There will be ten questions on the exam. Each question can be answered adequately in two or three sentences.
Questions will be selected from those listed here. The precise details of questions marked with an asterisk (*) may change—the phrases in italics will not necessarily be the same. Questions marked "!!" come in pairs; no more than one question from the pair will be included on the exam.
You may not use books, notes, or other sources during the exam itself. Bring an official examination book (blue book) to write your answers.
See the list of review questions for a summary of the in-class review topics.
- Describe at least two ways in which the principles of behaviorism still influence cognitive science.
- Give an example of a situation in which the theoretical ideas of Karl Lashley still influence cognitive science.
- Describe three important ways in which Alan Turing helped to develop a foundation for cognitive science.
- The notion of information processing is central to cognitive science. Define "information" in a way that would help someone who knows nothing about cognitive science understand what "information processing" means. ["Information is the opposite of uncertainty" will not suffice]
- Some cognitive scientists have invested many hours developing "intelligent machines" such as SHRDLU, SAM, even ELIZA, and most recently WATSON. These achievements may be technologically impressive, but from the perspective of a science, what is gained from these efforts?
- Few notions are more "subjective" or "mentalistic" than the concept of a mental image. Describe the key assumptions that researchers have made in order to study mental imagery scientifically.
- * How would someone using the three-level approach of David Marr or Keith Stanovich address the topic of depression?
- !! Present as effectively as you can the argument that one can study the mind without knowing anything about the brain.
- !! Present as effectively as you can the argument that in order to study the mind one must understand how the brain works.
- !! What is the most serious problem that is encountered when using data from brain damaged subjects to infer how psychological processes are implemented in the brain.
- !! What is the most serious problems that is encountered when using brain imaging data to infer how psychological processes are implemented in the brain.
- * Many studies have demonstrated that vision serves two independent functions ("what" and "where"). Using this research as a model, how might you test the theory that hearing also serves two independent functions?
- * Give an example of (a) a proximate and (b) an ultimate explanation for the finding that men are much more likely than women to use physical aggression to achieve their own goals.
- Based on the notion of the "selfish gene", explain why mothers readily sacrifice their own life to protect their offspring.
- !! Present as effectively as you can the argument that one can fully understand the mind only by using an evolutionary framework.
- !! Present as effectively as you can the argument that one can readily understand the mind without reference to evolution theory.
- * Provide a possible evolutionary explanation for the widespread existence of clinical depression. (The theory need not be correct, but should be plausible and testable)
- !! What is the most argument that would be given by someone who believes that to be successful cognitive science must seek integration among the various contributing disciplines?
- !! What is the most argument that would be given by someone who believes that cognitive science can proceed without integrating the various contributing disciplines?
- Explain (briefly) why the Cosmides & Tooby study using the "casava root" problem, and the Gigerenzer & Hug study using the "day off" problem, both support the cheat detector hypothesis proposed by Cosmides and Tooby.
- Most players in the prisoner's dilemma game will cooperate, and many use a tit-for-tat strategy. From an evolutionary perspective, under what conditions would cooperation be less likely to occur, and defection more likely? Why?
- What is the most important way in which the development of fMRI techniques has advanced cognitive science theories during the last two decades?
- What precautions should be taken when using fMRI data to test hypotheses?
- Give an example in cognitive science of a reductionist approach to theory.
- Why do many observers believe that intertheoretic reduction is not a promising approach for cognitive science?
- * Explain how Marr's three-level analysis of cognitive theory might be used to clarify the relationship between alternative theories of memory.
- What major difficulty does a theorist encounter when trying to model domain-general processes? Why is this difficulty not present when modeling modular processes?
- Why do proponents of the physical symbol system hypothesis regard the computer as a valuable tool for studying human thought processes?
- !! Explain why Newell and Simon believed that problem solving, using heuristic search processes, is the heart of intelligent behavior.
- !! Explain why some theorists believe that Newell and Simon were wrong to assert that problem solving, using heuristic search processes, is the heart of intelligent behavior.
- What is the most important role played by propositional attitudes in theories of human thought processes?
- How does Fodor's "language of thought" hypothesis extend the earlier work of Newell and Simon? (What does it add to Newell and Simon's theory?)
- !! Present as best you can the argument made by Searle and others that his Chinese Room example demonstrates the inadequacy of the physical symbol system hypothesis.
- !! Present as best you can the argument that Searle's Chinese Room example is not a valid critique of the physical symbol system hypothesis.
- Explain what makes the WHISPER system and the SHAKEY robot examples of physical symbol systems.
- * Explain what is meant by a means-end search strategy, using the behavior of the SHAKEY robot to illustrate the strategy.
- Many physical symbol system models of thought processes use production systems to describe thinking. Explain which features of a production system make it appropriate for this purpose.
- Physical symbol system models frequently use networks of associations to represent declarative knowledge. These networks have a structure much like that of natural language. Explain why this has proven to be useful for explaining memory processes.
- * Describe how a network of propositional associations would account for recognition ("Did you see X?") and recall ("What did you see?") in tests of memory.
- !! Summarize what you believe to be the most damaging criticism of physical symbol system theories, and explain why you believe it is significant.
- !! Summarize what you believe to be the most damaging criticism of connectionist network models, and explain why you believe it is significant.
- Describe the two or three differences between artificial neural networks and physical symbol systems that are theoretically most important.
- Explain how connectionist network models attempt to address the criticisms that have been made of physical symbol system models.
- Learning in the brain presumably consists of changes in the anatomical or electro-chemical properties of the synapse. How does learning occur in a connectionist network model?
- Explain why multilevel connectionist networks are necessary to fully explain human cognitive abilities. What problems arise when models use three or more layers of nodes?
- Although connectionist models were developed in an attempt to be more faithful to neural anatomy and physiology, the models are still different in many ways from the brain itself. What do you see as the most significant differences?
- !! Defend as best you can the argument that, having easily defeated the best players of Jeopardy, IBM's computer WATSON demonstrates the critical features of human intelligence.
- !! Defend as best you can the argument that, even though it easily defeated the best players of Jeopardy, IBM's computer WATSON does not exhibit critical features of human intelligence.