House On Hill

I wrote the music on this record for the trio that I led for roughly a decade, from 1994 to 2004, with Larry Grenadier and Jorge Rossy. Seven of the tracks come from a two-day session in 2002, where we recorded eighteen songs. We decided to split that material into originals of mine and interpretations of existing songs, and the latter group became Anything Goes, released in 2004. The originals are presented here with two songs from a more recent recording session, “August Ending” and “Fear and Trembling.” All of the songs were written between 2000 and 2002, and they form a time capsule of my writing then, and of the way the trio was playing together. To the extent that the music was conceived specifically for the three of us, the writing and playing are tied into each other.

Since the record is all original material, I thought I’d share my personal experience with jazz composition. The advantages of the idiom are the same things that make it problematic for me as a composer. The very condition that allows for expressivity implies its own limitation. The successful integration of composed and improvised material has always been a challenge for me. It warrants a discussion of form, or more specifically, the dialectic between the fixed form of the composed music and the (ideally) unfixed content of the improvised music.

In a strictly formal sense, the music on this record sits well within the “theme and variations” model that has long been the dominant approach of small ensembles in jazz. Defined succinctly, it goes as follows: The theme is stated first, often referred to as the “head.” Improvisation follows, often in the form of a solo, using the opening thematic material, whose structure is repeated indefinitely—like in a classical theme and variations setting—until the soloing is concluded. Generally, the theme is reprised at the end of the composition. Period.

That is an admittedly provisional definition of what takes place very often in jazz, but the approach itself has a provisional logic. In one sense, the beginning thematic material has a merely temporary role because it doesn’t develop further. The improvisation that ensues will usually constitute the bulk of the performance. On the other hand, the improvisation is bound to the initial thematic structure for its duration, repeating it over and over again. Within the bounds of functional harmony, two or more people cannot improvise the large-scale development of a theme, like we find in the exposition of a classical sonata’s allegro movement, because they cannot read each other’s mind. The bass player cannot know, for example, that the piano player wishes to modulate to another key; the piano player cannot know that the bass player wishes to stay a bit longer in the original key.

So, the theme and variations approach allows the soloist to correspond with the rest of the band. The efficacy of the approach lies in its expediency: It provides a quick and clear way for the soloist to improvise with a high degree of spontaneity. The harmonic material that underpins the solo, though, with its tension and resolution, will provide a narrative backdrop—a place of origin and a destination. A balance is reached between something fixed and something open-ended. For me, this technique has always been rewarding as an improviser, but can be confining as a composer.

It strongly implies a specific relationship between harmony and melody. In much of tonal music, the union of those two elements is an ideal. Harmony—simply two or more tones sounding in unison—is not so much an end in itself; rather, it is the outcome of two or more melodies taking place simultaneously. Melody has primacy always, and the relationship between two or more melodies creates harmony, which is secondary. In a Bach fugue, we can see this clearly:

Fig. 1 J.S. Bach, The Art of the Fugue, Contrapunctus 1, excerpt


Looking at the score horizontally—examining the music of each part as it moves across the page—we can see how the bass part, in gray, has a dual role. It harmonically underpins all four voices, providing the foundation that we associate with the bass, yet it also stands alone as a melody. Furthermore, the phrase of the bass voice is in fact the opening theme of the fugue.

The entire piece is made from the same stuff. To the extent that each part constantly reasserts the identity of the whole, the musical content, in the moment of utterance, immediately fulfills a formal role as well. Every note is ripe with implication; each voice has a multidimensional character, achieving several functions at once—melodic, harmonic, and formal. The texture of the fugue is appealingly plastic. With its four simultaneous melodies, the piece is constantly in motion. No voice is ever delegated to a mere static accompanying role. Yet within all that flux, a stringent formal economy is never forfeited.

When music became less contrapuntal, the model of polyphony that reached its apex in the music of Bach came to be seen by some in prelapsarian terms, as an idealized state of grace from which composition had fallen. The viewpoint still persists. Glenn Gould, for instance, championed Bach’s keyboard works, but did not hide his disregard for whole chunks of the classical piano canon, especially composers like Mozart or Chopin, whose piano music often divided melody and harmony into a single melodic line and a chordal accompaniment. This division may contribute to the pejorative “parlor music” tag that is sometimes attached to Chopin’s piano works. The assignation of melody to the right hand and harmony to the left hand that we find in a big portion of his piano music gives it a certain stylistic homogeneity. Mainstream jazz piano playing has for the most part followed this model of melody and chordal accompaniment in the right hand and left hand respectively. Of course, this division is not in itself a bad thing, but to the extent that it becomes a fixed stylistic procedure, it at least implies an expressive limitation.

This passage of Brahms is less overtly polyphonic than the Bach:

Fig. 2 Johannes Brahms, Clarinet Quintet in B Minor, Opus 115, 2nd Movement, excerpt


We hear the clarinet soaring above the other instruments; the division of melody and underlying harmony seems pronounced. Nevertheless, each voice below the clarinet has melodic integrity. The second violin and viola, both in gray, both use a three-beat rhythmic motif that ties together eighth-note triplets and eighth notes. This figure is played in tandem, canon-like.

How do we hear this music? The second violin and viola make a strong case to be heard as distinctive voices. But the rhythmic motif they both use, with its alternating tied triplets and eighth notes, is destabilizing, even more so because the two voices are set in overlapping tandem. We begin to hear them as an impressionistic blur of harmony. Brahms hides the melodies of the violin and viola; they are like an undercurrent in a stream that is not visible from the surface. It’s a kind of stealth polyphony.

One reason that Brahms is such a model for me is the way he straddles two epochs. He was a master of counterpoint, with its strict rules, yet his music expresses ardent, immediate emotion that we associate with the free flights of romanticism. The Sturm und Drang in his music is tempered by the rigor of its structure. He is fully a child of his time, yet reached back to an earlier epoch for inspiration. Bach’s music was the apotheosis of that epoch.

The expression of self-conscious irony in the music of Beethoven, Brahms’s more immediate predecessor, heralded more than just the Romantic period of classical music that ensued. It was nothing less than a musical response to the onset of modernity itself. Irony, taken here simply as the condition that arises when an object is placed in an opposing context, has no place in Bach’s great fugues, where there is never any question of context in each single voice. Each voice in the fugue can stand alone, yet is always in the service of the greater whole. To follow Hegel, the particular is subsumed in the universal. By the time Brahms came along, this relationship had already been fractured. The rupture had taken place at least two musical generations earlier. One could argue that it began before Romanticism in the High Classicism of a composer like Haydn. If Haydn’s greatest works epitomized symmetry and order, they often contained ironic comments on that order, in the form of musical practical jokes.

Brahms reached back one generation further past this worldly conceit of Classicism, to the music of Bach. Bach’s music did not pale in comparison to his predecessors because it lacked an ironic stance; on the contrary, he is like Shakespeare in that one can listen to his music and say, this whole genre could have stopped here and would seem complete; nothing need follow. Some of Brahms’s critics saw the way in which he incorporated Baroque gestures—sometimes more overtly, as in the final passacaglia movement of his last symphony, or the fugal sections of his German Requiem—as stylistic backpedaling.

Yet it gives Brahms’s music its multidimensional character: The Clarinet Quintet is one of his most unapologetically romantic works, rife with soaring melodies, gypsy-spirited cadenzas from the clarinet, and almost Wagnerian harmony at times. At the same time, in its formal coherency, compositional economy, and attention to contrapuntal detail, it speaks to us with an arresting austerity. Temperance, in this quintet as in so much of Brahms’s music, is still seen as a virtue. Amidst the emotional abandon of the musical content, its very structure points to an order that is immutable. The structure doesn’t take away from the emotional impact; it boosts it. The musical representation of structure and order is necessarily viewed from a distance, the distance of time between Brahms’s own epoch and Bach’s.

The felt distance gives Brahms’s music a two-tiered aspect. Often, the emotional effect—for me at least—is a cathartic feeling of tragedy and grace: the tragic impossibility of meeting with that non-ironic orderliness again, as a condition of modernity; and grace in the evidence that order nevertheless prevails even if it has distanced itself from us, as a salve for the meaninglessness that surrounds us. The teleological argument put forth in Bach’s music in his own time was a given: it posited an ultimate order to everything and answered with its own order; it was a manifestation of God’s perfection and was easily understood as such in Bach’s own time, especially in the church where the music was often presented. Brahms’s music, written around the time that Nietzsche pronounced the Death of God, was no mere nostalgic Rückblick; it was an act of defiance, a bold gambit in which he raised the stakes for himself. With all its romantic outpouring, his music still conveys its own kind of teleology in its very architecture.

If Brahms is a hero of mine because he straddles different epochs, how could one straddle two different epochs in the context of jazz composition and improvisation? What could a jazz musician, for example, take from someone like Brahms?

If we go back to that same passage of the Clarinet Quintet, it is also instructive to look at it vertically—to look at the harmony more as chords that sit under the melody—here reduced to a piano score:

Fig. 3 Piano reduction of fig. 2


The chord symbols above the staff are the kinds that are used in jazz. They are all a soloist would need to improvise on the music here. Or are they? In this piano reduction, the rich inner voice movement of the viola and second violin—Brahms’s stealth polyphony—has been swabbed away, and what’s left are the tones that act as “pivot points” within the figuration. These notes are pivots because of their strong triadic harmonic implication. With the exception of indicating a mode, which is less common, jazz chord symbols generally designate chords that begin as triads and then expand from there; thus the necessity of a reduction.

The primary limitation of this triad-based system of harmonic nomenclature is that it does not effectively account for inner voice movement—for all that great stuff between each chord. In figure three, the beautiful figuration in the inner voices is gone, and with it, much of the character of the composition. For example, we’ve forfeited the chromatic descent of the viola as it dips down to an F natural on its second note, seen in the original score in figure two. That mere stepwise motion in the first moments of the movement adds piquancy and color to the first two bars; take it away and we’re left with a rather pedestrian series of plagal cadences that sound more like a Lutheran hymn. Would we not want to go back and make a more detailed account of the passage in its original form, one that addresses the incompleteness of the chord symbols in the above reduction? It would offer the improviser more material to solo on, and it would be more accurate.

It would not make for a richer improvising experience, though; in fact, it would hinder a jazz soloist, and this is the rub: The more specified the inner voice movement becomes, the less room the soloist has to truly improvise. With each chord symbol, his actions become more dictated, his musical choices become more spelled out, and his freedom to make spontaneous musical decisions is increasingly usurped. As a hypothetical example, we need only illustrate the first bar of the Brahms in unreduced form, with chord symbols that attempt to address the melodic voice-leading in the score. We quickly see why it doesn’t work:

Fig. 4 Jazz notation of fig. 2, m. 1


It looks like a nightmare on the page—I’ve added arrows to show which symbol corresponds with which chord—and it is not possible to address all that information in real time. Furthermore, consider the harmonic implication of the sixth chord symbol, and the accompanying sounding notes that form its chord, both highlighted. This symbol for the E diminished triad, suspended over the cello’s tonic pedal point on the B, does not account for the clarinet’s melody note of D sharp that so beautifully suspends the tonic major mode over the other voices. The three middle voices do convey an E diminished triad, yet there is also an E minor chord when we pair two of them with the low B of the cello. There are several different ways to spell this with our chord symbols, yet none of them will be completely accurate, and all of them will be extraneous, because the music at this moment supersedes this kind of strictly triadic design.

As a jazz composer then, when we make a “chart” or “lead sheet,” we try to not fill up the page with those symbols in such a way that the chart is too dense, if it is to remain viable as a vehicle for improvisation. That is all well and good for the soloist, but by implication, it restricts the composer, most notably by discouraging the kind of melodic inner voice movement we see in the Brahms. The upshot of that is a form of composition with several potential limitations.

In general terms, the ideally open-ended quality of this kind of harmonic indication used in jazz paradoxically contributes to a certain stylistic homogeneity: namely, we hear the harmony that works most effectively with these chord symbols. Secondly, another phenomenon in jazz is the existence of a homogenized set of voicings that musicians playing chordal instruments use behind the soloist, or during their own solo. Finally, the necessarily vertical nature of the harmony in this system has led, through the years, to a whole school of piano playing that does not address voice leading—the melodic implication of each chord tone, from one to the next. Again, I don’t wish to say that these observations point to a poverty in the music itself per se, but to a tendency inherent in the method of expression.

In this appraisal, then, a preexisting fugal texture, as seen in the Bach, which prizes the simultaneous melodic activity of all the voices, offers the antithesis of harmonic freedom for an improviser to the extent that everything is already mapped out. It is redundant and ineffective to write out jazz chord symbols for such music. This is why classical music often does not lend itself to jazz improvisation. Trying to use a preexisting contrapuntal format as a vehicle for improvisation is like serving a steak with a big scoop of ice cream on top of it; it is adding to something that is already effectively complete.

Jazz improvisation that involves harmony, then, often favors a chordal texture, and the chords that are provided have a shorthand nature. The harmony is thought of vertically: In the jazz vernacular, a soloist is said to be “blowing over” or “on” the chords, and it’s a useful visual metaphor in seeing how the melodic content of the solo sits apart from those chords, on its own platform. The more shorthand those chord symbols are, the more freedom the soloist and the accompanying rhythm section have to fill them out and collectively improvise together. Songs that everyone knows and that form a loose canon of readily available vehicles for improvisation in a variety of contexts become standards because they lend themselves to this format most readily. They are, more often than not, simple in design.

To say that a jazz composition is simple is by no means pejorative; on the contrary, simplicity is treasured, but it is simplicity of a specific nature: the simplicity of the material that is used for improvisation. Write all the idiosyncratic melody you want for your initial melody, but if the harmony that lies under it is simple enough, you’ll never alienate the other people who play your tune when it comes time to improvise.

One of the greatest examples of this phenomenon, and one of my compositional heroes in jazz music, is Thelonious Monk. The melody on Monk’s “Evidence,” for example, is a profound study in rhythmic displacement. Here are the first eight bars:

Fig. 5 Thelonious Monk, “Evidence,” excerpt


Indeed, we almost need another name than melody here; it is more like the remaining broken shards of a melody that once existed. They are sparse, and land in what seem to be random, wrong-sounding places within the rhythmic meter. But after hearing the melody a few times, we realize that the melody has its own internal logic: it’s as if it’s set to a different meter than that of the existing 4/4, and the notes fall across the bar line on a separate tier that is nevertheless conjoined with the actual meter, albeit in an extremely syncopated manner.

Monk wrote some of the most ingeniously idiosyncratic melodies that were ever written in any genre of music, and within the jazz canon, they are also some of the more challenging ones to execute convincingly. “Evidence” is a particularly tricky head in terms of its rhythm; another example is “Trinkle Trinkle,” which might be the most supremely difficult essay of Monk’s. Yet his melodies are often grounded in simple harmony. (This is not always the case, of course, but Monk the harmonist, on tunes like “Pannonica” or “Monk’s Mood,” is a whole other subject.) The harmony of “Evidence,” for instance, is derived from “Just You, Just Me,” a popular song with facile harmony, written by Jesse Greer in the 1930s. Although the melody is unique to Monk, once we get to the improvisation, the chords that are blown over are well within the normative range of jazz—any musician with an understanding of functional harmony will be able to play something that corresponds with them; whether or not it sounds good does not rest on the chords themselves.

So we’re on common ground once the improvisation starts. Or are we? Let’s say that “Evidence” is played at a jam session, for instance. After everyone gets through the tricky head and breathes a collective sigh of relief, the trumpet player takes the first solo and does a Miles Davis-circa-1958 approach. The tenor saxophonist follows with a Wayne Shorter-circa-1963 thing. The pianist follows with a mix of Herbie Hancock and McCoy Tyner. It all works, but Monk’s music brings up a common dilemma in jazz: How much should the improviser address the tune? Is it okay to just get through the melody and then start playing your own grab bag of licks once the soloing starts? That approach may work well enough on a jazz standard like “Autumn Leaves,” but a Monk tune seems to ask more of the soloist, because what has just taken place on the head is so striking and full of meaning.

Here we’ve come back to the question of irony again, played out in a typical problematic scenario in jazz. The various styles that the soloists call on in that hypothetical Monk situation seem to step outside of the context of the composition. The question is whether this might constitute an aesthetic flaw. It would be troubling if it did, because you could argue that there is an ironic aspect to the whole phenomenon of normative jazz improvisation: When a player blows on a tune, he or she is commenting on that tune, removing himself from the original object, and, to varying extents, looking at it from a distance.

To me, the success of an improvisation depends not on how well it fits into the given context of the composition—after all, we don’t necessarily want to hear someone try to imitate Monk when playing a Monk tune. The success depends more on how much the solo transcends the context of the tune, making us forget about the whole question of context in the first place. That will rest ultimately on the fantasy and originality of the individual soloist. The aesthetic poverty of many jam sessions rests on a kind of weak irony: The players are completely out of context with each other, each playing his or her own bag, but that’s not by design; it’s just because they’re all playing what they already know, and what they know comes from an arbitrary variety of musical sources. That shouldn’t suggest that one should address a given style in jazz in one’s solos—that you should play like Johnny Hodges if you’re an alto saxophonist playing a Duke Ellington composition. To hold such a strategy up as a rule is to essentially give up improvising.

So the way to escape the problem of context is to create your own context. But how? It’s hard to be original as a composer in jazz; maybe it’s even harder to be an original soloist. It’s instructive to look at the way Monk fused his writing and his improvising together. The content of his melodies became fodder for his own solos; his solo vocabulary was not derived from fundamentally different stuff than that of his compositions. This might seem obvious and not worthy of mention; after all, riffing on the melody is one way in which jazz improvisation began to flower in its early stages. Monk was onto something else, though, and it involves the actual development of themes during his solo. By development, I mean that the musical content unfolds with a narrative logic; each idea springs from the previous one. This is already taking place in the initial written melody before Monk even solos, like in the first eight bars of the melody in “I Mean You”:

Fig. 6 Thelonious Monk, “I Mean You,” excerpt


The last three notes of the opening phrase are immediately developed before Monk goes any further. They act like a tail that has been cut off and changes its shape. In bar three, the order of the three notes is changed, and one interval is diminished. In bar four, continuing the thread of chromatic alteration, the whole “tail” shifts upward a half step, the rhythm is changed, and a note is added. Whenever something is added, it always expands directly from what has preceded it. In fact, this three-note motif that seems to spring from the last part of the opening phrase already has an earlier origin: the very first three notes of the tune—F, D, and C—respectively, from which so much intervallic material is derived throughout the course of the melody.

The way in which this organic development continues during Monk's solo suggests that when a song has a deeply embedded architecture like that of “I Mean You,” it will lend itself to formally richer solos, but only if the soloist is aware of the architecture and wishes to comment on it. When the composer is also the improviser, as is the case here, it’s a done deal. In this way, Monk has created his own context and there is no discrepancy between the composition and the improvisation. I make a lot out of this because Monk pointed a way for me through the challenge I mentioned at the beginning of these notes—the integration of composed and improvised material, as a jazz musician playing your own tunes. I take my cue from his method throughout this record to varying degrees, by incorporating the thematic material of my tunes into my solos in a variety of ways. Monk, like Bach in his time, managed to break through that dialectic of musical form and content. The two become one fused entity: The musical content of the initial melody becomes formal when it is used throughout the duration of the solo. It is no mere performative utterance; it is more architectural in nature. Monk set the bar for an approach to improvisation in which form itself becomes an expressive means.

The most immediate method of resolving the vagaries of context that I’ve discussed above is to have your own band, and I’d like to close with a few comments about that. Playing with Larry Grenadier and Jorge Rossy gave me a strong yet malleable context in which to write this music, and was an incredibly rewarding experience. I imagined them playing the music here as I was writing it, and the character of this music is determined to a large extent by their approach. For instance, “August Ending” or “Backyard” were viable for me because of a way that Jorge had, completely unique to him, of playing on them. On these tunes, I state the harmony under my melody as a constant stream of eighth notes, instead of the more typical approach of using punctuated chords. This thick texture of the piano part runs the risk of usurping the role of the drums. It is traditionally the drums that will provide a constant stream of rhythm within a band, yet there I am already playing eighth notes all the time. Jorge found a way of merging with what I was doing that gave these kind of tunes their particular shape and buoyancy. His drumming gave a flow to the music and kept the piano part from sounding too much like an opaque blanket that covered the whole sound by locking it into a fixed schema, and the looseness of his feel kept the whole thing from sounding fusion-like in a bad way. Jorge always managed to play free from a fixed repeated pattern, yet still with a deeply felt sense of the architecture of whatever tune we were playing.

This relates to one of the big challenges in jazz composition for me that I mentioned earlier: to not write too much, or not write in such a way that there is no longer room to improvise fluently. This became less of an issue from playing with Larry and Jorge for a long period of time: They both found ways to express themselves fully in the music, in part because the music was written with them in mind, and this allowed me to write more than I would in a comparatively generic jazz format. Another example of this “over-writing” is in the piano part of “Boomer.” The head of the tune was written out fairly explicitly; here are the first few bars:

Fig. 7 Brad Mehldau, “Boomer,” excerpt


This kind of figuration in the left hand, that also appears at the ending coda section of “Waiting for Eden,” is my attempt at that stealth polyphony of Brahms in a jazz improvisational context: it supplies the harmonic information that underpins the melodic content of the right hand, yet has a melodic flow in the stepwise movement of the sixteenth notes. As we played this tune and it developed in performance, I opted to keep the written left-hand figure as part of the solo section for roughly the first six or seven bars, blowing over it, before moving to a more chordal approach in the left hand for the remainder of the chorus; that’s the approach we take here. Throughout the record, there are compromises of this sort, where a provisional balance between the written material and the improvised sections is reached. It calls for a different strategy on each tune—on “Bealtine” or “Fear and Trembling,” for example, there is less actual written material than in “August Ending” or “Boomer”; consequently, the approach that the three of us collectively take is less idiosyncratic. The more that is written out, as a rule, the more idiosyncratic the approach becomes.

In these notes, I’ve tried to demonstrate some of the currents of thought that shaped the music here, informed by jazz heroes of mine like Thelonious Monk, Western classical music, and, of course, the most immediately felt presence of Larry and Jorge as we played together for several years. I hope that the listener enjoys this record as a time capsule of sorts. In my view, it represents a point that I reached with regard to composition, where several streams of influence coalesced into a broadly identifiable style that I would cautiously call my own. Finally, it also represents for me the apex of what Larry, Jorge, and I achieved together as a band.

– Brad Mehldau, March 2006