Abstract: The effect of observational learning on the acquisition of an operant response was examined in eighteen riding horses and ponies. The test horses were randomly divided into three groups of six and individually exposed to one of three treatments. An additional horse was trained as a demonstrator, to perform the operant response. The observer horses watched either the demonstrator performing the bin-opening response (Group D+B); the demonstrator standing passively (Group D); or the operant bin in the absence of the demonstrator (Group B). Observers had access to and were free to interact with an identical bin during testing. Observers in Groups D+B and D were socially familiar with the demonstrator. Each test horse was tested once a day for 10 days. An ANOVA revealed no significant differences between treatment groups in the number of responses or the time taken to reach the learning criterion. However, there were highly significant differences between breed types, with non-warmbloods performing more bouts of opening the bin and feeding (p=0.02), feeding from the bin sooner (p=0.01) and reaching the criterion for learning sooner than warmbloods (p=0.05). There was also a significant negative linear relationship between horses' ages and time spent investigating the bin, with younger horses performing more investigative behaviour (y=-3.08x+106.86; p=0.02).

Abstract: This experiment examines the cues used by cattle to discriminate between people, particularly the role played by facial cues. We trained and tested eight Holstein cows 5 days each week for 2 months. For each cow, we used two people, a rewarder and a non-rewarder, of different size and dressed in overalls of the same colour. The operant chamber was a large box within which stood the two people. The cow could see, smell and touch each person. A lever was placed in front of each person. When the cow pushed the lever in front of the rewarder, it received 75 g of concentrate and nothing when it pushed on the other one. For each test session, the cows made 10 choices. The placement of the people was determined randomly according to the Gellerman series. The success criterion was defined as at least eight correct choices out of 10 trials for two consecutive sessions (binomial law P<0.003). During the shaping, seven cows out of eight learned to press the lever to obtain the food. The cows were then tested in a series of 10 trials with only the rewarder present. Seven out of seven cows succeeded in reaching the success criterion. In experiment 1, both the rewarder and the non-rewarder were present and standing upright at normal height and in full view of the cow. Five out of seven cows achieved the success criterion. In experiment 2, the cows could see only the faces of the two people. None of the cows were able to reach the success criterion. In experiment 3, both people were present standing up and wearing identical masks that completely covered their heads. Five cows out of five achieved the success criterion. In experiment 4, we changed the relative height of the people. Five cows out of five succeeded when the two people stood so they were of equal height but with their faces visible. However, no cows succeeded when the people were both of equal height and had their faces covered. This study suggests that cows seem to use multiple cues to discriminate between people. Cows appear able to use either body height or the face to discriminate between people but use of the face alone is more difficult when the cows cannot see the rest of the body.

Abstract: Using horses, we investigated three aspects of the stimulus control of lever-pressing behavior: stimulus generalization, discrimination learning, and peak shift. Nine solid black circles, ranging in size from 0.5 in. to 4.5 in. (1.3 cm to 11.4 cm) served as stimuli. Each horse was shaped, using successive approximations, to press a rat lever with its lip in the presence of a positive stimulus, the 2.5-in. (6.4-cm) circle. Shaping proceeded quickly and was comparable to that of other laboratory organisms. After responding was maintained on a variable-interval 30-s schedule, stimulus generalization gradients were collected from 2 horses prior to discrimination training. During discrimination training, grain followed lever presses in the presence of a positive stimulus (a 2.5-in circle) and never followed lever presses in the presence of a negative stimulus (a 1.5-in. [3.8-cm] circle). Three horses met a criterion of zero responses to the negative stimulus in fewer than 15 sessions. Horses given stimulus generalization testing prior to discrimination training produced symmetrical gradients; horses given discrimination training prior to generalization testing produced asymmetrical gradients. The peak of these gradients shifted away from the negative stimulus. These results are consistent with discrimination, stimulus generalization, and peak-shift phenomena observed in other organisms.

Abstract: We present evidence that a dog (Philip, a 4-year-old tervueren) was able to use different human actions as samples against which to match his own behaviour. First, Philip was trained to repeat nine human-demonstrated actions on command ('Do it!'). When his performance was markedly over chance in response to demonstration by one person, testing with untrained action sequences and other demonstrators showed some ability to generalise his understanding of copying. In a second study, we presented Philip with a sequence of human actions, again using the 'Do as I do' paradigm. All demonstrated actions had basically the same structure: the owner picked up a bottle from one of six places; transferred it to one of the five other places and then commanded the dog ('Do it!'). We found that Philip duplicated the entire sequence of moving a specific object from one particular place to another more often than expected by chance. Although results point to significant limitations in his imitative abilities, it seems that the dog could have recognized the action sequence, on the basis of observation alone, in terms of the initial state, the means, and the goal. This suggests that dogs might acquire abilities by observation that enhance their success in complex socio-behavioural situations.

Abstract: “Clicker training” is a popularly promoted training method based on operant conditioning with the use of a secondary reinforcer (the clicker). While this method draws from theories of learning and is used widely, there has been little scientific investigation of its efficacy. We used 60 horses, Equus callabus, and assigned each horse to one of six reinforcement protocols. The reinforcement protocols involved combinations of reinforcers administered (primary versus secondary plus primary), schedule of reinforcement (continuous versus variable ratio), and reinforcers applied during extinction (none or secondary). There were no differences (P>=0.11) between horses which received a secondary reinforcer (click) followed by the primary reinforcer (food) and those which received only the primary reinforcer (food) in the number of trials required to train the horses to touch their noses to a plastic cone (operant response). There also were no differences (P>=0.12) between horses which received the secondary reinforcer plus primary reinforcer and those which received only the primary reinforcer in regards to the number of trials to extinction. We conclude that there is no difference in the amount of training required to learn the operant task or in the task's resistance to extinction between horses that received a secondary reinforcer followed by a primary reinforcer versus horses which received only a primary reinforcer.