comentário a “Nonconceptualism and content independence”: Embrace messiness in perceptual studies


Eros Moreira de Carvalho[1]


Referência do texto comentado: Rodríguez, Sebastián Sanhueza. Nonconceptualism and content independence. Trans/Form/Ação: revista de filosofia da Unesp, vol. 44, n. 2, 2021, p. 293 –303.


In his paper, “Nonconceptualism and content independence”, Sebastián Sanhueza Rodríguez sustains that the state nonconceptualism (SNC) is (1) independent from the content nonconceptualism (CNC), (2) compatible with nonrepresentationalist accounts of perceptual experience and (3) better situated to ground and explain the distinction between perceptual and thought. I wholly agree with (1) and even pursued it in my PhD dissertation (CARVALHO, 2007, Chapter 4) as well as (1) and (2) in a later paper (CARVALHO, 2016). Rodríguez did a very good job in the way he articulated and put forward (SNC). His final definition of (SNC) is perspicuous:

SNC*: For any perceptual experience E, any subject S, and any time t, E is nonconceptual, iff it is not the case that in order for S to undergo E, S must possess at t any particular concept (RODRÍGUEZ, 2021, p. XXx).


This definition has the advantage of making it clear the independence of (SNC*)[2] from the content thesis, the claim that perceptual experience has a content. “E”, in this definition, refers to a type of experience, whether it has a content or not. Thus, (SNC*) is compatible with nonrepresentationalist views of perceptual experience.

However, this definition may be too strong. Rodríguez (2021, p. Xxx) stays neutral as to what kind of concept-independence is relevant for (SNC*). One may think of at least three relevant kinds of dependence relation: causal, informational and constitutive. For instance, it may well be the case that the acquisition of a conceptual ability is an enabling condition for acquiring or exercising a particular discriminatory skill, without, however, being the case that this conceptual ability is constitutive or that it provides information for any experience that results from that skill. This case would not be taken as nonconceptual by (SNC*) because possession of a concept alone does not discriminate between causal, informational or constitutive dependence. For many debates in epistemology and in philosophy of mind, it matters what kind of concept-independence is at issue. A foundationalist my not be bothered that a concept is an enabling condition for a type of experience E provided that it is not constitutive of that experience. A cognitivist may not acknowledge any relevant cognitive penetration on perception if perception is not informationally dependent on higher cognition. We need to keep an eye on these discussions if we want a relevant formulation of (SNC). In the end, it may be better to drop the talk about possession of concepts in favor of a more precise relation of dependence or acknowledge that there may be different but equally relevant types of (SNC).

Rodríguez acknowledges that (SNC*), in order to be applied, requires also a suitable view of concepts:

[…] one still has to specify, first, a suitable notion of content for cognitive phenomena, and secondly, a notion of concept in relation to which perceptual phenomena could be characterized as concept-independent. But I take it that both representationalists and nonrepresentationalists concede that these tasks can and should be addressed (RODRÍGUEZ, 2021, p. Xxx).


However, I do not see that the question about what a concept is can be indefinitely postponed by anyone interested in this debate, especially because it is central for the very distinction between perception and thought. This distinction might even collapse depending on what concepts are taken to be. I mean, I don’t think that someone who advances a (SNC) position may at the same time stay reasonably unconcerned as to what concepts are. Consider for instance Alva Noë’s view on this issue. In Action in Perception (2004), Noë claims that perception is both conceptual and a matter of possessing the appropriate sensorimotor skills. This is possible because he takes sensorimotor skills as concepts: “[…] some practical skills–some sensorimotor skills–are simple concepts, or so I propose.” (Noë, 2004, p. 199). For instance, I may have a simple concept of displacement if I am able to make compensatory adjustments of my eyes, head and body to keep a moving object in the center of my visual field. A consequence of this view is that the distinction between perception and thought is blurred (NOË, 2012, p. 25). Accordingly, if it is said that I can think about the sofa in my living room because I am able to go to the living room and track changes in the sofa–I would notice if something were placed on it or if something happened to it–, then how different this thought capacity would be from the ability to perceive the sofa? Not so much (NOË, 2012, p. 22). Of course, Noë may face a lot of trouble in explaining language and reasoning since sensorimotor skills do not seem to fulfill two characteristics apparently essential to those two phenomena: compositionality and sistematicity (CARVALHO, 2016, p. 560). I will not pursue this path here. My point is that if the (SNC) vs. state conceptualism (SC) distinction is supposed to help to clarify the distinction between perception and thought, as Rodríguez (2021, p. Xxx) seems to argue, then this is not something the former distinction can do alone. For the desired result, the (SNC) vs. (SC) distinction should be worked out in tandem with a proper view of concepts. Clearly, this is not the case for Noë, he doesn’t expect these distinctions map onto each other. He can keep the (SNC) vs. (SC) distinction and even (SNC*), and then move to (SC*), the denial of (SNC*), as I think he should (CARVALHO, 2016, p. 562)[3]. Doesn’t this mean in the end that (SNC*) and (SC*) are neutral as to what concepts are? Again, it all depends on how one puts all these things to work together. It’s possible to stick with (SNC*) irrespective of how one conceives concepts, but then the (SNC*) vs. (SC*) distinction may turn out useless to clarify the distinction between perception and thought.

I’m in much agreement with Sebastián Rodríguez and pretty much appreciated his paper, but I have the impression that we part ways methodologically. He seems to have a more analytical approach, assuming maybe that the task of defining (SNC) may be approached independently of other issues in philosophy of mind and epistemology, whereas I think that these issues are deeply entangled, and we need a more organic approach to pursue them. If a definition of (SNC) is good or not depends on a lot of other issues, such as what a concept is, how perception is distinct from thought, whether perception and thought share the same kind of content and so on. There is no escape from this whole messiness.



CARVALHO, E. M. Em defesa da justificação perceptiva: desmitificando o Mito do Dado. 2007. Tese (Doutorado) – –Universidade Federal de Minas Gerais, 2007.

CARVALHO, E. M. An actionist approach to the justificational role of perceptual experience. Revista Portuguesa de Filosofia, v. 72, n. 2-3, 2016, p. 545-572.

NOË, A. Action in Perception. Cambridge, MA: The MIT Press, 2004.

NOË, A. Varities of Presence. Cambridge, MA: Harvard University Press, 2012.

Rodríguez, S. S. Nonconceptualism and content independence. Trans/Form/Ação: revista de filosofia da Unesp, v. 44, n. 2, 2021, p. 293 – 303.


Recebido: 11/8/2020

Aceito: 15/8/2020


[1] Professor de filosofia da Universidade Federal do Rio Grande do Sul (UFRGS), Porto Alegre, RS – Brasil e Bolsista de Produtividade do CNPq. ORCID: E-mail:

[2] “(SNC*)” refers to Rodríguez’s definition, “(SNC)” to any formulation of state nonconceptualism.

[3] He seems to be committed to content conceptualism in Action in Perception, but later (NOË, 2012), when he is more clear about the rejection of representationalism, (SC*) captures better his position.