Saturday, June 5, 2010

On "behavior" vs "behavioral"

Elisa doesn't want me to use sugary water as part of the training. Like all parents, she feels her kid shouldn't be hooked to sugar as we are. So I write to the pediatrician:
I would like to feed Dario, about four-month old, with tiny amounts of sugary water. That is, I would rub a bit of sugary water on his lips. The reason is that I am looking for behavioral reinforcers.

Could this be bad for his health? Does it risk giving him an addiction to sugar?
Their reply:
In regards to your email. We would not advise giving sugar water or using sugar water as a reward. In general, we do not advocate ever using food products as rewards for good behavior and at this age we would not suggest discipline or rewards as babies are not able to understand these concepts until 15 months of age. Behavior can be discussed at your next well appointment with your doctor.
I suspect I am in for a long lecture at the next visit. But I am preparing my ammunitions. I will tell them: "Does a puppy understand the concepts of discipline and rewards?" They will call the police.


Thursday, April 22, 2010

Extinction

Last February, I have mused on the different techniques to get your baby to sleep a full night. Animal training theory clearly ridiculed these theories. I was looking with disdain at all of them, including the colonel method (extinction).

I know two colonels: Elisa and myself.

One week ago, I came back from work having read some more of "Healthy Sleep Habits, Happy Child," by Mark Weissbluth, that recommended extinction. The rationale: it works, and it is the method that parents have an easiest time sticking to, because it will have effects very quickly (hence, positive feedback, for when you need to do it again after a sleep mess-up). Elisa, at the same time, had done some internet reading, in particular from Dario's pediatrician website. Claim: a baby is ready to do her nights at an early age. It's just the parents who are not ready.

At that time, Dario was being fed at 1:30am, 3:30am and 6am, roughly.

We looked deep inside ourselves, and thought that, if the conditions had been right, in a time of war, a war against an enemy set to destroy a good part of Humanity, that only the most resolved could stand a chance to defeat, if that had happened, it our president had asked for our help, had begged to put us in charge of a regiment, we thought, we would have accepted our responsibilities and the rank of colonel.

And thus, that same night, we moved Dario back to his room, closed behind him the steel doors of our hearts, and didn't answer his cries until the pre-specified time of 5am. Two nights later, he was sleeping seven hours in a row.

Last night, he slept nine hours and gulped down eight ounces of formula upon waking up at 5:20am. Two personal records that we will remind him of, with tears, for his twentieth birthday.

Where is animal training and positive feedback in all that? My belief system is shaken if not ground to bits.

With a way out: what we did was: "Ignore inconsequential undesirable behavior." We were using principles of animal training after all. Like any colonel who knows his men well.

Monday, April 5, 2010

Speech Therapy

Dinner with friends of Elisa. Alice is a speech therapist. This is all behavioral. You need to learn fine actions with your mouth, get used to placing your tongue here and there, pursing your lips.

There, timing is everything again. Alice tells us that as the kid speaks, he gets instant feedback with such words as "excellent" at the very precise time when he pronounces the sound correctly. You don't wait for the end of the sentence. She explains this method comes from behavioral studies, probably directly from Skinner. Warm fuzzy feeling all around.

Karen Pryor tells us the clicker works better than the voice for animals. Why is that so? She doesn't know why. It just does. She retells the story of a dog trainer who calls her irate: I tried your clicker thing. My dog learned twice as fast as when I use my voice. I took the clicker and smashed it to pieces!

It is so clear that it would work so much better for speech therapy too. Two hypotheses about the advantage of the clicker over the voice:
  • The voice carries too much information, and thus the organism spends a lot of time trying to interpret if "good" meant "goooood!!!" or "gud (but please, better next time)".
  • The voice's timing is not as precise as the hand. When the right behavior happens, it takes a tad longer to articulate "Good" than to pinch a little object between the thumb and the index finger.
What is the essence of comedy, you ask a friend. And as he is about to answer something: "Timing!" I tried to make that joke many times, but it usually falls flat: my timing is not as good as Borat's when he takes his class in comedy.

If a man answers

While I am still waiting for Dario to to give me a hint that he has noticed this recurrent click sound, that he has noticed food tends to come at the same time, while I am waiting -- and I am patient -- it is a good occasion to further my education.

Monica told us the other day of a movie. Chantal complains to her mother about her failing marriage. The mother proudly reveals her secret: the book "How to train man's best friend."

And it works (plot spoiler: until the husband is tipped off by an ungrateful friend of Chantal. What follows is what give the movie its title, "If a man answers", and is of no relevance to this blog).

What works? Very little is said, but this 1962 movie is resolutely modern.

1. When you call your dog, always make it a positive experience, and give lots of praise.

Chantal calls her husband Eugene to the bedroom (his studio is also his home), he arrives annoyed. She gives him displays of affection with a sexual undertone, and while you are here, hey, can you help me position these frames on the wall. Oh, perfect, I couldn't have done it without you.

In animal training, timing is everything. As soon as Eugene holds the frame on the wall: Oh perfect!

2. What to do if your dog is pulling on his leash? Let him drive you for a while, then finally pull him in the direction you want to go to. Unclear? Indeed.

Chantal can't get Eugene to go to Bloomingdale with her to choose some new drapes for home. So on she goes: breakfast in bed, and mon cheri (Chantal's mother is French), today is your day, I'll come with you anywhere you want to go. We follow an annoyed Chantal at the museum of photography (the movie is set in New York), at a camera store, then at a tobacco shop. Oh, mon cheri, we are two blocks from Bloomingdale, let's go. Why don't you go while I choose my tobacco blend? Unclear indeed.

Chantal calls her mother to get the missing instructions: at this time, you need to pull on the leash. Oh well, so much for positive reinforcement. But it works... until the ungrateful friend destroys this marital harmony.

These are alas the only two examples of the application of animal training. But still, it highlights the use of positive reinforcement and the strict application of operand conditioning. If the movie were to be done today, about fifteen years after clicker training has taken the dog world by storm, how would it go?


Saturday, March 27, 2010

Workshop day 1

The http://tagteach.com web site has a list of forthcoming workshops. One is in Vermont, five hours from New York. This weekend.

So, that's where I am, at the end of the first day, mostly a day of theory with a little bit of practice. Most of it, I knew from reading about it. Yet I didn't feel bored at all. It seems that Teresa, the presenter and co-founder of tagteach, has been using positive feedback techniques during the day, by giving us unexpected breaks.

Teresa is the gymnastics coach who was using clickers to teach her students. She is now full time in her company giving seminars and doing interventions in fisheries.

I laid down my cards while introducing myself this morning: I really need to be able to hold a handstand by April 15. I'll be clicked for that first thing in the morning tomorrow.

Dinner with six other of the participants. We talk mostly about dogs chasing squirrels and about raising kids (whether autistic or not). I mention clicking my six-week-old. Nobody is horrified. I feel I belong.

Saturday, March 20, 2010

Click for food

Being now the proud owners of 28 clickers (I gave away two to K, one to R, one to M, one to J), I started clicking.

Each time I insert the bottle into Dario's mouth, I click. Now, sometimes, Dario does a click with his tongue, a loud, nice 'tchak'. I am trying to shape him so that this is word for "more food, please". It seems to be the simplest possible shaping possible: he gives the cue, I do the behavior.

So far: no success. Need to remember to stop the training after a few minutes into the feeding: he still needs to be hungry enough to pay attention.

But so exciting.

TAGTeaching

Still reading Pryor's book, I learn that clickers are used in teaching. For autistic kids, but also in gymnastics, shoe-lacing, etc. Pryor has even co-created a company at http://tagteach.com. I am on my way to Vermont to take their workshop. This is where the handstand video comes from.

Parents complain: you can't treat our kids like animals. That's why the technique has been renamed 'TAGTeaching'. But somebody had a perfect answer to this common complaint:

The medications you give your kids -- they've been tested on animals first right? You wouldn't want them not to be tested on animals first. Well we did the same with this technique.
Beautiful.