The Jazz Repertoire

(with Robert R. Faulkner)

We have begun a study of the jazz repertoire, based first on our own experiences and memories, but more importantly on contemporary field work and interviews with working musicians.

We begin with a simple observation; every evening, and especially on weekends, all over North America, musicians who may not know each other at all and who, quite often, have never rehearsed together, perform in public for several hours, with a minimum of discussion or preparation. They perform well enough to satisfy the people who hired them and, quite often, well enough to satisfy themselves. Our research is designed to answer an equally simple question: how do they manage to do that?

The simple question has a simple answer: they can play together that way because they all know the same songs--they share a repertoire--so that all that is necessary is for someone to name one of the songs they know in common, perhaps mention a key, and then count off two bars--after which they play as though they had been playing together for years. But that simple answer, while more or less correct, is only the beginning of a complicated story. Because none of what we have just offered as the answer is exactly correct--more or less correct, in a general way, but differing on every specific occasion. And the differing occasions and their differing consequences let us see how this mechanism of producing concerted action really works.

The Basic Units of Analysis

Four elements furnish the raw material for the complications to follow:

1. Songs: The repertoire of jazz players (of working musicians in general) is made up of songs. A song is a melody with its accompanying harmonic structure (and, optionally, its lyrics).

2. Performers: Musicians (and singers) who perform songs in a variety of situations.

3. Performance situations: Musical performances take place in settings characterized by their location, their personnel, and the demands they make on the performers.

4. Working repertoire: Performers choose from the songs available to them to create a performance that takes account of the characteristics of the situation. Let that stand, for the moment, as our statement of what we are talking about, the elements and the process that make up the activities for which we are using "repertoire" as shorthand.


We've observed three kinds of songs played by musicians working in ensembles:

a) traditional songs, often created by the "folk" and preserved in the memories and activities of non-professional performers (e.g., "Happy Birthday").

b) songs written by professional songwriters for popular consumption. I'm not sure when this industry came into existence (in the U.S., maybe around 1900) but by the 1910s and beyond it was flourishing, producing thousands of songs a year.

c) Songs written by jazz players, mainly but not entirely for instrumental performance, thus mostly without lyrics.

Think of all of these songs--all the songs ever written or recorded or played anywhere, however they are stored and preserved--as constituting a reservoir of what performers could play if they wanted to and conditions allowed them to.

A song is a short composition with some repetitions that takes one of a few forms. There are variations but they are not great. Here are the main types:

The simplest form in the repertoire is the blues: twelve bars long, with a simple harmonic structure. The more or less traditional melodies are relatively unimportant, though musicians often compose melodies that then become a specific repertoire item. To say "Blues in B Flat," adding the key, gives enough information to the other players to make a creditable performance possible.

The most common form of the songs in this repertoire is: 32 bars, AABA or ABAB format (where each letter stands for an eight bar segment), with a range of no more than a tenth and usually less, and most of the melody moves scalewise or through arpeggios. (More notes became available as major melody notes over the decades.) When performers learn a song, they take this structure as the norm and note variations from it.

Harmonies conform to a few standard patterns: II-V-I, and the circle of fifths variations on that, for instance. Other variations are created by key changes (up a major or minor third are most common). Like melodies, harmonies got more complex and varied over the decades. There are other variations, which we won't go into here, other than to note that songs written for performance by jazz groups often depart even more from standard patterns.

In general, the songs are so formulaic that describing a song as "I Got Rhythm with a Honeysuckle bridge" is enough for other players to be able to play the song (given that at least one of them knows the melody, though even that is not completely necessary). Most songs are elaborate variations on a few templates. A person who knows the basic forms can play thousands of songs without difficulty.

To be available to performers, songs must be preserved and distributed. This is commonly done by publication, the printed song being available for purchase, and many places store amd/or sell old sheet music. Written songs can also be distributed informally, via xerox copies, as can songs which have been written but never published. Songs can also simply be preserved in memory (the Fahrenheit 451 method), as I preserve many of the songs of the 1930s in my memory. Finally, songs are preserved on recordings, which serve as audible scores.

To acquire a song in one of these ways takes certain skills--being able to read music or to reproduce a song after hearing it.


Performers play songs in performance situations, to which they bring a body of songs they have learned and practiced, which they can perform in a variety of ways. This is their personal repertoire.

Performers learn songs by reading them from written scores or by hearing them in public performances, on the radio or on recordings. What's available varies greatly. In the 1930s people could learn the constant stream of new songs from the radio; that became less possible as radio programming changed. Jazz players typically treat the recordings of well known players and singers as privileged sources of material (a song recorded by Chet Baker has a better chance of being heard and learned by players looking for new tunes than one recorded only by commercial dance bands).

To "know" a song completely means knowing the melody as written or played by the composer or an admired performer, and the underlying harmonies players typically use as a basis for accompaniment and improvisation. In theory, players know these things about every song they might play, but they often know only one, and that not very well. But, even then, the group can play the song, because the formulaic character of these songs makes it easy to guess at what's missing.

Players' repertoires vary idiosyncratically in how much and what they contain, and players change their opinions of songs as they become familiar with them, usually eventually tiring of them and replacing them with others in the "active" repertoire.

Playing Situations

Every playing situation makes specific demands for kinds of music on the musicians, who must have the skills and knowledge to do what's required.

Most performance events are audience oriented. In a public event, an audience pays (by buying tickets or drinks) to hear music and perhaps dance to it. What musicians must be able to play depends on the audience's tastes, as those are shaped by ethnic ("O Sole Mio." "My Yiddishe Mama," Does Your Mother Come From Ireland") or class culture, by generational experience (songs of the 30s or 60s) or even occupational peculiarities (e.g., gamblers). Audiences change and what a musician must know changes accordingly.

In a private event, the host of a party pays for the guests to be entertained. In either case, the people who pay usually know what they want and insist on it--this kind of song, that kind of dance. These demands may be quite idiosyncratic (the bride's father insists on the songs he asked for).

Some events, however, are musician-only events, where the performers' only concern is with what the other musicians present think. In such a setting, no one will insist on a tune musicians despise. Such situations let performers try out new things, adding to one's repertoire, experimenting with new formats, and setting up some new possibilities for collective work.

Working Repertoire

So performers, who have learned some but not all of the possible songs they might have learned, come together in playing situations where they have to play for a certain amount of time for whoever is there. The payoff for thinking of repertoire this way lies in the result you get when you specify the character of each of these inputs. This specific group of players playing in this specific place with its specific demands and using the specific repertoire that is, one way or another and at one level or another, available to them individually and collectively, will create, on the spot and perhaps for this one time only, a working repertoire: this working repertoire--the specific list of songs they play on that occasion.

In the most general case, a working repertoire has no temporal extension, it exists (analytically) for that one time only. In fact, of course, the same performers often work together in the same situation (or the same kind of situation) for long periods of time, and the working repertoire they develop becomes a resource they can call on again and again. They've worked out what they jointly know, what the situations demand or will tolerate, and have a list of tunes that only needs to be selected from and arranged sequentiallyfor them to produce an evening's performance.

The process of constructing a specific repertoire for an occasion means noticing that it's time to play, negotiating what to play, deciding whether they collectively know that tune or want to play it, whether (if they don't know it) they will try it anyway. The details of the negotiation tell us make explicit the constraints of the performing situation.

When a group plays together for many nights over many weeks, they will learn what the others know, what tunes they know in common, where the overlaps fail. They will also learn what others are willing to do. How much trouble are they willing to go to to learn something new? Will they take chances on something unfamiliar?.There is a process here too. One member of the group may be the de facto finder of new things which he then proposes to the others to learn and play on the job. Or that function might be shared.

The main elements of this negotiation are:

1.  Selecting elements from a pool of available resources.  

2.  Ordering and prioritizing the selected elements into a specific performance, a working set.

3. Adapting, tailoring and assimilating the preliminary set to the demands of the situation, to what audiences, employers, and other agents of social control require.

The working repertoire a group develops in this way consists of a more or less stable number of tunes they all know and are prepared to play, and a disposition (variable from group to group and maybe also from night to night)  to learn new things and try them out in the work situation.

The resulting repertoires can be classified according at least to three characteristics:

a) the number of tunes available as resources (the pool)

b) diversity: the variety of tunes the group is prepared to play, or how different the songs are that are selected from the pool

c) Variability of set and performance night after night (do we play nothing but bossa novas or bebop or popular tunes of the day, or do we play anything we want to?). How much room for variation in style of playing do the players have? Can they improvise freely, or must they stick to he melody as someone has recorded it, or to a solo someone has recorded?

Different kinds of repertoires result from the possible combinations of these elements. Here are some of them for which we have empirical examples (and after which they are named):

The "Tristano" (named for Becker's piano teacher Lennie Tristano, who followed this pattern strictly): A very few tunes (in Tristano's case, no more than a dozen) are the core set, played in no particular order. Everyone knows them. This repertoire is characterized by low turnover in tunes selected and few new tunes introduced; the old ones are sufficient to learn and improvise on.

The "Pinardi" (named for a trumpet player Faulkner knows in Amherst): New tunes, in a variety of styles, are continually introduced into the group, and a constantly changing selection is made from this variety. You may play a tune one week during several performances and then not play it again for months. The "set list," the tunes played in order on a specific occasion, varies greatly from night to night, with a constant introduction of new material.

The "Messer" (named for a guitarist Faulkner plays with frequently): The band works on a fixed set of tunes for a particular occasion or kind of occasion (a concert, a recording session), producing a well-rehearsed and highly competent version of that material.

The "MacLeod (or Club date)" (named for Bruce MacLeod, who described it in his book Club Date Musicians): A great variety of songs is tightly organized, for an entire evening, by a leader intent on pleasing the clients who have hired him for their party. A variant is the so-called "ghost band," which reproduces the performances of a band which no longer exists, but whose work is known from recordings (bands which thus mimic the performances of Duke Ellington, Count Basie, Glenn Miller, and others, continue to perform in this way today).

Each kind of repertoire works in a specific kind of playing situation. So the "Bobby Laine" works best in places like the Chicago clubs Becker worked in with him, where the band played seven hours every nights, for an audience which did not really listen, and where it was likely that the performers would be bored and thus looked for variety to keep their interest alive.


Such a way of thinking raises the researchable questions we are now trying to answer: how do musicians create, from the reservoir of materials they know in common, a working repertoire for the situation they find themselves performing in? What are the varieties of processes that occur and what are their results? And, finally, what does the detailed inspection of this case tell us about the general process of repertoire formation? Because "repertoire" has now become a common way of talking about "culture" and similar concepts, although the term is usually used metaphorically. What if we study it where it really is a repertoire in the strict sense?