University of California, Berkeley
Department of Psychology
Psychology 129/Cognitive Science 102
Scientific Approaches to Consciousness
Answer one (1) question in each of the following five (5) sections. Each question is worth 10 points, so that the exam is worth a total of 50 points. The front of each page should provide more than ample room for your answer – this is only a midterm exam, after all! – but the reverse of each page has been left blank for your convenience.
Do not provide long-winded answers. You have only 10 minutes, on average, for each question, and we are grading accordingly. Write legibly, or we won’t be able to appreciate how wonderful your answers are.
Be sure to provide your name and student ID# on each page.
Part 1. Choose one:
William James (1890) described five characteristics of consciousness apparent to him through introspection. Elaborate on each of these (2 points each).
James’s introspective analysis of consciousness ("the stream of thought") remains a touchstone for all later explorations. The definitions are more important than the names, but in any event the answer should elaborate at least a little on each of the following points:
Every thought is part of a personal consciousness.
Within each personal consciousness thought is always changing.
Within each personal consciousness, thought is sensibly continuous.
Human thought appears to deal with objects independent of itself.
Consciousness is always interested more in one part of its object than in another.
Owen Flanagan (1992) has listed four reasons why psychologists have been reluctant to discuss consciousness. List each of these (2 points total) and define them (2 points each).
Although interest in consciousness was revived with the cognitive revolution in psychology, the dirty secret is that many psychologists and cognitive scientists think they can get along perfectly well without taking an interest in consciousness at all – or, at least, that they should try to do so. The answer should elaborate at least a little on each of these points:
Positivistic reserve, emphasizing publicly observable behavior rather than private, subjective experience.
Piecemeal approach, understanding consciousness in a bottom-up fashion, studying discrete phenomena such as perception or attention without reference to conscious experience.
Conscious inessentialism, the idea that consciousness plays no functional role in behavior, and can be left out of descriptions and explanations of behavior.
The Epiphenomenalist Suspicion, the idea that conscious experience is real enough, but is merely a byproduct of cognitive functioning.
Part 2. Choose one:
What are qualia (3 points)? What is "intentionality" (3 points)? How does the distinction between qualia and intentionality relate to the structuralists’ injunctions against the stimulus error (2 points)? How does the distinction relate to the development of sensory psychophysics (2 points)?
Qualia are "raw feels", the phenomenal qualities of conscious experience. They are often held to be ineffable (indescribable), intrinsic (unanalyzable), private (cannot be compared between persons), and directly apprehended in consciousness (without any mediating inferences). Intentionality is "aboutness", and refers to the fact that states of mind always represent some feature of the world. As Searle puts it, intentionality is the way the mind relates us to the world. The stimulus error consists in ascribing meaning to one’s sensory experience -- seeing a "table" instead of a four-sided surface of various hues and intensities, for example. By strictly enjoining their introspective observers to avoid this "error", the structuralists effectively focused on qualia and ignored intentionality. Nevertheless, the structuralists’ introspective exercises had a lasting impact on psychophysics, and sensory psychology in general, by revealing much about the basic building blocks of sensory experience -- that all hues could be decomposed into black/white/gray and red/greeen/yellow/blue, for example, or that all tastes could be decomposed into sweet, sour, salty, and bitter (and soy sauce!).
What are some of the limitations of introspection as a technique for studying consciousness (6 points)? Specifically, what is the role of memory in introspection (4 points)?
The most important limitation of introspection, at least as practiced by the structuralists, is that it effectively focuses on qualia and leaves intentionality out of the picture. If, as Brentano held, intentionality is the mark of the mental, this was probably a really bad idea. Another problem is that the act of observing one’s own mental states may effectively change the state being observed, somewhat along the lines of Heisenberg’s uncertainty principle in physics. Perhaps the most important limitation is that introspection is not really direct, but instead relies on memory -- that, as James and others argued, "all introspection is retrospection". If so, then introspection isn’t really the on-line self-observation that it pretends to be. The delay between the observation and the report may induce some degree of forgetting, so that the observer, operating from memory, has forgotten what the sensory experience was really like; and it may also induce some errors through reconstruction, so that the observer, attempting to fill in the gaps in memory, infers what the sensory experience must have been like, instead of directly observing what it was really like.
Other problems, discussed by Farthing:
verbal description difficulties (if qualia really are ineffable, intrinsic, and private, they’re going to be hard to express to someone else in words);
censorship (suppression of what one actually observed)
experimental demands (observers may report what they believe the experimenter wants to hear)
lack of independent verification (science is supposed to be publicly verifiable, after all);
substitution of inferences for observations (the kind of thing that Nisbett & Wilson) complain about in their paper "On Telling More than We Can Know").
Part 3. Choose one:
Describe Nisbett and Wilson’s (1977) critique of introspection (4 points). What kind of evidence leads them to this view (3 points)? Is introspection necessarily invalid? Why or why not (3 points)?
Nisbett & Wilson basically elaborate on the point that introspection substitutes inferences for observations. They argue that while we believe that we have direct introspective access to the contents of our own minds, in fact we only infer these contents from our behavior -- the same way we infer the contents of other people’s minds from their behavior, instead of knowing them directly. Basically, their experiments involved (1) showing that people’s behavior was under the control of some feature of the environment; and then (2) showing that this feature rarely or ever was mentioned in people’s reports of why they did what they did. People’s explanations for their behavior were, apparently, just "a priori" theories about why they did what they did, or after-the-fact rationalizations, not some direct readout from their minds. Despite the power of N&B’s arguments, and the popularity of their paper, their arguments have come in for some criticism. For example, N&B’s research methods often militated against accurate reports, for example by entailing long retention intervals or failing to expose individual subjects to all conditions of their experiment. Just as important, there is a distinction between knowing the contents of one’s mind (e.g., whether you like someone) and knowing the processes generating those contents (e.g., the causal link between some attribute and likability). Just because someone can’t say why they like someone, doesn’t mean that their expressed preference is invalid. The fact of the matter is that introspections can be quite valid, as when subjects are given all the information necessary to reflect accurately on their internal judgment processes, they often do quite well.
Identify four (4) varieties of dualism (4 points) and four (4) varieties of monism (4 points). Consider the following (possibly apocryphal) quotation: "Someday I’ll be able to look in my brain and see what I’m thinking." What position on the mind-body problem is reflected in such a view? Why (2 points)?
There are lots of possibilities here, but the answer should elaborate at least a little (just a sentence will do) on any four of these items:
property dualism (the dual-aspect theory)
the dual-aspect theory (property dualism)
also Chalmer’s version of dual-aspect theory
and even Popper & Eccles’ tri-aspect interactionism
And on four of these items:
mentalistic monism, or idealism or immaterialism
the mind-stuff thery
materialistic monism, or materialism
the automaton theory
the functional behaviorism of Watson & Skinner
philosophical behaviorism, including logical positivism & linguistic philosophy
identity theory, including type and token versions
reductionism, including the Churchlands’ intertheoretic reductionism
With respect to the quote, a case can be made for any form of materialism, but also for certain forms of dualism, especially dual-aspect theories and psychophysical parallelism. But a dualist would have no reason to look in his brain to see what he’s thinking -- he’d already know what he’s thinking. Only some kind of materialist, especially of the sort who thinks that consciousness is illusory, would think he would learn anything new.
Part 4. Choose one:
Characterize Daniel Dennett’s position, as you understand it from the exchange in Searle’s book, in terms of Question 3B’s varieties of dualism and monism (5 points). Explain your answer (5 points).
Dennett is a materialist monist, and his views express both functionalism and eliminative materialism. Applying a verificationist perspective, he believes that qualia do not exist and that consciousness is an illusion, because neither can be verified publicly. As a first-person, subjective experience, it is out of bounds of a science that engages in third-person descriptions of objective facts. A scientific explanation of behavior will eliminate references to conscious mental states in favor of a description of brain processes. Like Watson and Skinner, Dennett believes that a science of the mind should confine itself to third-person, objective, descriptions of behavior. In any event, consciousness plays no role in the operation of the "virtual machine" that is the mind. For Dennett, as a functionalist, there is nothing special about consciousness – any "machine" that performs the same functions as the human mind has the same internal states, and that includes conscious mental states. Thus zombies are conscious, because they do everything that humans do, and humans are zombies too, because consciousness doesn’t matter.
Characterize David Chalmers’ position, as you understand it from the exchange in Searle’s book, in terms of Question 3B’s varieties of dualism and monism (5 points). Explain your answer (5 points).
Chalmers presents a blend of materialism and dualism. As a materialist, he is committed to the idea that mental states are causally related to physical states of the brain (and perhaps other information-processing machines as well). As a dualist, he is not prepared to reduce mental states out of existence or scientific discourse. In the final analysis, Chalmers’ double-aspect theory of information marks him as a property dualist, or a dual-aspect theorist generally. The physical aspect of information is an embodiment of some difference or change in the physical state of the world; the experiential aspect gives rise to conscious experience. Therefore, the information physically represented in the brain (or any other information-processing system) naturally gives rise to conscious experience, because conscious experience is one aspect of the state of being informative. Because information-processing is a function, and any system that processes information must (by virtue of the dual nature of information) has conscious experience (as one of those aspects), Chalmers also counts as a functionalist. In a way that is not so different from Dennett, Chalmers argues that any system that performs the same information-processing functions as the human mind will have conscious experiences, just like the human mind does. But while Dennett thinks that this renders consciousness illusory and irrelevant, Chalmers does not: consciousness cannot be reduced out of existence or out of scientific discourse, because it is an intrinsic property of information itself.
Part 5. Choose one:
What is the distinction between "early" and "late" selection as it applies to theories of attention (3 points)? What are the implications of this debate for the notion of "pre-attentive" or "pre-conscious" processing (3 points)? How do capacity theories of attention get around the "early vs. late" debate (4 points)?
Early- and late-selection theories of attention divide on the issue of where the attentional filter" or "attenuator" is placed – between physical and semantic analysis (early-selection), or after semantic analysis (late selection). Thus, they differ in terms of how much, and what kind, of information can be processed preattentively or preconsciously. Early-selection theories hold that preattentive processing is largely limited to aspects of physical structure, and that semantic analyses can occur only after attention has been devoted to a particular object selected on the basis of its physical attributes (spatial location, salience, etc.). Late-selection theories hold that semantic analysis as well as "perceptual" analysis is performed preattentively; attention is required for the selection of a response to the stimulus, not for selection of the stimulus itself. Capacity theories get around this notion by abandoning reference to any sort of filter or bottleneck: processing is flexible, depending on the cognitive resources available to the person and the cognitive resources demanded by the task(s) at hand. If processing has been automatized, that processing – whether "perceptual" or "semantic" in nature -- can be carried out preattentively, in the sense that it makes no demands on cognitive resources. Semantic processes may not be innately automatized, and they may be difficult to automatize, but once they are automatized they can be carried out preattentively.
How is automaticity defined in cognitive psychology (3 points). How do we know whether a process is automatic or effortful (3 points)? How does automaticity bear on our understanding of consciousness (4 points)?
Automaticity are defined as those that (1) are executed independent of the person’s conscious intentions; (2) make no (or very few) demands on cognitive resources; and (3) are performed outside of conscious awareness. Mostly, we know whether a process is automatic or not depending on whether it is independent of the cognitive load placed on the person by ongoing tasks. This is another way of saying that automatic processes do not interfere with other ongoing cognitive processes. Other criteria, such as independence of intention and lack of awareness, tend to be discounted because they are too "mentalistic" in nature. Independence is something that can be measured objectively. Nevertheless, automaticity bears on consciousness because it was the first, and most widely accepted, scientific approach to unconscious mental life. The fact that some processes are performed automatically means that there are some aspects of mental life that are unconscious, in the sense that they are performed outside of conscious awareness and conscious control.