Chris Eddy

My research explores interdisciplinary methods for improving education, in its broadest sense. When considering the capacity to learn as a very general notion, we can apply and study its principles in a variety of different contexts. My overarching research goals are to understand principles of learning, within their multiplicity of contexts, in such a way that I can contribute to the design of learning environments which support long term, conscious, sustainable evolution.

More concretely, I am doing research in various areas of philosophy, (e.g., philosophy of education, pragmatist epistemology, metaphysics, etc.), Habermasian social theory, developmental psychology, embodied theories of cognition, Bayesian inference, cultural evolution, and various subdomains within the complexity sciences.

Three of the questions I’m currently doing research on include:

– What are general epistemic principles of procedural rigor which have been tested and validated in various contexts of inquiry?

Charles Sanders Peirce, in “The Fixation of Belief”, noted that the methods of inquiry and justification we now take for granted, themselves required justification. And those methods gained validity through the experimentation with, and observed failures of, other methods. He lists, e.g., the methods of tenacity, authority, and appealing to a priori reasoning as modes of justification which we have seen fail to account for the claims they attempted to justify . As educators, we must be aware of, act in light of, and improve the methods of inquiry and justification which are currently being used in a community of inquirers.

– What is the relationship between Neo-Piagetian theories of cognitive development and Bayesian Inference models of learning?

Neo-Piagetian developmental theorists, have modeled cognitive development as a series of successive, hierarchical integrations of increasing cognitive complexity. In this paradigm, each conceptual element in a developmental “stage” acts as a coordinating basis for the previous elements while achieving skill proficiency at a lower level is a prerequisite for higher order skill integration. This process occurs as the organism faces (and is able to recognize) problems of increasing complexity. The human learns, and develops, through a process of acting on, receiving feedback from, and mentally modeling the environment.

This view has interesting alignments with research in enactivist cognitive science that is explicitly utilizing Bayesian theories of inference. In some areas of this research, the organism is said to have a certain ‘generative model’ of its environment. For the human, higher order neural structures coordinate the lower level neuron ensembles by sending ‘predictions’ of lower level state on the basis of a generative model. (These predictions can also serve as actions under this view; i.e., I can check the reliability of my generative model by trying to accomplish something.) Then the lower level architectures “send up” prediction errors to help correct the generative model. In this conception, each organism is enacting a cycle of: act on the environment -> receive feedback from the environment -> create an internal model of the environment. Through this cycle, the organism becomes more capacitated to interact in such a way that it can achieve its goals. 

While alignments between Piagetian and enactivist theories in cognitive science have been acknowledged, I don’t believe they have been sufficiently explored. This area of my work brings up very interesting questions in the study of cognitive development, artificial intelligence, and robotics.

– What are the qualities of a socio-cultural system which supports (and hinders) epistemically principled learning?

When we are thinking about learning, it becomes increasingly clear that social interaction is critical to the individual’s development and the evolution of the species. The resulting question is thus: what kind of social interactions result in an individual, and a culture, being epistemically capacitated to act on the basis of its values?

For example, there are a number of cultural mores and institutional policies which hinder the creation of an optimal learning environment. These can be as simple as having a boss who doesn’t want to have their opinion disagreed with. They can also be quite complex, such as the ways in which financial incentivization create contexts in which actors working on critical problems are disincentivized to share information with one another because they are operating in competing organizations. In either case, epistemically principled learning is being hindered due to cultural and/or institutional structures. In order to create learning environments that are commensurate to the problems we are facing in the 21st century, we must be able to identify pathological socio-cultural dynamics and design alternatives which account for them in a holistic way.

Contact Chris at chris.eddy@ronininstitute.org