Pär Sundström
Affiliation: Umeå University
Category: Philosophy
Keywords: nativism, empiricism, aphantasia, thought, sensory grounding
Date: Thursday 4th of September
Time: 16:00
Location: Room 161 (161)
View the full session: Concepts & Propositional Attitudes
Could something that never had the capacity to sense anything still think something? In a recent article, David Chalmers (2023) argues that such a “pure thinker” is possible but that there are significant limitations in what thoughts are available to them. Pure thinkers would be, he claims,
largely structuralist thinkers, at least where nonmental reality is concerned. … Pure thinkers will be able to entertain structural hypotheses about the external world, akin to the sort of hypotheses that science puts forward according to structural realism. But they will not be able to possess nonstructural concepts such as Mary’s full-blown concept of red when she leaves the room (2023, 35).
Chalmers’s view is part-nativist, part-empiricist (Nagel 1986 maintains a similar view). I shall argue in favour of a more nativist view: given that a pure thinker is possible in the first place, such a thinker could have nonstructural concepts, including Mary’s full-blown concept of red.
My argument builds on three “sense-thought independence theses”.
The first sense-thought independence thesis is that it is possible to have Mary’s full-blown concept of red at a time even if one does not sense red at that time. In fact, I shall argue, this is not just possible but typical for us. The relevant concept is a standing one. It is not a concept that we tend to acquire and lose as red goes in and out of our experiences.
The second sense-thought independence thesis is that it is possible to have Mary’s full-blown concept of red at a time even if one does not sense red at that time and did not in the past have a capacity to sense red. Take someone who possesses the full-blown concept of red but does not at the time sense red. The relevant concept possession is plausibly grounded in a certain brain state of that person. And it is in principle possible for someone to come to be in that brain state even if they do not at the time sense red and never in the past had the capacity to sense red. If that is right, then it is possible to have the concept even if one does not sense red at that time and did not in the past have a capacity to sense red. (This line of argument goes back to Unger 1966. It resembles and foreshadows Davidson’s 1987 more widely known “swampman” thought experiment.)
The first two sense-thought independence theses leave open that, to possess the full-blown concept red at a time one must at that time have a capacity to generate sensory imagery of red.
My third sense-thought independence thesis rules out this option. It says that one can have the full-blown concept red at a time even if does not have an ability to generate sensory imagery of red at that time. This thesis is, I shall argue, supported by recent evidence from aphantasics.
There are subjects who report that they do cannot generate any sensory imagery (Galton 1880; 1883, Faw, 2009, Zeman 2015; 2020). One possible explanation of these reports is that (i) the relevant subjects are just deviant reporters of their conscious lives and that they have both as much imagery and as much introspective access to their imagery as anyone. Another possible explanation is that (ii) the subjects fail to introspect sensory imagery that they in fact have, which may in turn be either because (ii:a) their introspective capacities are substandard or because (ii:b) their imagery is unconscious and for that reason unavailable even given standard introspective capacities. A third possible explanation is that (iii) the relevant subjects in fact cannot generate sensory imagery.
I shall argue that the explanation (iii) is the most plausible one, at least in the case of some subjects and imagery of red; i.e., there are some self-avowed aphantasics who cannot generate any sensory imagery of red, either voluntary or involuntary, conscious or unconscious. The alternative to this hypothesis is that all self-avowed aphantasics can generate some sensory imagery of red, either voluntary or involuntary, conscious or unconscious. I shall argue that the former hypothesis fits better with a range of recent experimental results, including Koegh and Pearson (2018; 2021), Wicken et al. (2021), Königsmark et al. (2021), Kay et al. (2022), and Chang et al (2025).
Moreover, I shall argue that some aphantasics who plausibly cannot generate any sensory imagery of red still have the full-blown concept of red. In part, lack of evidence against this hypothesis provides some support for it. This is so because, if the subjects in question had lacked the relevant concept, one would have expected some of them to issue reports along the lines of: “I don’t know what red looks like when I don’t see it. I keep rediscovering and forgetting it as red goes in and out of my perception”. But we do not have any such reports from self-avowed aphantasics. To the contrary, we find self-avowed total aphantasics reporting that they know perfectly well what red looks like even when they don’t perceive it (see for example Love 2024).
All of that provides significant support for the third sense-thought independence thesis that one can have the full-blown concept red at a time even if does not have an ability to generate sensory imagery of red at that time.
Tying these three sense-thought independence theses together, I shall show that they jointly support the conclusion that, if a pure thinker is possible at all, they can have a wider range of concepts than Chalmers supposes, including the full-blown concept of red and other “nonstructural concepts”.