Michel Treisman

Sensory scaling and the psychophysical law

Sensory scaling is analysed as a sensori-motor task for which two relations must be determined, the first, S = f(I), being the relation between I, the stimulus intensity, and S, a measure of the neural effect of the stimulus, and the second being the relation between the subjects response, R, and S. It is shown that for any scaling procedure the relation between R and S can be arbitrarily specified, and according to the choice made S = f(I) will be found to be either a power function or a logarithmic function; these two “psycho- physical laws” do not reflect empirical differences between scaling procedures but conventional differences in the assumptions made when interpreting these procedures. A model is then elaborated in which S = f(I) is fixed and the relation between R and S depends on the demands of the psychophysical task and the effects of previous learning. With this more complex model the empirical implications of the power function and log function differ; it is shown that if the latter is accepted explanations can be given for two major problems in scaling theory, the disparity between the functions given by category scaling and the direct methods, and the difference between prothetic and metathetic continua.

Predation and the evolution of gregariousness. I. Models for concealment and evasion

Abstract The contribution of predation to the evolution of gregariousness is examined, and it is suggested that the needs of both predator and prey to conceal themselves and to detect the other may have been a major factor directing the development of social behaviour. This implies that it is important to consider the sensory capacities and the strategies open to both sides. Some relevant features of sensory detection are discussed and illustrated by an experiment on visual summation. These features are taken into account in devising two models for the interaction between predator and prey which allow the relative advantages of different dispersion patterns to be assessed. The first model assumes that the preys main resource is concealment: it then appears that when detection will result in only one kill group formation is usually of advantage, but when the predator may destroy all or a proportion of a discovered group, it may be better to be solitary. A second model assumes that the prey have resources for evasion and defence. The occurrence of a kill will then be determined by priority of detection and under this constraint kill rate is lowered by group formation and falls with increasing group size.

The Magical Number Seven and Some Other Features of Category Scaling: Properties of a Model for Absolute Judgment

Abstract A model for absolute judgment is presented. This is derived from a theory of the processes which set and maintain response criteria ( M. Treisman & T. C. Williams, 1984 , Psychological Review , 91 , 68–111) which has been applied to a number of problems in psychophysics. A Thurstonian model is assumed as a basis for absolute judgment, and criterion-setting theory is applied to the criteria in this model. The question is then considered, to what extent can such a model account for the main findings that have been obtained in category scaling? The range of explanations that the model can provide is investigated by means of computer simulation. It appears that features such as the upper limit on information transmission as the number of stimuli increases, the effect of stimulus range on information transmission, the accuracy edge effect, the resolution edge effect, the central tendency of judgment, stimulus and response dependencies, and related phenomena can be reproduced by simulations of the model.

On the Relation Between Time Perception and the Timing of Motor Action: Evidence for a Temporal Oscillator Controlling the Timing of Movement

Studies of time estimation have provided evidence that human time perception is determined by an internal clock containing a temporal oscillator and have also provided estimates of the frequency of this oscillator (Treisman, Faulkner, Naish, & Brogan, 1992; Treisman & Brogan, 1992). These estimates were based on the observation that when the intervals to be estimated are accompanied by auditory clicks that recur at certain critical rates, perturbations in time estimation occur. To test the hypothesis that the mechanisms that underlie the perception of time and those that control the timing of motor performance are similar, analogous experiments were performed on motor timing, with the object of seeing whether evidence for a clock would be obtained and if so whether its properties resemble those of the time perception clock. The prediction was made that perturbations in motor timing would be seen at the same or similar critical auditory click rates. The experiments examined choice reaction time and typing. The results support the hypothesis that a temporal oscillator paces motor performance and that this oscillator is similar to the oscillator underlying time perception. They also provide an estimate of the characteristic frequency of the oscillator.

The internal clock: Electroencephalographic evidence for oscillatory processes underlying time perception

It has been proposed that temporal perception and performance depend on a biological source of temporal information. A model for a temporal oscillator put forward by Treisman, Faulkner, Naish, and Brogan (1990) predicted that if intense sensory pulses (such as auditory clicks) were presented to subjects at suitable rates they would perturb the frequency at which the temporal oscillator runs and so cause over- or underestimation of time. The resulting pattern of interference between sensory pulse rates and time judgments would depend on the frequency of the temporal oscillator and so might allow that frequency to be estimated. Such interference patterns were found using auditory clicks and visual flicker (Treisman & Brogan, 1992; Treisman et al., 1990). The present study examines time estimation together with the simultaneously recorded electroencephalogram to examine whether evidence of such an interference pattern can be found in the EEG. Alternative models for the organization of a temporal system consisting of an oscillator or multiple oscillators are considered and predictions derived from them relating to the EEG. An experiment was run in which time intervals were presented for estimation, auditory clicks being given during those intervals, and the EEG was recorded concurrently. Analyses of the EEG revealed interactions between auditory click rates and certain EEG components which parallel the interference patterns previously found. The overall pattern of EEG results is interpreted as favouring a model for the organization of the temporal system in which sets of click-sensitive oscillators spaced at intervals of about 12.8 Hz contribute to the EEG spectrum. These are taken to represent a series of harmonically spaced distributions of oscillators involved in time-keeping.

Temporal Rhythms and Cerebral Rhythmsa

A model for the internal clock is briefly described. It includes a temporal pacemaker whose rate determines time judgments, and whose frequency is affected by arousal specific to it. Three hypotheses relating time judgments and the alpha rhythm are considered: (a) They may be wholly independent, each reflecting the specific arousal of the mechanism determining it. (b) The alpha rhythm may be an index of a state of general arousal which also acts on the temporal pacemaker. Because of this common influence, the alpha frequency, and the proportion of alpha in the electroencephalogram, may be correlated with the speed of the temporal pacemaker. (c) The same pacemaker may be common to the internal clock and an alpha rhythm generator. Concurrent observations on alpha frequency, alpha prevalence, and temporal productions show that there are no simple relations between these measures such as might support the general arousal or common pacemaker hypotheses. However, relations are found between the variables. More or less regular oscillations occur in their values, some of which are common to two or more of the variables studied. These phenomena are further investigated and described.

Signal detection theory and Crozier's law: Derivation of a new sensory scaling procedure

Abstract The relation between sensory thresholds and the “sensory scale” is examined in the light of the threshold model given by signal detection theory. The problem is seen as that of determining the function relating E , the central effect of a stimulus serving as the decision axis when threshold judgments are made, and I , the physical intensity of the stimulus. Consideration of the model shows that when deviations from Webers law occur they may be accompanied by departures from Croziers law and that the occurrence and extent of the latter depend on the relation between E and I and can be used to define it. The argument is applied to some of the data on visual brightness discrimination in the literature. The results suggest that E may be related to I by a power function with an exponent of the order of 1.0. The disparity between this function and the psychophysical laws given by classical scaling procedures is discussed, and it is suggested that it may prove useful to hypothecate a distinction between central discriminal and metric processes.

Predation and the evolution of gregariousness. II. An economic model for predator-prey interaction

Abstract An earlier paper presented models which allow the relative advantages of different prey dispersions to be assessed for two cases: when the preys main defence is concealment, and when prey may evade the predator (Treisman 1975). With these as a basis, an economic model for predator-prey interaction is developed which takes explicit account of the conflicts between the need to maintain watch for a predator and the preys other needs. The model embodies arguments based on statistical decision theory which allow factors such as the degree of predation threat, the watchfulness of the prey and the number of watchers to be related to the expected value of the preys performance. The model provides a basis for predicting the optimal number of watchers or optimal group size under various conditions. The advantages of different group configurations are compared, and it is shown that animals which aggregate (and maintain a high sensory decision criterion) will have an advantage over solitary animals under certain conditions.

Bird song dialects, repertoire size, and kin association

Abstract The possession of large repertoires of songs by some species of songbirds is offered an explanation arising from a model of the evolutionary restraints which restrict aggression within a species (Treisman 1977). This showed that animals which settle near kin may obtain a selective advantage, provided both sides can recognize their relationship in the sense that they apply appropriately modified aggressive strategies. To ensure this recognition a genetic marker will not suffice. Two factors may favour such kin association: (a) there may be in some species a tendency to prefer territories near the place of hatching, (b) further guidance on where to settle may be provided by the observation of learned characteristics of the behaviour of neighbours where these are associated with place of origin (and therefore with relatedness). Dialect variations in song may serve this purpose, providing an indicator which can vary continuously in two dimensions and in relation to which a bird may position itself at the point of maximum familiarity. The greater the potential behavioural variability, the greater will be the scope for conveying this information by developing dialect variations. This advantage would lead to selection for the acquisition and expression of large repertoires of songs.

Changes in threshold level produced by a signal preceding or following the threshold stimulus

In a previous experiment (Howarth and Treisman, 1958) it was shown that when a warning preceded a stimulus by a fixed interval, the threshold level for the stimulus was higher the longer the fixed interval. A model of this effect was proposed, assuming that the threshold criterion adopted by the subject at any moment was modified by his estimate of the probability of a stimulus at that moment. Three predictions are derived and tested. It is found that: the standard deviation of the response is constant despite the shifts in threshold induced; the threshold continues to fall as the interval from warning to stimulus is decreased to zero, with the lowest threshold when the warning signal and the stimulus are simultaneous; and that a warning signal following the stimulus by an interval of less than 0 5 to 1 sec. also lowers the threshold.