King Oliver and Louis Armstrong in 1923, Part 5: "Dipper Mouth Blues" & What is Improvisation?

I hope you have been following the previous 4 parts of this series. Last time we discussed the authorship of “Dipper Mouth Blues,” the most famous piece by the King Oliver group with Louis Armstrong. They recorded it twice. Now, let’s go to the music. Today, we’ll focus on the clarinet solos by Johnny Dodds: Here is the clarinet solo from the first recording, April 6, 1923:

The first chorus start high and swoops low. The second, starting at 0:15, starts low and builds up.

And here is Dodds playing on June 23, about two-and-a-half months later. This performance is at a faster tempo:

The first chorus starts high and swoops low. It is exactly the first chorus from April 6. And the second chorus at 0:14….wait a second, it’s an exact repeat of the first chorus! He played the same chorus twice, instead of playing something different in the second chorus as he did before.

So, do you mean to tell me that Dodds was less free, as opposed to more free, after playing the song in public another three months, usually seven nights a week??!! How do you explain that?

Well, the standards—the aesthetic— about solos during the early days of jazz were precisely the opposite of what they are now. Today, if you play the same solo on a certain tune every time, the other musicians in the band will make fun of you, start to whistle along with you, and so on. But in the early days, they would simply say “I love your solo on that tune.” And by “your solo” they indicated that you were playing the same solo every time. But it was yours—you made it up, maybe at first you improvised it, and if you refined it over time, and even worked on it at home, so much the better. Ellington reportedly on occasion explicitly told one or more of his musicians that he liked the solo they played on a certain number, and that they should keep playing it just that way.

In other words, the entire concept of improvisation was different in the early days. If you took the time to work out a solo that you would play on a certain piece, that was considered a good thing, not cheating. And if you worked it out before a recording, that was even better! On a recording session, why take a chance that you might make a mistake? Remember, there was no such thing as splicing until tape was available in the late 1940s. If you made a bad mistake on a 78 rpm disc, the disc was often not usable, and it had to be done over. (Of course, that depended on how bad the mistake was. Sometimes, as I showed in my Lil Hardin essay, they would just leave the error in and move on to the next piece.) If a retake was required, that meant that the whole band and the recording staff would have to put in extra time, just for you. If you planned your solo at home, everybody appreciated it. Nobody would make fun of you for playing the same solo each time.

One of the few people to understand this early approach to soloing, and to write about it, is Robert Bowman, longtime professor of ethnomusicology at York University in Toronto. You may know him as a leading expert on soul music and Stax Records, but his 1982 Masters thesis was the most detailed study ever conducted of the 1923 Oliver recordings, with notated transcriptions. One of his conclusions was that the band would work on a new piece for a while, and in general, after some experimentation, each musician would settle on musical lines that seemed to work best. Those lines then became memorized parts of the piece, repeated both on the bandstand and on recordings, with small variations in the details.

Armstrong himself said just about the same thing, in a 1966 interview with Richard Meryman: “Once you got a certain solo that fits in the tune, and that’s it, you keep it. Only vary it two or three notes every time you play it.” (The interview was originally published in Life magazine, but this passage was only in the more complete version that was published as a short book.) In short, either Dodds himself, or Oliver, or both, must have said sometime after the first recording, “That first chorus is terrific. Why not just do it twice, instead of playing something different the second chorus?”

The bigger point, which I will talk about at greater length another time, is that the common stereotype that New Orleans players were more free, and more wild in their improvising, is what is known as a "primitivist myth." It's based on an idea that people who have less institutionalized education and training are more “free.” That is false in music, just as it is in most aspects of life. (Of course, there is also the issue that much training occurs outside of institutions, on the job and at home.) In fact, all of the evidence that we have of early jazz, and there is quite a lot—sheet music, interviews, recordings from before jazz, as well as early jazz recordings— points in the opposite direction: Early jazz musicians improvised cautiously, not with wild abandon. The idea that Buddy Bolden was a wild free improvisor is total and complete fantasy. The contribution of Louis Armstrong and Earl Hines was precisely that they improvised less cautiously than the first generation of jazz artists, and took more chances. The contribution of Lester Young was to introduce even more freedom, and when you get to Charlie Parker, you're talking about someone who can truly and spontaneously create three totally different takes on “Parker’s Mood” in 1948. Parker’s art represented a new standard that is basically still in place, so many years later—the expectation that every take should be a fresh, new improvisation.

Again, this is a huge topic, and I don't intend for this to be the full discussion of it. At a later time, I will present a series about what I call “the history of improvisation.” But in short, that's why Johnny Dodds on his second recording of “Dipper Mouth Blues” is more cautious, instead of more free.

Next time we will listen to the two recordings of “Dipper Mouth Blues” complete, and we’ll focus on both cornetists, among other things.

All the best,
Lewis