All musical traditions are unique, and all also evolve constantly. But, whatever culture we may belong to, our experience of music has a point of convergence from the early twentieth century onwards, brought about by the technology of recording, and the creation of the 78 rpm, then the 45 rpm “extended play”, and finally the 33 and 1/3 rpm long playing records. A record is, literally, a “record” of an event, as a chapter in a history is, or an entry in a diary.

Among other things, putting a piece of music or a performance “on record” makes retrieval possible – even if it was unheard in its day, we can celebrate it anew, because it irrefutably happened. Retrieval was possible once with compositions but not performances, which vanished with the performer. We don’t know what Tansen’s voice sounded like.

A record is also a commodity. In its capacity of being a product, it was always perishable, as products are. It contained within it the possibility of being replaced by another kind of commodification – though the recording, as a way of communicating and disseminating music, is here to stay, despite the fact that its nature, with digitisation, has altered radically.

Vocabulary of value

Finally, from the early twentieth century onwards, the record or recording also became crucial to the vocabulary of value. Earlier, you might have admired a performer, a performance, a composer, or a composition. In India, where the process of creation was cherished more than the artefact, and interpretation (what’s often called “improvisation”) seen to be a more fertile expression of the imagination than the static composition, it’s performer and performance that dominate discussions of musical excellence.

In the West, it’s the composer and composition. The latter, especially, comprises the “work”, and works in totality comprise a composer’s oeuvre, and oeuvres in totality a musical period or tradition. A composition, work, oeuvre, or tradition also add up, in the West, to something else: cultural property. (If a tradition’s energies – as with the khayal, for instance – are based on elaboration, the fleeting detail, and the evanescence of performance, it can’t be “owned” in the way a “great composition” can. Beethoven’s works are his property, as they are his culture’s. There’s no comparable sense of property in Indian classical music.)

All these tensions inform how we receive or listen to a recording today: as a commodity, or a piece of cultural property, or an unrepeatable occurrence, or a combination of all of these.

The album cover of Pink Floyd's Dark Side of the Moon.

There are studio recordings and live recordings. A pop or rock band – say, Pink Floyd – may do a recording in a studio, and then do a live version later of some of the songs on that record. A pop or rock band won’t improvise greatly: what you can hope for in a live version is a fresh but on the whole faithful replication of one of the hits – that is, cover versions by the band of its own songs. In Western classical music, notation ensures a great degree of continuity between performance and recording.

But, for an improviser, every performance – whether it’s in a studio or in a concert hall – is a new beginning. Playing outside the dead zone of the studio – which suppresses rather than enhances acoustics, where musicians play in soundproof compartments or at an emphatic social distance of ten feet from each other – can liberate improvisers, as with Paco de Lucia, John McLaughlin, and Al di Meola in Friday Night in San Francisco.

Or it can unfetter an artist who’s an improviser but has had to curb themselves in the studio-recorded song. I have Ella Fitzgerald in mind – her Berlin concert in 1960, especially Mack the Knife and the scat-driven momentum of How High the Moon.


There are certain performances that are premised entirely on discovery and unpredictability. That is, they run counter to the ethos of Western classical performance, of sheet music, and the predetermined programme. A recording may very occasionally capture such an exploration – as in Keith Jarrett’s Köln Concert, a series of improvisatory pieces of whose content he had no clear idea before the concert on January 24,1975. This is more than a live improvisation – it’s using the space of the performance and recording to incarnate process.

As with a North Indian classical concert, where the singer might clear their throat or stop to tune the tanpura, Jarrett eschews the finished-product air of the studio recording and thumps the piano and shouts as he plays. Jarrett is a master of his idiom; and this mastery allows him to pull of an unrehearsed performance. It’s the type of recording that provokes us to explore a new set of parameters with which to describe what exactly is significant about it.

Glenn Gould’s 1955 recording of Goldberg Variations too has the atmosphere – despite being, to all purposes, a non-improvisational, sheet-music-dependent, European classical performance – of being a recording one-off: contingent; of its time and place. Gould was 22; he was highly accomplished; his choice of material was eccentric and, technically, extremely demanding, and had hardly been broached before on the piano. Surprisingly, he got some leeway from Columbia Records, and they allowed him to proceed over six days in June. This recording is a reminder that technology, on certain occasions, might not only capture the composed and the rehearsed but the unforeseeable. Like the Köln Concert, Gould’s Goldberg Variations is not only about a performer or composer: it’s about a moment.


What Jarrett and in a less obvious sense Gould did in 1975 and 1955 respectively is what Indian classical musicians do with each performance: try out, and bring forth, their material as if for the first time. They are masters of their genre – say, khayal – and of raag and taal, but for the purposes of the performance they start from scratch, as Jarrett did: they don’t know beforehand its details and progressions. They aren’t trying to bring a pre-existing ideal to life, as a classical pianist is when they assiduously follow the sheet music before them; the performance – or recording – is inextricable from instinct, impulse, and risk. Every Indian classical recording is a Köln Concert.

Why do we then not have a language with which to assess the unique nature of recordings of Indian classical music? We speak of the greatness and accomplishment of certain artists; but we still haven’t plumbed the mystery of what they’ve actually done with those three or 30 minutes that’s different from a Western classical or pop musician. With YouTube, our idea of the range of journeys Indian musicians have undertaken outside the studio – available to us now via private recordings – has grown, but we’re still to enumerate the more extraordinary of these journeys.

Alongside the Köln Concert and Gould’s Goldberg Variations, I’d like to place DV Paluskar’s raag Shree – a no less remarkable record. I’m sure there are several other Indian classical performances that could be added to this list; for now, let me say that Paluskar’s recording is not just about him, or the raag, or the khayal – it’s about the moment of its conception.

It’s unpredictable from the start, though never aberrant; and it has a suggestion of unrepeatability. The note on the jacket of the record I bought in the late 1970s tells us that Paluskar “died on 26th October 1955 at the early age of 34 years”. (The cause of death, we now know, was encephalitis.) The note adds, after providing us with a biography, that on “one side of the record is offered a classical exposition entitled Raga Shri which the artiste recorded barely three weeks before his death and which remains the last recording of his brilliant though shortlived career”.

It’s become clearer recently that Paluskar recorded this composition by Hararang – Hari ke Charan Kamal (‘Hari’s lotus feet’) in jhaptal (a 10-beat time signature) – neither for the radio, which I thought was the case, nor for a live audience, but for a recording experiment: see the long quote from the HMV producer G N Joshi in the comments to the link below.

The long playing record still hadn’t come to India; Joshi wanted to capture a longer, more authentic exposition of the khayal than the 78 rpm allowed. Paluskar believed he wasn’t in good voice – you can hear the slight hoarseness – though he was in wonderful form. Joshi apparently assured Paluskar he’d never use the recording; this provides us with a hint as to its makeshift abandon. The small print on the vinyl record says it was “first published in 1962”: so it was, itself, an act of retrieval on HMV’s part seven years after Paluskar’s death.

I don’t want to make too much of the fact that this is Paluskar’s final recording: that’s not the main reason why it’s unique. But the recording’s provisionality is accidentally but inevitably linked to the death’s untimeliness. It is paradoxically alive, which is why it’s one of those rare records of a moment – Paluskar’s Shree is an unfolding to which we’re privy again and again.

Shree is a famous twilight raag, but it’s infrequently performed, partly because of the demands its strange progression makes on performer and listener. Its notes are sa re (flat) ga ma (sharp) pa dha (flat) and ni. That is, its second and sixth notes are flat, and its fourth note is sharp. Since Paluskar’s tonic is D# or Eb major, the notes of Shree in this particular recording in Western music would be D# E G A Bb B D. Given the clusters of notes in the raag separated only by a semitone - ni, sa, re (flat); ma (sharp), pa, dha (flat) – it’s anyway a difficult mode.

But Shree, compared to, say, Puriya Dhanashree, which has the same notes, derives its beauty from emphasising the flat second or re, the note next to the tonic, and swooping up to the main interval, the fifth or pancham, and coming back to re, denying us the sa or tonic’s resolution. The tonic and fifth comprise the basic relationship in a scale; in Shree, it’s the note after the tonic and the fifth that create the fundamental inner relationship and tension. With that one tiny tweak, we’re led from whatever representational qualities a scale might have deep into an experience of the non-representational.

Paluskar begins, unusually for him, with a short dhrupad-style nom-tom alap. Then he plunges into the jhaptal without any of his usual Gwalior-style patience and convention in progression. The alap, taankari, and exquisite layakari through which he possesses the raag feel unpremeditated.

I have two copies of this record. The second was given to me as wedding gift by the journalist Dharani Ghosh in 1991, in the last days of vinyl. So it was a well-known record. With the arrival of cassettes and CDs and, finally, downloads and streaming, it seemed to disappear without anyone noticing. I searched for it on YouTube several times in the last few years but never found it, probably because of copyright restrictions. When I downloaded the Gaana app, I found it buried, unannounced, beneath a calendar picture of Krishna.

It’s now on YouTube: two links were uploaded by Saregama on July and August 2019. A key moment in the history of modern performance is available here.


Amit Chaudhuri is a writer, a Hindustani classical vocalist, and a composer of crossover music. Listen to his music here.