Introduction to Cognitive Science, Assignment 1

Introduction to Cognitive Science, Assignment 1 I. reptile: Upon discussion of reptile, each member of the group created a different image: an alligator, a gecko and a snake. Since there is no common image for reptile, it suggests that it may be superordinate. Additionally, many of its subordinate categories are basic level, such as snake, which the group was able to accurately image. In the modern taxonomical hierarchy, reptile is equivalent to mammal, and they play a similar role in colloquial English. The fact that mammal is firmly superordinate establishes to a strong degree of certainty that reptile is as well. polo shirt: This category is clearly subordinate, as it contains a modifier to shirt, a much better candidate for being a basic level category. In addition to being shorter, shirt is more commonly used in English, as is evidenced by other categories that modify shirt, such as sweatshirt or t-shirt. surgeon: Although there can be an excellent argument made for this as a basic level category, we opted for a different style of categorization that places doctor at the basic level. Doctor contains a vast number of attributes, such as attended medical school, charges lots of money and has bad handwriting, just for starters. Surgeon only adds to this one attribute, the concept of opening up the body, tinkering around, and shutting it again. In addition, surgeon is not commonly used in day to day English, quite the contrary to phases such as “I have a doctor’s appointment”, or “I’m going to see the doctor”. to sleep: This is probably a basic level category, although as a group we experienced some difficulty identifying it as such. To begin with, we concluded that most common English phrases use sleep as a noun rather than a verb. For example, “I didn’t get much sleep last night”. As a verb, however, it generates a common image, and when compared to the surrounding categories, rest and types of sleeping (e.g., a nap), to sleep seems to have the most general motor program and contains the salient information regarding loss of consciousness, possible presence of dreams and other salient attributes. grain: As a group we felt the most difficulty stemming from categorization of grain. To begin with, we had issues characterizing exactly what a grain is, a symptom of oblivious urban living. To aid us in our judgment, we ran a small experiment to determine its use in common categorization by selecting subordinate categories and asking people to name the basic level. For example, dalmatian -> dog. Of 19 participants, 16 identified wheat as a grain, rather than a food or a crop, which is strong evidence for grain as a basic level category. paperback: Paperback is longer and less commonly used than the clear basic level choice in this case, book. It is subordinate to that category, and in fact, paperback only gives one more attribute to a book; categorization as a book implies the most relevant data, that it is bound, printed and so on and so forth. computer: Computer, despite its long size, is certainly a basic level category. It is used commonly in everyday life, and we all interact with computers the same way – through an input device such as a keyboard. This interpretation of computer, however, neglects any superordinate definition of computer, such as electronic machine that governs a system. One could talk about the computer which governs navigation on a space ship, for example. This is an example of the classic ‘oak oak’ phenomenon, and our classification regards the more specific use of computer. honda: The initial response in our group to this prompt was that a Honda is a car, although upon further discussion, we realized that this is not necessarily true. Regardless, Honda is a Japanese conglomerate that manufactures many goods (such as cars, motorcycles, etc.), but in each case referring to the Honda implies a specific type of something, making it a subordinate level category. Despite the fact that we cannot create a clear image for it, which is important for basic and lower level categories, Honda being a type of car places it lower on the hierarchy. SUV: SUV is a subordinate group, lying beneath car on the categorical hierarchy. Car is shorter, learned earlier, contains more of the relevant data (can be driven, holds passengers, has doors, etc.). Despite the growing popularity of the SUV, Tanisha still refers to hers as a car in daily conversation. “Where is your car parked?” is a common question, but “where is your SUV parked?” is far less common. kitten: Despite the early entrance of kitty/kitten into the lexicon (described often in lecture) and its application to a vast number of categories, as a group we felt that the classifications of non-adults was less representative of categorization as a whole. In the adult classification, kitten is certainly a subordinate category lying beneath cat. Cat contains most of the relevant data regarding a cat, with only the aspect of age and size introduced into the equation. II. The “things that people on foot use as containers to carry objects” is a very large category, which is not as clearly defined as one might think. Everything from a backpack to a kayak can be used by someone on foot as a container to carry an object, and the category theorists would no doubt be intrigued by such a category. Wittgenstein is famous for his work with family resemblances, and these criteria can be applied here, but not to great effect. His classic example of game utilizes these because there is no clearly articulated definition of what it means to be a game, and many games are very dissimilar. Here, however, there is a clearly articulated definition, and the concept of family relevance is not relevant. What is relevant, however, is his work with central and noncentral members, for although there is a clearly defined boundary for this category, a backpack is certainly more central than a kayak. Zadeh’s work with fuzzy sets does not shine much light on this category. Since membership is not graded in the sense that nothing is a partial container. Rather, Rosch’s work on prototype theory helps clarify this discrepancy. Although nothing can be a partial container, some examples of the categories are better examples of containers and would be more quickly identified as such. Her initial phase of analysis would not apply here, because this category has nothing to do with perceptual salience or memorability. Her concept of generalization, however, does apply from this phase, as many examples of containers come from a physical similarity. For example, anything with a compartment could be used as a container (while on foot), which could lead to purse if one first thought of backpack. Her second phase, in which she describes how prototypes reveal the internal structure of categories, is relevant, as is her third phase in which she relaxes the correlation between the two. Since some examples are so much better than others, it suggests that there is some internal structure to this category, if not defining clearly how the structure is represented. It is difficult to assess the internal representation of this category, however, because is not often referred to and many of its qualities are ad hoc, as described by Barsalou. This category requires much thought and despite its clear boundaries, a goal is implied that makes ad hoc considerations relevant. The goal determines the prototype, and this goal stimulates some sort of internal model in which one would have to transport things from one location to another on foot. This superimposed goal could be very common, such as carrying books from home to class, or getting an injured kitten up a flight of stairs. Despite the clear definition of the category, Barsalou would hold that the prototype effects stem from application of our cognitive models to theoretical situations. Barsalou’s other work on holistic categories does not apply here, because of the universal aspect of a container.