Training preschoolers to develop absolute musical pitch.

Originally published in Psychology, 2, 90-98, 1984.

S.G. Grebelnik

Translated by Christopher Aruffo

One of the most important aspects of personal development is the aesthetic self, in particular the development of musicianship.  To successfully meet the challenges of music education, one must know the psychological patterns of musical abilities.  Among the least explored of musical talents is absolute pitch.  In music psychology, absolute pitch is the ability to hear (passive AP) or produce (active AP) the pitch of musical sounds with comparison to a known reference tone. [6], [29].

Historically, there has been considerable disagreement on the nature and origin of the ability.  Some researchers believe that the ability is physiological [32], [46]; others consider the skill a kind of musical ability [29], [48].  The latter point of view is further divided in assessing the nature of AP; some have argued that it is a skill of vocal production [14], [43]; others saw it as a purely aural skill [15], [29], [35].  This discussion has centered around the question of absolute judgment; there is still debate about whether it is learned or appears genetically.  Some researchers believe it is a product of education [10], [24]; others think it is an inherent skill [41].  B.M. Teplov felt that absolute hearing, as with all other musical abilities, is acquired through musical practice, but only by those possessed of a genetic predisposition.  But to this day, the question remains:  exactly what advantage would a "genetic predisposition" provide?

In music, absolute hearing can be very important.  By allowing its possessor precise analysis of musical structure, absolute pitch makes possible the perception of complex sound [25], and aids in a refined perception of music, which is a prerequisite for meaningful musical experiences [29] and good musical taste.  In memory of musical sound, other things being equal, people with absolute pitch are 8-9 times more able to remember songs than those without [16].  This, of course, helps to quickly learn new music and expand one's repertoire.  Absolute pitch is of particular importance in composition.  While absolute pitch may not be necessary for the process of creating a piece of music, there is reason to believe that having AP is desirable for a professional composer [17].

Researching absolute pitch has, therefore, a certain practical significance.  As for its theoretical value, a psychological study of absolute hearing, with respect to its initial appearance, general characteristics, and development over time, might finally help dispel the myth of musical ability as a genetic endowment.

It has not yet been conclusively demonstrated that absolute hearing is a psychological ability.  While on the one hand, some researchers agree that it is an issue of mnemonic memory [23], [37], on the other hand there is data which suggests the ability is dependent on physiological senses [29], [47].  As formation of absolute pitch is connected with perception of tones during musical activity and simultaneously storing those tones in memory, our study approaches absolute pitch as a musical ability based on sense-memory, and we have tried to develop AP in preschool-age children who do not currently possess it.

The research shows three basic theories of learning absolute pitch.  1)  the natural development of an ability in an activity specific to its occurrence [12], [27]; 2) a physical ability as a result of specific actions [3], and 3) involuntary memory storage based on natural ability to recognize musical tones [8].

A.N. Leontev has written of the existence of two types of musical ability.  The first kind is a natural, innate ability, emerging from the human evolutionary response to sound, which serves as the biological precondition for training the second kind of ability which is learned skill.  The first type of ability, a purely physical response, is inborn; we know that humans can recognize differences in sound height immediately after birth [2].  However, a child will only be able to identify tones at approximately age three, after they have had some musical experience [29], [48].  In other words, absolute pitch ability requires some musical listening experience, and therefore is more closely related to the second kind of ability.

Therefore, unlike all other experimental researchers who have attempted to train this capability, we have used as our training materials not random and unmusical sounds, but melodies and musical texts.  The basic hypothesis of our research is the assumption that absolute pitch is a special kind of musical perception that is learned, and that the basis of its development is actively directed perception, recognition, memory, and labeling of musical sounds.

In addition, we have presumed that absolute pitch is formed by learning particular musical structures which become the basis for involuntary memory storage.  We have noticed that there are commonalities between absolute and relative pitch [39], [40]; attributes are encoded and identified [6], reference sounds can be remembered and recognized [5], and sounds are most easily perceived in the middle register [45].  We have further assumed that a musical ear develops from hearing structures to components, that is, relative perception develops before absolute.

It is known that passive and active AP do not always appear simultaneously; typically, passive AP is the first to manifest [29], [42].  B.M. Teplov wrote that absolute pitch undergoes a process of acquisition.  Therefore, to study the genesis of absolute hearing, it is necessary to understand the identification mechanism, which tends to compare perceived objects to benchmarks stored in memory [34].  Let us look briefly at how to analyze the perception of musical structures and the memory of melodic standards which we have used in our study.

It has been experimentally observed that for possessors of strong absolute pitch ability, their identification categories match the 12 pitches (octave) of the equal-tempered musical tuning system, such as is found on a piano [44].  It is also known that these categories are identified based on their significant distinguishing characteristics [21], [34].  The ratio of overtones remains the same for any isolated sound in a constant timbre; therefore, overtones cannot be considered useful attributes in identifying absolute pitch.  More likely, the attributes which enable the memory of benchmarks are found in the combination of equal-tempered musical sounds.  The selected identification tone-- the standard-- must first be associated with a second tone, creating a relationship, and to prevent transposition this second tone should also remain constant.

It is thereby possible to make absolute identifications from simple intervals, i.e. from a minor second up to a tritone [26], [48].  Using a standard tone we can build simple intervals both upwards and downwards.  However, a standard tone's relationships to some tones are identical (for example, a perfect fourth from sol to do and a perfect fourth from do to fa, etc)-- so for complete distinction between identifiable categories two attributes are insufficient.  It is thereby necessary to assume that, to identify each category, three attributes are combined; the three attributes form a "section" of a functional system of stimulus generalization [11] which coincides with a limit (3 objects) of absolute memory [16] and of simultaneous identification [20].

The third part of this combination, which we call a contour, is a characteristic separate from the constant interval measured in tones upwards or downwards.  Contour is an essential attribute of melody [4] whose elements are synthesized when listening [13] to be perceived as one attribute of a composition [31].  Identification of this complete entity, the components of which are relations between pitches, occurs without learning to compare sounds to each other [30] but as a recognition of an entire piece.

We know that it is possible to recognize objects without detecting all of their characteristics.  It is enough to identify some or even one of them [9], [11].  Often, identification can be made from a single dominant trait.  The dominant attribute-- the basis for association and generalization of objects-- in a musical work is its tonality; consequently, the desired standard tone would be the tonic.  But in practice, there are 24 musical tonalities, and only 12 identification categories, so the question arises:  what tonality should be used?

As we know, some absolute listeners are best able (or solely able) to recognize the white piano keys, while others may identify black or white keys with equal proficiency [29], [48].  Therefore, the two systems of identification (heptatonic white keys and pentatonic black keys) are linked together.  If recognizing the sound of any one white key does not interfere with recognizing another white key, then the tonalities of these keys are functionally related; the same applies to the sounds of black keys.

The standard tone C is a near relative of the tonalities a, G, e, and d.  For black keys, the standard tone should be f (which also has a primary relationship to C) which provides A-sharp, E-sharp, D-flat, and b-flat.  According to rules of tonality, the central tone may be displaced up or down by octaves and yet preserve its tonic function [28].  Having named these standard tones and their "seconds", we find that C and f are also linked to b and G-flat (which are closer to the standard tones than are B and F-sharp).

Now that the standard tone has been found for all 12 pitch categories, it seems desirable to describe the schematic traits of each model.  The primary goal of the schematic is comprehension of the double relationship, and abstraction from identical structures which can create fourth-fifth confusions (characteristic of some possessors of spontaneously-formed absolute pitch).  We have illustrated our double-reference schematics in Figure 1.

Fig. 1. Dual-relation schematic.

We selected melodies for the basic ten and two auxiliary tonalities (b-C and G-flat-f).  These melodies were Russian and Ukranian folk songs which already contained the required standards as their first notes.  Because of this, and because the melodies did not present any interfering influences, the standards were easily memorized.  Since melodies may be difficult to comprehend because of their multiple levels, we used educational harmony, giving functional categories to the melodic steps.  Here are examples of three songs containing references to do, re, and mi (Figures 2, 3, and 4).

Figure 2.  Russian folk song.

Figure 3.  Russian folk song.

Figure 4.  Ukranian folk song.

Over various years, the author of this work has led 7 different experimental groups in the formation of absolute pitch, in which a total of 34 children from ages 3-7 participated.  The purpose of the experiments was to create an adequate method for absolute pitch acquisition, i.e. a technique which, upon application, could create absolute hearing which would meet the standard criteria for absolute pitch ability.

Existing literature has set the following criteria for true absolute hearing.

1.  The number of correctly identified sounds (63% or higher accuracy) [33], [48].
2.  Absence of transposition errors [29].
3.  Generalization to different octaves [36], [48].
4.  Identifying the key signature of a musical composition [18], [38].
5.  Presence of the ability in normal musical activities [29].

The results of our experiments have shown that children differ considerably from each other in the degree of their formation of absolute pitch.  The level of absolute development was largely dependent on the level of development of their relative ear at the time training sessions began.  For this reason, an 8th experimental session was conducted, divided into two groups based on their relative abilities:  A (strong) and B (weak).

The training experiment was conducted with children aged between 3 years, 9 months and 4 years, 2 months, in the Frunze district of Leningrad, from October 1981 to May 1982.  There were five children in Group A and four children in Group B.  Training sessions were held twice a week for 30 minutes each; in total, there were 48 sessions dedicated to absolute pitch training and 16 sessions on the application of this ability to musical tasks.

Before launching the basic experiment, we conducted a preliminary experiment to determine whether any children in the experimental groups already possessed absolute pitch.  Because children very quickly learn the names of musical sounds, it only took two sessions for them to be able to remember the names and relate the names to white keys in the middle octave of the piano.  By the third session, children were asked to identify the sounds.  The instructor played a random sequence of 20 tones to be identified; the children's response time was not measured, as it is not considered a criterion of absolute pitch [29], [48].  To eliminate the possibility of relative identification, 5-10 distracting tones were played in between each identification trial.

Table 1 shows the results of the initial testing for the two groups:

Subject Trial (Identification tone)
  1 (F) 2 (F) 3 (E) 4 (E) 5 (B) 6 (F) 7 (D) 8 (C) 9 (B) 10 (A) 11 (E) 12 (A) 13 (G) 14 (C) 15 (B) 16 (G) 17 (C) 18 (A) 19 (D) 20 (E)
  Group A
Olya N. - + - - - - - - - - - - - - - - - - - -
Masha E. + - - - - - - + - - + - - - - - - - - -
Tanya M. - + - - + - + - - - - - - + - - - + - -
Alesha G. - - + - - - + - - - - - + - - - - - - -
Anfisa L. - - - + + - - - - + - - - - - - - - + -
  Group B
Olya S. - + + - - - + - - - - + - - + - + - - -
Anya T. - + - - - - - - - - - - - + - - - - - -
Roma M. - - - - - - - - - - - - - - - - - - - -
Alesha V. - + - - + - - - - - - - - - - - - + - +

The number of correct answers in group A varied from 5% (Olya N) to 25% (Tanya M), and in group B from 0% (Roma M) to 30% (Olya S).  According to the quantitative criterion, it is clear from our test that these children did not possess the ability.  Any correct answers from either group should be considered accidental.

In the development sessions, children of both groups were trained together.

The songs were introduced in the first session.  Each song was presented first by naming its title, then by explaining what it was singing about, and then by showing a picture which illustrated the song's text.  After this, the songs were played on the piano first in their original octaves and then, after repeating the same introductory explanations, in different octaves.

In the second session, the material of the previous phase was briefly reiterated.  The songs' dynamics were explained:  sustained or not, smooth or staccato.  These traits were analyzed along with meter, rhythm, and the basic melodic structure.  The children were asked to move their hand along with the melodic line, reinforcing the concept of melodic contour.

In the third session, the previous material was reviewed, and then harmony was added to each melody.  The general character of each piece, and the means of its expressiveness-- in particular, the effect of the harmonies, as compared to the melody without harmony-- was analyzed in detail.

In the fourth session, the third session material was repeated, as well as examining the specific notation of the harmonies.  Harmonic analysis (explored by seeing and hearing certain tonic melodies and scales) brought children to an understanding of harmonic stability, centrality of tonic harmony, and assignation of the basic tonic of a tonality.  Each harmonized tonal melody was presented for recognition in two or three octaves.

 There was a three-step process to make it possible to identify individual tones.  First, to present the basic melody of each model; secondly, to bring to the children's attention the complex traits that produce the desired tonality; and thirdly, to focus on the dominant traits in a mix of sounds.  To this end, a system of musical "hints" was developed, from easy to difficult; altogether there were 15 varieties of "hints" for each standard tonic (Figure 5).

Figure 5.  "Tips" for a pitch.
a:  melodic line alone
b:  harmonies for the melodic line
c:  reinforcement of the tonic sound.

After mastering the first two songs, each of the subsequent sessions included a "Playing and guessing" game.  The examinee was given a trial pitch; if the child was at a loss to answer, he was given "hints" until he provided a correct answer.  These games were one way of mastering the sensory traits of tonal patterns.  Gradually, the names of the songs were replaced by the names of the pitches (do, re, mi, etc).

At the end of the experiment, the children were individually tested as they had been at the start.  They were asked to identify individual tones, both white and black keys, from the three octaves centered in the middle of the piano.  Octave names were not required.  The researcher gave no feedback for incorrect answers.

Table 2 shows the post-test results for the two test groups:

Subject Trial (Identification tone)
  1 (D5) 2 (E4) 3 (Db6) 4 (C4) 5 (G3) 6 (Gb4) 7 (E5) 8 (Ab3) 9 (F5) 10 (Eb5) 11 (B2) 12 (F6) 13 (Bb4) 14 (Db4) 15 (A4) 16 (C5) 17 (E3) 18 (Eb4) 19 (B4) 20 (Bb4)
  Group A
Olya N. + A# Bb + + + + E D E + + + E# + F + C + +
Masha E. + + + + + + + + + + + + + + + + + + + +
Tanya M. + + + + + + + F B + + + A# + + B + + + +
Alesha G. Db G + + + + + Gb + D + + + + + A + + + +
Anfisa L. + G + + C + A + + + + E + + + + + + Db +
  Group B
Olya S. - + + - - - + + + - - - - + - - + - + -
Anya T. - + - + - - + - - - - - - - - - - - - +
Roma M. - - - - + + - - + - - + - - - - + - - -
Alesha V. - + + - - - + - - + + - + - - + + + + -

The number of correct answers in the strong group (A) varied from 60% (Olya N) to 100% (Masha E); in the weak group (B) from 20% (Anya T) to 50% (Alesha V).  Group A's average was 78% correct, while group B was 33.75%.

After the experiment and the post-test of isolated tones, we entered a new stage:  applying absolute pitch ability to normal musical activity.  Pitch identification is characterized by discrete perception of sequential sounds, and musical practice begins with association of those consecutive sounds into a single whole.  Development of absolute pitch should continue into sound relationships, and perceptions of those relationships in more complex music.

With the pitches well-known, and the children having learned to connect those sounds to graphic characters, we were able to begin singing the melodies.  The training pieces from the book Solfeggio [1] were used for dictation and sight singing.  Pieces were first drawn solely from do-major and then from other tonalities.

After 16 such training sessions, the children were again tested on isolated tones following the previously-described method.  The number of correct answers ranged, in Group A, from 65% (Olya N) to 100% (Masha E), and in Group B from 20% (Anya T) to 50% (Olya S).  Group A averaged 85% and Group B averaged 36.25%.  The examinees did not make transposition errors, and they successfully identified tones from different octaves.  The experiment thus shows that the production of music not only supports absolute hearing, but improves it.

The strong test group (A) shows, by quantitative and qualitative analysis, that they have developed true absolute hearing.

The weak test group (B) shows the beginnings of absolute ability, as children in this group made identification judgments without referencing previous tones, but quantitatively they are less developed.  This can be explained by the gradual development of the musical ear, such that relative pitch skill precedes formation of absolute pitch skill.  In this manner the data confirm the hypothesis of the study.

Examination of absolute pitch development during productive musical activity has shown that with our examinees, the discrete perception and reproduction of tonal sequences was gradually overcome; children could detect and sing within any tonality using active absolute pitch.

Tone deafness and active perfect pitch warrant special attention, and this is a task for our future research.  Future studies must show the concrete value of absolute listening to different kinds of music, in the same way as the musical benefits of solfeggio have already been explored.

The results of our experimental research can be used to provide a method not only of developing the active form of absolute pitch, but of managing and improving spontaneously formed absolute pitch, a need which is not currently met in the field of professional musical education. 


1.  Баева Н, Зебряк Т.  (1976).  Сольфеджио для I-II классов ДМШ.  Л., 86 с.

2.  Бронштейн А.И., Петрова Е.П., Брускина Л.М., Коменецкия А.Г.  (1959).  Материалы по исследованию слуха новорожденных и детей раннего грудного возраста. – В сб.: Проблемы физиологической акустики Л., IV, с. 114-122.

3.  Венгер Л.А.  (1976). 1.  О путях изучения генезиса сенсорных способностей.  2. О некоторых дискуссионных вопросах теории способностей. – В сб.: Генезис сенсорных способностей. 1. с. 4-18; 2. с. 233-248.

4.  Ворановскиц Г.О.  (1972).  возможности дальнейшего развития теории ладоьых функций. – В сб.; Проблемы лада с. 77-98.

5.  Войтинский Е.Я.  (1958).  Исследование точности «абсолютного» различения частоты звуковых сигналов у человека:  Автореф канд. дис – Л, c. 23

6.  Гарбузов Н.Л.  (1948).  Зонная природа звуковысотного слуха. – Л., 82 с.

7.  Готсдинер А.Л.  (1981).  О восприятии музыки и музыкальном слухе. – В сб. Роль музыки в эстетическом воспитании детей и юношества. Л., с. 41-57.

8.  Зинченко П.И.  (1961).  Непроизвольное запоминание. 522 с.

9.  Ительсон Л.Б.  (1971).  Парадоксы восприятия и экстраполяционные механизмы перцепции – Вопросы психолёгии, No. 1, с. 63-71

10.  Кауфман В.И.  (1947).  Абсолютный слух. – В сб.:  Труды Института по изучению мозга им В. М Бехтерева. Л., т. XVII, с. 169-170.

11.  Кольцова М.М.  (1967).  Обобщение как функция мозга. – Л., 173 с.

12.  Леонтьев К.Н.  (1965).  Проблемы развития психики – 545 с.

13.  Мазель Л.А.  (1962).  О мелодии. – 300 с.

14.  Майкапар С.М.  (1915).  Музыкальный слух, его значение, особенности и метод правильного развития.  СПб, 233 с.

15.  Мальцееа Е.А.  (1925).  Абсолютный слух и методы его развития Сб. работ физиолого-психологической секции ГИМНа. – вып. 1, с. 33-55.

16.  Миллер Дж.А.  (1964).  Магическое число семь плюс или минус два: (О некоторых пределах наи~ей способности перерабатывать информацию). – В кн.. Инженерная психология, с. 192-244.

17.  Морозов В.П.  (1965).  Вокальный слух и голос. – Л., 85 с.

18.  Иазаиконский Е.В.  (1983).  Слух Асафьева. – Советская музыка, No. 7, с. 81-89.

19.  Платоноа К.К.  (1972).  Проблемы способностей – – 310 с.

20.  Потапова А.Я.  (1970).  Об условиях, затрудняющих параллельное протекание опознавательных процессов. – Вопросы психологии, No. 5, с. 67-78.

21.  Потапова А.Я.  (1978).  Формирование опознавательных навыков в разных условиях обучения. – Вопросы психологии, No. 1, с. 81-89.

22.  Рагс Ю.Н., Назайканский Е.В.  (1966).  Музыкально-теоретические исследования и разработка теории слуха. – В сб.: Лаборатория музыкальной акустики. М, с. 10-29.

23.  Рагс Ю.Н.  (1981).  Слух музыкальный. – В кн.: Музыкальная энциклопедия. т. V, с. 102-105.

24.  Ржевкин С. Н. (1936).  Слух и речь в свете современных физических исследований. – Л, 307 с.

25.  Римский-Корсаков Н.А.  (1911).  Музыкальные статьи и заметки. – СПб, 223 с.

26.  Розенов Э.  (1903).   Новый способ развития музыкального слуха. – Известия С.-Петербургского общества музыкальных собраний. СПб, с. 12-13.

27.  Скребков С.  (1965).  Как трактовать тональностью – Советская музыка, No. 2. с 36-38.

28.  Рубикштейн С.Л.  (1959).  Принципы и пути развития психологии. – 344 с.

29.  Теплов Б.М.  (1947).  Психология музыкальных способностей. – 325 с.

30.  Теплов Б.М., Борисова М.Н.  (1957).  Чувствительность к различению и сенсорная память. – Вопросы психологии, No. 1, с. 61-76.

31.  Тюлин Ю.Н.  (1966).  Учение о гармонии – 3-е изд. – 212 с

32.  Фелициант А.Н.  (1962).  Вступительная статья. В кн.: Кондратьева А.И. Развитие музыкального слуха. с. 3-12.

33.  Хилыард Э.Р.  (1963).  Методы и приемы анали. за процесса научения. – В сб. Экспериментальная психология. с. 11-58.

34.  Шехтер М.С.  (1967).  Психологические проблемц узнавакия. – 210 с.

35.  Abraham, O.  (1901).  Das absolute tonbewußtseinSammelbände der Internationalen Musikgesellschaft, 3, 1-86.

36.  Bachem, A.  (1948).  Note on Neu's review of the literature on absolute pitch.  Psychological Bulletin, 45, 161-2.

37.  Bachem, A.  (1954).  Time factors in relative and absolute pitch determination.  Journal of the Acoustical Society of America, 26, 751-3.

38.  Bachem, A.  (1955).  Absolute pitch.  Journal of the Acoustical Society of America, 27, 1180-1185.

39.  Brady, P.T.  (1970).  Fixed-scale mechanism of absolute pitch.  Journal of the Acoustical Society of America, 48, 883-887.

40.  Eitz, C.  (1911).  Bausteine zum Schulgesangunterrichte im Sinne der Tonwortmetode.  Liepzig, Brietkopf und Härtel, s. 3-84.

41.  Gough, E.  (1922).  The effects of practice on judgments of absolute pitchArchives of Psychology, 7(47), 1-93.

42.  Köhler, W.  (1915).  Akustische Untersuchungen III.  Zeitschrift für Psychologie, 72, 1-192.

43.  Von Kries, J.  (1892).  Über das absolute GehörZeitschrift für Psychologie, 3, 257-79.

44.  Rakowski, A.  (1972).  Direct comparison of absolute and relative pitch.  In Bilsen, F.A. (ed), Symposium on hearing theory 105-108.  Eindhoven, The Netherlands: Instituut voor Perceptie Underzoek.

45.  Rakowski, A.  (1978).  Investigations of absolute pitch.  Warsaw, Kansas, p. 1-11.

46.  Stumpf, C.  (1890).  Tonpsychologie.  Band I, II.  Leipzig, I – 424 s, II – 549 s.

47.  Von Bekesy, G.  Zur Theorie des Hörens. Ober die eben werkbare Amplituden-- und Frequenzanderung eines Tones.  Physiol. Zeitschrift, 30, 721-45.

48.  Wellek, A.  (1938).  Das absolute Gehör und seine Typen.  Zeitschrift für Angewandte Psychologie und Charakterkunde-Beihefte, 83, 1-368.