Cognitive Science: An Introduction to the Science of the Mind by Jose Luis Bermudez

Chapter 1

• Section 1.1
  1. Explain the difference between reinforcement learning (operant conditioning) and associative learning (classical conditioning). Give your own example of each.
  2. Describe the experiment that led Tolman to postulate cognitive maps. Explain why the results of the experiment support the postulation.
  3. Explain the difference between a linearly and a hierarchically organized behavior. Why do hierarchically organized behaviors pose a problem for behaviorists?
  4. Explain the hypothesis of task analysis. Give your own example of a task analysis (it need not relate to cognitive science).
• Section 1.2
  1. Describe the basic design and functioning of a Turing machine.
  2. What is an algorithm?
  3. State the Church-Turing thesis.
• Section 1.3
  1. Explain the distinction between the deep (or phrase) and the surface structure of a sentence.
  2. What is the basic aim of transformational grammar?
  3. Construct a phrase structure tree for a sentence of your choosing.
• Section 1.4
  1. Describe the general format of one of Miller’s experiments on absolute judgments.
  2. Explain what chunking is. Give an example of it.
  3. According to Miller, why do we need to chunk information?
  4. Explain the phenomenon of selective attention. Give your own example of it.
• Section 1.5
  1. Explain how the concept of information runs through the episodes looked at in Chapter 1.

Chapter 2

• Section 2.1
  1. In what ways does SHRDLU go beyond a chatterbot?
  2. Give three reasons why SHRDLU was important in the birth of cognitive science.
  3. What are SHRDLU'S three main systems?
• Section 2.2
  1. What is the question at issue in the mental imagery debate?
  2. Describe Kosslyn’s experiments. How do they fit into the mental imagery debate?
• Section 2.3
  1. What are the goals of a computational level analysis?
  2. What does an algorithmic level analysis tell us?
  3. What does an implementational level analysis tell us?
  4. What is Marr’s computational level analysis for the visual system?
  5. What are the major steps in Marr’s algorithmic analysis for the visual system? Briefly describe each.

Chapter 3

• Section 3.1
  1. Explain the concept of multiple realizability. How does it figure into the argument that we can study cognition without studying the brain?
• Section 3.2
  1. Describe the basic methodology of a cross-lesion disconnection experiment. What are such experiments good at discovering?
  2. Explain Ungerleider and Mishkin’s distinction between the “what” and the “where” system. Where is each system located?
• Section 3.3
  1. Describe the key features of artificial neural networks.
  2. Describe the process of training a network.
  3. For what type of tasks are neural networks particularly suited? Give your own example of such a task.
• Section 3.4
  1. Describe the basics of how positron emission tomography (PET) works. What is directly measured?
  2. What are the main differences between the neurological and the cognitive models of single word reading?
  3. What were the conclusions of Petersen et al.’s study of single word reading?

Chapter 4

• Section 4.1
  1. Explain why the hexagon is not a good model of cognitive science.
• Section 4.2
  1. Describe the difference between how psychology and neuroscience are organized.
  2. What are the levels of organization in the brain?
• Section 4.3
  1. Describe the integration challenge in your own words.
• Section 4.4
  1. Why are logic and probability theory considered to be domain-general?
  2. Describe the original Wason selection task.
  3. What is a deontic conditional? Give your own example of one.
  4. Reconstruct Cosmides and Tooby’s argument for the cheater detection module.
  5. Describe a prisoner’s dilemma.
  6. What is the TIT FOR TAT strategy? Why do evolutionary psychologists think it might provide a way of explaining the emergence of cooperative behavior?
  7. Outline the local integration presented in Section 4.4.
• Section 4.5
  1. Why do we need the local integration between neural activity and the BOLD signal?
  2. Briefly describe how fMRI works. Compare it with PET.
  3. What are some hypotheses about the source of the BOLD signal?
  4. Outline the local integration presented in Section 4.5.

Chapter 5

• Section 5.1
  1. Describe the logical positivists’ model of intertheoretic reduction.
  2. Describe Cummins’s model of functional decomposition.
  3. Why do cognitive scientists believe that short-term and long-term memory are different systems?
  4. Describe Baddeley’s functional decomposition of the short-term memory system.
• Section 5.2
  1. Why does Marr's approach represent a possible solution to the integration challenge?
  2. What is the objection to thinking that Marr’s approach represents a global solution to the integration challenge? What is the reasoning that supports this objection (there are at least two points to be made here)?
  3. Explain the modular/non-modular distinction.
• Section 5.3
  1. Describe the mental architectures approach to the integration challenge. What are its three important questions?
  2. Distinguish mental architectures from cognitive architectures.

Chapter 6

• Section 6.1
  1. What is a law of qualitative structure?
  2. What is a physical symbol system? Explain the four key ideas.
  3. What is a recursive or inductive definition?
  4. According to Newell and Simon, which ability lies at the heart of intelligence?
  5. What is a search space? How do computer scientists often represent them?
  6. Explain the notion of a heuristic search technique.
  7. What is a universal Turing machine? Why does it help to illustrate the fourth requirement of a physical symbol system?
• Section 6.2
  1. State the language of thought hypothesis.
  2. What is a propositional attitude? Why does Fodor think that a consideration of propositional attitudes leads to the language of thought hypothesis?
  3. What is the puzzle of causation by content?
  4. Explain why Fodor thinks that the language of thought hypothesis solves the puzzle of causation by content.
  5. Explain the distinction between numbers and numerals. How does it relate to the puzzle of causation by content?
  6. What is the difference between the notions of logical deducibility and logical consequence?
  7. What are the soundness and completeness results? How do they bolster Fodor’s language of thought hypothesis?
• Section 6.3
  1. What claim does the Chinese room argument challenge?
  2. Describe the Chinese room thought experiment.
  3. Describe the Turing test.
  4. Explain the systems reply to the Chinese room argument.

Chapter 7

• Section 7.1
  1. What is the goal of expert systems research?
  2. What is the goal of machine learning research?
• Section 7.2
  1. Explain the basics of how ID3 works.
• Section 7.3
  1. Sketch the basic architecture of WHISPER. Explain the basics of how WHISPER works.
• Section 7.4
  1. What sort of knowledge does SHAKEY have about its environment and itself?
  2. What are the levels of SHAKEY’s software? Briefly describe each.
  3. What is the connection between SHAKEY and Lashley?

Chapter 8

• Section 8.1
  1. Explain why we do not currently have good resources for studying populations of neurons.
  2. Explain the basic structure and functioning of a neuron.
  3. Explain the basic structure and functioning of an artificial neuron. Compare it with the structure and functioning of a real neuron.
  4. Describe several different activation functions. Why do modelers use several different functions?
• Section 8.2
  1. Explain the precise mathematical notion of a function.
  2. What is a Boolean function?
  3. Explain the difference between OR and XOR. What is special about XOR?
  4. Explain what Hebbian learning is.
  5. Describe how the perceptron convergence rule works. What is a major difference between Hebbian learning and learning via the perceptron convergence rule?
  6. Explain the concept of linear separability. How does it pertain to artificial neural networks?
  7. Explain why a perceptron cannot compute a linearly separable function.
  8. Explain why the perceptron convergence rule cannot be applied to multi-layer networks.
• Section 8.3
  1. Describe the basic structure and functioning of a multi-layer network.
  2. Explain how the backpropagation algorithm works.
  3. Describe the worries about the biological plausibility of artificial neural networks.
  4. What is a local learning algorithm? Does it represent supervised or unsupervised learning?
  5. Explain what a competitive network is. How does it differ from a standard artificial neural network? For which sorts of task are competitive networks best suited?
• Section 8.4
  1. Explain the key features of information processing in artificial neural networks.
  2. Explain the difference between localist and distributed networks.

Chapter 9

• Section 9.1
  1. Explain the default hypothesis about what it is to understand a language.
  2. Explain the weak and strong notions of mastering a rule.
  3. Explain Fodor’s argument from language learning for the language of thought hypothesis.
• Section 9.2
  1. Describe the three stages in tense learning.
  2. How do Pinker and Prince explain the trajectory of tense learning?
  3. Describe the basic structure of Rumelhart and McClelland’s network for past tense acquisition.
  4. Describe how Rumelhart and McClelland’s network gets trained.
  5. What is Pinker and Prince’s criticism of the way that Rumelhart and McClelland model the over-regularization phenomenon?
  6. How did Plunkett and Marchman improve on Rumelhart and McClelland’s network?
• Section 9.3
  1. Describe the dishabituation paradigm.
  2. What did it demonstrate?
  3. What is folk physics?
  4. Describe Spelke’s four principles of folk physics.
• Section 9.4
  1. Describe how Munakata models the phenomenon of object permanence.
  2. What is a recurrent network?
  3. Describe the four stages children that go through with respect to the balance beam problem.
  4. Describe how McClelland and Jenkins model the stages of reasoning about the balance beam problem. What are the details of how they train their network?
  5. Outline the different ways in which physical symbol systems and neural networks supporters would view the information processing underlying infants’ physical reasoning.
• Section 9.5
  1. Describe the question of levels and explain why it is important.
  2. Reconstruct Fodor and Pylyshyn’s argument concerning artificial neural networks.

Chapter 10

• Section 10.1
  1. What is an agent? What is an agent architecture? What are the distinguishing features of an agent architecture?
  2. Describe a simple reflex agent. What are condition-action or production rules?
  3. What makes an agent architecture a cognitive system?
  4. What is a goal-based agent?
  5. What is a learning agent?
• Section 10.2
  1. Compare and contrast Fodor’s and Gall’s views. What does Fodor accept about Gall’s view? What does he reject?
  2. What is horizontal faculty psychology? What does the horizontal signify?
  3. What is a vertical cognitive faculty?
  4. What is a cognitive module? What are the characteristics of modular processing? Give some examples of candidates for cognitive modules.
  5. What is central processing? What does Fodor mean when he calls central processing Quinean and isotropic?
  6. Where do Fodor’s cognitive modules tend to fall in agent architectures?
  7. Explain Fodor’s First Law of the Nonexistence of Cognitive Science. What is the tension between the language of thought hypothesis and central processing?
  8. What are some ways of responding to Fodor’s argument for Fodor’s First Law of the Nonexistence of Cognitive Science?
• Section 10.3
  1. Compare and contrast Fodorian and Darwinian modules.
  2. Explain the “domain-specific bodies of knowledge” objection to the massive modularity hypothesis.
  3. Reconstruct the two arguments purporting to show that there must be domain-general processing.
• Section 10.4
  1. Explain the basic setup and functioning of ACT-R/PM.
  2. How does ACT-R/PM represent declarative and procedural knowledge, respectively?
  3. How does ACT-R/PM select which production rule should be active if there is a conflict?
  4. Which parts of ACT-R/PM are symbolic? Which are subsymbolic? Explain why.

Chapter 11

• Section 11.1
  1. Describe the difference between an anatomical and a cognitive map of the brain.
  2. What is the principle of segregation? What are the two ways to interpret it?
  3. What is tract tracing, and why would one do it?
  4. What are some pitfalls to avoid when looking at information about anatomical connectivity?
• Section 11.2
  1. What is the principle of integration?
  2. Describe the different ways of studying the brain while it is active. What are their strengths and weaknesses?
  3. Explain what an event-related potential (ERP) is.
• Section 11.3
  1. Explain the difference between early and late selection models of attention.
  2. Outline how researchers combined different types of study to advance our understanding of the locus of selection problem.
• Section 11.4
  1. What is visuospatial attention? What is covert visuospatial attention?
  2. What are the two main hypotheses about how visuospatial attention works?
  3. Describe a task that tests for spatial working memory.
  4. Describe the research that leads some to believe visuospatial attention is closely linked with spatial working memory.
  5. Describe the research that leads some to believe visuospatial attention is closely linked with preparatory motor control.
• Section 11.5
  1. Describe some of the reasons that one should be cautious when interpreting neuroimaging data.

Chapter 12

• Section 12.1
  1. Describe the stages of development in pretend play. What is common to all types of pretend play?
  2. What is a primary representation?
  3. Explain the three basic observations from which Leslie’s model of the pretense system begins. What does Leslie conclude from them?
  4. What is a metarepresentation?
  5. Describe Leslie’s model of the pretense system. How does he think it relates to the mindreading system?
• Section 12.2
  1. Why does the phenomenon of autism seem to support Leslie’s basic claim that pretend play involves metarepresentation?
  2. Describe the false belief task.
  3. What observation threatens Leslie’s claim that pretend play and mindreading depend on a common system?
• Section 12.3
  1. Describe Baron-Cohen’s model of the mindreading system. How does it develop? What are the functions of its components?
  2. What is joint visual attention?
• Section 12.4
  1. How does Leslie explain the time lag between the capacity for metarepresentation and the ability to pass the false belief task?
  2. Describe an experiment that seems to support Leslie’s selection processor hypothesis.
  3. What is counterfactual thinking? Is it metarepresentational?
• Section 12.5
  1. What is the basic difference between simulationist accounts of mindreading and the accounts of researchers such as Leslie and Perner?
  2. What is standard simulationism? How does introspection figure into the account?
  3. What is radical simulationism? How exactly is it different from standard simulationism?
• Section 12.5
  1. Describe the false photograph task.
  2. Describe the experiments that led Saxe and Kanwisher to believe that there is a neural system dedicated to mindreading.
  3. Explain the evidence that seems to show that low-level mindreading abilities depend on co-opted systems.
  4. What is a mirror neuron?

Chapter 13

• Section 13.1
  1. What is the dynamical systems hypothesis?
  2. What is a dynamical system? Give your own example of one.
  3. What are the key features of dynamical models?
  4. Explain the concept of a state space. What is a trajectory through state space?
  5. According to Van Gelder, what are the four important features of the Watt governor?
  6. What is a basin of attraction?
• Section 13.2
  1. Describe the two different ways of thinking about motor control. Illustrate them with your own example.
  2. Describe the evidence for Thelen and Smith’s dynamical approach to infant walking.
  3. What is the A-not-B error?
  4. How did Piaget explain the A-not-B error? Describe the other cognitive explanation of the error.
  5. Why do Smith and Thelen reject the cognitive explanations of the A-not-B error?
  6. Describe Smith and Thelen’s dynamic field model. What are the three types of input? What do Smith and Thelen think happens when infants make the A-not-B error?
  7. Explain why dynamical systems models may be compatible with more traditional information processing models of cognitive abilities.
• Section 13.3
  1. What is the principal objection that situated cognitive theorists have to more traditional approaches in cognitive science?
  2. Explain why situate cognition theorists criticize SHRDLU.
  3. Explain why situated cognition theorists criticize SHAKEY.
  4. What is Rodney Brooks’s main criticism of classic AI?
  5. What is the morphological computation movement in robotics?
• Section 13.4
  1. What is a subsumption architecture? Describe the three key features.
  2. How is a subsumption architecture different from a modular architecture? What are some potential advantages of subsumption architectures?
  3. Compare and contrast SHAKEY and Allen.
  4. What are the key features of behavior-based architectures? How do these features distinguish behavior-based architectures from other types of architecture?
  5. What is the distinction between a reactive rule and a behavior?
  6. What task does TOTO perform? How does it do it? What is significant about how it does it?
  7. What are the basis behaviors of the robots in the Nerd Herd? What are some emergent behaviors at the level of the group?