The Measurement of Music and the Cerebral Clock: A New Theory

Author(s): Ernst Pppel

Source: Leonardo, Vol. 22, No. 1, Art and the New Biology: Biological Forms and Patterns
(1989), pp. 83-89
Published by: The MIT Press
Stable URL: http://www.jstor.org/stable/1575145
Accessed: 06/03/2009 08:50

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=mitpress.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.

The MIT Press is collaborating with JSTOR to digitize, preserve and extend access to Leonardo.

http://www.jstor.org
The Measurement of Music and the
Cerebral Clock: A New Theory

Ernst Poppel

ABSTRACT

In his essay "On Conducting", Richard Wagner Ansermet in TheFundamentalsof

wrote: Music [3] or David Epstein in Be- How is it possiblethat,during
the performance of music,tempo
If one were to summarize what a conductor must do to assure yond Orpheus [4] attempted to can be controlledinsucha way
the correct performance of a piece of music, it consists of con- do, but rather from a scientific that,evenfora long-lasting piece,
tinuously setting the right tempo, because the choice and reg- point of view. I want to present a thereare onlyminutedifferencesof
ulation thereof shows us immediately whether the conductor few observations to demonstrate tempobetweenthe beginningand
has understood the musical composition or not [1]. the end?Fromthe biologicaland
that the possibility of controlling thisques-
psychological viewpoint
Bruno Walter wrote in "Musicand Its Performance": tempo in music depends on cer- tionis ratherdifficult to answer.A
tain mechanisms in the brain. newtheoryis proposedthattriesto
If I was not sure of the correct tempo of a musical phrase or linktempocontrolinmusicto the
Our perception of music, then,
episode for a long period of time, suddenly, as from the depths neuronalmechanismsof the brain
of my inner being, a decision occurred to me. Like in a mo- also has a biological basis-a
thatunderliehumantimeperception
ment of enlightenment, the right tempo appeared, and I had fundamental thesis in these re- andthe timingof intentional acts.
a feeling of security that freed me of doubt and, in most cases, flections. Elementary temporalexperiences-
this remained forever [2].
To illustrate the importance i.e. simultaneity,asynchrony, suc-
If tempo is so fundamental to music, then we must ask of cerebral processes in the per- cession,the subjectivepresentand
duration-arestructured withina
ourselves how we find the right tempo and how we maintain ception of music, it is first nec- hierarchical model.Oneaspectof
it once it has been found. How can we throw light on this essary to clarify how we handle thismodelis thatit allowsone to
problem from the standpoint of brain research and experi- time-how we perceive time and talkabouta 'brain-clock' thatunder-
mental psychology? Here, I would like to examine this prob- what it means to us. Immanuel lies, it is suggested,tempocontrol
lem for once not from a purely musical viewpoint, as Ernest Kant said, "Time and space are inmusic.
two sources of knowledge from
Fig. 1. Experimental evidence suggests that there is a neuronal os- which, a priori, different synthe-
cillator or 'brain-clock' that is responsible for the tempo control tic realizations can be obtained"
of behavioral sequences such as speech or the performance of
music. Time (t) is indicated by the arrow from left to right; the [5]. Time is a basic category in the human experience of re-
vertical bars indicate identical phases ('ticks') of the 'brain-clock'. ality. This raises the question of how time is made available
as an experience.
Neuronal
Oscillator("Brain-Clock") In the eleventh book of his Confessions[6], Augustine
I II I I I I I I I I I I I I I I I I I I I I I I IIII wrote: "Quid est ergo tempus? Si nemo ex me quaerat, scio;
t
si quaerenti explicare velim, nescio." (What is time? If no
one asks me, I know the answer; if I want to explain it to
Tempo Controlof behavioral
sequences by pulses someone who has asked, I do not know.) In spite of Augus-
of the "brainclock" tine's skeptical comment, I still dare to submit the question,
not of what time is, but of how time is being subjectively
made available to us, how we perceive and experience it.
Fig. 2. A neuronal oscillator controlling tempo in musical per-
formance would allow precisely controlled tempo changes (the Perhaps then it will be easier to answer the question of how
'Epstein Phenomenon'). If musical performance is coupled to a tempo is controlled in music-the problem that occupied
'brain-clock' and a tempo change in music is required, the next both Wagner and Walter.
tempo is numerically related to the initial tempo according to 'the Human temporal experience can be described by a hier-
strong laws of small numbers' (e.g. 2:1). Deviant tempo changes archical classification of subjective temporal phenomena
(e.g. 9:4) would violate this law because the control of the second
[7]. Five elementary experiences can be distinguished and
tempo by the 'brain-clock' would be much more complicated,
requiring additional neuronal mechanisms. categorized hierarchically: simultaneity, non-simultaneity
(asynchrony), succession, the subjective present (now) and
NeuronalOscillator ("Brain-Clock") duration. What these temporal phenomena are, on the sub-
I IIiII J Il l ll llll l l llIIl I IIi IIII lIl t jective level, and how they can be hierarchically ordered,
can be elucidated by various experimental observations.
2:1
"Epstein-
Tempo- I Phenomenon"
Control J* *
Ernst Poppel, Institut ffir Medizinische Psychologie, Goethestrasse 31/I,
JJ J J tXJ J X 9:4 D-8000 Munich 2, F.R.G.
Translated by Barbara Herzberger.
Received 16 March 1988.
Change of tempo

? 1989 ISAST
PergamonPressplc.PrintedinGreatBritain.
0024-094X/89
$3.00+0.00 LEONARDO, Vol. 22, No. 1, pp. 83-89, 1989 83

NeuronalOscillator ("Brain-Clock")
I I I I I I
_I - t
Tempo-L Player A i J J J j of temporal order requires a further
Control J cerebral mechanism than is necessary
J J J J
PlayerB I
Synchronization Ifthe phase angle differencebetween
"inphase" changes is too great the results may be:
a) aesthetic displeasure
b) experience of 2 tempi
Fig. 3. Neuronal oscillator of a conductor (t-arrowat the top; vertical bars indicating
'ticks' of the 'brain-clock') controlling the tempo of two players. Player A is thought to be
synchronized; player B lags behind.
TIME QUANTA- poral fusion of a succession of stimuli
in the visual or tactile realms have
EVIDENCEFOR A shown that the fusion threshold is ap-
'BRAIN-CLOCK' proximately 10 ms in the tactile system
and about 20 ms to 30 ms in the visual
First, let me demonstrate some find- system. These thresholds are depend-
ings on subjective simultaneity and ent on the specific stimuli selected. If
asynchrony. If, over earphones, a test the sensory systems hearing, touch
subject is given a brief stimulus lasting, and vision are compared, it is striking
for example, 1 ms (0.001 second) in that the transition from simultaneity
each ear, then when both stimuli are to asynchrony differs among them,
produced simultaneously, only one whereby the auditory system has by far
tone is perceived-in the middle of the lowest fusion threshold. The
the head. If a delay of, say, 1 ms occurs differences in fusion thresholds of the
between the two stimuli, still only one various sensory systems are probably
tone is perceived, although, objec- based on their mechanisms of trans-
tively speaking, the two stimuli are not duction. Transduction refers to the
simultaneous. Objective asynchrony is transformation of physical events (e.g.
not sufficient for subjective asyn-
light or sound) into action potentials
chrony. The tone, however, is no in the brain. Transduction is known to
longer perceived in the middle of the take considerably longer within the
subject's head but either on one side visual system than within the auditory
or on the other. Only if the temporal
system, which apparently results in a
difference between the two acoustic far less well differentiated perception
stimuli is 3 ms (4 ms to 5 ms in some of simultaneity in the visual realm than
test subjects) is the threshold of the in the auditory realm.
perception of asynchrony reached, During experiments involving the
and the subject hears a separate sound fusion threshold in the different sen-
in each ear.
sory systems, my colleagues J. Ilm-
If the same experiment is per-
berger, N. V. Steinbuichel and I in-
formed on patients with brain dam-
quired whether one or two stimuli had
age, e.g. after a stroke in the left cere- been perceived. When only a minimal
bral hemisphere with a resulting alteration of the experimental condi-
speech disturbance, the measured tions was made, it became obvious that
transition from synchrony to asyn-
simply changing the question could
chrony-the auditory fusion thresh- lead to a different test result. In the
old-is also between 3 ms and 5 ms. next experiment, we no longer asked
They have the same auditory fusion if one or two stimuli had been per-
threshold as healthy subjects. The cen- ceived; rather, the subject had to indi-
tral brain damage demonstrates no in- cate which stimulus came first and
fluence on the transition from syn- which second. The question was di-
chrony to asynchrony. One therefore rected at the order in which the per-
surmises that the ability to perceive ceived stimuli occurred. Although the
temporal differences among stimuli transition from simultaneity to asyn-
depends upon peripheral mecha- chrony took 3 ms to 5 ms in the audi-
nisms of the nervous system.
tory system, 30 ms to 50 ms were neces-
Experiments to determine tem- sary for the recognition of temporal
84 The Measurement of Music and the Cerebral Clock
PiPppel,
order under identical experimental
conditions. Apparently, the new ques-
tion activated a different cerebral
mechanism.
This thesis, i.e. that the recognition
for the recognition of asynchrony, has
been verified by experiments on pa-
tients who have suffered central brain
damage. In these experiments the
threshold for the recognition of tem-
poral order is raised to approximately
100 ms. In contrast to the auditory fu-
sion threshold, which is roughly the
same among healthy and brain-dam-
aged subjects, an obvious difference in
thresholds can be observed in the
recognition of the temporal order of
events.
A surprising result is that among
healthy subjects the threshold for the
recognition of temporal order in the
three sensory systems appears to be
the same for any one individual,
whereas the thresholds for the recog-
nition of asynchrony differ. This ob-
servation suggests that the recognition
of temporal order necessitates only
one cerebral mechanism, which is sim-
ilarly available to the three sensory sys-
tems investigated, whereas other, pre-
sumably peripheral, mechanisms are
responsible for the recognition of
asynchrony.
If events are to be placed in a tem-
poral order, indicating which came
first, etc., it is essential for the indi-
vidual events to be identified. A ques-
tion concerning the mechanisms re-
sponsible for the recognition of
temporal order involves determining
the minimal time required to identify
events and to make this information
consciously available. Experiments to
determine the minimal times that are
required to make a conscious identifi-
cation have shown that 30 ms to 50 ms
seem necessary.
This hypothesis has been substan-
tiated by investigations measuring re-
action times involved in making a
choice [8]. Above all, studies of the
intrahemispheric reaction times in
choice making, i.e. when stimuli were
presented to only one hemisphere of
the brain, illustrated that minimal
times of between 30 ms and 50 ms were
required to decide between stimuli. It
is theoretically possible that each
stimulus activates an oscillating pro-
cess in the brain, whereby, technically
speaking, this process is to be under-
stood as a relaxation oscillator. Such
neuronal oscillators are immediately
synchronized by the sudden appear-
ance of a stimulus. Presumably,
sensory stimuli lead to similar, if not
identical, oscillatory processes in the
various sensory systems. A visual, audi-
tory or tactile stimulus causes periodic
discharges in the stimulated neural
networks. This creates a temporal
framework, which enables events re-
corded by the various sensory systems
to be correlated. If there were no such
temporal framework, it theoretically
would be extremely difficult to com-
pare data in the various sensory sys-
tems and to coordinate them. I assume
that the periods of this oscillator,
which lie between 30 ms and 50 ms ac-
cording to experimental results, deter-
mine the fundamental rhythm charac-
terizing our mental activityand can be
defined as the minimal time required
to identify single events [9].
The discovery of such a neuronal
oscillator involves the simultaneous
discovery of a clock in the brain (Fig.
1). The introductory quotes by Wag-
ner and Walter, indicating the impor-
tance of the correct tempo in music,
can now be discussed on a neuro-
biological basis. The periodic pro-
cesses of the neuronal oscillator pre-
sumably have a constant frequency
under given conditions. It can be as-
sumed that tempo control in music is
based on the coupling of musical ex-
pression tojust such an oscillatory pro-
cess in the brain. This ensures that the
tempo of a piece of music can be main-
tained, for the neuronal oscillations
can be used like the ticking of a clock.
The relatively high frequency of this
neuronal oscillator guarantees that
variations in the periods of the oscilla-
tory process will have little effect on
the musical expression, since the
latter is manifested during a much
longer period of time. The availability
of a neuronal clock in the brain makes
it neurophysiologically possible to
maintain a constant tempo.
In 1985, the American musician,
conductor and composer David Ep-
stein investigated the precision with
which tempo is maintained [10]. By
examining tempo relationships within
musical pieces from different cultures,
he was able to demonstrate that, when
a new tempo is chosen, it is dependent
on the preceding tempo. If the tempo
speeds up, it becomes almost exactly
twice as fast; if it slows down, it decel-
erates to one-half or one-third of the
previous speed. This kind of tempo
change is only possible in the presence
of an effective, central tempo control.
According to my hypothesis, this is
provided by the periodic processes in
the neuronal networks of the brain.
Figure 2 illustrates this idea graphi-
cally. A clock in the brain enables tem-
po to be controlled in the presence of
two different tempi by ignoring one
pulse of the neuronal oscillator when
the tempo is changed. This leads to
the simple numerical relationships
between tempi when accelerated or
decelerated. This tempo relationship,
referred to as the 'Epstein Phenom-
enon', contrasts with tempo relation-
ships in which the simple numerical
relationships are ignored. Such tempo
relationships (the example provided
is 9:4) are not biological, and if forced
by the musician, they can produce aes-
thetically unpleasant experiences. A
departure from the time structure pre-
determined by the brain changes the
musical impression qualitatively. I as-
sume that this is detrimental to the
overall impression of the piece of
music.
These neuronal oscillations, which
are provided by the cerebral clock,
enable musicians to play together syn-
chronically. A predetermined tempo
ideally leads to temporal unison-to
synchronization of the cerebral clocks
of the participating musicians. It is oc-
casionally possible for disagreement
to exist among the musicians concern-
ing the tempo to be chosen. One musi-
cian may try to play more slowly than
the others. Figure 3 demonstrates
such a situation. The musicians con-
tinue to play in unison, but one is a
fraction of a second behind the others.
Such audible differences in the tem-
poral phase generally have one of two
consequences for the listener. Usually,
the listener finds the effect aestheti-
cally displeasing; however, the listener
can also find it surprising to perceive
two different tempi.
Recent neurological research has
revealed additional information,
55
Fig. 4. Duration
of intentional 50
acts (like a greet- 45
ing ceremony) 40
from subjects >35
from four dif-
D 30
30
ferent cultures.
Note the tem- 25
= 20
poral preference
of such acts up 15
to approximately 10
3 seconds (from 5
Ref. [15]).
0
0 1
P6ppel,The Measurement
namely that the temporal organiza-
tion of our perception, e.g. the recog-
nition of temporal order, is primarily
a function of the left cerebral hemi-
sphere. It can be demonstrated that in-
juries to the left cerebral hemisphere
lead to considerable changes in the
temporal process of perception. It has
also been observed that the right
hemisphere is essential for the appre-
ciation of music. As has been long
known, the right hemisphere is domi-
nant for spatial conceptions; it also
seems dominant for emotions [11].
Investigations in various laboratories
have shown that the right hemisphere
also plays an essential role in certain
aspects of musicality. Experiments
were conducted in which the left or
right cerebral hemisphere was tem-
porarily anaesthetized; the results
proved that the modulation of pitch,
in particular, is imparted by the right
hemisphere. Neurosurgical proce-
dures are sometimes carried out in
severe cases of epilepsy, which cannot
be adequately controlled pharmaco-
logically. These involve either severing
the corpus callosum (interhemi-
spheric commissural fibers) or resect-
ing a certain small portion of the brain
to prevent the epileptic focus from
spreading further. To avoid damaging
the speech center of the brain during
such operations, the Wada test is
administered preoperatively to deter-
mine which hemisphere houses the
function of speech. During the Wada
test, one of the two hemispheres is
functionally blocked for a few min-
utes. This allows the time needed to
test different functions, i.e. to deter-
mine if they are available when one
hemisphere has been anaesthetized.
During this diagnostic test, if the
patient is instructed to speak or to sing
when the left hemisphere is discon-
nected, speech or singing immediately
ceases. If the patient is instructed to
sing when the right hemisphere is tem-
- N-N=237
_i i
1l]l R
9
2 3 4 5 6 7 8
time [s]
w of Music and the Cerebral Clock 85
~~~~~~85
 porarily blocked, all temporal func-
tions of musical expression remain
intact, e.g. the patient still sings rhyth-
mically and can control the tempo
precisely. The patient, however, is no
longer able to modulate pitch. The
song is sung on only one note. This
leads one to believe that the right
hemisphere holds the key to pitch
modulation. This conviction is sup-
ported by numerous experiments on
patients with delimited right hemi-
spheric lesions, whose ability to modu-
late pitch has disappeared and whose
capacity to differentiate among vari-
ous pitches is limited.
This evidence presumably corre-
lates with right hemispheric domi-
nance in certain aspects of our emo-
tions. It is thus plausible that it is the
modulation in pitch that lends music
its particular emotional effect. A. Holl-
mann and I conducted a test to inves-
tigate the extent to which patients with
right hemispheric lesions can recog-
nize the emotional content of a piece
of music. We used songs that were
from an African culture but whose
emotional meaning would be recog-
nizable to any healthy member of our
culture. The patients we tested, how-
ever, were severely handicapped in
evaluating these songs. In particular,
they were no longer able to recognize
sorrow expressed in a song. This ob-
servation verifies the importance of
the right hemisphere for the recogni-
tion of a musically expressed emotion.
THE 3-SECOND
WINDOW OF
TEMPORAL
INTEGRATION
Returning to the hierarchical classifi-
cation of the subjective experience of
time, how far has this taxonomic sys-
tem progressed? Examination of syn-
chrony, asynchrony and succession
raises the question of whether these
three aspects alone are sufficient to
grasp the meaning of experienced
time.
A brief reflection reveals that an ad-
ditional mechanism must be necessary
to perceive time. Everyone knows
from personal experience that events
are not experienced in isolation, but
rather that single occurrences are re-
lated to one another and, normally,
create perceptive forms (Gestalten) in
which several events are combined.
This is made possible by a mechanism
86 Pippel, The Measurement of Music and the Cerebral Clock
in the brain that has an integrative integrating mechanism, which is 3 sec-
function. This integrating mechanism onds on the average.
can be demonstrated for the phenom- One must ask whether there exists
enon that I refer to as the subjective further evidence that the subjective
present, or 'now'. present or immediate consciousness
A simple experiment illustrates the can last only a few seconds. Many ex-
integration of single events into uni- periments on speech behavior have
fied concepts. If a metronome ticks been conducted to investigate this.
every second, it is easy for everyone to One can demonstrate, for example,
pick up a subjective beat. Everysecond that verbal units in spontaneous
beat can be given a subjective accent, speech are limited to about 3 seconds.
although all beats are identically loud. Tests concerning the recognition of
It probably also would be possible to rhythm can also be cited [13], al-
unite three successive beats into a sub- though they determined a somewhat
jective Gestalt by attributing emphasis shorter temporal limit. A further area
to every third beat, although this may that is especially illustrative of the ex-
be difficult for some. Trying to com- istence of a finite limit to conscious-
bine four, or even five, consecutive ness involves particular phenomena in
beats subjectively into a recognizable the arts.
form will prove even more difficult for The 3-second phenomenon can be
many. This test demonstrates that the verified in poetry as well as in music.
integration of successive events into Experiments that Fred Turner and I
perceptual forms has a temporal limit, conducted in 1983 using poems in
which seems to be a few seconds. Nu- different languages demonstrated
merous experiments, especially con- that verse lines preferably take only 3
cerning the temporal organization of seconds when spoken aloud [14]. Re-
vision, illustrate that the limit above gardless of the language involved, we
which we are no longer able to create discovered that there appears to be a
cognitive entities by integrating events universal temporal phenomenon,
is about 3 seconds [12]. These tem- which poets, presumably in all lan-
poral limits of integration can vary guages, have observed. Assuming that
individually, i.e. slightly longer or Latin and ancient Greek were spoken
shorter intervals are also possible. in a tempo similar to languages of to-
I would like to suggest applying the day, verse lines in antiquity observed
phenomenon of temporal integration the 3-second segmentation of speech.
to the formal definition of the subjec- Considering the wide variety of
tive present, or 'now'. That which is grammatical possibilities and cultural
subjectivelyavailable to us is so for only traditions, there is no recognizable
a few seconds. The mental availability reason for such an ubiquitous 3-
of something of which we are only tem- second segmentation. It would have
porarily conscious is determined by been easy to compose longer poetic
the temporal limitation of a central lines. The apparent reason is that the
3-second segmentation of the con-
Fig. 5. Visuali7ation of two brain mechanisms underlying tempo control and temporal
integration during musical performance. A neuronal oscillator provides 'ticks' of a 'brain-
clock' (t-arrowat the top). It is suggested that this 'brain-clock' underlies tempo control
and also provides the temporal framework for event detection. In addition, a neuronal
mechanism is assumed that integrates successive events into units. Experimental evidence
suggests that sensory stimuli lasting longer than the integration interval do not lead to the
subjective impression of a musical movement and thus of tempo. Only if several events
are recorded within one integration interval-using the 'brain-clock' as a device to do so-
will musical movement and thus tempo be perceptually available.
NeuronalOscillator ("Brain-Clock")
i I 1 IIII1 1111I I1 1IIi 111.III. t
L-I-. 1 I
^J , II II I1 1
I IntegrationIntervals
^. ~ @ # perception
Tempo- HHHHHHH HtHHHH HH of tempo
* Control t I No perception
durationof acoustic stimuli of tempo

William
Shakespeare(1564-1616)
Sonnet 18
Shall I compare thee to a summer's day' ~~~~?
Thou art more lovely and more temper ate.
Rough winds do shake the darling buds of May,
And summer's lease hath all too short a date:
Sometime too hot the eye of heaven shiines,
And often is his gold complexion dimmi'd;
And every fair from fair some time decliines,
By chance, or nature's changing course, ,untrimm'd;
But thy eternal summer shall not fade
Nor lose possession of that fair thou ow''st;
Nor shall Death brag thou wand'rest in his shade,
When in eternal lines to time thou growv'st.
So long as men can breathe or eyes can see,
So long lives this, and this gives life to thlee.
tents of our consciousness is typical of
the organization of our mental exist-
ence and that poets have implicitly
maintained this segmentation. I as-
sume, therefore, that this is a univer-
sal phenomenon, applicable to every-
one. This phenomenon is so stable
that it has not been altered by cultural
traditions, in spite of their many differ-
ent standards, which themselves have
evolved with time. To illustrate this
point (or, rather, to make it audible),
I offer a short poem (see Table 1),
which the reader can read aloud and
thereby time the lines.
An approximate 3-second segmen-
tation can also be observed in spon-
taneous behavior. My colleagues M.
Schleidt, I. Eibl-Eibesfeldt and I were
able to demonstrate an identical tem-
poral segmentation in intentional be-
havior in the cultures of four different
peoples (Europeans, Yanomami Indi-
ans, Kalahari Bushmen, and Tro-
briand Islanders) [15]. Ceremonial
greetings, playful gestures directed
toward others, and many other types
of behavior characterized by inten-
tionality also demonstrate a temporal
limit, which is about 3 seconds. Since
the temporal segmentation of the four
cultures is practically identical, the
data have been summarized in a single
histogram in Fig. 4. The preference
for a duration of action of a few sec-
onds is obvious.
A further area in which temporal
segmentation becomes apparent is
music. Musical themes often have a
temporal limit of about 3 seconds. The
famous theme from Beethoven's Fifth
Symphony or the Dutchman's theme
Table1. "Sonnet
18"byWilliam
Shakespeare.
~ Duringrecitation,
eachspokenline
lastsapproximately
3 seconds,which
correspondsto the
upperlimitof tem-
poralintegration
in
humanperception.
from Wagner's The Flying Dutchman
can serve as representative examples,
though any number of other examples
could be given [16]. At least in the
tradition of occidental music, there
seems to be a universal phenomenon
at work here that cannot be ignored
by the composer or the performing
musicians.
If one hears music in which the tem-
poral structure, as it is referred to
here, is abandoned, then the aesthetic
effect is also altered. Apparently,
mechanisms of the human brain de-
fine constraints that are also applied
for aesthetic evaluation. If the biologi-
cally predetermined temporal frame-
work is violated or ignored, the aes-
thetic frame of reference within which
the music generally is judged is also
changed.
Figure 5 shows how the cerebral
clock and temporal integration affect
the quality of certain auditory impres-
sions. The considerations shown here
in diagram form are supported by vari-
ous experimental observations. If dif-
ferent acoustic stimuli are offered
within an integration interval, the lis-
tener automatically perceives a tempo.
If the duration of individual stimuli is
longer than an integration interval,
the feeling of tempo disappears, and,
therefore, a musical movement can no
longer be recognized. This means
that, qualitatively, a new perceptual
quality is created if the temporal
framework for the integration of dif-
ferent events is exceeded. It can be as-
sumed that tempo variations appear-
ing within an integration interval
create the impression of a different
P6ppel,The Measurement of Music and the Cerebral Clock
musical movement, and that the per-
ceptual quality of a musical movement
is available only if several different
events can be integrated within one 3-
second window.
It is probably becoming increas-
ingly clear that humans are directed,
although unconsciously, by biological
constraints in their subjective attitude
toward a musical work of art. Although
artists do not orient themselves ac-
cording to known physiological brain
data during their creative activity,they
automatically and unconsciously heed
constraints predetermined by mecha-
nisms of the brain-temporal con-
straints in the case of music.
I should point out, however, that
the biological constraints I am discuss-
ing are not absolutely rigid. Some
people, when they hear of biological
constraints, have the impression that
these are unavoidable determinants,
and thus fear a loss of freedom. What
happens to artistic freedom?
This fear is dispersed by a law of
human perceptual organization. The
modern theory of perception assumes
that the perceived is not dependent
upon the constellation of stimuli
alone, but quite decidedly upon the
expectation (hypothesis) that the per-
ceiver has at a certain moment. The
following law can be formulated: Per-
ception is the acceptance or rejection
of a hypothesis that a subject has
formed concerning the condition of
the world (or him- or herself); the per-
ceiver structures, as it were, the stimuli
confronting him or her according to
the viewpoints of subjective relevance.
This rule of perception, which re-
lates to perceptual contents, seems to
apply also to the formal organization
of temporal perception. Figure 6 is a
schematic diagram of this concept for
the perception of music. I am acting
on the assumption that there is a cere-
bral clock that, as I set forth in this
paper, makes it possible to maintain a
constant tempo or to perceive musical
movement. This primary tempo con-
trol, predetermined by the brain, is
presumably variable due to what is ex-
pressed musically. A temporal integra-
tion interval of several seconds is re-
quired to express a musical theme; for
this neuronal integration to be carried
out optimally, it is conceivable that a
fine adjustment of the cerebral clock
is achieved via semantic feedback.
This variation can occur only within
limits, but it seems theoretically pos-
sible; experimental studies are needed
to test this thesis. I assume that a less
87

Semantic feedback on neuronal tempo control
NeuronalOscillator (primaryendogenous tempo control)
I I II I i I II I I I I1 I III I I I I I II I III I I II I I_ t which consists of interdependent con-
I II -... . The fact that this is an active process
neuronalintegrationinterval
(optimaltemporalwindow for a musical motive)
calib
iSecondary n of n l ments so that the meaning of the in-
S11111111 I1 I I I1 I I1 I I I I t dividual conscious concepts produces
Secondary calibrationof neuronaloscillation
Fig. 6. A neuronal oscillator provides a basic tempo control (t-arrowat the top), and a
neuronal integration interval provides a temporal window for musical expression. It is
suggested that there may exist a 'top-down' mechanism (within limits) that allows for the
fine tuning of tempo control ('secondary calibration'). A musical motive may require a
particular temporal window in order to be expressed properly. This may result in an
alteration of the period of the underlying neuronal oscillation. The implication of this
concept is that semantics (the musical motive) may alter in a self-referential manner the
formal, i.e. temporal, structure that is required to be expressed.
complex, logistic function of the brain ing time. If there is a great deal of
can be influenced by a more highly mental content, then the time is retro-
complex level, i.e. that of neuronal in- spectively considered long. If, how-
tegration. It should be reemphasized ever, during a certain time interval
that an alteration in the rhythmic beat, little mental activity is going on, then
which is predetermined by the neu- the time passed appears brief in retro-
ronal oscillator, is not arbitrary but spect. This approach, of course, also
can occur only within very narrow applies to music. The duration of
limits. These considerations show that eventful music is retrospectively
two temporal mechanisms, one with
judged differently than that of mon-
cycles in the range of 30 to 50 ms and otonous music.
one consisting of segments lasting ap- Here one must assume the pres-
proximately 3 seconds, interact to pro- ence of a completely different in-
duce the experience of musical move- tegrating mechanism-memory-in
ment or to make musical themes which information is cumulatively
mentally available. stored and whereby the stored infor-
mation can be made available with
respect to its duration. Memory, there-
MEMORY- fore, is a prerequisite for our experi-
ence of duration. Memory fulfills an
PREREQUISITEFOR
additional function regarding tem-
THE EXPERIENCEOF poral experience-it enables us to
DURATION prepare for future situations by re-
viewing past events to determine simi-
These reflections lead to the fourth larities.
stage in the hierarchical classification Finally, one ponders how it is pos-
of subjective time. I have taken a posi- sible for one's experience to be char-
tion on the elementary temporal acterized by subjective continuity. I
experiences: synchrony, asynchrony, have already established that inte-
succession and the subjective present.
gration mechanisms unite data into 3-
When speaking of subjective time, one second segments, which we become
has a further phenomenon in mind: aware of as 'pictures of the present'.
What mechanisms are involved when We must assume that further mecha-
certain time intervals are experienced nisms are at work that coordinate the
as having different durations? In The
meaningful contents of individual
Magic Mountain, Thomas Mann for- conscious experiences. That which we
mulated probably the most important
consciously experience is not inde-
observations on this subject [17]. It
pendent of that which we have already
seems that the mental content deter-
consciously experienced. That which
mines the subjective duration of pass- is available to memory becomes a part
88 P'ppel, The Measurement of Music and the Cerebral Clock
of our awareness together with the in-
formation provided by the act of per-
ception. A chain of sequential con-
scious elements is thereby created,
scious concepts.
of the brain can be deducted by ob-
serving patients with formal thought
disturbances. In extreme cases, schizo-
phrenic patients are no longer able to
correlate a sequence of conscious ele-
a rational chain of thought. Formal
thought disturbances seem to be
caused by the lack of ability to utilize
operative brain functions at the high-
est level, which are normally respon-
sible for the conceptual integration of
sequential elements of our conscious-
ness. Such patients have lost the conti-
nuity of their experience and thereby
the subjective impression of a tem-
poral flow, which is characteristic of
the experience of healthy individuals
and which Robert Musil described in
his novel TheMan WithoutCharacteris-
ticsin the following way:
The trainof time is a trainthat rolls
itsowntracksaheadof it. The riverof
time is a river that carries its own
banks along with it. The traveler
movesbetweenstablewallson a stable
floor;but the floor and the wallsare
moved, unnoticeably,by the move-
mentsof the travelerin a most lively
way[18].
IMPOSSIBLE MUSIC-
A FINAL PREDICTION
If one accepts the hypotheses and re-
flections presented here, then, in con-
clusion, an assumption can be formu-
lated concerning 'new' music, whose
performance overtaxes human capa-
bilities. If biological constraints are so
powerful that one cannot free oneself
of them, then one must, on principle,
be able to create new auditory realms
that can be realized only with modern
electronic methods. One must be able
to 'compose' impossible music that
cannot be performed by humans. I
maintain that musicians cannot devi-
ate from the fundamental tempo con-
trols. Music whose duration of sequen-
tial tones is randomly varied, whereby
the duration has to be brief in com-
parison with the integration interval of
about 3 seconds, cannot be performed
by a musician because a fundamental
temporal structure, which is predeter-
mined by the brain, would have to be
invalidated. That is hardly possible.
Music of this kind, played by a com-
puter (one must question if this is
music at all), would have to create,
qualitatively, a completely different
impression than 'normal' music. My
thesis thus is: For humans, there is in
music no freedom of tempo.
Acknowledgments
I am grateful to David Epstein for his many sug-
gestions. I wish to thank W. Simon for encourag-
ing me to write down my partially speculative re-
flections. Barbara Herzberger accepted the
difficult task of translating the German version of
the manuscript into English. Gabi de Langen and
Monika Herzog were very supportive during the
conception of the text and the production of il-
lustrations. The experiments on which I based my
observations were carried out with the support of
the German Research Foundation (DFG-PO 121-
13/1; project No. 6).
References and Notes
1. Richard Wagner, "Uber das Dirigieren",
quoted in Bruno Walter, Von derMusik und vom
Musizieren(S. Fischer Verlag, 1957).
2. Walter [1].
3. Ernest Ansermet, Die GrundlagenderMusik im
menschlichenBewusstsein(Munich: Piper, 1985).
4. David Epstein, BeyondOrpheus:Studiesin Musi-
cal Structure(Cambridge, MA: MIT Press, 1979).
5. Immanuel Kant, Kritik der reinen Vernunft
(1787; reprint Hamburg, 1956).
6. Augustinus, Bekenntnisse(Confessions) (397/
398; reprint Munich, 1955).
7. E. P6ppel, Grenzendes Bewusstseins.UberWirk-
lichkeitund Welterfahrung(Stuttgart:Deutsche Ver-
lagsanstalt, 1985). Translated into English under
the title Mindworks:Timeand ConsciousExperience
(Boston: Harcourt BraceJovanovich, 1988).
8. E. P6ppel, "Time Perception", in Handbookof
SensoryPhysiology,Vol. 8: Perception,R. Held, H.
Leibowitz, H.-L. Teuber, eds. (Berlin: Springer-
Verlag, 1978) pp. 713-729.
9. E. P6ppel, "ExcitabilityCycles in Central Inter-
mittency", Psychol.Forschung34, 1-9 (1970), and
Poppel [7] and [8].
P6ppel,The Measurement of Music and the Cerebral Clock
10. David Epstein, 'Tempo Relations: A Cross-
cultural Study", Music TheorySpectrum7, 34-71
(1985).
11. E. P6ppel, Lust und Schmerz. Grundlagen
menschlichen Erlebens und Verhaltens (Berlin:
Severin und Siedler, 1982).
12. P6ppel [7] and [11].
13. H. Feldmann, "DasWesen des Rhythmus im
Experiment an Geh6rlosen und Normalsinni-
und ZeitschriftfiirNeuro-
gen", ArchivfuirPsychiatrie
logie194, 36-61 (1955).
14. F. Turner and E. P6ppel, 'The Neural Lyre.
Poetic Meter, the Brain, and Time", Poetry,277-
309 (August 1983).
15. M. Schleidt, I. Eibl-Eibesfeldt and E. P6ppel,
"A Universal Constant in Temporal Segmenta-
tion of Human Short-term Behavior", Naturwis-
senschaften74, 289-290 (1987).
16. P6ppel [7].
17. Thomas Mann, Der Zauberberg
(1924; reprint
Frankfurt, 1967).
18. Robert Musil, Der Mann ohne Eigenschaften
(Hamburg: Rowohlt, 1978).
89