I investigate whether anorexia nervosa involves a kind of cognitive penetration by analyzing the results of Smeets' and Keizer's studies regarding body image distortion in patients suffering from anorexia nervosa (Smeets, 1999; Keizer, 2013). Because Smeets' study does not provide sufficient evidence for non-trivial cognitive penetration, namely that the results of her study do not show cognitive penetration to be anything more than a matter of focusing attention, iturn to the results of Keizer's 2013 study, which provides evidence for body image distortion occurring at an unconscious level. Interpreting Keizer's findings reveals a surprising result that, due to the plausible idea that the body schema and not the body image is the cause of body image distortion in anorexia nervosa, cognitive penetration is rejected in favor of what might be called cross-modal, somatic penetration in order to explain body image distortion in cases of anorexia nervosa.
One of the diagnostic criteria for anorexia nervosa (AN) is body image distortion (BID), which is characterized by the patient's inability to both judge her weight accurately and see her body accurately; instead, it is believed that the patient both judges and sees her body fatter than it actually is. Research has gradually come to postulate that two types of systems may be affected by BID: the body image and the body schema. The body image is a "system of perceptions, attitudes and beliefs pertaining to one's own body", while the body schema is a "system of motor capacities that function without awareness or perceptual monitoring" (Gallagher, 2008 p.24). The prevailing opinion about the cause of BID in AN can be summarized as follows: as a result of beliefs or attitudes about the body, the patient comes to see herself as fatter than she actually is (Bruch, 1962; Cash, 1997; Smeets 1999; Keizer, 2011; Riva, 2011; Favaro, 2012; Gaudio, 2014).
Philosophers call the idea that beliefs or attitudes can alter perception "cognitive penetration". Cognitive penetration is controversial and its exact definition differs slightly depending on whom you ask. So when discussing the possibility of cognitive penetration, it is crucial to distinguish significant from trivial accounts. For example, the idea that our perceptual experience is subject to top-down, attentional factors is trivial since it is accepted that there is nothing significant about the idea that perceptual experience changes based on how one focuses attention. In order for cognitive penetration to be significant, it must go beyond a mere shift in attention.
Macpherson defines cognitive penetration as "the alteration of the phenomenal character of perceptual experience by the state of one's cognitive system, for example by thoughts or beliefs" (Macpherson, 2012). I do not think that this definition is sufficient for establishing a nontrivial account of cognitive penetration, but it does provide a good starting point, so let us examine the definition further. The less controversial component is the causal component: the state of one's cognitive system must be the cause of the alteration of the phenomenal character of perceptual experience. But the details surrounding this component are not explicated, so what is meant by "cognitive system" is open to interpretation. In this paper, I will refrain from referring to "top-down" processes; instead, I will understand cognitive system to be "based on conceptual processing". The more controversial component is 'an alteration of the phenomenal character', which means that the perceptual alteration must be extensive enough to affect how an object appears in order for that alteration to be significant; meaning, it is not sufficient that an appraisal, a belief, or a judgment about an object has altered: the appearance itself must undergo alteration. As mentioned above, the alteration cannot be simply a matter of shifting attention.
To see this more clearly, consider Wittgenstein's duck/rabbit figure. Depending on how you focus your attention on the image, you will either see a duck or a rabbit. We might disagree as to whether a change in phenomenal character happens as the result of transitioning from one image to the next, from duck to rabbit or from rabbit to duck - some believe it does, some do not - but what is difficult to deny is the cause of this alteration: the alteration is caused by a shift in attentional focus. So seen, the duck/rabbit figure is a trivial account of cognitive penetration.
Empirical evidence for a non-trivial account of cognitive penetration is given by an experiment conducted by Delk and Fillenbaum (Delk and Fillenbaum, 1965). Participants were asked to adjust the color of a surface (call this the background color) to match the color of an object resting on the surface. The background color begins neutral and can be adjusted with a dial, whose turning gradually shifts the background color from yellow to red. The color of each object is always orange, but their shapes differ. The shapes are divided into two categories: geometrical shapes and shapes of commonly red objects such as hearts, lips, apples, etc. Evidence for nontrivial cognitive penetration comes as a result of the chosen background color when hearts, lips, and apples are placed on the surface. Even though these objects are the same color as the geometrical shapes, the background color is adjusted to exhibit a reddish hue. So it is postulated that the concept of heart, lips, apple, etc has an affect on how the object appears, has an effect on the phenomenal character of the perceptual experience of the objects' color. Furthermore, participants are unaware of having adjusted the background color to exhibit a reddish hue, which means it is unlikely that the cause of the alteration is a shift in attention, since a shift in attention is understood to be a conscious act.
Let us now discuss a prototypical case for cognitive penetration; I will draw upon this case throughout the paper. This case involves employing a conceptual capacity, such as nuanced color concepts, and that capacity's impact on perception. Imagine a painter who has mastered the nuances of color. Opposed to the lay, the painter possesses concepts for all those colors between red, orange, yellow, green, blue, indigo and violet. Instead of employing red to name the whole bandwidth of the lowest frequency of perceptible light, the painter employs granite, carmine, currant, etc. In possession of this conceptual capacity, the painter not only understands and recognizes these particular colors, but, so the theory goes, her experience of red is altered. Proponents of cognitive penetration believe that the painter's conceptual capacity alters the phenomenal character of the painter's experience of color. I have introduced this example not because I am convinced that is right, but because I believe that it provides a framework for asking the right questions about the existence of and mechanism behind cognitive penetration.
In this paper, I will use the results of Delk and Fillenbaum's color/shape experiment as the ideal to which any case for cognitive penetration should strive, and I will use the painter case to analyze BID in AN since for me it provides the framework for asking the right questions. The goal of this paper's analysis is to address the titular question: does anorexia nervosa involve a kind of cognitive penetration? That said, I wish to issue a caveat. It was my initial hypothesis that anorexia nervosa did indeed involve a kind of cognitive penetration, but after having understood more of the research, I have become convinced that this hypothesis is wrong. This is quite striking since it is based on the prevailing opinion about the cause of BID in AN; namely, that BID in AN is caused by patients' beliefs about their body. I will discuss in due time why I have chosen to reject the hypothesis, but now I would like to introduce the questions that will be raised in the conclusion: If the result of our discussion is correct and AN does not involve a kind of cognitive penetration, what can be said about the mechanism underlying BID in AN? And, what consequence does this have for the prevailing opinion about BID in AN? For now, let us turn to the main question.
Does AN involve a kind of Cognitive Penetration?
As mentioned above, the painter case is a prototypical case for cognitive penetration. The question I would like to ask at the outset of this discussion is whether a parallel can be drawn between what is going on in the mind of the painter and what is going on in the mind of the AN patient. In the painter case, it is concluded by proponents of cognitive penetration that the painter's employment of color concepts causes a change in the painter's experience of color. So we can infer from this conclusion that cognitive penetration involves some mechanism that alters the phenomenal character of sense data, which occurs as the result of the conceptual capacity employed by the agent. The argumentative crux of this claim is that concepts play a causal role in the alteration of phenomenal character. In line with the prevailing opinion of BID in AN, we are inclined to tell a similar story about what is going on in the mind of the patient. According to this prevailing opinion, the patient employs a concept, called the body image, that causes her to experience her body differently than a healthy person. So it is sound to think that the mechanism behind BID in AN is parallel to the cognitive penetration mechanism. In what follows, I will examine this parallel in more detail.
The field investigating AN is quite extensive and to do it justice would disrupt the scope of our discussion, so I will focus on the subgroup of the field investigating AN that examines the cause of BID. In understanding this subgroup, I have divided its history into three periods: early, modern, and contemporary. The early period begins with Bruch's 1962 paper and ends with the development of advanced brain imaging techniques, such as fMRI, occurring round about the turn of the century. But I will refrain from drawing upon modern sources since the results of brain imaging techniques are marred with controversy (see Gaudio, 2012 for a summary of the controversy). Beginning around 2009, the contemporary period of research into BID in AN stands out as having learned from the problems of the modern period and as a result turned to more novel attempts at understanding the cause of BID in AN, such as developing more nuanced models with more nuanced terminological distinctions.
Returning to the early period, Bruch believes that the cause of BID in AN is the patient's inner picture of her body, that this inner picture causes her to fail at "perceiving, recognizing or interpreting internal stimuli" (Bruch, 1962 p.189). In 1996, Cash echos Bruch's opinion: overestimation of body size reflects a cognitive judgment, not a perceptual bias (Cash, 1997). In 1999, Smeets argues that BID in AN is due to what she calls cognitive reconstruction, not perception (Smeets, 1999). All of these papers point to a common interpretation of BID in AN: body image distortion is caused by a top-down, high-level cognition, with or without an affective component. Inother words, patients 'see' themselves as fat because of their belief that they are fat. Let us take a moment to consider the prevailing opinion about BID in AN. Notice the scare quotes around see in the sentence. This is how researchers refer to the alteration of phenomenal character of perception: the body image alters how the patient 'sees' herself. This raises two questions. First, what do the scare quotes around see signify? And more interesting for the present discussion, can the scare quotes around see be dispensed with?
For AN to involve a kind of cognitive penetration, the scare quotes around see need to be, at least to some extent, dispensed with - but to what extent? As mentioned in the introduction, the alteration of experience by cognition needs to be non-trivial. When philosophers discuss cognitive penetration, they are essentially "weighing" the affect that cognition has on experience and arguing about where to place the "counterbalance". Think about cognitive penetration as operating along a scale from 1 to 10. Proponents of cognitive penetration wish to place the counterbalance at 1 (seeing), while opponents of cognitive penetration believe the counterbalance belongs at 10 (judging). Everyone accepts that cognition affects judgment, thus concluding so is tantamount to claiming that cognitive penetration does not exist. Now the question becomes whether we can move cognition's counterbalance down the scale. Setting the counterbalance at 10 means setting cognitive penetration at its most trivial, i.e. cognition affects judgment; while setting the counterbalance at 1 reflects setting cognitive penetration at its most significant, i.e. cognition affects seeing. Recall that the prerequisite for non-trivial cognitive penetration is an alteration in phenomenal character of perceptual experience. This means the counterbalance may not need to be placed at 1 to give an account of non-trivial cognitive penetration; perhaps there is some 2 or 3 that suffices. So the question becomes, do any of the early period investigations provide evidence for moving cognition's counterbalance closer to 1?
Of all the early investigations, Smeets' experiments reveal considerable evidence for moving cognition's counterbalance away from 10. At which number the counterbalance can be placed will now be the focus of our discussion. What makes Smeets' experiments so relevant for the present discussion is first her method of separating cognition from perception and her focus on differences in phenomenology between patients and controls. All of this is accomplished by devising two separate experiments. The first measures "sensitivity" and the second measures "bias". It is believed that measurements of bias reflect the cognitive while measurements of sensitivity reflect the perceptual component of BID in AN. Between patients and healthy controls, no significant difference is found in the first experiment regarding sensitivity. Thus Smeets concludes that there is no difference between what patients and controls see - see sans scare quotes. However, in the second experiment, patients, opposed to controls, do exhibit a significant bias toward reporting differences in thinness. For Smeets, the results of these experiments support her thesis that BID in AN arises during the reconstruction of the visual body image - reconstruction because the image is retrieved from memory. That said, the idea that memory causes BID in AN is disputed by contemporary findings (see Keizer, 2011). For that reason, I will separate Smeets' experiments' results from Smeets' conclusions.
In order to fix the setting of cognition's counterbalance, we need to determine whether and to what extent the results of Smeets' experiments give reason to claim that the phenomenal character of perceptual experience is altered by cognition in cases of BID in AN. As mentioned above, one of the reasons Smeets' experiments are so relevant for our discussion is how she employs a method of measuring what may be indicative of a difference in phenomenology. In order to show how Smeets was able to do this, Ineed to introduce the setupof both experiments. Before the experiments, photographs are taken of all participants, digital copies are made, and all but one of these copies are distorted to be either fatter or thinner than the original. In both experiments, patients and controls are asked to discriminate incremental differences in weight and size among the images.
In the first experiment, the participant is asked to discriminate the images as either thinner or fatter than her actual shape. This experiment is designed to test whether patients actually see themselves as fatter, whether the sense data have undergone alteration. If a patient actually sees herself as fatter, she should be less capable of determining whether distorted images of herself are different than her actual appearance and thus there should be a significant difference between patients and controls. But patients do not show any significant difference in the first experiment, which indicates that the visual sensory system of patients is functioning properly.
In the second experiment, there is a significant difference between patients and controls. By asking the participants to judge whether a random distorted image is the same as or different than a standard, non-distorted self-image, Smeets is able to measure the cognitive component of BID in AN, referred to as bias. This experiment shows that patients exhibit a bias towards reporting differences in thinness, compared to controls. Interestingly, patients are not merely more inclined to judge that there is a difference, but they are even more accurate at reporting differences in thinness. I will now examine to what extent these results pertain to our discussion of cognitive penetration in AN. But before I do, I would like to dismiss a possible misconception surrounding what I believe to be pertinent about Smeets' experiments.
 This case is largely analogous to Siegel's case of an expert 'seeing as' (Siegel 2013). A similar case of categorical color perception is used by Firestone and Scholl (Firestone and Scholl 2015), which has been shown to provide evidence of an alteration in phenomenal character of color experience (Webster & Kay 2011).
 Throughout the paper, I will be using "agent" in the discussion to denote a person whose psychosomatic health is not under question and thereby include with agent both patients and healthy persons.
 I would like to point out that though Smeets never uses the word 'phenomenology', it can be interpreted that her experiments turn upon detecting phenomenological differences.
 I will use "participant" to refer to both controls and patients in all of the experiment discussed in this paper.
- Quote paper
- John Dorsch (Author), 2016, Does Anorexia Nervosa involve a kind of Cognitive Penetration?, Munich, GRIN Verlag, https://www.grin.com/document/412272