In the second of a series of exclusive blog posts, Brad Mehldau
continues his conversation with the Carnegie Hall audience, revealing
his thoughts on composition and improvisation.
Creativity in Beethoven and Coltrane
Part Two: Who Needs a Good Melody Anyway?
The three-period schema we mentioned earlier partially fails us with
Beethoven because his later music develops in two opposing directions.
There is the tendency to retract and economize that we find already in
Op. 95 and also in the very last string quartet, but there is also a
move towards even greater breadth, found in the gargantuan
"Hammerklavier" Piano Sonata No. 29 in B-flat Major, Op. 106. In this
piece, Beethoven reaches far back, past the high Classicism of his
immediate predecessors Haydn and Mozart, to Bach. He returns to the
fugue, a musical form that had reached an apotheosis in Bach's music. He
employs it, though, to create a new kind of music, expressionistic in
its rejection of traditional beauty. In the fugal last movement of the
"Hammerklavier," a trill takes center stage, forming an integral part of
the fugue's theme:
By placing a mere ornament in the forefront of this brash fugal
texture, Beethoven is thumbing his nose at the banality of commonplace
musical gestures and gets away with it because of the imaginative
richness of everything that surrounds the trill as the fugue progresses.
He also presents the performer with one of the most famously difficult
essays ever written for piano. Any vestige of Viennese Classicism is
smashed to bits here, and Beethoven won a victory over the past,
present, and future: Nothing like this had been written for piano
before, and nothing ever will. The "Hammerklavier" is that rare piece of
art that transcends all previous means of expression and then remains
standing alone, towering over all that follows it. The sonatas of Liszt,
Schumann, Chopin, and Brahms in subsequent decades as well as of
20th-century composers like Prokofiev and Barber would take many cues
from the "Hammerklavier," but none would achieve its dichotomy of
archaism and wild, unbridled expression.
The banality of the fugue is humorous, but there is a menacing
quality to that humor because of its extreme difficulty, which often
comes from executing that trill amidst all the figuration that surrounds
it. Pianists can hear Beethoven laughing in his grave when they work on
this piece, and Beethoven's victory over his creative mortality here
has a mocking edge that again brings to mind 20th-century examples:
There is a distinctively sarcastic quality to this music at times that
one encounters, similarly, in some of Prokofiev's scores—one thinks of
the opening movements of his Sixth and Seventh sonatas in particular.
The difficulty of Beethoven's score is Joycean—one thinks of the
anecdote that Joyce was heard at all hours of the night from his room as
he wrote Finnegan's Wake, cackling with wicked glee. I'll bet
Beethoven had a similar sadistic pleasure as he composed this movement.
We know from his correspondence that Beethoven was consciously setting
out to give pianists something to chew on for a long time.
This direct confrontation with the banal is not an arbitrary
occurrence in Beethoven's later period. It is subsequently codified into
method in another late piano work, the "Diabelli" Variations, Op. 120.
Beethoven was given a theme composed by the Viennese music publisher
Anton Diabelli and asked to compose a variation on it. It was to be
included in a collection of variations from various composers on the
same theme, among them Schubert and Czerny. Famously, Beethoven first
refused to participate, but then took up the theme with a vengeance,
composing a huge set of variations that he eventually published as the
separate work we know today.
Diabelli's theme, in waltz time, is at first glance a trifle. Again,
as in the "Hammerklavier" fugue, mere ornamentation is one of its most
distinguishing features; in this case, it is the distinguishing
feature. Here are the first eight bars:
There is no melodic activity in the right hand here apart from the
quick curve of the opening right hand pickup phrase, repeated in the
dominant at the pickup to bar 5. Does this even qualify as melody,
though? It looks and acts more like an ornament. The first note of the
piece is an appoggiatura, and taken as a whole, the phrase is similar to
the ornament that appears first in the Baroque era, called a "cadence"
because of the way it forms a cadence in its shape. A cadence-type
ornament looks like this:
and sounds like this:
The difference between Diabelli's theme and the Baroque cadence is
when each lands on the tonic pitch. Because the tonic takes place
directly before the downbeat in the cadential ornament above,
it has the potential function of creating harmonic movement by leading
us somewhere else on the following downbeat. A simple series of
cadential ornaments, written out as they sound, illustrates how they can
keep things moving forward:
In the Diabelli theme, though, the opening melodic gesture lands
squarely on the C tonic, on the downbeat—there is nowhere else to go.
The effect is bland and stifling in a comic way. The piece has just
begun, and the first thing it is telling us is: "I have nothing more to
say!" Since this melody will not bring us anywhere else, the only way to
the dominant is to raise the whole shape by a step, as takes place at
the pickup to bar 5. This is unimaginative and sounds banal because
there is no development of the idea, but maybe that's just the point.
Commentators have scorned Diabelli's theme since it first appeared,
but others have pointed to its appealing generic quality. It's very
possible that Diabelli intentionally made his theme generic so there
would be room to let the composers use their imagination in the
variations to tell stories with true development. Whatever the case, the
theme took hold of Beethoven, and he created a set of piano variations
that is rivaled in magnitude and imagination only by Bach's Goldberg
Variations and Brahms' "Handel" Variations.
A good theme is often referred to as a gift, something that is
given to a composer—from the creator, from one's muse, etc. The way that
theme is worked out in the course of the piece will involve the will
and intelligence of the composer, but the initial theme often just comes
to him or her in a moment of inspiration. But what if it doesn't?
Beethoven demonstrates in the "Diabelli" Variations that he doesn't need
this gift—he will find inspiration elsewhere, and he will even use the
uninspired quality of someone else's theme to his expressionistic
advantage. This is a victory over banality that is achieved through
banality. And in the story of Beethoven the creative musician, it is a
victory over his mortality. He has avoided repeating himself. He thus
retains his relevancy and avoids dying a creative death.
By engaging in the banality of Diabelli's theme directly and
creating so much from it, Beethoven's music is deconstructive, calling
into question the importance of a good theme. (In part, one could argue
that he is deconstructing his own earlier output as well: There is a
tendency towards banality that we find even in some of Beethoven's
greatest themes—it is part of their character.) If such a paltry theme
as Diabelli's can inspire such a rich bounty of material from Beethoven,
the music itself is asking its listener, what constitutes good music?
The answer is not something that we point to in the music, i.e., the
melody, the harmony, or the rhythm. It is something that we find in the
one who creates the music. It is the imaginative gift of the composer,
more than whatever pre-made material he works with, that keeps us coming
back for more.
Theme and Variations: the Dumbassed-Genius Cousin of Musical
Now replace "composer" in the previous sentence with "improviser".
Imagine that instead of Beethoven composing variations on Diabelli's
theme, it's Charlie Parker improvising a solo on George Gershwin's "I
Got Rhythm." Parker and other jazz musicians were probably not
particularly intrigued by Gershwin's trite melody to "I Got
Rhythm"—although it has its time and place and it has its charms. They
were attracted by the organizing formal and harmonic characteristics of
the song, though—the simple AABA form, the square 32-bar length (after
they jettisoned Gershwin's short coda), and the yin-yang balance of
tonality that the tune has: We stay right around the home tonic for the
first two A sections. At the B section, we travel to harmony that is
remote from the tonal center. Then we return once more to the A section.
Here is how jazz musicians think of "I Got Rhythm":
Jazz musicians call this type of chord schema "rhythm changes."
("Changes" means "chords" in jazz terminology.) It is one of the basic
improvisatory structures of jazz—a vessel waiting to be filled, like the
sonnet for the poet, the bank heist for the filmmaker, or the Bildungsroman
for the novelist.
Rhythm changes in themselves are banal as well: When a musician
plays those chords of Gershwin's above, they have a hackneyed feel—it is
a distinctly 1930 kind of banality that you might describe as
repetitiously, relentlessly enthusiastic: "Come on everybody, let's be happy!"
You picture a musical revue, guys with top hats, and lots of forced
smiles. A jazz musician playing rhythm changes must find his or her way
to not succumb to their inherent corniness. An innovator like Charlie
Parker provides the strongest model we have: meeting the harmony head-on
with a new approach that was revelatory for American music. We'll look
at how and why later on.
Jazz musicians use other harmonic progressions that are less
repetitive than "I Got Rhythm" for their improvisatory flights. "Lover
Come Back to Me," "Just You, Just Me," and other popular songs of the
day were vehicles for Charlie Parker, Thelonious Monk, and others.
Sometimes they played the original melody, or sometimes they wrote a new
melody, and then they improvised over the structure of the song,
looping it around. For the layman: This is what's going on in much of
jazz for the last 80 years or so. On the rhythm changes example above, a
jazz musician creates melodies that are related to the chords or
"changes" there, and when the band gets to the last measure, they simply
go back to the beginning. Yes, they are improvising, but they are
improvising over that repeated structure. Formally, this is exactly like
the theme-and-variations model that Beethoven used earlier.
If you take the various common forms in classical music—the
sonata-allegro form, the minuet and its younger sibling the scherzo, the
rondo, what have you—almost all of them have some inherent drama built
into their structure. The sonata-allegro form is the most interesting as
a canvas by itself without paint: its theme-development-recapitulation
shape gives an imaginative composer a strong narrative arc to work with.
The dance-based minuet is much simpler in design, but the mere
inclusion in the trio section of new thematic material, usually in a
different key or mode, already gives a composer the potential for
contrast. Contrast, antagonism, and tension, or their opposites, unity,
accord and resolution, are all fundamental aspects of musical
storytelling in the same way that they are in any other narrative
medium. In the era of high Classicism, sonata-allegro form in particular
inaugurated a new kind of musical storytelling in which the large-scale
tonal relationships—particularly the fundamental dichotomy of tonic and
dominant—were exploited as a narrative means to an end.
Putting a theme through its paces, working it out, transforming it,
and all the while moving away from and then back towards the tonic home
base: This largely German impulse changed musical expression forever.
Put briefly, the impulse was to create something grander, but with
formal integrity: a large structure, not just a large sprawling mass.
This meant, for a composer like Beethoven, an inherently organic
structure, in which the tension and resolution on a micro-level, felt
in a singular melodic gesture, corresponded to the larger tonal
relationships within a movement—or even within a whole multi-movement
In a theme and variations, the tension and resolution that we hear
the first time through the initial theme is all we have. Going back to
the "Diabelli" Variations, let's look at the full opening theme:
The music begins on the tonic and then moves towards the dominant,
where it decisively lands at bar 16. This material is repeated. After
the repeat sign, we are moving back towards the tonic, which decisively
arrives in the last bar. And when we repeat that second part once more,
that's it. This tonal harmonic scheme will never be developed more; it
will simply be repeated. All of the variations will adhere to it, until
Beethoven reaches the exalted ending fugue.
In telling any musical story, the musician—the improviser, the
composer, the singer-songwriter, etc.—works with a dichotomy of identity
and difference. Both are necessary. The identity of a work is
established through some sort of repetition: A theme is initially
announced, and then a particular aspect of the theme is heard again.
What we call development is really a mix of repetition and something
different. The great difference between theme and variations and the
sonata-allegro formal approach is that in theme and variations there is
strictly no structural development after the initial statement
of the theme. What follows is a series of repetitions. The variation
that ensues is a kind of development, but it is development from the
top-down, so to speak: Melodic variation, rhythmic variation, but all
within an established harmonic and formal structure, which is repeated
over and over.
The impulse to constantly repeat the opening thematic material is,
in itself, unintelligent and narratively shortsighted. One relinquishes
any real possibility of structural development. Theme and variations
take a proto-copy-paste approach, and in the literature of high
Classicism, they have a unique identity—I think of them as the dumbassed
cousin of the other more exalted forms. Really, though, this cousin is
more complex; he is more of an idiot-savant figure. For while he repeats
the structure of his story over and over again, he uses his rich
imaginative gift to fill it with something continuously new. Yes, the
structure stays fixed, but once we are free from the burden of
actual structural development, so many unusual and downright
strange things can happen.
How exactly is structural development a burden? It isn't simply a
burden: It is the burden. It is why Brahms took so long with a first
symphony; it is why Chopin never composed one; it is why Schumann's
sonatas are not as popular as his other piano music. The Romantics that
followed Beethoven expanded on his expressive innovation but could not
match the formal integrity of his large-scale structures. Their most
realized contributions were often miniatures, and that made sense: The
greatness of composers like Schumann or Chopin (and already Schubert in
some of his later music) came in their ability to express something
fragmentary and fractured, and to let it stay like that. There is a
tragedy to this kind of fractured expression—it is the great, beautiful
tragedy of 19th-century art music.
To Justify or Not to Justify
The "Diabelli" Variations are full of strangeness. Early on in the
third variation, we encounter this:
The initial three-note pickup at the beginning is a lilting,
feminine shape that Beethoven uses throughout the variation, developing
it further after the repeat sign through the use of imitation in
different registers. It acts as a springboard, leading to cascading
chordal movement in all directions. But at bar 20, it is as if there is a
skip in the record, and we can't move forward: We hear only that three
note segment in the bass register, repeated, looped around continuously.
The effect is both humorous and mysterious, with the pianissimo dynamic
Schubert may have had the variation in his head when he went to
compose his last, exalted piano sonata five years later. It begins like
The trill there in the left hand at bar eight is the same kind of
weird non sequitur Beethoven used in Variation III—it is something
unannounced, unprepared-for, but most importantly, unjustified.
There is no good reason why that trill appears, in the same way that
there is no reason that Beethoven's figure in Variation III decides to
repeat itself, grumbling in the lower register.
Schubert wrote music in the realm of the non sequitur in a way that
is unsettling and sometimes simply terrifying in his last years. Take
the unexpected harmonic left turn in his song, "Der Döppelganger." When
the speaker discovers that the figure he is viewing is none other than
himself, Schubert gives us a chord on the stalt syllable of Gestalt,
on the second-to-last bar below, that is twisted and bizarre, and the
song is suddenly in an alienated, unhinged world that sounds more like
Vienna of the 20th century:
It's just plain weird, and it's so great.
One also thinks of the heart-wrenching middle section of the Adagio
movement of Schubert's String Quintet in C Major, which jabs us in the
gut after the impossible tenderness that came before it. We do not see
it coming, and even though it affects us deeply, we don't really
understand why it arrived. The feeling is: "What happened?" Or, there is
the hellish nihilism in the Andantino movement of his second-to-last
piano sonata, No. 20 in A Major (D. 959). Again, it comes fiercely,
without warning, from nowhere. Schubert was a guy who had visions at
night when he went to bed—bad dreams, but dreams full of ecstasy as
well—and he managed to get some down on paper before he left the earth.
That's the only way I can explain his music to myself when I hear it.
For me, there is no other music that is so dreamily beautiful and
terrifying at the same time.
Schubert found a way to "take it out," as jazz musicians would say,
and he may have taken a cue from Beethoven, or he may not have—we know
that he admired Beethoven. If he did, it is ironic, because with
Beethoven, even what looks to be a non sequitur is in fact justified if
we dig deeper: He is almost never just "taking it out" willy-nilly.
We'll look at that phenomenon later in his Ninth Symphony. With
Schubert, though, these outbursts in his music are purely chaotic, even
when you unpack them: They don't come from anywhere, and they don't lead
anywhere else. They're like seizures that pass.
Beethoven's way out in this case is through the
theme-and-variations form itself: Once the template is established in
the opening theme, he allows himself all sorts of utterances, like the
grumbling left hand of variation 3 above. The reason why one idea
follows another is not so important—there might not be a discernable
reason. The ideas may flow in a more stream-of-consciousness manner,
like a person talking continuously, saying what comes to him right at
This characteristic applies to the improvised jazz solo and gives
much of jazz its particular character. I would emphasize that the
stream-of-consciousness description does not denote
formlessness or arbitrariness in terms of the individual expression of
the soloist. There may well be a strong narrative line in the solo, and
the best jazz soloists are storytellers. The difference, though, between
this kind of theme-and-variations approach and the other more
"justified" style of composition that we find in sonata-allegro works,
is the viewpoint of the storyteller: In theme-and-variations approach,
the composer or jazz soloist is looking ahead constantly; whereas in the
sonata-allegro approach, the composer is constantly looking back,
seeing what he or she just did, and then building on that. The present
must always be justified by the past.
We could think of the two approaches in terms of building a house.
In theme and variations or jazz soloing, the frame of the house has
already been built, and the composer/soloist can decorate all he wants,
without worry that the structure will collapse. However (a big however):
He cannot stray from the frame of the house; he must work within its
borders. In the sonata-allegro approach, the composer must start from
scratch and build the house from the ground up - this is a lot more
work, and requires more diligence. He can't just build willy-nilly; he
has to make a house that will not just look good, but will also stay
standing. If he can pull that off, though, he has the reward of
building whatever kind of house he wants; he may follow his own course
at all times. There is no pre-existing template that he is
obliged to follow.
Theme and variations make a concession—they relinquish the autonomous
act of building that house from the ground up, and follow that
pre-existing template. They make this sacrifice in the service of a more
immediate kind of expression. The urge to make variations is a
pragmatic urge—the urge for a quicker means to an end; the desire for a
template that one can open and start filling with the creative,
chattering stream of material rustling in his or her head; a template
that will then quickly organize that chatter. The necessary dichotomous
identity of musical expression is then quickly established—the fixed
identity of the thematic structure announces itself to the listener
repeatedly, giving him or her a continual reference point, and the drama
and flux of difference and variety play out within that structure.
Which approach is deeper? Let's look at Beethovenian
justification in his very first piano sonata. Here is the beginning:
Everything comes from a two-part idea in the opening two measures,
bracketed above as Part A and Part B. They are yin and yang to each
other: Part A is more harmonic in nature, tracing the F-minor tonic
triad, while Part B is more melodic, moving stepwise. Part A is
ascending; Part B is descending. Part A is staccato; Part B is slurred.
The note values in Part A are uniform and square; those of Part B are
varied. The yin and yang will act as a springboard in two directions
throughout the movement, and the distinction between the two parts will
propel the musical narrative forward, as they differentiate themselves
from each other.
Everything comes out of what preceded it, as noted in the brackets
above the musical gestures. The first eight measures are a masterful
example of building tension: The initial idea is immediately recast in
the dominant harmony in bars 3 and 4. In bars 5 and 6, the first four
bars are shoved into half the space, which gives us a feeling of
insistence and tension. Repetition is tension for Beethoven in this
setting, and difference comes as a release of that tension. At bar 7, we
have a climax, and that release takes place in the descending scalewise
motion in the right hand after the rolled chord. Only here, for the
first time, do we hear distinct difference. Those four notes are eighth
notes, and that's a big deal: Until this point, absolutely no notes with
that metric value have shown their face. When Beethoven introduces them
here, they have a dual effect. With the hairpin diminuendo, they feel
like a retraction or a retreat. At the same time, we hear that those
notes are a slowed-down variation of Part B's triplets, expressing
themselves emphatically one more time. This simultaneous pulling-back
and insisting makes a wonderful moment, full of emotional ambiguity.
There are at least two ways for a pianist to play bars 7 and 8: more
insistent, more hesitant, or somewhere in between that reflects both of
those sentiments. Beginnings like these are what make Beethoven the
heavyweight champion of justification. Justification in his hands is
never obvious or easy—it is urgent; it is questioning and self-critical;
it is filled with import.
There is repetition and difference, repetition and difference.
Repetition builds tension here, and difference releases it, but in theme
and variations, the opposite is true: repetition is the norm and
represents stasis, while the constant variation superimposed on that
repetition provides the tension that keeps the listener occupied. These
two types of expression are fundamentally different. The one allows what
the other relinquishes. In the sonata-allegro form of the example
immediately above, the unfolding ideas follow no pre-existing grid, but
must continually justify themselves by what preceded them. In theme and
variations, those ideas may unfold as they please, but must always
adhere to the grid.
It's apples and oranges, but we could make one last observation
about the difference before leaving the topic. In the bulk of canonical
three- or four-movement classical works from the later part of the 18th
century through the 19th, and still in the 20th—such as sonatas for an
individual instrument or an instrument with accompaniment, chamber
music, and symphonies—the sonata-allegro movement will be first, and the
theme and variations movement will come after it, often as a finale,
but also often in a middle slow movement. This could have become the
norm for many reasons, but particularly after Beethoven, there is a
sense that a composer must prove himself compositionally worthy in that
first movement, and the way to do that was through the sonata-allegro
It's simply easier in many ways to compose theme and variations. I
don't have the data to back it up, but I'm sure that the actual amount
of time spent composing a theme-and-variations movement for all those
composers was less than the time it took to write a sonata-allegro
movement of equal length. That ease is why the approach benefits jazz
musicians because—here I might ruffle some feathers—jazz is not
fundamentally a composer's music. It's a form of music with some great
composers, and a lot of great improvisers. Don't get me wrong: I don't
mean to suggest that jazz musicians didn't write sonata-allegro pieces
because they couldn't, so they copped out and blew over repeated
thematic material. I mean that they had other designs.
The chief reason why theme and variations interest me here is
because they are the most significant formal device that jazz music took
from western classical music. This is not a coincidence. It speaks to
the pragmatic urge in jazz—the urge to accommodate the improviser. The
improviser is the protagonist, and most often he is decorating the
existing house versus building a new one from scratch. We watch him or
her with wonder, or we watch him or her with boredom. If you think about
a lot of modern jazz, you realize that, collectively, jazz musicians
threw all their eggs in one basket formally speaking, by submitting to
the theme and variations approach.
I am bracketing out all of the great compositional contributions
from people like Duke Ellington not because it is of lesser worth or
importance in my view, but because when we talk about the be-bop
revolution in small-group jazz, led by Charlie Parker, and then
everything that followed it—Miles Davis's groups in the 50s, John
Coltrane's group in the 60s, and many other great ensembles and
soloists, right up to the present day, when you walk into a jazz jam
session just about anywhere on the planet—we're talking about a
theme-and-variations approach, or, as a jazz musician would have it,
"head-solo-head out" (the "head" for jazz musicians is the theme). Jazz
musicians everywhere are all still trying to be little Beethovens:
They're trying to make their improvised variations imaginative and
If you think about that, it's really a ludicrous project: How is
someone going to arrange his or her notes in a more compelling way than
all the ways we've already heard—especially on older-than-dirt
structures like rhythm changes and blues? The reason why a lot of people
complain that jazz is boring is because, truthfully, a lot of it is.
And most of the time, the reason that it's boring is because that
soloist and the band he or she interacts with are not arranging their
notes in a fresh way. This may be an obvious point, but I raise it to
emphasize the nature of the creativity in jazz: It is not necessarily
expressed in the composition at all. Often the composition is just a
means to an improvisatory end. So much lays on the improvisation, on the
personalized variations of the material that the group comes up with in
the heat of the moment. If those variations aren't inspired, then the
banality of the composition—rhythm changes, for example—will be all
that's left. Then you want to race for the door.
This is not meant as a mystification of what jazz musicians do, nor
is it meant as a diminishment of what they have achieved. To me, it
speaks more to the character of their collective achievement, and now I
go back to this idea of avoiding a creative death: When a jazz musician
blows an uninspired solo, he or she dies right in front of us—and we die
with them, of boredom. When a jazz musician blows a great solo, though,
he or she avoids dying a creative death in real time, right in front of
us, for all to see. It's a victory and it's a thrill—it's the jazz
sublime. Later, we'll consider its nature more closely.