General Test - Sample questions

Written Test for Admissions, 2002-2003 (1 hour)

Instructions: Please carefully read the following passages before answering all the four questions.


I. This paper explores some foundational problems in the categorisation of qualitative states. I approach this issue as a philosopher. I will try to bring out, though, why the issue is essentially interdisciplinary. Philosophical thinking about qualitative states leads directly to Consideration of the techniques used in psychophysics and psychometrics - particularly the techniques used to provide an operational understanding of qualitative identity. Evaluating whether this operational understanding of qualitative identity is adequate, however, immediately intersects with several important areas of philosophical enquiry. The thoughts that I will sketch out are largely programmatic. They point towards defining an interdisciplinary research program.

Let me start with a brief sketch of the philosophical background. Philosophical naturalism is a doctrine, which can be spelled out in many different ways. At the heart of all such spellings out, however, are the two ideas (1) that human beings are physico-chemical organisms which differ from the rest of the biological world only in complexity, and (2) that the biological world is itself just part of the physical world (Horgan 1994). In philosophical circles (and elsewhere) a considerable amount of skepticism has been expressed about whether a naturalistic view of the world can properly accommodate certain aspects of conscious mental life. In particular, it is frequently argued that there can in principle be no naturalistic or scientific account of the purely qualitative character of certain mental states, like the experiences of colors, sounds, smells, pains and other sensations. The distinctive 'feel' of seeing a particular color, or of having a particularly awful headache cannot be captured or explained in the language of science (Block 1978, Jackson 1982). Qualities do not feature in the language of science, and no account of qualities in non-qualitative terms can be adequate.

Many if not most philosophers taking sides on this issue agree with the conditional claim that: if there are mental states with qualitative character then those states cannot be given a full scientific explanation. What they argue about is whether there are such mental states (Type 86). The upshot of the argument, not surprisingly, is a deadlock about how the basic facts are described.

One way of progressing beyond the deadlock would be to question the conditional. Is it really the case that the qualitative character of those mental states cannot be captured scientifically? Philosophers have been quick to jump to this conclusion on the basis of sketchy thought experiments. But reaching a more balanced judgment on this requires a close examination of the disciplines of psychophysics and sensory neurophysiology (Clark 1993). Here we have scientific disciplines which at least claim to offer a satisfactory quantitative treatment of qualitative states.

II. Let me for simplicity's sake concentrate on color experience. One might expect a satisfactory treatment of color experience to find neurophysiological correlates for experiences of different colors. Color experience would arguably be brought into the ambit of the naturalistic if neural mechanisms could be identified subserving different types of color experience. At the very least this is necessary if color experience is to be naturalistically accommodated. But any such account of course presupposes that we have some way of identifying experiences of different colors – that we have a way of categorising color experience. So, philosophers might be advised to ask the preliminary question: is such a categorisation possible?

The standard psychophysical technique (originated interestingly enough by the philosophers Rudolf Carnap and Nelson Goodman) for the categorisation of color experience is based on determining whether a subject thinks that two colored objects are or are not the same color (Goodman 1972, 1977). This in turn is tested operationally by determining whether or not he can discriminate between them. Discrimination is operationally defined in a statistical manner over classes of stimuli. Two classes of stimuli are indiscriminable when and only when a subject is unable to discriminate between them at a level significantly above chance. I will put to one side for the moment the tricky question of what exactly is to count as discrimination here. Stimuli can be proximal or distal. A proximal stimulus would be an event occurring at the sensory transducers and a distal stimulus a property of the external environment (such as wavelength or, most crudely, an actual physical object). The guiding idea here is that stimulus classes are defined in terms of paradigm instances of color - which may be a postbox, the wavelength spectra standardly correlated with light bouncing off postboxes, or the retinal events standardly correlated with the light bouncing off postboxes hitting the eye.

Suppose that, for a given subject and a given set of stimulus classes, we have a list of those stimulus classes that he can and cannot discriminate. For every pair of stimulus classes we know whether or not the subject can discriminate between them. How do we move from this to a categorisation of color experience? What is needed is an ordering of the stimulus classes. The relation of indiscriminability will not serve to define such an ordering, since it is non-transitive. Nonetheless a suitable relation can be defined on the basis of indiscriminability. The techniques to do this all employ overlap between the set of stimulus classes that are indiscriminable. Very roughly, the idea is to use these overlaps to define a relation of betweenness which will permit a partial order to be constructed. In the case of colors this ordering is multidimensional, with a dimension being added when it turns out to be impossible to fit a stimulus class into the ordering as it stands. More sophisticated techniques employ judgements of relative similarity to develop the relevant ordering - in the jargon, a multidimensional quality space (Schiffman, Reynolds and Young 1981).

The details of how this all works are complex and fortunately not directly relevant. It is the basic notion of indiscriminability that concerns me. For completeness, though, let me sketch out the role that this categorization plays in a naturalistic account of color experience. The different dimensions in the quality space each provide a differentiating property. Three such differentiating properties emerge in the case of color experience: these are, of course, brightness, hue and saturation. The general explanatory strategy is to provide a neurophysiological explanation for the structure of the color solid, and hence for the location within the color solid of any particular shade of color. The most popular current account is based on the idea of opponent processors (Hardin 1988, Clark 1993). The axes of the color solid are generated by opponent processors which compare the outputs of different receptors, responding positively to some parts of the spectrum and negatively to other in virtue of their different exhitatory and inhibitory connections to the three sorts of cones. Postulating opponent processors has had considerable success in explaining some of the peculiarities in the structure of the color solid, such as the presence of bluish green but the absence of reddish green and the anomalies in the similarity judgements made by different types of color-deficient subjects.

III. The overall picture is attractive. But it stands or falls with the basic operational notion on which it is constructed - the notion of indiscriminability - and in particular with the idea that indiscriminability can be used to define qualitative identity. We are trying to categorise different color experiences in such a way that we can map them on to what we know about the physical properties of stimuli and various mechanisms in the mind. If the method of categorisation is flawed then the project cannot get started. We have a basic pretheoretical notion of qualitative identity, applicable to color experience and to sensory qualities in general. The crucial question is: does the operational notion of indiscriminability capture that basic pretheoretical notion? I have no fixed views on what the answer to this question is. What I want to do is simply to present some prima facie puzzles or problems which any affirmative answer to the question must solve.

Let me break the question down a little further by posing two sub-questions:

(1) Does indiscriminability entail qualitative identity?

(2) Does discriminability entail qualitative non-identity (difference)?

Both these sub-questions must be answered affirmatively if the more general question is to have an affirmative answer.


Let me conclude. Any satisfactory scientific treatment of qualitative states like seeing colors needs to start from an operational way of categorizing the colors that things look to have. It would require an operational understanding of qualitative identity as applied to seeing colors. The principal candidate in the field is the notion of indiscriminability, but any attempt to use indiscriminability to define qualitative identity runs into problems in two different directions. First, there are problems with the idea that indiscriminability entails qualitative identity. Second, there are problems with the idea that discriminability entails qualitative non-identity. I am not saying that these problems are insuperable. I would be surprised if they were. But the two problems do seem to define an interdisciplinary research area that is of central importance not just to the study of similarity and categorisation in general but also to the scientific study of the mind.

Q1. What is the characteristic of "color" of an object that makes it a quale (qualitative) (singluar; qualia is plural) while, say, "length" of an object is not a quale? Does it have anything to do with the difference between primary & secondary properties?

Q2. Give your critical remarks on the author’s view of ‘scientific naturalism of the qualitative features in human experience’.

Q3. Can the author’s notion of ‘indiscriminability’ be used to scientifically define qualitative identity in experience?

Q4. With reference to the above passage, what are the possible approaches for the scientific study of the mind?

© 2016 Bhaktivedanta Institute
Template by QualityJoomlaTemplates
Guest Book