Content uploaded by Rhonda Boyle
Author content
All content in this area was uploaded by Rhonda Boyle on Dec 23, 2015
Content may be subject to copyright.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 1 of 79
1
PIANIST HAND SPANS: GENDER AND ETHNIC DIFFERENCES AND
IMPLICATIONS FOR PIANO PLAYING
Rhonda Boyle, Robin Boyle and Erica Booker
Hand span data was collected from 473 adult pianists and analysed using descriptive and
inferential statistical methods, focusing on differences according to gender, ethnicity and
level of acclaim. For comparative purposes, similar data was collected from 216 non-
pianist university students and 49 children and teenagers. Gender differences are
consistent with those found in earlier studies and ethnic differences are also significant
but smaller in magnitude. Highly acclaimed solo performers tend to have bigger hand
spans than others. ‘Small hands’ are defined in terms of hand span metrics, allowing
estimates of the proportions of pianists with ‘small hands’. The conclusion is that the
current ‘standard’ piano keyboard is too large ergonomically for a majority of pianists.
______________________________________________________________________
1.0 INTRODUCTION
Pianists are often heard to comment on the size of their hands or those of other pianists.
They may say that they have ‘small hands’ or ‘large hands’. ‘Small hands’ are generally
seen to be a disadvantage when playing the piano while ‘large hands’ are seen to be an
advantage. But what do we actually mean by ‘small hands’ or ‘large hands’? What hand
size dimensions are relevant to playing the piano comfortably and how can these can be
linked to the pain and injury that affects a significant proportion of pianists? What
proportion of pianists, and indeed members of the population at large, have ‘small
hands’?
Unlike many other musical instruments, most notably stringed instruments, pianos
effectively come with the same sized keyboard. The key question this paper addresses is
whether the current one-sized ‘standard’ or ‘conventional’ keyboard is ergonomically
suited to the varying hand sizes of the general piano playing population.
Until the late nineteenth century, piano keyboards were available in different sizes and
generally had smaller key widths than today. The current piano keyboard was
standardised in the 1880s, based on what suited male virtuosos of the time, such as
Liszt (Parakilas et.al., 1999; summarised by Booker & Boyle, 2011). The belief that the
1880s ‘standard’ keyboard size is somehow ‘sacred’, that it suits all pianists and that it
cannot be varied for either musical or technological reasons, has rarely been questioned
since that time.
However, today, the ‘one size fits all’ approach is increasingly being challenged, mainly
due to growing evidence from performing arts health research coupled with the recent
availability of ESPKs (ergonomically-scaled piano keyboards) with narrower keys. Such
keyboards effectively give ‘larger’ hands to pianists who otherwise experience enormous
frustration and risk pain and injury from attempting advanced repertoire with ‘small
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 2 of 79
2
hands’ on the standard keyboard. Pianists who experience such keyboards finally
understand the barriers caused by having hands too small for the conventional piano
keyboard. The need for ESPKs to become widely available is supported by evidence from
health professionals who are concerned about the high level of pain and injury among
pianists and from the direct experience of teachers and their students who have noted a
wide range of benefits when using ESPKs.
Within the general population it is clear that there is significant variation in hand size. In
this paper we provide statistical data to quantify this variation among pianists and non-
pianists. A definition of ‘small hands’ is proposed, drawing on research from various
fields and other evidence. We conclude that only a minority of adult pianists have ‘large
hands’, providing them with a significant advantage on the current keyboard in terms of
ease of playing, speed of learning, pianistic accomplishment, range of accessible
repertoire and reduced risk of pain and injury. Further, we conclude that the large
majority of people (considering children as well as adults) who have ‘small hands’ are
disadvantaged on the current keyboard and that they would benefit significantly from
having access to pianos with narrower keys (ESPKs).
2.0 STUDIES OF HAND SIZE
2.1 Previous studies of hand size
Hand size, as with other body characteristics such as weight and height, varies greatly
among the human population. The science of anthropometry, which deals with the
measurement of the human individual, has traditionally been used in anthropology to
correlate physical features with racial and psychological traits. Today, anthropometry
plays an important role in industrial and clothing design, ergonomics and architecture.
Much of the published anthropometrical data has been collected from specific
populations, such as defence force personnel or industrial workers, and used to analyse
physical requirements for specific tasks undertaken by those groups.
Measurements of the human body are normally categorised on the basis of gender and
ethnicity. Published data (for example, Pheasant 1988; Garrett, 1971) confirm what is
obvious for most physical characteristics – that there are variations between ethnic
groups and that males are generally bigger than females. General observation is
sufficient for one to realise that males have larger hands than females. From
measurements of many different features of the human hand in the US Army (Garrett,
1971), depending on the characteristic measured, differences between males and
females generally range between 10% and 20%.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 3 of 79
3
The authors have only found two previous statistical studies that relate specifically to
pianists’ hand spans and which looked at gender differences.
The most comprehensive set of data on hand size and other characteristics of pianists
dates back to the 1980s (Wagner, 1988). He measured 20 characteristics in hand size
and joint mobility for professional pianists, teachers and tertiary level students, and
found significantly higher mean values in males than females for virtually all hand size
variables. His measurements included active finger spans (i.e. maximum stretch when
the hand is spread out unaided on a flat surface) between the fingers, including the
thumb. Active finger spans can be directly related to a pianist’s capacity to play various
intervals and chords.
Steinbuhler (2004) measured pianists’ active thumb to fifth finger span at a music
teachers’ convention in the US. The data have been displayed visually since that time
1
but were not analysed quantitatively until research by two of the authors in 2009 (Boyle
& Boyle).
A third data set, held by the University of North Texas (UNT), has been discussed by
Yoshimura and Chesky (2009) but gender differences were not reported. They
commented on the extensive range of hand spans among nearly 400 UNT music
students, with the difference between the smallest and the largest being 4 inches (10
cm) – more than the width of four piano keys!
Several writers have noted that pianists of Asian ethnicity have smaller hands than those
of Caucasian origin (e.g. Sakai, 1992, 2008) but no data relating to ethnic differences
among pianists were found by the authors.
2.2 Previous research by the authors
The authors have previously undertaken research and reviewed published data and
literature relating to hand spans and piano playing, including:-
• Analysis of Steinbuhler’s (2004) data (Boyle & Boyle, 2009).
• Investigation of the relationship between active hand spans (based on
Steinbuhler and Wagner) and the capacity to reach certain intervals on the
conventional piano keyboard (Boyle & Boyle, 2009).
• Review of literature on the relationship between pain and injury and hand size,
and perceptions of pianists using ESPKs (Boyle & Boyle, 2009; Booker & Boyle,
2011).
• Survey of the experiences of adult pianists who play ESPKs (Boyle, 2012).
1
www.steinbuhler.com/html/our_research.html
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 4 of 79
4
• Summary of results of major international piano competitions; history of the
piano keyboard; analysis of the ‘barriers’ to change (Booker & Boyle, 2011).
• Exploration of the biomechanical and physiological factors relevant to hand
size/keyboard size (Boyle, 2013).
2.3 The authors’ study of hand size - aims and methodology
In 2011 the authors began collecting hand span measurements from pianists. The
principal intention was to gather sufficient data to make objective statements about
hand spans for pianists overall and about differences based on gender and ethnicity. The
data would enable direct comparison of results with those reported by Steinbuhler and
Wagner. An assessment was also made of the ‘status’ of each pianist, based on their
demonstrated success as a performer, to see if there was a possible relationship
between pianists’ level of accomplishment and hand span. Smaller databases were also
developed for children and teenagers, and for non-pianist university students, mainly for
comparative purposes.
The authors focused primarily on active hand span measurements, in particular:
• The 1-5 spans for left and right hands, defined as the distance from the outside
tip of the thumb to the outside tip of the little finger when the hand is at
maximum stretch on a flat surface.
• The 2-5 spans for left and right hands, defined as the distance from the outside
tip of the index finger to the outside tip of the little finger when the hand is at
maximum stretch on a flat surface.
It is essential to note that these two-dimensional metrics do not equate directly to
comfortable piano playing – obviously, it is not possible to play with a hand stretched flat
along the keyboard. However, the metrics were chosen in order to achieve consistency in
measurement from one person to the next, and to allow comparison of results with those
from previous studies.
The following table provides a summary of equivalent hand span stretches on a flat
surface in terms of white keys covered on a standard keyboard, measured from the
outside edges of the white keys.
Number of white keys
5
6
7
8
9
10
11
12
Inc
hes
4.6
5.5
6.5
7.4
8.3
9.3
10.2
11.1
Centimetres
11.7
14
16.5
18.8
21.2
23.5
25.9
28.3
The table can be used to put the authors’ hand span data into context, reflecting the
wide range of hand spans encountered among pianists. But as noted the measures do
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 5 of 79
5
not equate to comfortable playing (i.e. without any tension). For example, the table
shows that a span of 7.4 inches corresponds to the width of eight keys but a pianist with
that span will not be able to play the octave interval with any comfort. In Section 7.2 we
include a discussion of the minimum 1-5 and 2-5 spans required for comfortable playing.
The data collection process and the results of the subsequent statistical analysis are
summarised in this paper. The results are then examined in a wider context of
information from other sources, including performing arts medicine, biomechanics, piano
competition results and commentary from pianists about their experiences with ESPKs.
To draw conclusions about ‘all pianists’ world-wide one would need to randomly select a
sample from that group. Clearly this is unrealistic. Even sampling from pianists in a
particular country, state or city is impracticable. Hence, like the studies mentioned
above, the surveys undertaken by the authors were based on judgement sampling.
However, for much of the following analysis, it is assumed that each survey group can
be treated as an acceptable random sample.
Wagner (1988) collected many hand, wrist and arm measurements to a high level of
accuracy including spans between all fingers, hand widths and breadths, and passive
mobility measures. His measurement processes were time-consuming and required
specialised equipment. The authors decided on a relatively quick and simple data
gathering process involving just four hand span measurements, active thumb to fifth
finger and second to fifth finger spans for both left and right hands. Three databases
were developed.
The Adult Pianist database is the principal database and the main focus of the analysis in
this paper. It includes hand span measurements from 473 pianists, all of whom were
aged 18 years or over.
The Young Pianist database involves data from 49 children and teenagers between the
ages of 6 and 17. Most of these children and teenagers were students of one of the
authors, Erica Booker.
The Business Student database involves data collected as class-room exercises by two
business statistics academics at Monash and Deakin Universities (in Melbourne,
Australia). In collaboration with the authors, they collected hand span and other
pertinent data from their students for use in their teaching of introductory business
statistics during 2013 and 2014.
The following data were gathered for the Adult Pianist and the Young Pianist databases:-
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 6 of 79
6
1. Name
2. Gender
3. Handedness [Left or Right or Ambidextrous]
4. For left and right hands separately – active thumb to fifth finger and second to
fifth finger spans [cm or inches]
5. Ethnic background [Caucasian, Asian, Other/Mixed]
6. Any other musical instrument they play/have played.
Additional information sought for the Young Pianist database included age, date of
measurement and approximate AMEB level
2
.
For the Adult Pianist database, no strict acceptance criteria were adopted to define
‘pianist’ except that each participant had had piano lessons for at least a few years at
some stage in their life. The authors made their own judgement of the status (‘level of
acclaim’) of each pianist, based largely on published biographies. The three categories
adopted: ‘International’, ‘National’ and ‘Regional/Amateur’, are explained further in
section 3.5. Names were recorded on a confidential basis as a cross-check to ensure the
same pianist had not been included twice. Measurements were made in either
centimetres or inches using measuring tapes, rulers or a printed copy of Steinbuhler’s
‘Hand Gauge’
3
. In nearly all cases measurements were undertaken by the authors but for
approximately 30 Monash university piano students included in the Adult Pianist
database, hand spans were self-measured using a measuring sheet with verbal
instructions from one of the authors.
Those surveyed for the Adult Pianist database were mainly Australian pianists
encountered during the three-year period (2011-14) and included piano teachers and
academics attending piano pedagogy conferences and seminars, piano students at
university or undertaking private lessons, plus piano playing acquaintances known to the
authors. Several visiting professional performers were included in this database of 473,
plus a small number of piano students and teachers whom the authors met in the USA
and UK. The vast majority of pianists surveyed for the Adult Pianist database had
reached a reasonably high level of proficiency: piano professionals (teachers,
performers), those undertaking tertiary level classical piano studies or committed
amateurs learning from a teacher and/or involved in public performances.
The Business Student database was developed to provide the authors with a non-pianist
data set for comparative purposes. During tutorials, students recorded their own data on
a single-page A4 form which included a printed ‘measuring gauge’ marked in centimetres
2
AMEB – Australian Music Examinations Board
3
www.steinbuhler.com/html/handsizepage.html
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 7 of 79
7
(copy in Appendix 1). Hand spans were self-measured under the direction of tutors and
the A4 survey forms passed on to the authors for data entry. Names were not collected.
As many of these students were not pianists, a question was included as to whether they
had ever learnt the piano, and if so, whether they were still playing. (The completed
Excel database was subsequently used by the academics for their teaching purposes as a
practical example of statistics.)
Measurements for all three data sources were entered into Excel databases. Hand span
measurements in equivalent inches or centimetres were generated, and ratios of 1-5
(thumb to fifth finger) versus 2-5 (second to fifth finger) active spans were calculated.
While none of the three databases were based around strict random sample selection of
respondents, the authors have treated each database as being a random sample for
each group surveyed. All three data sets were analysed with statistical software using
both descriptive and inferential methods.
The results described in the following section (3) relate to the 473 pianists in the Adult
Pianist database. Analysis of the Business Students and Young Pianists databases are
described in sections 4 and 5 respectively.
3.0 SURVEY OF ADULT PIANISTS – ANALYSIS AND RESULTS
Left hand and right hand measurements of the 473 surveyed in the Adult Pianist
database were initially compared to see if there was any significant difference between
their left and right hand spans. The analysis indicated that, while it was uncommon for
individual pianists to have exactly the same measurements in both hands
4
, when
considered as a group, the statistical patterns revealed for right hand measurements
were emulated for the left hand. Accordingly, the authors made the decision to analyse
just the right hand 1-5 and 2-5 span results. Thus, the results described here focus on
right hand data only, with the exception of the analysis of left-right differences for
cellists and guitarists (section 3.8). Note also that Wagner (1988) found very little
difference between the left and right hands in most of his active hand span measures.
[Analysis by the authors of the smaller 1-5 span of each person (i.e. using the smaller of
their right and left hand spans) showed the statistical summary measures to be only
marginally smaller than for the right hand 1-5 span measures.]
Left-handed respondents in the survey were outnumbered by right-handed respondents
by a factor of 11 to 1. The general prevalence of left-handed people in the population is
4
L-R differences of between 0.1 and 0.3 inches were common; in some cases the differences were greater.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 8 of 79
8
10%.
5
A comparison of hand spans between left-handed and right-handed respondents
showed no significant differences in size between the two groups.
For the remainder of this paper:
• ‘1-5 span’ means right hand active span from the outside tip of the thumb to the
outside tip of the fifth finger
• ‘2-5 span’ means right hand active span from the outside tip of the second finger
to the outside tip of the fifth finger
• ‘active span’ means the maximum unaided stretch when the hand is spread out
on a flat surface.
3.1 All adult pianists as a group
How much variation in hand spans is there among pianists?
Key statistical summary measures were calculated for all 473 respondents in order to
understand the group overall. These are summarised in Table 1 in Appendix 2.
6
These
results show that there is significant variation in hand spans between pianists. The key
results to note are:
• The very large range from 6.4 to 10.8 inches (16.3 to 27.4 cm) for 1-5 spans.
This means that the respondent with the largest hand (a male) can span 4.4
inches (11.2 cm) more than respondent with the smallest hand (a female). This
is equivalent to about four and a half white keys on the standard keyboard.
• The very large range from 4.7 to 8.6 inches (11.9 to 21.9 cm) for 2-5 spans. This
distance of 3.9 inches (10 cm) is equivalent to approximately four white keys on
the standard keyboard. Again, the largest 2-5 span belongs to a male and the
smallest to a female but they were not the same two people as for the 1-5 span.
• The average overall (using the arithmetic mean) is 8.2 inches (21 cm) for the 1-5
span while the median (or middle value when the data are ranked in ascending
order) is the same at 8.2 inches (21 cm). The mean is 6.4 inches (16.2 cm) for
the 2-5 span with a median of 6.3 inches (16.0 cm). However, there is significant
variation either side of the means.
Figures 1 and 2 group all hand spans into 0.2 inch classes or groups within the Adult
Pianists database.
5
Wikipedia, 2014
6
A note on rounding: generally summary statistics are given to one decimal place, while some other measures,
for example, standard deviations and the difference between two means, are given to two decimal places. Any
apparent anomalies will be due to rounding.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 9 of 79
9
Figure 1: Adult pianists – 1-5 spans Figure 2: Adult pianists – 2-5 spans
0
10
20
30
40
50
60
5 6 7 8 9 10 11 12
Number
Right hand 1-5 span (inches)
All Adults (18+) | Frequency Interval
0
10
20
30
40
50
60
70
80
4 5 6 7 8 9 10
Right hand 2-5 span (inches)
All Adults (18+) | Frequency Interval
Number
The significant ranges for both hand span measures, as discussed above, are highlighted
by these graphs. However, the graphs also highlight the bimodal nature of the two
distributions, with two peaks, or high points, in each graph (at approximately 8 and 9
inches in Figure 1 and at approximately 6.4 and 6.6 inches in Figure 2).
Note that the results in Table 1 and Figures 1 and 2 do not differentiate between males
and females or between varying ethnic backgrounds. These factors are taken into
account in the following sub-sections and it is demonstrated that the first, and taller,
peak in each graph is attributable to the smaller hands of females while the second
peak, to the right, is due to the larger hands of males. (The ‘female’ peaks are taller
because almost twice as many females (314) as males (159) were surveyed but the
same bimodal shape would occur if the graphs plotted the percentage rather than the
number in each interval.)
3.2 Gender differences
Do male pianists have hand spans significantly larger than females?
Analysing males and females separately gives the statistics for the 1-5 spans shown in
Table 2 (Appendix 2). While there are approximately twice as many females as males, it
is the summary measures which are important. These results show that male pianist 1-5
hand spans tend to be significantly larger than for females. The key results to note from
Table 2 are:
• The arithmetic mean, the median and the quartiles all indicate a difference of
about 1 inch (2.5 cm) in favour of males. This difference is slightly more than the
width of one white piano key.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 10 of 79
10
• The smallest male span of 7.8 inches (19.7 cm) is close to the female average
value of 7.9 inches (20.1 cm) meaning that close to 100% of males surveyed
have hands bigger than the average female.
• In the sample group, 75% of males have 1-5 spans 8.5 inches (21.6 cm) or
greater (using the first quartile) while 75% of females have 1-5 spans of 8.3
inches (21 cm) or less (using the third quartile). Furthermore, the data show that
83.7% of males have hand spans bigger than 8.3 inches, meaning the top 83.7%
of males outrank the bottom 75% of females.
• Results of a two-tail hypothesis test (assuming random sampling), indicate that
this gender difference between arithmetic means is significant at the 5% level,
with a p-value of 0.0000.
• And we can be 95% confident that the average male 1-5 span is at least 0.88
inches (2.2 cm) bigger than the average female span and the difference could be
as much as 1.09 inches (2.8 cm).
The table also shows significant variation within each gender, the range from minimum
to maximum span being 3 inches (7.7 cm) for males and 3.1 inches (7.9 cm) for
females. This is a significant finding in its own right, and highly relevant to the discussion
later in this paper.
The disparity between males and females is clearly illustrated in the two frequency
curves shown in Figure 3, and also in the column chart (Figure 4) which groups 1-5
spans into four size categories.
Figure 3: Adult pianists – RH 1-5 span by gender, frequency curve
0
20
40
60
80
100
120
5 6 7 8 9 10 11 12
Right hand 1-5 span (inches)
Adults (18+) | Frequency Curve by Gender
Number
Males
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 11 of 79
11
Figure 4: Adult pianists – RH 1-5 span by gender, column chart
0
20
40
60
80
100
120
140
160
180
6.0 to < 7.0
7.0 to < 8.0
8.0 to < 9.0
9.0 to < 10.0
10.0 to 11.0
Right hand 1-5 span (inches)
Adults (18+) | Right hand 1-5 span (inches) v Gender
F
M
Number
The key factors to note are:
• Males are generally located to the right of females, that is, towards the larger
hand spans.
• The vertical line in Figure 3 shows that only a very small minority of men (2.0%
of all males in the sample) have 1-5 spans below the female mean of 7.9 inches
(20.1 cm).
• Conversely, the figure shows only a very small minority of women have 1-5 spans
above the male mean of 8.9 inches (22.6 cm).
Detailed cross-tabulations by gender are presented in Table 3 (Appendix 2). The
preponderance of males in the larger hand span groups is again highlighted. The ‘Percent
of Columns’ table shows that while only 2.5% of males have 1-5 spans less than 8
inches, 54.2% of females (3.2% plus 51.0%) have spans of less than 8 inches.
Conversely, 40.2% of males (35.8% plus 4.4%) compared with only 2.9 % of females
have 1-5 spans of 9 inches or more, with no females above 10 inches.
Summary statistics for 2-5 spans were also calculated for each gender and these are
presented in Table 4 (Appendix 2). These results show also show that males tend to
have significantly larger 2-5 spans than females. Key results to note are:
• Male 2-5 spans cover a greater range (3.6 inches, or 9.2 cm) than those of
females (3.0 inches, or 7.6 cm).
• The gender difference in the average values is again apparent with a mean of 6.7
inches (17.0 cm) for males and a mean of 6.2 inches (15.7 cm) for females, a
difference of 0.5 inches (about 1.3 cm).
• As expected, this difference is less in absolute terms than for the 1-5 span.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 12 of 79
12
• Results of a two-tail hypothesis test indicate that this gender difference in the two
means is significant at the 5% level, with a p-value of 0.0000.
• And we can be 95% confident that the average male 2-5 span is at least 0.42
inches (1.07 cm) bigger than the average female span and that the difference
could be as much as 0.62 inches (1.57 cm).
The table also shows significant variation within each gender. In particular, note that the
male 2-5 span range of 3.6 inches (9.2 cm) is greater than the male 1-5 span range of
3.0 inches (7.7 cm) as shown in Table 2.
The gender differences for the 2-5 span are illustrated in the two frequency curves
shown in Figure 5. While the disparity between males and females is not as great for the
1-5 span, the graph clearly shows males located to the right. Further analysis of the data
show that 73% of females have 2-5 spans less than 6.5 inches (16.5 cm), whiles 65% of
males have 2-5 spans of 6.5 inches or more.
Figure 5: Adult pianists – RH 2-5 span by gender
0
20
40
60
80
100
120
3 4 5 6 7 8 9 10
Right 2-5 span (inches)
Adults (18+) | Frequency Curve by Gender
Females
Males
Number
The relative flexibility of the second to fifth fingers in relation to total (thumb to fifth
finger) active hand span was assessed by calculating the ratio of 2-5 versus 1-5 values.
Table 5 (Appendix 2) shows these ratios for males and females separately. Key points to
note are:
• The mean 2-5:1-5 ratio for females (0.78) is higher than for males (0.75).
• This difference in the two sample means is significant at the 5% level with a p-
value of 0.0000.
• These results point to greater relative flexibility between non-thumb digits among
females, although this small statistical difference is of doubtful practical
significance.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 13 of 79
13
Figure 6 shows separate regression analysis plots of 1-5 versus 2-5 hand spans for
males and for females.
Figure 6: Adult pianists – RH 1-5 span versus 2-5 span by gender
4
4.5
5
5.5
6
6.5
7
7.5
8
8.5
9
6 7 8 9 10 11 12
Right hand 1-5 span (inches)
Adults (18+) | '1-5 span' v '2-5 span' by Gender with Linear Fit
Right hand 2-5 span (inches)
Females
Males
As would be expected, in both cases, the bigger the 1-5 span, generally the bigger the
2-5 span. The relationship is stronger for females with an R
2
value of 0.56, compared
with an R
2
of 0.39 for males. The bigger hands of males are demonstrated by the scatter
of points noticeably to the right of the scatter for females.
3.3 Ethnic differences by gender
Do hand spans vary according to ethnic background?
Of the 473 pianists surveyed, 124 were categorised as Asian and 332 as Caucasian. The
Asian category includes those originating from north-eastern Asia, south-eastern Asia or
southern Asia (Indian subcontinent, Sri Lanka). The Caucasian category includes those
identifying themselves to be of European or Middle Eastern/North African descent. The
majority of the Asian group were of Chinese ethnicity while the majority of the Caucasian
group had European ancestry.
Because of the very small sample size, those identified as southern African were omitted
from this analysis of ethnic differences. Those of mixed Asian/Caucasian descent were
also omitted from this analysis. After omitting 17 respondents, the remaining 456 were
categorised as follows:
Female Male Total
Asian 87 37 124
Caucasian 216 116 332
Total 303 153 456
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 14 of 79
14
Analysis of the Adult Pianist database (males and females combined) showed that the
adult Caucasians in the group have, on average, 1-5 spans 0.3 inches (7 cm) larger than
the adult Asians. The Caucasians have a bigger range, having both the smallest and
largest spans. The results clearly show that hand spans do vary according to ethnic
background.
Considering each gender separately, is it still the case that Caucasians have bigger
hands than Asians?
For this analysis, the database was first split by gender and then by ethnicity (Caucasian
versus Asian). Summary measures for the 1-5 and 2-5 spans of Asian and Caucasian
males are compared in Tables 6 and 7 (Appendix 2). These results show that 1-5 hand
spans for Caucasian males tend to be significantly larger than for Asian males, while
there is no significant difference in 2-5 spans. Key points to note are:
• For male 1-5 spans, arithmetic means are 9.0 inches (22.8 cm) for Caucasians
and 8.7 inches (22.0 cm) for Asians, a difference of 0.3 inches (0.8 cm).
• The difference in the two sample means is significant at the 5% level, with a p-
value of 0.0033.
• And we can be 95% confident that the average Caucasian male 1-5 span is at
least 0.11 inches (0.27 cm) bigger than the average Asian male span and the
difference could be as much as 0.52 inches (1.3 cm).
• In the case of 2-5 spans however, Asians have a slight advantage, but the
difference is small and not statistically significant.
• The range of hand spans is much greater among Caucasians, being most
pronounced for the 1-5 spans which show maxima of 10.8 inches (27.4 cm) for
Caucasians and 9.4 inches (24.0 cm) for Asians.
Figure 7 is a column chart which groups 1-5 values for males into four hand span
categories. The frequencies in each hand span group are expressed in terms of
percentage splits for each ethnic group, thereby accounting for the different numbers in
each of the two groups. It clearly shows the tendency for Caucasian male hands to be
bigger, with the greater proportion of Asian male 1-5 spans below 9 inches and no spans
above 10 inches.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 15 of 79
15
Figure 7: Adult pianists – Male RH 1-5 span by ethnicity (percent)
The same analysis by ethnicity was repeated for females. Summary measures for the 1-
5 and 2-5 spans of Asian and Caucasian females are compared in Tables 8 and 9
(Appendix 2). These results show that 1-5 hand spans for Caucasian females tend to be
significantly larger than for Asian females (though not as large a difference as for
males), while there is no significant difference in 2-5 spans between Caucasian and
Asian females. Key points to note are:
• For female 1-5 spans, arithmetic means are 8.0 inches (20.2 cm) for Caucasians
and 7.8 inches (19.8 cm) for Asians, a difference of 0.2 inches (0.4 cm) in favour
of Caucasians.
• The difference in the two sample means is significant at the 5% level, with a p-
value of 0.0127.
• And we can be 95% confident that the average Caucasian female 1-5 span is at
least 0.04 inches (0.1 cm) bigger than the average Asian female span and the
difference could be as much as 0.30 inches (0.76 cm).
• Looking at 2-5 spans, the mean and median results are the same for both female
groups.
• Once again, the range of values for both 1-5 and 2-5 spans is much greater for
Caucasians than for Asians.
Figure 8 illustrates these results in a column chart like that used above for males. It
clearly shows the greater proportion of Asian females with 1-5 spans of less than 8
inches (20.2 cm), indicating the Caucasian females tend to have bigger 1-5 spans than
Asian females.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 16 of 79
16
Figure 8: Adult pianists – Female RH 1-5 spans by ethnicity (percent)
The greater range of hand spans among Caucasians, whether male or female, (as
illustrated in Tables 6 to 9) is worth further investigation but beyond the scope of this
paper. The dataset seems to indicate that for both males and females, the very smallest
hands and the very largest hands are more likely to belong to those with a Caucasian
background.
The ratio of 2-5 to 1-5 spans is higher among Asians. (This follows logically from their 1-
5 spans being smaller than Caucasians but their 2-5 spans being very similar.) The mean
ratio for Asian females is 0.80 compared to Caucasian females of 0.78. The mean ratio
for Asian males is 0.78 compared to Caucasian males of 0.75. This difference is
significant statistically at the 5% level for both males and females.
3.4 Gender differences within ethnic groups
For each ethnic group, how do hand spans vary according to gender?
Tables 10 and 11 (Appendix 2) show gender differences in 1-5 and 2-5 spans for all
Caucasians in the sample group broken down by gender. These results show that
Caucasian male 1-5 hand spans tend to be significantly larger than for Caucasian
females. Key points to note are:
• The gender differences in Caucasian hand spans are similar to those for all adult
pianists (see section 3.2). In particular, Caucasian males have an average 1-5
span 1 inch (2.6 cm) greater than for Caucasian females.
• This difference in means is also statistically significant at the 5% level, with a p-
value of 0.0000.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 17 of 79
17
• And we can be 95% confident that the average Caucasian male 1-5 span is at
least 0.90 inches (2.3 cm) bigger than the average Caucasian female span and
the difference could be as much as 1.15 inches (2.9 cm).
• There is little difference between genders in the ranges for 1-5 spans, but for 2-5
spans, the range for males is significantly greater than for females.
The over-riding feature of the data is that Caucasian males as a group have significantly
larger hands than Caucasian females, illustrated in Figure 9. The vertical lines indicate
the mean (average) for each group and the graph clearly shows that the vast majority of
Caucasian females have 1-5 spans less than the mean for Caucasian males, while the
vast majority of Caucasian males have 1-5 spans greater than the mean for Caucasian
females. In fact, 97% of Caucasian females have 1-5 spans below the mean for
Caucasian males of 9.0 inches (22.8 cm), while only 1.7% of Caucasian males have 1-5
spans below the mean for Caucasian females of 8.0 inches (20.2 cm).
Figure 9: Adult pianists – Caucasian RH 1-5 span by gender
0
10
20
30
40
50
60
70
80
5 6 7 8 9 10 11 12
Right hand 1-5 span (inches)
Adults (18+) Caucasian | Frequency Curve by Gender
Number
Males
Females
Tables 12 and 13 (Appendix 2) show gender differences in 1-5 and 2-5 spans for all
Asians in the sample group. These results show that Asian male 1-5 hand spans tend to
be significantly larger than for Asian females. Key points to note are:
• The gender differences in Asian hand spans are similar to those for all adult
pianists (see section 3.2). In particular, Asian males have an average 1-5 span
0.9 inches (2.2 cm) greater than for Asian females.
• This difference is also statistically significant at the 5% level, with a p-value of
0.0000.
• And we can be 95% confident that the average Asian male 1-5 hand span is at
least 0.71 inches (1.8 cm) bigger than the average Asian female span and the
difference could be as much as 1.05 inches (2.7 cm).
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 18 of 79
18
• The range of values for 1-5 spans is greater for Asian females than Asian males,
but there is little gender difference in the range in the case of 2-5 spans. (This
pattern differs from that for Caucasians where males have a bigger range than
females.)
The overriding feature of the data is that, while it is not quite as pronounced as for
Caucasians, Asian males as a group have significantly larger hands than Asian females,
as illustrated in Figure 10. The vertical lines indicate the mean (average) for each group
and the graph clearly shows that the vast majority of Asian females have 1-5 spans less
than the mean for Asian males, while the vast majority of Asian males have 1-5 spans
greater than the mean for Asian females. In fact, 98% of Asian females have 1-5 spans
below the mean for Asian males of 8.7 inches (22.1 cm), while 97% of Asian males have
1-5 spans above the mean for Asian females of 7.8 inches (19.8 cm).
Figure 10: Adult pianists – Asian RH 1-5 span by gender
0
5
10
15
20
25
30
35
5 6 7 8 9 10
Right hand1-5 span (inches)
Adults (18+) Asian | Frequency Curve by Gender
Number
Females Males
A key finding from the analysis by gender and ethnicity described in sections 3.2 to 3.4
is that gender differences in both 1-5 and 2-5 spans are much greater in magnitude than
ethnic differences. To summarise:-
• Males generally have 1-5 spans about 1 inch (2.5 cm) greater than females. This
still applies whether analysing Caucasians by gender or analysing Asians by
gender.
• The difference between male Caucasian and male Asian 1-5 spans is considerably
less, with Caucasians on average 0.3 inches (0.8 cm) larger, while the difference
between Caucasian and Asian females is only 0.2 inches (0.5 cm).
• Male 2-5 spans are, on average, about 0.5 inches (1.3 cm) larger than females,
regardless of ethnicity.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 19 of 79
19
• The ethnic difference in 2-5 spans, both for males and for females, is very slight
and was not found to be statistically significant in the authors’ sample.
Figure 11 provides a useful means of comparing 1-5 spans for sub-groups of the Adult
Pianists. The top box plot clearly shows the generally larger spans of the Caucasian
males. Male Asians tend to have the next largest hands. Caucasian females have
noticeably smaller hands, with female Asians smaller still. The bottom box plot includes
all Caucasians and Asians together. Comparing the largest and smallest groups, the
sample mean for Caucasian males exceeds the sample mean for Asian females by 1.19
inches (3 cm) – a width significantly greater than one white piano key.
Figure 11: Adult pianists – RH 1-5 span by ethnicity and gender
6 7 8 9 10 11 12
Right hand 1-5 span (inches)
Adults (18+) | Multiple Box Plot by Ethnicity and Gender
Male-Caucasian:RH 1-5
Male-Asian:RH 1-5
Fem-Caucasian:RH 1-5
Fem
-
Asian :RH 1
-
5
All:RH 1-5
3.5 Hand span by level of acclaim
Do acclaimed pianists tend to have bigger hands?
Pianists in the Adult Pianist database were categorised according to ‘level of acclaim’
based on their success as a performer. Pianists with a certain level of acclaim were
identified by the authors at two levels.
Firstly, those with an ‘International’ profile based on:-
• Long-standing solo performing career, including in major concert venues around
the world, covering a wide range of repertoire
• Internationally acclaimed recordings covering a wide range of repertoire
• Prize-winner in major international piano competition(s).
Secondly, those with a ‘National’ profile based on:-
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 20 of 79
20
• Long-standing professional performing career in their home country, either as a
soloist, accompanist or chamber musician
• Nationally acclaimed recordings
• Prize-winner in significant national competition(s).
The remaining pianists in the database were classed as ‘Regional/Amateur’ based on
either:-
• The piano forming an important aspect of their adult lives, for example, as
teachers, occasional performers (for example, in community fund-raising
concerts) and/or as keen amateur recreational pianists, or
• The piano being part of their lives in the past, having learnt as a child or teenager
but no longer playing regularly.
The issue of repertoire range for ‘International’ performers is considered relevant to
hand span because repertoire from the Romantic and later periods tends to require a
larger span than repertoire from the Baroque or Classical periods (see later discussion in
section 7). Twelve pianists (10 male, 2 female) were identified as ‘International’, while
51 (30 male, 21 female) were identified as ‘National’.
Considering all adult pianists together, summary statistics for the 1-5 spans of the three
‘level of acclaim’ groups are compared in Table 14 (Appendix 2). These results show that
internationally acclaimed pianists tend to have bigger hands than the nationally
acclaimed, who in turn have bigger hands than the remaining ‘Regional/Amateur’
pianists.
The key results to note for the Internationals versus the other groups are:
• The International group has a maximum 1-5 span of 10.8 inches (27.4 cm) while
the minimum of 8.8 inches (22.4 cm) was close to the mean 1-5 span for all
males of 8.9 (22.6 cm) inches (Table 2).
• The minimum 1-5 span for the International group is bigger by 1.6 inches (4.1
cm) than the minimum span for the National group and bigger by 2.4 inches (6.1
cm) than the minimum span for the Regional/Amateur group.
• The mean and median 1-5 spans for the International group are approximately
1.0 inch (2.5 cm) larger than those measures for the National group and 1.3
inches (3.3 cm) larger than for the Regional/Amateur group.
• All of the differences in means are significant at the 5% level.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 21 of 79
21
• Further highlighting the significantly bigger hands of the internationally acclaimed
group is that all 12 Internationals have 1-5 spans bigger than approximately 75%
of Nationals and Regional/Amateurs combined (using the third quartiles).
The key results to note for the National group versus Regional/Amateur group are:
• The mean and median for the National group are about 0.3 of an inch (0.75 cm)
larger than for the Regional/Amateur group.
• The two quartiles for the National group are also higher by 0.2 to 0.4 inches (0.5
to 1 cm).
• However, the range of values for the Regional/Amateur group is greater than for
the National group, with the lowest minimum of all respondents but a maximum
close to the maximum for the International group. This very wide range for the
Regional/Amateur group is not surprising, given that it includes pianists from
many walks of life, many of whom do not perform regularly any more.
Table 15 (Appendix 2) gives cross-tabulations of level of acclaim by three 1-5 span
groups. Figure 11 illustrates the ‘Percent of columns’ information from Table 15 in the
form of a frequency chart. The table and chart clearly show the increasing proportions of
pianists of National and International status as hand span increases. Every acclaimed
pianist in the International group has a right 1-5 span of 8.5 inches (21.6 cm) or more.
And in fact, further analysis shows that 9 out of the 12 internationally acclaimed pianists
have 1-5 spans of 9.3 inches (23.6 cm) or more. The table clearly illustrates the pattern
of the higher the acclaim, the bigger the hand.
Figure 11: Adult pianists – RH 1-5 span by level of acclaim (percent)
Table 16 (Appendix 2) presents the same summary statistics for the three ‘level of
acclaim’ groups for 2-5 spans. Key findings from this analysis are:
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 22 of 79
22
• The differences in mean and median values between the three groups follow a
similar pattern to the 1-5 spans. That is, Internationals’ hands are larger than
Nationals which are in turn, larger than Regional/Amateurs. However, the
absolute differences are not as great as the 1-5 span differences.
• The difference between the means for International and National groups is
significant at the 5% level, whereas the National – Regional/Amateur difference is
significant at the 10% level.
• The range of values for the International group is only about half that of the
National and Regional/Amateur groups.
3.6 Level of acclaim by gender
In order to investigate gender differences in pianists’ status, Table 17 (Appendix 2)
presents cross tabulations of 1-5 spans by gender versus level of acclaim. Figure 12 is a
column chart illustrating the ‘Percent of columns’ cross tabulation which directly
compares the relative proportions of males and females within each of the three ‘level of
acclaim’ groups.
FIGURE 12: Adult pianists – Level of acclaim by gender (percent)
The proportions of Internationals, Nationals and others might not be representative of
the population of adult pianists at large due to the sampling methods used, however, the
results clearly show the higher proportions of males compared with females who have
reached a National level of acclaim, and even more obviously, the relatively high
proportion of males from this sample at the International level.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 23 of 79
23
From an analysis of 1-5 spans by level of acclaim for each gender separately, the data
show that:-
• For males, there is a significant difference (5% level) in hand spans between the
Internationals and Nationals and also between the Internationals and
Regional/Amateurs (in favour of Internationals in both cases). However, the
difference in hand spans between the National and Regional/Amateur groups is
not significant at the 10% level.
• As there were only two females in the International category, these were
combined with the National group for analysis. The difference in 1-5 spans
between this combined acclaimed group and the Regional/Amateur group was
found to be significant at the 5% level (in favour of Internationals/Nationals).
Figure 13, a dot chart like that compiled by David Steinbuhler (2004), clearly illustrates
the findings that Internationals tend to have significantly larger hands both overall and
within each gender. It shows the complete set of right hand 1-5 active hand span data
for all 473 adult pianists, with gender and level of acclaim illustrated by colour and
marker.
Figure 13: Adult pianists – right hand 1-5 spans by gender and level of acclaim
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 24 of 79
24
This graph shows a number of noteworthy features, some of which have already been
highlighted in earlier sub-sections:
• Males tend to be located towards the right, i.e., towards the larger 1-5 spans.
• The Internationally and Nationally acclaimed males are located in this top section.
• Females tend to be located towards the left, i.e., towards the smaller 1-5 spans.
• With one exception, the Internationally and Nationally acclaimed females are
located in the top section of the female distribution, around the female average 1-
5 span or higher. The two female Internationals are at 9.0 inches or higher.
Clearly, Figure 13 illustrates the authors’ findings that internationally acclaimed pianists
tend to have relatively large hand spans and as males tend to have larger hands, they
dominate the ranks of internationally acclaimed pianists. (Refer to discussion in section
7.1.5.).
The lone female at the ‘National’ level in the authors’ study who has a 1-5 span of only
7.2 inches (18.3 cm), later (in section 8) defined as ‘very small’ – is worthy of brief
discussion. She is an extremely talented young pianist, who is greatly assisted by having
an exceptionally wide 2-5 span for her hand size, in fact just within the ‘large hands’ 2-5
group according to the definitions in section 8. Her 2-5:1-5 ratio is approximately 0.85.
She has an eclectic repertoire including Mozart, selected works of Beethoven, Ravel and
late Brahms (Brahms and Ravel generally requiring significant modifications to the
score), and many carefully chosen contemporary works. She often performs with other
chamber musicians or orchestras. A pianist with her hand span of 7.2 inches would find
it impossible to perform extended fast passages of octaves or octave-based chords due
to the physical constraints (having to play octaves on the front edge of the white keys)
and build-up of pain that would become apparent due to the extreme stretch required. A
pianist with this span cannot play a ninth on the conventional keyboard.
3.7 Level of acclaim by ethnicity
Cross tabulations between level of acclaim and ethnicity have not been included here
because the small sample did not enable conclusions to be drawn with any confidence.
3.8 Differences between left and right hands for cellists and guitarists
There is a common belief among musicians that maintaining extreme joint positions from
playing an instrument and/or regular stretching exercises increases the flexibility in the
corresponding joints. For example, cello players are often thought to have larger left
hand than right hand spans.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 25 of 79
25
To test this belief, one question in the authors’ study asked each pianist which other
instruments they had played, if any. For the purposes of comparison with the results of
previous studies discussed below in section 6.3 (Kloeppel, 2000; Driscoll & Ackermann,
2012), those who nominated cello or guitar were compared with all other pianists in the
sample. For all 473 pianists, the results indicate a difference between left and right hand
1-5 span of only 0.03 inches (0.08 cm) on average in favour of the left hand. For the 65
pianists who had played the cello and/or guitar the average difference was 0.05 inches in
favour of the left. Looking at 2-5 span measures, the differences were slightly greater,
with a 0.11 inch difference for cellists, 0.02 inches for guitarists and 0.06 inches for all
other pianists, in favour of the left hand. These differences are all insignificant at the
10% level and in any case, are insignificant for piano playing in any practical sense. A
similar analysis of left-right hand differences among violin and viola players also found
no significant differences between this group and the adult pianists in general.
4.0 SURVEY OF BUSINESS STUDENTS – ANALYSIS AND COMPARISON WITH
ADULT PIANISTS
How do hand spans of pianists compare to the hand spans of non-pianists?
The only data the authors were able to find on hand spans for the population at large
date from the 1980s or earlier (see Section 6.5).
Given the lack of recent data, with the aid of two university academics the authors
collected data from a number of non-pianists, namely students in business statistics
subjects. The authors have assumed that this judgement sample can be treated as a
random sample for analysis purposes.
Thus, as described in section 2.3, similar hand span data as for the adult pianists were
gathered from business students at Deakin and Monash Universities. Some data
collection problems surfaced. With the Adult Pianist database, respondents were
generally closely supervised by one of the authors to ensure hands were properly
stretched. However, the academics were only able to give instructions from the front of
the class, hoping that students would make the measurements correctly and record
them accurately.
As reliability was expected to be less than for the main Adult Pianist database, a
thorough validation check of the student data was carried out by the authors prior to
analysis. As a result, several records were omitted due to missing data or measurements
that were not believable. For example, some students recorded identical values for their
1-5 and 2-5 spans while some other measurements were judged by the authors to be
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 26 of 79
26
abnormally small for adults. The measurements also show evidence of many students
rounding up or down to the nearest centimetre even though the measuring template
enabled them to measure to the nearest 2 millimetres.
Ultimately, data from 216 students were included in the analysis: 86 from Deakin
University and 130 from Monash University. All were students enrolled in business
courses, but while the gender and ethnic percentage mix may have been different to the
general university student population, there is no reason to believe the various sub-
groups in the sample differed from their larger sub-populations within the universities in
terms of hand size.
4.1 Analysis by university and semester
At each university, data were collected over two semesters. Table 18 (Appendix 3)
shows summary statistics for the 1-5 span measurements for all students in each
semester at each university. The results indicate that students in the Deakin semester B
group have significantly larger hand spans than for the other groups. It is noted that
89% of students in that group were Caucasian, whereas for the other three semester
groups Asians dominated with 75%, 85% and 94% of respondents. (Results from the
analysis of the Adult Pianist database discussed above indicate that Caucasians have
larger hand spans than Asians.) Figure 14 illustrates the results in the form of box plots
which highlight median and quartile values.
Figure 14: Business students – RH 1-5 span by campus/semester group
6 7 8 9 10 11
Right hand 1-5 span (inches)
Business Student s | Multiple Box Plot by Campus
Mon ash Sem (b):RH 1-5
Mon ash Sem (a):RH 1
-
5
Deakin Sem (b):RH 1
-
5
Deakin Sem (a):RH 1-5
All:RH 1-5
Table 19 (Appendix 3) shows summary statistics for the 2-5 span measurements. The
results show very little difference between the four groups.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 27 of 79
27
For further analysis, the four semester groups were combined to form a single Business
Student database of 216 respondents. Key findings follow; comparisons with the Adult
Pianist results are given in Section 4.6.
4.2 Gender differences
Gender differences were analysed in the same way as for the Adult Pianist database. In
the Business Student database there were 123 males and 93 females in total. Table 20
(Appendix 3) shows summary statistics for 1-5 spans for males and females.
These results show that male Business Student 1-5 hand spans tend to be significantly
larger than for females. Key findings from this analysis are:
• The arithmetic means are 8.3 inches (21.2 cm) for males and 7.6 inches (19.2
cm) for females.
• The difference in the two sample means of 0.76 inches (1.9 cm) is significant at
the 5% level.
• And we can be 95% confident that the average male business student 1-5 span is
at least 0.59 inches (1.5 cm) bigger than the average female 1-5 span and the
difference could be as much as 0.93 inches (2.4 cm).
Figure 15 illustrates the results in the form of a frequency chart. The location of males
towards the right highlights the difference between male and female 1-5 spans for the
business student data, with females generally having smaller hands.
Figure 15: Business students – RH 1-5 span by gender
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
5 6 7 8 9 10 11
Right hand 1-5 span (inches)
Business Students | Frequency Curve by Gender
Percent
Females Males
Table 21 (Appendix 3) shows similar summary statistics for 2-5 spans. Again, the
conclusion is that males in the Business Student database have significantly bigger 2-5
spans than females.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 28 of 79
28
4.3 Ethnic differences by gender
As for the Adult Pianist database, students were categorised as ‘Asian’, ‘Caucasian’ or
‘other/mixed’. The eight respondents in the last category were not included in the
analysis of ethnic differences in order to concentrate on the two major groups. Of the
females, 56 were of Asian descent and 30 of Caucasian descent. For males, 79 were
Asian and 42 Caucasian. Tables 22 and 23 (Appendix 3) compare 1-5 spans of Asians
and Caucasians, for males and females separately.
These results show that Caucasian male and Caucasian female Business Students tend to
have 1-5 hand spans significantly larger than their Asian gender counterparts. Key
results to note are:
• Caucasian males have a sample mean 0.7 inches (1.8 cm) greater than their
Asian counterparts.
• Caucasian females have a sample mean 0.28 inches (0.7 cm) greater than their
Asian counterparts.
• These differences in the means are both significant at the 5% level.
Figure 16 is a multiple box plot that summarises the 1-5 span results for the four
gender/ethnic groups. The top box plot clearly shows the generally larger 1-5 right spans
of the Caucasian males in the Business Student database. Male Asians tend to have the
next largest hands with Female Asians the smallest.
Figure 16: Business students – RH 1-5 span by ethnicity and gender
6 7 8 9 10 11
Right hands 1-5 span (inches)
Business Students| Multiple Box Plot by Ethnicity and Gender
Male-Caucasian:RH 1-5
Male-Asian:RH 1-5
Fem
-
Caucasian:RH 1
-
5
Fem
-
Asian :RH 1
-
5
All:RH 1-5
Further interesting results from this analysis are that:
• The sample mean for Caucasian males exceeds that for Asian females by 1.4
inches (3.5 cm). This difference is significant at the 5% level and we can be 95%
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 29 of 79
29
confident that, on average, Caucasian male 1-5 spans exceed Asian female 1-5
spans by at least 1.13 inches (2.9 cm) and by up to 1.58 inches (4 cm).
• The sample mean for Asian males exceeds that for Caucasian females by 0.4
inches (1.0 cm). This difference is significant at the 5% level and we can be 95%
confident that, on average, Asian male 1-5 spans exceed Caucasian female 1-5
spans by at least by 0.16 inches (0.4 cm) and by up to 0.59 inches (1.5 cm).
• Caucasian males in the Student database have only marginally smaller hands on
average than Caucasian males in the Adult Pianist database (0.2 inches or 0.51
cm less). However, Asian males in the Student database have noticeably smaller
hands than Asian males in the Adult Pianist database (0.7 inches or 1.8 cm less).
Analysis of 2-5 spans shows that ethnic differences are not statistically significant,
neither for females nor for males. While summary tables are not given, means and
standard deviations are included in section 4.6 (Table 25).
4.4 Pianists versus non-pianists
Students were asked if they had ever learnt to play the piano and if so, whether or not
they still played. Table 24 shows the results, grouped into the three classes.
Table 24: Business students – Piano playing history
Total % No. of males % No. of females
%
Total number of
responses
215 100%
123 100%
93 100%
Never learnt to
play
128 59% 89 73% 39 42%
Learnt but no
longer play
68 32% 24 20% 44 47%
Still play the
piano
19 9% 9 7% 10 11%
There are no noteworthy findings from statistical comparisons of the various groups.
Detailed information, such as AMEB level reached, was not sought from the university
students, as this was not suited to the data gathering process. It may be the case that
many of those who learnt but no longer play had only reached a relatively elementary
level. Those included in the Adult Pianist database were, on the other hand, virtually all
pianists who would have reached a high level of accomplishment, being either teachers,
university music students or professional or serious amateur performers. These results
are discussed further in section 6.5 which deals with how pianists’ hand spans might
compare with non-pianists.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 30 of 79
30
4.5 Height as a predictor of hand span
Do tall people tend to have bigger hands than short people?
In addition to hand span data, the Business Student survey also asked the students to
record their heights. The regression analysis plot of height versus 1-5 span, for all
students together, is shown in Figure 17.
FIGURE 17: Business students – RH 1-5 span versus height
5
6
7
8
9
10
11
55 60 65 70 75 80
Right hand 1-5 span (inches)
Height (inches)
Business Student s | Right hand 1-5 span v Height, Linear fit
This analysis indicates that there is a weak but clear relationship between a person’s
height and 1-5 hand span (R
2
=0.41; p=0.00). Thus, in general we would conclude that
although a person’s height provides a guide to their 1-5 span, it is not a perfect
predictor. For this data set we note the extreme example in the top left quadrant of a
person with a large 1-5 span (>9 inches) but with a height of only 63 inches (160 cm).
And from the bottom right quadrant we also note the many tall people with relatively
small hands, often smaller than much shorter people. As discussed in section 7.2, it is
quite common to hear people wrongly concluding that because a person is small that
they must have a small hand span (and conversely).
4.6 Comparison with Adult Pianist database
The results of the Business Student database analysis can now be compared with that of
the Adult Pianist database described in section 3.0. Table 25 (below) presents a selection
of means and standard deviations for various sub-groups. Relevant p-values from two-
tailed hypothesis tests are included for comparisons between the two databases. A p-
value of 0.05 is the normal benchmark used to indicate a significant difference between
arithmetic means, hence p-values below 0.05 indicate statistical significance at the 5%
level. Values between 0.05 and 0.1 indicate significance at the 10% level. Differences in
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 31 of 79
31
the 2-5 spans between the Adult Pianists and the Business Students are less likely to be
significant (indicated by high p-values) than differences in 1-5 spans (indicated by low p-
values).
Note that the Business Students sample comprises a much higher proportion of males
and also a higher proportion of Asians than the Adult Pianists sample.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 32 of 79
32
Table 25: Comparison between hand spans of Adult Pianists and Business Students
ADULT PIANIST DATABASE BUSINESS STUDENT DATABASE Test – diff.
between
means
Arithmetic
mean
Standard
deviation
Arithmetic mean Standard
deviation
P value
Sample Inches Cm Inches Cm Sample Inches Cm Inches Cm
All respondents 1-5 span* 473 8.2 21.0 0.71 1.81 216 8.0 20.3 0.72 1.82 0.0000
All respondents 2-5 span* 473 6.4 16.2 0.58 1.47 210 6.4 16.1 0.68 1.72 0.4371
Males 1-5 span 159 8.9 22.6 0.56 1.43 123 8.3 21.2 0.62 1.57 0.0000
Females 1-5 span 314 7.9 20.1 0.53 1.35 93 7.6 19.2 0.60 1.52 0.0000
Males 2-5 span 159 6.7 17.0 0.55 1.40 119 6.6 16.6 0.65 1.65 0.2761
Females 2-5 span 314 6.2 15.7 0.51 1.30 91 6.1 15.5 0.63 1.59 0.2209
Caucasian males 1-5 span 116 9.0 22.8 0.60 1.52 42 8.8 22.3 0.46 1.39 0.0637
Asian males 1-5 span 37 8.7 22.0 0.38 0.97 79 8.1 20.5 0.50 1.26 0.0000
Caucasian males 2-5 span 116 6.7 17.0 0.58 1.47 42 6.6 16.9 0.64 1.63 0.6341
Asian males 2-5 span 37 6.7 17.1 0.45 1.15 75 6.5 16.5 0.65 1.65 0.0427
Caucasian females 1-5 span 216 8.0 20.2 0.54 1.38 30 7.7 19.6 0.58 1.47 0.0278
Asian females 1-5 span 87 7.8 19.8 0.45 1.15 56 7.4 18.9 0.55 1.39 0.0000
Caucasian females 2-5 span 216 6.2 15.6 0.51 1.30 30 6.1 15.6 0.68 1.72 0.7055
Asian females 2-5 span 87 6.2 15.8 0.48 1.22 54 6.0 15.2 0.55 1.39 0.0132
‘International’ group 1-5 span 12 9.5 23.9 0.52 1.32 NA NA NA NA
‘National’ group 1-5 span 51 8.5 21.6 0.57 1.45 NA NA NA NA
‘Regional/Amateur’ 1-5 span 410 8.2 20.8 0.70 1.77 NA NA NA NA
‘International’ group 2-5 span 12 7.0 17.7 0.47 1.20 NA NA NA NA
‘National’ group 2-5 span 51 6.5 16.5 0.59 1.50 NA NA NA NA
‘Regional/Amateur’ 2-5 span 410 6.3 16.1 0.57 1.45 NA NA NA NA
*Note the different gender and ethnic mixes between the two groups. (NA = not applicable.)
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 33 of 79
33
Table 26 presents the mean RH 1-5 spans from all sample groups ranked in order, from
largest to smallest.
Table 26: Adult pianists and Business students – Ranking of
mean RH 1-5 spans
Sample group
Inches
Centimetres
ADB – All Internationals 1-5 span 9.5 24.0
ADB - Caucasian males 1-5 span 9.0 22.8
ADB - Males 1-5 span 8.9 22.6
SDB - Caucasian males 1-5 span 8.8 22.3
ADB - Asian males 1-5 span 8.7 22.0
ADB – All Nationals 1-5 span 8.5 21.6
SDB - Males 1-5 span 8.3 21.2
ADB - All adults 1-5 span 8.2 21.0
ADB – All Regional/Amateurs 1-5 8.2 20.8
SDB - Asian males 1-5 span 8.1 20.5
SDB - All adults 1-5 span 8.0 20.3
ADB - Caucasian females 1-5 span 8.0 20.2
ADB - Females 1-5 span 7.9 20.1
ADB - Asian females 1-5 span 7.8 19.8
SDB - Caucasian females 1-5 span 7.7 19.6
SDB - Females 1-5 span 7.6 19.2
SDB - Asian females 1-5 span 7.4 18.9
ADB=Adult Pianist database SDB=Business Student database
Major points to note from Tables 25 and 26 are:
• The group with the biggest span is the internationally acclaimed pianists, followed
by male Caucasians from the Adult Pianist database.
• The group with the smallest hands were Asian females from the Business Student
database.
• Those in the Business Student database had lower means than their counterparts
in the Adult Pianist database. This could partly be due to measurement problems
where the students were not made to stretch as much, combined with how they
recorded their own results. However it could also mean that pianists with larger
hands tend to continue playing and reach a higher level.
• For the Adult Pianists, Caucasian males have an advantage of about 1.2 inches (3
cm) on average over Asian females. For the Business Students, this difference is
even greater with Caucasian males having a 1.4 inch (3.6 cm) advantage over
Asian females.
• The large difference in means (0.7 inches, 1.8 cm) between Caucasian and Asian
males in the Business Student database is unlikely to reflect measurement issues
only (see discussion in section 6.5).
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 34 of 79
34
For 2-5 spans for each gender, the differences between ethnic groups are relatively
small or negligible.
The reasons why the mean hand spans of pianists might exceed those of non-pianists
and the general adult population are discussed further in section 6.5. The two main
possible reasons discussed there are: (1) the long term effect of playing the instrument
and/or stretching exercises, and (2) those with larger hands finding the task relatively
easier and achieving greater success than those with smaller hands, and hence being
more likely to continue playing (i.e. self-selection). For the comparison between the two
databases discussed here however, there is an additional factor which could also be
relevant - the different data gathering processes. While pianists in the main Adult Pianist
database were, in nearly all cases, measured directly by one of the authors, those in the
Business Student database measured themselves under the direction of business
statistics tutors. It is possible that the importance of stretching the hand to the
maximum extent along the measuring device was not adequately communicated and/or
some students did not listen intently or take the task seriously.
5.0 SURVEY OF YOUNG PIANISTS
The Young Pianist database with measurements from 49 children and teenagers (less
than 18 years) was not analysed in great detail due to the small sample size and the
non-representative make-up of ages, gender and ethnic background. There is also the
added obvious complication that their hands were, in most cases, still growing, and
growth rates will vary greatly among those of similar age.
Table 27 (Appendix 4) shows the students classified in three-year age groups broken
down by RH 1-5 span. Their actual ages ranged from 6 to 17 years. The 1-5 spans
ranged from 5.8 inches (14.7 cm) to 9.2 inches (23.4 cm). The youngest, a 6 year-old
girl, had the smallest span while the biggest span belonged to a 15 year-old boy.
The table clearly shows the predictable pattern that older children tend to have bigger
hands. This is illustrated graphically in Figure 18.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 35 of 79
35
Figure 18: Young pianists – RH 1-5 span by age and gender
5
5.5
6
6.5
7
7.5
8
8.5
9
9.5
5 7 9 11 13 15 17 19
Age (Years)
Young Pianists | RH 1-5 span v Age
RH 1-5 span (inches)
Females
Males
Figure 19 shows RH 1-5 spans versus AMEB
7
level.
Figure 19: Young pianists – RH 1-5 span by AMEB level and gender
0
2
4
6
8
10
12
5 6 7 8 9 10
RH 1-5 span (inches)
Young Pianists| Appox. AMEB level v RH 1-5 span
Approx. AMEB level
Females
Males
NB Level 9 corresponds to AMUS and level 10 corresponds to LMUS.
Further analysis of the Young Pianist database shows that some students reach a high
level of study while still young and with their hands still growing. For example, some
students in the database had achieved AMUS level (level 9 in Figure 19) while less than
15 years old (the youngest being 10) with 1-5 spans considerably less than 8 inches
(20.3 cm). The potential for injury among piano students who reach a high level at a
young age, but who must use a piano keyboard not designed for their hands, is an issue
worthy of further attention. (Refer to section 7.1.2.)
7
AMUS=Associate diploma; LMUS-Licentiate diploma
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 36 of 79
36
It is worth noting that 10 children, aged from 6 to 13 years, had 1-5 spans between 6
and 7 inches (15.2 and 17.8 cm). There were 10 adult women pianists in the Adult
Pianist database and 13 business students in the Business Student database with spans
in that range. Hence, the smallest adult female hands appear to be similar to those of
many children.
6.0 COMPARISON WITH OTHER DATA SETS ON HAND SIZE
6.1 Pianists’ hand spans and gender differences
As mentioned in section 1.1, Wagner’s survey and analysis (1984, 1988) is the most
comprehensive of any published relating to pianists. The majority of his sample of what
he called ‘professional’ pianists were teachers or students specialising in piano at the
Musikhochshule in Hannover, Germany. The school had an international reputation, with
Wagner’s subjects drawn from 28 countries.
As described in section 1.1, Wagner measured many hand size dimensions and
characteristics, including 1-5 and 2-5 spans for both hands and summarised the results
with key statistical measures. (He does not provide summary statistics for 2-5:1-5
ratios.)
A difference between the measuring processes used by Wagner and that used in the
authors’ study needs to be noted. Wagner measured active 1-5 spans from the middle of
the finger tips, while the authors’ study measured from outside edges of the finger or
thumb, consistent with Steinbuhler (2004) and other recent epidemiological and clinical
studies. Everything else being equal, Wagner’s measurement process should produce
results very close to the authors’ for the 1-5 spans, however, his results for the 2-5
spans would be slightly smaller than in the authors’ and other studies. This difference for
the 2-5 span is considered to be of the order of 0.2 inches (0.5 cm).
David Steinbuhler collected his data at the US Music Teachers National Association
(MTNA) 2004 National Conference, where attendees were invited to have their active 1-5
hand spans measured (for either their left or right hand). Of the 160 who participated,
90 were adult females, 66 were adult males and four were students not fully grown. The
distribution of their active 1-5 hand spans is shown in the chart on the Steinbuhler
website.
8
The difference in hand spans between genders is obvious from the chart.
Summary statistics from Steinbuhler’s data (excluding the four young students) were
derived previously by the authors (Boyle & Boyle, 2009).
Table 28 compares summary statistics for 1-5 spans for the Steinbuhler (2004), Wagner
(1988), and the authors’ Adult Pianist data.
8
www.steinbuhler.com/html/our_research.html
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 37 of 79
37
Table 28: Comparison of pianist 1-5 span data – inches (centimetres)
Males Females
Steinbuhler
(Mix of R
and L hands)
Wagner
(RH)
Boyle et al.
(RH)
Steinbuhler
(Mix of R
and L hands)
Wagner
(RH)
Boyle et
al. (RH)
Sample
size
66 111 159 90 105 314
Minimum 7.7 (19.6) 7.91
(20.1)
7.8 (19.7) 7.0 (17.8) 7.01
(17.8)
6.4 (16.3)
Maximum
10.2 (25.9) 10.24
(26.0)
10.8 (27.4) 8.9 (22.6) 9.33
(23.7)
9.5 (24.1)
Arithmetic
mean
8.9 (22.6) 8.93
(22.7)
8.9 (22.6) 7.9 (20.1) 8.17
(20.7)
7.9 (20.1)
Median 8.9 (22.6) 8.86
(22.5)
8.9 (22.6) 7.9 (20.1) 8.11
(20.6)
7.9 (20.1)
First
Quartile
8.5 (21.6) 8.62
(21.9)
8.5 (21.6) 7.5 (19.1) 7.83
(19.9)
7.6 (19.2)
Third
Quartile)
9.3 (23.6) 9.21
(23.4)
9.2 (23.4) 8.2 (20.8) 8.50
(21.6)
8.3 (21.0)
Range 2.5 (6.4) 2.32
(5.9)
3.0 (7.7) 1.9 (4.8) 2.32
(5.9)
3.1 (7.9)
Inter
Quartile
0.8 (2.0) 0.59
(1.5)
0.7 (1.8) 0.7 (1.8) 0.67
(1.7)
0.7 (1.8)
Standard
deviation
0.56 (1.42) 0.45
(1.14)
0.56 (1.43) 0.47 (1.19) 0.45
(1.14)
0.53
(1.35)
The major points from this comparison are:
• For males, the three sets of results are very similar for key summary measures,
in particular, the mean, median and quartile values.
• For females, the Steinbuhler and Boyle results are very similar for key summary
measures, with the Wagner results slightly higher. (The authors believe that the
Steinbuhler and Boyle data are more representative of adult female pianists – see
discussion below.)
• No females measured by Steinbuhler had hand spans of 9 inches or above.
• The range of values, for both males and females, were greatest for the authors’
survey, possibly a reflection of the larger sample size.
The authors believe that the higher female measures for Wagner’s data are explainable
by the fact that they were a select group of elite female students and therefore likely to
have bigger hands, combined with only 15% of his female students being Asian. Further
support for this contention is that females with a National or International level of
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 38 of 79
38
acclaim in the authors’ study have a mean 1-5 span of 8.2 inches, very close to
Wagner’s mean of 8.17 inches.
Table 29 compares Wagner’s 2-5 data with the authors’ results. Note that Steinbuhler
did not collect 2-5 data while Wagner did not measure 2-5 spans of all his pianists.
Table 29: Comparison of pianist RH 2-5 span data – inches (centimetres)
Males Females
Wagner Boyle et al. Wagner Boyle et al.
Sample size 44 159 60 314
Minimum 5.51 (14.0) 5.0 (12.7) 5.20 (13.2) 4.7 (11.9)
Maximum 7.52 (19.1) 8.6 (21.9) 7.12 (18.1) 7.7 (19.5)
Arithmetic mean 6.46 (16.4) 6.7 (17.0) 6.18 (15.7) 6.2 (15.7)
Median 6.42 (16.3) 6.8 (17.1) 6.18 (15.7) 6.2 (15.7)
First Quartile 6.10 (15.5) 6.3 (16.0) 5.80 (14.7) 5.9 (15.0)
Third Quartile 6.88 (17.5) 7.0 (17.8) 6.68 (17.0) 6.5 (16.5)
Range 2.01 (5.1) 3.6 (9.2) 1.93 (4.9) 3.0 (7.6)
Inter Quartile Range 0.78 (2.0) 0.7 (1.8) 0.88 (2.2) 0.6 (1.5)
Standard deviation 0.50 (1.27) 0.55 (1.40) 0.50 (1.27) 0.51 (1.30)
Key points to note are:
• Mean and median values in the authors’ study are slightly higher than Wagner’s
for males, for reasons given earlier relating to measurement procedures.
• For females, mean and median values are very similar in both studies, suggesting
that the different measurement procedure has compensated for the larger hands
of Wagner’s females.
• The range of values in the authors’ study is much greater than the ranges found
by Wagner both for males and for females (possibly due to the authors’ larger
sample).
6.2 Pianists’ hand spans and level of acclaim
In an earlier study, Wagner (1984) selected two ‘extreme’ groups of male pianists from
his database in order to compare their hand size characteristics:
• ‘Successful Performers’ – 26 well-known soloists and winners of international
competitions.
• ‘Problem Cases’ – 10 pianists who had struggled with technical problems over a
long period of time, such as lack of speed, manual dexterity or timing, and/or
those with injury problems.
While he found that there were no significant differences in hand shape measures (e.g.
length and breadth of hands) between the two groups, all active hand span measures
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 39 of 79
39
(involving all possible combinations of fingers) for the Successful Performers were
statistically significantly larger. They had a median 1-5 span of 9.2 inches (23.4 cm)
compared with 8.7 inches (22 cm) for the Problem Cases group (results for the
arithmetic mean were not provided).
For the 2-5 span comparisons (based on fewer measurements), the difference was 6.9
inches (17.5 cm) compared with 6.1 inches (15.5 cm), again significant statistically.
While not directly comparable with the three ‘levels of acclaim’ defined by the authors,
Wagner’s results support the authors’ findings that the most successful and acclaimed
pianists tend to have significantly bigger hand spans (sections 3.5 – 3.6).
6.3 Left-right hand differences
Wagner (1988) found very slight, but statistically significant, differences in some of his
active hand span measures in favour of the left hand. The data in the authors’ study
(discussed above in 3.8) also show slight differences in favour of the left hand but they
were not significant statistically. From a practical point of view, the differences in both
cases are negligible.
Kloeppel (2000) addressed the question of increased spans due to playing particular
instruments by measuring spans between fingers (excluding thumbs) and also finger
lengths for cellists, guitarists and a control group. Consistent with a common belief that
these particular musicians develop greater spans as a result of playing these instruments
over many years, she found slightly greater spreadability of the left hand than the right
hand, including the 2-5 span, among cellists compared with the control group. The same
was true for guitarists, although the 2-5 span was not measured for them given it was
not relevant to their playing. The greater spreadability was statistically significant for
cellists but not for guitarists. The slightly longer index finger in cellists (less than 1 mm)
was attributed to the development of calluses. She found no relationships with age of
starting to learn the cello. She concluded that although the changes over time in finger
spans were statistically significant in the case of cellists, they were negligible in any
practical sense (2-3 mm at most for the 2-5 span).
A recent Australian study (Driscoll & Ackermann, 2012) included left and right 1-5 hand
span measurements in an anthropometric study of 408 professional orchestral musicians
in Australia. They found the left hand span to be on average 0.2 inches (0.5 cm) larger
than the right, both for males and females. They also found significant differences
between the left (but not right) hand spans among different instrument groups, with
lower string players (cello and double bass) having the largest left hand spans (taking
gender differences into account). Left-right 1-5 span differences for lower string players
were 0.3 inches (0.8 cm) for males and 0.2 inches (0.6 cm) for females.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 40 of 79
40
6.4 Ethnic differences within the general population
It is often stated that people of Asian ethnicity have smaller hands than those of
Caucasian origin (e.g. Sakai, 1992, 2008, Furuya et al., 2006). Some comparative hand
anthropometry data has been published, e.g. Saengchaiya & Bunterngchit (2004), Nag,
Nag & Desai (2003) and Mandahawi et al. (2008), the results of which clearly show that
people of Asian descent have smaller hands than Caucasians.
Results of the study by Nag et al. indicate that the hands of Indian women are
significantly smaller than those of British, American and West Indian women –
particularly in terms of hand width. That study does include hand span data; the active
1-5 span of the sample of Indian women was only 6.8 inches (173 mm) and the active
2-5 span was 5.4 inches (137 mm).
There may be significant differences among different Asian groups. There is anecdotal
evidence that people from Korea and Northern China tend to be bigger (and have larger
hands) than those from southern China, Southeast Asia and Japan.
As discussed in 3.3, the authors’ study found the 1-5 spans of Caucasian pianists were
statistically significantly larger than for Asian pianists. But 2-5 spans were found to be
comparable for both ethnic groups, hence the 2-5:1-5 ratio is greater among Asians. The
Business Student survey (section 4.3) also showed a similar ethnic difference among a
general population of students.
The authors’ findings that Asian pianists have smaller 1-5 hand spans than Caucasian
pianists is consistent with findings from a wide range of surveys which indicate that
Asians have smaller hands overall.
6.5 Comparison of pianists with non-pianists
How do pianists’ hands compare to the population at large?
Much of the published anthropometrical data relating to human hands has not been
based on random samples of the adult population. Examples of surveys of particular
groups include measurements from armed forces personnel in the USA: e.g. Garrett
(1968, 1971), Greiner (1991), Donelson and Gordon (1996) and from industrial workers
in various countries: e.g. Nag, Nag & Desai (2003) and Saengchaiya & Bunterngchit
(2004). For Caucasian males generally, non-pianists tended to have broader hands than
pianists. Matzdorff (1968) and Garrett (1968) measured active hand spans thus allowing
direct comparison with findings from the authors’ survey and those by Steinbuhler and
Wagner.
Wagner (1988, p.117) notes that, based on studies prior to that time, musicians tend to
have greater finger spans than non-musicians. Of particular interest is Matzdorff’s
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 41 of 79
41
(1968) study which surveyed non-musicians in Germany. She measured 1-2, 1-3 and 1-
5 active spans of 373 men and 396 women of different ages. They were described as
middle class, and were all in the workforce employed in a variety of jobs. Wagner
compared her results with his own data and found that his pianists (both men and
women) had significantly higher mean values for the 1-5 span, and that the female
pianists had a significantly greater 1-3 span. There was no significant difference in 1-2
spans. He also found the variability of spans to be lower among Matzdorff’s non-
musicians.
In his specific study of male air force pilots (1968), Garrett measured active hand spans
– including both 1-5 and 2-5 spans. His mean 1-5 span was 8.5 inches (21.6 cm) and his
mean 2-5 span was 6.4 inches (16.3 cm). Table 30 summarises these findings. As it is
assumed most respondents in the Matzdorff and Garrett studies would have been
Caucasian (as in Wagner’s sample), for comparative purposes Table 30 includes the
means with those obtained for Caucasians in the authors’ Adult Pianist and Business
Student surveys.
Table 30: Comparison of hands spans of pianists with other groups
Total
sample
size (n)
Mean male
1-5 span
Mean female
1-5 span
Mean
male 2-5
span
Mean
female 2-5
span
Boyle et al
.
–
All Adult
Pianists
473
8.9 (22.6)
7.9 (20.1)
6.7 (17.0)
6.2 (15.7)
Boyle et al.
–
Adult
Pianists (Caucasian sub-
group)
332
9.0 (22.8)
8.0 (20.2)
6.7 (17.0)
6.2 (15.6)
Wagner 1988 (pianists)
214 (1
-
5)
104 (2-5)
8.93 (22.7)
8.17 (20.7)
6.5
(16.4)
-
S
teinbuhler 2004
(pianists)
156
8.9 (22.6)
7.9 (20.1)
-
-
Driscoll & Ackermann
2012 (orchestral
musicians)
408
LH:8.9 (22.7)
RH:8.7 (22.2)
LH: 8.1 (20.5)
RH: 7.9 (20.0)
Boyle et al.
–
All
Business Students
216 (1
-
5)
210 (2-5)
8.3 (21.2)
7.6 (19.2)
6.6
(16.6)
6.1 (15.5)
Boyle et al.
–
Business
Students (Caucasian
sub-group)
72
8.8 (22.3)
7.7 (19.6)
6.6 (16.8)
6.1 (15.5)
Garrett 1968 (
US
air
force pilots)
27
8.5 (21.6)
-
6.4 (16.3)
-
Matzdorff 1968 (German
adults)
769
8.5 (21.7)
7.7 (19.5)
-
-
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 42 of 79
42
These results support Wagner’s earlier finding that pianists tend to have larger hand
spans than non-pianists. Thus, note that the first four rows relating to surveys of pianists
show higher arithmetic means for 1-5 spans for both males and females than the last
four rows which are surveys of non-pianists. For the orchestral musicians (nearly all
Caucasian) in the fifth row of Table 30, male left hand 1-5 spans are similar to the
pianist groups and their right hand spans are 0.2 inches less while for females, their
right hand 1-5 spans are similar to the pianist groups but left hand spans are 0.2 inches
greater.
The question is, would one expect pianists as a group to have different hand spans to
the ‘general’ adult population or to specific sub-groups of non-pianists, with everything
else (such as ethnicity and socio-economic background) being equal?
One factor that might explain some of the difference between the Boyle, Steinbuhler and
Driscoll & Ackermann results compared with earlier studies is the documented increase
in the size of humans over time due to improved childhood nutrition.
9
This growth in
stature in wealthy countries is now believed to have levelled off. However, an increase in
mean hand span does not negate the significant variability in hand spans (in terms of
range and standard deviation). It will be argued in section 8 that a significant proportion
of today’s pianists have hands which are ‘too small’ for the current standard keyboard.
A further factor that could partially explain the significantly lower spans of Matzdorff’s
workers compared with Wagner’s pianists is the age mix of the sample groups; it is likely
that Matzdorff included a larger proportion of middle-aged adults compared with
Wagner. Anecdotal evidence suggests that hand spans contract somewhat with age.
Several authors, including Wagner (1984), Ortmann (1929) and Kloeppel (2000) have
found no evidence, nor scientific basis to support the contention that hand spans
increase significantly over time due to playing a musical instrument and/or stretching
exercises. Although Driscoll & Ackermann (2012) suggested that the significantly greater
left hand 1-5 spans among lower string players would most likely result from the
stretching of ligaments, muscles and tendons from many years of playing those
instruments, the magnitude of the average difference (0.2 inches) is still relatively small.
The lack of evidence of any significant left-right span differences among cello players in
the authors’ study could be attributed to the cello players not being professional or even
regular players, and/or their piano playing having resulted in a small increase in the
spans of both hands to the maximum extent possible.
It is certainly plausible that those with ‘small hands’ are more likely to give up playing
the piano at a relatively early age, while those with favourable hand spans find piano
9
www.scientificamerican.com/article/why-are-we-getting-taller/
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 43 of 79
43
playing comfortable and enjoyable, are more likely to achieve some success and are
therefore more encouraged to continue. This pattern of self-selection is evident in many
sports. The results in Table 30 suggest that self-selection does play a role in at least
partly explaining the smaller spans of non-musicians.
A further issue relevant to Caucasian males is the anecdotal evidence that some with
particularly thick fingers cannot play comfortably between the black keys on the
conventional keyboard. While this problem may well discourage those men from
continuing to play, it is not believed to be an issue that affects a significant proportion of
males
10
but further research on this issue is needed.
Finally, everything else being equal, it would be expected that females, with their
generally smaller hands, would be more likely than males to discontinue piano studies to
an advanced level. However, the fact that females predominate among adult pianists
may be explainable by other factors: perhaps they simply enjoy the piano more and/or
are more interested in pursuing a teaching career or perhaps are even more ardent than
men about a performance career. For males, the significant difference in 1-5 spans
among the business students (largely reflecting the much smaller spans among the
Asian students) compared with the adult male pianists suggests that the drop-out rate
for males with relatively small hands (even though they may have hands bigger than
most females) could be even higher than for females. Lack of career opportunities other
than teaching may be an important factor influencing their decisions.
Clearly, much of this assessment is speculative and further research comparing pianists
with non-pianists is necessary. However, in summing up these comparisons, the authors
believe that their data and the earlier findings of others indicate that adult pianists tend
to have bigger hand spans than the population at large. Certainly, the authors believe
there is no evidence suggesting that pianists have smaller hand spans than the
population at large.
Before applying all findings as presented so far in this paper, the question of ‘what is a
small hand?’ needs to be addressed.
7.0 WHAT IS A ‘SMALL HAND’ FOR A PIANIST?
The discussion to date has clearly shown that there is significant variation in adult
pianists’ hand spans. While the 473 pianists in the Adult Pianist data base had a mean 1-
5 span of 8.2 inches (21.0 cm), the smallest hand span in the group was just 6.4 inches
(16.3 cm) while the largest was 10.8 inches (27.4 cm). This difference between the
10
The authors conducted a brief ‘desk-top’ investigation by comparing anthropometrical summary data on
male finger widths with the widths of the gaps between the black keys on the conventional keyboard.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 44 of 79
44
smallest and largest of 4.4 inches (11.2 cm) is close to the width of five white keys on
the standard piano keyboard.
Clearly those with small hands will have different piano playing experiences to those with
large hands and to those who are in the middle range. Many pianists who have hand
spans in the middle range consider their hands too small for some advanced repertoire.
In order to reach a conclusion about what hand span is ‘too small’ in relation to the
current piano keyboard from an ergonomic perspective, the next section (7.1) presents
evidence from different fields to help shed light on the relationship between a pianist’s
hand span and their choice and performance of classical piano repertoire. In the section
following (7.2), an attempt is made to quantify the concepts ‘small hand’ and ‘large
hand’. Using these definitions and the data presented earlier in this paper, the
proportions of pianists affected are estimated in section 8.
7.1 The impact of ‘small hands’ on a pianist’s health, success and enjoyment – the
evidence
While not all classical pianists aspire to be performing artists, it is assumed that pianists
wish to gain maximum possible enjoyment from playing and maximum possible return
from their hours of study and practice. This would mean having access to a wide choice
of repertoire, avoiding pain and injury, and not having constant technical challenges
preventing full musical expression despite long hours of practice and good technique.
However every pianist today – whether they have small, large or middle-sized hands – is
effectively limited to the choice of just one sized instrument. Clearly some will cope
better than others: some find their hands are the right size to sit comfortably over the
notes they need to play while others are severely handicapped due to a mismatch
between the piano keyboard and their hand size for much of the repertoire they attempt.
The most obvious example to an observer is when a pianist finds it physically impossible
to reach the notes printed on the score, and alternative approaches such as ‘rolling’
chords are impracticable or musically inappropriate. But the inability to play notes as
written represents the ‘tip of the iceberg’; many of the technical challenges relating to
hand span issues are not at all obvious to observers, and are often not appreciated by
those with larger hands – who often have quite large hands by the time they are
attempting challenging repertoire.
The fact that the modern piano keyboard was standardised over a century ago to a size
that is ‘large’ in a historical context, based on the needs of certain European male
virtuosos (Parakilas et.al., 1999; summarised by Booker & Boyle, 2011) suggests that,
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 45 of 79
45
purely because males tend to have significantly larger hands, the current keyboard is
likely to be more suited to males than females. However, in Section 8, we also show that
there is a significant proportion of males for whom the keyboard is too large, not to
mention young pianists.
‘The one-size-fits-all approach has prevailed in the piano-keyboard world for nearly 100
years. And, like other one-size-fits-all systems, the largest was fitted, not the
average…Manufacturers were not about to make an instrument that would cause some
European Caucasian male who sat before it to say. “These keys are too small.” What
developed was a standard keyboard too small for nobody, but too large for many.’
(Christopher Donison, 2000, p.112).
What do the significant gender, ethnic and age differences in hand spans outlined in this
paper mean in terms of playing classical piano repertoire? The following discussion
supports the conclusion that the current keyboard is heavily skewed towards those with
larger hands (mostly males) and that the rest, the great majority, are disadvantaged as
a result. Further evidence presented also helps us to define a ‘small hand’ and a ‘large
hand’ based on the hand span needed to perform a wide range of repertoire with comfort
on the current keyboard. The common assumption that small-handed pianists somehow
‘compensate’ by adopting special techniques or tricks is not based on any scientific
evidence; instead the evidence is that those with small hands are more likely to play at a
lower standard and/or suffer more from pain and injury.
7.1.1 Underlying biomechanics and ergonomics
Since the pioneering work of Otto Ortmann early last century (1929), there have been
relatively few investigations of piano playing from the perspective of ergonomics.
Wristen (2000) has noted that small-handed pianists are at a higher risk of injury due to
greater degrees of lateral wrist motion, flexion, extension and deviation than for larger-
handed players. Large chords, octaves and arpeggios repeatedly force small hands and
fingers out of an ergonomically neutral, relaxed position.
More recently, Farias et al. (2002) investigated hand shapes among pianists. They
categorised hands in two classes: type A (wide hand) and type B (narrow hand). They
concluded that a pianist with a type B hand is at a disadvantage due to the higher
degree of wrist flexion needed to play large intervals than those with type A hands.
Women were found to be more likely to have Type B hands.
Muscles and joints operating near the middle of their range rather than close to their
limits will have a greater range of dynamic response. This means that those with large
hands are more likely to be operating within the optimum range, while those with small
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 46 of 79
46
hands are more likely to be operating at the extremes and therefore less efficiently due
to increased muscular tension. 'Fine dynamic gradation with the fingers in extreme
stretches is physiologically impossible' (Ortmann, 1929, p. 320).
Further, small hands are at a biomechanical disadvantage due to the need for increased
hand movement (compared with larger hands) in getting to the keys. Ortmann notes the
difficulty of controlling tonal intensity when there are rapid arm shifts, and that abrupt
and angular movements are associated with accents. This increased hand movement
also means loss of power and speed, as well as longer practice times just for the sake of
accuracy (Boyle, 2013).
According to Ortmann, technical mastery can be linked to the accuracy with which a
musical excerpt can be repeated, and this depends on a minimum of movement and
angular displacement. He concluded that hand size affects entire technique, going far
beyond whether or not certain notes can be physically played.
'The three factors of hand width, finger length and finger abduction ...., will explain a
surprisingly large number of technical difficulties that are often wrongly attributed to
defects of coordination or studentship'. (Ortmann, 1929, p. 313).
The impact of muscular fatigue on the central nervous system has been discussed by
Wagner (1984, p.169). He states: ‘Perhaps the most important effect of biomechanical
disadvantage lies in the increased load on the central nervous system.’ He says that
many musicians, despite the best training and preparation, never develop reliable
technique; they suspect it is biomechanically based, and subconsciously realise that they
have reached their limits in compensating for peripheral weaknesses. A pianist
attempting extended passages where extreme stretches are required is likely to
experience a progressive loss of power and control, with the additional mental effort
required just to keep playing limiting their ability to focus on musicality.
These issues have been discussed in more detail by Boyle (2013).
‘The ability to control the sounds at the piano, and this means producing lovely tone as
well as finely shaped phrases with a wide range of dynamics, depends to a large extent
on the ease with which we can play.’ (Cooke, 1985, p. 16).
7.1.2 Pain and Injury
Performance-related piano and injury is a significant issue affecting many pianists.
Several studies have found that female pianists run a significantly greater risk
(approximately 50% higher) of pain and injury than males. One example, Farias et al
(2002) found that, from a group of 222 pianists, 78% of females and 47% of males had
suffered from RSI.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 47 of 79
47
Given the gender difference of approximately one inch in the 1-5 span, this finding is not
surprising given that peer-reviewed epidemiological and clinical research on pianists
using the conventional keyboard has demonstrated that small hand spans are a
significant risk factor for pain and injury. Children and teenagers have also been found to
suffer from pain and injury related to piano playing (Ranelli et al., 2011).
More recently, with the availability of ESPKs (keyboards of different sizes), researchers
in the US have been able to experiment with the same pianists trying different sized
keyboards (for example, Wristen et al., 2006; Yoshimura & Chesky, 2009). These
experiments have found a reduction in pain and tension when the pianists move to
keyboards with narrower keys (in effect, reflecting the effect of getting ‘larger hands’)
and hence, indirectly supporting the demonstrated statistical association between pain or
injury and hand size. Keyboards used in these studies have included those with keys
approximately 7/8 the conventional width (i.e. with a 5.5 inch (14.1 cm) octave) and
those with keys approximately 15/16 the conventional width (i.e. with a 6.0 inch (15.2
cm) octave), in addition to the conventional keyboard which has a 6.5 inch (16.5 cm)
octave).
11
Yoshimura & Chesky (2009) documented levels of pain and tension reported by 35 piano
performance students associated with performing selected exercises (octave and chordal
scales) and an excerpt from Debussy’s L’Isle Joyeuse. The students performed the same
tasks on the conventional (6.5 inch octave) keyboard as well as on a 6.0 inch octave
keyboard. Students with hand spans above 22 cm (8.6 inches) experienced very little or
no pain on either keyboard. Those with the smallest hand spans reported significantly
higher levels of pain than those with larger spans. When the students (with spans less
than 8.6 inches) moved from the conventional to the smaller keyboard, the level of pain
was significantly reduced.
These results are consistent with the conclusions from clinical evidence reported by
Sakai (2002) that pain and tension is often related to the playing of octaves and large
chords, with small-handed pianists being more at risk.
This literature has been reviewed in more detail elsewhere (Boyle & Boyle, 2009; Boyle,
2013, and on www.smallpianokeyboards.org/pain-and-injury.html).
7.1.3 Performance quality
Very few studies have directly investigated aspects of the quality of a piano performance
and considered how this might relate to hand span. Ortmann (1929) referred to this
11
Refer: www.steinbuhler.com/html/our_research.html for definition of keyboard sizes.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 48 of 79
48
issue while Wagner (1984, p.169) discussed the impact of muscular fatigue on the
central nervous system.
One study (Lee, 1990) did not find hand ergonomic variables to be correlated with
performance quality measures (dynamic and temporal evenness), however, the excerpts
chosen (a simple five finger exercise and an arpeggio) did not involve wide stretches or
leaps, thus this lack of association in relation to finger spans is perhaps not surprising.
As mentioned below (section 7.1.5), female pianists appear to be far less affected, if at
all, when performing Baroque or early Classical repertoire such as Bach and Mozart, as
the musical figures tend to fall naturally ‘under the hand’ even for those with small
hands.
While not specifically about hand size, a more recent study of pianists by Goebl & Palmer
(2013) that used a three-dimensional motion capture system provides firm evidence for
the link between the efficiency of wrist/hand/finger movements and performance quality
(measured in terms of accuracy, precision of timing and control of tonal intensities).
The far greater movement of hands and wrists, and greater abduction of wrists and
fingers of small-handed pianists can readily be observed and has been noted by various
writers and mentioned in 7.1.1. For them, leaping and slanting strokes are more
frequently required to reach notes than for large-handed pianists. This reduction in hand
movement is obvious to a pianist transferring to a keyboard more compatible with their
hand size. The study by Goebl & Palmer provides an example of how 21
st
century
technology can be used to validate Ortmann’s conclusions (see 7.1.1).
As there are many factors that contribute to a quality performance, isolating the hand
span factor has been complicated by having just the one sized keyboard. This means
that in order to draw conclusions about the impact of hand span on performance quality
a statistical approach involving many pianists has been necessary. The availability of
ESPKs in recent years greatly increases the scope for research as the same pianist can
be tested on different sized keyboards to identify any change in performance quality.
7.1.4 The experiences of pianists using ESPKs
Although only a relatively small number of pianists around the world have experienced
ESPKs, their reactions are remarkably consistent. It is important to remember that a
pianist moving to a keyboard with narrower keys is, in effect, experiencing piano playing
as if they had ‘larger hands’ on the conventional keyboard. This means that a female
with an ‘average’ size span for their gender who moves from the 6.5 inch ‘standard’
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 49 of 79
49
keyboard to a 5.5 inch keyboard
12
suddenly discovers what it means to have an ‘average
male’ hand span on the conventional keyboard.
Pianists with small hands overwhelmingly report how much easier it is to play on a
keyboard with narrower keys and, almost without exception, they adapt to the smaller
size within hours. A key factor determining perceived ease and comfort is the extent to
which a particular figure falls comfortably ‘under the hand’.
A survey of pianists with spans of 8 inches (20.3 cm) or less who play or have played
ESPKs was reported by Boyle (2012). All pianists reported significant advantages after
transferring to a smaller keyboard (most played the DS5.5
TM
) – not only did problems of
pain and injury greatly diminish or disappear, but they found many advantages in
musical and technical outcomes, including increased power and speed when needed,
improved tonal and rhythmic control, greatly increased comfort with passages of octaves
and large chords, improved legato lines, greater speed of learning, as well as increased
accuracy and security. Unsurprisingly, they suddenly had access to an expanded
repertoire and were no longer forced to make repertoire choices based mainly on their
hand spans. Importantly, these pianists reported that they had a much greater ability to
focus on the music itself, rather than being mentally pre-occupied in overcoming
technical obstacles. These benefits are consistent with the benefits that would be
expected, based on the discussion in sections above (sections 7.1.1, 7.1.2 and 7.1.3).
Many of these experiences are described by Dr Carol Leone (2003, 2015), Chair of
Keyboard Studies at SMU Meadows School of the Arts, in Dallas, Texas. Leone has been
teaching and performing on DS5.5
TM
and DS6.0
TM
keyboards for many years. In her most
recent article (2015), she says:
'I often witness pianists place their hands for the first time on a keyboard that better
suits their hand span. How often the pianist spontaneously bursts into tears. A lifetime of
struggling with a seemingly insurmountable problem vanishes in the moment they
realise, "It's not me that is the problem; it is the instrument!" Following on that, the joy
of possibility overwhelms them.'
A (male) pianist who recently acquired a DS5.5
TM
keyboard sums up the feelings of
many:-
‘I have been playing on the piano and I am starting to grow quite accustomed to the
keyboard. Inevitably, of course, my first few hours of practicing has been wrought with
awkward mistakes, but also with astonishing revelations. It goes without saying that
12
All ESPKs used by pianists surveyed were made by Steinbuhler & Co. The 5.5 inch octave keyboard
(approximately 7/8 normal width) is labelled DS5.5
TM
.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 50 of 79
50
chords which were merely played in wishful daydreaming are quite comfortably reached.
That is not even the tip of the iceberg, however, as I saw, as I progressed in my
practicing, that the whole gamut of piano playing is infinitely easier with the right
keyboard size. Chord leaps, trills, scales, arpeggios, ascending and descending octaves
could be played with virtually no effort in comparison to what I had to labor through
hitherto. Consider, how I had neglected an Etude of Chopin's, after seeing that no matter
how much I practiced this one passage, it can only sound at best clumsy and
amateurish. Indeed, when I played it on the smaller keyboard, I thought to myself,
"Wow, I sound like Yefim Bronfman!"’
One of the authors, who has a 1-5 span of only 7 inches (17.8 cm), has been playing a
DS5.5
TM
keyboard for several years. On the conventional keyboard, she can only just play
octaves at the extreme front edge of the white keys which requires an uncomfortable
stretch and makes any rapid octave playing impossible. The DS5.5
TM
keyboard effectively
enlarges her active 1-5 hand span to 8.2 inches (20.3 cm), slightly above that of an
‘average’ female using the conventional keyboard. Although this keyboard opens up a
significant amount of repertoire that she is unable to play on the conventional keyboard,
she notices that fast octave playing is still associated with some unavoidable tension
which builds up in extended passages. She recently had the experience of playing an
even smaller keyboard (5.1 inch octave, DS5.1
TM
), which effectively enlarges her hand
span to about 8.9 inches (22.6 cm) – an ‘average male’ span. She found that any
tension and discomfort associated with such passages on this keyboard to be totally
absent, in addition to the many other benefits reported by other pianists who move to a
smaller (6.0 or 5.5 inch octave) keyboard.
There is anecdotal evidence that even pianists with 1-5 spans significantly above 8
inches also find significant benefits from playing ESPKs, including reduced tension and
the ability to play 10ths which requires a 1-5 span of at least 8.5 inches on the
conventional keyboard. In particular, from trials by many pianists playing keyboards of
different sizes since the 1990s, David Steinbuhler (ESPK manufacturer) has defined
keyboard ‘zones’ for different hand spans as summarised on the Steinbuhler website
13
.
This indicates that those with 1-5 spans below about 8.2 inches are likely to prefer a
DS5.5
TM
keyboard, while those with spans between about 8.2 and 8.8 inches are likely
prefer a DS6.0
TM
keyboard.
13
Refer to dot chart: www.steinbuhler.com/html/handsizepage.html.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 51 of 79
51
7.1.5 Piano competition results and discography
While females usually outnumber males studying piano performance in tertiary
institutions, it is males who dominate the lists of prize winners in major international
piano competitions and also the ranks of solo professional performers. A notable
exception relates to piano competitions specifically focused on Mozart or Bach, where
females dominate among prize-winners. In the nine major international competition
results reviewed (excluding Bach and Mozart competitions), 97 males compared with 19
females have won first prize. The question of ethnic background of female prize-winners
is also interesting. Of these 19 females, only three were of Asian ethnicity.
14
The same pattern of gender difference is not so apparent when one looks at violin
competition results, or indeed, from observing the gender mix today among string
players in world class symphony orchestras and chamber groups. The violin is an
instrument for which different sizes and variations in design are available and children
normally have the benefit of starting on smaller instruments.
For the piano, major performing and recording artists over the last 100 years have
overwhelmingly been male. Many of the best known female artists have been known
mainly for excellence in Baroque and early Classical repertoire. The rarity of award-
winning recordings by female pianists of complete sets of Romantic or later works (such
as the Chopin or Liszt Etudes, Prokofiev sonatas, or even the Beethoven sonatas) may
well reflect their need to be selective in their choice of repertoire.
In many careers, people reach their peak in terms of status and income in their 40s or
50s after two to three decades of commitment and women may be relatively
disadvantaged as a result of taking time out for child-rearing responsibilities. Clearly,
this would not be a significant factor affecting women succeeding in piano competitions,
where contestants are in their late teens, 20s or early 30s at most. Sustaining a long
term solo performing career, however, may be interrupted or affected to some extent by
child-rearing. But a further consideration is that pianists with hands that are not quite
large enough to play advanced Romantic and later repertoire without some tension being
present may find that they are unable to sustain a performing career over decades. The
known risk factors for pain and injury among pianists include not only hand size but also
intensity of practice.
7.2 Defining ‘small hands’
The previous section describes how ‘small hands’ have an impact on a pianist’s health,
success, performance quality and enjoyment, while the earlier sections reveal that there
14
www.smallpianokeyboards.org/piano-competitions.html
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 52 of 79
52
is significant variability in human (and pianist) hand size. We now return to the question
of the definition of a ‘small hand’ in relation to piano playing.
Many previous writers who have referred to ‘small hands’ have only given vague
indications of what is ‘large’, ‘average’ or ‘small’, and have rarely considered gender or
ethnic differences. Thus, the term is generally used loosely. Male pianists complaining
about their ‘small hands’ may well have hands larger than those of the vast majority of
females, but not as big as their acclaimed male counterparts; ‘small hands for a male’
would be a more appropriate description. Alicia de Larrocha is often assumed to have
had ‘small hands’ because she was very short (less than 5 feet tall), but she said herself
that she could reach a 10
th
in her heyday.
15
The authors’ regression and correlation
analysis (section 4.5) of height versus 1-5 span indicates that a person’s height may not
always be a good guide to hand span. Also, hands that may appear small from casual
observation (perhaps being thin and bony) may have a surprisingly large span.
In general, the term ‘small hands’ conveys the notion of relative disadvantage for
classical piano playing, especially at advanced and elite levels. The findings of the
authors’ study indicate that females are most affected, as only a very small percentage
of male pianists have hand spans less than the average female pianist. However, some
males do have ‘small hands’ based on the authors’ definition given below. And the
authors found that those of Asian ethnicity tend to have hands somewhat smaller than
those from Caucasian backgrounds.
What is perceived to be a ‘small hand’ for one person might seem to be a ‘large hand’ for
someone else. While the term ‘small hand’ might be best applied in the context of what a
pianist aspires to do, as discussed earlier it is assumed that any pianist would like to
have access to a wide range of repertoire without hand span limitations dictating
selection, to be able to overcome technical challenges and play to the best of their
ability, to minimise the risk of pain and injury and to gain maximum possible enjoyment
from playing.
7.2.1 Definitions by others
Kamolsiri (2002) defines small hands as those unable to reach a ninth and/or unable to
play an octave comfortably. Wristen et al. (2006) defines small-handed pianists as those
having an active 1-5 span of 8 inches (20 cm) or less. Farias et al. (2002) defines small
hands as those that cannot reach a tenth. The survey of pianists using ESPKs (Boyle,
2012) only included pianists with active 1-5 spans of 8 inches and below, all of whom
preferred the smaller keyboard over the conventional. And as discussed above, there is
anecdotal evidence that pianists with spans of more than 8 inches also prefer playing
15
www.nytimes.com/1995/11/23/garden/at-home-with-alicia-de-larrocha-a-pianissimo-star.html
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 53 of 79
53
ESPKs. In the future, assuming ESPKs were to become widespread, a more accurate
definition could be based on the hand size characteristics (including finger widths) of
those who prefer to play on the narrower keys of an ESPK rather than on the
conventional keyboard.
7.2.2 Comfortable octave playing and the ability to play larger intervals
What is meant by ‘comfortable playing’ in the case of octaves?
Octave playing is a fundamental requirement of the classical piano repertoire. Although a
pianist with a 1-5 span even below 7 inches can just play an octave on the front edge of
the keyboard, there is strong evidence (based on the discussion in 7.1.2 and 7.1.4) that
tension is likely to be present for 1-5 spans below about 8.5 or 8.6 inches.
The authors reported previously on a calibration of 1-5 hand spans and maximum
intervals playable, based on white key intervals (Boyle & Boyle, 2009). On this basis,
they defined the following approximate threshold 1-5 spans whereby a pianist can just
play (with considerable discomfort and not at speed) on the extreme outer edge of the
while keys:
• Octave – 6.7 inches
• Ninth – 7.6 inches
• Tenth – 8.5 inches
Considering octave playing in particular, a span of about 7.6 inches is required to be able
to move in from the outer edge of the white keys, but this generally does not allow the
elimination of tension in rapid playing of octaves and octave-based chords. Tension is
often associated with wrists being raised to avoid clipping adjoining white notes and to
gain additional power. Based on the work of Yoshimura & Chesky (2009) discussed
above, plus the personal experience of the authors (refer Boyle and Booker, 2011;
Boyle, 2013) and anecdotal evidence from other pianists, it is believed that a span of
around 8.5 inches is needed to totally eliminate tension in fast octave playing while
bridging the hand comfortably over the black keys.
Other aspects of octave playing also need to be considered: the ability to play legato
octaves (using 3
rd
and 4
th
fingers), ease with ‘cross-over’ octaves (right hand crossing
over to the bass, left hand to treble), and ability to minimise forward and backward hand
movement in fast chromatic octaves (such as in dotted rhythms found in polonaises).
Pianists with spans of less than about 8.5 inches are likely to experience greater
difficulty with these tasks.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 54 of 79
54
With a hand span of 8.5 inches, a pianist can normally just play a tenth. A ninth
becomes more comfortable, with the ability to move in from the outer edge of the
keyboard. As piano repertoire does not normally contain extended fast passages
involving ninths or tenths, the ability to eliminate tension when playing these larger
intervals is not nearly as critical as it is for octaves.
Based on the above, one of the authors (Boyle, 2013) has previously supported the
definition of a ‘small hand’ to be consistent with that of Farias et al. (2002), i.e. 1-5
spans of less than about 8.5 inches (21.6 cm). This is based on the presumption that
most classical pianists aspire to play a wide range of repertoire from Beethoven and
Schubert onward. The analysis of the hand span statistics described in this paper lends
further weight to this definition, in particular, on the basis of the internationally
acclaimed solo performers in the sample (who perform across a wide range of repertoire)
all having spans considerably larger than 8.5 inches.
7.2.3 Second to fifth finger spans
Prior to this study, spans between the second and fifth fingers in relation to piano
playing have not previously been considered by the authors. Clearly, this span (in
addition to stretches between all other digits) is important for the playing of chords,
arpeggios, broken chords and arpeggiated figures in general, and also for the ability to
maintain an unbroken legato line.
For the purposes of this paper, a quick calibration was done involving several pianists
with varying 2-5 spans. This suggests that a 2-5 span of around 6 inches (15.2 cm) is
required to play relatively comfortable major (white key) sixths and just play a seventh.
A span of at least 7 inches (17.8 cm) is required to be able to play a seventh
comfortably. (As noted previously for 1-5 spans, a 7 inch span can just manage an
octave on the edge of the keyboard.) Further more detailed work is needed in relation to
2-5 and other finger spans in the context of piano playing.
Advanced repertoire frequently includes chords (or broken chords) where a stretch of a
seventh is required. The authors therefore consider that a 2-5 span of less than 6 inches
should also be defined as a ‘small hand’ in relation to the current keyboard.
All pianists rated in the authors’ study as international performers had 2-5 spans above
6.5
16
inches, with most being close to 7 inches or higher.
16
Apart from the right hand 1-5 span of one pianist which was 6.2 inches.
12
th
Australasian Piano Pedagogy Conference Proceedings R. Boyle, R. Boyle & E. Booker
Page 55 of 79
55
7.2.4 Threshold spans defining ‘small’ versus ‘large’ hands
The hand span zone boundaries defined later reflect the ability to just play certain
intervals on the current standard keyboard. But of course piano playing is not only about
performing fast passages of octaves and large chords and being able to play ninths and
tenths. Many of the other difficulties faced by pianists with small hand spans, including
the higher risk of pain and injury, have been touched on in section 7 and in and other
papers (Boyle, 2012; Boyle 2013). Taking all of the above into account, the authors
believe that threshold values of 8.5 inches (21.6 cm) for the thumb to fifth finger span
and 6.0 inches (15.2 cm) for the second to fifth finger span are reasonable.
8.0 HOW MANY PIANISTS HAVE ‘SMALL HANDS’?
In section 7.2, we have defined ‘small hands’ in relation to playing the standard piano
keyboard as being:
• those with an active thumb to fifth finger (1-5) span of less than 8.5 inches, or
• those with a second to fifth finger span (2-5) of less than 6.0 inches.
In order to estimate proportions of pianists with spans in these ranges, further analysis
of the Adult Pianist database was undertaken. Given the closeness of these results to
those of Wagner and Steinbuhler, the authors have used thei