practical suggestion: memorizing music

in less than a week, univ peeps are facing the annual event: concerto competition.

as i worked with many different people in different levels, the one common difficult that arise to the top (always) is 'memorization.'

i personally feel that the requirement of memorization is old and antiquated, but as it will stay probably for another long while, it would be foolish to talk about its de/merits... sigh.

however, about the memorization process itself, though there isnt a simple way that fits everyone, i do think there are certain principles that helps to memorize music. so here's a small architectural approach.

1. divide your music in large to small chunk:
so all music, like stories, will have sections and sentences. break them down and look to see what their 'forms' are.  if it is an allegro-sonata form, see if there's any extra bits, such as coda(s), or extended cadences in unexpected places; for instance, extended instability in development is expected. however, an extended instability before the coda or return of the counter theme may be a surprise.

look carefully when the structure involves medium-sized cells in quasi-repetition, something like a rondo, where each fragment(s) is recognizeable as their own 'chunk' (if you hum it and you think it feels like a verse, then it's probably a section.

then in any variations of rondo or small binary-composite work, check for regular OR irregular repetition of the chunks.  sometimes, things can be organized as composite of binary and tertiary paris ( AA BCB AA, which breaks our expectation of AA BC or AA BB etc)

when there are differences, figure out what exact difference there may be- for instance, if the 'same' bar was used first time to return to the same place, then it will HAVE TO DO SOMETHING DIFFERENT to go to a different place the second time around. what is it? is it harmonic change? or melodic change? rhythmic? rhythmic phrase change?


2. move away from your instrument and SEE if you have them memorized.
i advocate making a blueprint of the score. here's an example:

that is a super-detailed one, but you get the gist.
get a long sheet of paper and start diving it up structurally so that you can actually compose the time-dependent model of the piece. if it can be scaled to reflect the duration of the phrases, even better.

once youve done it, go back and compare with the actual score. get a different coloured pen and be ruthless. then construct a new improved version.

see if you can practice from it. while you do that, record it. then listen to it with actual score, mark the discrepancy and fix weird stuff.

3. rather than focusing on the 'start' of the new sections, practice the hell out of the tails of sections. we tend to remember what happened in the first presentation of a thing 'A,' but it's the end of thing 'A prime' that will get us to B, for instance. most people practice in sequence, so naturally, one's impression of original form of A is stronger than A'. so practice that. then practice how A gets into B. and these are going to be rather short units, may be 4 bars or 8 bars.
do them slower than you want to. do them slow that it feels even boring. if it feels boring, see if you could attach some sort of biofeed back (as certain intervals and chords, shifts and fingerings will feel different, try to remember the 'feel' of it).

*this is a super 'frou-frou' version of analysis i did for a class. it's bit flowery, but i can still play that section looking at this. AHAHAHA.

go back to the blueprint on no.2, and see if you can consolidate the different of 'reading' the section change and 'playing' the section change. make up connections if necessary.

4. if you have similar sections, especially in rondo form or small-cell-composites, do simple analysis of both. you could do this simply by photocopying the section, print them out, tape them side by side and mark exactly what's different (use colours if you want; i do).  see if practicing those sections as 'first thing' of practice be any use, rather than staring from bar 1 or cadenza. you are freshest at the beginning of a session. use it to advantage. if you have the time, put the metronome on 'stupid slow' and use it as warm-up (intonation, articulation, bowing, breath exercise, chordal practice, whatever)

5. practice on your non-specialized instrument; i highly rec on piano or singing. singing cuz you can do it without one more 'thing.' if you do have access to piano, go through the piano part slowly with metronome (once again, on stupid), and sing your part. playing all notes is not important, but knowing all 'direction,' especially regarding directions of melodic/harmonic intentions will help you greatly (and it will also help with intonation and phrasing, articulation in top of sheer 'memorization.')

or listening to a recording (if you are playing with piano reduction, i may suggest that you listen to a 'lesser' version with a reduction, as standard reductions will have 'certain details' put in/ommited). and mark on your score what you HEAR as a relevant key for your part. often people mark from orchestral rec for piano reduction and sometimes, it is of no use as they didnt make it, or is impossible.

6. pick a time in the day, play through your thing and record it. sing through* all your rests if poss. and if you CAN, play the 'lead in' for your solo entry (in time please), this will help you realize what you 'depends' yourself on for the entry, counting rests* is different than singing through it, because counting and playing music uses separate (yes related but separate) part of the brain, and you will lose time as it will take time to reconnect those junctions. it's much easier if you sing through or silently play through the rest sections (than dead counting), as you will 1. still be in music (which will eliminate the 'dead face' during a tutti, and 2. singing is easier than counting.

listen to the recording after you went and made tea. make pact with yourself to not be so unreasonably angry with self. after all, it's just music. calm down. no one has know anything that happened in your own private session.

listening, make notes. where is the inaccuracy? if you video-tape, it's even better, as you can 'read' your own body language of 'uncertainty.' where you trip up is very unlikely the actual place you tripped. likely, it happened a few instance before that, and that has pushed you into losing balance, then to the eventual fall...

7. be nice to yourself.
if you dont like it, why are you doing it?
memory is repetition game and it does improve.
but simple repetition is a low-return gamble. yes. gamble. you are hoping for everything to go right.  dont gamble, plan. it's safer, easier and you will therefore play better.

here's album of sample analysis from my past and hope that helps.