Topographically well-defined classes of behaviour emitted by developmentally retarded subjects were monitored automatically
or by an observer. Classes of behaviour included lever-pressing
(a descriptive operant) and alternative behaviours (behaviours not
specified in a stimulus schedule). Experiments were performed jn
which various stimuli (e. g. slides, music, sweets, verbal approval,
etc) were presented according to various schedules of stimulation.
The latter included response-contingent, response- noncontingent
and extinction schedules. Several parameters of schedules and
stimuli (type of stimulus, duration of stimulus, schedule
requirements, etc) were varied systematically and the resulting
transition between elements of the behaviour stream were investigated.

In order to facilitate an in-depth analysis, many aspects
of the behaviour stream were examined. Several aspects of lever-pressing
including response rate, duration of post-stimulus pauses,
frequency of pauses in responding, and frequency of responding during
stimulus presentations were recorded. In addition the frequency,
total duration , bout-length, and temporal loci of alternative
behaviours were recorded.
The results revealed functional relationships between the
experimental maniuplations and various dimensions of lever-pressing
and alternative behaviours. Considerable variability was found
between and within subjects, the variability being the result of
complex, but lawful, interrelations bet ween elements of the
behaviour stream.
The experimental analysis of behaviour has traditionally
concentrated on the analysis of relationships between single
responses and other discrete events such as reinforcers. The
present results add to the growing evidence t hat it is necessary
to analyse behaviour in many dimensions and to consider behavioural
phenomena as a continuous stream of interrelated stimulus and
behavioural events. In this context the study of alternative
behaviours may enhance the analysis of behaviour and provide a
basis for the study of the orderly relations involved in the
behaviour of retarded persons even when, as is frequently the
case, such behaviour may appear haphazard.