Cog Sci #1
In Kosslyn's 1973 experiments on mental imagery, if subjects, who were asked to memorize an image of a plane, were then asked to focus on the tail of the remembered-image of the plane, we would expect it to take longer for them to confirm whether the plane had a pilot in its cockpit or had a propeller on its front end?
propeller on the front end
Chomsky's theory of syntax illustrates the hierarchical nature of lan-guage (as described by Lashley).
true
What important concept in cognitive science is exemplified by Chomsky's transformational grammar?
Algorithms.
Why are artificial neural networks so useful for cognitive science?
Because they provide a bridge between algorithm and implementation.
Present Shepard and Metzler's conclusion in your own words. Explain their reasoning. What sort of assumptions does it rest on?
If the time it takes to complete the Shepard and Metzler task is linearly related to the amount of rotation needed to map one figure onto the other, then the best explanation of the findings is that there are imagistic representations involved in mental rotation, as introspection would have it. This inference rests on assumptions about the time it would take to perform mental rotation using either imagistic or digital representation. In a digital scheme, processing time ought to depend merely on the amount of information that needs to be processed.
What prediction would someone make about Shepard and Metzler's shape rotation experiment if they thought that all information processing in the brain was digital?
That all rotation processes will take the same amount of time.
Marr described three different stages of information processing in the visual system. Which stage(s) did he think must be explained by his three levels for analyzing cognitive systems in order to provide a complete account of vision?
The primal sketch, the 2.5D sketch, and the 3D sketch.
What type of analysis is illustrated by Marr's three levels of explanation?
Top-down.
Which of the following most accurately describes what is measured by PET?
flow of blood
Explain in your own words what you think we can learn from programs such as ELIZA. Is it important that a person might be fooled by ELIZA into thinking that they were communicating with another human being?
- ELIZA can teach us about the mediated communicative processes assisted by writing. - ELIZA can teach us about the assumption that other entities are "cognitive" in the ways humans are "cognitive," e.g. anthropomorphism.
Some people have suggested that the ELIZA program mimics a conversation between _________ and ___________.
A psychotherapist and her patient
Which of these behaviors only requires serial processing, in the way behaviorists envisioned?
A. Playing golf. B. Writing a poem. C. Planning one's day. D. None of the above D
Which of the following are examples of cognitive systems that can be understood hierarchically?
A. Vision. B. Smell. C. Hearing. D. All of the above. D
Which of Marr's three stages is missing from Ungerleider and Mishkin's account of the visual system?
Algorithmic
Explain in your own words the difference between algorithmic and implementational explanations.
Algorithmic explanations specify the computational steps that a system carries out to achieve a goal, whereas implementational explanations describe the hardware that actually carries out those steps.
Explain in your own words why Marr drew the conclusions he did from Elizabeth Warrington's patients.
Both sets of Warrington's patients have a deficit in object recognition. For patients with right parietal lesions, the deficit stems from a lack of object-centered representations of the shape and spatial arrangement of objects in the environment. These patients have intact linguistic abilities, but they cannot name objects that are seen from a non-standard perspective. For patients with left parietal lesions, the deficit stems from a lack of linguistic abilities. These patients have intact object-centered representations of the shape and spatial arrangement of objects, but they cannot match these representations with names. Marr infers that an important function—perhaps the function—of the visual system is to construct object-centered representations of the shape and spatial arrangement of objects in the environment.
Give an example of a multiply realizable system.
Brake systems on automobiles are multiply realizable.
Can you think of other differences between digital representation and imagistic representation?
Digital representation uses a stock of basic elements (it is symbolic) and has rules for combining those elements into complex representations (it is compositional). Imagistic representation has neither of these features.
After-images are an example of a kind of ambiguity illusion.
False
It can perform the same function in many different types of physical systems.
False
SHRDLU's microworld is much more complex than the world of most human beings.
False
The visual system is the only cognitive system that can be broken down into three levels (as Marr did with vision).
False
One of the groups of rats in Tolman and Honzik's maze experiment went unrewarded for 10 days and was then rewarded after that. What prediction would a strict behaviorist make about these rats' performance after the first day of rewarded behavior? (or Day 11 of the study.)
It should be the same as the rewarded rats' performance on their first day of reinforcement.
Can you think of a way of explaining the results of Kosslyn's experiments without the hypothesis of imagistically encoded information?
Kosslyn's results might be due to demand characteristics—features of an experiment that reveal the experimenter's hypothesis to the participants. Participants might be inferring Kosslyn's hypothesis and acting (perhaps unconsciously) to confirm it. The effect might also be due to the way that digital information is organized in memory. Digital information about the objects might be stored according to location, in which case search time might correlate with distance from focus.
Explain in your own words why latent learning seems to be incompatible with the two basic assumptions of behaviorism.
Latent learning appears to take place in the absence of conditioning processes (association and reinforcement), and hence in the absence of conditioning. Consider the insight experiments. The rats that did not initially receive a reward were learning about their environment, but there were no conditioned stimuli being paired with unconditioned stimuli, and none of their behaviors were being rewarded or punished.
Why did some cognitive scientists interpret Shepard & Metzler's results as evidence of non-digital information-processing?
The processing time for digitally coded information should be a function only of the amount of information encoded.
Which of the following most accurately describes the two visual systems hypothesis, as proposed by Ungerleider and Mishkin?
The ventral system recognizes and identifies objects. The dorsal system locates objects in space.
example of a genuine algorithm, perhaps from elementary arithmetic or perhaps from everyday life.
There is an algorithm for navigating a four-way stop. There is also an algorithm for playing a faultless game of TIC TAC TOE.
What is an algorithm?
A finite set of rules that can be applied systematically to transform objects in circumscribed ways.
Explain in your own words Gregory's theory about function of qualia.
Because Gregory believes that experience (i.e., the past) plays such an important role in visual perception (as is the case when experience of seeing convex faces gives rise to the Hollow-Face Illusion), Gregory hypothesizes that qualia marks perceptions as being perceptions in the present (rather than the past).
Explain in your own words why the Church-Turing thesis entails that any computer running a program is simply a large and fast Turing machine.
Because a computer program is an algorithm, the Church-Turing thesis entails that there is a Turing machine that does whatever the computer is doing. The key is that these two machines are functionally equivalent. Given any input, both will return the same output. So, there is a sense in which the computer is just a better (faster, more efficient, more user-friendly) version of the Turing machine.
Marr concluded from Elizabeth Warrington's studies that the main job of the visual system is to "derive a representation of a shape." What level of analysis is Marr using to describe the visual system here?
Computational.
Tanya, a cognitive scientist, studies how memory works. Suppose she tells you that memory works to store and retrieve information. Which level of analyzing a cognitive system is she using to describe memory? (in Marr's terms.)
Computational.
Which crucial idea came out of applying the Turing machine to human minds?
Information processing is algorithmic.
Which of the following features of introspective psychology did behaviorism object to?
Introspective psychology claimed to be studying "inner" psychological states.
What is characteristic of a multiply realizable cognitive ability?
It can perform the same function in many different types of physical systems.
Give your own example of a hierarchically organized behavior.
Navigating a busy intersection or making a cup of coffee.
Which model of information processing appears to be supported by Petersen et al.'s studies of single-word processing?
Parallel
What characteristic feature of neural networks is illustrated by Gorman and Sejnowski's mine/rock detector?
Pattern detection
Which of the following features of language is missing from Winograd's SHRDLU?
Pragmatics
Explain in your own words why the experimental results (Tolman et al., 1946) seem to show that rats engage in place learning rather than response learning.
Rats in the first group had to make different sequences of movement depending on where (ie., North of South) they started the maze. Rats (in group 2) for whom the food reward was moved, so they never had to make different sequences of movements ran the maze more slowly than group 1.
Give examples of cognitive abilities that you think would lend themselves to being modeled by artificial neural networks.
Recognizing faces (especially when presented from various perspectives), or detecting emotions from visual and auditory cues.
Which crucial idea came out of Tolman's studies of cognitive maps and has continued to play an important role in cognitive science?
Representations.
Can you identify other ways in which SHRDLU falls short of modeling human understanding of natural language?
SHRDLU does not have to learn the meanings of words and its vocabulary is perfectly adapted to the highly artificial environment that it inhabits. It does not need to interact with other agents inside its environment and it is not part of a linguistic community.
Marr and Ungerleider and Mishkin both provided functional accounts of the visual system. What is the main difference in their approach to functional analysis?
Ungerleider and Mishkin focused more on the neural structures of the visual system.
Suppose I always walk South to get to the cafeteria but just moved to the other side of campus and now I have to walk North. If I were a rat in Tolman, Ritchie, and Kalish's study of spatial learning, what would I likely do? (Assume that trips to the cafeteria are rewarding for me, and I go every day.)
Walk North.
