The main problem in adopting the notion of constitution in the context of philosophy of mind is that it can have the flavor of creating or establishing reality; the expression ‘to constitute objects’ might suggest that the existence of objects depends on an activity of the mind; thus, it might seem that the adoption of this notion invites a form of idealism. Indeed, as I have pointed out above, there is an affinity between the notion of constitution and transcendental idealism in the work of Husserl. Kern shows that Husserl’s turn towards transcendental idealism was made possible by his account of genetic constitution.1 A similar point is made by Ingarden who characterizes the fundamental thesis of transcendental idealism in the following way: what is real is nothing but a constituted noematic unit (individual) of a special kind of sense which in its being and quality results from a set of experiences of a special kind and is quite impossible without them. (Ingarden 1975, 21) Since the account of constitution that I have outlined above is strongly influenced by Husserl’s phenomenology, I will now go on to discuss whether my account also implies a form of idealism. I will show that it neither implies nor invites idealism; rather it provides an interesting perspective on the realism-idealism debate. I will try to develop this perspective by considering a late-Wittgensteinian argument, according to which the position of the idealist or idealist sceptic, who has a general doubt concerning the existence of physical objects, cannot be formulated meaningfully.