Page 158 - Zoo Animal Learning and Training
P. 158

130  7  The Art of ‘Active’ Training

  VetBooks.ir    reinforces the next correct closer response to   7.3.7  Adding the Cue
            the goal behaviour.
                                                     Some trainers prefer to add the cue after the
              A common problem with shaping behav-
            iour in this manner results when a trainer   behaviour is performed at a high level of flu-
                                                     ency. In this manner the cue is associated with
            reinforces too many repetitions of one   only the mastered behaviour instead of
            approximation. Performing multiple repeti-  approximations leading up to the goal behav-
            tions of one approximation creates rein-  iour. For instance, teaching a giraffe to walk
            forcement history that can slow progress   into a chute or giraffe restraint devise (GRD)
            and make it more difficult to move from one   may involve shaping the behaviour through a
            approximation to the next. To create the best   baiting strategy involving the trainer holding
            flow of behaviour, a trainer should move to   browse just out of reach of the giraffe and
            the next approximation when the current   reinforcing each step closer to the GRD. After
            approximation is performed without hesita-  several repetitions the giraffe may walk
            tion. With this strategy animals may perform   straight into the GRD without seeing the
            approximations only one time when the    browse because it has learned through its
            trainer and animal are really in sync. If the   reinforcement history to expect the browse
            animal  hesitates,  the  trainer  can  always   once it is fully into the GRD. Now that the goal
            repeat that approximation or even move   behaviour is performed with a high level of
            back in the approximations to gain behav-  fluency, it is time to help the animal under-
            ioural momentum.                         stand reinforcers will only be available for the
              Shaping should involve letting the animal’s
            behaviour determine the size and speed a   GRD behaviour in certain conditions. The
                                                     trainer will  then begin  associating the cue,
            trainer should move through the approxima-  which is generally a visual or audible stimulus.
            tions. Through careful observation, a skilled   The trainer may say ‘Shift’ or point her finger
            trainer can recognise the subtle changes in   straight out to her side as the giraffe enters the
            the natural variation of an animal’s behaviour   GRD to receive a reinforcer. After the giraffe
            and reinforce small increases in movements   comes out of the GRD, the trainer can rein-
            towards a goal behaviour. Other times an   force the behaviour of approaching the trainer,
            animal might offer larger approximations   then cue the GRD behaviour. With each repe-
            that are more obvious to trainers. Either way,   tition the trainer starts a bit farther away from
            the animal’s behaviour combined with the   the GRD until she can give the cue across the
            trainer’s observational and mechanical skills   room and the animal will leave her to walk
            determines the speed of training.        across the room and enter the GRD.
              Animals can learn faster than most train-
                                                      Other trainers prefer to add the cue as they
            ers give them credit for, which is generally   shape the behaviour. For instance, some
            faster than most trainers are comfortable   mouth‐open behaviours are taught using a tar-
            training. Where one trainer might believe it   get on the top of an animal’s nose and then a
            will take a year to teach a primate to accept   second target on the animal’s chin. If these two
            an injection, another trainer might believe it   targets are the trainer’s finger and thumb, the
            will take two weeks. Some animals learn   prompts of spreading the finger and thumb
            faster than others and some trainers teach   gradually farther apart become the cue for the
            faster than others. However, the trainer who   open‐mouth behaviour. Even when shaping a
            believes it will take a year to teach the behav-  turn behaviour for a macaw, a lay down behav-
            iour will almost certainly take significantly   iour for a cheetah, or any other behaviour for
            longer to train the behaviour than the per-  that matter, some trainers add a verbal or hand
            son who believes it can happen in two weeks.   cue during the shaping process.
            Training at the animal’s speed is a valuable   The merits associated with adding the cue
            skill that is developed over time and with   after mastery of a behaviour or during the
            much practice.
   153   154   155   156   157   158   159   160   161   162   163