|
Description
|
Interval timing, the ability of organisms to discriminate durations ranging from seconds to minutes, is involved in many cognitive activities, from speech perception to learning, and may be linked to pathologies such as Parkinson's disease. Remarkably, the processes underlying interval timing share common properties across a wide range of species (Richelle & Lejeune, 1980), a fact that hints at the fundamental adaptive value of the ability to “tell time”. Several quantitative models have been proposed to account for interval timing. The most prominent is Scalar Expectancy Theory (SET). According to SET, a pacemaker-accumulator unit induces linear representations of time with scalar variability; one or more memory units store relevant intervals and durations, and a comparator yields the response output. The model’s influence on psychology and neuroscience has been so pronounced that it qualifies as a strong Null Hypothesis. More recent models include the PI’s Learning-to-Time (LeT) model and the Behavioral Economic Model (BEM). Stemming from an associative tradition, both models assume a series of time-dependent states that become coupled with responding through learning. They differ in the temporal dynamics of the states, linear in LeT, logarithmic in BEM. Given the positive results obtained by both models in their initial tests, they qualify as plausible Alternative Hypotheses. The primary goal of the current project is to contrast these three leading models of timing. Because the project will involve experiments with both humans and animals (pigeons), a secondary goal is to better understand the between-species similarities and differences in the basic processes of temporal regulation. To achieve those goals, we will use a variation of the classical psychophysical method of constant stimuli, the bisection task. A subject learns to choose one response following a short signal (S) and another response following a long signal (L). The responses are named Short and Long, respectively. Next, the subject is presented with novel signal durations and its choices are recorded. The function relating the proportion of Short choices to the duration of the test signal is the psychometric function, and the test duration for which the subject is indifferent between the two choices is the Point of Subjective Equality (PSE). Mathematical analyses revealed that the three models make different predictions regarding how the psychometric function and the PSE should vary with a) the payoff probabilities following the responses, b) the relative frequencies of the S and L signals, and c) the interaction between these two variables and the absolute values of S and L. Moreover, d) the models also predict different outcomes of a signal detection analysis of bisection data. When we conceive of, say, the L duration as signal and all the other durations as noise, we can define hits and false alarms and plot zROC curves. Under specific conditions, the models predict different slopes and intercepts for these zROC curves. We propose 6 studies, 3 with humans and 3 with pigeons, to test these predictions. The proposed studies will fill a major gap in the empirical domain of timing because we do not currently know how the preceding variables affect temporal discrimination. This missing knowledge is also important knowledge because it is knowledge that will help us to build better theories and models of timing. In fact, by contrasting the experimental results with the models' predictions we hope to identify the model components that deserve credit for the successful predictions or blame for unsuccessful ones. Retaining the former, eliminating the latter, and, when needed, imagining new model components, are the means we will use to move us further forward in the theoretical domain of timing. The project combines sophisticated theoretical analyses with straightforward empirical tests, and that is its major strength. It proposes a new approach to test the assumptions concerning temporal representation and decision rules of three leading models of timing, and that is its major novelty. Hence, the project is likely to advance significantly our understanding of the time sense of animals and people, and that may prove to be its major contribution.
|
