Research
Since switching my operant lab research from rats to humans I have worked primarily on stimulus equivalence.
I also ceased field work on seabirds after twenty years of my animal behaviour field course on Lundy though this was "inherited" by my elder son Thomas, who now edits the Journal of the Lundy Field Society and is actively researching the behaviour of kittiwakes there and at other sites (see also Collaborations)
Does language stem from stimulus equivalence or vice versa?
In stimulus equivalence (SE) trained relations between specific stimulus pairs, such as A1~B1 and B1~C1, can generate untrained pairs, including reversal of stimulus order, B1~A1 and C1~B1 known as symmetry, skipping across the shared stimulus, A1~C1 known as transitivity, and the combined test of both, C1~A1, often just called equivalence. It has proved very difficult to demonstrate symmetry or equivalence in animals other than man. The stimuli in an equivalence class, such as A=B=C, share functional properties. For example, if one, say A1, is established as a discriminative stimulus which must be present if a given new response is to be reinforced, B1 and C1 will automatically acquire the same control over that response without any specific training. Such SE classes resemble words and their meanings such as dog=chien=hund which are grouped with all the stimuli associated with the concept "dog", embodying symbolic representation. Some people think that the fusion of stimuli in an SE class is brought about by the participant naming the stimuli in the first place, and using language or similar symbolic processes to forge the new, experimental SE class in the first place. Others regard the basic processes of SE class formation as fundamental to the subsequent development of language.
We and others have devised various experimental procedures in an attempt to get around this problem of provenance, such as by comparing SE formation using nameable stimuli with stimuli contrived to be very difficult to name.
A new idea is to see if the manner in which trained relations are acquired influences things by presenting the stimulus pairs in the presence of a third stimulus only used in a single trial, and making the task one of spotting the novel stimulus rather than requiring the participant to learn an association between the other two, often repeated together. The test for the formation of symmetric, transitive, and equivalence relations, defining whether or not an SE class had been formed, could be given by asking participants if two stimuli "go together" or not, which has been shown to produce reliable responses in earlier experiments.
There are many experiments of this type still to be done, and various ways, including the measurement of reaction times, and protocol studies in which participants are asked to say aloud nonsense stimuli that come into their heads during learning or testing.
At what age are people able to do symmetry?
Symmetry is defined in Research interest 1, and surprising only humans seem readily able, if taught that "if A1, then B1" as in the typical trained relations supporting stimulus equivalence (SE), to invert this into "if B1, then A1", without specific extra training.
A pigeon would treat the two relations as totally separate: though it could learn either quite easily this would have no impact, positive or negative, on its also learning the other.
At what age does symmetry "kick in" in our species? We (Dr.Colin Bannard, Professor Richard Bentall and myself) are taking advantage both of the "supply" of infants to the Liverpool Language Learning Group, and the availability of eye-tracking equipment to see if we can demonstrate symmetry in children barely out of their first year. A negative result would simply indicate the limits of our procedures, but a positive result would be definitive. With some of each it may be possible to relate the emergence of symmetry to individuals' language development as otherwise logged by colleagues in the LLL group.
Research collaborations
Professor Thomas E. Dickins
Stimulus equivalence, symbols, and the Evolution of Language
Middlesex University
Conceptualising the probable adaptive value of stimulus equivalence formation, which is arguably unique to Homo sapiens, and it role in symbolisation and the evolution of language
Professor Erik Arntzen
Equivalence class formation from sight only of untaught baseline pairs
Oslo and Akershus University College (OAUC), Norway
Studying the derivation of stimulus equivalence classes simply by inspection of untaught baseline overlapping stimulus pairs: discriminating between within-class and between-class test pairs in relation to arrays changing from trial to trial.