The original French title of this book is a little awkward in English: Penser la musique aujourd’hui means “To think music today.” It implies that music—what it is and, above all, what it might and ought to be—can be conceived by the intelligence and that an act of thought can be a creative act of will. The publishers have decided to ignore these implications: “Boulez on Music Today by Pierre Boulez” has the advantage of getting Boulez’s name twice on the cover and of avoiding the mention of anything so repellent to the prospective customer as rational thought.

The publishers have also elected not to inform the reader that the book is unfinished; it is only volume I, and the expectations aroused by several references to a later chapter on musical form will remain frustrated. The book jacket is, in any case, singularly ill-informed: in the list of Boulez’s works given, not one is later than 1950, and Marteau sans maître and Pli selon pli, his most famous works—among the most famous works of any composer of the last twenty years—go astonishingly unmentioned.

The translators are Susan Bradshaw and Richard Rodney Bennett, a fine pianist and a very distinguished young composer respectively, and they have handsomely turned Boulez’s mandarin French into upper-class British. They have ventured on a small number of helpful notes, but it is amusing to be told that cartes du tendre (seventeenth-century allegorical maps of erotic psychology) are like Victorian moral diagrams. The opening twenty pages of polemic—without names, but it is easy to put them to the anonymous victims of Boulez’s wrath—are well rendered, although the translators naturally miss the repressed word play which growls beneath the surface of Boulez’s prose and gives it its tone.

The polemic is directed against the fashionable avant-garde trends of ten years ago in composition and analysis: the fascination with mathematical formulas divorced from all consideration of musical perception, the dadaist infantile exploitation of total chance, the use of machine noises impossible to integrate into the sound pattern of musical instruments—in short, any of the fetishes which were forcing music outside the control of the composer’s imagination when Boulez wrote this book, and of which many are still with us today.

Boulez starts from the premise that some form of serialism is an inescapable fact of life of contemporary music. What gives this premise some force is the existence of serial works composed over the past forty years that are not only of the highest quality but completely disparate in style. Schoenberg’s String Trio, Berg’s Violin Concerto, Boulez’s Marteau sans maître, Stravinsky’s Requiem, and Webern’s late cantatas all resemble one another in very little except their use of serial procedure. The increasing adoption by composers of some form of serial technique and its varied possibilities called for a theoretical basis wider than a mere explanation of the Second Viennese School.

Serialism is not, as is often supposed, a mechanical and facile procedure for producing unpopular works. It is a new way of organizing musical space. This space—the universe of notes that a composer may choose to write—is today bound at either end by the limits of the human ear. Between the highest and lowest frequencies, the space is divided by the module of the octave, which is subdivided (or “striated” as Boulez would say) into the twelve notes of the chromatic scale.

The composer needs a system to characterize this space, which, as described above, is like a series of vocal noises with no rules to bind them together into a language. From the sixteenth to the twentieth century, the space was organized by tonality which—within the module of the octave—made one triad the central one and arranged all others in a hierarchy subservient to it. (The central triad of the key or tonality of C major, for example, is C-E-G.) One basic rule guaranteed the coherence of the system: if a piece of music had more than a single tonality (and there was a dramatic advantage to setting up a polarity within a work), one of the tonalities must be central, and the piece must both begin and end in it.

By the end of the nineteenth century, this rule was abrogated so frequently that the system may be said to have largely broken down. Not only do works of Strauss and Mahler and many others begin and end often in different keys, but there are long passages in all the music from Wagner to Debussy with no specific central key-sense at all. The anarchy that followed for a dozen years from 1908-1920 was a partial (although never total) acceptance of a neutral musical space with unlimited possibilities: it was a great and fruitful period. Its masterpieces are the Rite of Spring and Erwartung.


Serialism is an attempt to characterize musical space without resorting to a centralized hierarchy of triads. It was based originally upon the German tradition of deriving a work of music from a short melodic kernel or motif. Applied only to pitch, as it was at first, it arranges the twelve notes within the octave into a particular basic order for each composition.

Serial technique does not, as many seem to think, prevent the composer from writing what his imagination prescribes. In the Lyric Suite, Berg was even able to reproduce the opening of the Prelude to Tristan und Isolde. By using the basic order inverted, in retrograde and inverted retrograde motion, and by transposing the series and fragmenting it, a composer is as free to write what he pleases as within the tonal system.

What serialism does is to characterize musical space by setting up “privileged” orders of notes. The groups found within the original series impose their identity through all the transformations of the series. It is false to assume that the series fixes the notes of a work in advance for the composer, and that choosing the series is the only creative and free act. What the series establishes is the relations between the notes: it is the relations that remain invariant through all the transformations, and they give the character to the musical space, a character that differs for each piece. The use and the exploitation of this character are as dependent as ever upon the composer’s genius—and upon his technique.

Many of even the greatest serial works (all of Schoenberg’s and Berg’s, for example) have been serial in pitch structure alone, while deriving their phrasing, rhythm, and large forms from the aesthetic developed for tonal music. This has not been as regrettable as some would have us believe, for it enabled composers to exploit the contradictions within the work to great dramatic effect. But it has entailed a clearly unstable situation which could not remain long without the now unsupported tonal elements giving way.

Boulez gives as broad and general a definition of “series” as coherently possible so that it may be extended easily to the entire field of musical thought. He is concerned to avoid the traditional limitation of series to pitch alone, for this leaves other aspects of music like tempo and timbre to imply incompatible aesthetics (although he is not the first to extend the concept of series in this way), and he also emphasizes the necessity of retaining a freedom of organization equal to that which existed in the past.

He is as concerned to discuss the possibility of expanding musical technique into regions almost totally unexploited as to classify the small part so far explored. He considers not only our present musical space divided by the module of the octave, but he also speculates on the possibility of other fixed modules, variable modules, and even of an absolutely continuous (“smooth”) space and the combination of smooth and striated systems. Similar considerations are applied to time, or tempo. The book is less a generalization from Boulez’s own experience than an infinitely extensible system intended to embrace any future developments and absorb them into a coherent tradition.

Many will naturally read this book hoping to find an explanation of Boulez’s own practice as a composer, and they will be only indirectly satisfied. The relation of a composer’s theory to his practice is never less than ambiguous, and it is rare for one unequivocally to dominate the other. The music can never be a simple illustration of the theory, and even when the theory is intended solely as an explanation of practice (or as an exercise in public relations), it inevitably trespasses beyond its self-imposed limits if only by virtue of being in a medium and mode of discourse other than the works it proposes to justify. Theory, by following the suggestions thrown out by the process of analysis itself, seems bound to over-rationalize the music it discusses, or at the least to bring it into an order and context that at some point transcends and overpowers the individual works.

A systematic theory, even one as improvised as this book of Boulez, can therefore be considered only indirectly in relation to the actual process of composition, and one is never the direct cause or the sufficient reason (as Pangloss would say) of the other. The tradition of the composer-theorist is a long one, but it is not easy to see how the music of more articulate composers, like Philipp Emanuel Bach, Schumann, or Wagner, differs because of their theoretical and critical writings from the works of more reticent composers.


Although Boulez’s theory tries not to judge the past but to give a frame for what may come, it is more profitable to read this book after a relative familiarity with his Pli selon pli or the Marteau sans maître. Several concepts in it—particularly in its present unfinished state—can only be understood in the light of Boulez’s own preoccupations as a composer.

The most interesting and important of these concepts is that of heterophony. Boulez’s treatment is complex and generalized. Heterophony is the simultaneous performance in two or more parts of the same melody (singer and flute, for example), in which one part modifies the melody by ornamentation or by a different rhythm. With the acceptance of unresolved dissonance in our century, the importance of harmony and counterpoint has considerably diminished, as they are traditionally based precisely on the resolution of dissonance, while heterophony, which skirts the question of dissonance, has taken on increased interest. It is fundamental to Oriental music, but Boulez claims that “in the Western tradition it has rarely been used even in an elementary state.”

It has, indeed, never been treated in Western musical theory before Boulez, but it is, nevertheless, an integral and indispensable part of Western music. Even if we eliminate canons and rounds (like “Row, row, row your boat”), which Boulez would consider as falling under the laws of counterpoint, and different figurations outlining the same chord, which may be considered directly responsible to harmony, there is still an important role for heterophony going back several centuries in our own classical tradition.

The most obvious case is the use of the same melody in voice and accompanying instruments (in Schumann’s songs and Wagner’s operas, for example) where the vocal rhythm differs in important details from the orchestral or piano version. Far from being directly responsible to the contrapuntal or harmonic structure, this is in almost all cases a violation of them. It even embodies what Boulez would call a structural principle, the clash of two systems of rhythm, speech rhythm and a more regular instrumental rhythm.

On a deeper level, the “laws” of counterpoint are not always the guiding principles in a classical polyphonic work. Strictly speaking, they forbid heterophony by prohibiting parallel octaves, fifths, and unisons. Yet this is a prohibition frequently subverted, not merely in detail or by simple doubling but often on a large scale. I am not alluding only to the common use of hidden octaves in music from Josquin to the present, but to the simultaneous presence of the same melodic line in a simple form in one part and in a free and elaborate form in another. The slow movement of Bach’s Italian Concerto is an example: the upper voice outlines the same basic pattern that we find in the bass, but luxuriantly ornamented and arranged so as to avoid exact parallelisms. Slightly out of phase, they are both different versions of the same model.

As in all other arts, there are forces in music that completely escape the grasp of official theory and heterophony has been one of them. It is only at a time of crisis, however, that this becomes apparent. If Boulez has not recognized the existence of these forces before his own century, this is because his own practice as a composer leads him in a very different direction. Heterophony is interesting to him not for the parallelism of the lines (as it was to Bach) but for the quality of their being out of phase. He is most interested in heterophony when the two parts embody independent rhythmic systems, one strict and one free, for example.

Boulez is often concerned with making the actual measurement of time play, by means of heterophony, a much more elaborate structural and even affective role. When in Eclat, for example, piano and vibraphone play the same rapid sequence, but one plays in strictly measured tempo and the other begins after a moment to accelerate, the sense of time is opened up, so to speak, and the absolute measurement is momentarily confounded. This turns a brief prestissimo into a static moment with an extraordinary tension.

Boulez himself recognizes that heterophony cannot be ultimately separated from homophony and polyphony and suggests a new conception of “part” (i.e. independent musical line) as a synthesis. But he himself has not elaborated this new conception, and it does not appear that even the unwritten chapter on musical form would have contained it. Instead of opposing heterophony to counterpoint, it might be more fruitful to consider independent polyphonic lines on a spectrum starting from total identity and progressing through various forms of parallelism to total freedom. Otherwise Boulez’s own music will escape the theory that he has elaborated for it.

It is, in any case, improbable that any theory will be totally adequate for music, particularly for works as yet unwritten. This of course does not imply that the attempts, such as this book, to generate such a theory are not an essential part of the creation and development of music.

This Issue

February 10, 1972