"The analogue laboratory," Hoffman writes, "gives students some firsthand experience in conditioning research. Under ordinary circumstances students have only limited access to live-animal experimentation. They read about experiments in conditioning and they may even observe portions of an experiment, but time in the full-scale laboratory is ordinarily scheduled only for qualified research workers.

"Before undertaking the construction of an analogue laboratory it is helpful to examine the problems Pavlov confronted at the beginning of his brilliant experiments. His initial observations were basically simple: when food is introduced into the mouth of a hungry dog, the animal begins to salivate. Pavlov also observed that occasionally the flow of saliva begins prior to the actual insertion of food into the dog's mouth. These 'spontaneous' secretions were exceptions to his previously formulated account of the mechanisms responsible for the digestive activity of dogs. Moreover, they suggested a technique for the experimental analysis of fundamental mechanisms involved in learning.

"Pavlov reasoned that certain unspecified stimuli were serving to evoke the salivary reaction. If this was indeed the case, it appeared that such stimuli had gained control over salivation by some process as yet little understood. Perhaps the sight of the experimenter or the sound of his movements reminded the dog of food. The explanation seemed obvious, and to a man of lesser scientific stature than Pavlov it might have been acceptable. To Pavlov, however, an explanation of this sort was both deceptive and distasteful. It was deceptive because it really explained nothing. To assume that the dog was 'reminded of food' was to substitute one mysterious process for another. The explanation was distasteful because it implied the action of mechanisms that were unlikely to provide a quantitative account of these puzzling responses. As a scientist Pavlov asked: 'What stimuli cause the salivation? How do they acquire their control of the behavior and what quantitative relations are involved?' He recognized that under normal circumstances the stimuli were constantly shifting and changing and could not be identified by an undisciplined search. To simplify the problem, Pavlov designed an experiment to control such variables as light, sound and other factors in the environment as well as the availability of food. The thoroughness of his preparations is indicated by the fact that when the dog in the experimental chamber was loosely restricted for 10 minutes, all spontaneous salivation ceased; the dog salivated only when food (in the form of a meat powder) was presented. In short, Pavlov had succeeded in gaining control of the animal's spontaneous secretions and could begin his analysis of the responsible mechanisms.

"The analogue laboratory [see illustration below right] was conceived to represent the control room of Pavlov's laboratory. When the button marked 'Meat powder' is depressed, for example, the meter on the face of the instrument promptly indicates an appropriate salivary response as though real meat powder were being presented to a living dog in a full-scale laboratory. The operation of the 'Tone' switch and those marked 'Green' and 'Yellow' simulate the presentation of a tone or of colored lights. As in the case of the meat-powder button, the operation of these controls evokes simulated responses that would be expected if equivalent stimuli were presented to a living animal.

"The apparatus is based on the electrical characteristics of circuits composed of resistances, capacitors and switches. The rate at which a capacitor acquires charge when it is connected to a battery or to a previously charged capacitor depends principally on its size and the resistance of the circuit, just as the time required to fill a container with water varies with its size and the resistance presented to the inflow by the plumbing. In the case of capacitors the charging rate is maximum at the beginning and diminishes as full charge is approached. Similarly, capacitors discharge at maximum rate when first connected to a load circuit; the rate diminishes as discharge approaches completion. Such nonlinear characteristics are observed in aspects of animal behavior. For example, a dog takes disproportionately greater interest in food at the beginning of its meal than it does after it has eaten. Good capacitors can also retain charge, much as the brain stores sense impressions, and can therefore be made to function as memory units. In the analogue laboratory an appropriate pattern of such characteristics has been built into a circuit that simulates the behavior of Pavlov's dog and the remote-control apparatus [see below left].

"The design represents a compromise between practical and theoretical considerations. The object was to devise an instrument that would duplicate a wide range of phenomena associated with conditioning, and to keep the cost low. The result is an apparatus that performs well, is inexpensive, rugged and easy to build. It will operate continuously for about three years on a single dry battery. The operation of the circuit centers in the capacitors C7 and C8, which charge continuously through the resistor R9. When the meat-powder switch is operated, C7 discharges through the meter and C8 through R6. When either a light switch or the tone switch is operated in conjunction with the meat powder switch, C7 discharges through the meter and the charge on CS is transferred through R6 to the capacitors controlled by that switch. Once a charge has been placed on the capacitors that are associated with a given switch, the experimenter can discharge the capacitor to the right through the meter by closing the switch. When the switch is released, the capacitor to the left discharges into the capacitor to the right until both attain equal charge.

"The operating procedure requires a watch equipped with a second hand, a notebook and graph paper for tabulating and analyzing results. With these materials at hand, depress the 'Clear' switch (S5) in conjunction with switches S1, S2 and S3. This simulates placing a naive dog in the experimental chamber. Then set the 'Hours of deprivation' dial to 10. This establishes the number of hours the simulated dog has been deprived of food. Now depress the meat-powder button. Note the prompt response of the meter. Record the peak amplitude of the response, arbitrarily taking the meter reading as the number of 'drops' of saliva. When the amplitude begins to fall, release the switch and after a few seconds 'present' the meat powder again. Observe that the second exposure to meat powder evokes somewhat less response. Next operate the tone switch and observe that no response follows. The sequence indicates that meat powder is an unconditioned stimulus for salivary response. It evokes salivation even though the experimenter has not trained the dog. Note also that the repeated presentation of meat powder to even a hungry dog-in this case one that has been deprived of food for 10 hours produces successively smaller responses. One would not expect the tone to evoke a response because in this sequence we are working with a dog that has not been conditioned to associate the sound of a bell with food. When plotted, the results resemble the top curve in the accompanying set of three schematic graphs [below right].

"The results differ when the tone is sounded a fraction of a second before meat powder is presented and continued so that both stimuli overlap in time. This sequence is shown in the second graph. The initial salivary response again appears with the presentation of meat powder. But the effects of the paired stimuli appear with the second trial. Tone now evokes salivation even before the meat powder is presented. This reaction becomes increasingly pronounced as the experiment continues. It is a conditioned response, one that will last for an appreciable period. In humans conditioned responses have persisted for at least 16 weeks and in dogs for more than two years.

"In the case of the analogue laboratory, time alone causes little reduction in the strength of a conditioned response, typically a decline of 10 per cent is observed after the apparatus has remained idle for 12 hours. As with living organisms, the strength of a conditioned response in the analogue laboratory can be reduced (extinguished) by withholding the unconditioned stimulus and repeatedly presenting the conditioned response, in this case by presenting the tone repeatedly in the absence of meat powder, as shown in the third graph. Observe that each presentation of tone evokes a diminished response. These reactions are similar to those observed in living organisms.

"To obtain an overview of the conditioning process clear the apparatus and, with the dog 10 hours hungry, depress first the tone button and then the meat button as in the first experiment. After 15 seconds depress the tone button again; when the conditioned response reaches peak amplitude, record it, quickly depress the meat button, hold both buttons down until the unconditioned response peaks and quickly release both buttons. Repeat every 15 seconds for eight additional trials. Plot the results. Then extinguish the conditioned response: depress the tone button alone until peak amplitude is reached and record that peak. Do this for 10 trials and plot the results. The acquisition and extinction plots should resemble the large curves in the accompanying graphs [above]. The colored curves placed beside them for comparison give the results of conditioning experiments with human subjects. In the experiments with humans a tone was paired with an aversive stimulus (a mild electric shock). The response consisted of changes in the electrical resistance of the skin, a phenomenon known as galvanic skin response. It is a sensitive index of emotional reactions that interests psychologists because it sheds light on the mechanisms underlying the learning of fear and anxiety. Observe that with the analogue laboratory, as in the experiment with humans, the largest effects during both conditioning and extinction occurred during the early trials.

"The analogue laboratory enables one to investigate the variables that affect the acquisition of a conditioned response. Experiments with the instrument show that the timing between the presentation of the two stimuli (the 'interstimulus interval') and the interval between presentations (the 'intertrial interval') [shown in color in middle graph in Figure 4], as well as the duration of the stimuli and the concurrent motivational level (the hunger of the dog), all affect conditioning substantially. Similar effects have been observed in experiments with live subjects. In general the instrument yields reproducible results when a given set of operations is repeated. Certain effects can be obscured by random errors in timing and recording, however. This is an advantage when the apparatus is used as a teaching device: it requires the experimenter to decide whether or not a given observation represents a fortuitous configuration of random errors, and it requires him to devise experiments that will reliably eliminate such results.

"One approach to coping with random error involves averaging data from identical experiments with several groups of subjects. For example, the effect of altering the intertrial interval can be explored by running an experiment on three groups of dogs. The analogue laboratory is cleared and set for '10 hours hungry,' and a series of runs is made with meat and tone for the first group of animals at an intertrial interval of three seconds. Identical runs are then made with groups at intertrial intervals of 10 and 20. seconds. If the results are carefully tabulated and plotted, the curves will resemble those of the accompanying graph [Figure 6]. The colored inset graph depicts the behavior of humans in a comparable experiment that required the subjects to learn to hold a stylus on a small disk fastened near the edge of a rotating turntable. In spite of gross differences in procedure the data are strikingly similar. Both experiments tell the same story: Spacing the trials facilitated learning, whereas massing the trials interfered.

"The device can also be used to explore the effects of the intertrial interval on the course of extinction, as shown in another graph [below]. The inset depicts the extinction of conditioning in humans, in this case a conditioned eyelid response. In both experiments the massing of trials appears to have speeded the process of extinction.

"What happens when a rest period is injected into a period of concentrated learning? Typically it exerts a facilitating effect, called 'reminiscence.' As can be observed with the analogue laboratory, a rest interval after a conditioned response has reached maximum intensity produces an initial rise in the amplitude of the conditioned response during subsequent conditioning trials, followed by a gradual decline to the pretest level. The effect is most pronounced when trials are massed and increases as the duration of the rest is increased. When a rest interval is introduced following a series of extinction trials, it is frequently observed that the response exhibits partial recovery. In general the magnitude of this phenomenon, known as spontaneous recovery, is influenced by the previous level of conditioning, the number of prior extinction trials and the length of the recovery interval.

"If a green or yellow light is used in place of the tone as the conditioned stimulus, the results remain virtually identical. Once a stimulus has acquired the capacity to elicit a conditioned response, however, it can be used to bring the response under the control of a second stimulus. This phenomenon is called higher-order conditioning. It was first described by Pavlov in 1927. Experiments on higher-order conditioning can be conducted by developing a conditioned response to tone and then using the tone, rather than meat powder, to condition a response to one of the lights.

"When a given stimulus has acquired the capacity to evoke a given response, stimuli like it frequently evoke the same response. This phenomenon, called stimulus generalization, was also described by Pavlov in 1927 and has since become the subject of numerous experiments. It can be examined with the analogue laboratory by conditioning a response to a given stimulus and then testing the response by presenting one of the other stimuli. In the analogue laboratory stimulus generalization occurs between the lights but not between the tone and the lights. This is consistent with Pavlov's observation that stimulus generalization can usually be expected between closely related stimuli.

"These experiments illustrate the variety and scope of research experience that can be provided by the analogue laboratory, but they do not exhaust its possibilities. Anyone who builds the device is likely to find his greatest satisfaction in designing, conducting and analyzing his own experiments. Since research in psychology (or in any scientific discipline) never proceeds best in an intellectual vacuum, it is well to do some reading in the field. Periodicals such as the Journal of Experimental Psychology, Comparative and Physiological Psychology and Experimental Analysis of Behavior are strongly recommended, as are the references listed in 'Bibliography'. Above all, do not make the mistake of assuming that an experiment run with the analogue laboratory is equivalent to one performed with an animal in a full-scale laboratory, and do not allow simulated results to strengthen or weaken confidence in conclusions based on conventional research. Any differences between 'real' results and simulated data serve to remind one that the analogue laboratory is not a perfect replica of reality."

Bibliography

BEHAVIOR THEORY AND CONDITIONING. Kenneth W. Spence. Yale University Press, 1956.

CONDITIONED REFLEXES: AN INVESTIGATION OF THE PHYSIOLOGICAL ACTIVITY OF THE CEREBRAL CORTEX. I. P. Pavlov. Oxford University Press, 1927.

PRINCIPLES OF BEHAVIOR. Clark L. Hull. Appleton-Century-Crofts. Merideth Publishing Co., 1943.

The Society for Amateur Scientists (SAS) is a nonprofit research and educational organization dedicated to helping people enrich their lives by following their passion to take part in scientific adventures of all kinds.