Film Exposes Dark Side of Orca Training

For decades, killer whales have been held up as poster children for the power of positive reinforcement and applied operant conditioning generally to produce reliable behavior without the use of force. Yet they suffer lives of abject deprivation, if not actual psychosis.

The maiden post to this blog was in response to the death of trainer Dawn Brancheau at Sea World’s Orlando park. It was the third death associated with the bull orca Tilikum, known to tourists as Shamu.

The newly released documentary Blackfish tells Tilikum’s story, and sheds light on the atrocities behind all those uplifting and profitable Sea World performances. For the record, I have not yet seen the film, but I’d hazard a guess that the message is that the magical relationship humans have achieved with these majestic creatures only seems magical to us.

To them, it is a tour of captivity, isolation, and abuse, no matter how many buckets of fish get tossed down their throats. Click here to listen to Jean-Michel Cousteau’s statement on keeping orcas for fun and profit.

Positive reinforcement based operant conditioning has proven utility both within and without the confines of zoos and amusement parks. I don’t deny that. Neither would I suggest that positive reinforcement was itself unethical. But I deny the legitimacy of extending the analogy between dogs and killer whales to the point of suggesting the best tools for engaging the latter must also be the best choice for training the former. And in so far as the management involved in captive marine mammal training is in fact abusive, there is real danger associated with modeling dog training after their example.

© Ruth Crisler and Spot Check, 2013.

Tags: , , , , ,

  1. Donna Soderstrom’s avatar

    I listened to an interview this am on NPR with a trainer involved in the film, and the director. The specific facts behind this appear to be outrageous. The whale had killed 2 other people, and was known to be dangerously aggressive.

    Ruth, you know I have always been of the same mindset as you about operant conditioning. Yes it can work and it has it’s uses. But, simply because it does not involve the infliction of physical pain does not mean it is humane.

    Take any intelligent creature, highly regulate it’s food intake, social interactions and mental stimulation, then tout your results as a model for training the family dog.

    How on earth did Skinner style operant conditioning come to be so revered? To me it is a souless tool. Why a “click” instead of the human voice?
    BTW Happy to hear Dr Brian Hare bringing to light some of the limitations of OC

    Reply

  2. Ruth Crisler’s avatar

    Soulless is a good word for it. As for why it’s held in such disproportionate esteem, I’d guess it has to do with the modern fascination with all things clinical.

    Reply

  3. Lori Ziemba’s avatar

    I think the reason it is so popular that we live in a world where everything must be nice and politically correct. Children are not disciplined, dogs are left to run amok, all because no one wants to look like a bad guy, and no one wants to admit that life includes pain. It’s like an entire generation never had to face any hard facts.

    I worked with dolphins in a public aquarium. While they were kept in an appalingly small encosure, they were at least fed to capacity 4 times a day. They would perform their “tricks” when they felt like it, and sometimes be naughty, because there was no coercion. I was told by the other biologists that all performing dolphins are fed 10% less than optimal to keep them hungry. THAT is why they perform. It’s not love, it’s not the power of positive. i see nothing positive in being kept hungry.

    Reply

    1. Ruth Crisler’s avatar

      Agreed. This is why it’s vital to look at an animal’s whole experience under training (and beyond). Fixating on one’s choice of Skinner quadrants is like obsessing over calories when dieting. A healthy outcome is by no means guaranteed.

      Reply

  4. Polly’s avatar

    “How on earth did Skinner style operant conditioning come to be so revered? To me it is a souless tool. Why a “click” instead of the human voice?”

    As an enthusiastic clicker dog trainer, I want to explain why the “click” is so effective in dog training. The sound of the click literally goes through part of the brain called the amygdala. This part of the brain processes emotional reactions and other things related to memory. Thus the clicker goes through a very primal part of the brain and makes learning more effective. That is why I love clicker training. I’ve never, and have never seen clicker trainers, use it in a way that is soulless and cruel. Clicker training is meant to be fun because it gives the dog (or animal) such clear communication! My chihuahua’s tail wags furiously when she sees me pull out the clicker because it is a game!

    But yes, indeed, the clicker is a TOOL. and any tool can be used humanely and cruelly. Just thought I’d add my two cents to this conversation. This documentary looks very very interesting.

    Reply

    1. Lori Ziemba’s avatar

      Can you cite references to any peer-reviewed, science or medical articles that state this amygdala reference?

      Reply

      1. Polly’s avatar

        :) I believe in the book Reaching the Animal’s Mind by Karen Pryor. But it is kind of a no brainer if you think about it. Here’s an article by Pryror about this: http://www.clickertraining.com/node/226.

        Reply

        1. Ruth Crisler’s avatar

          But Pryor is merely describing her personal hypothesis, not presenting scientific evidence.

          And to my mind, that hypothesis could apply equally well to any conditioned reward marker, verbal or non-verbal.

          Reply

          1. Polly’s avatar

            If you read the book, she goes into details on a conversation with a scientist about the amygdala and clicker. It’s quite convincing. Check it out sometime. :)

            Sure, verbal reward markers can work just as well- Michael Ellis and Ivan Balabonov mainly uses verbal markers if I remember right. I just love the clicker, I feel like I’ve gotten such great results from it and it’s been really amazing to watch dogs learn with it. I truly believe it goes through the brain in a much more primal manner with the amygdala but of course I’m open to being wrong.

          2. Lori Ziemba’s avatar

            Humans experience episodes in which the PNS is active as nice warm feelings, relaxation, contentment. Anytime that a previously neutral stimulus, like a clicker, or a kind word, gets paired with one of these parasympathetic reactions, through Classical (Respondent conditioning) the clicker acquires the ability to produce the same pleasant effects.

            What this is saying is that ANY neutral stimulus, when paired with a reward, will produce the desired effect.

            I would like to see real scientific evidence of this, using an eeg or MRI. If true, then using it the wrong way could also result in a fear response, since the amygdala tends to govern primitive emotions.

            My beef isn’t with clicker training, per se. I think it’s a good tool for teaching. When I teach a dog a new behavior, or put a behavior on cue, I use rewards. But after a behavior is learned, comes training, when you want the dog to be as reliable as possible under varying conditions. Some behaviors we might consider bad are actually much more rewarding to a dog than a food treat. I guess I’m old school, but I still believe a dog needs to know that it MUST obey, and that the trainer can enforce her commands. This doesn’t have to involve pain, but rather the dog must be made to understand that it does not have a choice to disobey.

          3. Ruth Crisler’s avatar

            Polly, yes, many high-level trainers use verbal markers exclusively or almost exclusively. I think the clicker has advantages and disadvantages myself. But in any case, marine mammal trainers (and some well-known dog trainers, like Chirag Patel) generally use whistles, as I understand it.

            To be clear, I’m a fan of marker training. But it’s not the be-all and end-all, in my opinion. It’s a technique, and a damn good one for building associations efficiently, whether positive or negative.

            My point was that we need to look beyond the tools and techniques a trainer chooses, in order to fully appreciate and properly assess the learner’s experience.

            If an abusive level of management is required to sustain “humane” choices, then it’s time to reevaluate.

          4. kaimeju’s avatar

            The clicker is popular (not only with pet trainers, but with psychologists studying animal behavior) because it is usually a novel sound (so it won’t be confused with anything else) and because it is consistent (verbal markers sound differently depending on tone of voice and who is saying them). No scientific studies needed: it’s just prima facie true, unless you have a dog that is deaf or has heard a clicker before. I don’t think anyone is claiming that they work equally well with each individual dog.

            Verbal and visual markers work just as well if used consistently. You could use a light flashing or a word or a hand signal. I don’t know of any pure positive trainers that only use clickers and never encourage their dogs verbally or with positive touch.

            Lots of straw men being built in this comment thread. If balanced trainers want to be taken seriously, they (we?) need to not fall into the bad habits that have led positive-only trainers to create gross caricatures and flimsy oversimplifications. It may not be your cup of tea, but at least take the time to understand it.

  5. Dale McCluskey’s avatar

    It is a identity crisis based on an ideology of self love. Learning theory is being exploited and misrepresented by this group of reprobate ideologues. Dogs, Dolphins and Whales are merely an extension of this ideology and devalued as a result. It is being embraced and worshiped by a generation of Narcissists who worship themselves.

    Reply

  6. Beth’s avatar

    John Rodgerson, who used to be the darling of the “purely positive” set complains that Americans have forgotten how to praise and communicate with our dogs. Instead, he finds too many using clickers and food to the exclusion of a real relationship.

    Reply

Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>


Hit Counter provided by Sign Holder