Perception: a discriminating capacity or a predictive capacity?

By John Joseph Dorsch

Abstract: Schellenberg (2016) believes that perception is an essentially discriminating capacity. On the basis of this belief, she develops the particularity argument, which holds that perceptual states are constituted by particulars. The first premise of the argument reads “If a subject S perceives a particular α, then S discriminates and singles out α” (p. 11). This premise can be rejected by arguing that it is possible to perceive a ganzfeld, viz. a homogeneous field without any particulars to discriminate. Schellenberg dismisses this counterargument by referring to the ganzfeld effect, which she describes as “experiencing a sense of blindness”, concluding that “we cannot in fact see a completely uniform wall that fills out our entire field of vision” (p. 13).

I discuss two challenges for Schellenberg’s dismissal of the ganzfeld. Participants in ganzfeld experiments report experiencing a sense of blindness only after approx. 30 minutes of exposure (Wackermann et al 2008). Surely it is plausible that viewers see the ganzfeld during this half hour before the sense of blindness is reported to occur. The second challenge is this: the ganzfeld effect only occasionally results in viewers reporting a sense of blindness; far more often, viewers report seeing hallucinations. I show that Schellenberg’s theory of perception as an essentially discriminating capacity cannot adequately respond to this challenge and argue that a more plausible alternative is the predictive processing theory of perception/action (Hohwy 2013; Clark 2016). I conclude with possible implications for Schellenberg’s particularity argument and her belief that perception is an essentially discriminating capacity.

 

Introduction

The ganzfeld effect can be easily reproduced. Take a white ping-pong ball and slice it in half; now place both halves over the viewer’s eyes, making sure her visual field is completely obscured; set a light source in front of the viewer and adjust illumination to be as uniform as possible. Within a short period of time, the viewer should begin hallucinating just like participants in ganzfeld experiments. These hallucinations typically develop in complexity throughout ganzfeld exposure; most participants report seeing a dense, monochromatic fog at first (Metzger 1930). As time progresses, more complex hallucinations appear, such as dreamlike imagery of people and places; sometimes participants report seeing “blackness” or experiencing a “sense of blindness” (Metzger 1930 ; Cohen 1960; Gibson 1979; Pütz 2006; Wackermann et al 2008).

Schellenberg draws upon the ganzfeld effect to bolster her belief that perception is an essentially discriminating capacity. She begins by claiming that if you perceive something, say, a table, it is only because you discriminate the table from its surround. She concludes from this that perceiving a homogeneous field, a ganzfeld, is impossible, since there would be nothing in a ganzfeld to discriminate. She believes further that this conclusion is supported by empirical evidence from ganzfeld experiments: some participants report experiencing a sense of blindness. That said, the ganzfeld effect is constituted by more than experiencing a sense of blindness; more often, ganzfeld viewers report seeing hallucinations.

The goal of this paper is to argue that you can, contrary to Schellenberg’s belief, see a ganzfeld. One possible challenge, which I call the argument from delay, is that a considerable amount of ganzfeld exposure is required before viewers report experiencing a sense of blindness. However, I will argue that Schellenberg may sufficiently counter this argument, provided an appeal is made to a system of object indexes (Leslie et al. 1998; Scholl and Pylyshyn 1999; Carey and Xu 2001; Scholl 2007). In the second section, I anticipate Schellenberg’s response to the argument from hallucination; i.e. participants in ganzfeld experiments more commonly report seeing hallucinations than experiencing a sense of blindness. That said, I show that this anticipated response fails to explain how hallucinations emerge. Therefore, I argue that an alternative theory of perception, called predictive processing, is needed to account for the ganzfeld effect. In the conclusion, I discuss what our discussion means for Schellenberg’s particularity argument and her belief that perception is an essentially discriminating capacity.

 

1. Argument from Delay

I will refer to Schellenberg’s view of perception as the neatly expressed maxim: no perception without discrimination, which shares similarities to Merleau-Ponty’s maxim: no perception without a figure-background structure. Merleau-Ponty would claim that perception can occur only on the basis of some discriminated figure vis-à-vis some discriminated background. Thus, Merleau-Ponty (1945) also denies the possibility of perceiving a ganzfeld, “A truly homogeneous area, offering nothing to perceive, cannot be given to any perception” (p. 4). In responding to arguments based on the ganzfeld, Merleau-Ponty has avenues at his disposal that Schellenberg does not. Merleau-Ponty might claim that even when viewing a ganzfeld, one still perceives a figure-background structure: the figure is the ganzfeld, while the background is provided by other sensory modalities, such as proprioception. Schellenberg cannot pursue this route, however, because her argument focuses on a single modality. Therefore, when confronted with arguments based on the ganzfeld, she emphasizes that viewers report experiencing a sense of blindness during ganzfeld exposure. For this section, I assume that a sense of blindness is the only ganzfeld effect viewers report; but assuming this does not dismiss the need to explain the delay between ganzfeld onset and ganzfeld effect (approx. 30m).

One might respond by claiming that the visual system requires time before the ganzfeld is perceived as a truly homogeneous field. This means, between ganzfeld onset and ganzfeld effect there is an interval, in which the visual system can still discriminate. Since the ganzfeld is a homogeneous field with nothing to discriminate, the source of discriminata cannot be the visual field. Instead, the source must be the visual system itself. Thus, I propose that the visual system continues to discriminate its own, what are called, object indexes (Leslie et al. 1998; Scholl and Pylyshyn 1999; Carey and Xu 2001; Scholl 2007).

Conjectured to underpin visual perception, object indexes function like pointers for perceptible objects by representing them and tracking them in the visual field. Imagine the following scenario. You see a completely homogeneous field before you. There are no differences in the field to discriminate, but you can still single out points in the field. Though each point appears identical to the others, each point is differentiated by the direction of your visual focus; e.g. you observe the top-left area, then you observe the bottom-right area. While shifting your focus from one area to the next, you continue to discriminate. It is claimed that these discriminated points are underpinned by a system of object indexes.

 

Though this response accounts for the delay in the ganzfeld effect, it fails to account for its very occurrence. An event is needed, wherein even discrimination underpinned by object indexes is no longer possible. Recall the scenario above. Notice that in order to discriminate the left from the right, or the top from the bottom, some distance between these areas must be perceived. And yet, it seems difficult to imagine how distance can be perceived in the absence of any perceptible difference, such as during ganzfeld exposure. This argument is consistent with empirical evidence from ganzfeld experiments: as ganzfeld exposure increases, participants’ ability to judge distance becomes progressively diminished (Metzger 1930). Therefore, it seems plausible that discriminating between top and bottom, left and right, becomes increasingly more difficult as discerning the distance between these areas becomes increasingly more difficult. Possibly, during prolonged ganzfeld exposure, the system of object indexes, which is enabling perception to continue to operate as a discriminating capacity, fails, once object indexes themselves, being dependent on some perceptible distance, can no longer act as aids to discrimination.

In responding to the argument from delay, Schellenberg, believing perception to be an essentially discriminating capacity, may appeal to a system of object indexes to account for the delay between ganzfeld onset and ganzfeld effect. This response thus posits two classes of discriminating capacities: allocentric and egocentric (see Klatzky 1998). The allocentric capacity discriminates based on a world-centered reference frame, i.e. how things in the world relate to each other. The egocentric capacity discriminates based on a self-centered reference frame, i.e. how the self relates to itself, the visual field and, if possible, things in that field. In the case of ganzfeld exposure, only the egocentric capacity can be employed, so points in the visual field are discriminated from other points by shifts in visual focus, all of which are underpinned by a system of object indexes, which is integrated into the visual system itself. That is not to say that object indexes only represent egocentrically; object indexes can represent allocentrically as well (Alæs et al. 2015). Instead, I mean to say that if things in the visual field can only be discriminated egocentrically, such as during ganzfeld exposure, this egocentric discriminating capacity is enabled by a system of object indexes. Once the discriminata of the egocentric capacity are no longer available, the ganzfeld fills the visual field, and, since perceiving a homogeneous field is believed to be impossible, the viewer experiences a sense of blindness. That said, the ganzfeld effect is more than a sense of blindness; more often, ganzfeld viewers report seeing hallucinations.

 

2. Argument from Hallucination

Discovering how ganzfeld exposure causes viewers to hallucinate has interested psychologists and philosophers for nearly a century now. In the late 1950’s, it was postulated that hallucinations emerge as the result of the breakdown of the perceptual system. Discussing possible reasons for the ganzfeld effect, Cohen (1957) conjectures, “…that the perceptual mechanism has evolved to cope with a differentiated field, and, in the absence of differentiation, there is a temporary breakdown of the mechanism” (p. 407). I believe Schellenberg would respond to the argument from hallucination in a similar fashion: ganzfeld hallucinations actually support the view no perception without discrimination; since there are no discriminata, it is plausible to think, in light of the emergence of hallucinations, that employing perceptual capacities generates discriminata. Though this response indicates that ganzfeld exposure leads to the generation of hallucinations, it does not explain how hallucinations are generated.

The theory of perception called predictive processing postulates that perceptual systems are essentially predictive capacities. If the environment is noisy, such that little in the environment can be reliably predicted, say, in a dense fog, where just about anything can emerge, then predictions about the environment will be processed as more reliable than sensory data from the environment. When the conditions are such that the sensory data are weighted as very unreliable, as is thought to be the case during ganzfeld exposure, the generation of predictions can lead to the emergence of hallucinations. So if the predictive processing theory of perception is correct, hallucinations do not result from the breakdown of the perceptual system but from its proper functioning. In other words, predictive processing maintains that hallucinations emerge during ganzfeld exposure because you can and do see a homogeneous field.

A predictive-processing perceptual system copes with the environment by continuously estimating its own uncertainty regarding sensory data. Through a balancing act of bottom-up influences, such as sensory data, and top-down influences, such as predictions, the perceptual system discloses percepts at the juncture between what is sensed and what is predicted (Hohwy 2013; Clark 2016). If sense data are processed as uncertain, less weight is given to errors resulting from the sense data: the less weight given to the errors, the less impact the data have on what is perceived. In other words, if the environment is processed as very unreliable, then the sensory data of the environment will have little impact on what is perceived. Instead, in these error prone environments, more weight is given to top-down influences that generate predictions.

 

This is how predictive processing explains the emergence of hallucinations during ganzfeld exposure. You might imagine seeing nothing but blue sky on a cloudless day. In that event, you would see a natural ganzfeld. That aside, the ganzfeld is a rare, natural occurrence. Since its occurrence is so uncommon, sensory data in the ganzfeld are processed as very unreliable. This means, the sensory data are inhibited from revising top-down predictions. This suppression of bottom-up sensory data leads to an increase in activity of top-down prediction data. Normally, predictions are regulated by sensory data, but, during ganzfeld exposure, errors resulting from sensory data are weighted so low that predictions begin to progressively regulate themselves and hallucinations emerge as a result—like a self-fulfilling prophecy.

This explanation can provide a response to the argument from delay. As mentioned above, the hallucinatory imagery often begins as a dense, monochromatic fog that slowly fills the visual field. Over time, more complex images appear within or outside of the fog. So the delay can be perhaps better understood as a slow progression from simple to more complex hallucinations as ganzfeld exposure increases. Thus, it is plausible to think that this progression develops in parallel with predictions generated by top-down processing: when predictions adapt to the unreliable environment and begin to regulate themselves, more complex hallucinations emerge.

In addition to the argument above, the predictive processing theory of perception, regarding its explanation for the emergence of hallucinations during ganzfeld exposure, is supported by neurological evidence. Pütz (2006) investigated the EEG correlates of ganzfeld induced imagery. When participants reported hallucinations, Pütz discovered an increase in alpha wave activity, which suggests “the retrieval, activation, and embedding of memory content in the ganzfeld imagery” (p. 175) and “[the inhibition] of the processing of optical and acoustical sensory input”, which “needs to be suppressed to allow for internally directed attention” (p. 177). This “internally direction attention” corresponds to the ganzfeld induced mental state, in which hallucinatory imagery emerge. If you assume a correlation between the phenomenology of experiencing hallucinatory imagery and the observed electrical activity in the brain, then it is plausible to think that Pütz’s findings support the conclusion that hallucinations emerge as the result of higher level processing coupled with the inhibition of lower level processing. Thus, regarding ganzfeld exposure, the case for predictive processing is supported by these findings.

 

Conclusion

I would like to summarize what this discussion means for Schellenberg’s particularity argument and her belief that perception is an essentially discriminating capacity. The alternative theory of perception, predictive processing, is able to account for hallucinations during ganzfeld exposure, though Schellenberg’s theory of perception cannot. One conclusion to be drawn from the predictive processing theory of perception is that we can and do see a homogeneous field. If this is correct, then the first premise of Schellenberg’s particularity argument is false and, furthermore, doubt is cast on her belief that perception is an essentially discriminating capacity; instead, perception may be an essentially predictive capacity. However, predictive processing, if correct, would need to explain why some ganzfeld viewers experience a sense of blindness, which I find difficult to imagine accounting for without appealing to perception as an essentially discriminating capacity. So regarding future investigation, perhaps a unification of the two theories should be sought.

Share your thoughts on this!

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s