SymBa: Symmetry in Baboons

Post-doctoral project "SymBa: Symmetry in Baboons"

-- Last update July 2022 --

Led by Thomas Chartier, advised by Joël Fagot
Funded by Marie Sklodowska-Curie Actions H2020 (Grant n°899068)
Project period: May 2020 - July 2022 
Official web page: https://cordis.europa.eu/project/id/899068 

 

Goals

In this 2-year project, we had proposed to try to resolve a paradox. On the one hand, non-human animals such as monkeys or birds are known to be endowed with advanced skills at high cognitive levels (metacognition, decision making, theory of mind, etc.). On the other hand, the same animals are thought to have a basic incapacity in a low cognitive levels, that of encoding bidirectional associations between simple stimuli - the capacity is called "Associative Symmetry". In contrast, humans appear to be proficient in associative symmetry and it is thought that this capacity confers them with an enhanced ability to acquire language, notably for the mental bidirectional (=symmetrical) relations between words and the objects/actions they refer to. Such a situation appears untenable scientifically and may represent a big deficiency in our understanding of how cognition, and brains in general, work. 

We set out to conduct a series of behavioral experiments in a non-human model species (the Guinea baboon), taking advantage of our world-unique computerized setup for automatized experiments, located in Rousset close to Marseille. In complement, we decided to conduct behavioral experiments on human participants in the lab as well as on the Internet, again taking advantage of the latest technology for automatized computerized data acquisition. The goal of these experiments was to test whether in baboons associative symmetry is really absent, and whether in humans it is as robust as generally reported. If both our hypothesis were true, this would call for reconsideration of the alleged cognitive gap of "Associative Symmetry" and question our understanding of human specificities. Next to empirical work, our aim was to also critically examine the evidence available in the human literature, as well as potentially come up with new theoretical propositions to integrate or findings in the existing data and theory. Success with this agenda would help resolve contradictions in our accounts of cognitive capacities in species other than humans, and contribute to improving our understanding of the basis of language acquisistion in humans.

Work carried out

To address the comparative question of symmetry, we have conducted a series of computerized, behavioral experiments in both baboons and humans. Of these, experiments 1, 2, 3 have tried to demonstrate the existence of associative symmetry in baboons (#1 was successful), experiment 4 tested humans and baboons in the same behavioral task to provide a fair comparison of cognitive performances, experiment A in humans investigated whether symmetry may depend on the behavioral task employed, experiment B in humans addressed the role of naming as a potential mediator of associative symmetry.

In parallel, bibliographical and theoretical work has been carried out. This included a vast survey of human experiments on symmetry (no such review existed before), and two further theoretical contributions (1 adressing the link between symmetry and language, the other developping a new theory to replace the concept of symmetry.

This 2-year project has provided the opportunity to train several students: 2 Master students from ENSAI Rennes (a reknown French college for Statistics), 3 Bachelor students from Aix-Marseille Université (1 of which was then recruited as data engineer during the project)

The quality of the work conducted during the project is attested in several ways.
At least six scientific articles have originated / will originate from the project: 3 experimental ones, 2 theoretical ones, 1 review (see list of publications below). Two of the supervised students (Dahmani, Sabatier) are co-authors in one of the article being prepared.
The work has led to presentations in major international scientific conferences: Naturally and Artificially Intelligent Systems 2020 (NAISys, Cold Spring Harbour, USA), Psychonomics Society 2021 (USA/virtual), Interdisciplinary Advances in Statistical Learning 2022 (San Sebastian, Spain), European Society for Cognitive Psychology 2022 (ESCoP, Lille, France). The work has also been presented in other institutes or workshops during several invited talks (Chunked ANR consortium, Grenoble, France, January 2022; Center for Interdisciplinary Research, Paris, France, May 2022; Institute for Language, Communication and the Brain, Marseille, France, June 2022; Center for Research on Animal Cognition, Toulouse, France, June 2022).

A final workshop has been organized to gather researchers interested in the project's topic. The 3-day, highly interdisciplinary event, called "From Associations to Cognition" was held in Marseille in July 2022, and gathered about 50 persons, including international speakers (Mexico, UK, USA). We managed to bring together experts and students from the fields of experimental psychology, modelling, linguistics, neurosciences, animal behavior and child development. It has been a great success and a friendly event, as well as an occasion to adequately disseminate the project's results through two oral presentations and one poster presentation.

Next to these core scientific activities, communication actions towards the general public have been engaged - though the plans were greatly disturbed by the Covid situation. A "general scientific knowledge" talk in front of the 2020 class of civil engineers at ENTPE in Lyon, France (a reknown French college for civil engineering - about 100 students) and in front of a lay audience in my hometown Rodez, France, within Rodez' Free TIme University (about 100 persons).

Progress achieved

A major contribution of the project is the reassessment of a long-held assumption in the field of cognitive sciences: humans would cognitively differ from non-humans in the so-called "associative symmetry" capacity, which would confer humans their proficiency in forming interchangeability relations between the words and the objects/actions they refer to. Empirical, bibliographical and theoretical contributions have allowed to cast doubt on the reality of this dogma, and propose alternative theoretical points of view to reconcile data and theory. Our conclusion so far is that non-human primates, at least baboons, do not seem to lack this crucial cognitive capacity. Moreover, we have shown that associative symmetry is far from being a hallmark of human cognition as generally stated. Our proposed vision of associations encompasses sequential information and statistical co-occurences in a dual account. Associative learning envisaged this way is more compatible with models of machine learning, of human hebbian learning, and with existing knowledge on neurobiological correlates of learning.

A wider implication of our work is that this proposition, which we are further developping, has the potential to reconcile 2 major historical schools of psychology, Behaviorism & Cognitivism, that have been opposed since the late 1950s and the Skinner/Chomsky famous dispute. This was also the motivation behind our final workshop, that aimed at bridging the gap between associations (behaviorism) and cognition (cognitivism).

List of publications

Chartier, T. F., & Fagot, J. (2022). Simultaneous learning of directional and non-directional stimulus relations in baboons (Papio papio). Learning & Behavior, 1-13. https://doi.org/10.3758/s13420-022-00522-8
Chartier, T. F., & Fagot, J. (2022). Associative symmetry: a divide between humans and nonhumans?. Trends in Cognitive Sciences, 4(26), 286-289. https://doi.org/10.1016/j.tics.2022.01.009
Chartier, T. F., & Dautriche, I. (submitted). Do backward associations have anything to say about language? 
Chartier, T. F., Tosatto, L. & Fagot, J. (in prep) Intersubject variability on bidirectional associations tasks in baboons and humans
Chartier, T. F., Dahmani, L., Sabatier, M., Rey, A., L. & Fagot, J. (in prep) Unsuspected variability and task-dependency in a purported hallmark of human cognition
Chartier, T. F., Rey, A., Tosatto, L. & Fagot, J. (in prep) Resolving the contradictions of unidirectiona lstimulus associations