BRONWYN SHERRIFF
WITS University, Department of Psychology
UNISA, Institute for Social and Health Sciences (ISHS)
SUPERVISOR: MR MICHAEL PITMAN
OVERVIEW
INTRODUCTION
TAXONOMY OF MUSICAL IMAGERY
AIMS & RATIONALE
METHODS
RESULTS
DISCUSSION
STRENGTHS & LIMITATIONS
FUTURE RECCOMENDATIONS
CONCLUSION
INTRODUCTION
 Psychological research concerning imagery is fairly heterogeneous in
nature owing to the seemingly surreptitious and internal character
 Methodological issues as well as paradigmatic debates  impinged on
the capacity of scholars to expound upon the various mechanisms and
nuances underpinning imagery
 Galton (1880) was one of the first individuals who attempted to
quantitatively investigate imagery meticulously and even considered
musical imagery in his self-report instrument although he primarily
focused on the visual imagery domain.
 Research related to auditory imagery has gained popularity in the last
two decades owing to the numerous technological advances in neuroimaging equipment that are currently accessible, compatible and
utilisable (Liikkanen, 2008; Kosslyn, Ganis & Thompson, 2006).
TAXONOMY OF MUSICAL IMAGERY
Auditory Imagery
Musical Imagery (MI) [hearing music in the mind’s ear ]
MI is both a subjective and conscious occurrence of internally generated or imagined musical, tuneful or
melodious sounds (with or without lyrical content) whether prompted or spontaneous, involving full
compositions or incomplete fragments, and is primarily in the absence of a directly corresponding sound
stimulus from the immediate or peripheral environment.
Voluntary MI
Individuals can consciously
and deliberately imagine a
musical composition or
musical sound
Involuntary MI
Spontaneously generated
MI that moves into the
imager’s conscious
awareness
MI occurring in
dream-states
Recent or repetitive,
overexposure to a
particular musical
composition
AIMS & RATIONALE
 MI is an understudied phenomenon in scholarly literature even though it
could provide valuable information concerning consciousness and elucidate
various aspects of memory and auditory processing.
 Existing MI research displays particular trends  neurological,
experimental and quantitative studies concerning MI appear to be the
dominant techniques.
 In the last five years (2005 – 2010), MI studies appear to have
demonstrated a methodological shift, and seem to have initiated a transition
in favour of non-experimental techniques and self-report approaches.
 Accordingly, the present study endeavoured to explore the four
deductively derived characteristics of MI and , include the following: (i) the
constancy, (ii) the nature, (iii) the content and, (iv) the corporeal
manifestations (Brown, 2006) associated with everyday MI episodes as they
occur in naturalistic environments.
 It also aimed to expand the existing literature on the topic in view of the
fact that there appears to be a dearth of consolidated information
concerning the qualitative properties and characteristics of MI.
This study aims to address the following gaps in the existing literature base:
i.
ii.
iii.
iv.
v.
vi.
To utilise mixed methods as a technique to elicit detailed and
exploratory data pertaining to MI which has not been previously applied
 the experience-sampling technique may conceivably compliment the
quantitative results described by other scholarly studies
To present findings for both musicians and non-musicians experiences
of everyday MI occurrences
To directly examine MI across individual’s daily lives, in order to
strengthen existing findings within the non-experimental paradigm of
MI
To add an element of diversity in MI research by examining a nonwestern/non-European sample of participants
To specifically present findings concerning both voluntary and
involuntary variants of MI in addition to discerning between the various
temporal subsets of experienced-sampled data which has not been
previously reported on in studies utilising longitudinal methodology
To investigate co-occurring phenomena such as motor responses
associated with MI and dual imagery modalities with the intention of
providing a thick description of the phenomenon.
METHOD
Sample and Sampling:
There were essentially two stages of exploratory observation used in this study; both
were non-experimental in nature (Patton, 2002) and utilised convenience sampling.
Sample Group
Gender
Frequency (percent)
Musician
Frequency (percent)
Female
58
(67.44%)
9
(56.25%)
6
(75%)
Male
28
(32.56%)
7
(43.75%)
2
(25%)
Yes
36
(41.38%)
8
(50%)
5
(62.5%)
No
51
(58.62%)
8
(50%)
3
(37.5%)
Sample
Group
Age Range
Mean
Standard
Deviation
Stage One
19 – 34 years
21.21 years
2.64 years
Stage Two
19 – 27 years
21.56 years
2.45 years
Pilot
19 – 34 years
23.75 years
4.46 years
Mode
(Frequency)
20
(29)
20
(4)
22
(3)
Stage One
Stage Two
Pilot
Sample size
(N)
87
16
8
Sample size
(N)
87
16
8
Procedure
 Ethical clearance (Ethics Protocol Number MPSYC/09/001 IH); permission from
appropriate WITS staff was granted
 Short presentations to invite undergraduate Psychology students to participate
 Distribution and collection of demographics questionnaire
 Piloting of SMIJ over two days (3 questionnaires per day)
 Distribution and collection of SMIQs  four different sampling periods in order to suit
participants’ needs over a five week interval.
 Analysis and write-up
Instruments:
I. Stage One (Survey)
II. Stage Two (Experience-sampling  SMIJ)
Ethical Considerations
 Informed consent and confidentiality (as well as anonymity)
 Participation in both stages of this study was voluntary  Participant Information
sheets for each stage of participation
 No obligation to complete any question(s) and withdraw without any consequence
arising
 All participants who submitted their demographic questionnaire and / or their SMIJ
were considered to have tacitly consented to participate in the study
 Participant debriefing  short summary of findings will be posted on an electronic
forum
RESULTS
91.95 % (80 out of 87) of the Stage One participants reported listening to
music every day
Frequency of MI (MI)
Once a month
Once a week
2 to 3 times a week
4 to 5 times a week
6 to 7 times a week
A few ( 1 to 3) times
everyday
Many (4 or more) times
everyday
Column Totals
Frequency
(Percent)
1
(1.15%)
3
(3.45%)
27
(31.03%)
10
(11.49%)
4
(4.60%)
28
(32.18%)
14
(16.09%)
87
Musician
Nonmusician
Female
Male
0
1
0
1
2
1
2
1
12
15
22
4
3
7
9
1
1
3
2
2
11
17
15
13
7
7
8
6
36
51
58
28
CONSTANCY AND NATURE OF MI
Nature of MI
Experienced-Sampled Imagery
(ESI)
Total number of reported
episodes
Percent
105
(of 226 MI reports)
46.46%
Involuntary ESI
89
84.76%
Delayed Experienced-Sampled
Imagery (DESI)
11
9.09%
Involuntary DESI
10
90.90%
Last Recalled Imagery (LRI)
66
54.54%
Musical Imagery On Waking
(MIOW)
44
42.72%
312
(of possible 336)
92.86%
Number of Completed SMIQs
• For 77.59% (90 out of 116) of the cases, participants reported having no
awareness of a personal association with their MI.
• The majority of MI incidences (93.97%) were familiar to the Stage Two
sample (109 out of 116 episodes).
• The volume of MI was predominantly reported to be soft in nature
(58.62%), and participants frequently remarked that one needed to
concentrate on it in order to hear it clearly.
• The co-occurrence of visual imagery along with MI tended to be a
relatively common phenomenon. The Stage Two sample reported
experiencing dual-imagery just below half the time (48.28%).
Characteristic
Melody
Lyrics
Rhythm
Mean
5.019
5.143
4.952
Standard
Deviation
1.676
2.054
1.740
Range
1-7
1-7
1-7
• These mean clarity ratings are all quite similar and they suggest that all
three elements were relatively clear
MI CONSTANCY
Individual MI Variations
4
3
2
1
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Stage Two Participants
Mean Daily MI Frequencies
 The figure above suggests that the experience of MI was fairly regular
over the twenty-one sampling periods. However, there are moderate
variations between individual participants
 Participants 1 and 15 had the most reported experiences of MI within a
seven day period both had 23 incidences
CONTENT OF MI
 Various genres of music were evident but most were familiar to the imager,
only a small number of MI episodes involved original compositions.
 Television theme songs and adverts were reasonably consistently
experienced across the Stage two sample
Content of MI
Stage Two sample
Vocal
(a)
Combination of vocal
& instrumental music
Instrumental
Repeated loop of a
musical fragment
(b) Full song
Part of a song
Not specified
Involuntary imagery
(c) Voluntary imagery
Not specified
Frequency
Total number of
observations
Percent
Frequency
58
32
105
105
55.24%
30.48%
58
32
15
59
105
105
14.29%
56.19%
15
59
19
26
1
89
15
1
105
105
105
105
105
105
18.10%
24.76%
0.95%
84.76%
14.29%
0.95%
19
26
1
89
15
1
Predictor
Variable
Chi-Squared
Test statistic
No. Of
observations
Musician
6.35
Absence of
external
music
6.67
p = .012
Absence of
others*
1.81
p = .009
Gender*
0.19
p = .179
p = .660
299
299
299
299
Unadjusted
Odds Ratios
1.85
.49
.72
1.11
p-value
.012
.012
.179
.660
95% CI
1.14 – 2.99
0.28 – 0.86
.44 – 1.16
.69 – 1.79
Pseudo R²
.0164
.0172
.0047
.0005
(85%)
(1.64%)
(50%)
(1.72%)
(28%)
(0.47%)
(11%)
(0.05%)
Pseudo R² is 0.0571 once all four predictors are included. This suggests that 5.71%
of the total variance in the occurrence of an MI episode (outcome variable) can be
explained by the predictor variables if they are simultaneously present.
TYPES OF INVOLUNTARY MI
In response to a question in Section A which asked “Did the music just ‘pop into your head’, or
would you say you thought about or chose the music? Please explain”.
(a) The MI
‘popped’ into
conscious
awareness:
“Just popped in
my head. Don’t
even know the
song. No I
couldn’t have
chosen it and its
[sic] not the kind
of music I listen
to.”
Participant 16
(b) Recent
exposure to an
externally heard
source of music
caused the
participant to
continue to
generate the MI
ex post facto
despite no
apparent wish to
do so:
“After I heard this
song in my car, I
just kept hearing
it in my head.”
Participant 5
(c) A particular
word or textual
cue triggered a
distinct case of
MI:
“The song
popped into my
head as a result
of a friend’s
status update on
facebook, which
used a single
line from the
song.”
Participant D
(d) A specific
thought
catalysed an
associative
reaction in the
form of MI:
“Popped into my
head because of
an association
with someone
that makes my
blood boil.”
Participant 7
CORPOREAL MANIFESTATIONS (CM):
 While the overall incidence of CM in the Stage Two sample was low (18.1%)
 62.5% of the participants (10 out of 16) reported CM accompanying their MI
on at least once over the experienced sampling period.
 Of these ten participants, six were musicians (60%) and four were nonmusicians (40%).
 CM experiences were reported to be predominantly involuntary or
unconscious responses (78.95%).
“[The] Movements were induced by [the] rhythm and melody of the music.”
Participant 15
“When I’m not actively listening to the song in my head but doing a different task, it’s
involuntary BUT obviously when I’m dancing around, it’s voluntary because I’m consciously
engaging in it.”
Participant C
 Participants also consistently suggested that CMs appeared to express an
aspect of the content of their imagery whether it was auditory or visual  i.e. MI
or dual imagery
“Humming feels like expressing the song in my head”
Participant D
“I’m tapping my feet and bopping my head according to the tune and the beat.”
Participant H
DISCUSSION
•The Stage One findings imply that, in general, the experience of MI was a relatively
common phenomenon amongst the participants. The cumulative frequencies
suggest that at least once a week or more regularly, 98.85% participants reportedly
experienced MI (n = 87).
•The self-reported everyday incidence of involuntary musical imagery (INMI) in
Liikkanen’s (2008) sample was similar to these findings: 91.7% of the participants
experienced MI at least once a week; 33.2% reported experiencing INMI ‘Every day’;
and 26.1% of the sample reported experiencing the phenomenon ‘Several times a
day’ (n = 11 904).
•Daily exposure to music seemed to be a more significant factor than any formal
musical experience or training. Approximately 92% (n = 87) of the sample listened
to music on a daily basis, the majority of which (59.77%) did so between one and
three hours per day.
•Similarly, 98.1% of Liikkanen’s (2008) sample reported listening to music on a
weekly basis which suggests that regular exposure to external sources of music
might be an equally significant factor regarding the consistent experience of MI
episodes.
 Stage Two results suggest that in 68.75% (11 out of 16) of the voluntary MI
cases, the MI had been heard from an external source on the same day;
additionally, in approximately half (51.69%  46 out of 89) of the involuntary
MI cases, participants were also exposed to the original composition, from an
external source, on the same day.
 Moreover, 79.05% (83 out of 105) of the ESI reports were heard by
participants from an external source within the same week relative to the
occurrence of their experienced-sampled MI episode.
 The odds ratio which was calculated does suggest that in this study,
musicians were 85% more likely to experience an episode of MI than nonmusicians based on the ESI (Experienced-Sampled Imagery) data. This may be
owing to the broad definition which was utilised to differentiate the two groups.
 The present study’s results imply that recent exposure to a tune in addition
to the “stickiness” of the music might enhance its tendency to become an
episode of MI given that more than one third (35.24%) of ESI reports were
classified as “stuck in memory after recent exposure” with 91.89% of these
‘sticky tunes’ MI having been heard on the same day.
 Bailes (2006) reports that 43% of her respondents had heard their MI from
an external source during the time lapse between DES signalling periods.
However, only 3.74% of Bailes’ (2007; 2006) cases were classified by the
participants as “sticky” tunes.
 In the current study, both samples were generally familiar with their MI,
whether it was classified as an ESI, DESI, MIOW or LRI case, and these
percentages ranged from 74.25 to 96.55.
 Bailes (2007) found comparable patterns regarding the familiarity of MI
episodes – 83.18% of the tunes were named by participants.
 Liikkanen’s research (2008, p.410) also demonstrated analogous
findings – “familiar lyrical music dominated over instrumental or new music
(76% vs. 26%)”.
 The present study examined the simultaneous occurrence of visual
imagery and found that two primary characteristics emerged.
 Firstly, music videos pertaining to the MI were among one of the most
persistent trends.
 Secondly, imagery classified as ‘other’ became a recurrent theme. In
general, these involved descriptions relating to a movie or television
programme or alternately, a scene that was in some way cognitively
associated with the MI.
 However, individual variations were apparent and this may suggest that
dual imagery is specific to each individual rather than being the norm.
STRENGTHS & LIMITATIONS
 Study was constrained by ethical protocols (use of incentives or extrinsic
reward systems for participation were prohibited)  Considering that there is a
high cost of compliance in a study of this nature, the rates of acquiescence were
pleasing
 Lack of generalisability (sample)  taking into account the sample size, age
range of the participants as well as the fact that all the participants shared the
commonality of being Psychology students
 Reliability of the two instruments not adequately evaluated  not within the
scope of the study given that the primary focus did not relate to the
psychometric properties of the instruments
 Traditional DES methodology uses a “sampling-then-interviewing process”
which is cyclical and spans over a number of days  Follow-up interviews
would have been advantageous in terms of clarifying the essential nature of
certain ambiguous responses as well as being able to adjudicate the
truthfulness of certain journal entries
 Potential reactivity to the instrument This is another characteristic
predicament related to studies utilising multiple observations as well as
activities which require repetitive introspection
 One participant remarked that as they became accustomed to the DES
task, they became more aware of MI over time.

Absence of inter-rater reliability estimates  thematic content analysis
FUTURE RECCOMENDATIONS
The following set of potential relationships or observations, at minimum, might serve to
facilitate future research endeavours relating to MI:
i.
ii.
the relationship between MI and Lamotrigine and / or other medications;
the natures and degree of personal (emotional and / or cognitive) associations in relation
to involuntary MI;
iii. whether the volume of MI is dependent upon the activity engaged in during an episode;
iv. exploring the vividness and clarity of various dimensions of MI;
v. the relationship of dual imagery with specific emphasis on which kind is the precursor;
vi. the statistical relationship between recent musical exposure and involuntary MI;
vii. the possible statistical relationship between the onset of MI whilst external auditory stimuli is
absent as well as when an imager is unaccompanied;
viii. the role and statistical relationship of musical training in relation to MI;
ix. the potential function of specific performing arts professions in relation to regular MI episodes
and/or corporeal manifestations;
x. the characteristics of truly spontaneous MI;
xi. exploring statistical or theoretical models of predictive variables which might explain a
larger percentage of the total variance of MI;
xii. explanatory frameworks concerning the cyclic and repetitive nature of MI;
xiii. the role, nature and prevalence of corporeal manifestations in relation to MI;
xiv. either fully or partially replicating the current study to validate of the findings using
diverse samples in various cultures; and
xv. establishing the psychometric properties of the instruments as well as their crosscultural utility.
CONCLUSION
 This study has provided exploratory evidence pertaining to the
characteristics of MI as it is experienced during everyday waking states.
 It has provided rigorous empirical validations to support the premise
that although everyday MI is experienced as involuntarily and seemingly
spontaneous in nature, it is in fact frequently owing to recent priming and
exposure from various external mediums.
 Although recency and repetition are factors which have been alluded to
in previous studies, it is reasonably patent from the findings in this study
that these dimensions appear to underpin everyday MI incidents.
 The prevalence of MI is consistently experienced by various individuals
regardless of their musical background even though musicians may have a
higher inclination toward MI occurrences.
 Daily exposure to music appears to be an equally significant factor in
relation to the incidence of MI during everyday activities especially given
the finding that most episodes of MI are usually familiar to the imager.
ACKNOWLEDGMENTS
 All the participants for their time, commitment to the study,
insightfulness and honesty; without whom, this study would not have been
possible
 My supervisor (Mr. Michael Pitman) for his guidance, inspiration,
empathy and assistance throughout the research process
 WITS University for the financial assistance in terms of the post-graduate
merit award bursary which helped make my post-graduate studies possible
 Friends and family for their constant support, encouragement, patience
and wisdom
 Prof. Brendon Barnes for his advice, his patience, willingness to assist
and his caring demeanour
THANK YOU FOR YOUR ATTENTION
COMMENTS AND FEEDBACK ARE WELCOME