This article comes from a Guitar Live challenge on my composition "Printemptation" from the album @-quartet. Initially, one of the participants was asking me how I got to where I am now in music. While answering him, I ended up describing my conception of music, and notably improvisation, as a natural language. It is this idea that I'm transcribing below.

I wrote: That's it. I hope that these considerations will help you getting started... note that when I wrote this piece, I did not think about all this. Music must express itself first, and can be subject to analysis, but afterwards !

Reaction from one of the participants: Now that is some beautiful thought that could very well be boldly advertised on the first page of the forum. But the question is, before letting the music express itself like this, one must reach a considerable level of technicity. It would be interesting to know how someone like you, who have reached that level, achieved this goal. I guess the answer will be "with a lot of work", but would that be a theoretical work, or rather an intuitive approach that sharpens progressively ?

Of course there is a lot of work, but my personal approach is intuitive before anything else, and theoretical only when it needs to be. This has not always been the case though: I started the other way around, but it nearly disgusted me (in fact, I think that if my musical fiber had not been that strong, I would have given up on music altogether at that time).

My parents put me at the conservatory when I was 5. I got two years of music theory without ever touching an instrument. It was like I was going to school on Wednesdays when my buddies were out playing. Deadly for a 5 years old kid. Next, I learned some instruments the classical way, where you practice scales without ever playing a real piece (or at most one every year, for the yearly show). I couldn't stand it either, and I think that's the point where people who don't really have music in their soul give up once and for all. After that, however, I discovered Jazz, improvisation, but also another way of teaching thanks to a guitar teacher. That's how I ended up playing the guitar and stuck to it. I'm passionate for Jazz and improvisation, but a guitar player almost by accident. I could have turned out to be a pianist or a drummer just as well. The instrument is just a means of expression to me. There are people who truely love their instrument, but that is not really my case. I worked with this teacher for 2 years, and have been a self-taught since then, for around 20 years now. My work is mostly intuitive, but be careful: being a self-taught doesn't prevent one from opening a book on theory when the need for it arises...

That was the synopsis. Now the details, both gory and glorious.

Improvisation as a natural language

My personal approach (but I try to apply it when I teach) could be summarized in one sentence: "Improvisation is a natural language", just like French or English. The only difference is that while music stimulates the lower layers of your cerebral cortex (emotions), languages are directed to the abstract layers, but otherwise, that's it. I know this might sound like a cliché, but this idea is actually much more subtle than it seems. Let's just take the example of my son Éliott, 2 years old (at the time of this writing), to illustrate the matter. He is currently right in the middle of the language explosion. What do we notice about him ?

  • First of all, he wants to communicate like crazy. He wants to show us things, objects, colors, everything new to his own eyes. He just can't stop babbling, telling us stuff, all the time. It is absolutely stunning how kids can be gossip at this development phase.
  • Next, he already understands practically everything that we say to him, and this, long before being able to talk by himself. That's actually normal, since he has been hearing us talking since he was born (as a matter of fact, even before that because hearing develops several months before actual birth). Having a 4 years old daughter (at the time of this writing) who's been there already, I'm still puzzled today by the understanding capabilities of the youngest children. Consider that when Éliott was 12 months old, we used to tell him "Hey sweetie, go put your socks into the laundry basket", and him, complied with the request, all proudly, the next second.
  • Finally, he's beginning to pronounce words that he has known for a long time, and there, the least I can say is that he trains himself every single second of his life. His language skills improve gradually, but I can also tell you that when he can't make himself clear, it drives him completely mad.

So what lesson is there to learn from this "experimental study" ? Well, the lesson is that the Éliott subject is a beginner musician, and that the work of a musician goes through 3 steps: semantic work, theoretical work, and practical/technical work. Unfortunately, people have a tendency to forget step 1... Please note that I'm not pretending that those steps are truely sequential (everything happens more or less at the same time), but it's better to describe them in that particular order.

Semantic work

If children didn't feel a vital need for expressing themselves, they would not speak. In the same way, in order to be an improviser worth listening to, you have to feel the need for expressing yourself, which means in particular having interesting things to say. That's the difference between a very good technician that throws ultra-high speed scales at your face but don't touch you otherwise, and the blues player who only knows how to use 5 notes, but overwhelms you with musical emotion. The former really has nothing to say whereas the latter has little vocabulary but a true meaning to express. It's a bit like a boring chatterer on the one hand, and somebody who only says 3 words every once in a while, but have you think about them for hours.

But how do we get to having something to say in music ? The answer is always the same, everywhere: culture. We are seldom geniuses (I mean we seldom have ideas of our own) but we are very often in agreement or disagreement with what we hear from others. Consequently, before pretending to have an interesting opinion on a particular topic, you must know what others have already said, and sort out the things you agree with, or not. In music, you have to listen, listen, and listen again. Personally, I'm a true music sponge. I probably know by heart my whole disc library, improvisations included, from all the musicians, note by note, even the drum solos. But more than that, I know how to sort things out: there are things that I understand or not, there are things I'm not very excited about (whether I understand them or not), and there are things that I'm desperate to be able to express. It is somewhat like when you're listening to somebody and suddenly go "But of course ! That's exactly what I think !" (and there you could add that it is said in a better way). That is the way your musical personality will develop.

I hope that now you understand why it is important that your musical expression conveys meaning, and how it can convey one. If this Éliott kid of mine had spent his 2 years of life in a white room all alone and with nothing around him, and if he knew how to talk, what the hell could he talk about ? Nothing. In order to learn how to talk or play music, you have to discover a world (mine is Jazz; Éliott's is his everyday environment) and to hear talking or playing (for Éliott, it's the people speaking French around him; for me, it's the musicians that I'm listening to).

Now, the problem with musical theory at the age of 5, or the problem with my piano teacher who had me practice scales all day long at the age of 7, is that it's like putting the cart before the horse. Why the hell would I want to know how to play a Eb minor scale ?! I got nothing to do with it. Why the hell would I want to learn how to use a hammer when I don't have any nails to stick ? I for one can only work out of urgency and frustration, when I feel the need for it. Practicing a scale just for practicing a scale, I never could.

Theoretical work

When Éliott discovers something new, he tells us: "there ! there !", and he shows us. Then, we explain to him: "yeah sweetie, that is the red car". During the year, we might need to repeat the same thing a thousand times, but that is actually important: he is consolidating his recent learnings. When this consolidation is satisfactory, he has done a theoretical work consisting in the linkage of a set of percepts (visual, auditory, olfactory etc) with the corresponding words, which are abstract labels. This association constitutes what is called a "concept" in cognitive science.

Suppose I'm listening to Mike Stern. At some point, I spot a lick which I don't understand, but that gives me an instantaneous erection. So I grab my computer, rip the lick in question, time-stretch it a little and work on it at a very low speed. There, I discover for instance that our fellow guitarist had fun going one half-tone below the tune's normal pitch, playing on the major/minor third (that's actually a classic in the blues) and eventually got back in pitch on the fifth. What I've just done here is theorizing by generalizing: it is possible (and amusing) to play "out" by minus one half-tone. It has a particular connotation to it, and you also have to know how to get back on your feet again.

Another example: I have 2 songs in mind, that I don't know why, sound similar in the improvisations ("convey the same meaning" in other words). So I analyze this and figure out that both tunes sound like blues (minor 7th harmony), and that consequently, the improvisers actually use the myxolydian mode. At this point, I "understood" what myxolydian is: I can feel how that sounds, and I know I like the sound of it. But be sure to note the order of events here: first, I was captivated by the sound of "something" (a mode). Next, I figured out which one it was, and now, the lesson is learned. If a teacher had told me "Here's the myxolydian mode; work on it for next time" without me ever having heard any Jazz-Rock for instance, I probably would not have done my homework.

Technical/practical work

Right now, Éliott knows what's a "red car". But that doesn't mean he's able to pronounce those two words. Far from it actually; this will take another year of practice. The same goes in music: your hearing must be much more advanced than your fingering at any time, otherwise, it means that you don't have anything interesting (or new) to work on.

Getting back at the Mike Stern lick again, now that I've got it dissected, I don't necessarily know how to play it yet, let alone at the same speed. So I work slowly, then faster, then faster etc. In the same vein, keeping the minus one half-tone idea in mind, I can try all possible combinations, transitions between the 2 pitches, see if that sounds cool, in order to get familiar with those new fingerings etc. Now for once, that is the dummiest job one could think of, but it is an important one, and above all, I'm doing it with a purpose in mind: I want to play that lick as fast as Mike Stern, and right now, I can't. Or, I want to sound myxolydian at places, but right now, my fingering is too weak.

To sum up, when I say "music has to express itself", I'm referring to the semantic step, which must be the very first one. You need to have something to say, first. When I see teachers going "here are the scales that fit your song; try to play on them", that gets on my nerves (I think I'm gonna make some nu friends here). Personally, I begin with asking the student "sing something for me". If the person has some culture, (s)he will do some acceptable vocal stuff, and in general, will be absolutely incapable of replicating it on the instrument (and there, we have something to work on). Otherwise, (s)he will produce something miserable and we will spend a few hours building a couple of phrases, listening to what other people would play, but all singing (that is why it rains a lot in Paris).

One last thing: theory is only a tool, not a goal. All musical theories (all theories actually) are shaky. The human brain acts as a massive memory, not as a computer processing logical information. Today you speak fluently your mother tongue, but who still remembers all the grammar rules that we learned at school ? When you build a sentence, do you think "so, I need to put the subject first, then the verb, etc", or does the sentence simply come out by itself ? In music, it's the same: modes, scales, sharp fifths and all that stuff are good for homework. However, in the heat of the moment, you must not think. You must let what's well digested and has become natural come out. And let me repeat that: your theoretical knowledge must be more advanced than your technical skills; otherwise, that doesn't cut it.

One last thing (the real last thing this time): I never really worked on a scale in my whole life (maybe I should have ;-). However, I do know my fingering in all positions, all modes etc. Why is that ? Because when working on pieces of actual music (things that you can whistle but wouldn't know how to play), you necessarily end up working on pieces of actual scales (or arpeggios). So the scales end up building in your fingers, piece by piece without you ever noticing it. That is a "bottom-up" approach if you will. In the same way, my daughter knows very well how to put out correct French sentences. However, she never studied French grammar; she's in her first year at school. Grammar is for later...