posted
July 16, 2004
NEW Q OF THE WEEK:
Define what you think this statement means:
"A
dog can learn different cues for the same behavior.
But it cannot learn different behaviors for the same cue."
Part
two of this Q of the week:
Since we have some latency on the part of some of our Ruffians,
we are moving to a Fixed Interval schedule. What
does this mean?
FIRST IN
(and only *lip*) - FROM TERI !!
Is this a trick question??
Pretty straight forward, I think.
A dog can learn different cues for the same behavior. A dog can learn a voice
commmand or hand signal for sit or any other cue for that matter (he may even
recognize one that his owner doesn't know he is even using) Body language is
used alot by agility dogs and dancers even. But they cannot learn different
behaviors for the same cue. Let's not confuse them anymore than we already do.
It's bad enough that the English language has words that sound the same but
mean different things according to context. I don't believe that dogs are much
in to context. Just keep it simple.....a cue should always get the same response
promptly.
Fixed interval schedule--- Snooze You Lose
The questions will appear at the same time every week (or whenever) No more
waiting, begging, adding on time, etc, etc.for answers. Answers must be rec'd
by a specified time. Or does it mean we will receive a reward for answering
the questions at a fixed interval. *giggle* I work for chocolate!! Teri .
YAY, TERI !!! She wins the jackpot!!
My answer:
"A
dog can learn different cues for the same behavior.
But it cannot learn different behaviors for the same cue."
Of course, we all teach
different cues for the same behavior:
Hand signal, verbal
cue, environmental cue -holding dinner dish or stop at door = "sit"
:
This can work to the handler's advantage or it can be detrimental to quality
of response, as in these often heard variations on a theme from your average
student:
Sam Come!, Come Sam, Here, Hurry, Come-on, Come-here, Get over here, Come-come-come, COOKIEEES!! (He always comes when I say "cookies" ...) = moving toward owner.
The dog always comes when you say 'cookies' cuz he always gets COOKIES! He comes slower when you say "get over here" as he usually gets chastised for some unknown reason when he hears the phrase, but since the owner is relentless and means it, he does come - but reluctantly. The consequence determines the behavior.
Unfortunately, since unknowing owners often switch back and forth between cues within the same sentence, with no real connection between the word and the desired response, no real training has taken place, and a lot of learned irrelevance. "Cookies" may be the only cue he's really LEARNED - as it's the only one that's consistently used with a reliably pleasant consequence.
Multiple cues for a behavior can complicate a desensitizing program - recognizing how many cues are conditioned in is essential to changing it: car door slam, footsteps on porch, other pet's alert, ding-dong, knock-knock, who's there, who is it, come in! = run to the door/bark/excitement
Different
behaviors - same cue:
Down! (lie down) Down! (don't jump up) Down! (get off the couch).
If the dog truly understood the cue, he would drop to his belly in all three
instances.
SPOT! (look at me)
SPOT! (stop that) SPOT! (leave it) ...
If the dog truly understood his name "Spot" as a 'watch me' cue, he
would look at the owner (effectively stopping what he was doing and turning
away from the forbidden thing or activity) - the problem is inconsistent criteria,
sloppy handler communication, too high a distraction level for the dog's ability
resulting in impatience or anger, and the downer of the implied "or else."
This can totally mess up using the dog's name as a positive attention cue!
I guess the REAL question
is - why would you want to have more than one cue?
For obvious reasons, the more ways we have to communicate, the bigger our dog's
vocabulary and understanding will be. When we have laryngitis or our hands are
full of groceries, we have an alternate way to communicate with our dog. Having
the option of verbal or handsignal, the ability to give very subtle cues, employing
chained behaviors that become reinforcers for the next behavior in the chain,
and contextual cues are all valuable for daily living and performance builders
when working with a dog as a team doing a variety of tasks from agility to obedience
to dance - it's how you achieve a level of training where the dog seems to know
what you want before you ask him.
Let's use walking on
leash as an example: "by me", "let's go", "heel",
"other side"
Same basic concept - walk near my leg and keep the leash slack - with different
levels of criteria, from relaxed walk to performance precision.
IF you use the same word for ALL variations of leash walking, the dog will struggle to know which level of performance you are expecting at any given time AND will be forced to use trial and error (offering the lowest level of response that works) or observe and take his cues from some other prompt: body posture, how you hold your hands or the leash, appearance or absence of treats or fanny pack, or environmental context. Lots of dogs walk nicely on the sidewalk and pull at the park not because of distractions, but because grass has become a release cue to move away and sniff.
You may be completely unaware of these accidently learned cues and risk a long history of accidental reinforcement and associations that you will have to un-do later after they've become habits. How many of our students have named "pulling" by repeating their loose-leash cue as the dog exerts tension, lunges ahead, and his nose hits the ground?
A new cue to fix a poorly
learned one?
If you have a substandard response with a long learning history or learned irrelevance
from owner nagging, you can (and should) re-teach it from scratch and give it
a new name.
Cue
vs. Command - they ARE different!
At the University of Texas they did a study comparing "old school vs new
school" methods. They taught the same dog two separate recalls. They used
free shaping with a clicker and treats for one cue and a different command given
in Swedish taught by hauling the dog in on a long line. The dog learned to respond
to both cues. However, the level of performance differed. When left on a stay
and called with the clicker-trained cue, the dog came head and tail up with
enthusiasm. When left on a stay and called in Swedish, the dog came slowly in
an arc, and showed calming signals. Same trainer, same dog, same exercise. Different
signal with different learning history, different response. The R+ cue became
a conditioned reinforcer in and of itself - the positive attitude became classically
conditioned into the cue!
Part two of this Q of the week:
Since we have some latency on the part of some of our Ruffians,
we are moving to a Fixed Interval schedule. What
does this mean?
Latency is the time delay between the cue or signal (in this case, the posting of the QoW) and the behavior (the Ruffian's most excellent answers). Fixed Interval is a predetermined time limit on a fixed schedule. By delaying the consequence (the posting of the Ruffian's answers) I have effectively punished the early answerers by making them wait and rewarded the dawdlers - actually increasing their dawdle time! SO in order to prompt a little faster response, we will from now on, be on a short time deadline of 3 days per question.
AND Teri got the bonus jackpot goodie. *G* But she
won't tell you what it was!!