(Dis)confirming theories of consciousness and their predictions: towards a Lakatosian consciousness science

The neuroscience of consciousness is accelerating empirically through adversarial collaborations that test rival theories. This paper argues that consciousness science should be paired with confirmation theory, the philosophical study of evidence and hypotheses, to understand how experiments challenge or validate theories. Using Lakatos's philosophy of science, the author proposes a model of theory-appraisal built on three criteria: the distinction between prediction and accommodation, the structural relevance of predictions, and the boldness of predictions. This Lakatosian model offers both normative and descriptive advantages, advancing the debate by considering theories' diachronic development, logical structure, and relationship with background knowledge.

Study at a glance

Abstract

Abstract The neuroscience of consciousness is undergoing a significant empirical acceleration thanks to several adversarial collaborations that intend to test different predictions of rival theories of consciousness. In this context, it is important to pair consciousness science with confirmation theory, the philosophical discipline that explores the interaction between evidence and hypotheses, in order to understand how exactly, and to what extent, specific experiments are challenging or validating theories of consciousness. In this paper, I examine this intricate relationship by adopting a Lakatosian lens. I propose that Lakatos’ philosophy of science can aid consciousness scientists to better interpret adversarial collaborations in consciousness science and, more generally, to develop a confirmation-theoretic model of theory-appraisal in this field. I do so by suggesting that such a model be built upon three Lakatos-inspired criteria for assessing the relationship between empirical evidence and theoretical predictions: (i) the model should represent the ‘distinction between prediction and accommodation’; (ii) the model should represent the ‘structural relevance’ of predictions; (iii) the model should represent the ‘boldness’ of the predictions. I argue that a Lakatosian model of theory-appraisal has both normative and descriptive virtues, and can move the debate forward by acknowledging that theory-appraisal needs to consider the diachronic development of theories, their logical structure, and their relationship with background beliefs and knowledge.