Consciously stop yourself leaning your body at all, and push on the 'bars on one side _gently_ (requires almost no perceptible movement).
If you're pushing on, say, the right hand side (or turning the wheel to point to the left), you will feel the bike go towards the RIGHT. This is counterintuitive to what you "know" about steering. But it's happening right now so you'll have to believe it!
Do it again, and when you feel the bike go towards the right, lean into it. You will find you can swerve quite violently to the side if you go for it.
When you get used to it, use this technique to chuck the bike into bends.
When this feels normal to you, and you can do it without concentrating, this is the point at which you realise you've been doing it ever since you started riding the bike! (OK, maybe you have more control and can turn & swerve harder now, but take my word for it: if you are riding a bike in any competent way, you have discovered countersteering for yourself).
p.s. now we can start arguing about centrifugal forces, displacing centres of gravity, whether or not it works on shaft drive bikes (it does 'cos I did it on a Guzzi and there's no way you ignore that sort of shaft effect:-). But don't worry because plenty of riders and bikes get on just fine without understanding what's happening!
You can initiate this lean in a number of ways, the most dramatic of which is known (confusingly in my view) as countersteering. This just uses the sames factors as above but what you do is shove the bars the oposite way to where you want to end up going while you are still upright. The result is that you are briefly cornering without leaning so you end up falling out of the corner neatly into the lean that you want in order to be able to corner the other way. "But what about the bars?" I hear you ask. Easy, the steering geometry, maybe with a little unconcious help from you, flips the wheel across to point the other way into the turn. Don't take my word for it, try it.
You are now leant over with the front wheel pointing into the corner, all the same rules still apply so you can pick the bike up by steering tighter into the bend or drop it down by steering less.
Where the rest of the world went wrong, an epistemological study of motofizzix delusions.
Why is it ever mentioned? I think it is probably used in motorcycle training courses to gain credibility and respect for the tutors and to prevent the trainees from getting over confident. I believe an extreme form is taught here on some Police courses as a low speed avoidance technique. I think people get confused between the bit of countersteering we all probably do without realising it and giving the bars a damn great yank to throw you over in an emergency, not something I would like to teach beginners.
Really counter steering. Some particularly sad souls believe that you actually steer the wrong way all the way through a turn. They invoke mysterious forces such as "the icecream cone effect" (see below) to provide a justification for how this might work. Probably induced by misunderstanding of the name "countersteering". If it were called "steerbanking" or something like that they confusion would probably not arise. Show them endless pictures of bikes cornering where you can see the alignment of the wheels. Only when the back end is sliding will the frame be the wrong side of the bars, even then the front wheel will be pointing round the corner as you would expect. Tell them to get a bike (Ed) and try it themselves, stick a straw on the top yoke to point at a mark on the tank, fairing and see what happens.
Gyroscopic forces. These are neither sufficient nor necessary to explain the phenomena. Since the bars flip back to the correct way when you actually start turning why don't you get picked up again. Can you really believe that a bike with ultra light wheels won't go round corners? When people switch to fancy lightweight racing wheels do they have difficulty steering? No, if anything it becomes easier because the main effect of gyro forces it to resist motion, a heavy front wheel makes it harder not easier to corner. Geoff Duke had great problems with his Gilera when they put a damned great drum on the front to try to get round the brake overheating problems of an enclosed front wheel. The other counter to gyro forces is that you get the same sort of behaviour from jetskis.
Icecream cone theory. This is widely used by people who seem to think that countersteering is the opposite of steering and thus need a way of actually getting the bike to go around the corner.
It seems to be largely based on a comment in an article by Kevin Cameron, a respected US bike writer, after visiting the Michelin factory. One of the things he mentions is camber thrust, an effect whereby the slightly different radii at the two sides of the contact patch create some torsion on the contact patch. The article doesn't say anything about the being a complete explanation of cornering forces, it is just one small effect which tyre companies think about.
This becomes simplified in the minds of some to "You have to different radii, like a cone, so you roll in a circle." First of all to talk about radii in a rubber tyre is fairly meaningless as at the contact patch the tyre is deformed from a circle anyway. Secondly what happens to the steering forces that apply to cars, tricycles and slow upright motorcycles, do they just vanish? Does the bike stop trying to follow the path of least resistance simply because the contact patch is having a bit of a writhe?
That's it. I gave up arguing this one on rec.motos a couple of years ago because there are some astounding obtuse people out there. I'm not claiming the above as a complete account of all factors involved, just a simple, adequate and internally consistent theory and some objections to some of the alternatives.
Dave Edmondson, only genuine with this label