Abstract:
The ability for non-human animals to plan for the future – such that they act in the present in
accordance with what will be best in the likely future – is an area of growing research
interest. Evidence suggests that ‘simple’ operant behaviour is future-oriented, implying that
behavioural control is fundamentally prospective. Evidence for future-oriented behaviour in
non-human animals is often surrounded with debate as to whether genuine future planning
abilities are demonstrated. Many demonstrations of non-human animal planning employ
species-specific behaviours, limiting the generality of conclusions. To further our knowledge
of the generality of planning across species, we investigated whether humans could learn to
plan in procedures commonly used with non-human animals, with arbitrary stimuli and
responses. In one experiment we replicated a procedure previously used with crows (Boeckle
et al., 2020) and varied the arbitrariness of the stimuli used with human participants. We
found that humans were successful in learning when trained in the same way as crows,
however they performed worse than crows during testing trials. Stimuli that were more
arbitrary resulted in slightly better performance during testing than stimuli that were less
arbitrary. In another experiment we compared the performance of pigeons and humans on a
task using arbitrary stimuli and responses. A response at the beginning of a trial was required
for a different response to produce a reinforcer at the end of a trial. We found similar patterns
of learning across both species. Both pigeons and humans required experience to learn the
procedure, and both performed better when a correct response at the beginning of a trial
caused an immediate stimulus change than when it did not. Both species learned to make
responses with temporally and spatially distant consequences. Our findings suggest that we
can study planning in humans and non-humans using arbitrary stimuli and responses,
however the arbitrariness of the stimuli and the task used may be critical in determining
whether experience generalizes to novel situations. These findings highlight a need to understand how experience (both within and outside the experiment) underpins planning and
prospective control.
Keywords: planning, future-oriented behaviour, prospective control, arbitrary stimuli