When I was a young girl, my grandmother would send gifts of books from Czechoslovakia. The books were filled with stunning moving pop-up illustrations. I learned a lot from those books. I learned how those illustrations popped up. I learned how one moving part operated another moving part. What I failed to learn was how to read Czech. My attention was so fixated on the illustrations that I memorized the words. I recited the story based on the illustration. I never focused on the letters. Illustrations overshadowed the letters.
Overshadowing is a well-researched part of dog training. One place it applies involves adding cues. (Commands for those who still use that term) Animals, when simultaneously given two or more cues are likely going to learn about the most salient – to the detriment of the others. Which facet carries the most weight depends on many factors.
Lured downs offer an example many can envision. Dogs see humans bending at the waist during the lure. If owners simultaneously say, “down” while luring or pointing, the word is unlikely to register. Despite dozens or hundreds of repetitions, the dog stays focused on the owner’s gestures. The word, “down,” becomes irrelevant background noise. It predicts nothing of importance because with each repetition, the movement gives the dog all the information it needs. Touching the ground comes to mean “lie down” rather than the word. Pointing is perfectly acceptable if one wants to touch their toes in order to ask the dog to lie down. I just don’t know many people who put it on their wish list.
Overshadowing affects many areas of training. This is not a luring problem. It can affect targeting, shaping and capturing. It is not a positive reinforcement based problem. Overshadowing has far reaching tendrils that reach pretty much everywhere.
Gestures do not need to be grandiose. It can be as subtle as the wiggle of a finger or a raised eyebrow. We all have dog training tells much like poker players have tells. It takes self-awareness to compensate. Often it takes a second set of eyes to spot them.
This does not mean hand signals are bad. Rather that few people want ridiculous signals. No one wants to touch their toes to get a down. It’s annoying. Holding an arm extended for “stay” isn’t useful. You can’t hold your hand up to sign stay and tie your shoe at the same time. If you train with a prolonged gesture, that’s what the dog will need in order to understand.
Good hand signals are functional. It takes effort to teach a great verbal cue. Teach both but don’t mash them together and expect the dog to recognize them individually. Don’t blame a dog for misunderstanding cues that were never made clear.
This might seem persnickety. Who cares if signals are mashed so long as the method is “positive?” Frankly, it is a problem when many trainers tell clients to mash signals and then encourage “balance” to clean things up. Many repetitions become justification that the dog ought to know better and do better without recognizing that the repetitions were flawed.
Clients may feel that positive reinforcement failed because the dog is not responding to verbal cues. This perceived failure again leads to “discipline.” Positive reinforcement didn’t fail. It worked exactly how science said it would. In the absence of another reasonable explanation “dominant, stubborn, willful” can sound reasonable to a frustrated owner.
That’s a real shame because overshadowing is pretty easy to deal with. Cues need to precede the behaviour without interference. Fade lures. Remove training aides during shaping or targeting. Get a clean behaviour. Then purposely add in the cue, and only the cue. Spend a little time highlighting that these words or signs that are important and have meaning. Do this in a thoughtful and planned way.
If we dismiss science, then we don’t learn about factors such as overshadowing. I suppose it must be easier to justify corrections and sleep at night if one remains willfully oblivious to such things. It is a choice to ridicule science and then justify corrections under the pretense of “balance.”
Learning the science is hard. It demands that we regularly rip off the bandage that protects our fragile egos. We need to find the thing we have yet to learn. It goes beyond lip service. We can’t just say, “We all have more to learn.” You have to do something about it.
Learn better. Do better. Get better results. Often those who suggest this path went down it before. At least I certainly have. There are walls you hit where it seems like something isn’t working right. Finding out why is worthwhile.
Individually we should all lose a little sleep wondering if there isn’t some variation of overshadowing playing havoc with our cues (probably). We should wonder if it impacts other areas of training (it does). We should wonder if there are other effects to learn about (there are.)
Owners absolutely don’t need to learn decades of research in order to train their dog. The trainers they hire should.
The takeaway for owners is twofold. First, don’t mash your cues. Second, science often does have all the answers….and it makes training a lot less frustrating when you work with it rather than against it.
Body language plays a big role in training dogs. It was about five years since I had a pup and now this is just the blog I need. Thanks!
Still thinking.
Re subconscious cues we are giving, this probably explains why many quite well trained dogs, simply don’t do well with another handler
I volunteered for several years at “Coffs Harbour Marine Magic” (was Coffs Harbour pet Porpoise Pool). The encouraged volunteers to come and ‘put the animals through their paces’. They said it was so the animals (dolphins, (eared)seals, seal lions and one leopard seal) did not become reliant on one particular trainer. It was great fun for me ad a wonderful learning experience.
But your article has made me realise that ‘becoming reliant on one particular trainer’ would be because of all the unconscious signals we give our animals. Each individual person would of course be giving different unconscious signals, so the animals would better learn the signal (the official one) which was common to ALL handlers.
🙂
Could be. And that one trainer would have a nice long reinforcement history with the animal. The volunteers, not so much. The animals would have to figure out that volunteers were subs for trainers, not subs for spectators. “Strange people mean…????”
The trainer, that’s probably a guarantee for loads of R+.
Sometimes, of course, the signal that the dog is actually reacting to, can be so subtle that we don’t even know we are giving it. 😦 The the dog gets accused of anticipating ;-(
I had a dog who used to seem to react when I just THOUGHT to cue something. On serious thinking I realised that he was responding to my very subtle (well I was unaware of until I carefully assessed what I was doing) movements of my lips, before giving a verbal cue.
I knew another dog who was perpetually bombed for anticipating the recall in trials. Her handler realised that before she lifted her arms to signal, she was raising one shoulder just a little before she actually lifted her arms.
All very Clever Hans! 🙂
How awesome that you figured out what the signals were!
Excellent. Thank you Yvette.
Thanks for the nice comment Elizabeth!
Reblogged this on Louise's Dog Blog and commented:
For anyone who is battling to instil verbal cues, this is a really, really useful article. Yvette’s example of learning a story by looking at the pictures is brilliant. … And I’ve learnt a new word “Overshadowing”.
Thanks for the re-blog! New concepts are awesome aren’t they? Love it when that happens!