Paul Berliner: On Improvisation

Paul Berliner is a Professor of Ethnomusicology at the John Hope Franklin Center for International and Interdisciplinary Studies, and jointly, the Music Department at Duke University.  He is the author of The Soul of Mbira and Thinking in Jazz: The Infinite Art of Improvisation (both University of Chicago Press).  With his expertise in both jazz and Shona mbira music from Zimbabwe, Paul was the man to talk with for Afropop Worldwide’s program, The Art of Improvisation: Part 2.  Banning Eyre met him in New York to discuss similarities and differences between these two music cultures, both of which value improvisation.  Here’s their conversation.

Banning Eyre:  In general terms, how is the act of improvisation different from or similar to improvisation for an mbira player and a jazz musician?

Paul Berliner: To begin with, when we make these kinds of comparisons between jazz and mbira music, it’s important to keep in mind the limits of generalization. That is, there are different creative practices associated with improvisation within each of these traditions.  The practices can differ from musician to musician, according to personal taste and ability.  They can differ in relation to conventions that have been established by different generations of performers, or in relation to different schools of performance.  They can also differ according to the precise musical arrangements that artists create for particular compositions, or for performances in different musical or social contexts.  Despite such variability, there are common issues pertinent to improvising in both the jazz tradition and the mbira music tradition.

One issue concerns the forms that contain musical ideas and give formal shape to improvisations.  In mainstream jazz, there are cyclical forms that guide improvisation, typically composed melody lines and their accompanying chord progressions, for example, a 12-bar blues or a 32-bar popular song. Jazz musicians treat compositions as the vehicles for their own inventions.  They serve as stimulating blueprints, springboards for the conception of new ideas.  Artists can draw harmonic, melodic or rhythmic material from the composed melody and chord progression.

Another issue has to do with the roles that individual band members play in different musical traditions.  In mainstream jazz, the distinction between the rhythm section and the soloist is important—what the bass player feels his or her job is, what piano players feel their job is, how drummers feel they should organize their parts in order to complement the bass player and the pianist, and the way they all work together to support the soloist and establish a groove. Different roles shape improvisations in particular ways, providing guidelines for band members creating their own parts in performance.

B.E.:  So each of those guys is making personal decisions, and the sum total of those decisions gives the group its character, and that’s why each group has its own character.  Is that it?

P.B:  Yes.  Jazz is created collectively, and part of the uniqueness of every performance has to do with the individual decisions artists make from beginning to end, and the way in which they converse with each other, inter-relating their parts.  An intimate discussion takes place between them in the language of music.

Still another issue is instrumentation.  In a conventional jazz group, the bass player takes care of the bottom within the musical space.  Perhaps the soloist occupies the space high above.  The pianist has the ability to move back and forth, joining musicians in their respective range, or filling in the space between their parts. The collective texture that artists create in jazz is partly a function of the group’s instrumentation.

In the Zimbabwean mbira music context, the same issues play out in a different way.    Composed pieces and harmonic-rhythmic cycles also provide the ground for improvisation, but their characteristics are different. Let’s take classical pieces for large Shona mbiras, like mbira dzavadzimu.  Rather than an independent melody line and chord progression of the kind used in jazz, mbira compositions minimally consist of two cyclical polyrhythmic/polyphonic parts, each embodying two or three interwoven lines. One part is called the kushaura or lead part. The second is the kutsinhira or following part. The two are interlocked in performance to create a complete representation of the composition.

An harmonic-rhythmic cycle provides the formal underpinning to mbira compositions, but the cycles are shorter than those typically used in jazz.  From a theorist’s perspective, one way to describe the form is a 16-beat cycle containing a sequence of 12 two-note chords. From the aural perspective of artists, the form serves as a timing device and provides harmonic guidelines for the conception of new ideas in relation to the composed parts of the mbira repertory.

Establishing further conditions for improvisation within an mbira group, the composed parts are performed by at least two musicians playing the same type of mbira (in the case of mbira dzavadzimu, an instrument with a three-octave range) who assume specialized roles as kushaura or kutsinhira players. The kushaura part commonly contains more of the melodic essence of the composition and places less emphasis on the bass. Also, there may be fewer variations associated with the kushaura part. In contrast, the kutsinhira player commonly places greater emphasis on the bass lines in creating variations and takes increased liberties in developing his or her part in the context of the ensemble. To facilitate the interlocking of the parts, the kushaura and kutsinhira players maintain different rhythmic positions with respect to the beat. Artists maintain their respective roles throughout the performance of piece, making decisions pertinent to the development of their own parts, and also to the relationship among the parts.

Here again, the differentiation of musical roles in relation to the overall texture of the music is important. Since an ensemble can have two, three, four, or five mbira players, each playing an instrument with the same range, artists have to be sensitive to the ways in which they treat musical time and space in creating their own parts– so as not to get in each other’s way. They strive to complement one another in a manner that is respectful, and effective musically.

B.E.:  So it is not that one of these is totally set and the other is improvising.  Both have aspects of improvisation within their roles.  Is that right?

P.B:  What’s interesting is that these roles can be defined in different ways, and that brings up the issue of musical arrangements—the decisions musicians make outside of performance that shape particular renditions of pieces.  You can have an arrangement in which, by decision, the kushaura player holds down the basic part, establishing the musical ground in relation to which other players vary their parts.  That represents one extreme, a situation in which the kushaura player restrains his or her part, while the kutsinhira player has great freedom of expression, weaving in and out of the first part and creating a stream of variations.

In both jazz and mbira music ensembles, you can think of a continuum of practices associated with improvisation.  At one end of the continuum, you increase the number of composed elements and minimize improvised elements.  At the other end of the continuum, you maximize improvised elements and minimize the composed elements.

B.E.:  That’s interesting.  In a jazz ensemble, you have more different instruments usually, so you have that timbre contrast.  Even if two people are playing the same thing, you would be less likely to confuse them.  Whereas if everybody is in the same range, you have to be more careful.

P.B:  Exactly.  Of course, these situations also arise in jazz.  For example, the pianist might play an active bass line that could potentially get in the way of the line that the bassist is developing.  More commonly, by decision, or according to the conventions for complementary interaction, a pianist will leave the bass line to the bass player (who emphasizes the roots of the chords and outlines the harmonic progression) and focus instead on harmonic coloration. It’s a matter of delineating and allocating different musical roles within the group.

B.E.:  The versions of the mbira piece “Nhemamusasa” on your Nonesuch Explorer release Shona Mbira Music really let you hear the different mbira parts, first the kushaura in isolation, then the kutsinhira added, and finally the hosho percussion.  Then there’s a second version with the two parts playing together.

P.B:  That’s right.  On that second track, the kushaura and kutsinhira players are both taking expressive liberties, so you get a sense of the interplay and conversation between them.  They unfold their parts in different ways, but in relation to one another, while the hosho (shaker) player keeps the beat. The third track presents the complete ensemble, featuring the famed singer Hakurotwi Mude, who brings in various improvisatory singing styles.  He plays off the mbira players, interweaving his part with theirs, and filling out the performance.  You hear Mude singing in the mahon’era style, the low, syllabic, riff-based style.  At other times, he switches to the upper voice, the huro style, the high, stylized form of yodelling, which adds tremendous poignancy to the music.  Then sometimes, he intercuts those passages kudeketera, the improvisational poetry.  Through his imagery, he introduces extra-musical layers of meaning to the performance. In one line, he refers to the “black tree,” a sacred tree beside which his ancestors worshipped. In another line, he scolds an ancestor a “witch ” who has caused him problems in life.  The artists develop the piece together, constantly playing off one another, talking together in the language of Shona mbira music.

Over the course of the three tracks, we move from a demonstration of a fixed kushaura part combined with a varied kutsinhira part, to a demonstration of variations created within both the kushaura and kutsinhira parts, and finally, to a performance by the complete ensemble which includes expressive freedom in all the parts.  On those Nonesuch CDs (The Soul of Mbira and Shona Mbira Music), there are also two tracks that reveal the ensemble’s conversation with the audience, dancers, and spirit mediums.  One of these tracks (on The Soul of Mbira), provides another version of “Nhemamusasa,” which was recorded in the larger musical and religious context of a bira, or ceremony for the ancestors.  Here you get a sense of the tremendous intensity of the music that occurs with audience participation in the context of collective worship.  I’ll talk more about such issues later.  But just on the level of musical performance, it is interesting to hear how flexible the structures associated with mbira pieces are—how radically different the ‘same’ piece can be in different social and musical contexts.

B.E.: You also talked about a jazz group where different roles and arrangements would be agreed upon among the players.  That’s often pretty explicit among jazz musicians, something they talk about and arrange in advance.  Do mbira players also discuss and agree upon roles for a particular performance, or would it simply be based on intuition, or status?  How to those decisions get made?

P.B:  In the mbira context, it happens different ways. It’s similar to jazz in terms of creative process.  Sometimes musicians get together in rehearsals and try out different things. Musicians really do talk about these things.  “Let’s try this kushaura with that kutsinhira.  Let’s pair different things together.  What do you like?  I like the way it’s interlocking in the upper voice, but the bass is not working there.  Can you compose a different bass?”  There is collective composition among players working together. In the context of the group, they create and cultivate specific performance practices, and that gives individuality to the sound of each group.

Distinctive practices can also be developed in performance. Artists can subtly dictate things to one another, giving each other cues and communicating their views about what works and what doesn’t.  In an mbira group, if one player changes something in the kushaura part, creating new lines that the kutsinhira player has never heard before, the latter may call out, “Ipapo!” literally, “There, that place!” meaning, “That’s great, what you did, stay there. It’s given me an idea.  Keep that going so I can play off it.” Players can direct their interaction that way, passing the inspiration back and forth, stimulating one another’s creative imaginations. Other times, when the music is really happening you’ll hear musicians praise one another, expressing their exhilaration at what they have achieved. “Pisa!” someone may shout out, meaning, “Burn!” as in, “This music is really hot!”

On the other hand, if you experiment with an idea that you think is interesting, and someone you respect glances over at you with a frown, that person is telling you he didn’t appreciate what you played—or that it got in the way of what he was playing. It can be devastating.  Or, it can be as subtle as somebody refusing to respond musically to an idea that you’ve introduced and invited him to take up. Beyond the normal course of communicating through the language of music within the group, there are modes of direction and criticism conducted through body language and sporadic verbalizations.

These interactions continue after performances. Suppose mbira players return from a exciting musical event in which unanticipated things happened in the group’s interplay and, in the moment, everybody responded well to it—creating new musical structures in the process.  Those can be the high points of improvisation. Afterwards, someone will say, “What did you do on that tune? I never heard that kushaura part before. That was fabulous.   It gave me an idea for a kutsinhira part, but I couldn’t quite finish it.  Let’s work it out in rehearsal so we can play the parts together in the future.”  Once people have been playing together for a long time and know each other’s musical personalities well, they can also sense what was successful during performances, and without discussion, strive to re-create aspects of their interaction again in subsequent performances.

Remember, when you hear a great improvisation, you are just getting the tip of the iceberg. Behind the artists’ expression in the moment lies years of musical thinking, training, and experience.

B.E.:  You talked about the spectrum of different degrees of freedom in different performances of the kushaura and kutsinhira parts. Can you give us some examples?

P.B:  Sure. On Shona Mbira Music, there are two versions of “Taireva.”  On “Taireva, version 1,” Cosmas Magaya plays the same basic kushaura part throughout.  Actually, at the very beginning, he plays a rhythmic variant, but overall, this is an example of how within a particular arrangement, one basic part, repeated, can be sufficient.  Meanwhile, the other musicians are varying their patterns in relation to the stable kushaura part. You can hear Mude singing, improvising his whistling parts, and drumming.

An interesting contrast is “Taireva, version 2” where you have greater interaction between the kushaura and kutsinhira players.  Cosmas, who is playing the kutsinhira part, takes interesting liberties with the bass, sometimes removing it altogether, which changes the texture of the music radically. Other times he creates surprise by reintroducing it, playing different kinds of bass lines and variations.  Meanwhile, Mude picks up fragments of the mbira parts, seizing upon particular rhythmic or melodic ideas which he develops in his vocal part.  At other times, he quotes the entire kutsinhira bass line in his high vocal register. As musicians pass ideas back and forth within the group, their interplay gives cohesion to the performance thematically.

B.E.:  That description could have been a description of the jazz performance.  It reflects the kinds of interactions, ideas, agreements, roles you find in jazz.  The two disciplines are coming up more similar then I would have expected.

P.B:  That’s right. Despite the features that distinguish the languages of jazz and mbira music, there are similarities.  You might think that because the interlocking aesthetic of mbira music dictates a very tight mesh between the kushaura and kutsinhira parts, that that might inhibit improvisers, compared with the parameters jazz musicians typically work within. Paradoxically, when mbira players establish a powerful groove within the complementary interlocking relationship, it generates a great feeling of mutual support which inspires greater freedom of expression. At such moments, artists describe a flood of ideas coming to them, and they begin departing from their  pre-composed parts to experiment with developing new figures, trying out new cross-rhythms, taking harmonic liberties.

B.E.:  What you just said might be about as close as we can come to a cross-cultural definition of improvisation among a group of musicians.  There is an emotional thing that happens, when the player almost surprises himself, or there’s a surprise shared among the musicians.  That becomes an emotional impetus that drives creativity forward and leads to new creation.

P.B:  Absolutely. Emotion and intellect are partners in the conception of musical ideas. These are common aspects of improvisation across different musical traditions.

B.E.:  Let’s come back to the issue of the contrasts.  In what other ways are they different, mbira and jazz, from the point of view of improvisation?

P.B: Well, the distinctive features of mbira music include sustained polyrhythmic patterns that provide the underpinning of compositions. Also, the constant interlocking relationship between the kushaura and kutsinhira players. Since each plays a polyphonic part, their combination produces different kinds of resultant figures at every level of the music’s texture. Typically, the pitches of the parts’ upper voices alternate every other pulse, while the pitches of the middle and bass voices form different kinds of resultant figures. When you’re playing mbira music, your attention is on maintaining perfect interlocking relationships at every pitch level.  The conversation between players takes place within this scheme. Sometimes this involves imitative interplay. If one player creates upper voice figures that leap to the highest pitch on the mbira and descend scale-wise at a very fast clip, the second player can create the same pattern one pulse behind. Together, they create descending patterns with a beautiful rippling effect.

These are simply different parameters than those that jazz artists typically work within. The mbira players’ tight interlocking relationship creates different conditions for improvisation than is found, say, in relationship of the jazz soloist to the rhythm section. The closest analogy in jazz might be the sustained musical relationships within the rhythm section through which bass player and drummer create a groove, and particular moments, say, in which the drummer and the pianist play explicit interlocking figures to delineate the piece’s form, like a turnaround at the end of  the chord progression.

B.E.:  It seems to me that the jazz soloist is freer than the improvising mbira player to really sail off.  It’s not that he’s ignoring the other musicians, but he is not constrained in the same way, right?  Mbira players are not comparable to, say, John Coltrane tearing loose on a wild solo, when his rhythm section isn’t.

P.B: That’s certainly true for the mbira players. One the other hand, the singers’ parts are more flexible, especially when they move away from text-bound vocals, using their voices as instruments, creating huro and mahon’era phrases. Singers like Hakurotwi Mude have the freedom to interweave their vocals among the mbira parts in various ways, drawing inspiration from the mbira figures and lines, but also extending and varying them.  Singers can take greater rhythmic and melodic liberties than their counterparts in the group.

Of course, if you are asking about the contrasting features of these traditions, anyone listening to samples of jazz or mbira music knows right away that they are different musical languages. Jazz musicians use a harmonic language that draws heavily on Western art music, which features a complex system of harmonic progressions, and they employ various kinds of African-American innovations harmonically and melodically.  Jazz musicians play instruments that support the performance of different kinds of scales and chromaticism.

In contrast, the 23-key mbira dzavadzimu is based on a seven-pitch scale, subject to different tunings. It has a three-octave range.  That is the diatonic tonal palate, the tonal universe, within which you are creating music—very different from a chromatic tonal system that divides the octave into twelve equal steps.

As we discussed earlier, the harmonic characteristics of mbira music are also different from those of jazz. From a theoretical standpoint, its underlying harmonic form can be viewed as a cycle of 12 two-note chords (root, fifth and sometimes, the third). The roots of the ‘chord’ sequence spell out the movement: 1-3-5; 1-3-6; 1-4-6; 2-4-6. The melodies of mbira music and the improvisations of mbira players are guided by the basic sequence. That’s not the end of it, however, because the classical mbira repertory makes use of the sequence transposed to seven pitch levels. With virtuoso skill, mbira players negotiate the features of the harmonic sequence aurally when creating variations—interpreting it at different transposition levels and in different mbira tunings.

B.E.:  Maybe we should say something about rhythm here to, about the polyrhythmic nature of the mbira music.

P.B:  One of the features that gives the world of imagination of mbira music its distinctive qualities is its polyrhythmic base.  Of course, polyrhythms are well known to jazz musicians too. In Thinking in Jazz Charlie Persip recalls the inspiration he took from drummer Elvin Jones’s performance style and he gives a wonderful description of the character of polyrhythms. On the one hand, Persip says, you have a triplet feeling that is relaxed:  ‘1, 2, 3; 1, 2, 3; 1, 2, 3…’. On the other hand, you have a duple feeling that forges ahead; ‘1, 2; 1, 2; 1, 2….’ like a march. When you combine those feelings, you get a complete rhythm that marches and still relaxes. Persip concludes, “That’s what makes the music sound so complex.”

The polyrhythms that jazz musicians employ at times to animate their music are actually built into the forms of many mbira compositions. Mbira dzavadzimu players commonly perform patterns with a triple feeling in the right hand while simultaneously performing patterns with a duple feeling in the left-hand, sustaining the relationship from the beginning to the end of performances.

That discipline is essential for mbira players. Polyrhythmic architecture provides the foundation for their world of imagination. It is this world that that they operate within, exploring all the possibilities for invention it enables. Ultimately, they use polyrhythmic resources to explore ideas that are melodic and harmonic in character, and to create rhythmic ideas that have increasingly-abstract relationships to the beat.

B. E.:  Polyrhythmic architecture.  I like that.  Can we talk a little bit about the worlds that created these two kinds of music?  You talked about jazz as existing at a crossroads of European art music and African-American traditions.  I would also say it is the context where individual free expression is a kind of bottom line of what culture is all about.  That is very different from the world that created mbira music.  Maybe you could contrast these context a little bit.  Mbira music is an ancient tradition.  It existed before Modernism, Cecil Rhodes, any of that.  Don’t you think the character and nature of mbira music is fundamentally the same now as it was even before Europeans came to Zimbabwe?

P.B:  Well, one of the factors which has had a conservative influence on mbira music in Zimbabwe is its role in traditional religious practice.  If your sense of identity as an mbira player is tied to performing at religious ceremonies, the focus of your musical life is on the classical repertory–the music passed down to the present from the ancestors. In the setting of the bira–the ceremony for the ancestors–the goal of mbira players is to bring about the possession of spirit mediums, who serve as advisers to other participants, counseling them and assisting them with their problems.  The theory is that if you want to call departed relatives to these ceremonies, the music that you play needs to be the music that they would have heard when they were in the world of the living.  A sense of history permeates the classical repertory.

Part of the skill of mbira players, and what distinguishes different individuals and groups is the affective power and speed with which they are able to bring about possession.  On the one hand, there is something conservative about the tradition and social context. It is looking backwards. It strives to preserve the music of the past.  On the other hand, conventional practices that invite artists re-create the classical repertory in novel ways that resonate with different mediums serve as a creative force.  Working within the frameworks of ancient compositions, artists look for individual ways of expressing themselves. They take pride in their unique creations while at the same time, pursuing the goal to call the spirits and successfully bring about possession.  To be bound to tradition and to be musically innovative is not a contradiction.

It’s also important to consider that in the context of the bira, the interplay between mbira players, singers, dancers, other participants also contributes uniqueness to the rendition of mbira compositions.

B.E.:  So I suppose there could be tension between the desire to express oneself, and the desire to bring about possession.  Sometimes, I understand, you have to play a song in a specific way that appeals to a particular spirit.  That’s pretty arcane, esoteric knowledge, isn’t it?

P.B: That true. Beyond all their technical knowledge of the music, mbira players acquire esoteric knowledge of the different personalities and tastes of spirits who possess particular mediums.  Over time, artists become expert at assessing the musical requirements of each situation.  But there are always capricious elements.  You never know exactly what’s going to happen at these events. And sometimes mbira players are not successful. At one event I heard musicians discuss a fine group that played to its utmost, but couldn’t call the spirits. A representative from the ceremony was sent to the village of great mbira player Bandambira to wake him from his sleep and implore him to return with his own group to help complete the bira.

Consider the connoisseurship of the artists. They master the classical mbira repertory. They cultivate the skill to create variations and new parts, so that they can perform with great fluency and imagination in these situations. Comparable connoisseurship is shared by worshippers at the ceremonies. They pick up on subtle details in musicians’ performances.  They respond to the ingenuity of the players and express their enthusiasm when new parts are introduced.

B.E.: Let’s talk more about the role of the audience.  That was a huge consideration in our conversation with Professor Racy about improvisation in Arabic music.  In that context, musicians are very much in conversation with the audience present.  Arabic musicians are trying to bring about a particular kind of emotional experience on the part of the listeners.  It seems there is something similar going on here.

P.B:  This is very important in the context of the bira.  Imagine the production of music in terms of multiple kinds of conversations.  We’ve used this metaphor before in discussing jazz.  A musical conversation takes place among players within the group. The characteristics of a piece as it unfolds in performance is a consequence of that conversation.

The interplay between the performers and the audience represents still another kind of conversation that can affect the development of the music.  If an audience is really listening, is really attuned to what artists are doing, that can offer encouragement to artists and inspiration for invention. In Thinking in Jazz, trombonist Curtis Fuller talks about how he likes to interact with the audience as a soloist.  He says he “feeds off the audience.”  Sometimes he deals with people in the audience who don’t know jazz very well, he says, so he’ll play something that a layman can understand.  He might quote a popular tune or throw a lick in his solo that’s really soulful and can get across.  If somebody in the audience responds with a shout, he knows that he’s touched that person, so he’ll develop that figure for a section of a solo.  As Fuller senses interest tapering off in the audience, he’ll move on to other musical ideas till he finds one that  engages the audience again.

In the Shona music context, similar interaction takes place as the mbira players develop  the piece, talking back and forth within the kushaura and kutsinhira parts. At the bira, when the music inspires members of the audience, seated on the floor of the roundhouse, they begin singing. They interject their own lines of poetry and, as they feel moved to do so, add mahon’era and huro vocal parts.  The growing sense of engagement and multiple layers of song can be exciting, as everybody contributes to the performance.  In turn, that fires up the mbira players, who put more energy into their playing.

Interplay can also be more specific, reminiscent of Curtis Fuller’s discussion. When Cosmas Magaya switches to his heavy bass patterns, the music surges ahead.  Suddenly, 4 or 5 people leap up from their seated positions and rush onto the dance floor. Their response can encourage the mbira players to sustain the effective parts and develop them, perhaps intensifying their emphasis on powerful bass lines.  Cosmas describes this as throwing firewood on the fire.  When you know you’ve got people involved, it’s no time to let the embers burn down.  The same thing can happen in the interaction between the mbira players and the singers in the audience.  If people start singing, adding the highly-emotional yodelling, and women begin ululating, cheering the musicians on, players may move from parts that are relaxed and meditative to parts that establish a heavy groove and really start to rock.  The texture of the music changes dramatically.

On the two tracks I recorded at a bira at Kwaramba Village (“Nhemamusasa” from The Soul of Mbira, and “Nyamaropa Yekutanga” from Shona Mbira Music), you hear great vocal performances from Mude, who is the featured singer, but also from the audience as individuals add their own vocals.  You hear the ululation of the women, and high, dental whistling from the men.  You hear makwa rhythmic figures interjected into the music by audience members clapping with cupped hands, a kind of surrogate drumming. You also can hear a softer rhythmic figures in the shangara-style of dancing—in which dancers ‘drum’ on the ground with the souls of feet,  treating the ground like the membrane of a drum.

Imagine the excitement created by all of these parts in the context of the bira.  The mbira players are at the center of this, attuned to changes occurring all around them.  They are watching people on the dance floor very carefully, and anything interesting that happens can inspire them to take the music in another direction.

All the while, the larger performance is leading to the possession of the mediums.  Those are the high points of the musical performance and the religious drama. Mediums commonly have stylized ways of announcing their possession. Their voices change.  Their way of singing and dancing changes. It can be dramatic, the huge burst of energy  that accompanies possession. Some mediums’ behavior takes on the characteristics of an epileptic seizure initially. Their bodies shoot out from a quiet seated position and writhe on the floor.  I knew one medium whose possession began that way, then he suddenly leapt to his feet and danced with such force, it galvanized everybody.  The mbira music was suddenly ratcheted up in intensity. Everybody began singing and pulling together. From all parts of the roundhouse, people poured out onto the dance floor.

Years ago, when I first went to Zimbabwe, one particular conversation alerted me of music’s importance in Shona religion.  I was studying with mbira players Cosmas Magaya and Luken Kwari under the guidance of Hakurotwi Mude. At the time, I was solely focused on the mbira. There was so much to learn and I was entranced by the music’s beauty. One day, after I finished practicing, Mude took me aside and said, “There is something I think you need to know. Our mbira is not played for pleasure. It is the way we pray to God.” Over time, as I became apprenticed to Mude’s group, Mhuri YekwaRwizi, and the band members shared the more intimate aspects of their religious practices with me, I came to understand more deeply what Mude was telling me.  It’s in the context of musical experiences at biras, like the performances you hear on the recordings, that you appreciate that participants are pushing themselves beyond the limits of what they would normally endure. They perform through the night without sleep to offer prayers to the ancestors and to assist in bringing about the possession of mediums so that serious issues can be resolved.

B.E.:  I spoke with a Sonrai musician about spirit possession ceremonies in the Sahara Desert.  And he talked about the way kurubu lute player will improvise in the emotional pitch of the ceremony, particularly in interaction with dancers, rather like what you just described.  But then if you take that kurubu player into a studio and say, “First play the melody, and then improvise,” he won’t understand.  In other words, the skill exists, but it is not easily transferred to a purely performance context, like a recording studio.  It seems that many mbira players do improvise in performance and recording situations outside the religious context.  That would seem to be a distinction between mbira players and the kurubu musicians.  Would you agree with that?

P.B:  It’s true that many Zimbabwean mbira players have had experience performing in different contexts, sacred as well as secular. In the latter case, they have played in local beerhalls and night clubs, and over the past few decades, toured on the world music circuit as well.  Regarding the kurubu lute players, I don’t know their music and performance practices, but there would seem to be several things suggested by the Sonrai musician’s remarks. One issue is the lute players’ lack of experience and discomfort with the sterile atmosphere of the recording studio where they are deprived of dancers with whom they normally interact in performance. That’s a separable issue from the players’ actual practices when rendering a melody and a separable issue again, from the local terms in which they think about or talk about their operations as musicians. It’s separable again from the ways in which outsiders may choose to classify the lute players’ practices.

Unless musicians and scholars initially share, or establish, a common definition of improvisation, the use of the term in these contexts can be confusing. This is one of the challenges for anyone interested in gaining a cross-cultural perspective on improvisation.

B.E.:  I hear you.  The issue of definition is huge.  We as Americans have a notion of improvisation that we have gotten largely from jazz culture, and we find resonances in these other cultures, and naturally lump everything under the heading improvisation.  But there is a danger there.

P.B:  That’s right.  We noted earlier in the communities we were discussing—jazz and mbira music—that, in fact, there were wide-ranging practices that could be encompassed by the term, improvisation.

B.E.:  How important is the expression of personality in the world of mbira players?  It seems to me that jazz players are often concerned with putting a personal stamp on their work, something that says, “This is me.”  Even in Mali, I could play a traditional Mande song that Djelimady Tounkara had taught me, and people would say, “Oh, that’s Djelimady, or Oh, that’s Batourou Sekou Kouyate.”  In other words, particular variations on a traditional song were associated with the players who had created them.  Players, and some sophisticated listeners, could listen and make those personal connections.  In that world, I found that artists were very interested in creating their own versions of shared, traditional songs.  I’m wondering if anything like that exists among mbira players.

P.B: There are various considerations for mbira players when it comes to self-expression and improvisation. One is that if you’re playing for long periods of time, you want to keep the music interesting for yourself as well as for the audience. So, while working within short cyclical parts in which repetition is an important element, players also look for inventive ways of transforming them.

Players feel great humility in the face of their musical and religious traditions, but they also take pride in the individual contributions that they make as artists. They render traditional structures in ways that reflect their own musical imaginations, adding to the foundation of the music that was passed down to them.

B.E.:  You mentioned that Cosmas is known for these particularly powerful bass lines.  Do mbira players generally try to develop characteristics of their work as a whole?  Do they think, “I want people to remember me for this characteristic or quality.”

P.B:  Yes. Sometimes when you’re sitting with mbira players, listening to different recordings, one musician will laugh with delight at the appearance of a familiar gesture within a particular part and compliment another musician, “Right away, I knew it was you who was playing.”  Artists appreciate one another for the unique patterns they contribute to the tradition.  So do audience members.  I’ve been present on occasions when people in the audience have called out for certain mbira players to replace other members of a group, because they missed their individual performance styles.  Sometimes, audience feedback is quite specific.  Someone will come up to Cosmas after a performance and say, “I really liked that part you played on ‘Nhemamusasa’ toward the beginning. I never heard that ‘Nhemamusasa’ before.  Please play that again for us sometime.” People have an appreciation of different individuals’ contributions to the music.

B.E.:  I’ve had an interesting experience with Mande music over the years in Mali, something that suggests the musical tradition is really alive and malleable.  When I went back there last year and played songs, familiar songs, the way I learned them back in 1995 and 96, musicians sometimes said, “Oh, but we don’t play that way anymore.  Now he played this way.”  There’s a sense that competition among musicians has steadily changed the commonly accepted ways of playing songs.  It’s a kind of collective evolution of the character of these very old songs.  Over the time the music seems to actually change.  Now, I haven’t studied this formally, but its something I’ve observed anecdotally.  Do you think that happens with the mbira music?

P.B:  Yes, there are situations in mbira groups in which friendly competition arises among musicians. They take pleasure in displaying their skills and imaginative abilities– challenging one another to imitate or otherwise respond to their ideas.

B.E.:  But do you think that changes the generally accepted sense of songs over time?  Imagine that we could hear recordings of “Nhemamusasa” as played in 1700, 1750, 1800, 1850, and so on right up to the present.  Do you think we would hear a steady process of change, or would there be a stable, recognizable version of the song that would run through all of that?  Obviously I’m asking you to speculate, but what you think?

P.B: Over the past 30 years, I’ve observed both continuities and changes within the classical mbira repertory. There are core patterns associated with important compositions  that have been passed with great fidelity from one generation to the next.  At the same time, I’ve observed players reworking traditional structures in performance, imbuing them with musical features that reflect their particular technical abilities and creative practices.

Historically, I believe that improvisation has played an important role in the development of the classical repertory. For example, Zimbabwean musicians describe several compositions as belonging to the “family” of Nyamaropa, one of the oldest mbira compositions. As a parent composition, Nyamaropa has spawned “Mahororo,” “Nhimutimu,” “Chipembere,” “Gorekore,” “Mandarindari” and others. My interpretation is that as early performers experimented with transforming the structures of “Nyamaropa,” they eventually created coherent variations with such distinctive features that the variations assumed lives as independent compositions.  I believe that this is a process that you find in different music-cultures in which improvisation is an important practice.

In Thinking in Jazz, for example, I also describe the perpetual interplay between composed and improvised ideas.  Jazz soloists, may begin with a composed piece as a springboard for invention.  Over years of improvising solos within the framework of the piece, they may find themselves developing habits of musical speech: particular musical figures begin to fall into place as cohesive chains of ideas. Eventually, as these structures become more elaborate, they can form the basis for a nearly through-composed solo, or even a new composition.  The process comes full circle when the new piece serves as a springboard for yet further invention.  Composition and improvisation are not separate worlds.  They are partners in the creation of new music.

B.E.:  Thank you, Paul.  This has been fascinating.

P.B:  Thank you.