Animal "reasoning, jackpots and food
Posted by: old_user on Dec-08-06 10:18 PM (EST)
I've spent a good part of my life training animals of different sorts, right now primarily dogs.
Maybe because of what most people been learned in intro biology classes they tend to think in terms of animals being capable of only simple stimulus/response types of learning.
If you work with them toward any sort of focused goal, you shortly begin to find that they are capable of learning many, many times more, in greater volume and to a higher degree of sublety than you ever initially imagined.
One of the most popular used to teach and reinforce a complex sequence of behaviors is a tool called a "jackpot".
This is basically giving a huge, beyond-belief reward if the dog, for example, successfully executes the desired behavior.
Small rewards reinforce modest progress. However, if the animal (usually at first mostly by chance) hits the nail directly on the head, the trainer opens up a huge stash of highly desirable food supply (or makes available a highly desired but witheld toy) and lets the dog have at it.
For example, a jackpot means the trainer might take off the bait bag that holds bits of chopped up steak, and instead of doling it out piece by piece, opens the bag and gives the entire contents to the dog to wolf down.
You don't do this every training sessions. Just at those times where the dog really gets it right in a spectacular way.
The result is that that incredible, unexpected reward will greatly increase the chance that the dog will try very hard to repeat that behavior...and so starts to imprint it HARD as a learned pattern.
Dogs and other animals will sometimes go to great extents to get jackpots. The effect is multiplied by a "gambling" effect created due to the fact that the payoff doesn't come on a regular schedule.
Animals can become addicted to "gambling" for jackpots just as easy as those lost souls caught up in the thralls of the one armed bandit.
They can learn behaviors that require several stages to get to the payoff.
AND they can make very novel adaptations when situations change or when the puzzle becomes harder.
This to me demonstrates that it is a sort of basic reasoning that begins to take place, rather than just rote learning or response to instinct.
I've seen many instances where an animal will take something they've learned in one context and add it to something they've learned in a separate context to self-create a new behavior to address a novel challenge.
I have absolutely no experience with bears, per se, but because many wild animals tend to be more intelligent than their domestic counterparts (domestic animals tend to have had their wits dulled to make them easier to handle), I can only guess that an animal such as a bear could very well be capable of constructing some fairly innovative and effective solutions to a problem as appetizing a food treasure trove hung from a tree.
It's actually, in fact, a relatively easy problem compared to other food finding challenges a bear might face on a daily basis. After all, food caches and garbage don't fight, don't hide themselves and don't run away.
To a bear's mind, it's probably a very fat, sluggish and energy rich find that is well worth a little investment.
Cheers!
 |  |