Where the term world music came from

In the summer of 1987, in the upstairs room of the Empress of Russia, a pub in north London, a small group of people who ran independent record labels tried to agree on a word. Their problem was a shelf. Shops did not know where to file the records their labels were releasing, records of Bulgarian choirs, Lagos dance bands, north Indian classical singers and Andean flute players, and a record that cannot be filed cannot be found, and a record that cannot be found cannot be sold. The word they settled on was world music, and with it, world music traditions that share no scale, no rhythm, no language and no history were gathered under one label. The label was never a description of the music. World music is a marketing term invented by the Western record trade, and the only honest way to write about the traditions filed under it is one musical system at a time, on that system’s own terms.

The people in that room worked for labels such as Hannibal, Earthworks, Stern’s and Triple Earth, small companies whose catalogues had outgrown every existing rack. Roger Armstrong of Ace Records is the man most often credited with proposing the winning words, and the alternatives on the table that day show what the trade thought it was doing. World beat was considered and set aside as too narrow, tropical was considered and set aside as too tied to one climate, and the phrase that survived did so precisely because it committed its users to nothing. It promised breadth and delivered a bin. A scholar had used the same two words earlier and for a different purpose. The ethnomusicologist Robert E. Brown had built a programme of study at Wesleyan University in the 1960s around the term world music, meaning by it the serious study of the world’s musical systems side by side. The London meeting borrowed the scholar’s phrase and put it to a shopkeeper’s work, and the gap between those two uses is the gap this article keeps in view.

The machinery followed quickly. The WOMAD festival, founded in 1980 and first held in 1982, had already shown that large audiences would gather for bills drawn from many continents, and through the following years the new category acquired a chart in Billboard magazine, a Grammy award of its own from 1992, festival stages, radio programmes and review pages. None of that machinery was dishonest in itself. It moved records across borders, paid musicians who had been shut out of the general racks, and put a singer from Cairo or a drummer from Accra within reach of a browser in a provincial record shop. But every act of filing is also an act of description, and this filing described nothing. The two words said where a record was not from, the Anglo American mainstream, and stayed silent about everything the record actually was. Understanding what the traditions inside the bin are requires abandoning the bin and taking up the systems one by one.

Why the category fails as a description

A category earns its keep by predicting something about the things inside it. Jazz predicts an improvising practice and a rhythmic feel, classical predicts a notated repertory and a concert institution, and either word tells a listener roughly what kind of attention the music will ask for. World music predicts nothing except exclusion. A Hindustani raga performance, a Cuban son and a Javanese gamelan share no pitch system, no rhythmic organisation, no performance setting and no account of what music is for. They were placed together because each of them failed to be something else, Western pop or Western classical, and a set defined by what it excludes has no internal logic, whatever its uses at the till. The category behaves like a miscellaneous drawer in a kitchen. Everything in the drawer is real, and the drawer tells you nothing about any of it.

Musicians objected early, and their objection was practical before it was principled. The American musician David Byrne set out the case in a New York Times essay in 1999 whose argument ran that the term flatters the listener while patronising the music, treating whole cultures as a flavouring for Western taste. Performers filed under the label reported the same grievances in different words across the trade press. Promoters expected a generic festival sound rather than the specific repertory an artist actually played. Journalists asked about geography when they would have asked a rock band about craft. Fees were set by the bin rather than by the work, and a virtuoso of the maqam could find himself billed as local colour on a stage where a mediocre guitarist was billed as an artist. The objection was never that audiences in London or New York should stay away. It was that the label taught those audiences to hear a place instead of a practice.

The commercial defence deserves to be heard in its own voice, because it is not a foolish one. Joe Boyd, whose Hannibal label was in that pub room, made the practical case for catch-all labels repeatedly across his career: a record filed nowhere is a record bought by no one, shelf space is the scarcest resource in the trade, and the bin financed recordings, tours and reissues that no scholarly argument would ever have paid for. On this account the label was a ladder. Musicians climbed it into general visibility and could then be filed, reviewed and booked as themselves. The critical answer accepts the ladder and disputes the destination. A ladder that delivers a Cairene master singer to a festival as an emblem of her continent, the critics replied, has not finished its work, and the filing habit outlived the emergency that created it. What would settle the argument is evidence about outcomes rather than intentions, and the evidence runs in both directions depending on the label, the decade and the tradition examined. This article therefore takes the only route that does not depend on the verdict. It describes the systems, because the systems are what the music is actually made of.

What a musical system actually is

A musical system is a connected set of decisions about pitch, time, timbre, form and transmission that a community treats as given. A performer raised inside a system can improvise for an hour without leaving it, and a listener raised inside it can hear the moment a boundary is crossed. The decisions interlock. The choice of how pitch is organised dictates what a melody can do, which in turn shapes the forms a performance takes, which in turn decides how the knowledge is passed on. Nothing in a system is arbitrary decoration on a universal core, because there is no universal core underneath. There are only systems, each coherent, each complete, and each largely invisible to the people inside it, who experience its rules simply as what music is.

The surest way to see a system is through its theory of error. Every tradition knows which mistakes are serious. A pitch placed slightly outside the permitted inflection inside a raga is a fault of one kind, judged by listeners instantly and almost physically, while the same displacement in a system built on fixed equal tempered pitches would pass unnoticed. A drummer who enters a fraction early against the bell pattern in a West African ensemble has broken something structural, where a soloist elsewhere might bend far wider and be praised for daring. Error reveals the load bearing walls. Where a tradition spends its strictness, on pitch, on the cycle, on the memorised text, on the relationship between leader and chorus, is where its real organisation lives, and a listener who learns to hear the errors has learned the system faster than one who memorises its ornaments.

The systems also differ in what they ask a newcomer to learn first. One tradition puts its grammar in the organisation of pitch and treats rhythm as the performer’s playground, another builds its strictest architecture in time and leaves pitch comparatively open, a third locates the art in ornament, in how a note is approached, bent and released rather than in which note is chosen. The table in this section gathers the traditions surveyed in this article into one view, with the organising pitch concept, the organising rhythm concept, the transmission method and a representative figure for each, so that the differences of structure stay visible instead of dissolving into a survey blur. A note on spelling before it is used. Transliteration in this article follows the forms heard in concert life and read on record sleeves, raga, tala and maqam without diacritic marks and personal names in the forms their owners used in print, because those are the forms a listener will meet in a shop or a programme note.

Tradition	Region	Organising Pitch Concept	Organising Rhythm Concept	Transmission Method	Representative Figure
Hindustani classical	North India and Pakistan	Raga	Tala cycle	Guru and student lineage	Ravi Shankar
Carnatic classical	South India	Raga within the melakarta scheme	Tala within the suladi scheme	Teacher lineage and sabha institutions	M. S. Subbulakshmi
Arab art music	Egypt and the Levant	Maqam	Iqa rhythmic mode and free taqsim	Master performer and conservatory	Umm Kulthum
Persian art music	Iran	Dastgah, learned through the radif	Improvised metre within gusheh models	Memorised radif from a master	Hossein Alizadeh
Turkish art music	Turkey and the Ottoman world	Makam	Usul cycle	Master lineage and radio institutions	Münir Nurettin Selçuk
West African drumming	Ghana and the wider region	Speech tone and drum timbre	Interlocking bell timeline	Hereditary and community apprenticeship	The master drummers of the Ewe tradition
Qawwali	Pakistan and North India	Raga inflected melody	Tala cycle with handclap patterns	Hereditary qawwal party	Nusrat Fateh Ali Khan
Cuban son and salsa	Cuba and New York	Diatonic harmony with montuno patterns	Clave pattern	Dance band apprenticeship	Celia Cruz
Brazilian bossa nova	Rio de Janeiro, Brazil	Extended song harmony	Samba rhythm condensed to guitar	Informal circle and radio	Antonio Carlos Jobim
Chinese classical	China	Mode keyed to instruments and tablature	Percussion patterns and free rhythm	Tablature books and master teaching	Wu Man
Japanese classical	Japan	Mode and fixed tuning systems	Jo-ha-kyu pacing and percussion cues	Iemoto licensing lineage	Yamaguchi Goro

A table like this one is a teaching instrument before it is anything else, and it flattens as it clarifies, which is why the sections that follow take each row apart in prose. Teachers who build a lesson around such a comparison and lay it out as slides can open the file in a browser-based presentation viewer to check how the comparison reads at a distance before it is shown to a class. The comparison earns its place only if it sends the listener back to the individual systems with better questions. The next four sections therefore step back from the named traditions and look at the four materials every system has to organise: pitch, time, ornament and improvisation, and then at the way all of it is handed down.

Scale, mode and the organisation of pitch

The piano in a Western room teaches a quiet lesson that most listeners never notice learning. Its octave is divided into twelve equal semitones, a compromise adopted across Europe between the eighteenth and nineteenth centuries so that music could move between keys without any of them sounding badly out of tune. The compromise is so complete that the twelve notes come to feel like the raw material of music itself, the atoms out of which all melody everywhere must be built. They are nothing of the kind. They are one system’s settlement of a problem every system has to solve in its own way: how to carve the continuous field of pitch into usable places, and how to arrange the relationships among those places so that melody can carry meaning.

Arab musicians organise pitch through the maqam, and their carving includes intervals the piano cannot produce, among them steps of roughly three quarters of a tone that give maqam hijaz and maqam bayati their unmistakable colour. These are not mistuned Western notes. They are exact pitches, aimed at precisely, judged by trained ears, and defended by theory that goes back centuries. Indian theory performed its own analysis when the treatise tradition divided the octave conceptually into twenty two shruti, small pitch positions from which the notes of a raga are selected and shaded. A raga uses a handful of pitches, but the shruti idea records something the piano hides: that a note in performance is a region with a centre of gravity, not a point, and that where a singer settles inside that region is part of the expression. Chinese and Japanese systems built their pitch worlds from other settlements again, from the fixed tunings of the qin and the koto, from modes keyed to instruments and to ceremonial functions, each internally exact and each untranslatable into the others without loss.

Underneath the variety sits a deeper divergence, and it concerns what pitch organisation is for. In the European concert tradition, pitch is organised vertically as well as horizontally: notes sound together in chords, chords move in progressions, and much of the drama of the music is the drama of that movement. In the systems surveyed in this article, pitch is organised almost entirely in the horizontal dimension, as melody unfolding against a stable reference. The reference might be a drone, as in Indian music, where the tanpura sounds the tonic and fifth without pause beneath everything the soloist does. It might be the recurring melodic pathway of a maqam, whose characteristic phrases tell the listener where home is. Either way, the expressive weight falls on line, inflection and return rather than on chord change, and a listener waiting for harmonic movement will hear stillness where an insider hears the whole argument.

Why does Indian classical music have no harmony?

Indian classical music organises pitch as melody against a drone rather than as chords in progression, so harmony never becomes its engine. The tanpura holds the tonic while the soloist spends the system’s complexity on melodic path, inflection and ornament. What sounds like an absence is a design: the interest lives in the line, not in simultaneous notes changing together.

The point is most easily heard as a division of labour. A Western orchestra distributes its pitch information across dozens of simultaneous parts, and its form is partly the story of how those parts agree and collide. An Indian classical performance concentrates its pitch information in a single melodic line of extraordinary density, supported by a drone that acts as a fixed sky and by percussion that organises time. Nothing is missing from the second arrangement that the first one has. The two systems simply spend their complexity in different places, one on vertical agreement, the other on horizontal depth, and the habit of calling the second arrangement undeveloped says more about the listener’s schooling than about the music. Pitch, then, is the first material a system organises, and the lesson of comparing its settlements is that none of them is the natural one. Each is a decision, and each decision opens some musical possibilities by closing others.

Rhythm cycles and the organisation of time

If pitch is where Western schooling most misleads the ear, time is where the world’s systems diverge most happily. The Western concert tradition measures time in bars of equal beats grouped by an accent that falls, with machine regularity, on the first beat of each group. Much of the music in this article measures time in cycles: spans of beats that recur, internally divided, with a marked point of arrival that performers and listeners feel as home. A cycle is not a bar repeated. It is a small circular geography that the performance travels around, and the pleasure of the form lies in departing from the arrival point, delaying it, disguising it and landing on it together.

The Indian tala, examined in its own section below, builds cycles of eight, ten, twelve, fourteen or sixteen beats, grouped asymmetrically and marked by claps and waves of the hand. West African ensemble music organises time differently again, around a short bell pattern that repeats without variation for the length of a piece while drum parts interlock above and below it, so that time is kept by a shared reference figure rather than by counting. Cuban dance music answers to the clave, a five stroke pattern in two halves that every part in the band must agree with, a key in the literal sense: phrases that cross it sound wrong to an insider in a way that is difficult to explain and impossible to unhear once learned. Ottoman art music developed usul, cycles that can run to dozens of beats, learned as patterned sequences of strong and weak strokes on the drum. Four continents produced four mechanisms, and all four answer the same need, to give long performances a temporal architecture that everyone on stage and many in the audience hold in common.

The comparison sharpens when the roles are reversed. A tala cycle of sixteen beats is sometimes described to Western students as four bars of four, and the description is not false, but it misses what the cycle does. The groups inside a tala are not equal in weight or gesture, the empty beat marked by a wave away from the palm carries a structural charge no bar line carries, and the first beat, the sam, is an event the whole performance anticipates rather than a grid line it passes. Cyclical time also changes the meaning of an ending. A performance organised in cycles does not conclude the way a sonata concludes, by argument completed. It settles, often by a patterned cadence, a tihai, in which a phrase is stated three times so that its final stroke lands exactly on the sam, and the satisfaction is geometric as much as expressive. Time in these systems is not a river the music floats down. It is a wheel the music rides, and the rider’s art is in the riding.

There is a social consequence buried in the mechanism, and it explains why cyclic systems so rarely need a conductor. Authority over time is lodged in a pattern that belongs to everyone, learned by children before they touch the lead instruments, kept by dancers’ feet as faithfully as by drummers’ hands. A Western orchestra centralises time in one man’s arm because its parts cannot each hold the whole design. A tala ensemble or a bell led drum choir distributes the design until every participant carries it, and the performance can then organise itself, entry by entry and cue by cue, without command. The cycle is not only a way of measuring music. It is a way of governing it.

Ornament and the note that is not fixed

Notation freezes pitch, and most of the systems in this article do not. Between the named notes of a raga or a maqam lies the territory where much of the music actually happens: the slide that arrives at a note from below, the oscillation that worries it, the grace that flicks across it and is gone. Indian practice gathers these under the name gamaka, and the family is large. A voice or a string can approach a pitch by a slow glide, can rock around it in a measured shake, can strike it with a hammered clarity, and each choice is grammatical, specified by the raga and judged by listeners who know which ornament belongs where. The written or named note is an address. The ornament is what the resident does at home.

This is why descriptions of these traditions that begin from the score or the scale miss the point so completely. Two singers can perform the same raga, honour the same pitch set and the same characteristic phrases, and be instantly distinguishable to an informed audience, because identity lives in the ornamentation, in the weight and speed of each glide and shake. The veena and the sarod realise gamaka through the string, the bansuri through breath and half covered holes, and the voice, which every Indian tradition treats as the primary instrument, realises them most fully of all. Arab singing carries the same lesson in its own terms. The melisma that stretches a single syllable across a curving line in a performance by Umm Kulthum is not decoration applied to a melody that exists without it. Remove it and the melody, as her audience understood melody, has been removed with it.

Ornament also carries information that notation systems elsewhere assign to other elements. In Indian practice, the way a note is taken can identify the raga as surely as the note itself, since certain ragas are recognised by a characteristic oscillation on a particular degree. In Arab practice, inflection distinguishes one regional school of a maqam from another, Cairo from Aleppo, and marks a singer’s lineage as plainly as an accent marks a speaker’s. A useful way to hold all of this is to stop thinking of ornament as something added to music and to think of the fixed note as the abstraction instead. The systems in this article were built by singers and players long before they were written about, and in the building, the gesture came first. The note that is not fixed is not a failure of precision. It is precision aimed at a moving target, and the aim is the art.

Ornament is also where teaching leaves its deepest fingerprints. A student learns gamaka by singing a phrase after the teacher until the shake sits in the throat rather than in the mind, and decades later, two students of the same master can be told apart by connoisseurs who never met either man, because the ornament each absorbed in youth hardened into a personal signature neither chose consciously. Styles, gharanas and regional schools are at bottom families of ornament, which is why arguments about correct practice inside these traditions so often come down to a disagreement, passionate and precise, about how one note should be taken.

Improvisation inside a system

Improvisation is the word Western listeners reach for when they meet these traditions, and it is usually misunderstood in a flattering direction. The alap with which a Hindustani performer opens a raga looks, from outside, like freedom: no pulse, no accompaniment, no fixed duration, the melody apparently invented on the spot. What the insider hears is closer to a proof being constructed in public. The performer introduces the raga’s pitches in their permitted relationships, establishes its characteristic phrases, tests its boundaries and only gradually assembles the full melodic personality, and every step is governed by grammar absorbed over years. The improvisation is real, and it is also rule bound to a degree that a jazz musician, the Western listener’s usual comparison, would recognise as familiar discipline in a foreign accent.

The Arab taqsim makes the same point in another grammar. A taqsim is an improvised instrumental meditation in a maqam, and its art lies in exploring the maqam’s regions, touching its characteristic notes, and modulating, by shared tetrachords and expected pathways, into neighbouring maqamat and home again. A poor taqsim is not one that breaks rules theatrically. It is one that wanders without rhetoric, that cannot make its departures feel earned or its returns feel inevitable. The audience for such playing is listening for judgment, and judgment is only possible where standards are shared. Improvisation inside a system is therefore best understood as fluency. A speaker improvises every sentence of ordinary conversation, and nobody calls conversation unruled, because grammar is what makes the improvisation intelligible. The raga’s grammar, the maqam’s sayr, the tala’s cycle, the bell pattern under an Ewe drum ensemble, all do for music what grammar does for speech.

There is a consequence for the listener that the following sections rely on. Because the improvisation is grammatical, it can be followed, and following it is the central pleasure these traditions offer their own audiences. The connoisseur at an Indian concert who murmurs at a phrase is responding to a decision, a risk taken inside rules he holds in his own ear. The dancer in a Ghanaian ensemble who changes step is answering a call from the master drum that the bell pattern frames. None of this requires the newcomer to master the grammar before listening, any more than enjoying a conversation in a second language requires fluency. It requires only the knowledge that grammar is there, working, and that the performer’s freedom is the visible surface of a deep agreement between the stage and the room.

Risk is the ingredient that keeps the agreement alive, and risk requires the possibility of failure. An improvisation inside a system can genuinely fail: the phrase can fall flat, the return to the sam can be fumbled in public, a modulation in a taqsim can strand the player in a maqam he cannot gracefully leave, and the audience’s knowledge of the rules is what makes such moments dangerous rather than merely untidy. The connoisseur’s memory raises the stakes further, because tonight’s performance is heard against every remembered performance of the same raga by the same master and by his rivals and his teacher. Improvisation of this kind is therefore an argument conducted with the tradition’s whole past, in the presence of witnesses qualified to judge it, and the instruments shape the argument as surely as the grammar does. What a bansuri can risk with breath, a sarod risks with the nail’s slide and a voice with the throat’s stamina, and each performer’s invention is partly the record of what his particular body and instrument have learned to dare.

Transmission, the teacher and the lineage

Every system must solve the problem of its own survival, and the solutions are as distinctive as the pitch worlds. Where knowledge lives mainly in sound and gesture, it cannot be posted or printed into safety. It must be lodged in people, and the institutions that lodge it are among the most elaborate each culture built. The Indian solution is the guru and shishya relationship, in which a student attaches himself to a master for years, living in the master’s orbit, absorbing not lessons in the Western timetabled sense but a way of handling the music that is demonstrated, corrected and demonstrated again. The student hears the same phrase a hundred times before being permitted to build on it, and what is transmitted in that repetition is precisely what notation cannot hold: weight, timing, ornament, taste.

Lineage gives the transmission a public form. In North India the gharana, named for a town or a court, identifies a stylistic house, a cluster of techniques, repertory and aesthetic preferences handed down through a line of teachers, and a musician’s gharana is part of his name in the informed listener’s ear. Japan built a parallel institution in the iemoto system, where a hereditary head of a school licenses teachers, controls the repertory and certifies students grade by grade, so that authority over the tradition is itself inherited and administered. In the Mandé world of West Africa, the jeli, the hereditary musician and historian attached to patron families, carries repertory, genealogy and praise in memory across generations, a living archive whose training begins in childhood inside the family. Three continents, three institutions, one recognition: this knowledge is too embodied to trust to paper.

Notation, where these traditions use it, plays a humbler role than the Western reader expects. The Chinese qin was served for a thousand years by tablature that tells the player’s fingers which string to stop and how to touch it, a set of playing instructions from which a trained musician realises a sound the page never specifies in pitch terms. Indian musicians developed mnemonic syllables and, in the modern period, notation systems for teaching and archiving, but the printed page remained a reminder for the initiated rather than a score for the stranger. The contrast with the European work concept, where the score is the music and the performance its rendering, could hardly be sharper. In the systems of this article, the performance is the music, the lineage is its library, and the page, where it exists at all, is a set of directions for finding a sound that only a taught body can produce. A listener who grasps that has grasped why these traditions guard their teaching so carefully, and why the teacher’s name appears in this article nearly as often as the music’s.

The modern conservatory has tested these arrangements without replacing them. State academies in Cairo, Chennai, Istanbul and Beijing teach the systems of this article to classes, with syllabuses, printed methods and examinations, and they preserve breadth admirably: repertories are catalogued, endangered styles are documented, and students from outside the hereditary lines gain entry that the old institutions would have refused them. What the classroom transmits less reliably is depth of judgement, the master’s ear passed hand to hand, and most serious musicians in these traditions therefore pass through both worlds, taking the academy’s credentials and the lineage’s authority together. The two institutions correct each other’s characteristic failures, and the health of a system can be measured by whether it keeps both alive.

Hindustani classical music

The classical music of North India and Pakistan is the northern of India’s two great systems, and its history is a history of courts, temples and, in the modern period, the concert hall and the radio studio. Its dominant vocal form is khayal, a name meaning imagination, which over the eighteenth and nineteenth centuries displaced the older and weightier dhrupad as the central vehicle for a singer’s art. A khayal performance takes a short composed song, the bandish, and treats it as a point of departure: the singer states it, worries it, departs from it into improvised passage work of increasing speed and daring, and returns, all inside one raga and one tala. Dhrupad survives with its own austere grandeur, its long unmetered alap and its compositions in slow, majestic cycles, maintained by hereditary lines of singers and rudra veena players who regard its disciplines as the root of the whole system.

The instrumental world mirrors the vocal one. The sitar and the sarod carry the plucked string tradition, the bansuri and shehnai the wind, the violin the bowed string adopted from Europe and thoroughly naturalised, and the tabla, a pair of hand drums of extraordinary articulacy, supplies the rhythmic architecture, with the harmonium or the bowed sarangi shadowing the soloist’s line. Behind many of the twentieth century’s defining careers stands one teaching household. Allauddin Khan, the formidable master of the Maihar court, trained a generation that redrew the map of the music, among them Ravi Shankar, whose life and music are traced in his own article, and his insistence on long apprenticeship inside a single discipline became the modern template for serious study. The ethnomusicologist Bonnie Wade, whose Music in India remains a standard orientation to both Indian systems, describes the northern tradition as a sister of the southern rather than a parent or a rival, sharing deep concepts while differing in repertory, terminology and temperament.

A Hindustani concert has a shape an audience of insiders reads like a narrative. The soloist opens with alap, slow and pulseless, assembling the raga phrase by phrase until its personality stands complete in the room. Pulse enters, the composition is stated, and the performance becomes a negotiation between soloist and tabla player conducted in public, each testing the other’s memory of the cycle, each aiming to arrive at the sam, the first beat, together after excursions that seemed certain to lose it. Listeners respond aloud at a landed phrase or a daring return, and the response is part of the event, not an interruption of it. What the newcomer should take from the form is its double nature. It is at once an inheritance of fixed compositions, lineages and rules centuries deep, and an art whose centre of gravity is a single musician thinking in sound, in real time, inside everything he has inherited.

The twentieth century gave the system two new institutions that between them redrew its professional life. All India Radio auditioned and graded its artists, and a performer’s grade governed his fees and his access to the national audience, so that the state broadcaster became an examiner no musician could ignore. The same broadcaster, for more than three decades from 1940, refused the harmonium a place in its classical broadcasts, judging its fixed tuning a corruption of the raga’s intonation, a ban that drove accompanists to the sarangi and remains the favourite anecdote of everyone who teaches the period. Recording did the opposite work, carrying regional styles into one another’s territories until a national concert circuit existed where the courts had once supported separate worlds. Dhrupad, meanwhile, kept its own counsel through the century, guarded by families such as the Dagars, whose unbroken teaching line presented the oldest stratum of the system to audiences that khayal’s brilliance had nearly persuaded to forget it.

Carnatic classical music

South India built its own classical system on many of the same deep concepts, raga and tala among them, and then took them in directions the north did not follow. Where Hindustani music prizes the long improvised unfolding of a small amount of composed material, Carnatic music rests on a vast repertory of composed songs, the kriti above all, and its concerts are architectures built from compositions, with improvisation deployed at chosen points inside the structure. The repertory owes its shape to three composers of the late eighteenth and early nineteenth centuries, Tyagaraja, Muthuswami Dikshitar and Syama Sastri, remembered together as the Trinity, whose kritis, composed in Telugu, Sanskrit and Tamil, remain the centre of the concert platform. A Carnatic musician’s education is in large part the absorption of this repertory, song by song, until the compositions become the vocabulary in which improvisation thinks.

The system’s theory is correspondingly systematic. Where the north organises its ragas through tradition and a scheme of parent groupings called thaat, the south elaborated the melakarta system, a complete logical grid of seventy two parent scales from which ragas are derived, set out in the seventeenth century by the theorist Venkatamakhin and refined by later hands into the form students learn. The percussion world is led by the mridangam, a barrel drum played with the fingers and palms, supported by instruments such as the ghatam, a clay pot, and the kanjira, a small frame drum, and a Carnatic percussionist’s art of accompanying, anticipating and answering the main artist is cultivated as a discipline of its own. The violin, arriving with European contact, was adapted so completely to the system’s ornamentation that it became the standard melodic accompaniment, a transfer credited to Baluswami Dikshitar in the early nineteenth century and absorbed without fuss into a tradition that otherwise guards its boundaries closely.

The social frame differs from the north’s as well. Carnatic music grew up in the temples and courts of the Tamil, Telugu, Kannada and Malayalam speaking south, with the court of Tanjore as one of its great nurseries, and in the modern city of Chennai it sustains a concentrated concert culture whose December season fills hundreds of halls with overlapping performances for weeks. The sabha, the membership concert society, is its characteristic institution, part promoter, part academy, part club, and the young musician’s progress is measured in sabha engagements. To the ear, the differences from the north announce themselves quickly: a brighter rhythmic intricacy, compositions arriving in succession rather than a single raga occupying an evening, improvisation that favours mathematical patterning in rhythm, and a devotional directness in the repertory’s stance toward its subjects. Bonnie Wade’s comparative treatment of the two Indian systems makes the essential point plainly. They are distinct musical civilisations sharing a vocabulary of concepts, and to call one a variant of the other is to misread both.

At the summit of the Carnatic concert stands a form that displays the system’s whole character in miniature. The ragam tanam pallavi takes a single line of a composition, the pallavi, and subjects it to the full battery of the art: the raga explored in free rhythm, then in measured phrases, then the line itself varied, redivided against the tala, and tossed between soloist and accompanists in rhythmic contest. A great pallavi performance is the southern equivalent of the northern soloist’s longest negotiation with the sam, except that the material under negotiation is composed, named and shared, and the audience, many of whom keep the tala visibly on their fingers throughout, judges the arithmetic as much as the melody. Around this centrepiece the concert arrays its kritis, its devotional close, and the percussionists’ own featured exchanges, in an order that sabha audiences know by heart and would not wish altered.

Raga explained

Everything in Indian classical music turns on the raga, and the raga is the concept newcomers most often get wrong by getting it almost right. A raga is usually introduced to Western readers as a scale, and the introduction fails because a scale is a ladder of notes while a raga is a personality. Two ragas can use identical pitches and remain unmistakably different beings, because what defines each of them is not the pitch collection but the behaviour: the routes melody is permitted to travel, the phrases that act as signatures, the notes that carry weight and the notes that are passed over quickly, the ornaments that belong and those that would be a solecism. The pitch set is the cast list. The raga is the play.

What is a raga?

A raga is a melodic framework used in Indian classical music: a chosen set of pitches together with prescribed ways of moving among them, characteristic phrases, notes of emphasis and repose, and inherited associations of mood and hour. Performers improvise and compose inside it. It is closer to a personality with rules of conduct than to a scale.

The anatomy can be taken apart. A raga has an ascent and a descent, aroha and avaroha, and they need not mirror each other: a raga may climb by one route and come home by another, omitting a note upward that it touches downward, and the asymmetry is part of its face. It has a pakad, a characteristic phrase or small family of phrases by which a knowledgeable listener identifies it within seconds, the melodic equivalent of a gait recognised across a room. It assigns roles among its notes, a vadi that acts as the centre of melodic gravity and a samvadi that answers it, with other notes ranked as resting places or as colours to be touched and left. And it specifies its ornaments, so that one raga demands a slow oscillation on a particular degree while its neighbour, built from the same pitches, forbids it. The standard demonstration of all this is the pair of ragas Bhupali and Deshkar, which draw on the same pentatonic pitch set. Sing the phrases of one with the emphases of the other and an informed audience hears the error at once, though no pitch has left the shared collection.

Around this grammar the tradition wraps associations that a modern account should report as the tradition’s own teaching rather than as acoustics. Ragas are assigned to hours of the day and night, dawn ragas and evening ragas and ragas for the small hours, and to seasons, and to rasas, the aesthetic flavours of mood that Indian arts theory enumerates. A performer who offers a dawn raga at midnight has not broken a law of physics, but he has spoken out of turn inside a culture of listening, and older audiences felt the impropriety keenly. Painting literalised the idea in the raga mala tradition, where ragas and their feminine counterparts, the raginis, were portrayed as persons in scenes, a king at his devotions, a lady waiting on a terrace, as if to say that a raga is someone before it is something. For the new listener, the practical summary is plain. Do not ask which notes a raga has. Ask how it moves, what it loves, and when it is at home, and the answers will be the raga.

The two Indian systems classify the family differently, and the difference is a small lesson in how traditions think. The south derives its ragas from the seventy two melakarta parents by rule, and calls the derived ragas janya, born ones, so that every raga has a stated ancestry inside a complete logical scheme. The north groups its ragas under ten thaat in the scheme Bhatkhande devised for teaching in the early twentieth century, a practical filing that many ragas fit awkwardly and that musicians use without mistaking for a genealogy. One system wanted a proof, the other wanted a shelf, and a raga sits inside both kinds of order without being exhausted by either.

Tala explained

The tala is Indian music’s answer to the organisation of time, and it is a cycle rather than a metre. A tala has a fixed number of beats, called matras, arranged in internal groups, and the performance travels around the cycle returning again and again to its first beat, the sam, which carries the structural weight of the whole design. The most performed tala of the north is tintal, sixteen beats in four groups of four. Other cycles in common use run to seven, eight, ten, twelve and fourteen beats, and the southern system builds its cycles from a different arithmetic, combining a small set of basic tala forms with choices of subdivision to generate a large family of possible cycles. What matters is not the number but the shape of the journey: where the groups fall, which beats are stressed, which are left empty, and how the return to sam is prepared.

The cycle is made physical in two ways. Performers and listeners keep tala with the hands, clapping some group openings and waving the palm away at the khali, the empty division, so that in tintal the beat at the start of the third group is marked by an absence, a gesture of departure inside the pattern that everyone present can see. And the tabla speaks the cycle in bols, syllables that name its strokes, dha, dhin, tin, na and their combinations, which function as a spoken notation learned by recitation before it is ever played. A student can chant the theka, the standard pattern of a tala, and the chanting fixes the cycle in the body more durably than counting does. For the listener, the discipline to acquire is modest and transformative. Count the cycle quietly, in groups, and feel for the sam. Once the sam becomes an event you anticipate, the soloist’s long detours acquire their suspense, because you know what home is and you are waiting, with the whole room, for the landing.

The landing has its own artistry. Soloist and drummer may circle the sam, imply it, overshoot it, until the tension is discharged in a tihai, a phrase repeated three times with calculated spacing so that its last stroke falls precisely on the sam, at which point the room exhales. Nothing in Western metric practice corresponds closely, because the bar line is a fence the music crosses without ceremony, while the sam is a rendezvous the music keeps. Notice, finally, what the cycle does to the idea of freedom discussed earlier. The improvising soloist of the north is not decorating a beat. He is navigating a clock that he shares with the drummer and the audience, and his boldest gestures are bold because the clock is unforgiving and everyone can hear it. The cycle is the strictest thing in the music, and it is the source of the music’s characteristic exhilaration.

One refinement of the cycle deserves the listener’s attention, because it overturns the assumption that everything begins at the beginning. Many compositions are set to enter not on the sam but at a calculated offset after it, a placement the southern tradition calls eduppu, so that the melody’s own phrase cycle and the tala’s cycle run out of phase for stretches of the performance, generating a patterned friction that resolves, by design, when the refrain comes home. The listener who has learned to feel the sam can then enjoy the higher pleasure of hearing two clocks disagree on purpose, and can understand why musicians speak of the tala not as a cage but as a partner, an independent order against which a melody defines itself.

The Arab maqam

The Arab world organises melody through the maqam, a system whose family resemblance to the raga is real and whose differences are the education. A maqam is not a scale, though it contains one. It is a melodic world with its own routes, gravities and customs, built from smaller units called ajnas, melodic cells of three, four or five notes that join at shared pitches to form the maqam’s full span. Each maqam has a sayr, a characteristic course that its melodies tend to travel, regions of the span where phrases gather, notes that act as resting points, and approved doorways into neighbouring maqamat, so that a skilled improviser can modulate and return by paths the informed listener recognises as the music’s geography. To learn a maqam is to learn a territory, not a ladder.

What is a maqam?

A maqam is the system that organises melody in Arab art music: a collection of pitches built from smaller units called ajnas, with a characteristic route that melody tends to travel, notes of emphasis and rest, and customary ways of moving into neighbouring maqamat. An improviser explores it the way a speaker uses a language, freely and within grammar.

The pitch material is where the maqam announces its independence most plainly. Several of its essential intervals fall between the keys of a piano, most famously the half flat steps, roughly three quarters of a tone, that define the colour of maqam rast, the tradition’s central maqam, and of bayati, the maqam of a thousand folk and classical melodies. These pitches are stable, taught, and judged with precision by the musicians who use them, and the theory that describes them has a long written history, from the medieval treatises of al Farabi onward. In the nineteenth century the Lebanese theorist Mikhail Mashaqa attempted a systematic account that divided the octave into twenty four equal quarter tones, a grid the music does not quite obey in practice, and in 1932 a congress of Arab music in Cairo, attended by musicians and scholars from across the Arab world and Europe, tried to settle questions of scale, notation and instrument building for a recording age. The argument the congress started, between fixing the system for modern institutions and keeping its flexible intonation, has never entirely closed, and it illustrates a theme of this article: systems meet modernity on their own terms or not at all.

Regional schools inflect the shared system. Egyptian practice, Levantine practice and the Iraqi tradition, whose maqam is also the name of a distinct sung genre with its own repertory and ensemble, differ in intonation habits, in favoured maqamat and in the weight given to improvisation against composition. The large forms built on the system include the wasla, a suite of pieces in one maqam, the sung qasida and the instrumental taqsim discussed earlier, and the great long songs of the Egyptian stage examined in the next section. For the newcomer, the handhold is the analogy with raga, used carefully. Both systems make melody carry the full expressive load, both are learned as behaviour rather than as pitch lists, and both reward the listener who stops asking what the notes are and starts asking where the melody is going and why it lingers where it does. The differences, of interval, of phrase, of the role of the ensemble, are exactly what make the comparison worth more than a sentence.

The maqam also governs the region’s sacred sound, which is part of why its grammar sits so deep in its listeners. Quranic recitation in the Arab world is performed with maqam melody under the rules of tajwid, and the muezzin’s call to prayer is sung in a maqam chosen for the hour and the place, while the pizmonim, the Hebrew hymns of the Syrian Jewish tradition, are assigned their maqam week by week in a published cycle. A child in Cairo or Aleppo absorbs the system’s intervals with language itself, long before anyone names a tetrachord, and the concert improviser’s art is to make that childhood knowledge speak at the highest level of reflection.

The Egyptian concert form

Egypt in the twentieth century built the Arab world’s most influential concert culture, and at its centre stood a form with few parallels anywhere: the long song, the ughniya, performed by a single singer with a large ensemble before an audience prepared to give it an entire evening. The older Arab concert had been the wasla, a suite of vocal and instrumental pieces in a single maqam, assembled in a recognised order and sized for the salon and the wedding hall. Radio, the record industry and the modern theatre stretched the suite’s logic until the evening contracted around one voice and one vast song, and the singer who defined the form was Umm Kulthum, whose concerts became national appointments in Egypt and, by broadcast, across the Arab world.

Her mature performances were built for duration. A concert might present two or three songs across an evening, and a single song, composed for her by Riyad al Sunbati or, from 1964, by Mohammed Abdel Wahab, could extend past an hour as she repeated lines, rephrased them, and answered the audience’s demands for more. The aesthetic governing this exchange has its own name, tarab, the state of moved absorption that singer and listeners build together, and it explains the form’s economics of repetition. A line sung once states the words. Sung again with altered inflection, after the hall has called for it, the line becomes an event the room has co authored, and the singer’s art lies in finding the new emphasis each return can bear. Virginia Danielson, the scholar whose study of Umm Kulthum and Egyptian society is the standard account in English, shows how these concerts functioned as a public ritual of a whole culture, carried live on Egyptian radio on the first Thursday of each month through the concert season, with streets and cafes organised around the broadcast.

The ensemble behind such a singer is a study in the system’s values. Violins in unison, cellos, the ud, the qanun zither, the nay flute and a percussion section articulate the composed fabric, while the singer moves within and against it, and the finest players were prized for the tact of their support, knowing when to fill a pause with a figure that hands the melody back enriched. Nothing in the form is improvised in the free sense and nothing is merely read. It is composed music held open at chosen joints so that the evening’s tarab can enter. For a listener arriving from the Indian sections of this article, the parallel and the contrast are both instructive. Cairo, like Delhi, built its concert art on a single melodic intelligence working over a stable modal ground. Cairo chose the composed song stretched by repetition and response, where the Hindustani stage chose the raga explored by invention, and the difference is a difference of temperament between two complete civilisations of listening.

The form’s personnel changed as its scale grew, and the change is written into the sound. The older takht, a handful of soloists on ud, qanun, nay, violin and riqq, accompanied singers in rooms where every player was audible as an individual. The radio age enlarged the takht into the firqa, a ranked string section of a size borrowed from European orchestras, and the balance of the art shifted: the singer now rode on a composed carpet, the instrumental interludes grew more arranged, and tarab had to be built at architectural scale rather than kindled in a circle. Danielson’s account is alert to what the audience policed through all of it. Listeners who had grown up with the takht judged the firqa’s precision against the older responsiveness, and the great singers were those who could make forty players breathe like five.

Qawwali and devotional practice

In the shrines and courtyards of South Asia, one musical form exists whose purpose is not performance in the concert sense at all. Qawwali is the devotional music of Sufi gatherings, sung at the assemblies called mehfil e sama, listening assemblies, held at the shrines of saints, above all at the shrine of Nizamuddin Auliya in Delhi, where the practice has been continuous for centuries. Its aim is the arousal of devotion, the turning of the listener toward God through sung poetry in Persian, Urdu, Hindi and Punjabi, and its musicians, the qawwals, are hereditary specialists whose art is a religious office as much as a profession. The tradition credits the thirteenth and fourteenth century poet and musician Amir Khusrau, disciple of Nizamuddin, with shaping much of the form’s repertory and method, and qawwals have honoured that descent ever since.

What is qawwali?

Qawwali is the devotional music of Sufi gatherings in South Asia, performed by a lead singer and party at shrines and assemblies. Built on repetition, handclaps and rising rhythmic intensity, it carries sacred poetry toward collective ecstasy, and its audience participates with responses, requests and offerings rather than listening in silence.

The performing party is built for escalation. A lead qawwal directs from the harmonium, a second singer supports and answers, a chorus repeats and reinforces, and tabla, dholak and massed handclaps lay down a rhythm that can tighten over the course of an item from processional calm to a driving intensity that carries the room with it. The performance is steered in real time. The lead singer watches the gathering, repeats a line that has struck the listeners, extends it, varies it, and chooses the next item according to the state of the assembly and the occasion, a saint’s festival, a wedding, a night of private devotion. Listeners respond as participants: they offer money at moments of intensity, presented to the singers and often dedicated onward to the shrine, they call out, they weep, and the boundary between performer and audience, absolute in the Western concert hall, is deliberately porous here because the theology of the gathering requires it. The ecstasy aimed at is communal property, not a soloist’s display.

The form’s modern history carried it far beyond the shrine without severing it from its root. Recording and film took qawwali across South Asia, concert promoters took it to the world’s stages, and the party led by Nusrat Fateh Ali Khan, whose place in this story is taken up later in this article, made the sound familiar to listeners who had never heard of Nizamuddin. What travelled was not a diluted essence but the form itself, its personnel, its escalation and its repertory of praise, set down in halls where the audience’s role had to be relearned by people encountering the convention for the first time. Qawwali therefore offers the clearest case in this article of a system whose meaning includes its setting. Move it and it survives, because its grammar is strong, but the grammar includes the listeners, and a recording of qawwali, however fine, is a document of a gathering whose other half is missing.

The Persian dastgah

Persian art music organises its melody through the dastgah, and it preserves its knowledge in one of the most remarkable teaching objects in any tradition: the radif. The radif is a fixed repertory of several hundred short model melodies, the gusheh, arranged inside the dastgah system and memorised entire by the student over years of study. To know the radif is to hold the tradition’s melodic vocabulary in the body, and improvisation in Persian music is the art of drawing on that vocabulary, sequencing its models, developing them and returning among them, so that a performance is a meditation conducted in inherited phrases. The version of the radif that anchors modern practice descends from Mirza Abdollah, the master of the tar and setar who died in 1918, whose ordering of the material was taken down by his students and became the canon against which later versions are measured.

The system built on the radif comprises seven dastgah and five avaz, related modal entities rather than scales in the Western sense, each with its own melodic tendencies, its own characteristic gusheh and its own expressive associations. Shur is generally treated as the mother of the system, the dastgah from which several of the avaz derive, and a student’s progress through the radif is also a progress through this family, learning which regions of melody belong together and how a performer may travel among them. The instruments of the tradition are intimate in scale: the tar, a long necked lute with a double bowl body, the setar, its smaller and quieter relative, the santur, a hammered dulcimer, the kamancheh spike fiddle and the nay, played in small ensembles or alone, in rooms where the music’s dynamic range, from near silence to sudden brilliance, makes sense.

What distinguishes the Persian case within this article’s survey is the explicitness of its bargain between memory and invention. The Indian improviser works inside a grammar; the Persian improviser works inside a memorised book of models, and the artistry lies in the handling, in how a gusheh is entered, lingered over, combined with its neighbour and left. Critics inside the tradition have argued for generations about where preservation ends and creativity begins, and about whether the radif’s authority shelters the music or constrains it, an argument conducted in the open by named masters and scholars and never settled, because the system itself holds both values at once. It wants the past kept exactly and the present performance alive, and it accepts the tension as the price of having both.

Singing occupies the system’s summit, and its modern master set the standard by which the art is judged. Mohammad Reza Shajarian, who rose to prominence in the 1970s, built his singing on a complete command of the radif joined to a voice of exceptional clarity, and his performances of the avaz, the improvised song that unfolds a dastgah through its gusheh, are treated by Iranian listeners the way a great khayal singer’s performances are treated in India: as the system thinking aloud. His art also displays the tradition’s signature ornament, the tahrir, a rapid oscillation produced in the throat that breaks a sustained note into shimmering pulses, an effect no other system in this article produces in quite that form and one that identifies a Persian singer within a phrase.

Turkish and Ottoman art music

The Ottoman court and the Sufi lodges of Anatolia and the Balkans developed an art music that shares a root with the Arab maqam and grew into a distinct system with its own name, makam, its own theory and its own repertory. The kinship is visible in the modal principle and in shared instrument names, and the differences are just as instructive. Turkish theory, codified in the twentieth century by Hüseyin Sadettin Arel and Suphi Ezgi, analysed the octave into fine divisions, fifty three commas in the theoretical count, and assigned the makamlar precise interval structures on paper that practice then inflects, a settlement between system and ear that runs through the tradition’s modern history. Turkish makamlar also carry strong conventions of seyir, melodic progression, with a makam characterised as much by whether its phrases gather in the upper register and descend, or build upward from the tonic, as by its pitch collection.

The forms are courtly in architecture. The fasıl is a suite of vocal and instrumental pieces in one makam, ordered from the improvisatory taksim at its opening through composed songs of increasing rhythmic animation toward dance pieces at its close. Instrumental composition reached a high polish in the peşrev, a multi section piece whose returning refrain organises long spans, and the Mevlevi order, the whirling dervishes of Konya, built the ayin, a composed ceremonial cycle of great length and gravity, for its rite of turning and listening. The rhythmic cycles, the usul, include some of the longest in any system surveyed here, patterns of twenty eight beats and more that the kudüm drummer articulates and the ensemble holds in common memory. The instruments declare the sound: the tanbur, a long necked lute with movable frets tied for the makam’s intervals, the ney, an end blown reed flute whose breathy attack is the tradition’s emblem, the ud, and the kemençe.

Ottoman music also demonstrates how a system crosses the boundary between court, lodge and city. The same makam might govern a sultan’s entertainment, a Mevlevi ceremony and an urban song, and the musicians moved among those settings carrying the repertory with them. In the twentieth century the Turkish radio became the system’s great conservatory and its great standardiser, employing the masters, broadcasting the repertory and settling disputed practice by the authority of the microphone. The tradition that resulted is at once deeply continuous with the Ottoman centuries and unmistakably modern in its institutions, a combination that the Arab sections of this article will recognise, and that the recording industries section below examines as a general pattern.

West African rhythmic practice

The drumming traditions of West Africa organise ensembles in a way that reverses the Western assumption about who leads. In the familiar European arrangement, a conductor or a lead part carries the design and the rest of the ensemble fills it out. In an Ewe ensemble from Ghana, to take the tradition most documented for outside readers, the design is distributed. A bell, the gankeke, plays a short pattern of seven strokes distributed across a twelve pulse cycle, and it plays that pattern without variation for the entire piece, because the pattern is the ground on which everyone else stands. Rattles and support drums add their own interlocking figures, each simple in itself, each landing in the gaps of the others, and only the master drum, the atsimevu, is free, and its freedom is conversational: it calls, comments, signals changes and answers the dancers.

The result is a texture in which complexity is an emergent property. No single part, heard alone, contains the music, and the beginner who tries to follow one line at a time will lose the whole, which lives in the relationships. The standard way into such a texture is through the bell, since the bell’s phrase is the shared reference that lets every other player, and every dancer, locate the same moment in the cycle. Dance is not an accompaniment to this music but one of its parts. The master drummer watches the dancers and plays to their feet, a lead dancer’s step triggers a drum call, the call reshapes the step, and the exchange is the performance’s living centre. John Chernoff, an American ethnomusicologist who learned to drum in Ghana, built his influential account of the music on exactly this observation: that the aesthetic values at work are social values, participation, responsiveness and the tension of parts held in productive disagreement, made audible.

The Mandé world to the west and north adds the hereditary musician to the picture. Among the Mandé peoples, the jeli, called griot in French, is a professional of music and memory, trained from childhood inside a family line, attached to patron families whose genealogies, praises and histories he carries. His instruments are the kora, a twenty one string harp lute whose rippling patterns support long sung narratives, the balafon, a wooden xylophone with gourd resonators, and the ngoni lute. Where Ewe drumming organises a community’s dancing, Mandé jeliya organises a community’s memory, and the two traditions together correct any notion that West African music is a single thing called drumming. It is a family of systems for doing different cultural work, rhythmically intricate in all its branches, and as diverse in function as the societies that built it.

The ensemble’s calendar explains its personnel. Ewe drumming belongs to occasions, funerals above all, festivals, and the recreational societies whose dances carry names, histories and costumes of their own, and a village’s drummers are office holders in that social machinery rather than entertainers hired into it. Women lead much of the singing, set the songs the drummers answer, and keep the rattle and handclap parts on which the texture’s lower storey stands, so that the celebrated interlock is a picture of the community itself, each constituency audible, none sufficient alone. Children enter the system at its simplest point, the bell or the rattle, and graduate across years toward the support drums and, for the few with the gift and the memory, the master drum. The training ladder is the social ladder, and a listener who watches who plays what, and who is still learning, is reading the society’s organisation chart rendered in rhythm.

Speech-tone instruments and the talking drum

Many West African languages are tonal: the pitch contour of a syllable helps decide what the word means, so that the same string of consonants and vowels, spoken high or falling, can name different things. This single linguistic fact produced one of the world’s most distinctive instruments. If pitch carries meaning in speech, then an instrument that can bend pitch can be made to speak. The Yoruba dùndún ensemble is led by the ìyá ìlù, the mother drum, an hourglass shaped pressure drum whose cords the player squeezes while striking, bending the pitch up and down in the contours of Yoruba speech. In skilled hands the drum reproduces the tone and rhythm of phrases, and listeners who know the language, and know the stock of proverbs, praise names and histories the drummer draws on, hear utterance, not imitation of utterance.

The drum’s repertory is the culture’s verbal treasury in another form. Praise poetry for lineages and patrons, proverbs deployed in argument, announcements and calls, all can be drummed, and in earlier centuries the Akan atumpan, a pair of kettle drums played by a master drummer at court, served as the chief’s speaking instrument, its phrases carrying statements across the courtyard and, in the days before amplification, across distances no unaided voice could cross. European observers named these instruments talking drums and the name, though loose, catches the essential point: the drumming is a channel of language, with grammar and idiom, and it can be misunderstood, quoted and answered like language.

The limits are as instructive as the powers. Drum speech works because the listener supplies most of the information. The drum gives tone contour and rhythm without consonants, and a hearer who does not already know the phrase, or the family of phrases it must belong to, hears elegant drumming and nothing more. Praise names are therefore conventional, proverb openings are signalled by context, and the drummer’s art includes judging how much the audience can complete. It is a communication system that runs on shared memory, which is why it belongs in this article beside the lineages and the radif: in each case the tradition stores its knowledge in people, and the instrument, whether it is a drum that speaks or a lute that carries a memorised canon, is only ever as literate as the community listening to it.

The principle outlived the era of long distance signalling and entered the recording studio intact. Yoruba popular music kept the talking drum at its centre: the juju bands of Lagos, and preeminently those of King Sunny Ade, weave dùndún phrases through electric guitars and synthesisers, and the phrases still speak, offering praise and comment that Yoruba listeners in the audience decode with pleasure while everyone else hears superb percussion. In Senegal the same hourglass drum, the tama, became the chattering voice of mbalax. A technology of language built for the courtyard proved, without changing its principle, to be a technology of commentary for the dance floor, which is as neat a demonstration as this article can offer of how a system survives: not by freezing its repertory, but by carrying its grammar into whatever room the age provides.

The diaspora transfer to the Americas

Between the sixteenth and nineteenth centuries, the Atlantic slave trade carried millions of Africans to the Americas, and carried their music in the only vessel available, memory. Instruments could be confiscated and were: after the Stono rebellion of 1739, South Carolina moved to ban the drums whose signals the authorities feared, and other slaveholding jurisdictions followed with restrictions of their own. What crossed anyway was deeper than any instrument. Rhythmic organisation built on interlocking parts and a shared timeline, call and response as a social form, the union of drumming with dance and ceremony, and ways of bending pitch and timing that European notation had no symbols for: these travelled in bodies and rebuilt themselves on new materials wherever communities could gather.

The survivals trace the transfer with unusual clarity. In the Sea Islands of Georgia and South Carolina, the ring shout preserved African sacred practice inside Christian worship, a counterclockwise circle of singers and shuffling feet that observers documented across centuries. In New Orleans, the Sunday gatherings of enslaved people at Congo Square kept drumming and dance in public view into the middle of the nineteenth century, in a city whose later music would carry the imprint worldwide. In Cuba, the batá drums of Lucumí religious practice retained Yoruba liturgical structure, repertory and language in consecrated form, and in Brazil the Candomblé houses preserved African nations’ drumming and song as the organising art of worship. Each case shows the same pattern: a system under maximum pressure keeps its grammar and rebuilds its surface, because the grammar was stored where it could not be confiscated.

The return journey matters as well. American and Caribbean recordings sailed back across the Atlantic in the twentieth century and were heard in Lagos, Accra and Kinshasa as news from relatives, feeding the modern styles examined in the next sections of this article. The transfer was never one way for long, and the longer account of how musicians, instruments and repertories moved around the world belongs with the series survey of music and migration through history, which takes up the migrations in their own right. What this article needs from the story is the mechanism. A musical system is portable because it is a set of agreements about pitch, time and conduct, and agreements can survive journeys that objects cannot. The bin that the record trade built in 1987 gathered the descendants of these crossings together with the traditions they came from, and labelled the whole gathering with two words that described none of it.

Instruments tell the same story when they are examined closely. The banjo, that emblem of American vernacular music, descends from the gourd bodied lutes of West Africa, rebuilt in the Americas from memory by enslaved craftsmen who had no pattern to copy but the one in their hands, and eighteenth century observers in the Caribbean described its ancestors under local names before the instrument acquired the form the world knows. The batá drums of Cuba press the point further: a consecrated set is not merely an instrument but a sacred presence, with its own name and its own protocols, carried across the Atlantic as a person is carried, by being remade without being redesigned. What the trade routes and the slave routes alike demonstrate is that the unit of musical survival is never the object. It is the agreement, about intervals, about cycles, about what an instrument is for, and agreements travel light.

Afrobeat as a modern synthesis

Afrobeat is the standing refutation of the idea that the traditions in the world music bin are anonymous inheritances. It has an author, a date and an address. The author was Fela Anikulapo Kuti, a Nigerian bandleader trained in London, formed in the highlife bands of Lagos, and converted as a musician and a citizen by a tour of the United States in 1969 and 1970, where his band heard American soul and jazz at close range and where he absorbed the politics of the Black Power movement. Returning to Lagos, he rebuilt his group as Africa 70, opened a club called the Shrine, and over the following years assembled a music that had not existed before: highlife melody, jazz harmony and soloing, the interlocking guitar figures of West African dance music, Yoruba rhythm and chant, and the groove discipline of American funk, fused into performances that could run for an hour on a single composition.

The synthesis had a co author whose name belongs in any honest account. Tony Allen, the band’s drummer, built the rhythmic engine, adapting the interlock of traditional drumming and the drive of the American kit players he admired into a style of playing that kept several patterns in motion at once while never hurrying, and Fela himself said in substance, repeatedly and in print, that without Allen there would have been no Afrobeat. Michael Veal, the American scholar whose life of Fela is the fullest account of the music’s making, shows how deliberately the form was engineered: baritone saxophones and a bank of guitars laying down repeating figures, keyboards and horns stating themes of symphonic patience, and over it Fela’s pidgin English singing, chosen so that his attacks on Nigeria’s military rulers and their civilian collaborators would be understood far beyond the Yoruba audience. The music’s length was part of the politics. It took the time it took, in defiance of the three minute single and of the state broadcaster alike.

Afrobeat’s confrontation with the Nigerian state was not a metaphor. Fela declared his communal compound, Kalakuta Republic, a separate sovereignty, and in 1977, after a song that mocked the army by name, soldiers raided and burned Kalakuta, an assault documented by journalists and by the official inquiry that followed, and one in which his mother, the veteran nationalist Funmilayo Ransome Kuti, was injured. The music that came out of the compound, confrontational, ecstatic and engineered for dancing, stands in this article as a modern, authored art built knowingly from traditional materials. To file it as world music in the sense of anonymous folk inheritance is to misdescribe it twice over. It is a twentieth century composition with named makers, and its roots in the older systems are a matter of craft, not of fate.

The form’s internal history after Allen’s departure at the end of the 1970s underlines how authored the synthesis was. Fela reorganised, fronted the band himself on saxophone and keyboards with the group renamed Egypt 80, and the music’s centre of gravity shifted toward his voice and his pronouncements, as if the engine’s designer had chosen to become its driver as well. Disciples and sons took the style abroad in later decades, and producers in Europe and America sampled and studied its patterns, but the repertory remained a catalogue of compositions attached to one man’s band, one compound and one confrontation with his government. Afrobeat is filed in this survey between the inherited systems on purpose. It shows the exact point at which tradition becomes authorship, and how much of the old grammar an author can carry into a music that did not exist before he made it.

Brazilian forms and bossa nova

Brazil’s place in this survey rests on a musical culture that absorbed African, European and indigenous elements under the pressure of the same Atlantic history examined earlier, and produced, among much else, two forms the whole world learned. Samba came first, gathering in Rio de Janeiro in the early twentieth century out of Afro Brazilian religious and social life, and organising itself around carnival, the neighbourhood escolas de samba, and a percussion architecture of great sophistication, in which the deep surdo drums mark the pulse and a battery of smaller instruments fills the spaces with interlocking figures. Choro, older still, kept an instrumental art of virtuoso improvisation alive in small ensembles of flute, guitar and cavaquinho, a chamber tradition that the record business largely ignored and musicians treasured.

Bossa nova, the new beat, arrived at the end of the 1950s as a deliberate simplification and deepening of samba by a small circle of Rio musicians and poets. João Gilberto, a singer and guitarist from Bahia, condensed the samba percussion battery into a single guitar pattern, a syncopated figure played with the fingers that implies the whole ensemble, and matched it with a voice of unprecedented quietness, singing just behind the beat as if in confidence. Antonio Carlos Jobim supplied the compositions, harmonically restless songs whose chord movements drew on both samba and the American songbook, and Vinicius de Moraes supplied many of the words. The record usually taken as the founding document, João Gilberto’s Chega de Saudade of 1958, announced the style complete, and Ruy Castro, the Brazilian journalist whose book on the bossa nova years is the movement’s standard chronicle, reconstructs the scene apartment by apartment: a few dozen young people in Rio’s south zone, trading records, chords and songs, largely ignored by the established industry until the music was already made.

The world’s route to bossa nova ran through jazz. Stan Getz and Charlie Byrd recorded Jazz Samba in 1962, Getz and Gilberto followed in 1964 with Jobim at the piano, and the style became, for a decade, the sound of cosmopolitan ease in North America and Europe. The ease was a mishearing. Underneath the soft surfaces sit harmonic movements of real sophistication and a rhythmic placement that punishes imitation, since Gilberto’s guitar figure and vocal delay are easy to copy badly and very hard to inhabit. Bossa nova matters to this article’s argument because it shows a tradition doing in reverse what the world music bin does by filing. It took specific local materials, samba rhythm, Portuguese song, Rio’s apartment culture, and made from them a form of such internal completeness that musicians everywhere adopted it as a system, with its own standards of error and its own masters.

The guitar figure deserves a closer description, because the whole style can be heard inside it. In a samba school, the pulse is distributed: the big surdo drums state the beat, cuícas, tambourines and agogô bells fill the subdivisions, and the harmony instruments move through the texture the drummers build. Gilberto’s right hand took that distribution onto one instrument. His thumb keeps a bass line that behaves like the surdo, stating the harmony’s roots on the pulse, while his fingers brush chords in syncopations that answer the thumb a fraction late, so that a single guitar implies a percussion section in conversation with a bass. Singing across that pattern, slightly behind where the ear expects the words to fall, he produced the impression of a man thinking aloud in rhythm. The technique is teachable and is taught, in Rio and far beyond it, which is the mark of a system: it can be handed on, judged and failed at.

Tango and the Argentine line

Tango was born in the port districts of Buenos Aires and Montevideo in the late nineteenth century, in a population of immigrants, labourers and musicians thrown together by the Atlantic economy, and it carries its parentage in its rhythms: the habanera pattern arriving from Cuba by way of Spain, the milonga of the pampas, African drumming survivals in the candombe of Uruguay, and European salon dances, all compressed into a music for close partnered dancing in crowded rooms. Its emblematic instrument arrived by accident. The bandoneón, a large button accordion, was designed in Germany by Heinrich Band in the 1840s as a portable substitute for the church organ in poor parishes. Exported to the River Plate, it found its destiny in tango, where its wheezing, vocal attack became the sound of the form’s longing, and Argentine players made it a virtuoso instrument of formidable difficulty.

For its first decades tango was instrumental dance music, played by small orquestas típicas of bandoneóns, violins, piano and bass. The sung tango, the tango canción, dates from 1917, when Carlos Gardel recorded Mi Noche Triste and showed that the dance’s melancholy could be carried by a voice and a lyric of the streets, written in the lunfardo slang of Buenos Aires. Gardel became the form’s first international star, its defining singer and its first recording export, and his death in an air crash in 1935, at the height of his fame, fixed him in the Argentine imagination as the voice of the city’s golden age. The dance halls of the 1940s then raised the great orchestras, each with a rhythmic signature dancers could identify in bars, and tango’s social architecture, the milonga evening with its codes of invitation and its tanda sets of dances, became as systematic as any raga’s etiquette.

The system’s modern argument arrived with Astor Piazzolla, a bandoneonist formed in the dance orchestras who studied composition with Nadia Boulanger in Paris and returned determined to write tango for the concert hall. His nuevo tango kept the bandoneón, the rhythmic cells and the melancholy, and rebuilt the harmony and form around them, to the fury of traditionalists who held that tango you cannot dance to has broken the contract. The dispute is this article’s recurring question in Argentine dress: whether a system’s identity lives in its social function or in its materials. Piazzolla’s music prevailed in the world’s concert halls while the dance floors kept the old orchestras, and the system, like a language with two registers, learned to hold both.

The dance orchestras of the golden age repay a listener’s attention the way a gharana does, because each was a school with a signature the dancers read instantly. Juan D’Arienzo drove his band at a bright, relentless tempo that filled floors and earned him the title of king of the beat. Aníbal Troilo, himself a bandoneonist of the first rank, built a warmer and more singing sound, the orchestra as a choir of phrases. Osvaldo Pugliese developed a dramatic rhythmic stamp, a heavy marcato his followers named la yumba, and his orchestra’s intensity made it a byword for tango played as if everything depended on it. A milonga evening that programmes all three in sequence is teaching its dancers the system’s internal variety, three grammars of the same embrace, and regulars argue their allegiances with the passion of listeners anywhere who have learned to hear difference inside a shared form.

Son, salsa and the Cuban line

Cuban son is the root form of a family that covers half the dance music of the Americas. It grew in the eastern province of Oriente in the late nineteenth century, in the hills where Spanish song and guitar met African drumming and call and response, and its classic instrument is the tres, a small guitar with three doubled courses of strings that plays the repeating melodic figures, the guajeos, around which the music turns. A son ensemble sets those figures against the clave, the five stroke pattern, in two halves of three and two strokes, that every musician in the band must agree with at every moment. Clave is not a rhythm in the decorative sense. It is the key of the whole architecture, the fixed reference against which phrases lean, and a player whose phrase crosses it has made the Cuban equivalent of a wrong note in a raga: an error of orientation that every insider hears.

The form’s great moderniser was Arsenio Rodríguez, a blind tres player and composer who in the 1940s expanded the septeto into the conjunto, adding the conga drum, a piano and extra trumpets, and deepening the music’s African rhythmic base while sharpening its arrangements. From Havana the family travelled. Mambo, built by Pérez Prado and the Havana arrangers on the son’s chassis with big band brass, conquered Mexico City and New York in the 1950s, and in New York the music of Cuban and Puerto Rican dance bands converged into what the trade, in the 1970s, learned to sell as salsa. The word was another label decision, like the one in the London pub: salsa named no new system, it named a market, and its great institution, Fania Records, founded in 1964 by Johnny Pacheco and Jerry Masucci, gathered the city’s bandleaders under one roof and sold their music to the world as a single hot category.

The voice the category could not contain was Celia Cruz, who had sung with Havana’s Sonora Matancera before leaving Cuba in 1960 and who became, with Fania’s bands and in her own right, the defining singer of the music, a performer whose rhythmic precision and improvising choruses set the standard her peers measured themselves against; her life and music are traced in her own article. Peter Manuel, the ethnomusicologist whose work on Caribbean music is a standard reference, reads the Cuban story as a demonstration of how dance musics survive the industries that package them: the labels change every generation, son, mambo, salsa, while the clave, the guajeo and the call and response persist underneath, transmitted player to player in rehearsal rooms. It is the same lesson the Indian and Arab sections taught, in dance tempo. Systems outlive the words used to sell them.

The salsa rhythm section shows the family’s grammar at its most concentrated. The piano plays the montuno, a syncopated chordal figure that circles the harmony without stating it plainly, the bass answers with the tumbao, a pattern that characteristically avoids the downbeat and leans into the next chord before it arrives, and congas and timbales divide the remaining space between foundation and firework. Singers improvise the soneos, the call and response choruses against the band’s coro, in the direct line of descent from the son’s estribillo. Every element is an inheritance with a name and a function, and a young player in New York or Cali learns the style the way a tala student learns the theka, by absorbing the patterns from masters until his own variations become possible. The label on the record sleeve is the last thing the music is about.

Chinese classical traditions

China’s classical traditions begin, for this article’s purposes, with an instrument that embodies the argument about notation made earlier. The guqin, a seven string zither associated for two thousand years with scholars and recluses, is played from tablature called jianzipu, a notation that records which string to sound, where to stop it and which finger technique to use, while leaving the resulting pitch to be realised by a player who already knows the style. The earliest surviving handbook of this repertory, the Shenqi Mipu of 1425, preserves pieces attributed to much earlier centuries, transmitted hand to hand and fixed in print without ever becoming self sufficient: a stranger with the book and no teacher owns instructions for music he cannot yet hear. The qin’s aesthetic prizes the sounds between the notes, slides, harmonics and silences, and its repertory of programmatic pieces, evoking water, mountains and historical encounters, makes it the most explicitly philosophical instrument in this survey.

Around the qin stands a wider civilisation of sound. The pipa, a short necked lute held upright, carries narrative and martial repertories of brilliant attack. The dizi and xiao flutes, the sheng mouth organ, whose free reeds influenced the European harmonium and accordion, and the erhu, a two string fiddle that entered Chinese music from Central Asia, all serve regional ensembles such as the Jiangnan sizhu, the silk and bamboo clubs of the Shanghai region, where amateurs gather to play a shared repertory in heterophonic texture, each instrument ornamenting the same melody in its own way. Peking opera built a theatre music of strident colour, led by the high jinghu fiddle and an assertive percussion battery whose patterns cue actors’ movements and fights, while at the ceremonial summit stood yayue, the ritual music of court and Confucian temple, where the correct performance of ancient pieces was itself an act of government. Confucian thought held that a state’s music revealed its moral condition, an idea with no real parallel elsewhere in this article: the system as a branch of statecraft.

Two features distinguish the Chinese case for the comparative reader. The first is the depth of the written record, treatises, tablatures and dynastic histories of music reaching back more than two millennia, which gives the historian documents where other traditions offer only memory. The second is heterophony as an ensemble ideal: where Western practice prizes agreement among parts, the sizhu texture prizes simultaneous individuality, every player realising the common melody through personal ornament, so that the ensemble sounds like a conversation about a tune rather than a rendering of it. Both features prepare the listener for how differently a system can distribute the same musical labour.

The pipa’s repertory makes a related point about classification. Its pieces are traditionally divided into wen, civil works of lyric character, and wu, martial works of narrative violence, in which the player deploys a battery of percussive techniques, snapped strings, rolled tremolos and imitations of battle, to tell of sieges and ambushes without a word sung. A single instrument thus holds both the salon and the battlefield, and the division is not between pieces but between hands: the same techniques, weighted differently, produce the intimate and the spectacular. Chinese musical thought is comfortable with such pairings, yin beside yang, and a listener who has grasped the wen and wu of the pipa has a key to the aesthetic of balance that runs from the qin’s quietest harmonics to the opera stage’s loudest gongs.

Japanese classical traditions

Japan preserved at its imperial court a repertory that its musicians describe as among the oldest continuously performed orchestral musics in the world. Gagaku, elegant music, arrived from China and Korea more than a thousand years ago and was maintained by hereditary court musicians through every change of regime, so that its slow, stately pieces, the togaku repertory traced to Tang dynasty China, its wind and string sonorities and its sho mouth organ sustaining motionless clusters of tone, constitute a living archive of an Asia that vanished elsewhere. Preservation on that scale required an institution, and gagaku’s musicians supplied it: families attached to the court and the great shrines, teaching by oral transmission within the household, guarding tempo and gesture as carefully as pitch.

The shakuhachi, an end blown bamboo flute, took the opposite path, from ensemble service into solitude. In the Edo period the Fuke sect of Zen Buddhism adopted the instrument as a tool of meditation, and its komuso monks, licensed by the state and travelling the country with baskets covering their heads, developed the honkyoku, solo pieces of extraordinary breath control and pitch flexibility, in which the noise of the breath is part of the music and a single tone, bent and shaded, can occupy a minute of attention. The practice they called suizen, blowing Zen, treats the instrument the way the qin treats philosophy, as a discipline of the person before it is a performance for others. Beside these stand the narrative and chamber traditions: the biwa carrying epic tales, the koto zither and the shamisen lute joining the shakuhachi in sankyoku trio music, and the theatre musics of noh and kabuki, each with its own instruments, cues and hereditary schools.

Transmission in Japan was systematised through the iemoto arrangement met earlier in this article, where the head of a school, the iemoto, holds authority over repertory, teaching licences and the right to perform under the school’s name. Students purchase progression through graded pieces, receive names and permissions, and owe loyalty upward through the structure. The system preserved styles with remarkable fidelity across centuries and also froze them, since innovation required the founder of a new school or the permission of the old one. For this article’s comparative argument, Japan offers the clearest case of transmission as deliberate policy: a culture that decided its classical musics were treasures to be administered, and built the administration to match.

The theatre shows the administration at work in miniature. Noh, the masked drama preserved since the fourteenth century, is accompanied by the hayashi ensemble of flute and hand drums, whose players punctuate the action with kakegoe, stylised shouts that function as cues and as part of the music’s fabric, so that the boundary between signal and sound dissolves. The drummers’ parts are learned as oral patterns, syllables sung in the lesson before the drum is touched, in a transmission method that mirrors the Indian bol and the qin master’s finger instructions: the body first, the theory after. Japan’s schools carried this method across seven centuries for the noh, and the result is a theatre music whose tempos and cries a fourteenth century audience would still recognise as correct, maintained not by notation but by an unbroken chain of licensed bodies.

The Indian film music industry as a mass form

India produced the largest film industry in the world by output, and it built that industry on song. From Alam Ara, the first Indian sound film of 1931, the songs were not interludes but the product: audiences came for them, memorised them, and measured the film by them, and the industry’s economics came to rest on music released in advance as the film’s advertisement and its insurance. No other national cinema organised itself this way at this scale, and the result was a popular music with its own institutions, its own star system and its own musical grammar, answerable neither to the classical traditions surveyed earlier nor to the Western pop it freely borrowed from.

Why is Indian film music its own tradition?

Indian film music is its own tradition because it runs on its own star system and economics: a small company of playback singers became more famous than most actors, songs were composed to sell each film before release, and the musical grammar mixes Indian classical and folk material with outside styles on its own terms.

The system’s defining invention was playback. In 1935, working on the Bengali film Bhagyachakra and its Hindi version Dhoop Chhaon, the music director Rai Chand Boral and the singer and composer Pankaj Mullick put into practice the technique of recording the song first and having the actor mime to it on set. The consequences reorganised the industry. Singing and acting became separate professions united only on screen, a playback singer’s voice could serve a hundred faces, and a star singer could outlast generations of actors. The music director became an auteur in his own right, commanding orchestras that mixed harmonium and tabla with violin sections, Hawaiian guitar, Latin percussion and whatever else the decade offered, and the film song developed its own architecture of refrain and verse, engineered for the radio and the record as much as for the screen.

Radio wrote the next chapter, and it is a story about gatekeepers. In 1952 the minister for information and broadcasting, B. V. Keskar, judging film music a debased influence, sharply restricted it on All India Radio. The audience simply moved: to Radio Ceylon, broadcasting from across the water, whose Binaca Geetmala, a weekly countdown of film songs presented by Ameen Sayani, became one of the most popular programmes in Asia and demonstrated that the public’s attachment to film song was stronger than the state’s disapproval. Peter Manuel, whose Cassette Culture is the standard study of North Indian popular music’s modern transformation, then documented the sequel: the cheap cassette of the 1980s broke the film industry’s near monopoly, letting regional, devotional and independent producers reach audiences directly, and multiplying the number of musical economies that could sustain themselves. Gregory Booth, who has written extensively on India’s music industries and its brass band culture, makes the complementary point about the wedding bands and processional musicians who carried film melodies into every street celebration, an unpaid distribution network no marketer could have designed. Film music in India is therefore not a soundtrack business. It is the country’s popular music, full stop, with classical and folk traditions flowing into it and out of it at every border.

The Bombay model was replicated, with local accents, by the Tamil and Telugu industries of the south, and the system’s later history can be read in two careers. Ilaiyaraaja, the dominant Tamil composer from the middle 1970s, fused the Carnatic and folk materials of his upbringing with full orchestral writing learned from the film scores of the West, producing hundreds of film scores whose songs saturated the southern market. A generation later, A. R. Rahman, beginning with the Tamil film Roja in 1992, rebuilt the film song around digital production, programming rhythms and studio layering that the older orchestral system had no way to make, and carried the result to the world’s attention, winning two Academy Awards for his work on Slumdog Millionaire in 2009. The playback division of labour survived both revolutions intact. A system that can absorb the orchestra and the laptop in successive generations, while keeping its singer at the centre and its songs at the box office’s throat, has earned the name of tradition in full.

Arab and South Asian recording industries

The gramophone reached Cairo and Bombay within a few years of its commercial birth, and it changed what a tradition was. In 1902 the Gramophone Company’s recording expedition under Fred Gaisberg recorded hundreds of Indian performers in Calcutta, among them Gauhar Jaan, who became the first Indian recording star, and the three minute limit of the 78 rpm disc side began its quiet discipline over forms built for the long evening. Egyptian companies and the local agents of European labels, Odeon and the Lebanese founded Baidaphon among them, built catalogues of Arab art music from the first years of the century, and Umm Kulthum’s recording career, begun in the middle 1920s, grew alongside the industry until singer and industry were supporting each other’s authority. A performance that had existed only in a room now existed in a catalogue, comparable, repeatable and saleable across borders the performer would never cross.

The trade’s effect on the systems was double, and the two cases required by honest treatment can be stated plainly. The preservation case runs that the companies, whatever their motives, paid masters, recorded repertories that oral transmission was beginning to drop, and left archives without which much of this article could not have been written; that radio and records carried a village singer’s art to national audiences and made stars, with star incomes, out of hereditary musicians whom the old patronage had kept poor. The critical case answers that the trade also selected and froze: it decided which regional style would stand for a nation, shortened forms to fit its formats, paid for what it could sell and let the rest go unrecorded, and taught audiences to mistake the catalogue for the tradition. Both cases are true of different repertories at different times, and the evidence that would weigh them, session ledgers, sales figures and the unrecorded repertories that vanished, survives only in part.

South Asia replayed the pattern at each technological turn. The cassette era that Peter Manuel documented decentralised production in the 1980s, shrank the distance between performer and audience, and let devotional, regional and political musics flourish outside the film industry’s gate. In the Arab world, state radio and then satellite television played the centralising role the cassette played elsewhere, carrying the Egyptian style across the region until it functioned as a pan Arab standard against which local styles were judged. The pattern generalises beyond both regions. Every recording technology arrives as a neutral carrier and behaves as an editor, promoting the forms that suit it, and the traditions that thrive are the ones whose custodians learn to use the editor without letting it rewrite the book.

The reissue trade completes the circle back to the London pub. The unlikely bestseller that helped provoke the 1987 meeting, the Bulgarian choir record, was itself a reissue: recordings made for Bulgarian radio across earlier decades, compiled by the Swiss ethnomusicologist Marcel Cellier and licensed westward, so that the founding document of the world music boom was an archive dressed as a discovery. Much of the bin’s catalogue worked that way, older masters remastered, field recordings recontextualised, radio tapes given sleeves and stories. None of this diminishes the pleasure the records gave or the incomes some of them generated. It does mean that the category was built, from its first hit onward, out of the recording industry’s past as much as out of the world’s traditions, and a clear eyed listener will keep the two ingredients separate.

Where Lata Mangeshkar sits in the story

The playback system described above produced one figure whose career measures its full span. Lata Mangeshkar began as a child performer in her father’s Marathi theatre company, sang her first film songs as a teenager in the early 1940s, and after her father’s death supported her family with the work. Her breakthrough is conventionally dated to 1949 and the film Mahal, where her song for the composer Khemchand Prakash, a voice calling through a great house as if from another world, made her name across India in a matter of weeks. From that point the history of the Hindi film song and the history of her voice become difficult to separate. For decades she was the first choice of the industry’s leading music directors, from Naushad and S. D. Burman through Shankar Jaikishan and R. D. Burman to Laxmikant Pyarelal, and the sound of the screen heroine, across generations of actresses, was in large part the sound of her singing, light, exact and apparently effortless at any altitude of pitch.

Her craft repays description in this article’s terms. Playback singing is a discipline of ventriloquism: the singer must match the actress’s persona, the character’s age and the scene’s emotion, while delivering melody shaped from classical training, folk idiom and whatever the music director’s arrangement demands. Mangeshkar’s distinction lay in diction in many languages, in intonation of a precision engineers remarked on, and in an ability to sound young, which extended her reign across an unnatural span. Published estimates of her recorded output vary so widely, from several thousand songs to figures many times larger, that no total can be printed as fact, and this article prints none: the absence of an audited catalogue for the most recorded voice in her industry is itself a comment on the industry. What is documented is the standing. India gave her its highest civilian honour, the Bharat Ratna, in 2001, and her life and career are traced in her own article in this series.

The working method behind the voice was famously exact. Singers in the industry tell of her rehearsing a song until its turns were reflexes, arriving at the studio with the lyric’s meaning settled line by line, and recording at a speed that left orchestras half her age struggling to match her preparation. Her duets trace the system’s partnerships: with Rafi, a dialogue of equals that defined the romantic duet for a generation of films, and with Kishore Kumar, a lighter and more playful exchange that showed her comic timing. She sang, by the industry’s count, in more than thirty languages, a figure that belongs with the disputed output totals as a measure of reputation rather than of audited fact, and she kept her place at the centre of the profession across changes of technology from the 78 to the compact disc, adapting her art to each without altering its character. A playback industry is a machine for replacing voices. Hers was the voice it could not replace.

Where Mohammed Rafi and Kishore Kumar sit

The male half of the playback story is usually told through two singers whose gifts were opposite in kind. Mohammed Rafi, born in Punjab and trained in the classical discipline this article has described, was playback’s complete musician: a singer of formidable range and control who could reshape his voice to the actor before the camera, devout and restrained for Dilip Kumar, urbane for Dev Anand, flamboyant for Shammi Kapoor, whose physical exuberance Rafi matched phrase for phrase. His partnership with the composer Naushad established the template of the classically grounded playback singer, and his recorded repertory runs from devotional songs to qawwali styled numbers to the great romantic solos of the 1950s and 1960s. Rafi died in Bombay in 1980, aged fifty five, and the scale of public mourning measured what the playback system had made possible: a voice without a famous face had become one of the most beloved presences in the country’s life. His life and career are traced in his own article.

Kishore Kumar arrived by the other road. The younger brother of the star Ashok Kumar, he began as a comic actor who sang his own songs, largely untrained in the formal sense, and developed a style built on yodelling picked up from imported records, on rhythmic mischief and on an actor’s sense of the line as dialogue. Where Rafi embodied mastery, Kishore embodied character, and the industry found it needed both. His voice became the sound of Rajesh Khanna’s superstardom after the success of Aradhana in 1969, and then of Amitabh Bachchan’s angry young man of the 1970s, a heavier persona carried by the same unmistakable instrument. He died in 1987, aged fifty eight, having directed, acted, composed and sung across four decades. Placed side by side in this survey, the two men define the playback system’s range: a tradition this article has traced through lineages and radifs here shows itself able to honour the guru trained master and the self taught original in the same breath, and to build an industry on both. Kishore Kumar’s life and career are traced in his own article.

A third voice completes the era’s map and prevents the pairing from hardening into a fable. Mukesh, the singer associated above all with the actor Raj Kapoor, built a career on an apparently artless directness, a narrower range used with such emotional plainness that his songs of parting and endurance became standards of a different kind. Industry memory organises the three men as a trinity of temperaments, Rafi the master, Kishore the character, Mukesh the heart, and the simplification survives because it captures something real about how the playback system divided its labour among voices. Between them they sang the inner lives of three generations of film heroes, and the fact that the heroes are remembered partly through voices not their own is the playback system’s strangest and most durable achievement.

Where Nusrat Fateh Ali Khan sits

Qawwali’s journey from the shrine to the world stage has a single central figure. Nusrat Fateh Ali Khan was born in Faisalabad in 1948 into a family of qawwals, led the family party from his teens after the deaths of his father and his uncles within a short span, and rebuilt its name over two decades of performance at shrines and gatherings across Pakistan before the outside world heard him. What the outside world heard, when the WOMAD circuit and Peter Gabriel’s Real World label carried him to Western festivals and studios in the 1980s, was a voice of exceptional power and stamina organised by the full qawwali discipline: the slow building of an item, the repetition of a line until it became a hammer, the handclaps and harmonium and chorus of his party driving the assembly toward the state the form exists to produce. Recordings such as Shahen Shah and the Mustt Mustt album of 1990, made with the Canadian musician Michael Brook, presented the party’s art in studio form, and his singing reached film audiences in Pakistan and India as well. His life and career are traced in his own article.

His position in this article’s argument is precise. The world music trade did not invent Nusrat Fateh Ali Khan, package a raw folk essence, or dilute his form for export. It encountered a master of a complete system, working at full strength inside a hereditary discipline, and its machinery of festivals and labels did what the preservation case in the recording section claims at its best: it carried an intact art to listeners who would never have reached the shrine. He died in London in 1997, aged forty eight, leaving a recorded legacy large enough to keep the argument going. The shrine gatherings continued without him, as the system intends. The party is the institution, and the institution survives its greatest voice.

The technique behind the effect deserves a paragraph of its own, because recordings preserve it and listeners can learn to hear it. Khan sang from the harmonium he played himself, conducting the party with nods and lifted hands, and his improvisations borrowed the sargam patterns of classical singing, runs of notes named by their solfa syllables and delivered at speeds that turned melody into rhythm. His breath seemed to renew itself inside the phrase, letting him sustain an intensity across spans that left audiences describing performances of three and four hours as the normal measure of his generosity. Behind him the party worked as an organism: the second voice shadowing, the chorus answering, the tabla and dholak tightening the cycle, the handclaps of twenty men landing as one instrument. Other qawwals had the repertory. The combination of that voice, that stamina and that party discipline, held together for two decades, is what the world stage encountered, and it has not been assembled since in the same strength.

Where Umm Kulthum sits

Egypt’s long song found its voice in a singer from a Delta village, born in a year her childhood records leave uncertain, most often given as 1904, whose training was not musical in the conservatory sense at all. Umm Kulthum learned to sing as a child in her father’s small troupe, performing religious songs at village celebrations, and the foundation of her art was tajwid, the disciplined recitation of the Quran, which taught her the breath control, the diction and the command of Arabic vowel colour that critics would praise for the rest of her life. She reached Cairo in the early 1920s, studied with the leading musicians of the city, and by the 1930s stood at the centre of Egyptian musical life, a position she held for four decades.

Her art was the art of the long form described in the Egyptian concert section: songs composed for her by Zakariyya Ahmad, by Riyad al Sunbati across a partnership of decades, and from 1964 by Mohammed Abdel Wahab, performed on the first Thursday of the month and carried by radio to an Arab world that arranged its evenings around the broadcast. Virginia Danielson’s research establishes the scale of the phenomenon and its texture: the audience of connoisseurs who demanded repetitions and received them, the singer’s command of the maqam system deployed across spans of time no record side could hold, and the public role that followed, in which her concerts raised money for Egypt after the defeat of 1967 and her voice functioned as a national institution. She died in Cairo in 1975, and the crowds at her funeral were estimated in the millions. Her recordings, nearly all of them documents of an art built for duration, remain the measure of the form, and her life and career are traced in her own article.

The making of that authority began in disciplined obscurity. In Cairo she attached herself to exacting teachers, among them the singer and composer Abu al Ila Muhammad, who drilled her in the classical repertory and in the delivery of the qasida, the formal ode whose long lines test breath and diction as no popular song can. The ud player and composer Mohammed al Qasabgi joined her takht and stayed for decades, modernising its sound from inside while guarding its standards, one of several master musicians content to spend a career in her service because the service was the summit of the profession. She acted and sang in six films between 1936 and 1947, then withdrew from the screen at the height of her fame in it, judging rightly that the concert and the broadcast were where her power lay. The episode is a small masterclass in the management of a tradition: she tested the new medium, took its measure, and kept only what the long form needed.

Where Fela Kuti, Jobim and Marley sit

Three more figures mark three different doors out of the bin. Fela Kuti, whose construction of Afrobeat was examined in its own section above, represents the author as system builder: a musician who studied the materials of several traditions, highlife, jazz, soul and Yoruba practice, and engineered a form with its own standards, its own virtuosi and its own politics, so that the category African music had to stretch around a single man’s invention. His place in the story is proof that the traditions the trade files as inheritances are also, in every generation, compositions.

Antonio Carlos Jobim stands at the opposite door, the composer absorbed into another system’s repertory. His songs, written in Rio with Portuguese words for a local style, travelled so completely that they became standards of jazz, learned from lead sheets by musicians who had never heard the original recordings, harmonised and reharmonised by players on every continent. The 1967 album he made with Frank Sinatra marks the formal moment of transfer, a Brazilian composer and an American singer meeting as equals inside Jobim’s material. His life and career are traced in his own article. A tune that survives that journey has done what a raga’s characteristic phrase does: it has become a piece of grammar in a system far from home.

Bob Marley travelled the third road, the route from a local studio system to a world audience through deliberate packaging. Reggae had been built in Kingston’s studios out of mento, rhythm and blues and the sound system culture of dances, and Marley’s group the Wailers served their apprenticeship inside that system before Island Records, under Chris Blackwell, presented them to the rock market in the early 1970s with the production, artwork and touring that market understood. The songs carried Rastafari faith, the memory of slavery and exile, and a politics of dignity stated in the plainest language, and they entered the world’s common stock of melody to a degree no other Jamaican artist has matched. Marley died in 1981, aged thirty six, and the industry that grew around his catalogue afterwards is a study in what this article has called the trade behaving as an editor. His life and career are traced in his own article. Between Fela’s self built republic, Jobim’s absorption into jazz and Marley’s posthumous industry, the modern figure’s whole range of fates inside the world music economy is on display.

Set the three side by side and a colder pattern shows beneath the romances. Fela owned his stage, his compound and his confrontation, and the market reached his music only on his terms, at the price of permanent war with his government. Jobim owned his copyrights, and the market reached his music by adopting it, paying him the rent of a standard every time a band anywhere opened the book at his page. Marley’s catalogue outgrew its maker and became an asset administered by others, its earnings detached from any stage he would ever stand on. Author, rentier, estate: the three relations a modern musician can have to the trade that files him. The bin on the record shop shelf never hinted at any of them, which is the final reason this article has told the story through systems and the people who mastered them instead.

How to listen across world music traditions

The argument of this article can be converted into a method, and the method is short enough to state plainly. Begin with the organising concept of whichever system you approach, not with its famous names. If the music is cyclic, learn to feel the cycle, the sam in a tala, the bell pattern in an Ewe ensemble, the clave in a son, before following the melody, because the cycle is the ground the performers themselves are standing on and the melody is weather above it. If the music is modal, learn what its mode loves: the phrases a raga returns to, the route a maqam travels, and listen for the returns rather than for development in the European sense, the statement, variation and departure that other systems prize instead.

Work inside one system before comparing across systems. Two performances of the same raga by different singers teach more in an hour than a tour of ten traditions, because the differences isolate the art: the same grammar, two minds, and everything audible between them is judgement. Track one part at a time on the second hearing, the tabla’s negotiation with the soloist, the support drums under the master drum, the qanun’s fills behind a singer, and then let the texture reassemble. Keep notes as you go, with the date, the raga or maqam if you can name it, and the moment something happened that you did not expect, because the unexpected moment is the system showing you a rule you had not yet heard. A listening diary kept in an offline encrypted notebook protects that record from loss and from other eyes; the VaultBook download page is where the notebook application is obtained, and a year’s entries in it will teach a listener more about his own hearing than a shelf of survey books.

Above all, resist the bin’s way of listening, which hears geography where it should hear grammar. The musics in this article crossed the world long before a London pub named them, along routes of trade, faith, empire and migration whose fuller story is told in the companion survey of how music crossed borders, and they arrived as complete systems, each with its masters, its errors and its arguments about itself. The two words on the rack were a beginning, a way for a browser to find a first record. The systems are the destination, and they repay, on their own terms, exactly the attention that any great music repays: specific, patient and increasingly informed.

One habit remains to be named, because it converts method into membership. Every tradition in this article already has its audience, meeting in sabha halls and shrine courtyards, at milongas and festivals and on the dance floors of a dozen cities, and the newcomer’s fastest education is to join that audience on its own ground. Attend where the insiders attend, in the diaspora communities whose celebrations keep these systems alive in London, Toronto, New York and Paris as faithfully as in Chennai or Cairo, and let the room teach you its etiquette: when it applauds, what it asks to hear again, which error makes it wince. A listener formed by such rooms stops hearing the world’s music as a shelf of alternatives to his own. He hears neighbours, each keeping a house he is welcome to visit and in no position to rearrange, and that change of posture, from consumer of the bin to guest of the systems, is the real subject this long survey has been circling since the pub in Islington.

Frequently Asked Questions

Q: What is world music?

World music is a category invented by the Western record trade, agreed at a meeting of independent label staff in a London pub in the summer of 1987, to solve a practical problem: shops had no rack for the Bulgarian choirs, African dance bands, Indian classical singers and Latin American groups whose records the small labels were releasing. The term names a shelf, not a sound. The traditions filed under it share no scale, no rhythm system, no performance setting and no theory of what music is; they were grouped by exclusion, as everything that was neither Western pop nor Western classical. A scholar, Robert E. Brown, had used the same words at Wesleyan University in the 1960s for the comparative study of musical systems, and the honest use of the term still points that way: not a genre to listen to, but a reminder that many complete musical systems exist and that each must be understood on its own terms.

Q: Why do musicians dislike the term world music?

The objection is partly practical and partly about respect. Practically, musicians filed under the term found that the label set their fees, their billing and their reviewers’ questions. Promoters expected a generic festival atmosphere rather than a specific repertory, journalists asked about the performer’s country when they would have asked a rock musician about technique, and a master of the maqam or the raga could be presented as regional colour beside lesser artists presented as individuals. David Byrne stated the principled case in a New York Times essay in 1999, arguing that the term flatters the Western listener while patronising the music, as if whole cultures were a flavouring. Defenders of the category answer that the shelf space it won financed recordings and tours that would otherwise never have happened. Many musicians accept that bargain and dislike the words anyway, because the bin describes where their music is not from and stays silent about everything it is.

Q: What is the difference between a raga and a scale?

A scale is a set of pitches arranged in order, a ladder that any melody in the system may climb up or down. A raga contains a pitch set but is defined by behaviour rather than by inventory. It prescribes an ascent and a descent that need not mirror each other, characteristic phrases by which it can be recognised within seconds, notes that carry structural weight and notes that are only touched in passing, and ornaments that belong to it and would be errors in its neighbour. Two ragas, Bhupali and Deshkar, draw on the same pentatonic collection, and an informed listener never confuses them, because their phrases, emphases and gestures differ completely. The difference matters practically: learning a scale tells you which notes are available, while learning a raga tells you what the music means by them, how it moves, where it rests and when it is at home. A raga is closer to a personality with rules of conduct than to a ladder of notes.

Q: How does a tala cycle work?

A tala is a fixed cycle of beats, called matras, grouped internally into divisions, which the performance travels around repeatedly. The northern cycle tintal has sixteen beats in four groups of four; other common cycles run to seven, eight, ten, twelve or fourteen beats, and the southern system generates its cycles from a different arithmetic. The first beat of the cycle, the sam, is the point of arrival, and performers mark the internal divisions with claps and, at the empty division called khali, a wave of the hand away from the palm. The tabla speaks the cycle in syllables called bols, whose standard pattern, the theka, a student learns by recitation. Soloists improvise excursions of any length, but every excursion is measured against the cycle, and the performance’s characteristic thrill is the return: a phrase repeated three times with calculated spacing, a tihai, so that its final stroke lands exactly on the sam before the whole room.

Q: What is the difference between Hindustani and Carnatic music?

The two classical systems of India share deep concepts, raga and tala among them, and differ in almost everything built on top. Hindustani music, the northern system, centres on the long improvisation of a single raga around a short composition, in forms such as khayal and the older dhrupad, performed by a soloist with tabla and a melodic accompanist. Carnatic music, the southern system, rests on a vast repertory of composed songs, the kritis of Tyagaraja, Muthuswami Dikshitar and Syama Sastri, and builds concerts from compositions, with improvisation at chosen points. Their theories differ, the north grouping ragas by thaat, the south deriving them from the seventy two melakarta parent scales. Their percussion differs, tabla against mridangam, and their concert cultures differ, the northern mehfil tradition against the southern sabha societies and the Chennai December season. The ethnomusicologist Bonnie Wade treats them as sister civilisations sharing a vocabulary, and that is the accurate picture: related, distinct, and misread whenever one is described as a variant of the other.

Q: How is Indian classical music taught?

Indian classical music is taught primarily by residence in a master’s orbit rather than by enrolment in a curriculum. In the guru and shishya relationship, the student attaches himself to a teacher for years, hears phrases demonstrated and corrected hundreds of times, and absorbs the things notation cannot carry: weight, timing, ornament and taste. Early training is vocal even for instrumentalists, since the voice is treated as the primary instrument, and rhythmic cycles are learned by recitation of tabla bols and by keeping tala with the hands before they are performed. Lineage gives the teaching its public identity through the gharana, the stylistic house named for a town or court whose techniques and repertory a line of teachers hands down. Modern India adds conservatories, university degrees, printed notation and examinations, and serious students use all of them, but the summit of the system remains what it was: a long apprenticeship in which a master’s judgement is transferred, phrase by phrase, into a student’s body.

Q: What makes a maqam different from a mode?

A Western mode, in the way the term is usually taught, is a scale pattern: a rotation of a parent collection, defined by its sequence of tones and semitones. A maqam is defined by much more. It is built from smaller melodic units called ajnas, cells of three to five notes joined at shared pitches, and it carries a characteristic route, a sayr, that its melodies tend to travel, with regions where phrases gather, notes of emphasis and rest, and customary doorways into neighbouring maqamat. Its essential intervals may fall between the keys of a piano, the half flat steps of roughly three quarters of a tone that colour maqam rast and bayati, so the maqam cannot even be written fully in staff notation without special signs. The difference is the same one that separates a raga from a scale. A mode tells the player which notes exist. A maqam tells the musician how melody behaves inside them, where it lingers, how it departs and how it comes home.

Q: How long is a traditional Arab concert performance?

A traditional Arab concert was an evening’s whole occupation, and its units were long. The older form, the wasla, was a suite of vocal and instrumental pieces in a single maqam that could occupy an hour or more by itself. In the twentieth century Egyptian form that grew from it, a concert by Umm Kulthum typically presented two or three songs across an entire evening, and a single song could extend past an hour as she repeated lines at the audience’s demand, rephrasing each return until the hall’s state of tarab, the moved absorption the form exists to create, had been fully built. Her concerts were given on the first Thursday of each month through the season and broadcast live on Egyptian radio, with public life in Cairo adjusting itself around them. Recordings compress this art severely, since the three minute side of a 78 rpm disc held only a fragment of a performance, which is why the surviving concert tapes, rather than the studio discs, are the form’s true documents.

Q: What is the role of the audience in qawwali?

In qawwali the audience is a participant with duties, not a body of spectators. The gathering, a mehfil e sama, assembles to be moved toward devotion, and listeners answer the music directly: they call out at lines that strike them, ask for repetitions, rise and sway, and at moments of intensity offer money, which is presented to the singers and often dedicated onward to the shrine. The lead qawwal conducts the performance by reading the room, repeating and extending whichever line has landed, choosing the next item according to the occasion and the state of the assembly, so that no two performances of the same repertory take the same shape. This participation is theological, not decorative: the ecstasy the form aims at is communal, and a silent, motionless audience would represent a failure of the event rather than good manners. Concert performances outside the shrine keep the conventions in softened form, and experienced listeners still respond aloud, because the music was built to be answered.

Q: How does West African polyrhythm work?

West African polyrhythm works by layering simple patterns that interlock, so that the complexity lives in the relationships rather than in any single part. In an Ewe ensemble from Ghana, a bell called gankeke plays a short pattern of seven strokes distributed across a twelve pulse cycle and repeats it without variation for the whole piece. Rattles, drums and handclaps each add a pattern that falls into the bell’s gaps, and the support drums hold figures of their own, while only the master drum improvises, calling to the dancers, commenting and signalling changes. Because the parts are offset against one another, the ear hears cross rhythms, groupings of two against three and phrases that seem to float over the cycle, none of which any single player is performing. The dancer is part of the mechanism: steps articulate the cycle in the body, and the master drummer plays to the lead dancer’s feet. The practical way in is through the bell. Find its pattern, hold it, and the rest of the texture assembles itself around it.

Q: What is Afrobeat and who created it?

Afrobeat is a form of Nigerian dance music created in Lagos in the years around 1970 by the bandleader Fela Anikulapo Kuti, with his drummer Tony Allen as the rhythmic co author of the sound. It fuses the melody and guitar patterns of West African highlife, the harmony and soloing of jazz, the groove discipline of American soul and funk, and Yoruba rhythm and chant, arranged for a large band of guitars, keyboards, horns and percussion in compositions that run far beyond the length of a pop single, often for the full side of a record and longer in performance. Kuti sang in pidgin English so that his attacks on Nigeria’s military governments and their collaborators would be understood across ethnic lines, and his Lagos club, the Shrine, and his communal compound, Kalakuta Republic, made the music the centre of an alternative public life. Michael Veal’s biography of Fela gives the fullest account of the making. Afrobeat is not a folk inheritance: it is a modern synthesis with named makers, a date and an address.

Q: How did bossa nova develop?

Bossa nova developed in Rio de Janeiro in the late 1950s inside a small circle of musicians and poets who set out to refine samba rather than replace it. Its foundation is samba’s percussion battery condensed into one guitar pattern, the syncopated finger style perfected by João Gilberto, who also set the vocal manner, quiet, precise and just behind the beat. Antonio Carlos Jobim supplied compositions whose harmony drew on both samba and the American songbook, and Vinicius de Moraes supplied many of the words, the two having worked together on the play Orfeu da Conceição in 1956. João Gilberto’s record Chega de Saudade in 1958 announced the style complete, and the journalist Ruy Castro, whose chronicle of the bossa nova years is the standard account, shows how small the founding scene was, a few dozen young people trading chords and songs in Rio’s south zone. Jazz carried the music abroad through Stan Getz and Charlie Byrd in 1962 and the Getz and Gilberto album of 1964, but the form was finished before the world heard it.

Q: Where should I start with Indian classical music?

Start with one raga heard whole, in a single performance long enough for the form to unfold, rather than with a survey of excerpts. For the northern instrumental tradition, Ravi Shankar’s early sitar recordings, beginning with the sessions he made in London in 1956, present the alap’s slow assembly of a raga and the accelerating interplay with tabla in performances of manageable length, and his concert recordings with the tabla player Alla Rakha show the partnership at its most responsive. For the northern vocal tradition, a khayal performance lets you hear the bandish and its improvised departures, and for the south, a Carnatic concert recording in the kutcheri format introduces the kriti repertory and the mridangam’s art of accompaniment. Whichever door you choose, read nothing technical first. Listen once for the whole shape, once following the percussion, and once for the returns to the sam or the refrain, and let a single tradition, northern or southern, become familiar before you take up the other.

Q: Should I learn the theory before listening?

No, with one qualification. The theory of these systems was built to describe what practised listeners hear, and it yields its meaning fastest when it answers a question your ear has already asked. A newcomer who begins with the melakarta grid or the rules of the maqam will memorise a map of a country he has not visited. The qualification is that a small amount of orientation pays for itself immediately: knowing that a tala is a cycle with a point of arrival called sam, or that a raga is recognised by phrases rather than by its pitch list, changes what you are able to notice on a first hearing. The productive order is to listen with one concept in hand, then read for the next question the listening raised, then listen again. That alternation, experience and explanation correcting each other, is how the traditions teach their own students, compressed in the guru’s room into demonstration first and analysis afterwards.

Q: How do I count an unfamiliar rhythm cycle?

Begin by finding the pattern that does not change. Usually one part, a bell, a drum pattern, the tabla’s theka or a handclap figure, repeats exactly, and that part is your clock. Tap along until its length is in your body rather than in your arithmetic, then notice its internal grouping, since cycles are built from unequal groups, fours and threes and twos joined in a set order, and the groups, not the raw beat count, are what the performers feel. Next find the arrival point, the beat the ensemble treats as home, the sam in Indian music or the stroke the dancers follow in a West African piece, and test your counting by whether the music’s moments of resolution land where your count says home should be. If they do not, your grouping is wrong; adjust the groups rather than the tempo. South Indian listeners keep tala with the fingers and palm, and copying that method, groups tapped on successive fingers, is the fastest way to make a cycle permanent.

Q: Which Umm Kulthum recording should I hear first?

Hear a complete concert performance rather than a studio excerpt, because her art was built for duration and survives whole only in the live documents. The strongest single recommendation is Al Atlal, The Ruins, composed for her by Riyad al Sunbati and first performed in 1966, in which a long poem of return to a lost beloved’s dwelling places is carried through an hour of melody that Egyptian audiences received as the summit of her powers. For a first hearing, the 1964 Inta Omri, her first collaboration with the composer Mohammed Abdel Wahab, is nearly as rewarding and somewhat easier to enter, its melody more immediately graspable while the performance still shows the form’s signature method, lines repeated and rephrased as the hall demands them. Virginia Danielson’s study of the singer and Egyptian society explains the world these concerts belonged to. In either case, give the performance an evening and do not excerpt it. The repetitions are the content, not a delay before it.

Q: How do I recognise a sitar from a sarod by ear?

Listen first for brightness and shimmer. The sitar has a long neck, a gourd body and a bank of sympathetic strings that vibrate in answer to the played notes, so its sound carries a bright, buzzing halo, and its player bends notes by pulling the string sideways along high curved frets, producing long sliding ornaments. The sarod is shorter and heavier, with a skin covered resonator and a smooth metal fingerboard without frets, played with the nails of the left hand, and its sound is deeper, rounder and more percussive, with a drier attack and slides that feel weightier and more vocal. In a fast passage the difference sharpens: the sitar sparkles and rings, the sarod speaks in a lower, grainier register with each note more deliberately placed. The two instruments also tend to occupy different temperaments in performance, the sitar brilliant and quicksilver, the sarod grave and architectural, and once you have heard a master of each in the same raga, the distinction becomes one your ear keeps without effort.

Q: Why did film songs become India’s popular music?

Film songs became India’s popular music because the Indian film industry was built on song from its first sound film, Alam Ara in 1931, and because every institution that followed amplified the arrangement. Songs sold the film in advance, filled the radio hours listeners demanded, and were carried by records, cassettes and wedding bands into every town and village, while no rival popular industry existed at national scale to compete with them. The playback system then gave the songs star voices of their own, singers such as Lata Mangeshkar and Mohammed Rafi whose fame rivalled the actors’. When the state broadcaster restricted film music in 1952, the audience moved to Radio Ceylon’s Binaca Geetmala countdown rather than give the songs up, demonstrating where the country’s musical loyalty sat. Peter Manuel’s research on the cassette era shows the sequel: cheap tapes broke the film industry’s monopoly and let regional and devotional musics flourish, but they flourished in a marketplace the film song had created and continued to dominate.

Q: Is playback singing the same as recording a song?

No. Recording a song ordinarily means capturing a singer’s performance as the finished artefact, the record the public will hear. Playback singing is a specialised division of labour inside film production: the singer records the song in a studio first, before or apart from the filming, and an actor later performs the scene on camera while miming to that recording, so that the voice the audience hears and the face it watches belong to two different artists. The technique was put into practice in Indian cinema in 1935 by Rai Chand Boral and Pankaj Mullick, and it reorganised the industry around the split. It demands specific skills: the singer must match the actor’s persona, the character’s age and the scene’s emotion, phrase for a mouth that is not singing, and deliver in a single studio pass a performance that will be judged against the actor’s face forever. Its consequence was a star system of voices, singers more famous than most of the faces that wore their songs.

Q: Is it wrong to listen to music from another culture?

No, and the question deserves a more useful form, which is how to listen well. Listening across cultures is how these traditions have always grown: Indian film orchestras absorbed the violin and the Hawaiian guitar, Cuban son absorbed the piano, and Fela Kuti built Afrobeat partly from American records carried to Lagos. What separates engagement from mere consumption is conduct. Learn the music’s own terms rather than renaming it in yours, credit the named artists and lineages rather than treating the tradition as anonymous atmosphere, pay for recordings and concerts where payment is possible, and let informed insiders correct your hearing, since every tradition in this article maintains teachers whose authority settles questions an outsider cannot. There are settings where participation has limits set by the community itself, sacred repertories and closed ceremonies among them, and those limits are part of the music’s own rules, to be respected as you would respect the rules of a raga. Curiosity conducted with that respect is not appropriation. It is the audience every tradition hopes for.