ORCA and the University of North Texas are proud to present:
The 2nd annual
Art and Science of Animal Training Conference:
Innovations and Refinements
Saturday, February 6th, 2010
8:30 a.m. – 6:00 p.m.
This year’s title is Innovations and Refinements. The field of animal training is continually changing–just when we think we’ve finally understood the secret to behavior, we learn something new! Our guests of honor will share with us some of their own innovations in animal training.
Program
Robert Epstein (Keynote Speaker): Engineering Complex and Novel Behavior in Animals
Ken Ramirez: Wanted: Animal Training Consultant–those good with animals need not apply!
Steve Martin: Training Outside the (Skinner) Box
Kay Laurence: Assessment: Love it or hate it, it makes you grow
Alexandra Kurland: What Horses Teach Us About Cues: The Loopy Training Strategy
Bob Bailey: Innovation: A Case History
Abstracts
Robert Epstein (Keynote Speaker): Engineering Complex and Novel Behavior in Animals
Animal training usually consists of applying positive and negative consequences to behavior, strengthening some behavior and weakening other behavior until you get a desired performance. This is what psychologists call operant conditioning. But rewards and punishments only strengthen or weaken behavior that’s already occurring. You can’t get NEW behavior unless you know how to wait strategically; in a procedure called “shaping,” you wait for a bit of new behavior to occur and then reinforce it, and then you wait again until you get MORE new behavior, and so on. But here’s a surprise: When you’re not training—which is most of the time!—you’re really waiting, and while you’re waiting, almost ALL behavior that occurs is new in some sense. The new behavior that occurs most of the time in virtually all animals (including people) is called “generative,” and laboratory experiments show that this kind of behavior is both orderly and predictable. It HAS to be orderly and predictable, or shaping couldn’t work! First developed in the 1980s, Generativity Theory is a predictive, mathematical theory of generative behavior—in other words, of most of the behavior that occurs in real organisms most of the time. It has proved to be useful in both predicting and engineering complex behaviors in animals that is so extraordinary that it appears to have features of human “higher mental processes.” The theory can also be applied to improve the way we train animals.
Ken Ramirez: Wanted: Animal Training Consultant–those good with animals need not apply!
The title is, of course, somewhat facetious! To be a good animal trainer, one does need to understand training and be good with animals. However, sometimes the most important skills needed to solve behavioral problems are not animal training skills. It is people skills, observational skills, and organizational skills that can be the key to finding solutions to behavioral problems. Before tackling a behavioral problem with the household pet or a large zoo animal several factors need to be considered. This presentation will focus on those factors that need to be addressed first when trying to solve animal related problems. A review of various case studies will help to reveal how to start out with the right tools and why animal skills may not be the only talent required.
Steve Martin: Training Outside the (Skinner) Box
Training is training no matter where it occurs. Whether you train a rat inside a completely controlled environment with no distractions, or train a rat to run a specific path in a show performed in a 5,000 seat amphitheater, you teach that rat to make decisions based on antecedent or consequence experiences. Lab animals perform in controlled environments to limit distractions and establish controlled experiments that provide more consistent data. But, how much is the behavior of the animal inside the experiment chamber influenced by events that occur outside the chamber? How do distant antecedents influence the performance of a behavior? Are setting events always recognized and acknowledged in the experiments? What influence do motivating operations other than food have on the performance of the behavior? To answer these and other questions we will look at some of the most important factors that guide the training we do with free-flight birds and other animals in zoological exhibits.
Kay Laurence: Assessment: Love it or hate it, it makes you grow
Assessment can be defined as “the sampling of some aspect of a person’s or dog’s behaviour at a particular moment. At Learning About Dogs we operate a Competency Assessment Scheme. It is designed to provide a structured curriculum to develop an individual’s skills, and through assessment against set criteria, progress can be measured. Assessment can range from very informal minute to minute assessment of the environment, the training plan, the progress etc; to quite formal programs for certification. Assessment identifies prior learning, strengths and weaknesses, aptitudes, achievements and areas in need of development. Assessment can determine the skills of the teacher trainer, and whether the learner has reached the outcome. It shapes your way of thinking about what you and your dog have learned. You learn to look for evidence of learning, evidence of skills and competency. Assessment is not judgement. The difference between assessment and judgement is a different mindset, different approach and a different outcome. Learning to assess, yourself, your dog, or your class is a key skill without which your development can be seriously restricted.
Alexandra Kurland: What Horses Teach Us About Cues: The Loopy Training Strategy
We often think of each training step as a single sentence: Cue => behavior=> click => reinforcer. In teaching complex behaviors we need to present this sentence not just once, but many times. In doing so we end up training in repeating loops. You can recognize good clicker training by the presence of clean loops. Mix in poisoned cues and the loops fall apart. Clean loops tell us when we can add in a new criteria. Tightening up a loop gives you effective strategies for eliminating unwanted behaviors and for creating fast, efficient shaping plans. Loops are everywhere! Once you learn hot to recognize them, you’ll see they’ve always been part of your training. Awareness creates more effective use. Loops help handlers resolve any poisoned cue effect that may have crept into their training. They let you problem solve more quickly and efficiently. Well-crafted loops turn the behaviors we teach into cues our animals use to shape our behavior!
Bob Bailey: Innovation: A Case History
What is innovation in animal training? During the 1940s Skinner’s behavioral innovation, called operant conditioning, was leaving the psychology laboratory and transforming into a powerful methodology for animal training. His innovation continued commercially with Keller and Marian Breland. They founded Animal Behavior Enterprises (ABE) in 1943 and commercially exploited the new behavioral technology. Careful experimentation and data collection were at the heart of the Brelands’ technical accomplishments, but it was ABE’s innovations and training efficiencies which made it commercially successful for nearly half a century. One example of ABE’s novel application of operant conditioning was the use of discriminative stimuli to maintain long duration behaviors. Beginning with rats in small boxes in 1965, and extending to dolphins and dogs and cats operating in free environments today, the Keep Going Signal (KGS) has been useful to maintain long duration behavior. This presentation describes ABE’s development of the KGS and some of its applications and addresses issues concerning training innovation.