Quite a big word, subliminal. I had just learnt it from reading Andy Soltis's STUDYING CHESS MADE EASY.
I would recommend this book for the serious player wanting answers on how to improve his/her chess through the various ways of studying chess. There are many methods which I find I had in common with the ones he explained in his book.
The one I liked best was Subliminal learning - using methods to input chess knowledge to the mind subconsciously. How can this be done? Reading chess books and analysing every diagram one comes across. In his book he gives countless examples of top players doing that, Each time the book is read, different thoughts and perceptions are picked up by the player. One such book I find which does this is Nimzowitsch's MY SYSTEM. You can read this book at different phases and it tells you different things.
The other way which Soltis does not talk about and is my own invention, is Chess Game watching. You will need a game-viewer or Chess database like chessbase to do this. This manner of learning I find is applicable to learning openings and middlegame plans right out of the opening, particularly useful in theoretical openings like the Ruy Lopez and Sicilian.
Select an opening variation of say 15 moves deep. Select about 50 games where the same 15-moves were made and White wins subsequently later. Try to keep the games short, say about 35 moves max. Load them onto the viewer and get it to replay the game with an interval of 1 sec a move. Watch the game like a movie. Once the game is ended, the viewer will automatically start the next game. Sit through this 50 game movie and I assure you, your mind will pick up many useful bits of instruction as to how White went about winning his games.
Now use the same variation and select 50 wins by Black. Go through the process and now you will learn exactly how Black defended and won, subconsciously. Your eyes and mind learns at a speed way beyond your imagination.
I gathered that this is how human beings derive what and what not to do when playing computer games, not just by memorising patterns of steps but subconsciously programming and reprogramming themselves to do the right thing till it works out to perfection. Then they proudly defeat the game. However, computer gamers are now re-introducing the human element into the domain and allowing humans to pit their wits and knowledge against each other, so there is a wider dimension of plans that can counter archtypical setups and steps. The computer gamers are learning from chess after all, so why shouldn't we learn from them in how to learn ?