For much of the twentieth century, the prevailing account of jazz improvisation cast the improviser as a solitary inventor, conjuring melodic lines that had never before existed and would never recur. This romantic picture, congenial to critics who prized spontaneity as the music's defining virtue, treated the recorded solo as a fossilized instant of pure creation. Yet the picture proved difficult to reconcile with what analysts began to notice once they transcribed and compared large bodies of recorded solos: improvisers, even the most celebrated, reuse. Recurrent melodic figures - the trade calls them licks or formulas - surface across a player's career, transposed to fit new harmonic settings but recognizably the same. The discovery unsettled the romantic account, for a phrase deployed on a hundred occasions can hardly be described as invented on any one of them. Some scholars concluded that the romantic account should simply be discarded. On their view, improvisation is better understood as the fluent recombination of a stock of learned material, closer to the extemporized oratory of a practiced speaker than to the sudden visitation of the muse. The analogy is instructive: a skilled orator does not compose each sentence from nothing but assembles familiar phrases, rhetorical turns, and cadences into speech that is nonetheless responsive to the moment. On this account the formula is not a failure of creativity but its very medium; without a shared vocabulary the improviser could neither be understood by fellow musicians nor build the long-range coherence that distinguishes a memorable solo from a string of unrelated gestures. This corrective, however, risks overreaching. To show that improvisers draw on formulas is not to show that improvisation reduces to their retrieval. The decisive question is not whether a phrase has been used before but how it is placed: the same formula can fall flat or electrify depending on its harmonic timing, its relation to what the rhythm section is doing, and its distance from the phrase that preceded it. A transcription flattens these dimensions, recording which notes were played while discarding the pressures under which the choice was made. What the formula theorists have established is that the raw materials of a solo are largely inherited; what they have not established, and what their method cannot by itself establish, is that the arrangement of those materials is likewise a matter of retrieval rather than judgment. The most defensible position, then, treats the two accounts as describing different strata of a single act. At the level of vocabulary the improviser is a curator of the already-said; at the level of syntax - the ordering of phrases in real time against a shifting harmonic and rhythmic ground - the improviser exercises precisely the responsiveness the romantic account prized, though not the invention from nothing it imagined. The romantic critics erred in locating creativity in the novelty of the material; the formula theorists, in supposing that inherited material forecloses creative choice. Improvisation is neither invention from nothing nor recitation from memory, but the disciplined placement of the familiar.