Smart but not so sweet: Bayan Northcott wonders why newly invented instruments have hardly ever succeeded in establishing themselves in Western music

Music, too, has its mad scientists, and one of the more colourful of them died in Moscow this month at the age of 97. Leon Theremin attained a passing fame between the wars with his invention of an early electronic instrument, duly dubbed the Theremin. This generated a single continuous tone, the pitch and volume of which could be varied by the performer's hands hovering, respectively, around a vertical antenna and a horizontal loop. Patented and manufactured in the United States, its eerie glissandi found their way into a number of Hollywood film scores and even Edgard Varese's futuristic Ecuatorial (1934).

But these were not enough to save the thing from ultimately declining into a mere music-hall stunt - and those who attended the British premiere of David Del Tredici's Lewis Carroll fantasy Final Alice, in 1982, are unlikely to forget the mystic passes of the aged performer the BBC had somehow found to render the Theremin effects the composer had naughtily demanded to suggest Alice's fluctuations in size.

In fact, the gadget was by no means the first attempt to break away from traditional acoustic instruments. Even before the invention of the amplifier, 19th-century American experiments in generating and transmitting musical tones over the telephone system had already culminated by 1900 in Thaddeus Cahill's so-called Telharmonium: a mammoth contraption played from organ consoles and weighing 200 tons. Nor was the Theremin by any means the last of its kind, being effectively replaced by the invention of the more precise Ondes Martenot in 1928. Yet no sooner had the latter gained a toe-hold in the orchestra, through Messiaen's involvement of its celestial howlings in Turangalla, than younger composers were lured onwards by the successive post-war developments of the tape-recorder, synthesiser, sampler and computer-controlled sound.

And so it goes? Sibelius once remarked to a pupil that 'orchestration is the discomfiture of absolute idealism', meaning that the free play of the sonorous imagination is constantly up against the limitations in range, volume and agility of actual instruments. In one sense, the whole history of Western musical instruments, their evolution, standardisation and refinement, could be interpreted as an endless quest to bridge the gap between the actual and the ideal. So the skirling medieval fiddle was amplified into the piercingly sweet Stradivarius violin, and the pungent, out-of-doors shawm mollified into the plaintive modern oboe; so the succession of serpent, ophicleide and cimbasso has in turn been displaced (still not altogether satisfactorily) by the tuba in the search for an ever more sonorous orchestral bass.

It is striking, indeed, how many instruments of the modern symphony orchestra can still trace some sort of lineage back to ancient times; surprising, on the other hand, how comparatively few seem to have been invented from scratch in more recent times. No doubt the piano is the great exception: its distinctive action was virtually devised in 1709 by a single maker, Bartolomeo Cristofori - little though he can have foreseen the instrument's ubiquitous acceptance by the beginning of the 19th century. Then again, the 20th-century percussion section has been steadily expanded by all manner of novelties of which the swivel-drum rototoms and John Cage's water gong are only two of the more enterprising examples.

All the same, the failure rate of would-be inventors seems to have been inordinately high. Among the myriad manufactures of the 19th-century Belgian, Adolphe Sax - including a project to suspend a giant pipe organ over Paris - only the saxophone and, in military bands, the saxhorns, have lasted. And pace James Wood's aspirations for his recently founded Centre for Microtonal Music, at the Guildhall School of Music and Drama, earlier attempts this century to create fresh possibilities by cramming more pitches into an octave than the evenly spaced 12 semitones of conventional Western tuning have proved none too fertile. 'That's the kind of experiment that is forever being made - and forgotten in next to no time,' remarked Sibelius of the indeed little-remembered music that the Czech, Alois Haba, wrote for quarter-tone and even sixth-tone instruments.

Yet the real puzzle is more fundamental. Why, after several decades now of ever-accelerating technological sophistication, are electronic instruments and methods of sound manipulation not well on the way, as early prophets such as Varese hoped, to superseding traditional orchestral instruments which, as Boulez often complains, have developed little since the 19th century? Part of the answer is that the orchestra has proved unexpectedly adaptable; that, after toiling in their electronic laboratories to devise unprecedented sounds, composers have found they could draw the same effects from complex blendings of acoustic instruments - resulting in a whole new range of cloud textures and sonic frissons in the orchestral works of such composers as Ligeti and Berio. But part of the answer is surely artistic. Electro-acoustic music has undoubtedly achieved a handful of classics since Varese's interspersing of taped musique concrete between the instrumental sections of his Deserts (1954) or Stockhausen's interplay of choirboy voices and electronically generated sounds in Gesang der Junglinge (1956). Yet it would be idle to pretend that the yield to date anywhere near fulfils the kind of expectation implied by Messiaen's dictum, even before the electro-acoustic age had really got under way, that 'an abundance of technical means enables the heart to expand freely'.

And the explanation for this surely lies in Stravinsky's counter-assertion that art proceeds not from possibilities but from choice. 'I shall go even further,' he remarked in his 1939 Poetics of Music. 'My freedom will be so much the greater and more meaningful the more narrowly I limit my field of action . . . Whatever diminishes constraint diminishes strength. The more constraints one imposes, the more one frees oneself from the chains that shackle the spirit.' In other words, far from inhibiting invention, the very limitations of traditional instruments, tuning and the compositional materials that can be wrested from them constitute both a challenge to a composer to create something new and a measure of the degree to which he succeeds. From this point of view, the devising of instruments capable of ever more subtle subdivisions of pitch, rhythm and dynamics, would tend to be retrograde, subverting possibilities of strong contrast in ever smoother degrees of transition. And the invention, perhaps no longer so far off, of some ultimate electronic gadget wired to the brain, and capable of translating whatever the free play of the sonorous imagination could come up with into instant reality, would for Stravinsky have constituted the ultimate threat: by removing all constraint, also removing any criterion for why one compositional choice is better than another.

Actually, he was not so conservative as this might make him sound; believing that the retuning of scales probably would open new expressive possibilities for Western music; but selectively and in specific contexts, not through some systematic all-or-nothing revolution. In the end, new instrument makers like everyone else come up against the limits of perception and their artistic consequences. The ear is able to follow the minute microtonal inflections of line of an Indian sitar player precisely because these are thrown into relief against an unchanging drone; import those inflections into every level of a harmonic texture and one would be liable to end up with the aural equivalent of squeezing all the brightest coloured plasticines together - and producing an undifferentiated grey. Once again, it was Sibelius in old age - dutifully keeping up with broadcasts of the latest trends - who underlined the point, after some particularly complex and doubtless grey offering, that true personality can show itself in just five notes.