Philosophy of Music Improvisation

I think Day misses the idea of soloing and improvisation in his article. First, the difference in ending a concerto piece and a jazz solo, which has been stated by others and Day himself is that an ending of a concerto piece actually concludes the song. However, the ending of a jazz solo is just a conclusion of a section of the song. I view jazz songs as conversations amongst the musicians performing the piece and the audience. Whatever the subject matter is and the structural components of the song, it is the musicians job to discuss through their instruments the tune, society, and emotion. I believe musicians speak with one another on stage and their solos are their opportunities to have to the floor, speaking through their instruments, playing whatever comes to mind. There might be a call and response with others and again this an example of musicians communicating with one another.

I do not agree with Day when he says that, "the end of the solo is, by and large, no particular concern for the improvising artist". I  think that it is the exact opposite case. In speaking, we focus a lot on the first and last sentences that come out of our mouths. We want to capture the attention of audience with the first sentence and also we want to conclude with something that will have a lasting affect on the audience. SImilar in soloing, we want to capture the ears of the audience and also we want to end with something that makes the listener nod in agreement. It is the musician's job to decide what that musical phrase will be and there is no telling if their ending is successful or not. 

I guess what I am saying is that, a solo is a small component of the song, it is the chance for the trumpet or saxophone or whatever the instrument is to have a say in the conversation or story. The ending of the solo is the musicians time to direct the audience's attention back to the overall conversation.

"personal utterance is given to the soloist, while the orchestra speaks for the community through its chorae or its passacaglia." - Kerman.  

Is the goal of Jazz Music to fulfill a compositional ideal? 

One of the problems I found with William Day's argument in "The Ends of Improvisation" was that he, like many of the authors of our musical philosophy readings, limited the scope of his musical analysis of jazz to the standards set by Western Classical music.  In my opinion, this is a close-minded method of examining the modes of a musical work because, in the vast realm of all music, Western Classical music represents a miniscule proportion of the whole.  Day essentially says that the prototypical (Western Classical) musical work mirrors the processes of life in that it sustains itself between a beginning and an end; the end, Day argues, is paramount to the structural ontology of the work because it is the end goal of the work - all things proceed toward their end.  

In the case of Jazz improvisation, however, Day seems to find that the goal of the soloist is to never propose an ending - to create a kind of rhythmic flow in an attempt to suggest infinity.  

I find Day's ideas to be wholly irrelevant to the philosophy of music, and instead to pertain to the philosophy of Western Classical music.  After all, there are so many varying musical genres that feature immense differences in the structure of the work, particularly in how they approach their ending.  Day's argument is irrelevant because he doesn't consult any examples other than Western Classical music by which to measure improvisational Jazz.  

From my perspective, improvisation essentially occurs when any musical work is instanced; i.e. when Beethoven sat down to write his 9th Symphony, he was improvising (by using chord structures and harmonies that he knew would compliment one another in a specific, intentional way), but he was doing so in a very slow, procedural manner.  In jazz, the soloist is essentially doing the same thing that Beethoven does, but on-the-spot in a spontaneous instancing of the work.  The fact that the structure of solos vary or that they can run for ambiguous amounts of time is irrelevant and does not in any way affect their status as musical works - they are very much musical works, they may just never be instanced again.  

I disagree with Day's objection to Williams's approach to analysing Charlie Parker's solos, namely that it is a mistake to analyse this as a completed work, for two reasons.

1) As we discussed in class, the classical works that we take now "verbatim" were originally written in a culture of improvisation.  The music written down was meant to provide a structure that the artists could use as a base to create sounds that showed off the talents of the vocal or instrumental performer.  Taking this history into account, what it unique to jazz is that when a musician is supposed to improvise based on certain characteristics of the piece (for instance, a basic chord progression,) it is usually noted in the score.  I think this is a trivial difference between the two kinds of pieces.

2) Jazz artists, as well as dancers who do contact improvisation and actors who do comedic improvisation, practice this art.  The performer who is improvising has the entire performance leading up to her solo, as well as countless hours practicing this solo, to reflect upon and gather inspiration from during her performance.  This very much mirrors the hours a composer will spend spitting out different lyrical phrases.  A composer may play a piece, perform it (or hear it performed,) all the way through, and at the end of that performance decide that something needs to change for the next performance.  And a jazz artist, who has soloed many times during rehearsals leading up to opening night, might have a very good sense of where repeating a phrase from the written music will sound appropriate and tie things together well.

In the very least, Williams's "apologetic" comments and Day's following differentiation between planned and improvised music misrepresent and underestimate the amount of variation the composer expects in each performance of the piece and the amount of practice and skill that goes into an improvisational piece, and his musically analytical comments about the Parker piece could just as easily misrepresent the intent of a composer in a completely planned piece.

Perhaps the composer merely wrote down what she thought would sound good, and the language we use merely helps us point out where we believe she achieved that goal.  

Day's essay seems funny to me in that he sets up this constraints for himself --to focus on the endings of jazz solos-- to ultimately say that they aren't really endings. The difference between the end of a concerto movement as opposed to the end of a jazz solo seems pretty obvious, in that the former is literally and ending and the latter doesn't actually end the song. Spencer's idea of continuing a story sounds right to me. I see the usefulness of the comparison that he's making when he drives the point home saying basically that the jazz solo is not only not and ending but it is also not a movement toward an ending. It's not even gesturing toward the end. (Certainly contrasting with Western-classical concerti in that sense). Day's final point that jazz solos are "a way of continually proposing a conversation, the end of which is always to discover whether you and I share a world" seems like a vast generalization to me. Like Hilary I don't get jazz, but these seems to be more variability in intent for performers of jazz. The dynamic between soloist and the big band and other soloist could definitely vary from piece to piece. I think Day at the end might be imposing a "program" on all of jazz that doesn't necessarily fit every piece. 

It seems that there has been a lot of discussion regarding the ontology of a work of jazz.  One question raised is how a piece of jazz music could be ontologically similar to another if they are both improvised in drastically different ways.  I believe that in jazz there is a great deal of emphasis on the melodic material.  In standard jazz performance practice it is expected to play the melody of the "standard" version of the song before delving into improvisatory phrases.  Rhythm, timbre, instrumentation of the ensemble, even harmonic material all become flexible when interpreting jazz standards.  If you take a standard and add a 5/4 meter, change it from a ballad to an uptempo swing, add 2 guitars instead of a saxaphone, and throw in some crunchy tritone substitutions, and many inexperienced listeners would have a very hard time believing that the same piece is being played.  Many times, modern interpretations of jazz standards are unrecognizable due to these loose standards of compatibility.  However, as long as the melodic material of the original song is present (and as long as the intention of the performer is to perform the same piece), it will generally be considered the same song.  This is why a contrafact can exist: the melodic material changes over identical harmonic material, which, to jazz aficionados, completely changes the nature of the song.

On a side note, I would also like to ask Farris what he considers "good jazz", as jazz is a genre with very flexible standards. "Swing and a hot manner of playing" is only pertinent to a very small sub-category of jazz, a small slice of a bigger picture.  I have a hard time understanding how a piece of jazz can fail on a functional level the way that a chair can fail to support a human being.  Additionally, if someone asked me to sit in a chair and it broke, I would call it a bad chair, not a "non-chair".  If I bought a car and it broke down in the lot, I would be furious at the poor quality of the vehicle, but I would never go back to the dealership and say, "this is not a car."

What I found particularly interesting about Day’s pieces on jazz improvisation is the idea that improvisation is a way of proposing a conversation to both the other musicians and to the listener himself. The instantaneous nature of the improvised piece further allows for the capturing of an “inimitable rush of spirit” that would otherwise be likely lost in the traditional composing process. Given the unchecked personal communicative nature of the improvised Jazz solo it would seem that the improvised jazz solo is a musical work more strongly identified with the musician/creator than any other musical work. In the case of other rock, pop, classical music etc, covers, remixes and repeat performances are considered fully acceptable and in fact often readily invited. The improvised jazz solo however is something that seems to belong solely to musician himself and thus meant only to be played by him, (As evidenced by Thelonius Monk’s instructions, “Don’t do that… I’m the piano player, you you’re your part, I’m accompanying you. Don’t pick up on my things”), and in fact probably only once by him. Going back to Moore’s essay, this would thus seem to be a case where the provenance of the work is as important to its identification as the musical structure. An improvised jazz solo played Thelonius Monk when re-performed by some other musician at a different time certainly loses much of it’s communicable and expressive power. The meaning and aesthetic beauty of an improvised jazz solo is thus inextricably linked to the spirit and mind of the improviser himself and consequently the improvised jazz solo’s identity is dependent upon the improvisers identity.

I do not have the musical background to know exactly what Spencer was talking about in the first part of his response in relation to sonatas, but his idea of "continuing a story" seems right to me. Day says near the end of his article that the jazz solo is "a way of continually proposing a conversation," and I think this is how we think of it intuitively. An improvised part has a lot of the musician's identity in it, so when different musicians play off of each other, it is very much like a conversation.

I do think Joey brings up a great point about musical works as well, though we might need to take the example of the two performers in his second to last paragraph a little further. I think many people would be willing to say that the two performers are playing the same song even if there is heavy improvisation. Granted, the song in this case would be very thin, allowing for such heavy improvisation, but it would be the same "song" still. There is the question of "How thin is too thin?"which I think Joey is spot on to bring up, but I do think one of the things we want to preserve in our intuitions is that two different performances of one song can exist. The improvised sections do provide a unique aspect to the songs, which provides an intentionally different sound structure, but I think most of us want to say that there is enough similar about the performances that we can call them the same song. Otherwise, I fear we might be forced to go down the road of calling every performance of anything, improvised or not, a different song every time. This would lead you to say that if you make a mistake while performing a song, you've just performed a different song. There is, of course, in the case of improvisation, the intent to create something different, but the improvisation is still in the context of the "same song," so it could be said they still intend to play the same song.

To preface, I don't really "get" jazz, so I am offering my comments with a heap of salt.

I, like Joey suggests, have the intuition that two "takes" of an improvisation on some song are substantially different. Although, from what I know, jazz musicians would like to call them the same song. To me, it seems more like the improvising musician quotes the "song" in his improvisation on it. He takes bits and pieces, but writes a new song around it. But then it also seems odd to say "writes a ... song" in this context. If I am right in thinking that the improvising musician plays a new song, then there was no pre-existing work... but, usually, when we think of "writing" a song, we think of planning, and, most importantly, the intention for it to be played by others. But this may not be an obstacle at all, but rather a flaw in our concept of "composition".

Now, it's at least intuitiveness to think that the work and the performance could come about at the same time, rather than the work pre-existing the performance. So, even if they are works, is their first appearance a "performance"? The ones that come after (playing the improvisation from a transcription, etc.) definitely are performances, but the calling the first sounding a performance seems shaky, as the thing being performed does not yet exist.

I wanted to respond to Day's article "the Ends of Improvisation," and his intital, then sustained, linking and contrasting of an improvised jazz solo with an ending of a concerto.  First, I agree with Day that there is a certain sense of intentionality about the endings of jazz solos, and that the improvisers are generally able to give their solos a sense of distinct, formulaic ending- if that were their intention.  However, I think his discussion might make a bit more sense, and the lack of an ending to a jazz solo need less defending, if held in the context of, say, the development section of a sonata as opposed to the end of a concerto.  The development is a section in the sonata where the composer often ventures into different keys, fragments and modifies themes from the music before it, and eventually transitions back into the last movement, whose responsibility is to end the piece.  

The statement "When you solo, you tell a story," from what I can discern, is partly true.  However, it might be modified a bit by saying "When you solo, oftentimes you continue a story."  I think the major difference between an improvised jazz solo and a concerto ending, besides the obvious improvisatory aspect, is that the solo (or at least some of the solos that Day references) are not actually ending the song- they are continuing them.  Like a development, they bridge a span of music between events, whether from one solo to another, or from the standard tune back into the standard tune.  Taken in this sense, it is less essential for the solo to complete a story, but rather for it to continue a story.  Or, in other words, it less important for a solo to have a ucharacteristic sense of ending than it is for it to provide contrast and color for the piece while transitioning between different parts of the song. 

Having spent a few weeks now on the metaphysics of music, we have struggled to find any satisfactory identification of "musical works." With its complicated history and dependence upon improvisation, jazz further complicates the questions of musical individuation raised by Prof Moore's paper.

Brown raises the possibility that the ontology of jazz might be consistent with our general picture of musical ontology. Hence, jazz performances can be instances of a certain sound structure. This sound structure might be dictated by something like a canonic form. However, Brown finds a problem when recognizing that there are jazz tunes with similar harmonic structures yet a unique title (contrafacts). Also, with improvisation, perhaps something like a canonic form is too "thin" for its performances to be instances of it. The problem is, how do we draw a line between "thin" musical works, and a sound structure that is too "thin" for all performances to be one instance of the same work. Surely, there is something in common between the performances of "Sweet Georgia Brown" (as Samia discusses), but does our intuition really believe they are the same song? Don't improvised performance's provide something intuitively new and unique?

Prof Moore's paper also states that provenance factors in individuates musical works, and this is made clear by Day's article. Day claims that a jazz improviser draws their originality from three different types of histories. As Day says, "Improvisation always occurs in some context" (101). Provenance seems to be even more important in jazz that in other musical genres.

However, I wonder if our intuition about musical works as described in Prof Moore's paper is incomplete with reference to jazz. I might need a ridiculous hypothetical to make this point, but that seems consistent with Prof Moore's paper. Consider two different worlds where the same performer comes  on stage. This performer has the same provenance and performs the same canonic form in both cases, yet will be doing a serious amount of improvising. Doesn't our intuition say these might be substantially different songs?

It seems a nice coincidence that Wittgenstein would be mentioned in conjunction with a discussion about the difficulties in nailing down the essence of a word. Day mentions different parts of the Investigations but I think it is clear that Brown shows that "jazz" has many of the same problems as Wittgenstein's "games."

In answer to the question, “what is jazz?”, Holdeir describes its essence: "an inseparable but extremely variable mixture of relaxation and tension deployed at the same moment;" i.e. swing and the hot manner of playing. Lee Brown objects that Holdeir’s view conflates the question, “Is this jazz?” with the question, “Is this good jazz?” Brown’s objection, however, assumes wrongly that those two questions are not deeply related.

Ask yourself: how does a man know whether he is looking at a chair or something like a chair but not quite a chair? It seems that in his mind are a set of conditions an object must meet to be a chair. Perhaps the condition is simply: the object supports one sitting human being. Now, imagine someone presents you with a wooden object and tells you to sit in it. As you attempt to sit in the wooden object, it breaks apart and you hit the floor, perturbed. I presume you would say that object, though it may have resembled a chair, either was not a chair, or was a very very bad chair. Here we notice then, that the conditions that make an object a very bad chair are the conditions that make it also not really a chair, in a fully robust sense of the term. In other words, if we take a good entity or activity to be one that manifests its essence or characteristic activity successfully, then there is a close relationship between a good instance of a kind and a recognizable instance of a kind. As a person plays jazz with increasing deficiency, what they are playing becomes less recognizably jazz. The success conditions of an activity involve its essential conditions.

Note: extremely rambly. sorry.

Brown's article -- "Jazz" -- raises a problem for Day's article, "Knowing as Instancing." In writing about the intertwined histories of jazz music and sound recordings, Brown writes about the overlooked distinction between live and recorded improvisation: "With live improvised music one is... responding to the performer's on-the-spot choices and actions that generate these details [of the unfolding music]." But with a recording, "one soon learns to anticipate precisely how a stretch of familiar recorded improvised music is going to go at any given point." I think Brown is correct to point out the phenomenological difference between listening to live improvisation and recorded improvisation. But I also think that this distinction marks the two types of improvisation as ontologically different in a way that goes beyond Kania's distinction between tracks and live performances. If part of the essence of improvisation, as all the articles seem to agree, is spontaneity, then a recording of improvisation necessarily precludes that spontaneity. Although Brown focuses on the experience of the listener who has come to expect certain aspects of the recorded improvisation after listening several times, I think that the distinction exists regardless of whether or not the recording is heard at all. All that is required for the recorded improvisation to not be spontaneous is that the first and original instantiation of the improvisation has already existed. Moreover, the improvising artist has already "done" the improvising, such that there exists at least one person - and not necessarily the listener - who knows how the improvisation will go.

That got a bit tangential -- sorry. I want to bring that back to the beginning of Day's article, where he lays out three conditions for the tension and resolution of improvised music: familiarity with jazz history and tradition; history of specific tunes; and development of a performer's musical style. Day describes the second condition in the following way: "the tunes that typically serve as the vehicle for a performance will themselves often have a history of definitive performances" (101). But I'm not sure the "history of definitive performances" exists outside of the improvisatory tradition. It seems like Day is saying that before a tune can be improvised, it first has to exist in multiple definitive states. It seems like this can't be the case for some tunes which are composed as improvisations. If it were, then Day would have to allow for a "definitive" improvisation, which seems contrary to the spirit of improvisation.
Another problem with the "history of definitive performances" relates to Day's distinction between recorded and live improvisations. In his "history," Day would probably include several different performances of the same tune (with or without improvisation), and it is his choice of "definitive performances" that I find troubling. To stick to the example of a piece composed as an improvisation, one problem would be the (necessary) variations in each performance that makes up the history of definitive performances. The only way I can think of avoiding this problem -- except allowing for different improvisatory performances to count as "definitive performances" (which runs into the problem of basing knowledge of improvisation/ ability to improvise on an improvisation) -- is to refer to the same performance as creating a history of definitive performances. This sounds unintuitive, but I think it is accurate. I have the Benny Goodman Quartet's version of Sweet Georgia Brown on my itunes, even though it was written in the 1920s by Ben Bernie and Maceo Pinkard (?) and performed by countless other musicians. As a result, the "Sweet Georgia Brown" I imagine is not the jazz standard, but Benny Goodman's recording. And perhaps Day would say to this that I am not *really* familiar with the tune, even though I have listened to it (let's say) 100 times. The problem with the recording picking out the tune here is that improvisatory jazz tunes are meant to be performed differently each time, so a recording not only does injustice to the improvisation, but to the tune itself.

Maybe it's not that Brown presents a problem to Day's theory, but that both authors pick out the same problem in different ways. Because a recording of an improvisation does not contain a crucial element of spontaneity (in the way that a recording of a pop song does contain the crucial elements of said pop song), we cannot treat them as similar. Day requires that a jazz innovator be familiar with a tune before being able to improvise (another problem for pieces that are created using improvisation: with what tune should the innovator be familiar if she is improvising a whole new song?), but this familiarity cannot come through listening to the same recording of an improvised tune.

(samia's response)

Note: I wrote this before Professor Moore posted the required readings, and it turns out- the reading isn't required after all... Still, I think it applies to what we'll be talking about tomorrow, so I'll leave it as is.

I've found these readings on Jazz really interesting, because they force limiting cases that in many ways break down traditional notions of musical ontology. I was particularly struck by the argument given by Young and Matheson on pg 197, in which they discuss the startling fact that according to their ontological system, two pieces can be played identically yet still be ontologically distinct. It's a claim that I've held in the past in reference to historical differences (take the Hawaiian Hammer, for instance), but I'm not confident that I want to hold it for Jazz, which makes me want to question exactly what it is that specifically defining ontology is supposed to do. In the case of Hawaiian Hammer, it's clear that the two composers are living in drastically different cultural worlds, and so if I walk into a room in which someone is playing a piece that I know is either the Hammerklavier or the Hawaiian Hammer, then I'll hear the song differently once someone tells me which one it is that's being performed, listening for the different styles which nevertheless ended up creating the same sound structure. However,the case of Jazz doesn't quite parallel that classical story, because Jazz is such a contained musical movement (or at least, for the case of what Lee Brown calls "Classical Jazz"), existing for a short period that is now essentially over, as I understand it (note: I don't know anything about modern Jazz and how it differs from this "classical jazz", so this may be flawed).

In other words, it would be like one of Beethoven's contemporaries, two months after Beethoven wrote the Hammerklavier, writing a piece that is identical to the Hammerklavier. This seems almost impossible, because of course, Beethoven's contemporaries would have heard the Hammerklavier played, so they couldn't possibly write it; but it is much more conceivable a thought experiment in Jazz, because the works are so thin, and every performance so divergent. In other words, if two jazz musicians wrote essentially identical, but nevertheless very thin, works at about the same time, and there existed recordings of each being played that involved identical sound structures, would they then still be different works? My intuition is fuzzy, but I think in the end, it depends on what ontology is meant to clarify. If the question is, will the two sound to  a well-informed listener as different works, if they knew that the works were written independently, I want to say no, I don't think they would: everything a person is listening for in the first is, well, the same as everything he is listening for in the second, because they are two works that came out of the same exact cultural background (it would be similar to the example of "Round Midnight" given in the reading, only Twin Earth has exactly the same cultural history as Earth does, so that they're effectively identical- aren't the set of instructions, then, the same?). They are, after all, identical manifestations of the works, if not the same works- the improvisations are cultural as well as the works, and everywhere the culture is the same. On the other hand, if ontology has nothing to do with the perceiver at all, but only involves the actions of the performers, who happen to be looking off of two "different" scores, then I guess the authors view that they are distinct is inevitable, but I'm not very comfortable with that conclusion.

The readings for Tuesday are:

"Jazz" and "Improvisation" by Lee Brown (these were handed out in class)

"Knowing as Instancing" (in the reader) and "The Ends of Improvisation" (handed out) by William Day

Please read all four. The first two are really overview articles. My hope is that we'll spend some time on Day's claims about the values of improvisation. The initial sections of "Knowing as Instancing" are difficult. Day's presentation refers in a compressed an complicated way to certain ideas of Kant and of Emerson. Please push through these, get as much as you can, so that you can see how his views apply to the pieces by Lenny Tristano and Charlie Parker. 

I have posted (in separate posts, unfortunately) most of the recordings that Day discusses in his two articles.