Abstract

This paper asks a basic but largely neglected question: which behaviors genuinely qualify as instances of joint attention, and why? I argue that the current literature relies on an heterogeneous explanandum: very different behaviors are treated as paradigmatic cases of jointness, without a unified criterion for what makes attention joint rather than merely individual and parallel. I propose an exclusion criterion in the form of a canonical description of joint attention. In an enactivist-phenomenological spirit, I treat joint attention not as a capacity or mental state, but as an episodic phenomenon of social interaction in which subjects engage in a task whose success essentially depends on a specific kind of co-dependent and shared exercise of attention: an interwoven attentional exercise. I first use thought experiments (one-way mirrors, mutual spying and declarative pointing) to argue that attentional co-dependence alone does not warrant invoking ‘jointness’. I then examine real experimental cooperative linguistic scenarios, drawing on Merleau-Pontian reading and Clark and Krych’s Lego® task, to argue that genuine joint attention requires interwoven contributions. I close by distinguishing this proposal from close accounts and sketching a different conception of attentional overtness.