Tatum's Dissonant, "Avant-Garde" Side, Part 1, + Bonuses! (Updated); His Method & His Predecessors

With bonuses for paying subscribers!

(NOTES: I revised and enlarged the discussion of the chord from “Yesterdays,” thanks to a question from subscriber Peter Feng.)

(Paying Subscribers, at the bottom of this article you will find your latest bonus!)

I’ve learned from experience that once people decide that they know something, it’s extremely difficult to get them to rethink it. (Of course, this could be applied to many situations! But let’s stick with jazz.) I’ve noticed, for example, that a few of the responses to my posts, even though they were positive, were from persons who clearly had not listened to a single one of the audio examples. They assumed that they didn’t need to listen, because they already knew “everything” about the artist in question!

So let me begin by saying, no matter how well you know the music of Tatum, please listen to every musical excerpt in this post—all of them together will take you maybe five minutes to hear, so it’s not a huge time commitment. And prepare to be shocked!

Everybody knows that Art Tatum (1909-1956) was “one of the greatest jazz pianists of all time,” or “one of the greatest technical virtuosos in jazz,” or something to that effect. If you google his name you get phrases like that. But read further, and the contradictions begin to pile up. You’ll learn that he was heavily influential, but you will also read the opposite — that his style was so personal and technical that few musicians emulated him. My late colleague Gunther Schuller argued as much in his book The Swing Era, noting that the beboppers “pretty much ignored Tatum,” which is absolutely false. (The Swing Era is monumental and essential, but it’s also a maddeningly inconsistent book.)

Another contradiction: Some applaud Tatum as supremely inventive, while others say that he was boringly repetitive, and that he barely improvised. French composer and jazz historian André Hodeir made the latter case in a famous article of 1955; Schuller argues both sides somehow.

There are many published transcriptions of Tatum’s recorded work, and even a book of Tatum’s fast runs (Riccardo Scivales’ The Right Hand According to Tatum, 1998). In 1994, James Lester published a biography, Too Marvelous For Words, but it’s mostly about the life, more than the music. (There are also at least three unpublished Ph.D. dissertations on Tatum, but they include general music discussion, mostly without notation.) Schuller is perhaps the only published author who analyzes what Tatum does, rather than simply documenting what notes he plays.

Outside of Schuller, the main thing we read about Tatum is that he played very fast. But if the only remarkable things about Tatum were his speed and technique, I’m sorry, but that would not have been much of an accomplishment. Every year, classical conservatories turn out graduates at or beyond Tatum’s technical level — including pianists who can easily play transcriptions of Tatum’s solos at the correct tempos.

Overall, the consensus seems to be that Tatum was an amazing technician who specialized in original and imaginative — but mostly memorized — arrangements of popular songs. Really?? Then stop right there!—because, as I will show, that is the opposite of what the recorded evidence shows. (When I make a statement about “recorded evidence,” it is always based on my having listened to every recording by the artist in question. Yes, I said every one, including bootlegs and unissued items.) I see Tatum as a highly experimental artist who was confined by having to play standard show tunes all his life. When he sounds bored and facile, it’s because he is indeed bored. But when he’s on... Wow! Significantly, many of Tatum’s most daring recordings are from live performances, not intended for release — settings in which he felt free to experiment.

Tatum's music is full of astonishing details that the casual listener could easily miss—and apparently most people have missed them. I've learned that in order to fully appreciate him, you have to listen to a lot of his recordings, intently and without distraction. Do that, and you’ll start to hear three amazing things he does:

1. THICK, DISSONANT CHORD VOICINGS. (“Voicings” are the many ways that any chord can be played.) Tatum often plays chords so rich and full of dissonances that they sound like “clusters” of notes, rather than conventional chords. He does this most often on dominant chords, because those chords can support many “extensions” (that is, added upper notes). So here’s a suggestion: On your second listening to any Tatum recording, hit “pause” at every chord that sounds interesting. Let it ring in your mind, then back up and hear it again. Here is an example, of the kind of “cluster” or “crush chord” that I’m talking about. This one is near the beginning of a 1955 version of “Tenderly,” from the double album 20th Century Piano Genius. What is the ringing chord at the end of this 3-second excerpt?! Play it a few times (and if you can notate it, please send it to me):

Here is another voicing that he used regularly: It appears in the song “Yesterdays,” at the beginning of the 9th measure, where a vocalist would sing “Olden days.” I usually play an A dominant chord with a sharp 4 (some would say sharp 11—D#, aka Eb), at this point. The right hand melody note is F, the sharp 5 (that is, E#), so many musicians would include that in the chord as well. Tatum’s chord has those notes, A, Eb and F—but also a whole lot more! First he hits with the left hand a low A, then a higher A. Then he continues to hold the second A, and plays these notes—please read from the bottom up:

Left hand: (A still ringing) Eb G Db Right hand: F Ab Db Eb F

Theoretically, the only note here that is really dissonant is the Ab. But that’s on paper—that’s not how it sounds. As pianist and educator Dave Frank has explained, Tatum often voiced chords in such a way as to intensify their richness and complexity (what I call “thickness”), by spreading out the notes in certain ways, and doubling some notes (for example, this one has Db and Eb in both hands).

In this case Tatum has divided the notes in such a way that it sounds like an A dominant chord (with a raised fourth) and a Db major chord on top of it. Let me play it for you. The left hand is quite a stretch (it seems that he could reach a 12th in each hand—that’s, for example, from C up to the second G above it). It appears that he keeps his left pinky on the A, but I had to let go of the pinky to play the chord for you on my Yamaha digital piano. I played it three times, the last time with the notes played separately:

Now, here is Tatum playing that chord in “Yesterdays” on a TV variety show in 1954—the chord comes at 0:27. And here is someone’s transcription of that TV performance, played on a synthesizer using MIDI; the chord is at 0:17. Finally, this performance of “Yesterdays” begins quite differently, but he plays the same chord at 0:19.

I’m sure that I’m driving you crazy with this chord by now! But the point is, this is quite a wild chord. James P. Johnson, Fats Waller, and Teddy Wilson are some of the greatest musicians of all time, but they did not ever play chords like this. And I’m telling you, this was typical for Tatum. Even today, few pianists regularly use voicings that are as “out” as Tatum’s.

2. RADICAL CHORD SUBSTITUTIONS. He was the master of “substitutions” — playing a different sequence of chords behind the melody than is usually played. He did this a lot, and his substitute chord progressions are often radically far “out.” We will hear several examples.

3. PLAYING MELODY NOTES OUT OF THE KEY! In Tatum’s generation, pianists did not usually play flowing, saxophone-like improvisations in the right hand. After all, this was well before Bud Powell, Hank Jones and the like. But on rare occasions, Tatum did create such lines, sometimes incorporating notes outside of the key, creating a striking effect.

OK—let’s hear some examples, mostly in chronological order, and let’s begin with an instance of #3, playing notes outside the key. The earliest example I know of Tatum doing this comes at the end of “After You’ve Gone,” recorded in New York City on August 24, 1934. Listen Please:

First, at 0:11, is a very strange chord sequence. And at the very end, starting at 0:25, did you hear the right hand?! Say what??! He ends on a G major chord in the left hand, but the right hand briefly suggests the scales of Eb major, Bb major, E major, and something like Eb minor!! Here is a transcription played on a synthesizer using MIDI—the passage above starts around 2:19. And here is the entire recording with a notated transcription. I’ll share the very end of that below.

In just a bit I’ll talk more about what he played there. But first, here’s something interesting: He recorded two more takes of this song for the same company, Decca, on October 9, 1934. One take has never been issued and most likely no longer exists, but the other take was released. This version is quite similar to the one from August, except for the important difference that the ending has no dissonances. Here is that ending, the one that does not employ bitonality:

So, why did the recording company (aka the “label”) ask Tatum to redo it? Sorry to say, nobody knows, not even jazz historian Steven Lasker, who has done more research on Tatum’s Decca recordings than anybody. I initially thought it must have been because the ending of the August version was too far out. But Steven pointed out to me that six other numbers were redone on that October date, and those had no such outrageous moments. However, Steven also noted that the Decca label in America had only just begun, and the August session was one of their very first recordings. He notes that perhaps the label heads and the recording engineers felt that they were producing higher-fidelity product in October than in August, which would be reason enough to re-record them. However, that is an educated guess, and even if it’s correct, I would add that it’s still conceivable that they said to Art, “As long as we’re redoing ‘After You’ve Gone’ anyway, to get better sound, can you give it a more standard ending?”

But let’s move on to the big question: What inspired Tatum to come up with such wild musical ideas? This was unprecedented in jazz, not even found in the innovative solo recordings of Earl Hines. However—it was somewhat “current events” in classical music. After all, this was the era of the fantastic innovations of Debussy (1862-1918), Ravel (1875-1937), Schoenberg (1874-1951), Stravinsky (1882-1971), and many more. And there were a number of composers who specialized in short piano miniatures that incorporated the modern sounds. For example, listen to and look at this little repetitive passage from a musical impression of a “Rainbow Trout” flitting around in the water. Written by English composer Cyril Scott (1879-1970), the sheet music was published in 1916, and the audio was recorded by the composer in 1928:

(Paying subscribers, scroll down for the complete sheet music and audio of Scott’s delightful piece. THANK YOU as always for your support.)

I’m sure you can hear what I would call a “family resemblance” between Tatum’s ending and Scott’s piece—that is, they don’t use the same notes at all, but there is some similarity in the textures and in the use of dissonance. Scott’s passage is based on two diminished chords—the top notes move down from C# to Bb to G to E, while the bottom notes are a fourth lower. Under that the left hand plays a sort of E major chord. (In other works such as his first piano concerto, Scott demonstrated other possibilities of diminished chords.)

On the other hand, Tatum’s passage is more of an unstructured exploration of different scale fragments, as I described above, not a worked-out pattern like Scott’s. Here is a transcription of the last four bars of the Tatum that you heard above. (I don’t agree with all the accidentals—that is, flats versus sharps—but the pitches seem to be correct.):

Despite the differences, I think we can assume that Tatum knew of pieces like Scott’s. He did not have to know the “Rainbow Trout" specifically, because there were many other composers of that time who were writing short piano pieces that pushed the boundaries of tonality. But it’s no coincidence that Downbeat, October 1935 wrote of one of Tatum’s reharmonizations, “Cyril Scott would have admired that succession of chords”! At the time, Scott was well-known enough that a writer could make that reference and readers would “get” it.

Many of these composers have been forgotten, and in fact Cyril Scott himself was almost totally ignored for about the last 30 years of his life. (Happily, his music has been rediscovered since about 2000, and pieces that were never even performed have now been recorded.) A full examination of music from the early 1900s would involve learning about many composers who were very much emulated at the time, but are now obscure. (For another example, it appears that Bix Beiderbecke admired the composer Eastwood Lane, who is now unknown.)

How would Tatum have learned such pieces? You need to know that he was not blind—he was what we today might call “legally blind.” But he did learn from recordings, more than from sheet music. In fact, that same issue of Downbeat, October 1935, states that he was “in the process of teaching his wife, Ruby, to play the piano. He believes she will be an invaluable aid to him when she can read the chords to him, or more intricate compositions, and save him the difficulty with which he works now, with the assistance of powerful glasses.”

(This October 1935 issue of Downbeat contained the first two feature articles ever written about Tatum. Before that were only a few short record reviews, and listings of where he was playing. For paying subscribers, the relevant pages are below—with my thanks!)

Let’s move on. Here is Tatum playing “I’ve Got A Right to Sing the Blues,” pre-recorded for radio broadcast in Hollywood, probably in January 1940 (not NYC, August 1939 as usually listed; this is also from Lasker). It proceeds in an excellent but unremarkable manner until the very end, when Tatum breaks out into some totally unexpected chords, followed by a concluding passage of bitonality:

Daring stuff for a jazz artist of that time! This is more evidence—and I have plenty more—that Tatum’s “real self,” if you will, was considerably more radical than what people generally assume. I feel safe in assuming that in the privacy of his home, Tatum would have enjoyed experimenting at length with such juxtapositions of keys, taking his time with ideas that he only played for a few seconds on recordings. But you’ve only heard a few examples so far. There’s much, much more to come!

All the best,

Lewis

P.S. I posted a much shorter and very different version of this at WBGO.org in 2017 and that one was edited by Nate Chinen—thank you, Nate!