Gat: The Fixed Composition at the Heart of Sitar Playing
Everything before this point — the slow unfolding of alap, jod, and jhala — has been improvised, with no tabla and no fixed cycle to return to. The moment the tabla enters, that changes. The performer settles into a short, memorized melodic composition set to a specific taal, and everything that follows — every improvised phrase, every fast run — keeps circling back to it. That fixed composition is the gat, and it's arguably the single most sitar-specific idea in this whole tradition: gats are instrumental compositions, written for the instrument's own idiom, not adaptations of vocal material.
The mukhda: a phrase you keep coming home to
A gat's defining feature is its mukhda (literally "face") — a short, fixed opening phrase, precisely placed against the taal, that the performer returns to again and again between passages of improvisation. Every time a mukhda lands, it resolves cleanly onto the sam — the first, most important beat of the cycle — giving both performer and listener a recurring point of arrival. If you've used this site's Songs feature, you've already seen this structure labeled directly: notation lines marked "Mukhda" and its later "Mukhda (reprise)" are exactly this idea, borrowed into a filmi-song context. Some gats also have a second section, the antara, that moves into a higher register before returning to the mukhda — the same antara/mukhda structure that shows up in those same Songs.
Two named traditions, two different tempos
Sitar repertoire draws heavily on two named categories of gat, traditionally associated with two different composers and, more usefully for a listener, two different tempo characters:
- Masitkhani gat — a slow (vilambit) composition, usually set in teentaal, historically attributed to the musician Masit Khan. Masitkhani gats tend to be melodically spacious, leaving a lot of room for ornamentation — meend and gamak especially — and for a performer to develop the raga slowly within the taal, almost as a continuation of the alap's exploratory mood but now with a cycle to land on.
- Razakhani gat — a fast (drut) composition, also commonly in teentaal, attributed to Ghulam Raza Khan. Where a Masitkhani gat gives room to linger, a Razakhani gat is built for speed and rhythmic precision — fast mizrab patterns, quick melodic runs (taans), and a more percussive relationship to the taal.
As with a lot of gharana-era attribution (see our piece on gharanas), the exact historical details behind these two names — who these composers were precisely, and the full chain of transmission — rest heavily on oral tradition and aren't uniformly agreed on across sources. What's not in dispute is the pattern itself: a huge share of sitar recitals move from a slow Masitkhani-style gat into a fast Razakhani-style one, mirroring the alap-to-jhala arc at a smaller, taal-bound scale.
Not a song, a scaffold
It's tempting to think of a gat the way you'd think of a song's melody — something performed roughly as written. That undersells it. A gat functions more like a scaffold: the mukhda and antara are fixed and returned to precisely, but the space between them is where a performer improvises taans (fast melodic runs) and elaborates the raga further, all while staying anchored to the taal. The gat isn't the performance — it's the structure the rest of the performance is built around.
Hear the shape yourself
The clearest way to feel what a mukhda actually does is to follow one in real time. Practice this: open any piece in Songs and watch how the notation highlights jump back to the same "Mukhda" line each time it recurs — then open that song in Compose and try isolating just the mukhda phrase with loop selection, so you can hear it land on the sam over and over, the way a live tabla-and-sitar performance would.