528063

research-article2014
POM0010.1177/0305735614528063Psychology of MusicCha

Article

Psychology of Music

The Takadimi system reconsidered:

2015, Vol. 43(4) 563­–577
© The Author(s) 2014
Reprints and permissions:
Its psychological foundations and sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/0305735614528063
some proposals for improvement pom.sagepub.com

Jee-Weon Cha

Abstract
The Takadimi system of rhythm pedagogy codified by Hoffman, Pelto, and White is supposedly an
effective means to teach both dictation and reading. This article evaluates the potential of Takadimi
as a method for rhythm dictation and rhythm reading, considering its strengths and weaknesses
in light of psychology, music theory, and music theory pedagogy. Takadimi’s main forte is that it
is based on the psychologically and neurobiologically valid notion of the beat; most of Takadimi’s
shortcomings arise from its apathy towards implicit beats. After exploring the psychological reality of
the implicit beat, this study considers how an awareness of the implicit beat can make the Takadimi
system a more consistent and accurate rhythm solmization method than it currently is.

Keywords
aural skills, implicit beat, rhythm perception, rhythm solmization syllables, sight reading, Takadimi

Codified and improved by Hoffman, Pelto, and White (1996) in the 1990s, the Takadimi system
of rhythm pedagogy is, according to Karpinski (2000), an authority on music theory pedagogy,
the “most systematic and comprehensive system of rhythmic solmization in use today” (p. 81).1
Unlike notation-oriented (or symbol-specific) systems that assign a specific syllable to each note
value, Takadimi is a functional system in which all durations are labeled in relation to the beat,
i.e., according to their location or function within a beat, regardless of their notated value
(Karpinski, 2000). Takadimi was conceived, above all, to promote a sound-to-symbol approach
– that is to say, as a teaching device that would help students learn dictation, a process in which
musical sounds are converted into visual symbols: “It [Takadimi] . . . moves from the perceived
sound through a process of translation and verbal description, before addressing issues of
graphic description. In other words, identification and labeling occur before notation” (Hoffman
et al., 1996, p. 28).2

Grinnell College, USA

Corresponding author:
Jee-Weon Cha, Grinnell College, Department of Music, 1108 Park St, Grinnell, IA 50112, USA.
Email: [email protected]

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

564 Psychology of Music 43(4)

The Takadimi system has also been conceptualised as a symbol-to-sound approach that can
facilitate sight singing/reading, a process in which musical symbols are rendered into auditory
sounds. Karpinski (2000), for example, takes the system as a method that “require[s] that read-
ers understand the proportional metric values of the notes and rests as they read” (p. 169;
emphasis added). Regarding Takadimi as a pedagogic tool in rhythm sight singing/reading is
not without foundation since it is none other than Hoffman et al. (1996) themselves who sug-
gested such a possibility: “At the center of our method is a set of rhythm syllables for use in sight
singing and theory classes at all levels of instruction” (p. 13; emphasis added). The Takadimi
system is, then, meant to mediate between auralization and visualization, operating as both
sound-to-symbol and symbol-to-sound strategies towards more holistic rhythm education.
Indeed, Ester, Scheib, and Inks (2006) claim that Takadimi is an effective means to teach both
reading and notation required by Article 5 (“Reading and Notating Music”) of the National
Standards for Music Education implemented by the National Association for Music Education
(NAfME).
One may ask at this point: Is Ester et al.’s claim true? If it is, to what extent? This article
evaluates the promise of Takadimi as a method for modeling both the sound-to-symbol process
and the symbol-to-sound process – central to complex activities of rhythm dictation and rhythm
sight singing/reading respectively – considering its strengths and weaknesses in light of a vari-
ety of perspectives such as psychology, music theory, and music theory pedagogy. After a review
of the definitions of the beat (on which Takadimi is based), as well as a discussion of the psycho-
logical foundations of the beat, an introduction to the workings of the Takadimi system is pre-
sented. The psychological reality and import of the implicit beat will then be explored. Finally, I
will examine how a fully studied awareness of the implicit beat can contribute to improve the
system, making it a more consistent and accurate rhythm solmization method than it currently
is.

Beat, pulse, meter

The presence of a beat is one of the most important temporal features of metric music. Beat has
been variously defined as the “basic pulse underlying mensural music, that is, the temporal unit
of a composition” (Sadie, 2001, p. 20), a “unit of measurement of rhythmic pulse of music”
(Kennedy, 1994, p. 70), an “evenly spaced or regular pulsation” (Gauldin, 2004, p. 20), and a
“regular series of pulsations that divides a period of time into equal parts” (Houlahan & Tacka,
2011, p. 9). Without reference to metric context, these definitions are potentially misleading, if
not completely erroneous: while all beats are, in principle, regular pulsations, not all regular
pulsations are beats, and whether regular pulsations are beats or not is determined by metric
context. Beats are instrumental in meter formation. A meter in music is determined by two
criteria closely related to beats: 1) how beats are divided – the meter type is either simple or com-
pound, depending upon whether beats are split into twos or threes – and 2) how beats are com-
bined into a larger, recurring pattern – the meter type is duple or triple, depending upon whether
beats are grouped into sets of two or three.3 In a word, beats are the standard by which lower
level (beat divisions) and higher level (measures) of metric activities are gauged and, depending
upon metric context, regular pulsations may be interpreted as measures, beats, beat divisions,
or beat subdivisions.
Notwithstanding that the above definitions do not acknowledge the difference between
pulses and beats – and that the words “pulses” and “beats” are often used interchangeably –
some scholars offer a refined distinction of one from the other, which may help achieve termi-
nological precision. Karpinski (2000) explains that pulses are regularly recurring points in time,

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

Cha 565

while beats are durations between two adjacent pulses. Kramer (1988) makes a similar distinc-
tion, using a different set of terms: “timepoints” and “timespans” (p. 82). The relation between
pulse and beat in a temporal dimension is similar to that of point to line in a two-dimensional
Euclidean space. Pulses cannot be divided into smaller units since, as such, they have no dura-
tions. Beats, on the other hand, can be divided into smaller units such as beat divisions or can
be combined into larger units such as measures. From this, the beat might be redefined as the
time span between two adjacent pulses in a temporal continuum of isochronous points, which has the
potential to be divided into twos or threes and to be grouped into units of two or three.

Psychological, evolutional, and neurological foundations

of the beat
The definition of the beat given above is still somewhat abstract, omitting a crucial element of
analysis: the perceptual aspect of the beat. Just as pitch perception is enabled only for a certain
frequency range, beat perception is limited to durations ranging from approximately 200 ms
(300 beats per minute; bpm) to 2,000 ms (30 bpm) (London, 2004, p. 31).4 Durations shorter
than the lower limit of the spectrum are perceived as an undifferentiated event; durations
longer than the upper limit are perceived as a series of unrelated events. Taking into account
these lower and upper limits of beat perception, I suggest the following revised definition of a
beat: the time span between two adjacent pulses in a temporal continuum of isochronous points, which
in general has the potential to be divided into twos or threes and to be grouped into patterns of two or
three, and which falls within a range from approximately 200 ms to 2,000 ms.
Worthy of attention in this context is that the duration of 600 ms, or the tempo of 100 bpm,
carries special perceptual weight. Fraisse (1982) claims that the tempo of 100 bpm represents
both spontaneous tempo, the speed of the beat considered by subjects to be the most natural, and
preferred tempo, the tempo judged neither too slow nor too fast (p. 153; London, 2004, p. 31). In
spite of individual differences in these tempos, subjects tend to tap at around 600 ms on aver-
age, without recorded musical stimuli. Prominence of the 600-ms span is closely associated
with the rate at which people move in general and with the average walking tempo of humans
in particular. Interestingly, Tan, Pfordresher, and Harré (2010) report that differences in walk-
ing rate are paralleled by differences in tapping tempo. According to Drake, Jones, and Baruch
(2000), the spontaneous tapping rate of the older participants in their study of rhythmic
attending is, on average, slower than that of the younger ones: “Whereas the adults tapped on
average at about every 600 ms, confirming Fraisse’s predictions concerning an optimal tempo
range, children tapped considerably faster, with 4- and 6-year-olds tapping every 400 ms” (pp.
265–266). London (2004) interprets this tempo change with age as a reflection in the chang-
ing anthropometry during different developmental stages.
The connection of music and bodily movement is psychophysiologically fundamental
(Burger, Saarikallio, Luck, Thompson, & Toiviainen, 2013; Lahav, Saltzman, & Schlaug, 2007;
Sedlmeier, Weigelt, & Walther, 2011) and might have existed as long as music itself. Honing
(2012) presumes that the ability to perceive the steady beats and to use them to synchronize
movements would have played a pivotal role in the origins of music. In fact, multiple theories for
the origins and evolution of music point to a meaningful relationship between music and move-
ment. According to Darwin, music arose from the evolutionary process of sexual selection:
superior coordination and physical capability that underlay superior musical performance
were regarded as sexually attractive traits (Darwin, 1871; Miller, 1999). Another theory based
on survival benefits – rather than sexual selection – suggests that music emerged as a means of
training coordinated movement: performance of rhythmic elements of music could engage and

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

566 Psychology of Music 43(4)

hone skills of controlled movement and thus benefit important tasks such as hunting and herd-
ing (Thompson, 2009). The interaction of music and movement, then, could have contributed
significantly to the history of humans as a species.
While many discussions of the origins of music are overwhelmingly speculative, some studies
do address how music and movement interact in the brain (Chen, Penhune, & Zatorre, 2008,
2009; Grahn, 2009; Grahn & Brett, 2007). Additionally, significant evidence points to the role of
the beat in the relationship of music and movement at the neuroscientific level. It is commonly
acknowledged that spontaneous synchronized movement stems from the presence and percep-
tion of beats. For example, Grahn and Brett (2007) report that, for both musicians and non-musi-
cians, the accuracy of rhythm reproduction is improved when rhythmic sequences have regular,
recurring accents, and these beat-based rhythms (“metric simple” in their terms) activate the pre-
supplementary motor area (pre-SMA), the supplementary motor area (SMA), and the basal gan-
glia – brain areas known to be involved in movement production and coordination.
Chen et al. (2008) further illustrate inherent neurological enmeshments between music and
movement. The researchers conducted two fMRI (functional magnetic resonance imaging)
studies as participants tapped along with musical rhythms: subjects in the first experiment
knew in advance that they would have to synchronize with the tapping; subjects in the second
experiment knew that they would be hearing tapping but did not know that they would be
required to tap, so that no motor planning or rehearsal was possible. Despite this incongruity,
brain regions involved in motor activity such as the SMA, the mid-premotor cortex (mid-PMC),
and the cerebellum were activated in both groups of participants. Rhythmic stimuli appear to
trigger these regions automatically. To supplement their findings, the authors observe that
rhythm sneaks into our everyday lives independent of our exposure to music: “Many of our
daily actions such as walking are not only executed in a rhythmical manner but also generate
sounds that highlight the progression of events” (Chen et al., 2008, pp. 2851–2852).
The aforementioned studies by Grahn and Brett (2007) and by Chen et al. (2008) put forth a
persuasive case for the existence of a certain relationship between musical beats and movement
at the neurobiological level, a meaningful association that stands behind, and provides addi-
tional value to, the easily observable connections between the two. It is these interactions that
can explain why tapping one’s finger and bobbing one’s head to the beat is so natural, why walk-
ing in step with a beat seems to be an almost unconscious process, and why it is so difficult to
dance in the absence of musical accompaniment. The cardinal beauty of Takadimi is that it is
based on a phenomenon deep-rooted both in human history and in the human brain – the beat.

Takadimi, a beat-oriented system of rhythm pedagogy

The Takadimi system concerns onsets (or “attack points” in Hoffman et al.’s 1996 terminology) of
sounds at the beat, beat-division, and beat-subdivision levels, employing separate sets of sylla-
bles for simple meter and compound meter. Irrespective of the meter type, the onset of the beat
is labeled as “ta.” In simple meter, the onsets of the two beat divisions are called “ta” and “di,” and
the onsets of the four beat subdivisions are “ta,” “ka,” “di,” and “mi.” In compound meter, the
three beat divisions are intoned as “ta,” “ki,” and “da,” and the onsets of the six beat subdivisions
are “ta,” “va,” “ki,” “di,” “da,” and “ma.” Figure 1, adapted from Figure 1 in Ester et al. (2006),
shows how Takadimi syllables work at and within the beat in simple and compound meters.
Since the syllables “ta di” denote two equal divisions of a beat, they represent not only regu-
lar beat divisions in simple meter, but also simple beat divisions in compound meter (to be writ-
ten as duplets). The syllables “ta ki da,” normally used for regular beat divisions in compound
meter, are used for compound beat divisions in simple meter (triplets). Similarly, the syllables
“ta ka di mi” work both for regular beat subdivisions in simple meter and for quadruplets in
Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016
Cha 567

Figure 1.The Takadimi syllables at the beat, beat-division, and beat-subdivision levels in simple and
compound meters.

Figure 2.  The Takadimi syllables for two, three, four, and six divisions of a beat in simple and compound
meters.

Figure 3.  The Takadimi syllables for three rhythmic figures that consist of three uneven durations within
the beat in a simple meter.

compound meter; the syllables “ta va ki di da ma” are used both for regular beat subdivisions
in compound meter and for sextuplets in simple meter. Figure 2 illustrates syllables for two,
three, four, and six equal divisions of a beat in simple and compound meters.
Syllables in this system are assigned according to the “discrete one-to-one mapping onto
metric attack points” (Hoffman et al., 1996, p. 16) and therefore the number of syllables must
match the number of attack points, or onsets. For example, when there are three uneven durations
within the beat in a simple meter, three conceivable permutations exist in each of which there
must be three syllables: “ta di mi,” “ta ka mi,” and “ta ka di” (see Figure 3). Figure 4, adapted from
Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016
568 Psychology of Music 43(4)

Figure 4.  A rhythm pattern that contains tied-over beats (marked with *) and rests (+).

Figure 9 in Hoffman et al. (1996), shows 17 onsets and 17 matching syllables. Notably, beats
tied over from the previous sound (marked with * in Figure 4) and rests (+) do not receive any
syllable inasmuch as they are not onsets – a crucial point I will return to later. Owing to the fact
that tied-over beats and rests are unsyllabled, it is impossible to accurately deduce the number
of beats by examining the syllables alone. For instance, there are eight beats in Figure 4, but
“ta,” the syllable that represents a beat, appears only twice. Takadimi is a beat-based system – in
the sense that it assigns the syllable “ta” to each beat irrespective of the note value – that, some-
what ironically, downplays beats in deference to onsets.

Psychological reality of the implicit beat

Barnes and Jones (2000) conducted an experiment that might shed light on the psychological
nature and significance of what the Takadimi system fails to appreciate: the implicit beat. They
had participants listen to what they called “induction IOIs” (I, consisting of a sequence of seven
isochronous interonset intervals equal to 600 ms) followed by one “standard IOI” (S; 600 ms,
600 ± 21 ms, or 600 ± 76 ms, i.e., slightly longer or shorter than 600 ms) and one “compari-
son IOI” (C; standard IOI ± ΔT, where the ratio of ΔT/S is 0.12, i.e., slightly longer or shorter
than S) (pp. 259–260). The stimulus sequence used in their experiment is shown in Figure 5,
adapted from Figure 1 in Barnes and Jones (2000). Subjects were asked to judge the duration of
a comparison IOI as shorter than, the same as, or longer than the preceding standard IOI, with
the instruction to disregard induction IOIs. The authors found that accuracy in estimating the
comparison IOI was greatest when the standard IOI was 600 ms and least when it was 600 ±
76 ms; in other words, participants judged S and C pairs far more accurately when S matched I
than when S departed from I. These findings suggest that the activity of induction IOIs entrained
– and was assimilated by – the listeners. The psychological presence of these implicit IOIs, not
explicitly present but implied by the initial sequence of IOIs,5 prevailed over the subsequent
standard and comparison IOIs. In spite of the instruction to ignore them, they strongly influ-
enced participants’ estimation, resulting in a biased judgment of the relative difference between
S and I.
A series of studies by McAuley and colleagues (Jones & McAuley, 2005; McAuley, Frater,
Janke, & Miller, 2006; McAuley & Jones, 2003; Miller & McAuley, 2005) have harnessed experi-
ments similar to the one by Barnes and Jones (2000) in design, albeit with different purposes.
In an experiment on the perception of tempo changes, McAuley et al. (2006) presented partici-
pants with three different sequences of tones and asked them to decide whether each sequence
was accelerating or decelerating. The three types of sequences are shown in Figure 6, adapted
from Figure 1 in McAuley et al. (2006). Sequence #1 is a five-tone sequence consisting of three
600-ms intervals followed by a slightly shortened, 500-ms final interval. Sequence #2 is a
four-tone sequence consisting of a 600-ms interval followed by a 1200-ms interval and a

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

Cha 569

Figure 5.  The stimulus sequence used in one of the experiments conducted by Barnes and Jones (2000).

Figure 6.  The three sequences of tones used in one of the experiments conducted by McAuley et al.
(2006).

variable final interval of 600 ms ± ΔT, where ΔT = 2%, 6%, 10%, or 14% of 600 ms. Sequence
#3 is a five-tone sequence consisting of a pair of 300-ms intervals followed by a 1200-ms
interval and a variable final interval of 600 ms ± ΔT.
The second and third types of sequences are of concern to us. The second, four-tone sequence
yielded few individual differences among participants. Most of them felt that the sequence was
speeding up when the final interval was slightly shorter than 600 ms, and that the sequence
was slowing down when the final interval was slightly longer than 600 ms. This result verifies
that the subjects took the initial 600-ms interval as the beat and compared it to the final inter-
val, and that they interpolated a missing beat between the second and third tones.
For the third, five-tone sequence, participants’ responses were more complicated. Most of
them sensed that the sequence was slowing down when the final interval was in fact slightly
longer than 600 ms. The five-tone sequence with the final interval of 600 − ΔT, however, elic-
ited two opposite perceptions, indicating pronounced individual differences: some participants
responded that the sequence was slowing down, while others judged the identical stimulus to
be speeding up. Participants who returned a verdict of “slowing down” appear to have detected

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

570 Psychology of Music 43(4)

Figure 7.  A rhythm pattern with a tied-over beat (*) and a rest (+).

a 300-ms interval (derived from the first two 300-ms interval of the sequence) as the beat,
which means that they interpolated three missing beats between the third and fourth tone of
the sequence; participants who passed the “speeding up” response must have heard a 600-ms
interval as the beat, interpolating one missing beat between the third and fourth tone of the
sequence (pp. 568–569). In both types of participants, we find the exciting conclusion that
humans can and do perceive beats even when they are not physically present.6 An implicit beat
– “implicit interval [perceived] as the beat” (p. 569) – is induced by means of mental interpola-
tion. Just as important as these fascinating results are the implications for enhancement of the
Takadimi system.

Some proposals for improvement of the Takadimi system

In order for a rhythm solmization system to be a compelling method for dictation (sound-to-
symbol), it is required that, for the given beat value, a sequence of syllables (s-sequence hereafter)
representing a specific rhythm pattern be mapped to exactly one unique sequence of notes
(n-sequence). Namely, an s-sequence may not be paired with more than one n-sequence. For
example, the s-sequence “ta di ta di” in meter must have a unique n-sequence “ ” and noth-
ing else; if the same s-sequence were allowed to stand for two or more different n-sequences,
then that s-sequence would be open to more than one interpretation, making it impossible to
decide unambiguously which n-sequence is the correct one. For a rhythm solmization system
for sight singing/reading (symbol-to-sound), the order of the mapping process must be reversed:
for the given beat value, an n-sequence must be paired with exactly one unique s-sequence that
represents a specific rhythm pattern. As Takadimi has been developed as a system effective for
teaching both rhythm dictation and rhythm singing/reading, an s-sequence must be mapped to
exactly one unique n-sequence, and vice versa; namely, there must be a bijection, or one-to-one
correspondence, between n-sequences and s-sequences (and, of course, a one-to-one correspond-
ence between s-sequences and specific rhythm patterns).
As Figure 4 shows, in Takadimi, an onset-oriented system, syllables are assigned neither to
tied-over beats nor to rests: in Figure 7, reproduced from Figure 8 in Hoffman et al. (1996),
the second beat of m. 1 (a tied-over beat; *) and the second beat of m. 4 (a rest; +) are unsyl-
labled. The problem here is that a single s-sequence can represent two or more different
n-sequences in the given meter. The s-sequence “ta ta ta di” works not only for the first two
measures of Figure 7 but also for other n-sequences such as those illustrated in Figure 8.
Similarly, the s-sequence “ta ka mi di mi ta” represents the n-sequence in mm. 4–5 of Figure 7, but
the same s-sequence can embody quite different n-sequences (see Figure 9). The current
Takadimi system tolerates a single s-sequence being mapped to more than one n-sequence and
is therefore unable to function as a dependable method for dictation. Similarly, it is impossible to
determine precisely which rhythm pattern is referred to by an s-sequence as the s-sequence
represents more than one rhythm pattern – hence Takadimi’s unsuitability to serve as a reliable
system for sight singing/reading.

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

Cha 571

Figure 8(a) and 8(b).  An example of two different n-sequences that share the same s-sequence.

Figure 9(a) and 9(b).  Another example of two different n-sequences that share the same s-sequence.

I propose a modified Takadimi system in order to eliminate this deficiency. Taking explicit
cognizance of implicit beats – which are, after all, psychologically real, as implicated by Barnes
and Jones (2000) and McAuley et al. (2006) among others – the modified Takadimi is con-
ceived to be a strongly beat-based system in which a syllable is assigned to each and every beat,
whether explicit or implicit, as opposed to a weakly beat-based system that overlooks implicit
beats and maintains the privilege for explicit beats.
Implicit beats can be divided into two types: 1) a tied-over beat and 2) a rest on the onset of a
beat. A beat tied over from the previous beat, beat division, or beat subdivision gets the syllable
“–ah.”7 The syllabifying of mm. 1–2 of Figure 7, Figure 8(a), and Figure 8(b) in the modified
Takadimi system is shown in Figure 10, with a comparison to the original system. While in the
original Takadimi system the same s-sequence can translate as multiple n-sequences, in the
new system a one-to-one correspondence between an s-sequence and an n-sequence is assured,
as shown in Figure 11.
A rest that falls on the onset of a beat is labeled with the syllable “(ta).” I apply this rule to
rests at both the beat-division and beat-subdivision levels; that is, besides “(ta),” syllables “(ka),”
Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016
572 Psychology of Music 43(4)

Figure 10(a), 10(b), and 10(c).  The syllables in the modified Takadimi system for mm. 1-2 of Figure 7,
Figure 8(a), and Figure 8(b), with a comparison to the original system.

“(di),” “(mi),” “(va),” “(ki),” “(di),” “(da),” and “(ma)” are possible, and, when these parenthe-
sized syllables are vocally verbalized, the phoneme “sh” [∫] will be added to the end of the syl-
lables that correspond to their attack points and implicit beats: “tash,” “kash,” “dish,” “mish,”
“tash-ahsh,” and so on. The s-sequences in both the modified and original systems for mm. 4–5
of Figure 7, Figure 9(a), and Figure 9(b) are shown in Figure 12.
S-sequences in the original Takadimi system can show us the number of onsets, but not the
number of beats they contain. S-sequences in the modified system enable us to reason both the
number of beats and the number of onsets they contain: the number of beats is determined by
counting all occurrences of “ta” (explicit beats), “(ta)” (rests on the beat), and “–ah” (tied-over
beats); the number of onsets should equal the number of all syllables minus the number of
tied-over beats and syllables in parentheses. The s-sequence of Figure 12(b), “ta ka mi–ah di mi
(ta) ta,” indicates that it has four beats (“ta,” “–ah,” “(ta),” and “ta”) and six onsets (“ta,” “ka,”
“mi,” “di,” “mi,” and “ta”).
A four-page “Takadimi Short Guide” by Hoffman (2009) includes examples of dotted rhythm
patterns, which are lacking in Hoffman et al.’s (1996) article. According to Hoffman (2009),
the n-sequence “ ” – provided the beat value is a quarter note – is mapped to the s-sequence
“ta di” (p. i, mm. 2-3 of the example on the bottom). It must be clear by now that this is a dubi-
ous practice: in a system that is apathetic about implicit beats, the s-sequence “ta di” can repre-
sent not one but numerous n-sequences such as “ ,” “ ,” and “ .” Houlahan and Tacka
(2011) attempt to fix this problem by ascribing three different s-sequences to the three different
n-sequences: “ ” is labeled as “ta di,” “ ” as “ta–di,” and “ ” as “ta di–” (pp. 16, 83–88). We
should note that they use the dashes purely for orthographic purposes, without indicating how

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

Cha 573

Figure 11.  A one-to-one correspondence between an s-sequence and an n-sequence in the modified
system, with a comparison to the original system.

Figure 12(a), 12(b), and 12(c).  The syllables in the modified Takadimi system for mm. 4-5 of Figure 7,
Figure 9(a), and Figure 9(b), with a comparison to the original system.

to articulate them vocally, meaning that, while these new s-sequences have a certain degree of
differentiability when they are written, they are indistinguishable from one another when ver-
bally declaimed. The modified system I propose consistently applies “–ah” to implicit beats,
resulting in “ta–ah di” and “ta di–ah” for “ ” and “ ” respectively.8 The n-sequence “ ” in a
meter that has a quarter note as the beat value, a typical syncopated rhythm pattern, would be

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

574 Psychology of Music 43(4)

syllabified as “ta di–ah di” in this proposed, modified system, instead of “ta di–di” (Houlahan and
Tacka, 2011, pp. 89–92) or “ta di di” (Hoffman et al., 1996, p. 26, m. 3 of Figure 16; Hoffman,
2009, p. i, m. 7 of the example at the bottom & p. iii, m. 2 of the example at the top).

Conclusion
The Takadimi system is a tool of outstanding merit for rhythm instruction, anchored (whether
intended or not) to psychological and neurobiological bases supporting the primacy of beats
in rhythm perception and reproduction.9 It allows a limited number of common rhythm pat-
terns to be learned as aural patterns reinforced by syllables at the beat, beat-division, and
beat-subdivision levels. Depending upon whether the task is rhythm dictation or rhythm
singing, those syllables that represent prevalent sound patterns can be translated into either
symbols or sounds. In the words of Hoffman et al. (1996), “through application of the sys-
tem, rhythm emerges as specific syllabified patterns that can be spoken consistently and
accurately within the span of a beat, regardless of prevailing metric divisions” (p. 16). This
very sentence, however, reveals Takadimi’s own limitations: sound patterns can be intoned
“consistently and accurately” as far as the duration of each tone is “within the span of a
beat.” In the original, onset-oriented Takadimi system, any s-sequence containing a duration
longer than a beat is doomed to multiple n-sequences; syllables can translate neither to
sounds nor to symbols consistently and accurately. In sum, Hoffman et al.’s ways of using
syllables for tied-over beats and rests obscure their otherwise laudable rhythm solmization
system. The modified Takadimi system that I propose treats all beats and rests as a rhythmic
entity by assigning the syllable “–ah” and the regular syllable in parentheses to tied-over
beats and rests respectively, which guarantees a one-to-one correspondence between
s-sequences and n-sequences.
One direction for future research would be to enhance the Takadimi system further for irreg-
ular divisions and complex meters. Hoffman et al. (1996) hold that the original Takadimi sys-
tem is capable of accommodating complex rhythms such as “irregular divisions” and
“asymmetric meters”10 (pp. 14–15, 19–20; Hoffman, 2009, p. iii). However, the repurposing of
syllables Hoffman et al. suggest for irregular divisions (e.g., “ta ka di mi ti” for quintuplets such
as “ ” in ) and complex meters (“ta di ta ki da” for “ ” in [ ]) violates the principle of
one-to-one correspondence between s-sequences and n-sequences, since, in either case, the
same s-sequence would be shared by more than one n-sequence (“ta ka di mi” for both “ ” and
“ ” in ; “ta di ta ki da” for both “ ” in and “ ” in [ ]). One might devise logically
sound solutions to problems such as irregular divisions and complex meters by introducing
new syllables.11 The risk, however, is that asking students to learn a number of new syllables
and to develop fluency in their application in real time would impose a serious cognitive burden.
Due to the relative rarity of irregular divisions and complex meters in the common-practice
musical literature and the excessive complexity of introducing even more syllables, it might be
best to abandon rhythm syllables altogether at this point in the learning sequence. Nevertheless,
I would not completely shut my musical mind to the possibility of a theoretically solid and prac-
tically efficacious rhythm syllable system that accommodates those advanced rhythms with
irrefutable perspicuity, which has yet to appear.

Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-
profit sectors.

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

Cha 575

Notes
  1. Takadimi originates in North Indian rhythm training. For the Indian origins of the system and its
earlier appearances in the West, see Hoffman, Pelto, and White (1996), p. 14.
  2. It is not a coincidence that Houlahan and Tacka (2011), ardent advocates of an approach that
“explores concepts . . . through the sound of music before moving to an explanation of how symbols
are used to represent the sound” (p. xviii), have included the Takadimi system in their music funda-
mentals textbook. Its title is emblematic of their philosophy: From sound to symbol.
  3. While it is commonplace to consider that beats can be combined into units of four (as well as into
units of two or three), units of four result from pairs of two. In other words, the most basic divisions
and combinations of beats are only twos and threes. See Karpinski (2000) and Lester (1986).
  4. A duration range for beat perception depends, to certain extent, upon the cognitive dispositions,
attentions, and body measurements of participants in the research. For slightly different duration
ranges, see Fraisse (1982); Parncutt (1994); Tan, Pfordresher, and Harré (2010); and Thompson
(2009).
  5. One may suspect that we cannot know whether subjects thought of the 600-ms IOI as a beat. It is tell-
ing, however, that Barnes and Jones used 600-ms intervals (or 100 bpm), which represent spontaneous
and preferred tempo, as induction IOIs in their experiment. In addition, McAuley and Jones (2003)
specified that the intervals they were dealing with (which they called “short intervals”) were “time
intervals between approximately 200 and 2,000 ms” (p. 1102), a range exactly corresponding to the
lower and upper limits of beat perception. Thus, it seems reasonable to interpret these IOIs as beats.
  6. See Grahn and McAuley (2009) for a neuroscientific follow-up study suggesting that, despite indi-
vidual differences, people in general, and “strong beat-perceivers” (p. 1897) in particular, recognize
implicit beats without much mental effort.
  7. Hoffman et al. (1996) largely ignore implicit beats. They attend tied-over beats only when they are
dealing with the “complex syncopations” for which they use syllables in parentheses (Hoffman et al.,
1996, p. 20, mm. 3–5 of Figure 10). There is only one case in which they assign a syllable to a rest
that falls on the onset of a beat (Hoffman et al., 1996, p. 20, m. 5 of Figure 10). Since both a tied-
over beat and a rest on a beat receive the same syllable “(ta)” in their example, there is no way to
fathom whether “(ta)” refers to a tied-over beat or a rest on a beat. At any rate, assigning a syllable
to tied-over beats and rests, which must be considered as an exception in their system, contradicts
their principle of “discrete one-to-one mapping onto metric attack points” (Hoffman et al., 1996, p.
16). Houlahan and Tacka (2011) utilize a method similar to the one I am proposing here: “When
we hear one sound over two beats, we label it with the rhythm syllable ta–ah. When we hear one
sound over four beats, we label it with the rhythm syllable ta–ah–ah–ah” (p. 16). However, they do not
consistently honor this principle of ascribing “–ah” to tied-over beats. See the paragraph discussing
Houlahan and Tacka’s treatment of dotted rhythm patterns below.
  8. It is unfortunate that Houlahan and Tacka forgo the practice of assigning “–ah” to implicit beats.
  9. For the importance of a beat-based rhythm pedagogy, see Ester et al. (2006).
10. Hoffman et al.’s “asymmetric meters” are actually what should be called complex meters – meters that
have beats of unequal lengths within a measure.
11. For example, we could propose “ta” plus four new syllables for quintuplets. Gordon (1993, pp. 284–
285) uses two different sets of syllables, “du de du da di” and “du be du ba bi,” to distinguish between “
” in and “ ” in [ ] (see also Gordon, 2009, p. 62).

References
Barnes, R., & Jones, M. R. (2000). Expectancy, attention, and time. Cognitive Psychology, 41, 254–311.
Burger, B., Saarikallio, S., Luck, G., Thompson, M. R., & Toiviainen, P. (2013). Relationships between
perceived emotions in music and music-induced movement. Music Perception: An Interdisciplinary
Journal, 30(5), 517–533.
Chen, J. L., Penhune, V. B., & Zatorre, R. J. (2008). Listening to musical rhythms recruits motor regions of
the brain. Cerebral Cortex, 18, 2844–2854.

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

576 Psychology of Music 43(4)

Chen, J. L., Penhune, V. B., & Zatorre, R. J. (2009). The role of auditory and premotor cortex in sensorimo-
tor transformations. Annals of the New York Academy of Sciences, 1169 [The Neurosciences and Music
III – Disorders and Plasticity], 15–34.
Darwin, C. (1871). The descent of man, and selection in relation to sex. London, UK: John Murray.
Drake, C., Jones, M. R., & Baruch, C. (2000). The development of rhythmic attending in auditory
sequences: Attunement, referent period, focal attending. Cognition, 77(3), 251–288.
Ester, D. P., Scheib, J. W., & Inks, K. J. (2006). Takadimi: A rhythm system for all ages. Music Educators
Journal, 93(2), 60–65.
Fraisse, P. (1982). Rhythm and tempo. In D. Deutsch (Ed.), The psychology of music (pp. 149–180). New
York, NY: Academic Press.
Gauldin, R. (2004). Harmonic practice in tonal music (2nd ed.). New York, NY: W. W. Norton.
Gordon, E. E. (1993). Learning sequences in music. Chicago, IL: GIA Publications.
Gordon, E. E. (2009). Taking a reasonable and honest look at tonal solfege and rhythm solfege. Chicago, IL: GIA
Publications.
Grahn, J. A. (2009). The role of the basal ganglia in beat perception: Neuroimaging and neuropsychologi-
cal investigations. Annals of the New York Academy of Sciences 1169 [The Neurosciences and Music III
– Disorders and Plasticity], 35–45.
Grahn, J. A., & McAuley, J. D. (2009). Neural basis of individual differences in beat perception. NeuroImage,
47, 1894–1903.
Grahn, J. A., & Brett, M. (2007). Rhythm and beat perception in motor areas of the brain. Journal of
Cognitive Neuroscience, 19, 893–906.
Hoffman, R. (2009). Learning rhythm with the Takadimi system of rhythm solfege. Retrieved from http://
www.takadimi.net/shortGuide.html
Hoffman, R., Pelto, W., & White, J. W. (1996). Takadimi: A beat-oriented system of rhythm pedagogy.
Journal of Music Theory Pedagogy, 10, 7–30.
Honing, H. (2012). Structure and interpretation of rhythm in music. In D. Deutsch (Ed.), The psychology
of music (3rd ed., pp. 369–404). New York, NY: Academic Press.
Houlahan, M., & Tacka, P. (2011). From sound to symbol: Fundamentals of music (2nd ed.). New York, NY:
Oxford University Press.
Jones, M. R., & McAuley, J. D. (2005). Time judgments in global temporal contexts. Perception &
Psychophysics, 67(3), 398–417.
Karpinski, G. S. (2000). Aural skills acquisition: The development of listening, reading, and performing skills in
college-level musician. New York, NY: Oxford University Press.
Kennedy, M. (1994). The Oxford dictionary of music (2nd ed.). New York, NY: Oxford University Press.
Kramer, J. (1988). The time of music: New meanings, new temporalities, new listening strategies. New York,
NY: Schirmer Books.
Lahav, A., Saltzman, E., & Schlaug, G. (2007). Action representation of sound: Audiomotor recognition
network while listening to newly acquired actions. Journal of Neuroscience, 27(2), 308–314.
Lester, J. (1986). The rhythms of tonal music. Hillsdale, NY: Pendragon Press.
London, J. (2004). Hearing in time: Psychological aspects of musical meter. New York, NY: Oxford University
Press.
McAuley, J. D., Frater, D., Janke, K., & Miller, N. S. (2006). Detecting changes in timing: Evidence for two
modes of listening. The Proceedings of the 9th International Conference on Music Perception and
Cognition, 566–573.
McAuley, J. D., & Jones, M. R. (2003). Modeling effects of rhythmic context on perceived duration: A
comparison of interval and entrainment approaches to short-interval timing. Journal of Experimental
Psychology: Human Perception and Performance, 29(6), 1102–1125.
Miller, G. (1999). Evolution of human music through sexual selection. In N. L. Wallin, B. Merker, & S.
Brown (Eds.), The origins of music (pp. 329–360). Cambridge, MA: The MIT Press.
Miller, N. S., & McAuley, J. D. (2005). Tempo sensitivity in isochronous tone sequences: The multiple-look
model revisited. Perception & Psychophysics, 67(7), 1150–1160.

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016

Cha 577

Parncutt, R. (1994). A perceptual model of pulse salience and metrical accent in musical rhythms. Music
Perception: An Interdisciplinary Journal, 11(4), 409–464.
Sadie, S. (Ed.). (2001). The new Grove dictionary of music and musicians (2nd ed., Vol. 3). London, UK:
Macmillan.
Sedlmeier, P., Weigelt, O., & Walther, E. (2011). Music is in the muscle: How embodied cognition may
influence music preferences. Music Perception: An Interdisciplinary Journal, 28(3), 297–306.
Tan, S.-L., Pfordresher, P., & Harré, R. (2010). Psychology of music: From sound to significance. New York,
NY: Psychology Press.
Thompson, W. F. (2009). Music, thought, and feeling: Understanding the psychology of music. New York, NY:
Oxford University Press.

Downloaded from pom.sagepub.com at LOYOLA UNIVERSITY CHICAGO on April 19, 2016