Ethical self-control is conceived of as self-controlled responding under concurrent contingencies involving (conflicts of) consequences for the individual, and consequences for the group. The study assessed the production of ethical self-control repertoires in laboratory microcultures under four different macrocontingency arrangements. The experiment was held with eleven groups of four college students each. They were exposed to a task that required a choice of odd or even rows in an 8x8 matrix. Odd rows produced higher individual reinforcements and delayed aversive consequences for the group, thus being labeled impulsive selfish choices; even rows produced lower reinforcements for the individual participant, but positive delayed consequences for the group, thus labeled ethical selfcontrolled choices. Each group was exposed to only one condition. In the first condition, each participant was exposed alone to the task, producing high rates of impulsive selfish choices. In the second condition, the four participants were exposed to the task together, with access to one another's choices and being allowed to talk. The result was a high rate of ethical self-controlled choices. In the third condition, participants were exposed to the task together, could talk, but had no direct access to each other´s choices, which also resulted in a high rate of ethical self-controlled choices. In the fourth condition, participants were exposed to the task together, but could not talk, and had no access to each other´s choices. Results from this experiment show a higher rate of impulsive selfish choices. The data on the four conditions suggest that the possibility of verbal interaction has more effect on the emergence of ethical selfcontrolled responses than access to each other´s responses.

We describe a software program that was developed to study cultural selection using a procedure that is analogous to the free operant. The experimental manipulations allow the use of individual behavioral patterns that are produced by classic schedules of reinforcement to be used as a baseline for experimental demands that are related to social behavior. Two to 4 participants can work simultaneously on the experimental task. Responses on a single manipulandum can lead to different consequences for individual behavior and for participants' coordinated, interlocking behaviors. The software is called "Free Culturant" because participants can emit responses at any time, without intervention from the apparatus or experimenter, and it can assess the selection of cultural units that are composed of sets of interlocking behavioral contingencies that generate a common consequence. We present a pilot study and discuss the software's contributions to advancements in cultural evolution research.

No abstract