Riyaz: What Tabla Masters Say About Practice, and What the Evidence Shows

Students ask how many hours they should practice. It is a fair question with no single answer — even among the players best placed to give one. Ustad Ahmed Jan Thirakwa is remembered for sixteen-hour days. Pandit Aneesh Pradhan, one of the most respected players working today, keeps no fixed schedule at all. Both are speaking honestly about a life given to the instrument, and the space between them is where this article begins.

The hours, as documented

These are the hours claims I could verify at their source. Every other figure in circulation failed that test, and there are many.

Two details in that table matter more than the numbers.

Pt. Swapan Chaudhuri's nineteen years of four-hour evenings were taleem, meaning instruction with his guru present. That is a different activity from solitary practice. It is also the only figure here with a duration, a frequency and a span attached to it, which makes it the most useful entry in the table.

The second detail concerns Ustad Zakir Hussain, whose training is the most retold story in tabla. He told NPR in 2015: "I was woken up at 2:30 in the morning, and that's when we sat and talk rhythm." He added that this happened "every night for four years." The sessions were conversation. Alla Rakha talked to him about rhythm, about tradition, about the masters. Hussain described the playing separately, to the National Endowment for the Arts in 2016, as what he did afterwards to "put into practice what had been drilled into my brain between two-thirty and six in the morning."

The famous night hours were transmission. The riyaz came later in the day. Lose that distinction and you are left with the impression of a child made to play for three and a half hours before dawn.

What did Alla Rakha require of his own students? His regimen, reported by his son, was roughly two hours before bed, ninety minutes before school, and two hours after returning home. That is about five and a half hours, in three sessions, around a school day. It is more useful than any sixteen-hour legend, and it comes from the same household.

James Kippen's 1988 study of the Lucknow gharana supplies the distinction most advice skips. Riyaz, he writes, is where one puts into practice all one has learned. Taleem is transmission. Riyaz is assimilation. They are separate activities, and conflating them is what makes Chaudhuri's four evening hours look comparable to Sahai's twelve solitary ones. Kippen also records that his own riyaz was structured in two daily periods, on Afaq Husain's instruction, which is a sourced example of session design.

Where the masters differ

The difference does not run between traditionalists and moderns. It runs through individual careers, and sometimes through single interviews.

Sanju Sahai practiced twelve hours a day in a rented room away from the city. In the same conversation he says: "Riyaz should not be forced." He gets more specific: "If you only want to practice for 20 minutes then practice for that long." He also rejects the direction the instrument's culture has taken. "Speed and volume have too much importance in modern times, and this can kill the music." And on whether hours are sufficient: "I can play 20 hours a day in this lifetime, but won't be half as good as my dad was."

Nikhil Banerjee said "just practice 14 hours a day minimum." Elsewhere in the same interviews he says: "You must first practice technique, then you must forget it."

Zakir Hussain told a Berklee class in 2019 to learn every solo and standard, then "get on the stage and forget all that." Speaking to the NEA he described his father's approach in terms that contradict the disciplinarian image entirely: "it wasn't like he was going to force me to practice."

Ustad Amjad Ali Khan states the counter-position most plainly. "I don't practice every day, but mentally most of the time I am thinking about music." Speaking to CNN in 2005 he was blunter about what hours cannot buy: hard work and practice are fine, he said, but without being blessed you cannot reach your goal.

Aneesh Pradhan says: "I do not have a fixed riyaaz schedule, but am involved with music most of the time."

Pandit Suresh Talwalkar takes on a different orthodoxy. Teachers, he says, let students believe extraordinary technique comes only from natural talent, and he disagrees completely.

Swapan Chaudhuri, whose own training was imposed rather than chosen, is candid about it. "From then on my father, a strict disciplinarian, made me practice tabla. I had no choice." He adds that even on the day of his final university exams, his father insisted he practice.

Set these together and the picture is a tradition in conversation with itself, in public, about hours, speed, schedules and talent. It does not speak with one voice on riyaz, whatever a single confident source might suggest.

What masters actually do

The argument about hours has crowded out the more useful question, which is what happens inside a session. Here the sources are more consistent, and more practical.

Say it before you play it

Padhant is the recitation of bols aloud, away from the instrument. The word comes from padhna, to read or recite. It is the most widely shared habit in tabla training, expected of students and of performers, and in solo performance a player will often recite a composition before rendering it. Kathak dancers do the same thing before executing a pattern in footwork. Carnatic music has a direct analogue in konnakol.

Padhant is not a memory aid bolted onto practice. The bol is the sound. Each syllable names a specific stroke and encodes the timbre it should produce, so reciting correctly is a rehearsal of the intended sound and not only of the sequence. This is why teaching sources treat recitation as the route into stroke quality, and why a student who cannot say a phrase cleanly will rarely play it cleanly.

I found no source describing padhant as a graded stage with a fixed duration. It reads as a continuous discipline, which is a gap in the documentation and not a contradiction between sources.

The stroke is the subject, not the sequence

Nikas is the quality of sound production: how a stroke is made, as distinct from which bols are played. It is trained through the distinction between open and closed strokes. Khula strokes are struck sharply, at the kinar, for a ringing and resonant tone. Band strokes are struck onto the syahi with the last three fingers for a dry, contained sound. This framework appeared consistently across independent sources with no contradiction, which is unusual in this literature.

The practical consequence is that two players can produce the same bols and not be playing the same thing. Nikas is what the difference is called.

Making paltas is a skill with a method

A kaida is a theme with a formal system of variation. A palta is one permutation of its phrase, built by rearranging the theme's own material while keeping its identity intact. Prastar is the broader combinatorial idea.

Shawn Mativetsky, a professional player in the Sharda Sahai lineage who teaches at McGill, has published the arithmetic. He learned it, he writes, from "my senior guru-bhai, Bob Becker, of the renowned percussion ensemble, Nexus." For a phrase with repeated elements the number of distinct arrangements is the factorial of the phrase length divided by the factorials of the repeats, and he works a small example through in full.

His caution is as valuable as the formula. "In performance, we would never play through ALL permutations, as it would soon become predictable and boring." He defends the exercise anyway, for technical preparation and for discovering rhythms a player would not otherwise reach. That is a clean statement of what a drill is for, from someone with standing to make it.

Teaching sources describe an intermediate step worth noting: students build paltas and rehearse them on the palm, memorising the bols, before taking them to the tabla. Padhant again, doing the work.

The laya ladder

Layakari drilling has a documented progression. A kaida's theme is played at ekgun, then dugun, then chaugun or atgun, with the timing of each step depending on the composition and the student.

A separate and harder discipline holds the base tempo fixed and moves the theka through ekgun, dergun, dugun, tigun and chaugun without letting the underlying pulse shift. Sources note that dergun at three against two, and tigun, are markedly harder than the even doublings. Any player who has tried this knows it, and it is rarely written down. Cross-taal work follows, superimposing one taal's theka inside another's cycle.

What is taught first, and what is played first

One syllabus-level fact is worth more than a dozen appeals to tradition. The first-year examination of the Akhil Bharatiya Gandharva Mahavidyalaya Mandal already includes kaida, alongside term definitions and the taught taals. Kaida is beginner material.

A related error is common and documented. In a solo, peshkar comes early, at slow tempo, before kaida and its variations. But students typically learn peshkar at intermediate level, because its ornamentation depends on fundamentals that kaida drilling builds. Performance order is not teaching order.

Beyond that, David Courtney is explicit that no single universal hierarchy exists, and that sequencing varies by gharana. Anyone presenting a fixed curriculum for the whole instrument is describing one lineage and calling it the tradition.

The lehra exists so you can practice alone

The lehra is a repeating melodic cycle. Its function in tabla practice is uncontested across sources: it lets a player work on compositions and solo material without a second musician present. That is a narrower and more useful claim than the general debate about drones, which belongs to vocal training and is genuinely unsettled. Teachers disagree about whether a vocal student should practice against a harmonium or a tanpura, and how soon to move between them.

Watching, accompanying, and memorising without paper

Three practices appear in masters' accounts and almost never in written advice.

Pandit Yogesh Samsi, describing twenty-three years with Alla Rakha, says his guru "would make me sit in front of him and practise." He adds that he also watched his guru practice and learned through observation alone. Observation is treated as practice, not as waiting.

Swapan Chaudhuri widens the definition further, against the assumption that riyaz means solitude: "Accompanying itself is also a part of one's practice. As you keep playing, you improve." He is equally direct about how he expects material to be held. Of his own students: "If you want to study with me... no writing! You have to memorize and play."

Samsi locates the discipline outside the individual entirely. "You need a good environment and a good guru," he says, using the word samarth, capable. Sanju Sahai describes the Benares version of that environment, where neighbours "would complain if they could not hear your riyaz." Practice was socially enforced before it was ever self-managed.

What the tradition actually transmitted

The scholarship suggests the hours argument may be the wrong argument.

Dard Neuman, writing in Ethnomusicology in 2012, describes a pedagogy that resisted a century of attempts to convert it into theory and notation. Ustads withheld technical nomenclature deliberately. Students practiced rags without being told which rag they were practicing, and sang notes without note names. What mattered was repetition sustained long enough that the hands moved without direction from a deliberating mind.

Suresh Talwalkar's account of his own training is the most extreme version I found. Speaking to The Indian Express in 2018, he said: "I was never asked to play the tabla during the nine years of my training, a decision I disliked." Nine years of instruction before the instrument.

Emily Sayers filmed just over nineteen hours of vocal taleem across six Indian cities and published the analysis in 2025. Her findings cut against the romance of improvisation. Students rarely improvise spontaneously. Learning proceeds by imitation and memorisation, and creative recomposition emerges from inexact replication of fixed material. Musicians accumulate roughly four hundred to fifteen hundred fixed compositions across a career. Improvisation grows out of that accumulation and not out of instruction in grammar.

Daniel Neuman recorded how riyaz was verified before anyone kept a log. Musicians inspected his hands. Callus and a hardened ridge at the cuticle were the evidence that practice had occurred. He also observed that learning how to practice correctly was itself a taught stage. Students were not assumed to arrive knowing it.

One thing, many times

If the sources converge anywhere, it is here, and it is not on hours.

David Clarke and Vijay Rajput state the principle directly in their 2024 handbook, which is open access and therefore checkable: "in any riyāz session, one should not practise many items a few times, but one of them very many times." Depth over coverage. That single sentence contradicts how most students actually spend an hour.

The counts attached to this are literal. Clarke and Rajput cite Gerry Farrell on a palta "which one of my teachers told me to play one thousand times (and he wasn't speaking figuratively!)". The vocalist Pandit Ajoy Chakrabarty has described practicing the same taans five to six hundred times through a night. Some gurus have students count repetitions on a mala, which as Clarke and Rajput note "usually has 109 beads, and is conventionally used when chanting mantras". The counting device is borrowed from a discipline built on repetition.

Read against that, Suresh Talwalkar's scale stops looking idiosyncratic. He told The Hindu in 2022 that a pattern played 108 times will not be forgotten, that 500 brings mastery, and that at 1,008 a tej, an energy, enters the playing. The numbers are close to the mala's own count, and they sit inside a tradition that has been counting in hundreds for a long time.

None of these figures is a measured finding, and none should be read as one. What they establish together is that senior practitioners across instruments structure practice by count instead of by clock, and that the counts are large.

Pandit Arvind Parikh, a senior disciple of Ustad Vilayat Khan, supplies the only repetition threshold I found with a stated mechanism. After roughly fifteen to twenty repetitions of a palta, he says, the conscious mind begins to wander while the hands continue. That split is the objective, not a failure of concentration: the conscious mind is freed to invent while the trained part executes. It is the same claim Dard Neuman makes from the scholarly side, arrived at from the inside.

Farrell adds a point that should reassure any student who suspects they have outgrown their exercises. The routine material is "practised at all stages of learning, by beginner and maestro alike, the only real difference being the speed of execution and level of sophistication". Nobody graduates from paltas.

Two structural ideas worth stealing

Two accounts describe how practice is organised, and both are more useful than a target number.

Pandit Sanjoy Bandopadhyay has written about his taleem under Pt. Radhika Mohan Maitra. The lesson was ten minutes long, once a week, and the door "opened at 8:59 and closed at 9:01, timed to All India Radio". He would play the previous week's lesson for fifteen or twenty minutes, receive ten minutes of new instruction, then spend five to six hours daily for a week expanding those ten minutes into twenty minutes of refined music. He also notes what was forbidden: "No writing, no recording—only memory and assimilation."

That ratio is the clearest statement of the riyaz contract I found anywhere. The input is tiny. The work of turning it into music is enormous, and it belongs to the student.

The second idea is a readiness test rather than a schedule. Marathi practice literature from the Bhartiya Sangeet Kalapith tradition sets an over-preparation ratio: a composition that will run three and a half to four minutes in performance should be sustainable for ten to twelve minutes at maximum tempo in practice. It also describes a forty-day stamina ladder, beginning at five minutes on day one at a settled top tempo and adding a minute a day, arriving at around thirty continuous minutes by day forty. The same tradition names murakka, short and deliberately awkward phrase-sets built to attack one problem syllable.

These are implementable today, which distinguishes them from almost everything else in the literature.

The dissenters on hours are more distinguished than you would guess

The anti-grind position is not a modern accommodation. It has canonical backing.

Deepak Raja records that Ustad Amir Khan's father rejected the ten-and-twelve-hour regime that was then normal, and allowed him three sittings of one hour a day. The reasoning given is that mechanical repetition is not learning, and that practice pays only while melodic imagination stays active. Amir Khan later said that half of his riyaz was thinking rather than singing. For one of the most revered vocalists of the twentieth century, that is a substantial claim to have on the record.

Arvind Parikh, who reports the Imdadkhani practice of measuring riyaz by the burning of a candle, roughly two to two and a half hours, and of stacking two to four candles for an eight to ten hour day, nevertheless states the limiting principle plainly: "There is no general prescription which could apply to all students." He holds that the guru must set each student's riyaz according to their education, background and worldly responsibilities.

Amit Chaudhuri, a trained khayal singer as well as a novelist, has written that he practiced every other day, insists on regular breaks from practice, and argues that the aims of practice cannot be reached by force of will.

Vidushi Shubha Mudgal locates the purpose somewhere other than accumulation. Riyaz, she has said, "is something one did in solitude. It is meant to examine your limitations, introspect in brutal honesty." Practice as diagnosis, not as display, which is a pointed thing to say in an era of posting one's practice.

A protocol, from a player who published one

Most masters describe conditions. Arvind Parikh describes a procedure, and it is the most complete practice protocol I found from any Hindustani musician.

Play one raga from beginning to end as though performing it before an audience. The weak and confused passages announce themselves under those conditions in a way they will not during fragmentary practice. Then stay with that same raga for a prolonged period, on the order of six months, until the material is embedded.

He pairs this with sequencing advice that runs against intuition. Start a new student on the faster form at a reduced tempo and raise it gradually, because a faster-moving melody is easier to remember than a slow, long-drawn one, and slow tempo is genuinely hard for beginners to hold. He also gives an objective test for how slow a vilambit tempo should be: the speed at which the player can actually execute chaugun taans.

Shawn Mativetsky offers the best definition of "learned" in this literature, and it is not a count. A phrase is looped until it is integrated on four axes: memorisation, technique, tone, and musicality. His own guru's standard, he reports, was to practice a million times and perform once.

What the research supports

Most of what a student reads about practice science is a distortion of the underlying work.

The ten-thousand-hour rule is the clearest case. Ericsson's 1993 study of violinists in Berlin found that the strongest players had accumulated more solitary, effortful, feedback-driven practice than their peers. It found no threshold. Ericsson said so himself in 2016, pointing out that half his elite group had fewer than ten thousand hours by age twenty. The popular version dropped the word deliberate and added a magic number.

A 2014 meta-analysis by Macnamara, Hambrick and Oswald put figures on it. Deliberate practice explained about 21 percent of performance variance in music and roughly 19 percent across domains. The detail that matters most: in studies that used practice logs instead of asking people to recall their practice, the figure fell to about 5 percent. The strongest early evidence came from the least reliable measurement.

The most useful single study for a practicing musician is smaller and more specific. Duke, Simmons and Cash watched seventeen advanced pianists learn a difficult passage from a Shostakovich concerto, then tested them the next day. Practice time predicted nothing. Neither did the number of trials, nor the number of complete run-throughs. What predicted retention was how the players handled errors: locating the problem, isolating it, changing tempo around it, stopping once corrected. The paper is titled "It's Not How Much; It's How."

The largest relevant study is more uncomfortable. Hallam and colleagues surveyed 3,325 young players in 2012. Months of instruction was the best predictor of expertise. Once that was controlled for, the number of good strategies a player adopted stopped predicting anything. Two specific strategies still did: using recordings, and using a metronome, which for a tabla player is the lehra doing the same job. And the finding that should reorganise most practice advice, including much of this article: avoiding ineffective strategies predicted expertise more strongly than adopting effective ones.

Three further results are solid enough to build on. Sleep consolidates motor learning, and practice scheduled close to sleep shows measurable next-day benefit. Mental rehearsal without the instrument produces a real improvement, replicated across three meta-analyses at effect sizes near 0.4 to 0.5, always smaller than physical practice. And percussionists outperform non-percussionists at reproducing rhythm only when permitted to move; the advantage disappears in listening-only conditions. That last result is the closest thing to evidence that physical riyaz cannot be replaced by listening.

The mental rehearsal finding invites a connection to padhant, and here, unusually, there is tabla-specific evidence.

Kai Siedenburg, Shawn Mativetsky and Stephen McAdams published a study of tabla memory in Psychomusicology in 2016. They tested tabla students and musicians unfamiliar with tabla on their ability to recognise sequences of bols and of drum strokes, comparing idiomatic sequences against the same material reversed, randomised, and randomly selected. Performance fell steadily as sequences departed from idiomatic order, which the authors read as evidence for chunking in auditory recall.

The finding that matters for practice is the second one. The difference between tabla players and other musicians appeared only for idiomatic sequences of bols. It did not appear for the drum strokes. Expertise showed up in memory for the spoken syllables, and not in memory for the sounds of the instrument.

That is an empirical argument for recitation, on tabla, from a study co-authored by a working tabla player. It does not test a practice regimen, and it does not prove that padhant causes the advantage. But it locates tabla memory in the verbal layer, which is exactly where the tradition has always put it.

Two findings translate into concrete session design. Long motor sequences are stored as chunks, and the measurable signature is that transitions within a chunk are fast while transitions between chunks are slow. Practice concatenates chunks into larger units, and sleep appears to act on that process. The seams between memorised segments are the fragile points, which argues for drilling transitions specifically instead of running whole compositions. This comes from laboratory work on short finger sequences, so treat it as a mechanism and not as a tested musical prescription.

Self-recording measurably improves the accuracy of self-assessment. Students hear on playback what they miss live, and for judging sound, audio feedback is slightly better than video, which adds visual information irrelevant to the task. This is the rare case where the research and Hallam's large survey point the same way: recording yourself is worth the trouble.

Where the evidence contradicts the advice

Four widely repeated rules are weaker than their confidence suggests.

Slow practice is the first. Every teacher says to start slow and build. The controlled evidence for slow practice as such is thin. Survey work by Allingham and Wöllner found gradual tempo increase is the commonest strategy musicians report, while noting the empirical case for it in applied musical settings is weak and that transfer from slow to fast is not automatic. Searching for a slow-practice statement from a named tabla master, I found none. The closest attributable source in Hindustani music is Pandit Ravi Shankar, writing in My Music, My Life that exercises should be practiced very slowly at first. So the most universally taught rule in tabla has almost no citation record, in the tradition or in the laboratory. That does not make it wrong. It makes the certainty with which it is asserted unearned.

Spaced practice is the second. Distributed study beats massed study across a large body of learning research, and that result is routinely imported into music advice. When Wiseheart, D'Souza and Chae tested it directly on piano learning in 2017, they found no spacing effect on any measure. The authors are careful about why: their intervals ran from zero to fifteen minutes, possibly too short for forgetting to occur before restudy. The honest summary is that spacing is well established for verbal material and unproven for musical motor skill.

Interleaving is the third, and here I have to report a genuine conflict. Rotating between pieces is supposed to beat drilling one thing at a time, and the general contextual-interference literature supports that. Music-specific results do not agree with each other. A 2016 study of ten clarinetists found a marginal advantage for interleaving, and noted that most participants preferred blocked practice even though interleaving helped them more. A 2013 piano study found the same illusion, with players judging fixed-order practice more effective while random order produced better recall. But the newest and best-powered study, Mathias and Goldman in 2024, tested nineteen advanced violinists and found the reverse: at twenty-four-hour retention, blocked practice declined significantly less than either interleaved condition, with a sizeable effect. Three studies, two directions. Anyone who tells you to rotate your kaidas because science says so is overstating a contested result.

Self-testing is the fourth. The testing effect is among the best-replicated findings in cognitive psychology for verbal material. The one direct study on piano melody memorisation found no benefit at short delays. The question is open and not a settled win.

Two more claims deserve a shorter treatment. Warm-up rests on a pilot study with sixteen participants that found no unadjusted difference in pain between groups, and the broader injury-prevention literature says protocols are too inconsistent to establish a causal role. Overlearning past the point of mastery has strong support in laboratory motor and verbal tasks and essentially no direct evidence in musical repertoire retention.

The cost nobody mentions

An article that opens with sixteen-hour days has an obligation here.

Playing-related musculoskeletal disorders are not a marginal risk for percussionists. A survey of 279 percussionists found 77 percent reported at least one, rising to 89 percent among keyboard percussionists and 74 percent among those playing membranophones. A separate survey of 831 drummers found 68 percent reported a lifetime disorder and 59 percent reported more than one. The commonest diagnoses are tendinitis and carpal tunnel syndrome, concentrated in the wrist and upper limb, with the lower back also affected. Musician's dystonia was substantially more common among professionals than amateurs.

No study of tabla players specifically appears in this literature, which is its own problem. The mechanism is not exotic: repetitive loading of the hands and wrists over long sessions, often without progression control.

The one prevention principle that clinicians repeat is to keep weekly increases in practice volume under roughly ten percent, which allows recovery to keep pace. That figure is a heuristic borrowed from sports-medicine overuse research and has not been tested on musicians, so treat it as prudent rather than proven.

The relevant point for anyone reading a sixteen-hour anecdote as a target is that the risk of injury from long hours is documented, while the benefit of long hours is not.

What nobody knows

No study has tested a riyaz regimen. Nobody has taken tabla players, varied their hours or their tempo discipline or their session structure, and measured what happened. Every claim in this article about how to practice is imported from work on Western instrumentalists or from general motor learning research.

The exception is worth stating precisely, because it is the only tabla-specific empirical result here. Siedenburg, Mativetsky and McAdams tested tabla memory, not tabla practice, and found the expertise effect sitting in the spoken bols. That is evidence about how tabla knowledge is stored. It is not evidence about how many hours to practice, and it would be dishonest to stretch it.

So when anyone tells you the science says something specific about riyaz hours, they are extrapolating. That includes me, and I have tried to say so at each point.

The chilla is better documented than I expected when I started, and I am revising a claim I would otherwise have made. Forty days of secluded practice, from the Persian chehal for forty, with roots in Sufi retreat. Sadanand Naimpalli, a tabla author, describes it as the most result-oriented riyaz system and gives ten to twelve hours daily across the forty days, with a start around four in the morning. A 2024 paper in Swar Sindhu repeats the same figures. Mulla Afsar Khan adds a constraint that sounds like practice rather than legend: the same place, at the same time, each day. Amir Hussain Khan is reported to have held that real stamina comes only from practicing kaida under chilla conditions.

Zakir Hussain's own first chilla, around age sixteen at the Haji Malang shrine and lasting sixteen days instead of forty, comes to me through summaries of his autobiography, so I label it testimony at one remove. And the tabla player Pandit Samir Chatterjee has published a day-by-day journal of his own forty-day chilla, which is the only first-person account of its kind I am aware of. I have not read it. Anyone writing seriously about chilla should.

What remains true is narrower than my earlier position. The chilla is real, described consistently, and practiced within living memory. Its history as a systematic institution across gharanas is still thinly evidenced, and one widely circulated source on the subject now redirects to an unrelated commercial site.

The famous stories belong in the same category. The scorpion, the chilli in the wound, the hair tied to a ceiling ring, the sixteen-hour room. Daniel Neuman reads this material as a moral charter for perseverance, not as a record of events. Ravi Shankar's own book reports his guru practicing sixteen to twenty hours with his hair tied to prevent sleep, which is testimony from a named source and still not documentation.

The claim that musicians practice less today has no support I could find. Max Katz's work suggests the golden age of rigorous ustads is partly a retrospective construction, and that institutional teaching arose from a genuine supply crisis, not from decline. James Kippen has shown that notated tabla pedagogy has a datable modern origin in a 1903 text written for Paluskar's school, which is to say that a good deal of what students receive as immemorial tradition is younger than the gramophone.

Practice hours themselves have no scholarly measurement in this tradition. Every figure in the table at the top of this article is testimony.

What follows

The evidence does not deliver a schedule. It supports a smaller number of statements that can be made without embarrassment.

Duration is the weakest predictor anyone has measured. Error handling is the strongest. Sessions built around locating and isolating problems outperform sessions built around accumulating minutes. That is a measured result, not an opinion.

Eliminating bad practice matters more than collecting good techniques. That is the uncomfortable finding from the largest study in the field, and it argues for auditing what you already do before adding anything.

Recording yourself is supported twice over, by a large survey and by controlled work on self-assessment. Most players will hear faults on playback that they cannot hear while playing.

The tradition's own most concrete regimen, from Alla Rakha's household, is three sessions totalling about five and a half hours around a school day. The sixteen-hour figures come from a different kind of source and should be read differently.

Taleem and riyaz are separate activities, and comparing players who conflate them is meaningless.

Practice one thing many times instead of many things a few times each. This is the one prescription the sources converge on, it is stated plainly in an open-access handbook, and it is the opposite of how most students fill an hour.

Recitation carries more weight than its reputation. Padhant is the entry point to stroke quality and the rehearsal method for paltas. It is the one practice the whole tradition agrees on, and it is the only element of riyaz with tabla-specific empirical support behind it.

Depth of repetition beats breadth of coverage, and the counts in the tradition are large. Whether you take Talwalkar's 108 or a mala's beads or Farrell's thousand, the instruction is the same and it is not a metaphor.

Test readiness rather than counting minutes. If a four-minute piece cannot survive ten minutes at full tempo in practice, it is not ready.

Long hours are not the mark of seriousness they are taken for. Amir Khan practiced three hours a day in three sittings and said half of his riyaz was thinking. Arvind Parikh, who can quantify riyaz in candles, still holds that no general prescription fits all students.

Drill the seams. Transitions between memorised segments are where sequences fail, by mechanism and by experience.

At least one senior master structures practice by repetition count instead of by clock. Talwalkar's 108, 500 and 1,008 are worth trying as an experiment, on his authority and not on anyone's data.

Long hours carry a documented injury risk and an undocumented benefit. Increase load gradually.

Beyond that, the most accomplished practitioners of this instrument disagree with each other about hours, speed, schedules and talent. A student who is told otherwise is being sold a certainty that does not exist.

Reviewed 26 July 2026.

References

1. K. A. Ericsson, R. T. Krampe and C. Tesch-Römer, "The Role of Deliberate Practice in the Acquisition of Expert Performance," Psychological Review 100/3 (1993); with Ericsson's own corrections to the popular reading, interviews, 2016.
2. B. N. Macnamara, D. Z. Hambrick and F. L. Oswald, "Deliberate Practice and Performance in Music, Games, Sports, Education, and Professions: A Meta-Analysis," Psychological Science (2014).
3. R. A. Duke, A. L. Simmons and C. D. Cash, "It's Not How Much; It's How: Characteristics of Practice Behavior and Retention of Performance Skills," Journal of Research in Music Education 56/4 (2009), 310–321.
4. S. Hallam, T. Rinta, M. Varvarigou, A. Creech, I. Papageorgi, T. Gomes and J. Lanipekun, "The development of practising strategies in young people," Psychology of Music 40/5 (2012).
5. T. Mathias and A. Goldman, "How Does Increasing Contextual Interference in a Musical Practice Session Affect Acquisition and Retention?" Journal of Research in Music Education (2024).
6. C. Carter and J. A. Grahn, "Optimizing Music Learning: Exploring How Blocked and Interleaved Practice Schedules Affect Advanced Performance," Frontiers in Psychology (2016).
7. M. Wiseheart, A. A. D'Souza and J. Chae, "Lack of spacing effects during piano learning," PLOS ONE (2017).
8. J. Allingham and C. Wöllner, "Slow practice and tempo-management strategies in instrumental music learning," Psychology of Music (2022).
9. J. M. Silveira and R. Gavin, "The effect of audio recording and playback on self-assessment among middle school instrumental music students," Psychology of Music (2016).
10. W. B. Verwey, "Concatenating familiar movement sequences," Acta Psychologica (2001); K. Sakai, K. Kitaguchi and O. Hikosaka, "Chunking during human visuomotor sequence learning," Experimental Brain Research (2003).
11. Playing-related musculoskeletal disorders among percussionists, University of North Texas Musician Health Survey (N=279); and "Injury Prevention Considerations for Drum Kit Performance" (N=831).
12. E. Sayers, "Pedagogical strategies for the development of improvisation and composition in North Indian classical music," Frontiers in Psychology (2025).
13. Dard Neuman, "Pedagogy, Practice, and Embodied Creativity in Hindustani Music," Ethnomusicology 56/3 (2012), 426–449.
14. Daniel M. Neuman, The Life of Music in North India (Wayne State University Press, 1980; University of Chicago Press, 1990), chapter 2.
15. James Kippen, The Tabla of Lucknow: A Cultural Analysis of a Musical Tradition (Cambridge University Press, 1988), chapter 6; and Gurudev's Drumming Legacy (Ashgate, 2006).
16. Max Katz, "The fly-by-night ustad: Problems of music education in North India, 1863–1915," Modern Asian Studies 59/2 (2025), 309–345.
17. Shawn Mativetsky, "Permutations, a tabla player's best friend," shawnmativetsky.com; "Understanding the Tabla Solo Structure," Percussive Notes, September 2017, 76–78; and interview, 22 October 2020. 17a. K. Siedenburg, S. Mativetsky and S. McAdams, "Auditory and Verbal Memory in North Indian Tabla Drumming," Psychomusicology: Music, Mind, and Brain 26/4 (2016), 327–336. 17b. David Clarke and Vijay Rajput, Rāgs Around the Clock (Open Book Publishers, 2024), §1.12 "Riyaz." Open access. Includes Gerry Farrell, "Teaching Indian Music in the West," British Journal of Music Education 3/3 (1986), and Ajoy Chakrabarty, Shrutinandan (2002). 17c. Sanjoy Bandopadhyay, "Thoughts on how to connect Music Education with NEP-2020," sanjoybandopadhyay.com, September 2025. 17d. Arvind Parikh, FAQ corpus, panditarvindparikh.org. 17e. Deepak Raja on Ameer Khan, swaratala.blogspot.com, August 2011. 17f. Bhartiya Sangeet Kalapith, riyaz method articles, sangeetjagat.omkarsangeet.org, August 2023. 17g. R. K. Badasheshi, "Importance of Riyaz and Its Forms," Swar Sindhu 12/3 (2024), 424–427. 17h. Sadanand Naimpalli, Theory and Practice of Tabla (Popular Prakashan, 2005), riyaz chapter; and Tabla for Advanced Students, preface. 17i. Amit Chaudhuri, essay on riyaaz, Scroll.in; and Finding the Raga (2021). 17j. Shubha Mudgal, reported remarks, Gomantak Times, 23 September 2022. 17k. Samir Chatterjee, Those Forty Days (a day-by-day journal of a forty-day chilla). Not consulted.
18. David Courtney, cyclic forms and glossary entries, chandrakantha.com.
19. Zakir Hussain, interview, NPR Morning Edition, 8 January 2015; Art Talk, National Endowment for the Arts, 12 May 2016; Berklee master class, December 2019.
20. Suresh Talwalkar, interviews, The Hindu, 14 July 2022; and The Indian Express, 14 December 2018.
21. Sanju Sahai, interview by George Howlett, Darbar Living Traditions.
22. Swapan Chaudhuri, interview, Chitresh Das Institute, May 2017; and Jan Haag profile, raga.com.
23. Yogesh Samsi, interviews, The Daily Star, 29 December 2016; and Free Press Journal, 15 February 2025.
24. Aneesh Pradhan, interview, Free Press Journal, 13 April 2024.
25. Amjad Ali Khan, interview by Graham Reid, Elsewhere, 2019; and CNN TalkAsia, 14 January 2005.
26. Ravi Shankar, My Music, My Life (1968).
27. Akhil Bharatiya Gandharva Mahavidyalaya Mandal, examination structure, Prarambhik level.