The concept of affordances is rapidly gaining traction in the philosophy of mind and cognitive sciences. Affordances are opportunities for action provided by the environment. An important open question is whether affordances can be used to explain mental action such as attention, counting, and imagination. In this paper, we critically discuss McClelland’s (‘The Mental Affordance Hypothesis’, 2020, Mind, 129(514), pp. 401–427) mental affordance hypothesis. While we agree that the affordance concept can be fruitfully employed to explain mental action, we argue that McClelland’s mental affordance hypothesis contain remnants of a Cartesian understanding of the mind. By discussing the theoretical framework of the affordance competition hypothesis, we sketch an alternative research program based on the principles of embodied cognition that evades the Cartesian worries. We show how paradigmatic mental acts, such as imagination, counting, and arithmetic, are dependent on sensorimotor interaction with an affording environment. Rather than make a clear distinction between bodily and mental action, the mental affordances highlight the embodied nature of our mental action. We think that in developing our alternative research program on mental affordances, we can maintain many of the excellent insights of McClelland’s account without reintroducing the very distinctions that affordances were supposed to overcome.

Key words: affordance, mental action, action selection, attention.

The main aim of Section 1 is to get to grips with the radical element of Radical Enactivism (RadEn): the conviction that basic cognition is to be explained in a thoroughly non-representational fashion. I first contrast the Enactivist approach with the traditional representational approach to cognition. Thereafter, I introduce RadEn’s notion of representation in terms of content. I then describe RadEn’s worries about naturalising content by relying on the notion of information, in terms of which the contemporary debate is set. I end this section with Hutto and Myin’s formulation of the Hard Problem of Content. In Section 2, I introduce Radical Enactivism’s distinction between basic, contentless, cognition and linguistically mediated thought, which is mirrored in the distinction between basic minds and linguistic superminds. Basic cognition is described in terms of biosemiotics: an Enactivist makeover of teleosemantics – an approach that accounts for content in terms of functions of organisms. I further describe how RadEn describes cognition that deals with objects that are absent in terms of recreative imaginings. I then introduce Hutto’s account of linguistic content, which is based on Davidsonian interpretationalism: content can only be ascribed through a process of radical interpretation, which functions as an intensionality test that can only be passed by the fine-grained responding of linguistically competent creatures. In Section 3, I raise some initial worries for RadEn’s distinction between basic and linguistic cognition. I argue that linguistic mediation is not necessary for adopting a propositional attitude if we accept the intensionality test as a criterion. Thereafter, I argue that even if we adopt a linguistically mediated propositional attitude, the symbols of natural language are not the intentional objects of our attitudes. It is only when we talk about language that we shift our attention to the linguistic symbols and they become intentional objects. After that, I argue that, using the the resources available to RadEn, it will be impossible to make sense of the notion of a proposition. In Section 4, I trace the problems stated in Section 3 to problematic assumptions about language. I argue that the root problem is that Hutto conceives of our linguistic competence as being made possible by an abstract system of symbols, and that conceiving of language in this way robs us of the possibility to explain how language works. I show that Davidson, who is the main source of inspiration for Hutto’s account of language, never meant his theory of meaning as a theory that could also explain our linguistic abilities. Although Hutto extends the Davidsonian approach to include the acquisition of words through social triangulation, this extension does not explain how language works. After that I give some general objections to thinking of language in terms of an abstract symbol system. I argue that the main function of language is not to represent the world, but to coordinate behaviour. These considerations will provide the basis for showing the implications of relinquishing an abstract symbol system view of language and the alleged distinction between the basic and supermind, which I consider in the final Section of this thesis.

This paper proposes an ecological-enactive account of utterances of concrete words – words used to indicate observable situations, events, objects, or characteristics. Building on the education of attention model of learning, utterances of concrete words are defined as attentional actions: a repeatable form of behaviour performed by a person to indicate (i.e. point out) a particular aspect of the current situation to someone in order to achieve something. Based on recent empirical evidence on categorical colour perception, attentional actions are proposed to constrain the ongoing phenotypic reorganisation of persons into task-specific devices. The paper ends by situating the proposed account in a wider theoretical perspective on language. This paper serves two purposes: first, it undermines the scope objection against the ecological-enactive approach, and second, it provides a novel explanation for recent empirical evidence with respect to the role of language in categorical colour perception

Key words: attentional action, enactivism, ecological psychology, word learning, word-meaning, categorical perception.