European Journal of Neurology 2006, 13: 1378–1380

doi:10.1111/j.1468-1331.2006.01470.x

SHORT COMMUNICATION

Abnormal musical pacemaker in a patient with musical hallucinations
I. Biran and I. Steiner
Department of Neurology, Hadassah University Medical Center, Jerusalem, Israel

Received 26 June 2005
Accepted 10 October 2005

Music creation requires a highly orchestrated temporal pattern. The study of a patient
with repetitive musical hallucinations enabled us to examine temporal pacemakers in
music production. Here, we show that the pattern of faster silent production of a
chosen tune compared with its production aloud was reversed when the patient produced the hallucinatory tune. This observation might suggest the utilization of a
pacemaker(s), which functions differentially during the disease.

Introduction
The creation and the estimation of time durations are
an intriguing and complex cognitive phenomenon. Do
we possess an internal clock or a pacer? Is it affected
during the disease and how? As music requires a
highly orchestrated temporal pattern of hundreds of
milliseconds, one may postulate that an internal time
keeper is an essential component of its generation.
Accordingly, the creation of music might serve as a
research vehicle to examine the subjective estimation
of time durations and the possible role of internal
pacemakers.
Musical hallucinations have a prevalence of 0.16–
2.5% [1], usually occur spontaneously and sometimes
are associated with structural brain disease, hearing
impairment, psychiatric disorders, epilepsy and intoxication [1–4]. Patients report of tunes repeatedly playing
within their head in the absence of corresponding
external auditory stimulus.
A patient with repetitive musical hallucinations
enabled us to study the internal estimation of time
durations related to a single event under normal
conditions and during disease.

Case report
A 74-year-old right-handed female without musical
training presented with 2 months duration of musical
hallucinations. She denied other auditory complaints
aside from hearing loss secondary to Menier’s disease.
Segments of European classical music heard by her
recently would repeatedly play in her head till it would
stop or be replaced by another tune. The length of the
repeating musical fragments varied, but usually lasted
<30 s. Her past medical history was notable for two
episodes of transient global amnesia (TGA) 23 and
Correspondence: Iftah Biran MD, Department of Neurology, Hadassah University Medical Center, P.O. Box 12000, Jerusalem, 91120,
Israel (tel.: 972 50 8573457; fax: 972 2 6242850; e-mail: i_biran@
hotmail.com.)

1378

12 years prior to the appearance of the musical hallucinations. A brain magnetic resonance imaging (MRI)
taken during the second episode demonstrated a right
temporal cyst (Fig. 1). Although she was an avid listener to classic European music, she could not read
music nor play any musical instrument.
Her general and neurological examinations were
normal and no cognitive abnormality was present. She
scored 30/30 on the Mini Mental Status Examination
and 96/100 on the Addenbrooke’s mental evaluation
[5]. Electroencephalogram was normal. Brain computed
tomography (CT) scan did not demonstrate any change
from several previous neuroimaging studies (MRI and
CT). Audiometry showed severe bilateral sensoryneural deficit. A therapeutic trial with phenobarbital
did not affect her auditory hallucinations.

Experimental Study
The patient and 11 healthy controls without any history
of neurological or hearing impairment and mean age of
44 ± 7 were examined.
Subjects were asked to produce tunes by singing or
humming, and we recorded the production time. The
tunes, which were either tunes of choice (Ôchosen tuneÕ)
or hallucinatory tunes, were produced in the following
three tasks.
Intentional Production of a Chosen Tune

We recorded the time required for both the patient and
the control subjects to produce a tune chosen by each of
them under two conditions: (i) tune produced intentionally silent (imagery production of the tune [6]) and
(ii) tune produced intentionally aloud. The subjects had
to choose a tune of 4–10 s, and the duration was verified prior to the start of the experiment. There were no
restrictions as to the nature or source of the tune. The
silent and aloud conditions were assigned to six randomly presented blocks with 10 repetitions each, and a
total of 30 recordings for each condition.

Ó 2006 EFNS

Clinical spectrum of musical hallucinations

1379

Figure 1 Magnetic resonance imaging
study of the patient following the last
TGA episode, 13 years prior to presentation, demonstrating a right temporal cyst.
(A, Axial FLAIR; B, Axial T2; C, Axial
T1 with contrast).

Intentional Production of a Hallucinatory Tune

We recorded the time required for the patient to experience or produce her current hallucinatory tune. In this
task, only the patient participated as only she experienced
hallucinatory tunes. Similar to the previous task, there
were two conditions in this task: (i) the hallucinatory
tune produced intentionally silent and (ii) the hallucinatory tune produced intentionally aloud. The conditions
were blocked similar to the recording in previous task.
Intentional and Spontaneous Production of a
Hallucinatory Tune

We recorded the time required for the patient to
experience spontaneously or produce intentionally her
current hallucinatory tune. Accordingly, there were
three conditions in this task: (i) the hallucinatory tune
produced intentionally silent, (ii) the hallucinatory tune
produced intentionally aloud and (iii) the hallucinatory
tune produced spontaneously (the time required for the
patient to experience the hallucinatory tune). The three
conditions were assigned to 18 counterbalanced blocks
with five recordings in each block and a total of 30
recordings for each condition.
Time durations required to complete a tune were
recorded using a G4 Mac computer and the PSYSCOPE
software by hitting the ÔMÕ key at the start and the ÔNÕ
key at the completion of each tune [7].

Table 1 Mean time required to complete a tune of choice aloud and
silent by the patient and the control subjects
Chosen tune (mean time in
ms ± SD)
Silent

Aloud

P

Aloud/silent

Controls 6382 ± 1724 7318 ± 1972 P < 0.0001 1.16 (95% CI:
1.02–1.23)
Patient
7741 ± 388 8384 ± 212 P < 0.0001 1.08

silently. The ratio for controls was 1.16 (95% CI )1.02–
1.23) and for the patient 1.08, within the CI of the
controls. Therefore, there was no difference between the
performance of the patient and the controls in this task.
These findings suggested that: (i) production of a
musical tune silently takes less time than its production
aloud; (ii) there is a fairly constant adherence to time in
repeated production of an identical tune.
Intentional production of a hallucinatory tune

Intentional production of a hallucinatory tune by the
patient was statistically significant slower with the
production of the hallucinatory tune silent compared
with its production aloud (8094 ± 1189 ms,
6850 ± 213 ms, t ¼ 5.63, P < 0.0001). The ratio
between producing the tune aloud to its production
silently was 0.85 and is outside the 95% CI for controls
in the first part of the experiment. Hence, the pattern of
faster silent production of a tune was inverted.

Results
Intentional production of a chosen tune (Table 1).

Intentional and spontaneous production of a
hallucinatory tune (Fig. 2)

In controls and the patient, intentional silent production of a chosen tune was statistically significant shorter
than the singing aloud of the same tune. Control subjects needed 6382 ± 1724 ms to complete a tune
silently and 7318 ± 1972 ms to produce it aloud
(P < 0.0001); the patient required 7741 ± 388 ms, and
8384 ± 212 ms, respectively (P < 0.0001).
In order to account for the differences in reaction
times amongst the subjects, we calculated for each the
ratio between producing the tune aloud to producing it

The time required for the patient to experience a hallucinatory tune spontaneously was 10,707 ± 566 ms
compared with 10,369 ± 1263 ms when it was produced intentionally silent and 9885 ± 395 ms when it
was produced intentionally aloud. ANOVA revealed a
condition effect (F ¼ 6.38, P ¼ 0.0027). Post hoc analysis demonstrated that the spontaneous production of
the hallucinatory tune was significantly slower than the
production of the same tune voluntarily and intentionally aloud (P < 0.0001).

Ó 2006 EFNS European Journal of Neurology 13, 1378–1380

1380

I. Biran and I. Steiner

Figure 2 Mean reaction time in milliseconds to complete a hallucinatory tune in the three conditions of part 3. (*P < 0.0001).

Discussion
Biological pacemakers are part of an internal clock system, similar to those designed by man, and are expected
to control event durations or temporal intervals in a
precise and predicted pattern. The performance of the
controls and the patient in the first part of the experiments confirmed the existence of such a time keeper. As
expected, both patient and controls were slower with the
actual production of a tune compared with its mental
imagery and the ratio between the two conditions was in a
narrow range (95% CI )1.02 to 1.23). The existence of
such a pacemaker was fostered by the finding that even
when this constant pattern was entirely reversed as the
patient was asked to perform the unsolicited hallucinatory tune, the complete adherence to time durations
during the repetitions remained unchanged.
This reversal of pattern can imply that the hallucinatory tune is activated by an independent and aberrant
neural mechanisms with a different time keeper, which
follows other rules (longer durations for musical imagery compared with actual production of a tune), or
that the hallucinatory tune itself differs from the original tune and thus cannot be compared. Alternatively,
this result could represent the action of a single pacemaker within a frame of an internal clock: as time
processing is influenced by attention and working
memory, it could be argued that the timing of a hallucinatory tune requires different memory and attentional
resources, especially so, when the tune is attended to
and not produced, as in the third part of the experiment. Differential allocation of neural resources for the
hallucinatory tune may therefore bring about this differential pattern [8], and is in accordance with the
observation that irregular and unpredicted appearance
of a stimuli influence the reaction times [9]. It could also
be argued that non-central nervous system factors, such
as heart rate, could have caused the effect observed.

However, as the different conditions were presented in
random blocks, we believe that this is not the case.
The functional locus of the impairment in our patient
is not clear. It could be argued that the abnormal pattern observed arises at the level of execution and that
the abnormal pacemaker is involved only in motor
execution. However, as the patient showed a reversal of
the timing ratios of the actual and imagery production
of the hallucinatory tune, the pacemaker is probably
involved in more than just tasks of motor production.
The anatomical source of the hallucinations in our
patient is unknown. It is intriguing that the patient had
a structural damage in the region responsible for
auditory sensation and interpretation. Our findings
could theoretically also represent secondary functional
changes in auditory processing in response to peripheral
hearing loss.
Regardless of the primary abnormality, our findings
suggest that the music production under normal conditions and during disease is dependent upon the differential action of an internal pacemaker. Further
studies, looking at other variables such as rhythm and
musical motifs in the timing of the production of
musical extracts can shed more light on this issue.

References
1. Evers S, Ellger T. The clinical spectrum of musical
hallucinations. Journal of the Neurological Sciences
2004; 227: 55–65.
2. Kasai K, Asada T, Yumoto M, Takeya J, Matsuda H.
Evidence for functional abnormality in the right auditory
cortex during musical hallucinations. Lancet 1999; 354:
1703–1704.
3. Cerrato P, Imperiale D, Giraudo M, et al. Complex
musical hallucinosis in a professional musician with a left
subcortical haemorrhage. Journal of Neurology, Neurosurgery and Psychiatry 2001; 71: 280–281.
4. Fischer EC, Marchie A, Norris M. Musical and auditory
hallucinations: a spectrum. Psychiatry and Clinical Neurosciences 2004; 58: 96–98.
5. Mathuranath PS, Nestor PJ, Berrios GE, Rakowicz W,
Hodges JR. A brief cognitive test battery to differentiate
Alzheimer’s disease and frontotemporal dementia. Neurology 2000; 55: 1613–1620.
6. Halpern AR. Cerebral substrates of musical imagery.
Annals of the New York Academy of Sciences 2001; 930:
179–192.
7. Cohen JD, MacWhinney B, Flatt M, Provost J. A new
graphic interactive environment for designing psychology
experiments. Behavioral Research Methods, Instruments,
and Computers 1993; 25: 257–271.
8. Rao SM, Mayer AR, Harrington DL. The evolution of
brain activation during temporal processing. Nature Neuroscience 2001; 4: 317–323.
9. Grosjean M, Rosenbaum DA, Elsinger C. Timing
and reaction time. Journal of Experimental Psychology:
General 2001; 130: 256–272.

Ó 2006 EFNS European Journal of Neurology 13, 1378–1380