Jazz improvisation

Why does a jazz soloist sound freer inside a fixed chord scheme than with no rules at all?

Hand a competent player a twelve-bar blues and they will sound inventive. Hand the same player a blank canvas — no key, no tempo, no form, play anything — and what often comes out is hesitant, or busy, or oddly featureless. This is backwards on the face of it. The second situation contains every option the first contains and vastly more.

So either the constraint is doing something we have not accounted for, or "freedom" is not the same quantity as "number of available options". Which is it?

Begin by separating two things the word freedom carries. One is the size of the option set. The other is the capacity to make a choice that actually lands — one the player can execute, the band can follow, and a listener can hear as a choice. Removing every rule maximises the first. Ask what it does to the second.

Consider the clock first, because it is unforgiving. Improvisation happens at tempo. A soloist at a medium swing tempo places several notes a second and has no interval in which to deliberate. Whatever they play must come from something already prepared — shapes, figures, ways of getting from one chord to the next, practised until they are available without thought. A fixed chord scheme is what makes that preparation applicable: a figure worked out over one common progression is usable every time that progression appears, in any tune, in any key. Remove the known harmony and the stored material has nothing to attach to; the player is thrown back on what can be computed in a fraction of a second, which is far less than they know.

Now the second party to the problem: the soloist is not alone. Bass, drums and piano must be somewhere compatible at every instant, with no conductor and no score. They could manage by listening alone — but only reactively, always slightly behind. A shared chord scheme lets each player predict where the others will be several bars ahead rather than infer where they just were. It is a coordination device, and a cheap one, carried in memory rather than negotiated in the moment.

Then the point I think is strongest, and least obvious. Expression in music is largely deviation, and a deviation is perceptible only against something expected. A note outside the chord sounds outside only because there is a chord. Delaying a resolution is expressive only where a resolution was due. The bent note, the substitution, the phrase that starts late and catches up — each is a departure, and departures need a baseline. Where nothing is expected, nothing can be surprising, and a freedom the listener cannot perceive as freedom does not reach them at all.

The form supplies one thing more: a clock at a larger scale. A twelve-bar blues or a thirty-two-bar song form gives repeating boundaries, so players know when to hand over, how to build across successive choruses, and where an ending can be. Shape becomes available, rather than merely duration.

Does the argument survive its obvious counterexample? Free jazz exists — Ornette Coleman, Cecil Taylor, the later Coltrane groups — and it is not chaos. But look at what it does rather than what it claims. It replaces the chord scheme with other structure: motivic development, agreements about register and timbre, conventions about who leads and who supports, a tonal centre held collectively by ear. So the honest conclusion is not that constraint is necessary in general, but that some shared structure is, and that the chord scheme is an unusually efficient one — learnable, portable between tunes, transmissible without rehearsal. It is also telling that free playing demands more continuous listening rather than less, which is why it is mostly done by players who had already internalised the changes.

The refinement that ties the strands together is that it is not constraint as such which helps, but shared constraint.

That distinction matters, because the popular version of the idea — creativity thrives on limits — is far looser than the reasoning above supports. An arbitrary private restriction is a discipline, and may be useful for practice, but it does none of the three jobs identified here. It does not let the bassist predict you. It does not make your stored vocabulary retrievable. It does not give the listener a baseline against which your choices register. What earns its keep is a frame the player, the band and the audience hold at once — which turns "constraints enable freedom" into something more specific and more testable: constraints that are common knowledge enable coordination, retrieval, and legibility.

Two honest qualifications. First, the balance between recall and generation inside a solo is genuinely contested. Analyses of recorded solos find a great deal of recurring, transposed material, which supports the vocabulary account; how much of the rest is assembled fresh, and what it would even mean to draw that line, is unsettled. Take the claim as "prepared material is essential", not "solos are recall".

Second, the general psychological literature on whether constraints improve creative output is mixed and heavily task-dependent, and this explanation deliberately does not rest on it. The argument is drawn from features specific to the practice: playing at tempo with no time to deliberate, coordinating an ensemble with no conductor, and expressing through deviation from a baseline the audience shares.

• How a rhythm section negotiates a substitution nobody agreed in advance.
• Whether the same three jobs — coordination, retrieval, legibility — explain form in improvised traditions with no harmony at all, such as raga or maqam.

Every term the collection defines is gathered in the glossary.