-
So, if you've decided to work on a bunch of ideas, lets say several variations of an idea in different positions and different keys etc, do you find you work on one "chunk" at a time, or a bit of each?
By serial learning I mean mastering one idea before moving on to the next and not refreshing until all the chunks have been mastered first as opposed to parallel learning which would mean working a little on each chunk without mastering any and continuing to revisit each unmastered chunk until they all improve enough to move on.
And yes, I know the answer from many of you will be "both", but I'm assuming if we're honest we favour one approach a little more than the other. I know that for me (I think) I progress faster with the serial approach, although I can take it too far where If I leave it too long to refresh something I thought I had under control weeks earlier, then I need much remedial work to bring it back.
What about you guys?
-
08-23-2018 10:33 AM
-
Depends. Quite a lot of ideas are not worth to master every possible way. And also, there are so many variations/ideas/ways to play, that it takes million years to master them all. And probably I'd still choke on stage when going to show them in public
Some stuff I just plough through - all positions and keys without worrying too much about actually ever using them like that. With a few ideas I can sit for 10-20 mins and get a single thing (one key, one pos) groovy, for some exercises I reserve a day occasionally and not even "master" anything.. Probably have wasted a lot of time like this but it's just a hobby for me. So what
-
When It comes to learning lines, I take in some material, play it, and the let it go and move on. Start on something else. Rinse and repeat. The ideas I needed from the material will be there and I can always revisit things I really found I liked, later.
-
Thanks for your take. Anyone else?
-
That's about it for me too.
Originally Posted by DS71
I used to obsess a bit about playing fast but found that after a certain amount of serial attention my execution would get clumsier. So I started moving onto different things quicker and not worrying so much about beating my one day younger self.
What do you know, things stopped being clumsy. We should always be careful not to bore our subconscious rigid, it WILL sulk and you don't want that because that's where all the real power is.
D.
-
Yep Spent years on working speed and developed a lot of technique. It serves me well. But these days I rarely think about it. Usually much more interested in feel than speed these days. Originally Posted by Freel
-
First of all...*
Second, all the old advice on this thing is the same, usually put in a form like this:
Better to learn many to an incomplete medium level than a few to complete perfection (songs, techniques, scales, fingerings, chord types, etc...)
So without all the fine print about distinguishing definitions of serial/parallel, it is "parallel" learning that you want to gain because:
- the bits that you learn interact and compound (you learn to integrate things despite their being incomplete)
- you learn to play more effectively with incomplete information (as when performing with others)
- you learn to switch and swap quickly among ideas in the moment (as when evaluating which ideas to select while improvising)
- your knowledge and technical skills are more general (you learn to recognize more contexts for application)
- your musical grasp is further (like a net - not a continuous manifold (it's full of empty spaces!), but it successfully catches things as far as you can cast it)
* I would suggest that the serial/parallel learning idea is the wrong perspective; it's all serial - the only difference is the period of time between switching to something else. If you work on item 3 until completion, then move to item 4 and complete that, then go back to item 2 that you skipped yesterday and complete that, you may call that serial. But if you don't complete item 3 before moving to work on item 4, which you also don't complete before going back to work some on item 2, all those are serial, too. If you switch around among the items, working on each for a little while, you are not parallel multitasking; you are just switching among serial tasks faster. Completion of a task before moving on to the next one is not what determines serial/parallel... the attention of consciousness does not do parallel - it only looks or feels like it if the switching among tasks is fast enough.
-
Some good points, although when you say the conventional wisdom is: "Better to learn many to an incomplete medium level than a few to complete perfection (songs, techniques, scales, fingerings, chord types, etc...)" - I'd say that will depend on who you ask as some will say the opposite, i.e, better to learn a few things "really well" than lot of things half baked... Originally Posted by pauln
But yeah, I hear you on the main gist of this take, that we need to also practice integrating all the parts even in their unfinished state. This is important, perhaps, because practicing integrating stuff can be considered as important as just learning the "stuff" itself in isolation.
This, however, has not been my path, and so it's taking longer to integrate all aspects of improv, but at least the "stuff" I have learned feels solid and almost automatic. Hopefully we all get there in the end, regardless of what order we learn things. Would I arrive sooner with another approach? Maybe, but the serial way seems a better fit for me. Diff'rent strokes an all that...Last edited by princeplanet; 08-26-2018 at 07:23 AM.
-
To me it's fishing with a spear vs fishing with a net. It may be personally rewarding to master the precision and focus of spearing fish one at a time, but the trusty net is going to put a lot more fish on the grill... two kinds of appetite.
When I want to investigate, examine, analyze, and internalize a specific element, issue, problem, etc., I take my spear and use my focus to capture the specific target. For gaining ground on general progress, I throw a diffuse wide net.
-
Anyone else out there prefer the "serial" approach? (for want of a better term...)
-
I certainly do. I don't know that it's necessarily better . I believe it is for some things for sure. Originally Posted by princeplanet
-
Is it better for some things? Or just better for some people? I know I find it hard to cease working on something until I have it licked, so that means one manageable task at a time. This is either how I came into the world hard wired, or how I became after many years of (stupidly) obsessing over scales and arps. Or maybe a bit of both... Originally Posted by matt.guitarteacher
-
Yeah. I think so. Blessing and a curse. I have to admit that I never understood the idea of "maintenance", the way some people talk about it. Once I've spent enough time with something, there's a level of retention that isn't going to be lost. Originally Posted by princeplanet
Of course, there's also a lot of stuff I've never covered much at all.
-
The main distinction I see between the serial/parallel approach to learning a tune is the applicability of what you have afterwards. If the serial approach finds the first chord, then the second, etc., when you find the last chord what you have is one way of playing it. How likely are you to explore other ways to play it if your plan and methodology was to learn to play it "right" from beginning to end?
The parallel method is more like a series of progressive approaches to the song that results in multiple ways you know to play it. For example, say you set out to learn "Just Friends"...
You might start by hacking through it using just simple chords like this:
Cmaj7 | Cm7 | Bm7 | D#(7) | Am7 - D7 | Bm7 - Em7 | A7 | D7 - C#(7)
Cmaj7 | Cm7 | Bm7 | D#(7) | Am7 - D7 | Bm7 - Em7 | Am7 | D7-G#(7) - Gmaj7 | D7 - C#(7)
Then changing some of the chords more toward the sound of the song:
C69 | Cm7 - Cm6 | Bm7 | D#(9) | Am7 - D9 | Bm9 - E7sus4 | A13 | D11 - C#(9b5)
C69 | Cm7 - Cm6 | Bm7 | D#(9) | Am7 - D9 | Bm9 - E7sus4 | Am7 | D9-G#(7) - G69 | D11 - C#(9b5)
Then fine tuning the chords some more:
C69 | Cm7 - Cm6 | Gmaj7 | D#(13) | Am9 - D9 | Bm9 - E7sus4 | A13 | D11 - C#(7#11sus2)
C69 | Cm7 - Cm6 | Gmaj7 | D#(13) | Am9 - D9 | Bm9 - E7sus4 | Am7 | D9-G#(7#11sus2) - G69 | D11 - C#(7#11sus2)
Then exploring some more movements and additional changes:
C69 | Cm9->C9sus4 - F13/A | Gmaj7/B | D#(13)/A# | Am9->A9sus4 - D13/F# | Bm9 - B9sus4 | A13/G | D11 - C#(7#11sus2)
C69 | Cm9->C9sus4 - F13/A | Gmaj7/B | D#(13)/A# | Am9->A9sus4 - D13/F# | Bm9 - B9sus4 | A7#9sus4 | D13-G#(7#11sus2) - G69 | D11 - C#(7#11sus2)
The value of the difference between knowing one way and knowing a few ways is application flexibility in performance. If I am accompanying a soloist that is playing simple "vanilla" lines, I can help his stuff with harmonic support by making his lines sound much more interesting over the third or fourth version of the chords. On the other hand, if the soloist is sophisticated and needs lots of harmonic freedom, I can down shift to a reduced version, which might be a mixture of changes among the four versions. Likewise, depending on the bassist, I can play the non-root chord bottoms or not. Same with accompanying singers; some love to sing over the lush chords with deeper harmonies, others like simpler accompaniment.
Another difference if that the serial strategy may have you waiting a while before your work has reached the end of the song and then the result is "ready for prime time" whereas the parallel strategy provides playable versions of the whole song early in its development. Additionally, when it comes time to solo, the different development versions helps explore ideas.
-
I am with Pauln absolutely on that. I would rather work out a three chord version of a tune and then work towards what I hear on the recordings I admire. That way I'll find out what I like and maybe a little about how it was done on the way.
I think we want to build heuristics and unconsciously. There is a famous experiment ( like a lot of things that aren't easily sold hard to find on google recently) that measured the performance of two groups of children throwing bean bags into a box. One group were given a fixed target and a fixed distance, the other was encouraged to have a variety of both.
After their preparation both teams were measured on the standard of the fixed target and distance. You might expect that the team who had solely been specialised in their preparation to do best at the specialised task. The results were emphatically the opposite.
It's always better to PLAY.
If I find something in a piece that I can't easily or reliably do then I will work out ways of 'taking it for a walk', different string sets, keys, play it on my mandolin if it is in single notes, ANYTHING but the serial approach.
If I am in a hurry and want to solve it in one key with one fingering then I will vary the rhythm, I have system to do that but still fun and the system has enough room to make choices and have variety.
The principle advantage that I've found since adopting this approach is that the tough spots no longer fill me with dread as they join onto the rest of the performance more organically and haven't been narrowly overworked to the point of jinxing.
Having said all that I find it perfectly reasonable that Matt and Princeplanet and many other people reading may have a lot of interesting variety to their 'serial' approach that was absent in mine. I come from a classical guitar background and it took a while to learn that the approach to practicing classical pieces traditionally taught can an encumbrance to the pursuit of freedom.
The best idea is probably to cast our nets wide, then draw them in at our leisure, never rush.
D.
-
Of course! I learned classical gtr as well early on, and because I had to cram to get to 8th grade to get me into Uni, I definitely learned pieces one bar at a time... That habit was carried on to the way I learned songs when I played in Rock bands, so it's perhaps no wonder that it feels comfortable for me to be still learning this way. Originally Posted by Freel
-
Cheers for the thoughtful and interesting post (cool Just Friends variations!). I guess the serial/parallel thing for me is not just about learning tunes, in fact it has more to do with the way I acquire devices. For example, I may take a piece of language (say, the bebop "lick", the Honeysuckle motif, or a Pat Martino line etc) and create a long line of continuous 8th notes by extending the idea in all 5 positions so that all the available notes (from lowest available to highest available) in a given position are used and the "line" fits evenly into either one or 2 bars. I have a dozen devices I've developed for every main chord quality (Maj, Min, Dom, Alt Dom etc) in both directions and in 5 positions. So 12 x 2 x 5 = 120 things I practice, and just in one key! Naturally I run things through other keys, but it's really the same things in shifted positions, so it's still down to only 120 "chunks". Originally Posted by pauln
The way to get the organic variations is to be able to start the idea from any string or note, on any part of the bar and to be able to insert any length of rests at will. This, for me, takes care of what I call my prepared "filler" stuff as opposed to the other way less contrived stuff which is the 100% improvised "melodies" where I play more slowly and just following where my ear takes me. So kind of like 50% Coltrane and 50% Miles!
(haha, well that's the idea anyhow...).
Anyway, the serial bit for me comes down to the reps it takes for me to get each device flying off the fingers as easy as speaking, so I can then weave in the other organic (musical!) variations blended with the pure "melody-on-the-fly" stuff. So chops first, then make music with them. This doesn't mean 6 months of 8th note etudes followed by 6 months of actual improv practice. It's more like 2 weeks chops building and 2 or 3 days actual improv.
Not how I'd like it, I'd like a week of each, or ideally swapping every half hour until I feel like I can just improv all the time and not need any more new chops. Nice plan, but I'm never happy enough with the control of the chops dept. They're tricky and uptempo and a real workout for the right hand (not very guitaristic, unfortunately). If I forced myself to do a week of each (about as "parallel" as I can get!), then the prelearned stuff always feels under ready, so it's back to the woodshed. The players I like have their chops down, I'm not really into the slow and searching style alone, I like to contrast it with the quicksilver stuff.
Not saying that my serial approach is the only way to get what I'm after, just that for my specific goals I can't see myself having any other approach. I need to repeat things hundreds of times, if not thousands. I wish it didn't take as long, and for some lucky people I'm sure it doesn't (bastards!). It would be amazing to be able to start incorporating tricky stuff into 100% improv forays after just a few run throughs, and I'm happy to concede that parallel learners are probably more talented or cleverer than I will ever be. It's a grind, the way I'm doing it, that's the truth - but I can hack it, I'm a patient (if not efficient) bugger...
-
Yeah I wasn't talking about tunes.
I tend to compartmentalize in different ways. For me, there's "skills I'm working on" , and then there's "tunes". You use one to reinforce the other.
In classical teaching they kind of separate things into categories like technique, theory and repertory.
Anyway, if you're working on tunes , the tune IS the object itself - different ways of getting through it etc... more about the whole then the parts. But that's what technical work is FOR, to me anyway: it's just giving myself permission to work on something separately, maybe even outside of the context of "real music" for a time, if I need that focus.
The degree to which a musician separates working tunes versus technical is largely personal in my experience. I mean, I've actually met people who want to play guitar but aren't really interested in real MUSIC -like songs/tunes/pieces - at all. I really can't teach to that. My mind doesn't have a context for even understanding that way of thinking. I guess I can respect it, but someone else can do that.
-
great topic. i’m from classical background, so serial is more natural. i like to work in say 4 bars and cone up with 4 or 5 ways to play it (solo guitar context) before i move onto next part. i feel like i am creating, constructing. as if i had to just run through the tune over and over I would just feel like im faking it until i come up with some ideas.
for lines and licks same way because once i get it in subconcious it always pops up naturally in my actual playing.
of course the answer is always both. i don’t think serial is better i just lean that way
-
I've talked about learning stuff before trying it in an improv setting, but I'd like to be more specific about how I even commit stuff to memory.... and actually my main reason for putting this thread up was to bring that up specifically. To refer back to the way I will learn , say, the above mentioned 120 chunks, I remind you that they are in categories:
- several ideas for a specific chord type, up to 32 notes per chunk.
- several ideas for the other chord types
- several ideas for each chord types in the other direction (ascending instead of descending)
- all the above in 5 different positions for each key
... then to make it harder, practicing starting each device in different ways (note or beat etc)
- quite separately, I may devote a few consecutive hours every once in a while to "free wheeling", just off the cuff making up melodies without thinking of what the notes are.
- finally, putting it all together against a tune, the melodic playing linked together with prelearned language.
Then I'll listen to Sonny Rollins, or Wayne Shorter, or Joe Henderson etc ... sigh, and resolve to hit the shed again even harder...
Yeah, it's a very narrow focus, but it's what I want to do and gradually I am narrowing the divide between what I hear in my head and what comes out through the hands. Technically, and perhaps harmonically it's reasonably advanced, but it's horse before the cart because I am far from being a complete player that could easily improvise against anything on the spot. So I'll find myself compartmentalising one specific challenge at a time until I can do it eyes closed while thinking about something else. Only then will I move on to the same idea in a different position (same key). Because the stuff can get very stretchy and chromatic and requires careful attention from the right hand (each idea needs different combos of alt picking, mini sweeping etc), there's no way I can make progress moving from one idea to another without nailing each one first without any insecurity or hesitation. If I did a little here and a little there, I find my right hand forgets the required nuances for each idea. For example, some things are only playable at tempo if I do non intuitive things like play an upstroke stroke on a downbeat, or left hand shifting where there has to be an awkward finger cross or something. If I don't "burn in" these moves until they feel automatic, then I mess them up in a different way each time. If the stuff was simpler or more diatonic, or easier to address with alt picking throughout, then that would be different, but the lines I tend to like just don't lay that easily on the guitar, so I can't do it that way.
Further, I can personally vouch for the 21 day rule, and I tend to run with a small number of ideas and burn them in for 3 weeks before moving to something else. I pretty much know what I'll be practicing for the next few years, but I'd like to focus more on fully improvising against full progressions because obviously that's what it's all about! But if I do too much of it, I worry I'm developing habits and falling on things I can do, not the things I hear, or even the things I commit to muscle memory. It's incredible how long it takes for the prelearned stuff to come out in a way that sounds uncontrived, or not too forced.
So I wonder if any of you guys can relate to any of this? I suppose on a simpler level we all had to learn our inversions, arps and scales in different positions and keys, right? Were you taught to do a little of each as you were learning, or did you do it in serial chunks? Beyond that, when learning to integrate lines, patterns, devices, language, licks etc into your general improv, did you find a way to strike a good balance between time spent freewheeling vs peppering in your pet lines? Do you feel that time spent really just improvising is more useful than developing an arsenal of prelearned language? How would Charlie Parker honestly answer these questions?
-
I think I might get back to parallel learning because I read a few studies a long while ago that claim that doing that sort of thing will help benefit you more in the longer run despite not feeling that much improvement in the short term. I think it was called interleaved practicing? Might be a good idea to work on 3 new tunes at a time, 6 minutes each continuously switching back and forth between them... !
-
Now that some of the serial people have explained more detail of their process I am seeing something interesting.
Apart from parallel being I think a more natural process for ear players (where you progressively "hear" your way to the song, a layer at a time), I have always thought that the parallel process had a particular advantage - that you came to know the whole thing, knowing the most important parts the best.
Imagine listening to a solo that you wanted to transcribe. You already know its peaks of expression, so you learn the peaks first and work your way "down" as if each layer is an elevation on a topographical map... you learn the whole song for each elevation.
Now imagine learning a solo one bar at a time... playing through the first bar until you have it, then playing through the first bar into the next until you have that and have it integrated with what came before. By the time you have learned the whole thing, how many times have you played the front end of the solo, and what is your confidence level? Compare that to the back end of the solo.
What I used to think was that the parallel approach lends itself to automatically concentrating the most confidence at the most important peaks (which is true, and is a great way to feel when performing) and that the serial approach would tend to create a profile of confidence that progressively diminished through the solo (because each subsequent section had been played less)...
But what I notice is that the serial posters are detailing much longer periods of working up the sections than I thought (three weeks!?). What I mean is that the assumed difference above is going to wash out as the number of play-throughs increases; I just had no idea of the time and number of cycles.
We call hearing external music and figuring it out "transcribing", and hearing internal music in one's head and figuring it out "composing"... to me these feel identical, use the same fast internal processes.
-
Yup. Jazz musicians, or aspiring ones like me, are real lucky we don't need fancy 'memory palace' nonsense to learn tunes. The tune IS the memory palace. Everything can be bolted onto it and given sense and made memorable through it. That is if we CHOOSE, which means working by ear. Originally Posted by pauln
Reading a chart and memorising it is really difficult until you start memorising and internalisng using TUNES, not concepts read in a book and not counting bars.
I get a lot of people, and sadly not just beginners, who are counting bars to REMEMBER where the chord changes are in simple pop tunes. Someone suggested it in a video on youtube..... As a result they are not listening to the tune and are LOST all the time.
We work on a tune and I urge them to stop doing this and listen to the tune and change when the tune does, they instantly improve. Then the sad thing happens, they come back and they are counting bars again. I would burn most books and gag most teachers.
Still, consumerism is the new religion. Why learn how to play a tune when you can buy another course in ii V Is, and why learn to get the most out of a guitar when you can put another one on the wall, or your virtual wall. Why write a lick or transcribe it by ear when you can have someone finger it for you so you won't learn anything about how to listen or how to play. Why learn to improvise when you can run patterns. Why sing over chord substitutions to activate your musicality when you can read a book about them, and another and another. So maybe the vast majority of practice time being wasted is a good thing, maybe we need legions of illiterates to keep Mel Bay churning out four volume tab heavy editions of books that used to be available in a useful and concise single form when the author was still alive. Robots and investors get happy and people have an excuse to work more hours for another corporation to get the money to buy more toys that they can't play and all the other stuff they need a bigger mortgage for the bigger house it sits in.
All griping aside though, most serial guys are outlining a fairly varied approach, we are dealing with loose definitions here.
D.
-
Yes, well I thought I should go into more detail in my last couple of posts given that there are many different ways to compare serial to parallel. I think I prefer the term jazznylon mentions "interleaved" learning, and I can see how that would benefit learning tunes, transcribing solos, or even learning fundamentals (scales arps etc). But for me, the way I'm trying to build a body of language is much more difficult than just learning a few dozen scales or arps. It just takes me longer for it to become "automatic".
I listen to horn and piano players mostly and I hear a ton of re used language in my fave players. I have to wonder how they integrated their pet ideas into their solos. Like how much is off the cuff and how much is cocked prefab devices ready to be unleashed in many varied ways. Bird is a good example here because he has been analysed in many different ways. The Owens dissertation finds 240 odd pet phrases, and even shows what they are and how they get cleverly disguised. But having such an arsenal of lines locked and loaded is certainly no guarantee that they will come out the way they do when the greats play. And the Clifford practice tape - we hear all his devices condensed into a solid half hour or so of non stop lines. No doubt these guys put a lot of time into "burning in" their stuff.
But the real mystery is always, how the heck did they practice improvising? Obviously they also spent much time on the over arching gestalt, the story telling, the phrasing, the motivic development etc etc. To be that masterly, does one need to be practicing "story telling" from the get go? Before one has his bag full o' licks? Play the best melodies you can, even as a beginner, and gradually work in your pet devices is a disguised way? Maybe this is the better approach, or even the only approach, and possibly easier on a horn or piano than guitar because we have to synchronise our pick hand which might slow us down, and obsess over technique at the expense of story telling?
Many of you won't agree, but I feel that there are hundreds of horn and piano players from all eras who are masterful soloists, seamless weavers of lines and spontaneous melody. Yet really only a handful of guitarists are in this rarified league, Wes being the obvious poster boy. I'm sensing that to play in this way one would had to have learned all things "in parallel", not practicing technical things for 20 years and then improvising for the next 10. But it's pretty darn hard to bring that to our instrument and hence we have either the slow and deliberate melodic players in the Jim Hall mould, or the post Coltrane pattern shredders
... (not that there's anything wrong with either approach...). The guys who had it all (including "modern" cats like Sco and Metheny) are rare, and we should listen very carefully to what they have to say about this discussion, in fact, what have they said?
(* I'll answer my own question - They "said" it all in their recordings. Perhaps, like all great art, somethings should remain a mystery.)
'Nuff said....Last edited by princeplanet; 08-31-2018 at 01:56 AM.
-
Wes was very honest and to the point.
About practicing, he stated that he never practiced anything that he would not play in a song. Think about that - about how much that excludes from the usual guidance of most lesson plans, methods, teachers, and "common knowledge".
About "technical knowledge" or theory, after performing a tune on a talk show, his host asked him the first chord of the song he had just played. Wes said, "I dunno... I just cool."
Similarly, in an old VHS video lesson (now on YouTube), Joe Pass was asked the name of a chord he had just played. He stared at his left hand for a full five seconds before he was able to reply that it was a G13th (clearly he was not using chord names to play).
My experience improvising is that it does not feel like it has any level of effort. I don't have to search and struggle to find, extract, or invent things to play. When I hear music I'm presented with a whole bunch of musical ideas (as sounds) coming from my mind's ear, all of them contending to be played. I just let the best one pass through, so the selection process is just a natural feeling of which of them is most appropriate, effective, and beautiful, and that is dependent on the mood, the "story" (what I have already played), the song, the style, the audience, etc.
The curious part is that I hear them and am testing them for how they will sound with the harmony that is coming up ahead in the music. Seems impossible, but I am comparing different possible sounds at the same time against a future background (so all this is in my mind's ear) all while playing what I chose earlier against what is happening in the present. The more I think about it, the more I realize that the sounds I'm comparing are actually longer in duration than the selection process cycle, yet I hear all the sounds from end to end kind of at the same time against the harmonic background that will be happening in the very near future... Kind of like talking (saying words in the present) and thinking (planning what you are about to say next) at the same time; feels natural and effortless. I'm pretty sure that musical memory does not operate at all the way people think it does.
Reply With Quote
Just Friends -- or keeping things simple
Today, 12:04 AM in Improvisation