In current jazz practice, performance and theory are closely intertwined. This is good news for the music theorist who teaches jazz students at a conservatory. Jazz students never fail to see the point of theory subjects; they are abundantly aware that active mastery and knowledge of subjects that traditionally belong to the domain of music theory are necessary for them to perform and create music. Ear training is evidently the first and central subject. This is arguably true for every student at a conservatory, but in jazz performance practice, it is virtually impossible to function without a well-developed and “pro-active” ear. Instant recognition of harmonic patterns and chord formations that are sometimes rather complex is a necessary skill if one is to hold one’s ground on the bandstand. Furthermore, auditive skills are important for studying recorded performances.

A close second is the theory of harmony. Jazz performance often requires performers to realize the harmony themselves. In performance, the chord progression is usually predetermined, represented by chord symbols. But exactly how the progressions are played, the actual (foreground) details of the harmony, is negotiated by the performers. This holds true for the vertical distribution of the notes (the voicings) and the voice leading but also includes the substitution, discarding, or addition of chords. Furthermore, “harmony” and “harmonic theory” are broadly understood. In other words, many elements that are not traditionally seen as such are conceived as belonging to the realm of harmony. Melody, for example, is frequently considered in terms of which scales to play in conjunction with which chords.[1] All performers must deal with harmony, most notably, of course, in the improvisations.

Developing a vocabulary for improvisation is central to jazz education. There exists a wealth of pedagogical materials that focus on phrase and chord patterns and chord scales, such as etude-like books or play-alongs. However, this may result in a narrowing of scope. Studying patterns and similar material may diminish the perception of their function in a specific, larger musical context. It may come at the cost of an awareness of moment-transcending events or of matters of which the immediate practical application is less obvious. Issues such as the varying importance and function of harmonies within progressions, the role of a specific (melodic) passage in the complete course of events, the development of melodic ideas, or the relationship between the improvised material and the composition itself, just to name a few, often do not receive proper attention.[2] Therefore, it is one of the most important tasks of jazz pedagogues, certainly at conservatories, to increase awareness of the full breadth of musical phenomena.

At the jazz department of my own institution, the Conservatorium van Amsterdam, improvisation is, first and foremost, taught in the individual “main subject” (Hauptfach) lesson and, furthermore, in ensembles or specialized courses and electives. The theory program is practice-oriented and evidently also addresses improvisation but is not primarily focused on it.[3] A weekly three-hour class forms the core of the theory program (one hour equaling fifty minutes). This class consists of ear training, harmony, analysis, and basic arranging.[4] Although it is not possible for all topics, the aim is to integrate the subject matter as much as possible. For example, ear training “dictations” are usually based on recordings, which subsequently can be analyzed or used to introduce new theoretical topics. Furthermore, analysis can provide opportunities for an integrated approach. For instance, we can engage those aspects of harmony that the Roman numerals, usually employed for harmonic analysis in jazz, do not cover. Moreover, under the umbrella of analysis, we can study in context what in mainstream jazz pedagogy will so often be broken down into smaller segments. We can address the particular, rather than the general, and in so doing, deepen the aural understanding of music.

In this article, I present an example of such a class. Central is Clifford Brown’s improvisation on September Song, a composition by Kurt Weill from the 1938 Broadway musical Knickerbocker Holiday. The performance appears on the 1954 recording Sarah Vaughan.[5] Given the central artistic position of solo improvisations, primarily in small group performances, taking a solo as a point of departure for analysis in the theory classroom is an obvious choice. The analysis of this specific solo provides an opportunity to discuss a host of issues that are important but easily fall between the cracks and to introduce specific theoretical concepts. The class also involves preliminary analysis and transcription work, which already provides numerous interesting topics.

A “live” analysis class has a specific dynamic; for the sake of readability, however, I will cast this paper mostly as a written analysis, interspersed with commentary regarding the pedagogy. Depending on the available time, the class can consist of one or more lessons. Some tasks may be added or dropped, and the order of the topics may vary, allowing the class to be organized in various ways. The connection to the practice will always be a topic, and mainly takes the form of reconstructing considerations that may have formed the basis for the choices made on the recording. While such questions can be problematic in some contexts, they are part and parcel of the practical study of jazz.[6]

This article focuses on how this type of class can complement and support the study of harmony and melody. This is not to say that other very important aspects of jazz performance, such as timing and articulation, could not be addressed in the context of this specific solo. They most certainly could, and I will, on occasion, incorporate some discussion of these matters.[7]

The Preliminary Work

Prior to anything else, we must prepare the material on which we base the analysis. First, we need to establish the more or less “original” melody and harmonization of the composition. When an existing piece, such as a standard, is the basis of an improvisation, familiarity with and understanding of the musical material is required, for instance, for arriving at one’s own interpretation of a piece — an important part of the artistic process. Creating solid versions is therefore an essential skill that can be cultivated and developed in the theory classroom.[8] In this case, the endeavor is necessary to better understand the artistic choices in the Sarah Vaughan version.

Over time, a great many standards and other core pieces from the repertoire have become available in the lead sheet format commonly used in jazz; these are not always reliable, though. For example, it is often impossible to determine how a version was created, and the risk that there are outright errors is considerable. No lead sheet can thus form the point of departure without a check. Many show tunes were published in a version for voice and piano. Together with the earliest recording(s) they are important sources.[9] For our purposes here, we use the version for voice and piano of Weill’s song.[10] Example 1 shows the first eight measures of the chorus, transposed to the key of Vaughan’s recording, A major.[11]

Example 1: Kurt Weill, September Song, mm. 1–8

At this point, depending on the time available, a number of aspects can be incorporated into the class. For instance, we could consider the relationship between the text and the music,[12] or an extra task could be to transpose the published C major version to the key of the Vaughan version, A major. Whatever the choices may be, the notated music, especially the harmony, is discussed in detail.

Of particular interest are the opening measures. The harmony in m. 1 is a minor added 6th chord, with the (unstable) 6th in the melody. This is surprising after the verse, which is in major throughout, and the upbeat line in the melody that consists of an incomplete arpeggio of the major tonic (––). The major-to-minor change provides a sudden shift in the atmosphere. M. 2 provides even more instability and melancholy with the introduction of the in the melody. In m. 3, the tonic chord appears as a major chord again, with in the melody. Evidently, at first, the opening sonority will be perceived as, and arguably is, a tonic minor 6th chord. But we can also interpret the –– line and shift of to as two slowly operating voices, suspending the tonic’s third (C/B to C) and fifth (F to F to E), respectively. We can, therefore, consider the first four bars of the tune as actually only harmonized with the tonic, A. Example 2 shows the voices in the lower staves; the –– line appears both in the melody and a middle voice. Thus, the passage forms an excellent opportunity for discussing the importance of voice leading and introducing the principle of prolongation as well as the various levels of harmony (foreground, middle ground, etc.).

Example 2: Voice leading in mm. 1–3

The passage also provides an occasion to discuss the appearance of a minor tonic in a major-key piece. The chord symbol notation used in lead sheets may easily lead to analyses in which the chord quality is simply translated into a Roman numeral. Thus, the first bar would be analyzed as “I minor” as an instance of “modal interchange.” However, such an analysis does not reflect the “how?” — the crucial role of the voice leading in the passage. Also, the “modal interchange” label easily obscures the “why?” — the ambiguity of key and juxtaposition of minor and major reflecting the melancholy and ambivalence of the text. The passage offers opportunities of exploring how changing the quality of a key’s third, in principle, entails a modulation and remains an exception in the tonal repertoire. It could invite discussion of pieces in which this also occurs or in which this only seems to be the case. An example of the former could be Autumn in New York, which starts in major but, after some earlier hints, ends in minor. The Man I Love could serve as an example of the latter: The major tonic in m. 1 changes into a minor chord in m. 2, which is not tonic but functions as a subdominant in relation to the next chord. It is my experience that the students usually volunteer many more examples, which enriches the discussion of the topic.

To better articulate the issue of the voice leading, it is helpful to have every student prepare a lead sheet version of the composition. We then encounter the problem of there being no satisfactory solution for the opening measures, especially in the second bar. Here, we have no alternative but to notate the symbol as F/A. Even if we are aware that the slash-chord notation can be misleading and that the bass note may, in fact, be the root, the symbol F/A will still produce the (psychological) effect that the chord is F. Example 3 shows such a lead sheet version in the key of A.[13]

Example 3: Lead sheet translation of the published version of Kurt Weill’s song

For the remainder, the chorus is not particularly harmonically dense, and the harmonic procedures are not complicated.

Transcription and Notation

Musical notation provides us with something tangible — a means to communicate in class. Therefore, we will use a transcription of Clifford Brown’s solo as the basis for the further analysis. Notation captures only so much of the music; the transcription will mostly serve as a mnemonic device to evoke the auditive memory of (specific parts of) the performance. Making a transcription partly involves analysis as decisions need to be made regarding notation. In terms of pitch, this usually concerns whether to write ghost notes or not, estimating how to write inflected pitches, and how to write very fast passages that are difficult to hear. Most decisions, though, involve the notation of the rhythm. I take the view that the notation should reflect how we hear the music as closely as possible, regardless of early or late timing. In this case, one specifically important decision concerning the notation of the rhythm must be made. The piece is played very slowly, even for a ballad. At the end of the head, a change is prepared to double-time feel, first in the drums (starting around 2:20).[14] At the beginning of Clifford Brown’s solo, the double-time feel has taken precedence. Since the notation of the rhythm is so closely connected to the articulation, in my transcription, I chose to write in double time. Thus, two measures in the transcription equal one measure in the original song. In this way, the notation best reflects the triplet swing feel of the line. It should be noted that because bassist Joe Benjamin stays in two (relative to the double-time feel), the actual pulse remains somewhat ambiguous. This ambiguity is one of the major factors in the creation of the relaxed yet melancholic atmosphere of the performance.

Of course, preferably, all students transcribe the complete solo. This is not merely a means to integrate a form of ear training into the class; it is by far the best way to become intimately familiar with the performance. Additionally, comparing the various transcriptions in class can give rise to useful discussions. It also provides the best handle for addressing details of Brown’s performance. Especially if there are trumpet players in the group, the latter can be very profitable: if they are willing to play examples (which is not always the case), they can provide tremendous insight into the actual difficulty, power, and expressiveness of Brown's playing and make many extra details audible to non-trumpet players.[15] Since I focus on other elements in this paper, in my transcription, shown in Example 4, I have included only a few ghost notes and an occasional indication of timing (mostly swing vs. straight). Most of Brown’s inflections, grace notes, microtiming, etc., have not been included. Of course, not every student will transcribe equally quickly and accurately. To what extent the transcription should be prepared by the students themselves depends on the group composition and the specific focus of the course (more analysis-oriented or more integrated). Either way, students should be very familiar with the recording and should have listened intensively before the next analytic step.[16]

Example 4: Transcription of Clifford Brown’s solo, mm. 1–32 (starting ± 2:40) and mm. 48–64 (starting ± 4:20)

Clifford Brown’s solo

Now we have analyzed and familiarized ourselves with the original song and prepared the transcription, we can turn to the solo. As discussed above, many events in the first four bars of the original song, such as the apparent change of minor to major, can be regarded as voice-leading phenomena, and the harmonic function is essentially tonic. This alone is reason enough for reharmonization, but the very slow tempo in this performance makes it all the more necessary. The first eight measures (sixteen in the transcription) have been reharmonized. The ambiguous opening tonic chord has been replaced with IV, appearing as a “blues-IV” (D7). This chord not only fits the style nicely but also adds to the ambiguity because of its instability in the key. Another interesting aspect of this reharmonization is that the song’s horizontal line –– is one that occurs very often in jazz harmony, usually in combination with the harmonization IV-IVmd-I (or variants thereof).[17] Thus, – has strong associations with the subdominant in jazz harmony. The D7 remains until the end of m. 4, where it is followed by a brief V. The addition of V, however brief and underplayed, points to the stylistic preference within the style for authentic rather than plagal progressions.

Brown’s opening statement is a smart paraphrase of the original melody’s first line. As can be seen in Example 5, he recasts the upbeat into a complete and embellished arpeggio of Amaj7. The long upbeat is followed by what I consider the main motif in m. 1 (labeled a), which is subsequently repeated twice (a and a’). What exactly constitutes the main motif is open to interpretation. It can be heard in different ways, as usually transpires in the student’s transcriptions. Some hear a motif of two, and others a motif of three notes. In fact, both are true, since Brown ghosts the third note, rendering it ambiguous. While techniques such as ghosting notes are common, and determine the rhythm of lines to a great extent, in this case, we can also interpret it as a more structural event. We may note that the motif, including the ghosted note, corresponds to the three-note tone repetition in the original melody. That the third note should be ghosted is interesting, as Brown’s solo will mostly feature the two-note tone repetition, as we will see. Moreover, the original melody’s triple tone repetition returns in another way. In the opening phrase, the motif appears three times. As can be seen in Example 5, each repetition corresponds to a melody tone, at the proper location in the hypermeasure. The second repetition (a’) varies the motif. It is both fragmented (limited to two core pitches) and augmented (from eighth notes to quarter notes). Thus, the opening has a classic aaa’ set-up. The final pitch in m. 2, the B, is the odd one out. It does not appear in the original melody, and in terms of range, it stands out within the line. The two metrical dimensions of the line, one corresponding to the hypermeasures and the other to the surface time feel, remain important throughout Brown’s performance.

Example 5: Transcription of Brown’s opening statement and main motif (a)

Frequently, the apparent complexity of the material that the detailed analysis highlights leads to discussions about whether this fragment was improvised or premeditated. This discussion should not be brushed over light-heartedly because it foregrounds some of the tacit beliefs that students may hold about the nature and the aesthetics of improvisation in a jazz context. An element that may advance the discussion can be found in the introduction of the performance. Here, the flute already plays a motif that is derived from the same part of the melody, which has a bearing on the listener’s perception of Brown’s opening. We can also make excursions to other such paraphrases and their subsequent importance in an improvisation, which are not exceptional in the style. A famous instance is Parker’s opening in the 1947 version of Embraceable You (take 1). While it can be heard as a quote from A Table in the Corner, it is also a retrograde of the first measures of Gershwin’s melody.[18]

In mm. 3–5, Clifford Brown plays a varied repetition of the opening motif, now beginning on beat 3 (see Ex. 5). One instance of motif a is dropped and a tail is added to the phrase, its ending a variation of motif a’. As mentioned, the B of the opening phrase in m. 2 stands out. It is unstable, both in relation to the chord (13 of D7) and in the key ( in A major). In m. 4, the corresponding pitch is now a C, and forms a Sekundgang, a structural melodic progression with the B from m. 2.[19] If the students are not yet familiar with this term, this is a good opportunity to introduce it. It is an important phenomenon, which can be clearly demonstrated here because it allows us to account for the appearance of the B and C. Owing to the Sekundgang, we are more likely to interpret the C as a B, which sounds even more unstable than the B (enhanced by the brief E7 at the end of m. 4). The B-B Sekundgang creates an expectation of resolution to C, and indeed, the subsequent descending arpeggio in the melody in m. 4 “transports” the melody to the middle register, after which the B resolves to the C, supported by the tonic chord (m. 5, beat 1). This should lead to a release of the tension, which is unfortunate at this point because of the long tonic function. However, the one seemingly unsatisfactory element in the passage is the resolution of the B in the wrong octave. Precisely this aids in sustaining the energy in these measures. The loss of energy is also circumvented by the use of prolongation techniques in mm. 6–7. The tonic is elaborated with the passing chord Bm7 and the tonic substitute Cmin7, and the low C in the melody with a mere scale, descending from A back to C. The line provides a contrapuntal motion to the ascending bass and is made interesting by Brown’s various techniques of execution and embellishment.

The C does, however, subsequently appear in the correct octave, appearing first in m. 8 as an upbeat in combination with the F7.[20] The actual moment of resolution of the B to C follows in m. 9. At this point, the harmony changes to B7, which destabilizes the C significantly. In this way, the expected resolution occurs, but the musical energy does not decline. In sum, the / to line from the beginning of September Song is a prominent feature of Clifford Brown’s opening measures in an “extended” version, starting on , and then moving to . Subsequently, the horizontal tension resolves with the low at a point that corresponds to the original moment of resolution, and with the high at the arrival of the B7. The latter’s pronounced and relatively sustained appearance matches the original melody’s pronounced and sustained high B. Such a replacement of an important melodic pitch with a more “hip” version, in this case, the 9 instead of the 8 of B7, is a common feature of the style. From the C in m. 9, the structural, overarching Sekundgang will then descend.

Example 6: Sekundgang in mm. 9–11

The phrase that appears in mm. 9–11 on a deeper structural melodic level sustains the top note C during its three measures. This phrase is very free and diverse and provides contrast to the opening measures through its virtuosity and capriciousness. It provides ample room for discussion about the notation of the rhythm when transcribed. My notation, as shown in Example 6, is not the only option.[21] The line itself is also overarched by a Sekundgang, which appears in a manner that is very typical for bebop. The apex notes appear irregularly and at a faster pace as the line progresses. Thus, the Sekundgang engenders a second rhythmical layer, which provides much of the attraction and swing of the line, quite typical of the style. The faster pace at the end creates a suggestion of acceleration. Yet another Sekundgang appears at the bottom of the line, operating at a much slower pace.[22]

An additional function of the phrase in mm. 9–11 is the octave transfer. It is a distinct feature of the style that phrases such as this negotiate the registers (as can be seen previously in mm. 4–5).[23] Brown brings the line down from the high C to the low B. The low B forms a long-range enclosure together with the G in m. 12, and finally resolves to the A in m. 13 (shown in Ex. 7). In the high register, the high C at the beginning of m. 9 descends to the high C in m. 12. Here, Clifford Brown uses a new motif (labeled x) with contrasting material. Because of its triplets and the use of the , it has a strong blues connotation. Similar to the first measures, Brown thus “works” in two registers, but there is a clear hierarchy between the two, with the high register being the primary one. As before, the horizontal and harmonic tension resolve in the low register first. That the tension is still unresolved in the high register sustains the musical energy.

Example 7: Long-range enclosure B-G-A and new, contrasting motif (x) in mm. 11–13

The high C from m. 12 will eventually resolve to the A.[24] Where exactly one experiences the eventual resolution is a matter of interpretation. It first appears in the same bar as the low C (m. 13, beat 4) and could be heard to resolve there (indicated by the dashed line in Ex. 7). It is also possible to hear the resolution only in m. 15, beat 3, as will be discussed below. We should take note that m. 13 corresponds to m. 7 in the original, which means that, in principle, the melodic and harmonic tension (should) resolve here. And indeed, as we have seen, Clifford Brown plays both the low and high A in the melody. However, even though the tonic harmony is not significantly reharmonized, the passage neither works as an ending nor sounds static. This is because, in m. 13, Clifford Brown begins a process of acceleration. The acceleration is brought about by a change from the relaxed triplet feel to straight eighths and to the use of fragmentation — motif a is reduced to its core (two notes, see Ex. 8). Ultimately only the core motif remains, which, moreover, is played in double time (m. 15, beat 3). As the A is placed on the relative downbeat, in combination with the tonic harmony, its structural melodic arrival can also be experienced here. At the same time, the repeated motif over the now static tonic imbues the music with a sense of “being stuck.”

Example 8: The development of motif a in mm. 15–17

Crucial is the subsequent change of harmony to D7 in m. 16 (Ex. 8). The seemingly dead-end street, so to speak, in which the main motif found itself, turns out to have an exit after all. The D7 destabilizes the A, which becomes the 5 in the chord. The harmony “frees” the core motif, as it were. For such a situation, I use the expression “harmonic freebie.” This is obviously not an official term, but one that will hopefully stick. Its purpose is to make students aware that one can let the harmony do the work. It is as though Clifford Brown responds to the “escape”, for from this point onwards, the motifs again become longer and gradually more melodic. In the passage Brown combines the main motif a with material that relates to motif x. This material thus becomes a more structural, recurring counterpart to the main motif. The fact that we consider this material as structurally significant may lead to some discussion. Evidently, the material is quite generic. It is important, though, to point out that also generic material can gain more weight, for instance, when it recurs, as is the case here. In mm. 18–19, Brown plays a variation of m. 17 (including the upbeat motif). Therefore, we will hear the passage from m. 13 up and until m. 19 as belonging together. This means that Clifford Brown is overriding the form, because in m. 17 the repeat of the A-section has begun. Such a procedure is quite common in the style.

Example 9: a. Variation of main motif a (m. 21); b. Ghost notes

In mm. 18–20, Clifford Brown briefly plays a repeated motif reminiscent of the upbeat to m. 1, with a light triplet feel. In m. 21, he returns to the straight eighths in a passage based on a new variation of the main motif (Ex. 9a). This variation consists of an upbeat and the original three-note version of motif a, as the lower brackets in the example show. The three-note core rhythm itself is constructed from the two-note version of motif a, once in eighth notes and once in a double-time feel, using 16th notes (Ex. 9a, small brackets). Interestingly, similar to the opening measures, Brown uses the ghosted notes as well (Ex. 9b).[25] In addition to the tone repetition, the opening phrase’s ascending perfect fourth (m. 2) also returns here. Again, multiple “speeds” are operating simultaneously: the two in the respective versions of the motive and one more created by the Sekundgang on beats 2 and 4 (see Ex. 10).

The line itself operates in a quasi-two-part manner at first (mm. 21–23). The stationary E (a pedal of sorts) alternates with a quite regularly placed top line (Ex. 10). The top notes ascend from A through B to C, and eventually to the E.[26] From the moment the high E is reached in m. 23, the melody becomes more capricious, and the repeated E dissolves into the line as well. The top line descends more than an octave from E to D. The target note D already occurs at the very weak last 16th of m. 24, but the structural arrival takes place in m. 25, beat 4.

Example 10: Sekundgang creates sense of acceleration in mm. 21–25

In m. 26, Clifford Brown returns to the original triplet swing pulse. The effect is that of slowing down, which installs an expectation of closure. This is reinforced by the descending Sekundgang C (m. 26), B (m. 27), A, G, F (m. 28), to E (m. 29). The melodic content is not very specific, with only light touches of previous material, such as the gesture from m. 12 and motif a (see Ex. 4 above and Ex. 11b). The landing on the E occurs together with the arrival of the tonic chord, but since E is the fifth of the chord, there is only a tentative sense of closure.[27] The phrase in mm. 30–31 functions as a kind of summary but also descends further to the C. Whether one hears the C or the E-flat as the structural target note in the passage is open to interpretation. What is clear is that Clifford Brown does not end on the tonic note A.

Example 11: Instances of unemphatic return of earlier material supplying coherence: a. mm. 4–5; b. mm. 26–27; c. mm. 30–31; d. mm. 49–50

At this point, the chorus is only halfway. Nevertheless, another soloist takes over, flutist Herbie Mann. A switch of soloist occurring at a formal point of the chorus other than the top is not unusual, and certainly not in slow ballads. However, a subsequent return to the previous soloist to play the final 16 bars, as happens here, is less common. In a sense, this makes the flute solo part of the trumpet solo. Brown’s return makes clear why he did not end on the in m. 31; the closure at that point was only tentative. That this section concludes the complete chorus is certainly something to listen to and discuss in class, also because some transcriptions do not include it.[28] (For reasons of space we will only focus on Brown’s playing here.)

The rhythm section’s playing in the passage supports the winding down. The final section is of a quite laid-back nature, both in content and execution. Brown uses a relaxed tone and easy-going triplet feel, underscored by the fact that pianist Jimmy Jones plays less in a double-time feel than in the previous parts. Brown’s musical material is deceptively simple and generic. However, Brown does make use of earlier material, especially the main motif. It occurs at the end of most of the phrases in the augmentation (a’), but rather unemphatically. That the listener perceives it as the main motif is a result of Brown’s handling of the motif throughout the improvisation. After all, in and of itself, this material is completely generic. Example 11 shows some of the stages of this development after the opening statement. The fact that such generic material can be “structural” or “main subject matter” in a solo, can be a steppingstone to discussing the relationship between style and originality.

In the passage Clifford Brown actually plays the same –– ascent from B to C as in the opening measures, but, again, in this section in a more relaxed manner. That this section functions as a “wrapping up” of the complete 64-bar chorus, becomes even clearer with the augmentation of the rhythm in mm. 42–46. Virtually all that occurs in these bars is the –– descent. The durational proportion of the notes may seem exceptional in relation to the complete solo, but in relation to the underlying proportions of the chorus (the hypermeasures), they are not.

Clifford Brown first lands on the A in the “proper” register and, by octave transfer, also closes an octave lower. The descent employs the main motif in such a way that even the double-time feel recurs in this codetta of sorts.

Example 12: Closing section (mm. 49–64)

Resumé and some final thoughts

As discussed, harmonic theory is central in jazz pedagogy and is conceived in very broad terms. It provides the basis not only for realizing chord progressions but also for playing melodies. The focus of mainstream pedagogy is mostly on chords and scales. However, because of its importance for practice, it is crucial to teach harmonic theory in depth and to incorporate all its facets. We should also realize that harmonic theory, because of its importance for performance, feeds back into the practice.

In the foregoing, I have shown how we can address aspects of harmony and improvisation beyond the musical moment through analysis, focusing, in this case, on Clifford Brown’s improvisation on September Song. It enables us to engage with the structure of a solo, the development of the material, long-range connections, the hierarchy of the material, and the like. The detailed discussion of the development of the motivic material in Clifford Brown’s solo articulates its scope and rhetorical impact. I consider such detailed analysis an essential part of any theory lesson because, time and time again, these insights prove to be eye-openers for students, even if they were already familiar with or had previously studied the solo under consideration.

The opening motif provides an avenue to discuss improvisational invention from the perspective of the composition. What does the piece offer in terms of creative possibilities, or potential pitfalls (such as the extended tonic passage)? The less important motif x provides an occasion to consider that generic material can be structurally significant in a solo. Furthermore, it can spur discussions of which material we actually hear as derived from motif x.

Analysis can also form an opportunity to address more general aspects of melody and harmony, especially those that mainstream pedagogy does not always thoroughly consider. To this end, an integrated approach, in which analysis is combined with harmony and ear training, is especially apposite. Comparing the harmony of the original piano-vocal sheet music and the reharmonizations in the A-sections in the recording makes it possible to address aspects of harmony that can easily be overlooked, such as voice leading, structural versus embellishing progressions, harmonic layers, or the rhythmical dimension of the number of chords per bar. It also enables us to consider which musical reasons may have underpinned some of the decisions made in the performance and how the material of the solo relates to the underlying harmony in terms of structural versus embellishing moments.

Of course, producing analyses such as the one presented above are only initial steps, but important steps nonetheless. The ultimate goal is for students to be able to integrate these insights into their own playing. We can help students take that step by suggesting how to create practical exercises. This can be supported by assignments that are more practice-oriented but can still be dealt with in the theory lessons. Based on the analysis above, one assignment could be to ask students to record their own improvisations and analyze them for aspects addressed in the analysis, such as large-scale and small-scale Sekundgang, the interconnectedness of motivic material, or the status and impact of generic material within the context of the improvisation. Another is to design a motif, motif group, or phrase based on the melody of, for example, a standard, which could serve as the basis of an improvisation. Yet another assignment might be to produce a reharmonization of a piece, but to radically change the tempo, to very slow as in this recording, or contrariwise. By discussing these assignments in class we can cultivate a broad analytical attitude, including toward one’s own playing — an attitude that is indispensable in a living musical culture.