William Walton once advised all sensitive composers 'to die at the age of 37': at least they would escape the 'critical damnation' of failing to live up to their early promise. Happily, Rameau was not around to heed this tongue-in-cheek advice. Had he done so, he would never have become one of the most important figures in French musical history. By that age he had yet to compose some nine-tenths of his music or to publish the theoretical writings that would transform the way in which the scientific basis of music is understood.

Who was Jean-Philippe Rameau?

Rameau was, in fact, the classic late developer. At school he was a drop-out, and his first 40 years were spent mainly in the obscurity of the French provinces where he held a succession of organist posts. During that time he composed little more than one slim volume of keyboard pieces (the Pieces de clavecin of 1706) and a few other 'bagatelles', as he later described them: motets, cantatas and drinking songs.

Not until he was 39 - two years beyond Walton's cut-off point –did Rameau begin to produce the works by which he became best known. First to appear was the Traite de l'harmonie (1722). This epoch-making, 450-page tome set out theories of chord-formation and much else that have formed the basis of harmony teaching ever since. To supervise the printing – evidently a fraught experience - the composer moved to Paris, where he remained for the rest of his life (he died shortly before his 81 st birthday).

Within two years of his arrival in Paris, Rameau turned once again to the harpsichord. The success of his Pieces de clavessin of 1724 encouraged him to issue a further volume, the Nouvelles suites de pieces de clavecin, in 1729 or 30. Compared with his youthful set of 1706, these collections reveal a huge increase in technical difficulty and a broader range of moods. Following Couperin, most of whose harpsichord works had now appeared, Rameau included more pieces with evocative titles: 'The Conversation of the Muses', 'The Cyclops'. But the muscular energy of his keyboard music is far removed from the enigmatic introspection of Couperin's, as is his fondness for the formal dances (Allemandes, Courantes and the like) which were now less fashionable.

At around the time of his Nouvelles suites, Rameau published a volume of Cantates fran,coises. These works, like his earlier cantatas, employ that happy synthesis of French and Italian styles which characterises the genre. Yet there are hints of the dramatic power of his future stage works, notably in the searing dissonances of Le bergerff'dele.

Still, such small-scale works – for all their skill and imagination – would never earn him the kind of reputation as a composer that he enjoyed as a theorist. For that, he would have to turn to opera, something which had preoccupied him since the age of 12 but for which he only now felt ready. In 1732 he turned for a libretto to the abbe Simon-Joseph Pellegrin, who had recently provided Monteclair with the text of his biblical opera Jephte, a work that had deeply impressed Rameau. Pellegrin, though universally reviled in literary circles, understood his craft. His libretto for Rameau's Hippolyte et Aricie is one of the strongest the composer ever set. It owes its ancestry to tragedies by Euripides and Seneca but most of all to Racine's Phedre. The theme was a bold one: the incestuous love of Phaedra for her stepson Hippolytus. It was ideally suited to Rameau's talents, and he responded with music of astounding power. This coup d'essai was also his coup de maitre.

Hippolyte was first performed in 1733, a week after Rameau turned 50. Few at that premiere can have been prepared for what they heard. Voltaire was certainly not alone in declaring himself 'disgusted' (though he soon became one of the composer's champions). Opera goers were used to a repertory dominated by the tragedies of Lully, many of them at least half-a-century old. Despite their dramatic effectivenes, these works were often harmonically bland and orchestrally unadventurous. By contrast, Rameau's harmonic intensity and sheer vigour proved disconcerting. His music was judged excessively italianate, unnatural - in a word, baroque (a derogatory term at that time: Hippolyte has the dubious distinction of being the first musical work to which it is known to have been applied).

Inevitably, two rival camps soon formed - the conservative lullistes and the composer's growing number of supporters, the ramistes or ramoneurs (chimneysweeps). The dispute between them raged throughout the 1730s and had not fully abated by 1749, when a cabal disrupted the first run of Zoroastre. Rameau, taken aback by the strength of feeling, threatened to abandon opera. But he soon realised that the lullistes, though vociferous, were losing support, while his supporters hailed him as 'the Orpheus of our century'

Once his operatic career had been launched, Rameau never looked back. l lis operas of the 1730s include many of his finest works. These include two further tragedies. Castor et Pollux (1737) is unusual in concentrating less on romantic entanglements than on the brotherly love of the eponymous twins; Dardanus (1739), though based on a feeble plot, contains some of the most astonishing music Rameau ever wrote. There are also the delightful opera-ballets Les Indes galantes (1735) and Les fe^tes d'Hebe (1739) the one set in modern times and contrasting European and other cultures (not always to Europe's advantage), the other more conservative in its source material but saturated with glorious music.

The period 1740-5 was slack by Rameau's mature standards. The composer had evidently quarrelled with the Opera management: he told a young hopeful who offered him a libretto entitled Orphee et Euridice that he had 'abandoned the theatrical lyre' because of 'certain dissatisfactions'. How sad that Rameau never set this libretto, a version of the Orpheus myth to place alongside Monteverdi's or Gluck's. Sad, too, that his setting of Voltaire's mould-breaking Samson, a libretto which eventually inspired Saint-Saens's Samson et Dalila, was abandoned because its biblical theme and its author's anti-establishment reputation were judged too controversial.

In 1745, after a change in the Paris Opera management, Rameau's productivity revived. Between then and 1749 he completed no fewer than nine operas, including Platee, Pigmalion and Zoroastre. His works dominated the stage, to the extent that in 1749 the Opera was instructed not to schedule more than two per year, for fear of discouraging other composers.

This period saw the first of many commissions from Louis XV's court. Most were intended to mark some political event. La princesse de Navarre, for example, was written for the Dauphin's wedding (1745), as was Platee, an extraordinary comedy involving the thwarted loves of an ugly marsh nymph: a tactless choice, since the Dauphin's bride was apparently 'not well served by nature'. Les surprises de l 'amour (1748), which celebrates the Treaty of Aix-la-Chapelle, was written for the amateur theatricals organised by the king's mistress, Mme de Pompadour. 'Occasional' as these works were, Rameau almost invariably breathed new life into their conventional allegories and mythological plots. If few generate quite the emotional and dramatic force of his first tragedies, they are nevertheless teeming with lively and attractive music. Above all, Rameau's virtuoso handling of the orchestra and the expressiveness of his ballet music at this time were seldom equalled in his time.

Around 1750 Rameau briefly enjoyed a period of almost universal esteem as the foremost French composer of the day. In his theoretical work, which he pursued alongside his composition, he had the support of most intellectuals, at home and abroad. But his final decade was bedevilled by further controversy. He was involved in acrimonious disputes about music theory, while during the notorious Querelle des Bouffons (1752-4) this formerly 'excessively italianate' composer found himself attacked by Italophiles as arch-representative, with Lully, of the conservative French tradition.

In his seventies, despite failing health, Rameau found the strength to publish a succession of theoretical works and to complete two of his most astonishing operas, Les Paladins and Les Boreades. This last was scheduled for performance at court in 1763; it was rehearsed that year but then inexplicably abandoned. The plot may be conventional but it is swept along by music so compelling that it is hard to believe the composer was almost 80 when he completed it.

Within a decade of his death in 1764 Rameau's music had largely fallen out of fashion. Of all the major composers of his generation - Bach, Handel, Scarlatti, Telemann – he has been the slowest to gain widespread recognition among today's musical public. Several factors are responsible. Only since the revival of period instruments has it been possible to recreate the sonorities and performing style his music needs. His operas are, moreover, notoriously expensive to stage: they require a large orchestra, chorus and corps de ballet, and only make their full effect if they bedazzle the spectator with elaborate scenic effects.

Prejudice has also played its part. For two centuries, Anglo-Saxon historians from Burney onwards have revealed a lack of sympathy. As recently as 1928 Cecil Gray could write that Rameau's music is 'unenterprising, stiff, archaic, lacking in spontaneity and charm'. Even now, English and German histories and encyclopaedias – with the glowing exception of The New Grove and its sister publications -tend to devote disproportionately little space to Rameau and the French Baroque.

Today we are learning better. A generation of performers have rediscovered the secrets of Rameau's elusive idiom, and conductors of the stature of John Eliot Gardiner, William Christie and Marc Minkowski have brought the operas and other works glowingly to life. We had the Royal Opera House's Platee last year and Christie's Zoroastre at this year's Proms. We eagerly await Sir Simon Rattle's Les Boreades next year. Rameau's time has surely come.

We named Jean-Philippe Rameau one of the best French composers of all time

Graham Sadler