Causal Theory of Perception

H.P. Grice's Causal Theory of Perception was advanced as a place-holder, advanced in order to discuss the differing roles of semantics and pragamatics of 'looks as if' locutions. However, it became very influential in the theory of perception, where it has typically been treated as a completely articulated analysis of direct-object uses of 'see.' In this paper, I examine several counter-intuitive (and somewhat sensational) applications of the Causal Theory and show how they depend on such an understanding. I propose that if we fill our our understanding of visual perception in terms of the modalities and functional benefits of vision, these applications can be corrected. On this understanding, some sensory states will turn out not to be material object seeings because they are either not created in the normal way or fail to provide information in the normal way.

In an important late paper, ‘The Role of the Imagination in Kant’s Theory of Experience’, Sellars brings together ideas about the complex nature of perceptual consciousness and the content of perceptual demonstratives. In a development of his previous ideas about perception, he clarifies the key role played by the imagination in integrating the conceptual and sensory (or phenomenal) components of perceptual experience. I propose a modification of Sellars’s views on the imagination, and show how the resulting conception explains the different ways in which experiences can be conceptualised. I then discuss how the account enables us to understand exactly how, according to the Sellarsian critical realist analysis of experience, we are able to make demonstrative judgements about physical objects, while avoiding a problematic appeal to neo-Russellian notions of acquaintance.

Key words:  perceptual experience; causal theory of perception; critical realism; Wilfrid Sellars; demonstrative reference; acquaintance; imagination; Kant

In this paper I will try to outline a Metaphysics of Perception that takes for granted one of the central thesis of the metaphysical doctrine called Indirect Realism. Firstly, I will introduce the central thesis of Indirect Realism and then a special version of the Causal Theory of Perception that modifies in some fundamental respect one of the most influential version of Causal Theory of Perception designed by William Child.

Several psychological theories of perception claim that the possible is something we perceive. Typically, the ecological approach holds that perception provides direct access to opportunities for action potentiated by the structures of the environment. Is such a claim legitimate? Can we actually perceive the possible? If perception plugs our mind into reality, should we not be freed from perception to gain access to possibilities? Therefore, shouldn’t the possible be conceived rather than perceived, something we have knowledge of, not something we see?
Answering these questions requires examining whether the thesis that the possible is perceived is consistent with: (a) the phenomenological structure of the perceived object, (b) our models of the perception mechanisms, especially the standard causal model. This inquiry will show that if, in a strict phenomenological sense, the possible is not something appearing in the same way as perceived objectities, it nevertheless contributes to determine the organization and semantics of the perceived environment. The possible as such is not seen, but it gives structure, form and meaning to the world that appears in perception.

FRENCH ABSTRACT
Différentes théories psychologiques de la perception soutiennent aujourd’hui que le possible est quelque chose que nous percevons. Pour l’approche écologique par exemple, la perception fournit un accès direct aux possibilités d’action potentialisées par les structures de l’environnement. Cette idée est-elle légitime ? Peut-on réellement percevoir le possible ? Si la perception nous branche sur le réel, ne faut-il pas justement se libérer de la perception pour accéder au possible ? Dès lors, le possible n’est-il pas plutôt pensé que perçu, l’objet d’un savoir et non pas d’un voir ?
Répondre à ces questions nécessite d’examiner si la thèse que le possible est perçu est en cohérence avec : (a) la structure phénoménologique de l’objet perçu ; (b) nos modèles du fonctionnement de la perception, en particulier le modèle causal standard. Cet examen permet de montrer que si, dans un sens phénoménologique strict, le possible n’est pas quelque chose qui apparaît sur le mode de l’objectité perçue, il contribue néanmoins à déterminer l’organisation et la sémantique de l’environnement perçu. Le possible comme tel n’est pas perçu, mais il confère structure, forme et sens au monde qui apparaît dans la perception.

Many philosophers believe that there is a causal condition on perception, and that this condition is a conceptual truth about perception. A highly influential argument for this claim is based on intuitive responses to Gricean-style thought experiments. Do the folk share the intuitions of philosophers? Roberts et al. (2016) presented participants with two kinds of cases: Blocker cases (similar to Grice's case involving a mirror and a  pillar) and Non-Blocker cases (similar to Grice’s case involving a clock and brain stimulation). They found that a substantial minority agreed that seeing occurs in the Non-Blocker cases, and that in the Blocker cases significantly less agreed that seeing occurs. They thus hypothesized that folk intuitions better align with a no blocker condition than with a causal condition. This paper continues this line of enquiry with two new experiments. The paper investigates the generality and robustness of Roberts et al.’s findings by expanding the sense modalities tested from only vision to audition and olfaction as well. The paper also uses Gricean-style thought experiments as a case study for investigating the “reflection defense” against the negative project in experimental philosophy. Our results replicate and extend Roberts et al.'s study and support their hypothesis that folk intuitions better align with a no blocker condition. They also provide an empirical reason to doubt the reflection defense.

We can hear silence because silence, an absence of sound, causes our hearing of it. Advocating this position, Roy Sorensen puts to use his own theory of the direct perception of absences. Sorensen’s theory, which relies on two theories of perception (the causal theory of perception and the theory of non-epistemic perception), certainly has its appeal. However, it also has its problematic aspects. On my reading, a weak point of his theory is that it does not provide a criterion for the identification of what exactly we hear. By elaborating this objection in detail, I intend to demonstrate that Sorensen’s theory (i) does not concern direct (non-epistemic) perception, and (ii) does not show that silence is causally efficient. Therefore, (iii) it fails to show that silence is the genuine negative object of hearing. I conclude by giving two further reasons for why the ontology that underpins Sorensen’s theory should not be endorsed.

Das Ziel dieses Aufsatzes ist es erstens, die Unterscheidung zwischen dem direkten und indirekten Realismus hinsichtlich der Wahrnehmung zu erläutern und zweitens, die weit verbreitete Ansicht, der direkte Realismus sei mit der Kausaltheorie der Wahrnehmung unvereinbar, zu widerlegen. Es lassen sich fünf Argumente für die Inkompatibilität des direkten Realismus mit der Kausaltheorie der Wahrnehmung unterscheiden. Keines dieser Argumente ist stichhaltig.

Two arguments Paul Snowdon has brought against the causal theory of perception are examined. One involves the claim that, based on the phenomenology of perceptual situations, it cannot be the case that perception is an essentially causal concept. The other is a reductio, according to which causal theorists’ arguments imply that a proposition Snowdon takes to be obviously non-causal (A is married to B) can be analyzed into some sort of indefinite ‘spousal connection’ plus a causal ingredient. I conclude that neither argument is sound. The reason that Snowdon’s critiques fail is that, since causal theories need not be about ‘effect ends’ that are internally manifest to perceivers, no such ostensibly separable, non-causal property as it being to S as if he were perceiving O need be an essential element in a causal theory of perception.