Philosophy

Artificial Linguistic Mind and Natural Linguistic Mind: From Classical Cognitivism to Embodied Models The emergence of latest generation AI systems has revitalized the longstanding debate concerning the affinities and divergences between human and artificial intelligence, with particular emphasis on linguistic and communicative competencies. Whereas traditional symbolic AI—often referred to as GOFAI (Good Old-Fashioned Artificial Intelligence)—implicitly or explicitly drew upon the representationalist framework of classical cognitivism (remaining faithful to the mind-as-computer metaphor), contemporary AI systems align with an anti-representational conception of mind. Rather than operating on predefined structures and rules formulated by linguistic theorists, these systems "learn" autonomously from vast textual corpora. Nonetheless, both paradigms remain fundamentally disembodied, as artificial systems lack the biological embodiment that situates human cognition in the world. The comparison between the seemingly intelligent behavior of machines—which, at best, possess intelligence only in a metaphorical sense—and the naturally grounded intelligence of human users provides a critical lens through which to examine the underlying models of mind, body, and language embedded in Foundation Models. This comparison invites reflection in light of recent developments in linguistic cognition, especially within the frameworks of social cognition theories, cultural semiotics, and cognitive semiotics.

It is required a basic philosophical knowledge, preferably in the fields of semiotics, linguistics and of philosophy of language, corresponding to the level of the bachelor’s degree in philosophy.

1. Legrenzi, P. (2024), L'intelligenza del futuro. Perché gli algoritmi non ci sostituiranno, Milano, Mondadori. 2. Deacon, T. (1997), The symbolic species: The symbolic species: the co-evolution of language and the brain, W.W. Norton & Co (trad. it. La specie simbolica. Coevoluzione di cervello e capacità linguistiche, Giovanni Fioriti editore 2001). 3. Schilhab, T., Stjernfelt, T.,Deacon, T. (2012), The Symbolic Species Evolved, Springer (saggi scelti forniti dalla docente). 4. Mahowald, Kyle & Ivanova, Anna & Blank, Idan & Kanwisher, Nancy & Tenenbaum, Joshua & Fedorenko, Evelina. (2023). Dissociating language and thought in large language models: a cognitive perspective. 10.48550/arXiv.2301.06627. Further readings and references will be suggested during the lectures and, where possible, made available in the Google Classroom. Both attending and non-attending students will be invited to enroll in the course’s Google Classroom platform.

Attendance is not mandatory but strongly recommended.

The exam consists of an oral interview (70% of the final grade) and a written paper (30% of the final grade), both in English. Attending students may agree with the instructor to produce a short paper (maximum 15,000 characters) providing a critical analysis of one of the texts included in the exam syllabus. The paper must be submitted by e-mail to the instructor’s institutional address at least one week before the chosen exam date. Alternatively, the paper may be replaced by an in-class presentation following the flipped-classroom format. If students do not opt for either the short paper or the in-class presentation, the exam will be conducted entirely as an oral interview.

The first part of the course will mainly consist of frontal lessons, while the second part will stimulate open discussion, also increasing a direct involvement through the flipped-classroom methodology.