Dr. Damjana Bratuz
Home Page < Essays

Bla Bartk - On Bartk's Improvisations and the "Pippa Principle"

A long time ago I read a statement by Wilhelm Furtwngler to the effect that, unlike modern music, classical music seemed to remain impervious to the onslaught of the amateur performer. The explanation, or the implication, was — if memory serves me correctly — that by virtue of their very structure, their built-in relationships, works by Mozart and Beethoven allowed their greatness to remain recognizable and undiminished, in spite of the amateur's inadequacy; both could be heard clearly.

Such recognition is possible only when familiarity with a certain order enables the listener to perceive it, regardless of the execution a composition is subjected to. The same situation occurs in the realm of professional interpretation. A new, different ordering of musical language, whose codes are unfamiliar or are not yet shared by the composer-performer-listener trinomial, requires a new mode, different skills, of listening. For this to develop, a composer obviously depends on the ear of the performer. However, to hear or, better, entendre, with its double meaning of hearing and comprehending, is not as natural and absolute a capability as someone with an educated ear may assume. Much has been written in modern art criticism to show the fallacy of the theory of the "innocent eye." The fact that there exists no "innocent ear" is something most performers are simply not aware of; what the eye sees on the score, what the ear hears in the music, and what the technical apparatus translates into sound, are all born out of reflexes carried over from music previously learned. Rare is the performing artist whose sensitivity and imagination are guided by proper observation, who is aware of unfamiliar relationships, who possesses a translator's skill and perception to decode new orders of musical thought, and finally, who is capable of giving them the appropriate voice.

The performer approaching an unfamiliar style is very much in the position of someone learning a foreign language. Previous skill is of no avail. Art historian EH. Gombrich writes:

When confronted with the task of saying "Lisbeth," a child who had learned to say "papa" and "mama" produced the compromise "Pippa" — a transposition of the sounds he heard into the limited phonemes of his language. What we call a "foreign accent" is nothing but an extension of this "Pippa principle." The foreigner imitates the sounds of the new language as far as the phonemes of his native tongue allow. The motor habits acquired early in life will not only condition his speech but also the way be "hears" the language. His original schemata have conditioned him to watch out for certain distinctive features while ignoring other variations of sound as irrelevant, and nothing proves harder than articulating the world of sound afresh . . . . An accent, we suspect, has many similarities to those all-pervading qualities we call "style." ( 1 )

Most piano students and concert artists have been conditioned by nineteenth-century repertory and carry what has become to there a native accent into the repertory of the preceding centuries as well as of the twentieth. Although in Western musical tradition certain features of prosody and rhythm are shared by different schools and individual composers (not unlike those which occur in the spoken languages), such shifts of musical thinking and such a reordering of relations have taken place in history, that no musical ear could have failed to perceive them — or so it seemed to the composer. We know from Delacroix how resentful Chopin was when his music was praised for the less relevant qualities, while his real contribution, that is, the simultaneity of different rhythmic orbits with their resulting tensions, remained unheard.( 2 ) To this day, most performers have no awareness of this dimension in Chopin's music. We know from Debussy how much he disapproved of what he called the "cocktail" concept of music:

Je m'efforce d'employer chaque timbre I'tat de puret; comme Mozart par example. Et, de I vient qu'on ne sait plus jouer du Mozart. On a trop appris mlanger les timbres; les faire ressortir par des ombres ou des masses, sans les faire jouer avec leur valeurs memes. ( 3 )

But the blurring of lines and layers into an amalgam of sound is the very way Debussy's music is still heard, for the most part, today, in the name of a vagueness which exists solely in the perception of the performer.

Yet, it is the music of Bartk which appears to have suffered most through the misreading and misinterpretation of performers — Western and, paradoxically, Hungarian also. So much is this so, that when the re-issue of a Bartk piano recording comes along to one hears in it not only a totally different reading, but indeed a different kind of music. In my experience, the force of the "Pippa principle" is so apparent that the playing of a Bartk recorded performance often results in such categorical statements by teachers and students as, "Bartk was obviously one of those composers who could not play his own music."

"Let my music speak for itself," Bartk answered his first biographer, Denijs Dille, when the latter suggested that perhaps some explanation for the listeners would be in order. But the chance for Bartk's music to be intelligible was less than it had been for that of Chopin or Debussy, for the new mode of perception in compositional and imaginative matters it called for required familiarity with a prosody that was alien to Western ears.

No analysis of the score or technical know-how can possibly supply a pianist with the capacity to interpret, that is, to translate from a notation which he can only read and hear in terms of traditional symbols and references. For instance, he will naturally, unthinkingly, lead short values toward longer ones, see to it that up-beats fall onto the next down-beat, will distribute accents of lesser or greater intensity, or shape phrases and round them off with appropriate nuances. These commonplaces are understood as part of musicianship itself, and not as part of a musical heritage based on a distinctive prosody. In this heritage, the musical discourse follows linguistic properties common to Indo-European languages: the anacrusis, a rhythm conceived as a succession of accented and unaccented sounds and inflection. Bartk, however, adopted in his musical prosody the metrical properties of a Ianguage which is not Indo-European. Hungarian has no anacrusis — every word is stressed on the first vowel, first syllable — and the rhythm is quantitative, i.e. composed of short and long syllables. ( 4 ) Bartk had no tradition to follow; he adopted the heightened version of the Hungarian and other non-Western idioms contained in folk music and used them as a structural buildingblock. In a single lifetime, his music filled the gap which derived from the absence of a long tradition, creating the full cycle of evolution which Western music had taken centuries to achieve. Bartk did not use folk music for its exterior signs of national characteristics, but as an element of musical structure in a new order of relationships, in new dimensions of time, space and proportion.

Bartk's Improvisations Op. 20 represent what he termed the "second way" of composing with authentic folk material, i.e. using it as a "motto" surrounded by, embedded in and juxtaposed to totally alien, invented, "daring" musical substance. ( 5 ) The eight tunes of the Improvisations contain the three types of Hungarian rhythm which Bartk described: ( 6 )parlando-rubato, requiring a speech-conditioned articulation of longer and shorter syllables, indicated in No. 3 of the Improvisations and implied elsewhere, e.g. in No. 7, by dashes above a succession of notes; tempo giusto rhythm, of a stricter kind, yet preserving always the flexibility of the dance gesture, as exemplified in Bartk's recorded performances; and the dotted rhythm the most alien to Western tradition, in which the main stress occurs on the shorter note. ( 7 ) In the score of the Improvisations, all the given stresses convey a particular quality of attack and duration which is abstracted from the spoken sound of the text. The lack of anacrusis pull is clearly marked by staccato dots, short rests, or slurs. Familiarity with the composer's collections of folk material acquaints one with the four-line structure of each "motto" as well as with their syllabic formations. The lines of Nos. 1, 2, 5, 6, 7 and 8, form an ABCD structure which speaks for the antiquity of the tunes. Those of Nos. 3 and 4 have an ABA'B' structure, further indication of their Eastern connections.

Both careful obeying of the signs and intellectual probing remain powerless to activate those gestures, that diction, that freedom of articulation which are intrinsic to the music. Instinctively a "foreigner" will either read into the score what it does not contain or less than it does, or he will distort it. There is for instance a sort of dynamic equality to the articulation of the Hungarian folk-derived idiom, not unlike that of Baroque keyboard music, in which agogic flexibility replaces, for expressive purposes, Romantic inflection and nuance. Any search for the key to this style in the content of the text —of what the tune is about — is based on ignorance as well as naiveté, since all musical properties contained in the spoken work — the accentuation, the length of the vowels — have been absorbed into the musical material, transformed and clearly notated. Like language, music is speech, in a sense, before it is sign. Therefore Bartk never allowed his son and pupil Peter to play anything that he had not learned to sing first. The inner ear decides all.

Familiarity with the language Bartk drew from in his Improvisations generates that shock of recognition that comes upon hearing his performances of Nos. 1, 2, 6, 7 and 8, in the recent re-issue of an old recordings. ( 8 ) One is reminded of Emerson's words:

Only so much do I know, as I have lived. Instantly we know whose words are loaded with life, and whose not.

Often we hear a foreigner speak our language in the most accurate, precise manner, and yet it does not sound "right." Most students play unfamiliar styles like an actor who learns the lines phonetically without being able to speak the language himself. If I mentioned above that even Hungarian performers often misinterpret Bartk's music, that happens because few have been exposed to the musical language of their village culture, such as Bartk knew it.

But it is in a comparison of the reading of the Improvisations by a superior musician with Bartk's own that one can best study the mischief of the "Pippa" principle." Few contemporary performers possess the wide-ranging preparation, the integrity, or the skill of the American pianist Charles Rosen. And yet he betrays his unawareness of Bartk's world- behind-the-notes in his commentary to his recording of the Improvisations ( 9 ) when he states that "they are still in the rhapsodic tradition of the Liszt virtuoso arrangements of gypsy music." He ignores the Bartkian prosodic quality an the articulation it requires, and believes instead that:

There is a greater attempt to control the rhapsodic [italics mine] quality: the continual changes of tempo are marked metronomically, sometimes with indications that last less than half a second. However, rubato and capriccioso throughout are evidence that complete control was more than could be hoped for. ( 10 )

We are back to the non-innocent ear. Established images, in this case Lisztian, (whose pseudo-Hungarian connotations are discussed by Bartk in several essays), are mistaken for perception; the composer's attempt to notate a living rhythm is seen instead as an attempt to freeze, to structure, its flow.

In the five Improvisations performed by Bartk there is first of all a clear rendering of the fourline structure of the tune. The different tensions created by each line play one upon the other; the highest degree of intensity converges onto the third line, which is separated from the preceding one by a slight caesura, notated by Bartk in his original transcriptions by an apostrophe. In Improvisation No. I an apostrophe as well as a vertical line appear at the end of each four-line statement. The postlude effect of the last four bars is clearly conveyed in the composer's performance. There is an unmistakable initial stress to each first syllable-note of the six contained in each line. The "diction" is flexible and elastic. The "alien" territory in which the ancient tune is embedded is articulated with a resonant, harmonious and all-enveloping tone.

Mr. Rosen articulates each tune, as if it consisted of a long unbroken line: the melody proceeds without revealing its inner construction and the last beats usually carry over as upbeats. In the first Improvisation the accompaniment is subdued and not treated according to the of "second way." The melody carries on to the end with no feeling of postlude.

The Molto capriccioso of the second Improvisation does not refer to a "gypsy" style of performance. Its tempo giusto rhythm points to a dance tune, as in Nos. 4, 5, and 8. In the Universal edition, which gives the original tunes (the "mottos") on the first page, it is stated that the text for No. 2 was missing. It is conceivable that, because of the overall humourous character and boisterous treatment, the text belonged to the so-called"men's songs," whose words were generally naughty and teasing. The accelerandos and tempo changes reflect a delivery quite familiar to certain folk poetry which can be heard on various recordings and which serve as a good example of this style. ( 11 ) In Bartk's performance of Improvisation No. 2, each sign of articulation — and every bar contains several — is rendered in a truly speech-like manner; each line starts with a clear stress on the first syllable; each is separated from the preceding, with all the choreography of the inner design made audible. Mr. Rosen's reading preserves an up-beat character in the movement of the tune and disregards the agogic significance of the pesante.

The idiomatic rendering of the figures (No. 3) and (No. 6) escapes the Western ear. Mr. Rosen hears a cymbalom in Improvisation No. 6, while the peasant flute, which is really there, is not recognized. We know Bartk's articulation of this particular style already from his recording of Evening in Transylvania: ( 12 ) an instrumental — not speech-conditioned — rubato, with its own peculiar agogic freedom.

No. 7 was dedicated to the memory of Debussy who, in a similar way, had drawn from the prosody of the French language to create a new order in the language of music. In paying homage to a revered colleague, Bartk, musicien hongrois, selected an ancient, melismatic, "Asiatic" tune. In his recording, Bartk articulates each "cell" of the material in a parlando manner, with each sign conveying that sense of experience which is contained in the human voice, echoing rhythms and connotations which defy analysis and verbalization. The articulation of Mr. Rosen again favours a leading of anacrusis toward gravitational points, which would be logical only in Western tradition. The syllabic nature and elasticity of the melodic structure are not perceived. Because the four-line structure of the tunes does not guide the articulation, all significant fragments and components remain submerged in a general long line.

Improvisation No. 8 is based again on one of those men's songs whose text, in early editions, has sometimes been defined as to unfit to print. ( 13 ) The English translation of this particular song can be found in Bartk's collection Hungarian Folk Music, No. 46. ( 14 ) It is therefore delightful to discover in Bartk's performance of the tune all the subtle references of the con grazia, bar 9. The flexibility of the articulation preserves the gesture of the dance; the poco a poco accelerando, similar in spirit to the one in No. 2, and the tenuto with its elastic lengthening of syllables in the final Maestoso, are true lessons in musical prosody, both in its compositional and its performing aspect. None of these elements are present in the reading of Mr. Rosen, albeit his performance has a remarkable polish to it.

Because of the lack of insight into the properties of the ancient idiom contained in the folk tunes, the entire concept of these compositions becomes distorted. It is a concept based on the clash of two distinctive and opposing musical languages and on the tensions of two contrasting modes: the archaic and the modern, the permanent and the improvisatory. Each preserves its identity in a play of penetration and interaction; each energizes the other while being enhanced itself, like the shifting interplay of figure and background, of inner and outer, of yes and no. It is the culmination of Bartk's labours with folk material, and he never attempted anything similar again.

There remains a certain validity to performances given from the "outsider 'sit point of view, but to hear Bartk's playing of his music is to enter a world of unknown beauties. The re-issues of his recorded performances and the recent publication of his essays ( 15 ) should help to reveal, and to further, the understanding of his world.


  1. E.H. Gombrich, Art and Illusion: A Study of the Psychology of Pictorial Respresentation, Bollingen Series XXXV/5 (Princeton University Press, 1972), 364. [ Back to Text ]

  2. Anton Ehrenzweig, The Hidden Order of Art: A Study in the Psychology of Artistic Imagination (London: Paladin, 1973), 103-104. [ Back to Text ]

  3. Quoted in L'Universe sonore de Debussy by Stefan Jarocinski, Debussy et l'evolution de la musique au XXe sicle (Paris: Centre National de la Recherche Scientifique, 1965). [ Back to Text ]

  4. cf. Harvard Lectures (1943) in Bla Bartk Essays, Selected and Edited by Benjamin Suchoff (London: Faber, 1976), 383 ff. [ Back to Text ]

  5. The "first way" being a less complex harmonization of a tune, the third a sublimated, invented, folklore-based musical "tongue." [ Back to Text ]

  6. Bartk Essays. [ Back to Text ]

  7. Ibid. [ Back to Text ]

  8. Bartk plays Bartk, Turnabout/Vox Historical series, THS-65019. [ Back to Text ]

  9. Liszt-Bartk, Charles Rosen pianist, Epic Monaural LC 3878. [ Back to Text ]

  10. Ibid. [ Back to Text ]

  11. cf. Földédesanyàm, Népköltészet, Magyar Records, Vol. 1. [ Back to Text ]

  12. Bartk plays Bartk, Turn- about/Vox TB 4159. [ Back to Text ]

  13. cf. Schirmer's edition of For Children. [ Back to Text ]

  14. Béla Bartk, Das Ungarische Volkslied, Ethnomusicologische Schriften, Faksimile Nachdrucke, Denijs Dille ed. (Budapest: Editio Musica; Mainz: B. Schott's Söhne, 1965). [ Back to Text ]

  15. Bartk Essays. [ Back to Text ]


  Damjana Bratu TOP

new concept design - web design london, ontario