Psychology of Timing Research Paper

The study of time perception was already well developed when William James wrote his textbook, The Principles of Psychology. In a chapter entitled, ‘The perception of time,’ James (1890) provided definitions of subjective concepts such as the duration of the perceived present, and reviewed the extensive psychophysical data on time perception available primarily from European laboratories. With the psychophysical method a person makes a psychological judgment based upon the characteristics of the time interval between two physical stimuli. James describes the use of this method to identify such values as the intervals between two stimuli that are the most accurately perceived, the shortest interval between two stimuli that could be perceived to be two distinct stimuli, and the longest interval between two stimuli that can be perceived as a unit. By the middle of the twentieth century, most research on timing was based on psychophysical research with human participants (Woodrow 1951).

The perception of time was not a main focus of studies of animal learning in the first half of the twentieth century. The first laboratory study of the determinants of the behavior of rats was conducted by Small (1900), and rats became the most frequently used animal in psychological research. In his Handbook of Psychological Research on the Rat, Munn (1950) did not describe any experiment that was explicitly about timing.

But many of the experiments on learning involved variations in time intervals between stimuli, responses, and reinforcements that affected behavior. The results of such experiments have been important in the development of the current understanding of time perception.

1. Timing And Timed Performance

Classical conditioning (also called Pavlovian conditioning) is a procedure in which a reinforcement, such as food, is delivered contingent upon the time of occurrence of a previous stimulus or reinforcement.

The simplest example is one in which food is given at a regular interval and the animal develops an anticipatory response to the time of arrival of the next food. This phenomenon, known as temporal conditioning, was demonstrated in experiments on salivary conditioning of dogs described by Pavlov (1927). If food were given to a dog regularly at 30-min intervals, the dog would begin to salivate shortly before the time of occurrence of the next food delivery. This was objectively determined by counting the number of drops of saliva in each 30-s interval. The importance of such observations was that it demonstrated that time since the occurrence of food, or some biological effects that were correlated with that time, could serve as a conditioned stimulus.

A more typical example of classical conditioning is one in which food is given at a regular interval following the onset of a stimulus, and there is a long irregular interval between successive food deliveries. This also results in an anticipatory response to the time of arrival of the next food, but the time since the onset of the stimulus (rather than the previous food) is the relevant interval. This phenomenon, known as delay conditioning, was also demonstrated in the experiments of Pavlov (1927). It provided additional evidence that time from any discriminable event could serve as a conditioned stimulus.

The temporal control of behavior in classical conditioning has been demonstrated with many different species, stimuli, responses, and reinforcers over a wide range of time intervals. It occurs in a wide range of classical conditioning procedures, including those in which the time intervals are not constant. Variations in the time intervals between stimuli and reinforcers produce corresponding variations in the time of conditioned responses.

Instrumental conditioning (also called operant conditioning) is a procedure in which a reinforcement, such as food, is delivered contingent upon a response, although it may also be contingent upon the time of occurrence of a previous stimulus or reinforcement.

A simple example is one in which food is made available at a regular interval after the last food delivery, and it is delivered following the next response. The animal develops an anticipatory response to the time of availability of the next food. This procedure, known as a fixed interval schedule of reinforcement, was described by Skinner (1938). If food is made available to a rat in a lever box one minute after the previous food delivery, the rat will be increasingly likely to press the lever as the time of food availability approaches (as signaled by the last food). Again, the importance of such observations was that it demonstrated that time since the occurrence of food, or some biological effects that were correlated with that time, could serve as a conditioned stimulus.

There are many other examples of timing in operant conditioning. If food is made available a fixed time after the onset of a stimulus and delivered following the first lever response after it is available, the rat will be increasingly likely to press the lever as the time of food availability approaches (as signaled by the last stimulus onset). Many other operant conditioning procedures with rats, pigeons, and other animals also demonstrate temporal control of behavior. These include the ability of animals to delay responding if only long interresponse intervals are reinforced, the ability to hold down a lever for an interval that corresponds approximately to the duration necessary to secure reinforcement, and many others.

Two procedures that have been used explicitly to study the timing ability of animals are the peak procedure and the bisection procedure. Versions of both of these tasks have been devised for human participants, and the results have supported the conclusion that the same principles of timing apply to humans and other animals (Church 1993).

In the peak procedure, reinforcement is delivered following the first response after a fixed interval of time after the onset of a stimulus on some occasions; on others there is no reinforcement and the stimulus continues for a long interval. Response rate increases as a function of time from stimulus onset until approximately the time that food is sometimes delivered, and then it decreases. The location of the peak of the response rate function provides a measure of the expected time of reinforcement, and the spread of the function provides a measure of the certainty. The peak procedure has been regarded as a temporal production procedure in which responses occur at the time of expected reinforcement.

In the bisection procedure, a stimulus is presented for a short time (such as 2 s) or a long time (such as 8 s), and then two responses are made available (such as a left and right lever). Reinforcement is delivered if the rat presses the left lever following the 2-s stimulus or if it presses the right lever following the 8-s stimulus. Then five or more stimuli of intermediate duration are presented, and responses following these durations are not reinforced. The psychophysical function relating the probability of a right response following each of the stimuli is typically an increasing function with an ogival shape. The duration that correspondence to indifference between the two alternative responses provides a measure of the psychological middle. It is usually approximately at the geometric mean of the extreme durations. The bisection procedure has been regarded as a temporal perception procedure in which the animal reports whether a presented interval was short or long.

Estimates of the absolute duration of short intervals are more accurate than of long intervals. But estimates of the relative duration of short and long intervals are about equally accurate. For example, the error in judgment of a 10-s interval and a 60-s interval is approximately the same if the error is calculated as a proportion of the interval. This is an example of Weber’s law for timing (Gibbon 1977). In peak procedure, the bisection procedure, and many other procedures, both the mean and the standard deviation increase linearly with the temporal interval. Thus the ratio of the standard deviation to the mean, the coefficient of variation, remains constant. A more general principle has been called ‘superposition.’ This refers to the fact that the entire functions relating the behavioral measure to time in the peak procedure, the bisection procedure, and many other procedures are very similar if time, and the behavioral measure, are scaled in relative units (Gibbon 1991).

The ability of animals to time intervals in the range of seconds to minutes is now well established, and the quantitative effect of variations in the conditions on performance in different timing tasks is now well established. These results have provided a guide for the development of psychological explanations about the mechanism of timing. For the evaluation of quantitative psychological models of timing, it would be useful if archives of raw data from timing experiments were available. Such resources are becoming increasingly feasible to establish.

2. Psychological Explanations

A psychological process model of timing and time perception necessarily involves intervening variables. These intervening variables can generally be classified as being related to perception, memory, and decision processes. A formal model of the process consists of rules connecting the input to one or more of the intervening variables, one or more of the intervening variables to the output, and, perhaps, the intervening variables to each other. In a quantitative model of the process, the rules are specified mathematically.

The input to a psychological process model of timing is a physical duration of time that begins with some event, such as the onset or termination of a stimulus, response or reinforcement. This event initiates some internal or behavioral process that changes in a regular way with time, a process that is often referred to as an ‘internal clock.’ Many types of internal representations of physical time have been proposed. One proposal is that a pacemaker emits pulses according to some distribution, the event that begins the timing process closes a switch that permits the pulses to enter an accumulator, and the psychological representation of time is the accumulation of pulses. This leads to an approximately linear relationship between psychological and physical time. Another proposal is that the event begins a continuous, nonlinear process, such as an exponentially decaying memory trace, and the magnitude of the trace provides a nonlinear relationship between psychological and physical time. Other functions to convert physical to psychological time have also been proposed.

The internal clock may consist of multiple processes that are initiated by an event. The multiple processes may be a several periodic or nonperiodic functions with different characteristics that are initiated at the same time, or several processes that are initiated serially. The representation of time with such clocks is usually a vector of heights of each of the functions. There has been no consensus regarding the nature of the internal clock. Any of them can account for the behavioral data qualitatively, and none of them can be ruled out as neurally implausible.

A psychological process model of timing also requires some memory of previously reinforced intervals. For example, the animal must remember that it was reinforced 20 s after the onset of a stimulus, or it must remember that a left lever response was followed by food after a 2-s stimulus. One proposal is that temporal memory consists of an unorganized collection of examples; another is that it consists of a continuum of strengths that correspond with particular times after an event. In addition to a structure, a psychological model of timing requires a rule for storage of new information and a rule for retrieval of information. There has been no consensus regarding the nature of temporal memory based upon either the behavioral or neural data.

Finally, a psychological process model of timing requires some decision about whether or not to respond, or, if multiple responses are available, which response to make. The response decision is based upon a comparison of perception of the current time with the memory of the reinforced time. If the perceived current time is near enough to the remembered time of reinforcement, a response will occur.

Many quantitative models of timing have been proposed that can be described in terms of the three processes of temporal perception, memory, and decision (Church and Kirkpatrick 2000). Some of them are mechanistic, but others consider the animal to be a computational device that operates on symbolic representations (Gallistel 1990). The models differ in their quantitative representation of each of the parts. Scalar timing theory (Gibbon et al. 1984) is based upon a perceptual representation of time that has a pacemaker that emits pulses that are switched into an accumulator, a memory representation of time that is based on an unorganized collection of examples, and a decision that is based on a ratio comparison of the perceptual representation with the a single sample from temporal memory. In contrast, the Behavioral theory of timing and related models (Killeen and Fetterman 1988, Machado 1997) are based upon a perceptual representation of time that is a cascade of functions, a memory representation of time that is based on the strength of each of the functions at the time of reinforcement, and a decision that is based on the product of the perceptual and memory representation. The various quantitative models of timing have been found to produce reasonable fits to the data in different timing procedures, but none of them make accurate quantitative predictions in all procedures. The development of a single quantitative model of timing that makes accurate quantitative predictions for a large number of behavioral measures in a wide range of procedures is essential. Such a model is likely to contain modules for temporal perception, memory, and decision processes that included parts of current theories.

3. Neural Explanations

The perception of time is quite different from the perception of physical stimuli such as lights, sounds, and chemicals. In the case of vision, audition, olfaction, taste, pain, and other sensory systems, there are physical stimuli that impinge upon receptors specialized for particular physical stimuli. But time, like location, is an attribute of any stimulus and there are no known specialized receptors. A perceived time refers to the time since the onset of a physical stimulus, the interval between two physical stimuli, or the duration of a physical stimulus. Thus, it is possible that the psychological explanations that propose general mechanisms for timing are incorrect. Instead, there are specialized mechanisms for the perception of the duration of visual stimuli, auditory stimuli, and others, based upon the unique properties of each modality. One argument in favor of some general processes is based upon the success of the cross-modal transfer experiment. Rats trained to make a temporal discrimination in one modality (such as a 2-s noise vs. an 8-s noise) can be tested on a temporal discrimination in another modality (such as a 2-s light vs. an 8-s light). The test can either use the same response assignments for the short and long stimuli or use a reversed assignment. During testing the rats learned significantly more rapidly with the same assignment of responses to the short and long stimuli than with reversed assignments (Roberts 1982). This demonstrates that there is something general about a 2-s and an 8-s stimulus in different modalities.

Although it is clear that the brain is the biological mechanism responsible for interval timing, the specific neural processes involved are not well understood. To obtain a general understanding of the neurotransmitter mechanisms and their sites of action, the primary methods have been to study the effects of brain lesions and systemic drug administration on behavior in timing procedures. The goal is to obtain some degree of neural and behavioral selectivity. Neural selectivity refers to the specificity of the brain structure or receptors affected, and behavioral selectivity refers to the specificity of the observed effects. With the development of timing procedures with well-established quantitative functional relationships between independent and dependent variables, and quantitative psychological models of the timing process that consist of three or more separate modules, it has become possible to search for biological manipulations that affect one module but not others. On the basis of systemic injection of dopamine agonists and antagonists, the speed of the intvernal clock appears to be controlled by activity of D² dopamine receptors. Similar research with systemic injection of cholinergic agonists and antagonists produced results that suggested that the memory storage appears to be controlled by the level of activity of the cholinergic system (Meck 1996). Other research with neurotoxin lesions has identified the clock effects with structures in the basal ganglia, and the temporal memory and attentional effects with structures related to the frontal cortex.

Bibliography:

1. Church R M 1993 Human models of animal behavior. Psycho- logical Science 4: 170–3
2. Church R M, Kirkpatrick K 2000 Theories of conditioning and timing. In: Mowrer R R, Klein S B (eds.) Handbook of Contemporary Learning Theories. Erlbaum, Hillsdale, NJ, pp. 211–53
3. James W 1890 Principles of Psychology. MacMillan, London, Vol. 1
4. Gallistel C R 1990 The Organization of Learning. MIT Press, Cambridge, MA
5. Gibbon J 1991 Origins of scalar timing. Learning and Motivation 22: 3–38
6. Gibbon J 1977 Scalar expectancy theory and Weber’s law in animal timing. Psychological Review 84: 279–325
7. Gibbon J, Church R M, Meck W H 1984 Scalar timing in memory. In: Gibbon J, Allan L (eds.) Timing and time perception. Annals of the New York Academy of Science 423: 52–77
8. Killeen P R, Fetterman J G 1988 A behavioral theory of timing. Psychological Review 95: 274–95
9. Machado A 1997 Learning the temporal dynamics of behavior. Psychological Review 104: 241–65
10. Meck W H 1996 Neuropharmacology of timing and time perception. Cognitive Brain Research 3: 227–42
11. Munn N 1950 Handbook of Psychological Research on the Rat. Houghton Mifflin, Boston
12. Pavlov I 1927 Conditioned Reflexes. Oxford University Press, Oxford, UK
13. Roberts S 1982 Cross-modal use of an internal clock. Journal of Experimental Psychology: Animal Behavior Processes 8: 2–22
14. Skinner B F 1938 The Behavior of Organisims. Appleton Century-Crofts, New York
15. Small W S 1900 An experimental study of the mental processes of the rat. American Journal of Psychology 11: 133–65
16. Woodrow H 1951 Time perception. In: Stevens S S (ed.) Handbook Of Experimental Psychology. Wiley, New York, pp. 1224–36