Karen Pryor made clicker training popular with her book Don't shoot the dog. In it, after delving extensively in what it means to shape a behaviour with conditioned reinforcement, she lists 8 methods ("the only ones there are") to get rid of unwanted behaviour. The first four are more or less cruel (to yourself, in the case of #4), and the last four are more or less humane, but she states that every method has its place (except #2, punishment, which doesn't works—even when it seems to).
She also lists examples of every method for common situations, like dealing with messy roommates, barking dogs, irritable husbands, faulty tennis swings, aaaaand, cats jumping on the table, which I include next to her description of each method:
Method 1: “Shoot the animal.” This definitely works. You will never
have to deal with that particular behavior in that particular subject
Cat gets on the kitchen table: Keep the cat outdoors or get rid of it.
Method 2: Punishment. Everybody’s favorite, in spite of the fact that
it almost never really works.
Cat gets on the kitchen table: Strike it and/or chase it out of the
Method 3: Negative reinforcement. Removing something unpleasant when a
desired behavior occurs.
Cat gets on the kitchen table: Put cellophane tape, sticky side up, on
the kitchen table.
Method 4: Extinction; letting the behavior go away by itself.
Cat gets on the kitchen table: Ignore the behavior. It will not go
away, but you may succeed in extinguishing your own objections to cat
hair in your food.
Method 5: Train an incompatible behavior. This method is especially
useful for athletes and pet owners.)
Cat gets on the kitchen table: Train the cat to sit on a kitchen chair
for petting and food reward. An eager or hungry cat may hit that chair
so hard it slides halfway across the kitchen, but still the cat is
where you want it, not on the table.
Method 6: Put the behavior on cue. (Then you never give the cue. This
is the dolphin trainer’s most elegant method of getting rid of
Cat gets on the kitchen table: Train it to jump up on the table on cue
and also to jump down on cue (this impresses guests). You can then
shape the length of time it has to wait for the cue (all day,
Method 7: “Shape the absence”. Reinforce anything and everything that
is not the undesired behavior. (A kindly way to turn disagreeable
relatives into agreeable relatives.)
Cat gets on the kitchen table: Rewarding the cat for periods of
staying off the table is practical only if you keep the kitchen door
closed when you're not home so the cat can't indulge in the behavior
Method 8: Change the motivation. This is the fundamental and most
kindly method of all.
Why do cats get on the table? (1) to look for food, so put the food
away; (2) cats like to lounge in a high place where they can see
what’s going on. Arrange a shelf or a pedestal higher than the
tabletop, close enough so you can pet the cat, and offering a good
view of the kitchen, and the cat may well prefer it.
As Karen says, almost everyone resorts to #2, Punishment, even though it's mostly useless, for cats or anyone else, humans included: with or without social structure, with or without ingrained fear (how effective is the war on drugs?).
The thing is, when the punishment is applied during the act, it can inadvertently work like #3, Negative reinforcement: if every time you do something you immediately get an annoying electrical shock, burn, water squirt or whatever, you'll quickly and forcibly relate the act with the punishment (Freud's association by simultaneity: neurons that fire together wire together) and you'll become a Clockwork Orange, reeling away from the act (and the punisher), though not as dramatically as in the movie. Much (most?) of our behaviour already works by association: you already know not to put your hand in the fire; just as well, it can be hard to walk over burning coals, even when you know they are (mostly) harmless.
Karen herself used the bottle squirt to stop a dog from scavenging in the trash, but in a more sophisticated way: she put a few drops of vanilla extract in the water, so as to make its smell identifiable (and more annoying to a sensitive nose), and fired it (with a duly noted cringe) at the dog, and also at the trash cans. After just a few "association shots", the dog stopped playing with the trash —and eating vanilla ice-cream.
So again, every method has its place (except punishment/disciplining), but if you value your humanity, if not your furry friends' well-being, you could start from bottom to top, even though the humane methods require more effort and creativity...
...or rather, precisely because of that. If you go through the effort of training your pet with positive reinforcement, you'll find out that it goes both ways, that it'll become more playful and creative, and even demanding of her needs, now that there's a clear way to communicate.
For example, I have not been consistent enough to control my dog's barking with #6, but she just loves me to try... and specially, likes to figure out new behaviours —you can see the light bulb shine in her eyes when she gets it. So, I'm delighted with the result, even if she still wakes me up at the middle of the night... (but we're making slow progress on that).
The process of behaviour shaping is too simple (you already use it, inadvertedly, when you clap, smile or scream "well done!"): you associate a word, noise or gesture to stand for "you're doing good!", and every time your trainee does good, you use the signal (event marker) and immediately stop the game to offer a reward (always, even if you used it by mistake!). This works for anyone, from fish to children.
Whistles and clickers are very practical because they are loud, clear and hard to use by mistake (unlike clapping or saying "good!"). You work in short sessions of around 20 reinforcements, or whatever your friend can get through without getting bored (try to end the sessions in a high note, specially a break-through, so she stays interested). First you associate the signal with the reward with two sessions of using it and giving a treat. Then you reinforce easy stuff, like looking towards your face, sitting or other tricks she already does, and build increasingly complex behaviour. Two things that are hard to do right: this is a game of warm and cold where you shape (reward) every step towards the desired behaviour. You must read the situation: sometimes she will learn to do something in a few steps; others you will have to reward every little improvement, or even backtrack; then again, other times you will fool yourself thinking something happened when it didn't —
once I thought she had learned basic math, much alike that horse that could count... but only when his owner was around.
And this takes me to the second hard thing: if you're trying to communicate to someone that doesn't speak your language —any of them!— don't talk, with words, face or body, unless it's an established or developing signal. If you do, you'll just make things more confusing for her. So keep a poker face, and body, and use only purposeful signals (including encouragement and reward words). And I do mean poker: you have to watch for unconscious ticks as signals, or she will read and associate them, just like the genius horse, that knew when to stop counting by some minute signal that not even his owner was aware of doing.
Playing games like this, you both will get loads of fun and understanding, or even a therapeutic challenge in life, which can improve the mood of confined pets immensely.
Don't shoot the dog gives you a good idea of how and why reinforcing works, but even though it's peppered with examples, it's no step-by-step guide, so one of those could make it easier get started, if you wanted to try your hand at it.
(But naughty pet or not, Don't shoot is a great read to understand our behaviour at its most basic level —fav takeover from the book: no one should try their hand at parenting without first training a chicken).