According to Skinner’s causal model of selection by consequences, human behavior is a
product of three levels of selection: phylogeny, ontogeny and culture. Empiric
investigations of the third level just recently begun in behavior analysis. In the theoretic
field, Glenn introduced the concept of Metacontingency to describe functional relations
between interlocked reinforcement contingencies and an aggregated outcome
responsible for the selection of the interlock. In laboratory, a pioneer work by Vichi,
reproduced a metacontingency using a procedure adapted from experimental studies in
sociology. Vichi suggests that the interlocking behaviors of a small group of people
could be modified by the aggregated outcome produced by the interlock, in this way,
characterizing a metacontingency. The present work is a replication of Vichi’s study,
with the objective to verify if interlocked behavioral contingencies can in fact be
selected by an aggregated outcome contingent to the behaviors of people of a small
group microculture. The participants were eight undergraduate students, divided into
two groups of four, who accomplished a group task. The task consisted in a problem to
solve by choosing a cell in a matrix composed of 8 columns and 8 rows, containing
positive and negative signs. On each trial, the participants chose one row and the
experimenter chose one column. A positive sign in the intersection of the chosen row
and column resulted in gains for the group; a negative sign resulted in losses. The
column chosen by the experimenter was contingent to the way in which the gains were
distributed by the group (equally or unequally) in the immediately anterior trial. In
experimental condition A, the positive sign was contingent to an equal distribution of
gains, and in the experimental condition B, the positive sign was contingent to an
unequal distribution of gains. Group 1 presented 43% of correct choices (the
participants distributed the gains accordingly to the experimental condition imposed),
and the group 2 made 19% correct choices. These results showed that procedures which
use contingent consequences (win or lose in a trial) without contiguity with the
interlock, make it difficult to select such interlock. However, interlocked contingencies
of reinforcement were selected by its aggregated outcome under variables not controlled
in the experiment. This phenomenon can be characterized as an experimental analogous
of a metacontingency. The procedure, possible improvements of the procedure and the
complexity of the experimental task are discussed. Emergent superstitious rule patterns
of behavior are also discussed.
|
