 3, we see the the type of cue producing the effect but was avoiding

119
http://www.ac-psych.org
 Roger Chaffin and Topher Logan

Figure 3.
The record of practice during Session 9 of bars 77-92. The record reads from bottom to top, with each line representing the
playing of the music shown below. Each time the pianist stopped and started again the record begins again on the next line
up. The starting places correspond with the location of the performance cues for the passage. In Session 9, the pianist was
setting up the performance cues by using them as starting places.2

Table 2.
Significant effects (p<.05) of musical structure and performance cues on the frequency of Starts (S), Stops (E) and Repeti-
tions (R) as a function of practice session. Positive effects are listed in blue and capital letters, negative effects in red and
lowercase.2

Sessions
1-6 7-8 9-16 17 20-24 26-27 28-36 37-50
Musical Structure
Begin Section S E S S S S S S
Ends Section R E
Switch S E R E R
Performance Cues
Basic Cue S S E R S R R r
Bar after Basic Cue R S E R E R R
Interpretive Cue S R E S R S
Bar before Interpretive R S E R R
Expressive Cue S r
Bar before Expressive r r e r e
Bar after Expressive e e
R2 Starts .23 .22 .28 .16 .08 .15 .12 .11
R2 Stops .16 .29 .17 .05 .07 .06 .14 .07
R2 Repeats .10 .44 .42 .20 .06 .18 .10 .18

stopping, starting, or repeating them. Negative effects with effects of structure throughout. We next describe
occurred mainly for expressive cues and indicate that this pattern of effects in more detail.
changes of expressive intention were practiced by
playing through them without stopping. Formal structure
The practice data summarized in Table 2 showed The pianist started and/or stopped at section bound-
that the pianist was thinking about the formal structure aries (beginnings and ends of sections) more often than
throughout the learning process and attended to each at other locations in every session set except sessions
type of performance cue in turn. The progression was 7-8. The effects show that the pianist was thinking of
similar to the one we have already seen in the com- the structure throughout, suggesting that she used
ments: first basic, then interpretive, then expressive, the hierarchical organization as a retrieval structure

120
http://www.ac-psych.org
Practicing perfection

(Chaffin & Imreh, 2002; Miklaszewski, 1989, Williamon There was a similar progression for starts during
& Valentine, 2002). In sessions 7-8, the pianist needed sessions 1-17 but not thereafter. Practice segments
to practice putting sections together for the first time started at basic performance cues in sessions 1-6 and
after the section-by-section practice of sessions 1-6 and 7-8 as she practiced the piece in sections and put it
avoided starting and stopping at section boundaries in together for the first time. Next, as automaticity de-
order to rehearse the transitions between sections. veloped in sessions 9-16, interpretive cues were add-
Attention to musical structure is also evident in the ed to basic. Then, when she memorized the piece in
practice of switches, which are points of structural session 17, she started at interpretive and expressive
ambiguity where the same passage occurs at different cues, shifting her attention to the main musical turn-
points in the piece. Switches were a focus of attention ing points. The progression did not continue in later
at the beginning and end of the learning process. In sessions. Relearning the piece again in sessions 26-27
sessions 1-6, switches received more starts, stops, and after a 3-month break, starts were once again at in-
repetitions than other bars. Switches were not singled terpretive cues and this continued in the final polishing
out for attention again until sessions 28-36, when play- in sessions 37-50.
ing stopped at switches once again. We have already The progression of effects for repeats was more
quoted the pianist’s comment about the difficulty of complex. A progression from basic to interpretive to
remembering “which way the switch would go”. When expressive was interrupted at two points by a return
retrieval from memory did not keep up with the new, to the practice of basic cues. Initially, the effects were
faster tempo in sessions 28-36, it was at switches that similar to those for starts: passages containing basic
playing was interrupted. The solution was more prac- cues were repeated more than others in sessions 1-6
tice of switches and this took place in sessions 37-50 and 7-8 and then basic and interpretive cues were re-
when switches were once again repeated more than peated more in sessions 9-16. In session 17, however,
other features. effects for starts and repetitions diverged; as the pian-
ist practiced performing from memory for the first time
Performance cues effects of interpretive cues disappeared and basic cues
The progression in the effects of performance cues re-emerged. The progression resumed in sessions
(first basic, then interpretive, then expressive; see Table 20-24, as she prepared for her first public performance:
2) took a different form for each measure, reflecting interpretive cues were repeated more and expressive
differences in the factors responsible for starting, stop- cues less. These effects continued in sessions 26-27,
ping, and repeating. Changes in where playing starts as the piece was relearned after a 3-month break and
reflect changes in the mental landmarks for the piece. then practice of interpretive cues disappeared during
Changes in where playing stops also reflect changes in further polishing of the piece in sessions 28-36, leav-
landmarks as well changes in the problems interfering ing expressive cues as the main focus of attention.
with fluent performance. Changes in which performance Finally, in sessions 37-50 the negative effect of ex-
cues are repeated reflect changing goals for practice. pressive cues was accompanied by a negative effect
The progression was clearest for stops. First, in for basic cues. In the final polishing of the piece, the
sessions 7-8 and 9-16, as the pianist developed the pianist apparently practiced the retrieval cues for both
ability to play through the entire piece, she stopped expression and technique by playing fluently through
at basic cues more often than at other locations. The them without interruption.
memory for what she needed to do was not coming In sum, the pattern of effects indicate that, in gen-
to mind quickly enough and she stopped to make eral, the pianist attended first to basic, then to interpre-
sure that crucial details of technique were executed tive, and finally to expressive cues with starts, stops,
as planned. Then, in sessions 20-24, as she prepared and repetitions showing the progression in different
for the first performance, she stopped at interpretive ways. The progression was simplest for stops (basic in
cues, indicating that it was now interpretive gestures sessions 7-16, interpretive in 20-24, expressive in 28-
that were not coming to mind quickly enough. Finally, 50). For starts the same progression was evident until
in sessions 28-36 and 37-50 during the final polishing, the piece had been memorized in session 17 (basic in
it was expressive cues that affected stops. The effects 1-16, interpretive in 9-17, expressive in session 17).
were negative, indicating that the pianist did not stop For repeats, the pattern was interrupted by the reap-
at expressive cues, reflecting the fact that to practice pearance of effects of basic cues in sessions 17 and
expression it is necessary to play through emotional 37-50 (basic in 1-17, interpretive in 9-16 and again in
turning points without interruption. 20-27, expressive in 20-50). Despite the differences,

121
http://www.ac-psych.org
 Roger Chaffin and Topher Logan

each measure showed basic performance cues being Chaffin & Imreh, 2002; Chaffin et al., 2002, p. 214).
practiced earlier and expressive cues later. The serial position effects tell us that the pianist’s
The practice data provide behavioral confirmation memory was organized primarily in terms of formal
of the pianist’s report that she practiced performance structure and expressive cues, even after a more than
cues and that she focused on expressive cues in her two-year absence from performing the piece. Recall
final preparation for performance. They do not, how- was best for bars marked by structural boundaries and
ever, show that the pianist continued to think about tapered off with each successive bar. Expressive and
performance cues during performance. For evidence interpretive cues produced similar effects, although
of this point we turn to her recall of the score after the the effect for interpretive cues was not significant.
final performance. Basic cues, on the other hand, produced the opposite
effect. Recall was worse for the bar containing the cue
and improved over the next three bars. The effect sug-
Effects of performance cues on
gests that the role of basic cues was different from
recall of the score
that for expressive cues. Basic cues represent details
Free recall reflects the organization of information in about planned movements that were not called for in
memory. Recall of an ordered series is generally better the written recall task and that apparently interfered
for the first item in the series and declines with each with it. Structural and expressive cues, in contrast,
succeeding item (Broadbent, Cooper, & Broadbent, were linked to the conceptual representation of the
1978; Raaijmakers & Shiffrin, 1981; Roediger & piece that was required for writing out the score.
Crowder, 1976). Serial position effects are thus a re- The main landmarks in her memory for the piece more
flection of how information is organized in memory. than two years later were the structural and expressive
Performers generally report that during the final prep- cues. The serial position effects for these cues support
aration for performance, they focus more on musical the pianist’s report that during the final performance
expression and less on technique. If these reports she was thinking mainly about musical expression with
are accurate, we would expect to find serial position basic and interpretive performance cues on the fringes
effects for structural boundaries and expressive and of her awareness, available in case things went wrong.
interpretive cues. Better recall of the score at points There is, however, another possible explanation. The
where these cues are reported would indicate that they finished performance could have been entirely auto-
functioned as retrieval cues. Declining recall in follow- matic. On this view, performance cues functioned as
ing bars would suggest that memory was organized in retrieval cues as a result of extended practice but did
associative chains starting at each retrieval cue. not otherwise control the performance. To evaluate the
Twenty-seven months after the recording session the pianist’s claim that she used structural and expressive
pianist in the Presto study was unexpectedly asked to write cues to monitor and direct her performance, we turn
out the score from memory. Her memory was remarkably next to the polished performances at the end of the
good, 65% accurate (Chaffin & Imreh, 1997; Chaffin et learning process and in the recording studio.
al., 2002, p. 212). More interesting, though, were the ef-
fects of the different kinds of performance cues.
Tempo variation during
The proportion of notes correctly recalled in each
performance
bar provided a measure of the probability of recall.
(Table 3 shows the effect of serial position with respect For evidence that performance cues guide and control
to each type of performance cue on recall of the score. polished performance, we need to compare their ef-

Table 3.
Probability of correctly recalling the score decreased with distance from section boundaries and expressive cues and increased
with distance from basic cues

Serial position: Distance from cue in number of bars

Type of Performance Cue 1 2 3 4 5-8
Structural Boundary .97 .90 .87 .69 .28
Expressive .85 .85 .74 .43 .00
Basic .68 .77 .78 .77 .46
Interpretive .75 .78 .61 .00

122
http://www.ac-psych.org
Practicing perfection

Figure 4.
Profiles of tempo in beats/min, plotted across bars for mean practice and the CD performance (last three bars omitted for
tempo and first and last bars omitted), showing the locations of two of the downward octave jumps and ascending 8th note
scales (indicated by red arrows) and the C and D themes (indicated by highlighting and brackets).

fects on different performances. Systematic differences pianist’s intended performance in the practice sessions
between performances attributable to performance that preceded it. The absence of live audiences for all of
cues would support the proposals that these cues are the performances lessens the chance of finding differ-
responsible for the differences between performances ences as does averaging across practice performances.
that musicians commonly report, and that skilled per- Finding differences between these performances would
formance involves the strategic management of the suggest, therefore, that differences are likely any time
variability inevitable in any activity (Latash, Scholz & the same piece is performed repeatedly. (Differences
Schoner, 2002; Todorov & Jordan, 2002; Van Orden between the seven practice performances are reported
et al., 2003). On the other hand, the absence of sta- by Chaffin et al., 2006. The present comparison has
tistically significant differences between performances not been reported elsewhere and provides a more con-
would suggest that any differences that do occur are servative test of the prediction of differences between
not a product of systematic control process but simply repeated performances.)
reflect the random variability that pervades any activ- We might expect attention to basic performance
ity. On this account, the impression of many perform- cues to be reflected in slower tempi, to allow time to
ers that each performance is unique is an illusion; the monitor the details of execution represented by these
differences are not a product of artistic inspiration or cues. Predictions for the effects of expressive and
the skilled management of variability but of random interpretive performance cues are more complex be-
noise. On this view, performers remember the random cause their effects depend on the nature of the musi-
variations that happen to be musically significant and cal gesture intended. The predominant feeling evoked
forget those that are not (Chaffin et al., 2006). by the Presto is of headlong, rushing forward. For a
We will illustrate the type of analysis needed to piece of this sort, we expected that expressive cues,
evaluate these claims by comparing the pianist’s per- representing higher levels of emotional intensity (as
formance of the Presto in the recording studio, released reported by the pianist) would be associated with fast-
on CD, with the mean of seven polished performances er tempi. For interpretive cues, we empirically iden-
in the practice studio in the two weeks leading up to tified four additional musical gestures by identifying
the final performance. The comparison approximates fluctuations in tempo that were consistently related to
the common situation of a pianist giving repeat per- features of the music. We then looked for differences
formances in different venues. Differences would show between performances in the effects of the gestures.
that the professionally important CD performance dif- All of the musical gestures identified involved features
fered from the most reliable measure available of the that occurred at multiple points in the piece. Any ef-

123
http://www.ac-psych.org
 Roger Chaffin and Topher Logan

fects cannot, therefore, be readily attributed to the

idiosyncratic playing of a single bar. Rather, significant
effects would indicate that a musical gesture involved
consistent changes at multiple points in the piece.
Likewise, differences between performances in the ef-
fect of a gesture would reflect consistent differences
in the same direction at multiple points.
Figure 5.
Score of bars 1-3 showing musical features that elicited two
Description of the performances of the pianist’s interpretive gestures: the octave jump down
in bar 1 and the ascending 8th note scale in bar 2.
As required by the score and by Baroque performance
practice, the tempo was fast and variation minimal wave structure evident in both performances in Figure
(Mean = 141.5 and 139.0 beats/min for the practice 4. Two sections containing the light, dancing C theme
and CD performances respectively; SD = 4.4 and 4.6 (bars 77-92 and 155-166; highlighted and indicated
beats/min respectively). The pianist played a mean by brackets in Figure 4) were faster while the complex
of approximately 14 notes per second throughout the fugue in the D theme that marks the center of the piece
entire piece, while the standard deviations indicated (bars 104-112) was slower.
that the average variation in tempo was well below An additional gesture was identified on the basis of
the 20 ms threshold required for detection (Clarke, the pianist’s reports of her decisions about phrasing.
1989). Although values for the two performances were When these were compared with the fluctuations in
very similar, tempo was reliably faster in the practice tempo, it was evident that tempi were slower in bars
performances, t(207) = 12.80, p < .001, something where the pianist reported beginnings of phrases. These
that the pianist had anticipated in session 42 when she musical gestures are listed in Table 4 along with the
noted that she might slow down for the CD perform- other predictors used in the analysis of tempo (basic
ance to “make it sound easy” (Chaffin et al., 2002, and expressive performance cues, expressive intensity,
p. 154). section bounds, switches and technical difficulties).
Table 4 also shows the source of each predictor (pian-
ist’s report or our visual inspection of the correspond-
Identification of musical gestures
ences between the profiles and musical score just de-
The tempo profiles for the CD and mean practice per- scribed), the number of bars in which each gesture or
formances, shown in Figure 4, contain marked similari- musical feature occurred, and the range of values used
ties (r = .79). For example, there were sharp down and to code it. All gestures occurred in multiple locations in
up spikes (indicated by red arrows in Figure 4) at the the piece so their effects were unlikely to be due to the
beginning (bar 13-14) and end (bar 199-200). These idiosyncratic playing of a single bar.
spikes coincide with the beginning of the A theme which
is introduced by a downward octave jump followed by
Factors affecting performance
an ascending scale of eighth notes (see Figure 5). These
two musical features contribute to the predominant feel- Effects of the predictors on the two performances are
ing that the Presto seems to evoke of headlong, forward summarized in Table 5 (left and middle panels). Positive
motion. The downward jump seems to evoke the feeling effects indicate increases in tempo; negative effects
of leaping downward, while the upward scale gives an indicate decreases. The table also summarizes differ-
impression of rushing back up again. These musical im- ences between the mean practice and CD performances
ages were accentuated by the pianist. By prolonging the (right hand panel). The regression coefficients in the
first note of the jump (downward spikes), she seemed right hand panel of Table 5 represent effects in the CD
to reinforce the sensation of leaping down and the im- performance that were not present in mean practice
pact of landing. By compressing the time taken by for performance.
the upward scale (upward spikes), she emphasized the The most important result for our purposes was that
feeling of rushing upward. The A theme returns repeat- there were differences between the CD and practice
edly throughout the piece, and the repetition of these performances attributable to expressive and to basic
two gestures is responsible for many of the upward and performance cues. The differences support the pro-
downward spikes in both performances. A third musi- posal that performance cues were used to control the
cal gesture was responsible for some of the large-scale performances. Passages where the pianist reported

124
http://www.ac-psych.org
Practicing perfection

Table 4.
Description of predictors used in analysis of tempo indicating how each was identified (source), the number of bars involved
(max 210) and the range of values used to code presence/absence in a bar.

Number of bars
Description Source Values coded
involved
Performance cues
Expressive performance Change in expressive
Pianist 171 0-2
cues intention
Level of expressive
Expressive intensity Pianist 210 1-5
intensity
Attention to technique
Basic performance cues Pianist 125 0-3
during performance
Musical structure
First and last bars of
Section begins or ends Pianist 137 0,1
subsections
Similar musical material
Switches Pianist 142 0-3
repeats
Technique required
Technical difficulties Pianist 126 0,1
repetition during practice
Interpretive Gestures
Octave jump before or
Octave jump down Investigator 111 0,1
at start of A theme
Octave scake of 8th notes
8th note upward scale Investigator 115 0,1
before or after Jump:A
Ranking of tempo of
Theme Pianist/Investigator 210 1-4
A,B,C & D themes
Phrasing Starts of phrases Pianist 188 0-7

that she intended to play with more expressive in- details were executed correctly. In summary, the dif-
tensity were marked by faster tempi. The effect was ferences between the two performances suggest that
present in both practice and CD performances but the pianist altered her performance in the recording
was larger in the CD performance, F(1,192) = 26.14, studio by playing more expressively and less cautiously
p < .01. The pianist apparently used faster tempi to and that these adjustments were achieved through the
convey more intense musical feelings (cf. Juslin, 2000, use of performance cues.
2003) and exaggerated this gesture more in the CD The differences between the CD and practice per-
than in the practice performance. The slightly slower formances are difficult to detect but may be discern-
tempo used in the CD performance may have allowed able in the overall “feel” of the performance. Readers
more scope for expressive variation (Repp, 1995). The can judge for themselves for measures 184-195 where
second difference was in the slowing in bars before the CD performance seems to us slightly more expres-
basic performance cues which was more pronounced sive but not noticeably faster than a representative
in the mean practice than in the CD performance, practice performance of the same passage. Sound files
F(1,192) = 28.49, p < .01. In six of the seven practice for these selections can be found at http://csa.uconn.
performances the pianist was trying to produce a video- edu/research/public/mm184CD.mp3 and http://csa.
recording to be played to professional audiences during uconn.edu/research/public/mm184practice.mp3.
talks about the research. She was concerned with giv- There were no differences between the two perform-
ing a note-perfect performance and commented that ances in the interpretive gestures. Effects of all four
all of the performances sounded “cautious” as a result. gestures were significant for both the CD and practice
In the recording studio, in contrast, she could afford to performances, confirming the description of these ges-
take more risks because mistakes could be removed by tures given above. First, the pianist’s accentuation of
editing. Slowing in bars before basic performance cues the downward octave jump at the start of the A theme
in the practice performances was, therefore, probably was reflected in a negative (slowing) effect. Second,
due to taking a little more time to ensure that technical the accentuation of the upward rush of the ascend-

125
http://www.ac-psych.org
 Roger Chaffin and Topher Logan

Table 5.
Regression coefficients and R2 for the effects of musical gestures, performance cues, and basic features on the mean of prac-
tice performances in sessions 42-50 and the CD performance and differences between the two.

Mean practice Difference of CD from

CD performance
performance mean practice performance
Predictor Variable Tempo Tempo Tempo
Mean practice performance -1.058***
Expressive performance cues
Expressive cue -0.411 -0.783 -0.151
Bar before expressive cue -0.900 -0.741 -0.127
Expressive intensity -0.148 -0.527** -0.170**
Basic performance cues
Basic cue -0.764~ -0.363 -0.233
Bar before basic cue -1.502*** -0.546 -0.538*
Musical Structure
Begin section -0.097 -0.616 -0.341
End section -1.767* -1.459~ -0.248
Switch -0.052 -0.477 -0.197
Bar before switch -1.181** -1.238* -0.042
Technical difficulties
Technical difficulties -0.422 -0.135 -0.287
Interpretive gestures
Octave jump down -5.812*** -6.255*** -0.131
8th-note upward scale -5.019*** -5.576*** -0.031
Theme -1.633*** -1.457*** -0.178
Phrasing -1.759** -1.956** -0.010
R 2
-0.50*** -0.45*** -0.92***

* p < .05, ** p < .01, *** p < .001

+ Significance levels for predictors based on interaction of predictor with performance in a 2x14 mixed ANOVA.
See text for explanation.

ing scale resulted in a positive (accelerating) effect switches was what made the Presto hard to play from
on tempo. Third, the differences between the faster memory. The tendency to overcome difficulties by play-
C theme and the slower D theme were significant in ing faster (and louder) is very familiar to musicians,
both performances. Fourth, beginnings of phrases were although it has not, to our knowledge, been explored
marked by slower tempi in both performances. The use by psychologists. It is possible that increasing tempo,
of a short pause or “breath” to mark the beginning of amplitude, or force helps overcome inertia in the motor
a new phrase is a common interpretive device (De Poli, system (Stins & Michaels, 1999). It is interesting to
Roda & Vidolin, 1998; Repp, 1998). compare the effects of switches and basic performance
There were also no differences between the two per- cues. Basic cues included a subset of switches – those
formances in the effects of musical structure. The slowing most likely to give problems. In spite of the overlap,
at the ends of sections that occurred in both perform- the effects on tempo were different. Tempo increased
ances is a common expressive device (Clarke, 1999, before switches and decreased before basic cues. At
Gabrielsson & Lindström, 2001; Palmer, 1989, 1997). basic cues, the pianist was thinking about the upcom-
The effect was statistically reliable only in the mean prac- ing difficulty and slowed down accordingly. Switches
tice performance, but did not differ significantly from the that were not marked by basic cues, in contrast, were
trend in the same direction in the CD performance. places where the pianist did not anticipate problems.
The faster tempo in bars before switches can be at- In these cases, she apparently relied instead on the
tributed to the tendency to play faster in difficult pas- automaticity of motor memory, increasing the tempo to
sages. The pianist reported that keeping track of the help push through any residual difficulty.

126
http://www.ac-psych.org
Practicing perfection

CONCLUSION that by session 24, when she was ready for her first
public performance, she was attending mostly to in-
Solo recitals in the Western classical music tradi- terpretive cues. Only when preparation was almost
tion place extraordinary demands on the performer. complete, between sessions 31 and 32, were ex-
Performances must be practiced to the point that they pressive cues mentioned explicitly for the first time.
can be delivered automatically in order to ensure reli- The practice data showed a similar progression with
ability under the pressures of the concert stage. At the practice focusing first on basic, then on interpretive,
same time, the performance must remain fresh and and finally on expressive cues, with musical struc-
flexible enough to communicate emotionally with the ture influencing practice throughout.
audience and permit recovery from inevitable mis- The fact that the pianist paid more attention to ex-
takes. We suggest that the integration of automatic pressive cues in later practice sessions is suggestive,
motor performance and cognitive control needed to but does not necessarily prove that expression was the
provide this flexibility is achieved through the practice main focus of attention during performances. Practice
of performance cues. Use of performance cues is an may have been designed to ensure that actions elicited
attentional strategy that maintains conscious control by expressive cues occurred automatically during per-
of a highly automated performance and provides the formance. The conclusion that the pianist was attend-
musical spontaneity that is valued by both performers ing to expressive cues during performance is, however,
and audiences. Spontaneity is a byproduct of musi- supported by two additional types of evidence: effects
cians’ ability to use performance cues to adapt a highly of expressive cues on recall of the score more than two
prepared performance to the requirements and oppor- years later and differences between performances in
tunities of each performance. the effects of expressive cues.
When a performer has to think mostly of basic cues First, expressive and structural cues provided the
dealing with matters of technique, the possibilities for main landmarks of the pianist’s memory for the piece
musically creative variation are limited. When a per- two years later, suggesting that these cues had also
former is focused on interpretive cues and is thinking served as landmarks when she last played the Presto.
about what the music sounds like, the opportunities Second, there was a small but statistically reliable dif-
for creativity are greater but still limited. The goal of ference in the effects of expressive and basic cues in
performance is to evoke musical feelings and this is the CD performance compared to practice perform-
best achieved when the performer focuses on expres- ances in the days leading up to the recording session.
sion. An expressively spontaneous performance is, The pianist increased the tempo in passages where
therefore, most likely when the performer is focusing she wanted to increase expressive intensity, and did
on expressive cues and the musical structure that sup- so more in the CD than in the practice performances.
ports them. This allows the artist to adjust the per- The pianist also slowed the tempo slightly approaching
formance to the unique opportunities and demands of basic performance cues where some point of technique
the occasion to achieve the maximum possible impact required attention during performance. This effect
on the audience. was larger in the practice than in the CD performance
We base this account of how a performance is and probably reflects the pianist’s concern with pro-
prepared on longitudinal case studies of experi- ducing a note-perfect performance in these sessions.
enced soloists preparing for performance (Ginsborg The effects of expressive and basic performance cues
et al., 2006; Lisboa et al., 2004; Noice et al., in on polished performances, thus, support the pianist’s
press) which we have illustrated here with a pianist report that she attended to expressive cues during
learning the Presto (Chaffin & Imreh, 2001, 2002, performance. The differences between performances
Chaffin et al., 2002; Chaffin et al., 2003, 2006). support our claim that experienced performers use
The analysis was based initially on the performer’s performance cues to monitor and control the highly
self-reports after practice performances. We then practiced, automatic motor sequences developed dur-
looked at behavioral evidence to test her account. ing the long months of practice. Additional research
The pianist’s spontaneous descriptions of what she is needed to understand how planned use of perform-
was attending to during practice indicate that dur- ance cues described here relates to the incremental
ing her first practice performance without the score planning during performance described by Palmer
in session 12 she was focused on structure, that and Pfordresher (2003). The idea that event retrieval
the next time she played from memory, in session relies on metrical similarity and serial proximity may
17, she was thinking mostly about basic cues, and help in understanding why some musical decisions

127
http://www.ac-psych.org
 Roger Chaffin and Topher Logan

made during practice become performance cues while previous reports in including only musical structure and
others do not. performance cues as predictors.
Concert soloists provide an interesting test of the
principles of expert memory (Ericsson & Kintsch, Author Notes
1995) because they make their living performing from We would like thank Gabriela Imreh for contribut-
memory. Like expert memorists in other fields, the ing both the ideas on which the research was based
concert pianist in the study we have described engaged and the data to test them, Mary Crawford for helpful
in extended practice of a retrieval scheme to ensure discussions, Colleen Chen and Anthony Lemieux for
that recall occurred rapidly and automatically. This re- work on the tempo measurements, Ben Chaffin for
liance on an explicit, conceptual memory is somewhat programming help, and for work on the practice data:
surprising, given the importance of implicit, motor and Ellie Corbett, Jennifer Culler, Elizabeth Dohm, Helene
auditory memory in piano performance. However, the Govin, Julie Konik, Amelia McCloskey, Sandra Paez,
pianist went to great lengths to ensure that she could Alethea Pap, and Aaron Williamon.
rely on an explicit, conceptual memory. The formal
structure of the music provided a hierarchical retrieval References
structure, organized into sections and sub-sections, Broadbent, D.E., Cooper, P.J., and Broadbent, M.H.
with expressive phrases containing basic and interpre- (1978). A comparison of hierarchical matrix retrieval
tive performance cues making up the bottom levels schemes in recall. Journal of Experimental Psychology:
(Figure 1). The pianist engaged in prolonged practice Human Learning and Memory, 4, 486-497.
to bring the operation of this retrieval scheme up Chaffin, R., and Imreh, G. (1997). “Pulling teeth and
to the pace of the motor performance. This explicit, torture”: Musical memory and problem solving.
conceptual representation allowed her to focus on ex- Thinking and Reasoning: Special Issue on Expert
pressive goals during performance, while also keeping Thinking, 3, 315-336.
track of where she was so that the performance would Chaffin, R., Imreh, G. and Crawford, M. (2002).
unfold as planned. If things did begin to go wrong, Practicing perfection: Memory and piano perform-
the conceptual representation provided a means to ance. Mahwah, NJ: Lawrence Erlbaum Associates.
recover. Chaffin, R. and Imreh, G. (2001). A comparison of
This study, and other case studies like it, extends practice and self-report as sources of information
the principles of expert memory from domains that about the goals of expert practice. Psychology of
rely almost entirely on explicit memory (e.g., for ma- Music 29, 39-69.
terial such as digit strings, to a performing art where Chaffin, R. and Imreh, G. (2002). Practicing perfection:
it might be possible to rely on automatic, implicit Piano performance as expert memory. Psychological
motor skills in place of explicit memorization. The Science, 13, 342-349.
experienced musicians we have studied supplement Chaffin, R., Imreh, G., Lemieux, A., and Chen, C.
their implicit memories of motor sequences with ex- (2003). “Seeing the big picture”: Piano practice as
plicit conceptual memories that provide them with a expert problem solving. Music Perception, 20, 461-
mental map that is used to monitor and modify the 485.
highly practiced motor sequences. This combination Chaffin, R., Lemieux, A., and Chen, C. (2004) “It’s dif-
of explicit and implicit memory may be a hallmark of ferent each time I play”: Why highly polished per-
expertise in domains such as musical performance formances vary. In S. D. Lipscomb, R. Ashley, R. O.
that require high levels of both motor dexterity and Gjerdingen and P. Webster (Eds.) Proceedings of the
aesthetic sensibility. 8th International Conference on Music Perception and
Cognition. Adelaide, Australia: Causal Productions.
Notes Chaffin, R., Lemieux, A., & Chen, C. (2006). Spontaneity
1
It is this smooth integration of retrieval that Ericsson and creativity in highly practiced performance.
and Kintsch (1995) refer to as “long-term working In I. Deliège & G.A. Wiggins. Musical creativity:
memory. Multidisciplinary research in theory and practice (pp.
2
The analyses summarized here are similar to those 200-218). London: Psychology Press.
reported by Chaffin & Imreh (2002) in not separating Chase, W.G., and Ericsson, K.A. (1982). Skilled and
practice into two types (work vs runs) as in Chaffin working memory. In G.H. Bower (Ed.), The psychol-
et al., (2002), but are similar to the latter in dividing ogy of learning and motivation 16: 1-58. New York:
sessions into 8 groups. The analyses are different from Academic Press.

128
http://www.ac-psych.org
Practicing perfection

Chase, W. G., & Simon, H. A. (1973). The mind’s eye in Hallam, S. (1995b). Professional musicians’ orientation
chess. In W. G. Chase (Ed.), Visual information process- to practice: Implications for teaching. British Journal
ing (pp. 215-281). New York: Academic Press. of Music Education, 12, 3-19.
Chi, M. T. H., Feltovich, P. J., & Glaser, R. (1981). Halpern, A. R., and Bower, G., H. (1982). Musical ex-
Categorization and representation of physics prob- pertise and melodic structure in memory for musical
lems by experts and novices. Cognitive Science, 5, notation. American Journal of Psychology, 95, 31-
121-125. 50.
Clarke, E. (1999). Rhythm and timing in music. In: D. Hinson, M. (1987). Guide to the Pianist’s Repertoire.
Deutsch (Ed.), The Psychology of Music (pp. 473– Bloomington: Indiana University Press
500). San Diego, CA: Academic Press. Imreh, G. pianist. (1996). J.S. Bach [CD]. New York:
Clarke, E. (1989). The perception of expressive timing Connoisseur Society.
in music. Psychological Research, 51, 2–9 Imreh, G. and Chaffin, R. (1996/97) Understanding
Cooke, J. F. (1999). Great pianists on piano playing: and developing musical memory: The views of a con-
Godowsky, Hofmann, Lhévinne, Paderewski and 24 cert pianist and a cognitive psychologist, American
other legendary performers. Toronto: Dover. (origi- Music Teacher, 46, (3): 20-24, 67.
nally published 1913, expanded edition published Juslin, P.N. (2000). Cue utilization in communication of
1917) emotion in music performance: Relating performance
DePoli, G., Roda, A. and Vidolin, A. (1998). Note-by- to perception. Journal of Experimental Psychology:
note analysis of the influence of expressive intentions Human Perception and Performance, 26, 1797-1813.
and musical structure in violin performance. Journal
of New Music Research, 27, 293-321. Juslin, P.N. (2003). Communication of emotions in
Ericsson, K. A. and Kintsch, W. (1995). Long-term work- vocal expression and music performance: Different
ing memory. Psychological Review, 10, 211-245. channels, same code? Psychological Bulletin, 129,
Ericsson, K. A., and Oliver, W. L. (1989). A methodology 770-814
for assessing the detailed structure of memory skills. Krampe, R.T. & Ericsson, K.A. (1996). Maintaining ex-
In A. M. Colley, and J. R. Beech (Eds.), Acquisition cellence: Deliberate practice and elite performance
and performance of cognitive skills (pp. 193-215). in young and older pianists. Journal of Experimental
Chichester: Wiley. Psychology: General, 25, 331-359.
Gabrielsson, A. (1999). The performance of music. In Latash, M. L., Scholz, J. P., & Schoner, G. (2002). Motor
Deutsch D. (Ed.), The psychology of music (2nd Ed.) control strategies revealed in the structure of motor
(pp. 579-602). San Diego: Academic Press. variability. Exercise and Sports Science Reviews, 30,
Gabrielsson, A. and Lindström, E. (2001). The influ- 26-31.
ence of musical structure on emotion. In P. N. Juslin Lisboa, T., Chaffin, R., Schiaroli, A.G., Barrera, A.
and J.A. Sloboda (Eds.), Music and emotion: Theory (2004). Investigating practice and performance
and research. New York: Oxford University Press. on the cello. In S. D. Lipscomb, R. Ashley, R. O.
Ginsborg, J., Chaffin, R. & Nicholson, G. (2006). Gjerdingen and P. Webster (Eds.), Proceedings of the
Shared performance cues in singing and conducting: 8th International Conference on Music Perception &
A content analysis of talk during practice. Psychology Cognition, Evanston, IL, 2004. Adelaide, Australia:
of Music, 34, 167-194. Causal Productions.
Gilden, D.L. (2001). Cognitive emissions of 1/f noise. Mach, E. (1991). Great contemporary pianists speak
Psychological Review, 108, 33-56. for themselves. New York: Dover.
Glaser, R., & Chi, M. (1988). Overview. In M. Chi, R. Mandler, G., and Pearlstone, Z. (1966). Free and con-
Glaser, & M. Farr (Eds.), The nature of expertise (pp. strained concept learning and subsequent recall. Journal
xv-xxviii). Hillsdale, NJ: Erlbaum. of Verbal Learning and Verbal Behavior, 5, 126-131.
Gobet, F., & Simon, H. A. (1996). The roles of rec- Miklaszewski, K. (1989). A case study of a pianist pre-
ognition processes and look-ahead search in time- paring a musical performance. Psychology of Music,
constrained expert problem solving: Evidence from 17, 95-109.
grand-master-level chess. Psychological Science, 7, Miller, G. A. (1956). The magical number seven, plus or
52-55. minus two: Some limits on our capacity for process-
Hallam, S. (1995a). Professional musicians’ ap- ing information. Psychological Review, 63, 81-97.
proaches to the learning and interpretation of music. Neuhaus, H. (1973). The art of piano playing. New
Psychology of Music, 23, 111-128. York: Praeger Publishers Inc.

129
http://www.ac-psych.org
 Roger Chaffin and Topher Logan

Noice, H., Chaffin, R., Jeffrey, J. & Noice, T. (in press). Shaffer, L.H. (1984). Timing in solo and duet piano
Memorization by a jazz pianist: A case study. Music performances. Quarterly Journal of Experimental
Psychology. Psychology, 36A, 577-595.
Noyle, L. J. (Ed.). (1987). Pianists on piano playing: Shaffer, L.H., Clarke, E.F., Todd, N.P. (1985). Meter and
Interviews with twelve concert pianists. Metuchen, rhythm in piano playing. Cognition, 20, 61-77.
NJ: Scarecrow. Stins, F. J., & Michaels, C. F. (1999). Strategy differences
Palmer, C., and Pfordresher, P. (2003). Incremental in oscillatory tracking: Stimulus–hand versus stimu-
planning in sequence production. Psychological lus manipulandum coupling. Journal of Experimental
Review, 110, 683-712. Psychology: Human Perception and Performance. 25,
Palmer, C. (1997). Music performance. Annual Review 1793–1812.
of Psychology, 48, 115-138. Sundberg, J., Friberg, A. & Frydén, L. (1991). Threshold
Palmer, C. (1989). Mapping musical thought to and preference quantities of rules for music perform-
musical performance. Journal of Experimental ance. Music Perception, 9, 71-92.
Psychology: Human Perception and Performance, Thompson, C. P.,. Cowan, T. M & Frieman, J. (1993).
15, 331-346. Memory Search by a Memorist. Hillsdale, NJ:
Raaijmakers, J. G., & Shiffrin, R. M. (1981). Search of Lawrence Erlbaum Associates.
associative memory. Psychological Review, 88, 93- Todorov, E. & Jordan, M.I. (2002). Optimal feedback
134. control as a theory of motor coordination. Nature
Repp, B. H. (1992). Probing the cognitive representation of Neuroscience, 5, 1226-1235.
musical time: Structural constraints on the perception of Tulving, E. (1962). Subjective organization in free re-
timing perturbations. Cognition, 44, 241–281. call of “unrelated” words. Psychological Review, 69,
Repp, B. H. (1995). Quantitative effects of global 344-354.
tempo on expressive timing in music performance: Van Orden, G.C., Holden, J.G. Turvey, M.T. (2003).
Some perceptual evidence. Music Perception, 13, Self-organization of cognitive performance. Journal of
39-57. Experimental Psychology: General, 132, 331-350.
Repp, B. H. (1998). Obligatory “expectations” of ex- Williamon, A., and Valentine, E. (2002). The role of
pressive timing induced by perception of musical retrieval structures in memorizing music. Cognitive
structure. Psychological Research, 61, 33 - 43 Psychology, 44, 1-32.
Roediger, H. L., III, and Crowder, R. C. (1976). A Williamon, A., Valentine, E. & Valentine, J. (2002).
serial position effect in recall of United States Shifting the focus of attention between levels of
presidents. Bulletin of the Psychonomic Society, 8, musical structure. European Journal of Cognitive
275-278. Psychology, 14, 493-520.

130
http://www.ac-psych.org