Please read the chapter from your text that deals with the Verbal Representation of Knowledge. Next, find three topics from the chapter that are of interest to you and discuss what your text has to say on those topics. Next indicate two topics that you would like me to cover in more depth in class.
Let me know if you have any questions.
The first interest that came from the chapter is the fact that there are five main models of representational knowledge. They are the clustering model, the set-theoretical model, the semantic feature-comparison model, the network model, and the cognitive neuroscience model. The clustering model states just that, that typically concepts are organized into clusters. An example of this is when you are looking for words that are similar to each other (example: cabbage and lettuce). This was done by Bousfield and Bower. The set-theoretical model states that concepts are remembered in sets or collections. This can be shown either by category or characteristics. An example of category would be when someone says “bird” and a person would think of different types of birds (like robin, wrens, eagles). An example of characteristics would be when we say a bird is characterized by feathers, flight, and wings. This work was done by Meyer. The semantic feature-comparison model, done by E. Smith and Rosch, states that concepts are remembered as a set of semantic features. These features have to do with the item’s meaning and they include, defining features or essential components, and characteristic features. The network model suggests that each memory is independent, but is connected in a network. There are two main models of this which are TLC (teachable language comprehender) and HAM (human associative memory). The last is cognitive neuroscience model, this is where knowledge is shown in organization of neural networks.
The next section of the text that really interested me was the part that is titled “what amnesic patients tell us when they forget.” This section is about a patient who was diagnosed with encephalitis (condition in which the brain becomes inflamed). From this disease his memory span only lasted a few seconds. He quickly forgets what he just ate or what he said, but his musical ability seems to be intact. He can learn new songs, conduct a choir, etc. This showed us that some parts of the brain stores facts (names, images, and events) and others store how we do things. There are currently two types of amnesia. Retrograde amnesia (disorder of retrieval) is the first and it allows us to recall information acquired prior to the onset of the disorder. The other is anterograde amnesia (disorder of encoding information), which is the loss of information presented after the onset of the memory disorder. Both of these can either be temporary or permanent. Permanent amnesia can also be brought on by severe consumption of alcohol. This disorder is known as Korsakoff’s syndrome – results in bilateral damage to the diencephalon.
The last thing I found interesting in the chapter had to deal with memory consolidation. This section first talked about Muller and Pilzecker and the fact that they found that memory for new material was interrupted by the learning of new information shortly after the original learning. They essentially found that memories needed a certain period of time to be fixed, otherwise, they fade away. The text also stated that certain drugs can block short-term memory (seconds to hours) while other drugs can block long-term memory (hours to months). This shows that the memories are independent of each other.
I would like to learn more about specific studies done with language and knowledge and would like to learn more about TLC and HAM.
Terms: clustering model, the set-theoretical model, the semantic feature-comparison model, the network model, cognitive neuroscience model, TLC, HAM, encephalitis, Retrograde amnesia, anterograde amnesia, Korsakoff’s syndrome, memory consolidation, STM, LTM
Since this is a new chapter, I decided to start at the beginning and chose to write about the different models of the representation of knowledge. I like how the chapter starts off by telling exactly what is meant by knowledge. The text defines it as, “the storage, integration, and organization of information in memory (Solso, MacLin & MacLin, 2005, pp. 261).” This is important to understand before learning of the various models, to know that knowledge is information that has already been processed and that the models then explain how the information is interpreted.
The first model discussed in the chapter is the clustering model. Much like it sounds, the clustering model states that concepts tend to be organized in clusters. An example the book provides is that categorically similar words, such as cabbage, lettuce, and spinach are easily recalled together.
The next model the book talks about is the set-theoretical model. According to this model, semantic concepts are represented by sets of elements, or collections of information. The set can include instances of a category and also attributes, or properties of a category. Under this model, there are two types of relationships that a set can have. The first, universal affirmative, is when all members of one category are subsumed in another category. For example, “all canaries are birds.” The second, particular affirmative, states that only a portion of the members of one category make up the member of another category. For example, “some birds are canaries.”
The third model listed in the text is the semantic feature-comparison model. This model states that the meaning behind a word is “represented by a set of semantic features” (pp. 268). These features that contribute to the meaning of the word can be described by two features. Defining features are “essential aspects of the word” (pp. 268). In other terms, the word would not have the same meaning without these features. The second, characteristic features, are “only accidental aspects” (pp. 268). An example would be when someone uses a “loosely speaking” comparison, for example, “a butterfly is a bird” (pp. 268). A butterfly does not have the defining features of a bird, however, loosely speaking, it has the same characteristics as a bird.
The next model is the network model. In this model, knowledge exists in memory as independent unites connected to a network. Take the canary example, the canary is a smaller set of the larder category “bird.” The “canary” category can also be broken down into its own meaning being “a yellow bird that can sing” (pp. 270). This model is helpful because humans make these kind of associations all the time when retrieving information.
The final model, the cognitive neuroscience model, was talked about the least in the text. A summary of its application is that this model “represents knowledge in the organization of neural networks.” In other words, the knowledge that is gained comes from making connections between the individual units.
Terms: representation of knowledge models; knowledge; clustering model; set-theoretical model; universal affirmative (UA); particular affirmative (PA); semantic feature-comparison model; defining features; characteristic features; network model; cognitive neuroscience model
I would like to learn a little bit more about the last two types of models since they seem to be a little more complex.
I was unable to find a chapter that discussed verbal representation of knowledge however, I did read a chapter discussing the mental representation and manipulation of knowledge. There was various concepts presented in this chapter that discussed how knowledge is represented in our minds. We are able to present our knowledge in two forms, declarative and procedural. For example, knowledge such as our birthday, names, and descriptions of things are declarative. On the other hand, procedural is knowledge of procedures such as tying our shoelaces, driving a car, and adding and subtracting numbers. Throughout this chapter, further concepts discussed how our knowledge is demonstrated through various mental representations including mental imagery, pictures/words, symbols, mental models, ect.
The first topic I found most interesting was the mental models concept. My text describes mental models as knowledge structures that we construct to understand and explain our experiences. An example of a mental model would be our description of how a plane is able to fly. Our model depends on the physical laws and as well as our own beliefs we have of them. After reading about this concept, I've learned that experience is necessary to have more accurate mental models. For example, students who participant in an activity are likely to be more accurate when answering questions about the activity versus students who simply made predictions concerning the activity. Overall, mental models help us organize and understand our knowledge through our experiences.
The next to concepts I found interesting in this chapter were Visual images and Spatial images. According the the text, visual and Spatial images are forms of representation that seem to answer certain kinds of questions for knowledge use. For example, we use our visual imagery to represent visual characteristics such as colors and shapes. Our spatial imagery represents features such as depth dimensions, distances, and orientations. Within our text, evidence suggests that our cognitive processes use both visual appearance of objects as well as the spatial structure of objects in order to have accurate knowledge of images. For example, we demonstrate knowledge of object labels and object information because we have symbolic knowledge (visual imagery.) On the other hand, we also demonstrate the ability to manipulate, understand sizes, and features of objects due to our spatial imagery. Therefore, both forms of imagery play a role in how our knowledge is used.
Since my text didn't seem to have a chapter discussing the verbal representation of knowledge I'll be interested to learn more about this idea in class. However, if we do discuss the mental representation of knowledge I'd like to learn more about the theories that play a role in this concept. More specifically, the dual-code theory and the propositional theory.
Terms: Declarative knowledge, Procedural knowledge, mental maps, visual and spatial images
I had a chapter that was titled "knowing and using language" which talked about how we produce language and certain mistakes we make with language. A lot of this chapter talks about how we actually produce language, not so much the processes in the brain that we go through to produce language, but I will try to talk about topics that apply the best with the class.
The first thing I wanted to talk about was certain problems that arise when communicating. There are certain problems we come across between the visual representation of language and the actual act of speaking language. These terms mainly apply to writing, but can be used to understand verbal language. There are certain things that we call content problems, where we have concern of the matter of what to say. We have a ton of knowledge on a topic and all of this information comes into play when we are solving a problem of what exactly to say. With the term "problem space" we have a certain mental representation of the possible steps that might be taken to reach the solution to what needs to be said. Besides content problems, there are rhetorical problems. These concern the matter of how to say what needs to be said. Within this problem space, we struggle with constructing information to be said that is clear, persuasive, interesting, etc. These are problems we all face when either writing or speaking with others on a topic.
There are current differences between adults and children and their ability to communicate through writing and verbally. Bereiter and Scardamalia found that there is a different planning process when comparing adults and children. When children try to come up with what to say they usually establish a goal to retrieve an idea from memory. Once they have an idea they translate it into a scentence, they continue to repeat this over and over, referred to as knowledge-telling. In this approach they are thinking and then saying what they want others to know. For adults or more mature individuals they use both their content knowledge from the past as well as discourse knowledge and also sample from the environment at hand. This called a knowledge-transforming approach, but adults also incorporate a sense of knowledge-telling in their writing/speaking as well.
The last thing I wanted to talk about was grammatical encoding. There are two stages in this process: functional processing and positional processing. During the first stage (functional) the lemmas (abstract rep. of a word that specifies its semantic and grammatical features) are first selected for use in a sentence to convey a concept that the speaker has in mind at the time. During this process substituion errors can occur where a person may put the wrong word in a sentence, like referring to a tulip instead of a rose. Most of these errors tend to stay in the same class, like how those are both flowers, 99.7% of errors. The second step of functional is to assign the lemma to a gramatical function such as a the subject of a sentence.
The second stage, positional processing, then takes over. This determines the order in which the words selected for the sentence will be used. Although the grammatical function has already been set for the sentence, word order is still open for interpretation and needs to be set. The order of word production is determined by the phrase structure of the sentence to be said. The last step in this process it to attach the inflections that are needed to make sure that the sentence is grammatical. A lot of times -ing will be added to words incorrectly and are the most common mistake in structuring a sentence.
I don't really have specific concepts that I want to know about. I would like to know, however, more about the process that goes on when trying to come up with language. I want to know specifically what may be happening in the brain. My text talked more about the process to create speech with the mouth, ect, but I want to learn more about the thinking process.
terms: positional processing, functional processing, gramatical encoding, content problems, problem space, rhetorical problems, knowledge-telling, knowledge-transforming
I chose a chapter which title is nit original for this task: “Knowledge”. It does not talk directly about verbal representation of knowledge, but at the same time this topic is touched in the chapter as well.
The concept of knowledge is strongly connected with the issue of categorization in our brain. Indeed, to a certain extend we can say that knowledge is categorization, as the last one represents our reflection of the environment; the result of analyzed sensory input. Once we define that an object belongs to a certain category we already know a lot about its basic features. Neurological researches provide a lot of evidence that different categories are processed or stored in different parts of the brain. Such cases as visual agnosia or the research with category-specific neurons uncover some mysteries but at the same time open new questions.
There are several approaches to the problem of categorization. First and the least successful one is the definitional approach to categorization. According to this idea, we decide that an object belongs to a certain category, if it fits the category’s definition. While this idea works perfectly for some things, it seems pretty doubtful for others. Some geometrical figures, for example are easy to define, but abstract things like “glory” are a bit more complicated.
Another influential approach is based on the idea of similarity. Again, there are two ways to look at the problem: using prototypes of exemplars.
The idea of prototype was introduced by Eleanor Rosch in the 1970s. A prototype is “a basic draft” of any or at least the majority of the objects in the category. For example, in case of the category “dogs” it would not be any particular breed, but rather a typical, average dog, which at the same time possesses all the distinctive features of the objects at the category (has fure, can bark, is a mammal, popular as a pet, etc.). it is necessary to point out that all the objects in the category can be divided in groups according to the level of prototypicality. Thus for the category “birds” a sparrow would have high prototypicality (as it closely resembles the category prototype), and a penguin would on the contrary have low prototypicality (as it has low resemblance with the prototype).
The exemplar approach suggests that “the heart of the category” is not a prototype, but an example of the category. Thus in case of “birds” it can actually be a sparrow or an eagle and for “dogs” it can be even your own dog. Thus exemplars are the actual objects that a person has encountered previously in life.
Till the present time it has not been decided finally which of the approaches works best. So far it is believed that people are more likely to use both of them. A very interesting question that actually deserves a separate blog post is the influence of culture and experience on categorization and a choice of prototypes and exemplars.
One more interesting aspect of the topic is the hierarchical structure of the categories. Rosch defines three level in the hierarchy: superordinate (i.e. furniture), basic (chair, table, etc.) and subordinate (kitchen chair, dining table, etc.). What is interesting is that the basic level seems to be more important or useful for our mind. If you ask the participants to list all the furniture objects, they would start with the basic-level ones. If you ask them to name pictures, they would also give the names from the basic category – “guitar” instead of “musical instrument”. Note that the situation might be different in case if the person is an expert in the related field – thus a musician would more likely to name a picture using the subordinate level (like “bass guitar”). So knowledge can actually affect categorization.
The latest and in my opinion one of the most productive approaches is the semantic network. It represents the idea that concepts are arranged in the networks in a way they are organized in mind. All the objects are connected by links of various lengths depending on how typical and relevant the object is to the category. This one of the most flexible models that allows us to explain the results of many researches.
Terms used: semantic networks, categorization, prototype, exemplar, definition approach, superordinate, basic, subordinate level
Topics for class: semantic networks, any evince of categorization from animal cognition
After reading Solso's chapter on the representation of knowledge, I was interested in a few specific models. The first model I took an interest in is the semantic feature-comparison model. The basic assumption of this model is that words can be broken down into a set of semantic features. They consist of both defining features and characteristic features. Validation of words rely more heavily on the essential defining features of the word rather than the characteristics. The textbook uses birds as an example. The word bird is typically used when we are describing something similar to the prototype we have in our heads.
Another interesting topic relates directly with the semantic feature-comparison model of knowledge representation. In my textbook it provided a short paragraph and an informative table on linguistic hedges. Keeping consistent with the bird example, bats share similar characteristics with birds but are not actually birds. The use of linguistic terms like 'loosely speaking' or 'technically speaking' “expand [our] conceptual representations,” (Solso, 2001). So true statements are accounted for by both defining features and characteristic features. A statement using a linguistic hedge like “loosely speaking” is based only on characteristic features, whereas a statement using 'technically speaking' is based on defining features but not the characteristic ones. Therefore a bat is a bird, characteristically speaking. And an ostrich is defined as a bird despite the lack of bird-like characteristics.
The third interesting point in the chapter was about the network model. The network model is difficult for me to explain, partly because I was having trouble understanding one component of it. It is basically a network of relationships composed of knowledge (words) that is stored and retrieved using the associations between two independent units. I found this model interesting because it applies the idea of structure to knowledge and incorporates the time it takes for retrieval to occur. (teachable language comprehender, TLC) I would, however, like to learn more about human associative memory (HAM) in class.
I would like learn more about amnesia, specifically Korsakoff's syndrome. After reading the quote “Many times this [Korsakoff's syndrome] is brought on by a final alcoholic bash in which the patient 'drinks himself blind,'” I find myself curious the extent of alcohol consumption required at a time or in a lifetime for this condition to be a problem.
Terms: linguistic hedge, semantic feature-comparison model, korsakoff's syndrome, network model, HAM, TLC
Cognitive Psychology 6th ed. Robert L. Solso
My textbook labeled this topic as representations of language. At first I found most of the information somewhat confusing. What made the most sense to me was the section that referred to universal grammar. Universal grammar is defined as that which refers to genetically determined knowledge of human language that allows children in all cultures to rapidly acquire the language to which they are exposed. This can best be explained by what is known as the language acquisition device or LAD for short. The LAD is the innate mechanism that presumably analyzes the linguistic inputs to which we are exposed and adjusts its parameters to fit that language (Kellogg 2003). There are two factors that underlie the construction of the LAD. These factors depend on the amount of linguistic input they receive as a child and the parameter settings they create. For example, some kids are raised in a neglectful environment where they are not taught language either by a parent or are completely ignored all together. Disabilities may also hinder a child from learning language. A child who is born deaf is not able to learn language and if they are not taught sign language they tend to create their own gestural language. Therefore the child’ LAD will only allow for minimal language development. This is referred to as absence of input. People who have tried to learn a second language as an adult know that the LAD creates a difficult barrier. As a child the LAD is more open to learning new languages, possibly a second language, anytime previous to puberty.
There are a few examples of deprived children who have never learned language properly. One example is of the girl that was discovered who was beaten for making any noise and was constantly alone in a closet most of her childhood. When she was found people believed she was severely mentally retarded, but in fact she was just prevented from learning language and was punished when she attempted.
Parameter settings deal with how a child knows how to start speaking successfully within the parameters of the language they are introduced to. For example, an English speaking child will know to use pronouns instead of omitting their usage as a child learning Italian might because the Italian language is permissible as omitting pronoun use before verbs. What I would like to discuss more in class deals with two theories. I want to understand the differences and similarities between the two specific viewpoints that are known as either the symbolic or connectionist’s frames of mind.
Terms: Universal grammar, parameter settings, language acquisition device, absence of input
One of the chapters in my cognitive book that covers Verbal Representation of Knowledge is called “ Knowing and Using Language” the specific section is called “Though and Language”. The three topics that I will cover are three hypotheses that aim to explain how we use language in our thought processes. The three hypotheses are; the identity hypothesis, the modularity hypothesis, and the general resource hypothesis.
I will start by summing up the identity hypothesis. This ideas was advanced by behaviorists, they said that thought is equal with sub vocal speech. In other words our thoughts are just inner monologs. Researchers suggested that if we use sub vocal speech to think then muscles in our vocal tract would move when we are thinking. They found mixed results when testing this. I think this idea makes sense however, because in this class we use an evolutionary approach to explain things.. I think that over years of ‘evolving’ or adapting we would have weaned ourselves off of extra muscle use. Even though we may have used our vocal muscle when thinking long ago I think that our body would recognize that it was unnecessary.
It is like the video we watched in class on Thursday - one of the small story-lines included a monkey controlling a machine with its thoughts and hand movements. When the monkey realized that they did not have to be using its hand any more he used only one process to control the machine (his mind… Which is a crazy concept to think about - that’s for another blog!). But, again the animal recognized that it did not need to have both processes working at the same time - when only one was needed.
The second hypothesis we’ll look at is called the modularity hypothesis. This hypothesis states that language is independent of thought and is served by a specialized module of language-specific representations and processes. In other words language has its own module for functioning. A module is a set of processes that are automatic, fast and are separate from other process in the brain - they are localized in one area of the brain.
In this hypothesis the other processes going on in the brain and other information in the brain cannot have any effect on language processes that are going on in the brain. It is an automatic function in the brain… so is it that once we learn our language then those process cannot be interrupted? What happens when we learn a new language? Does this module exist because we are genetically predisposed to have language information in our brains/as a part of our lives? Another general question I have is how does one develop a module in their brain - why is this process completely separate from other processes?
The third hypothesis that is called the linguistic relativity hypothesis which states that thought depends on the properties of language. So thought is shaped by the nature of language. This is an argument that came up in my Psychology of Gender Differences class last year. Our professor said that our thoughts are diminished by a diminished vocabulary. When we want to express a thought the first way that comes to mind is through speech. If we do not practice this or know enough words to convey what we are thinking will we stop thinking in that way?
Researchers who are interested in this hypothesis have suggested that cultures that use different languages will perceived the world differently from others. Because their language defines how they see the world - our language defines how we see the world. There are several differences among languages. Some words exist in one country/language and not in another. Examples of this were given in my cognitive textbook. They used words that defined colors. In some African cultures/languages they only use two words to define colors - one word for warm colors and one word for cool colors. Then there are place like in the United States where we have multiple words to define various colors. So when we perceive different colors in America our brain will recall it based on the terms that we know them as if we have a greater vocabulary of color-base words it makes sense that we would be able to recognize more colors. We would know more colors than languages that only have two words for colors. Though not much research has supported this when tested it still seems to be valid argument.
Two areas of further research that I would like to know more about from this chapter include Broca’s aphasia and Wernicke’s area. These two areas refer to fluidity of speech. In general I am interested in identifying the areas of the brain and their function. Just to map out these neural processes that we have been discussing in class and reading about.
Terms: Identity hypothesis, Sub vocal speech, Adapt, Linguistic relativity hypothesis,
Firstly I have to admit that I didn’t care very much for this chapter. I’m not sure way, perhaps it is just not an area that interests me but I had a very hard time reading it and even harder time walking away with the ideas the chapter gave. But I’m going to try to write this blog and hopefully put it together better that way.
The first thing that caught and somewhat kept my attention in this chapter was the semantic feature-comparison model. The text explains that a lexical unit can be represented by both defining features and characteristic features. The text goes on to explain the differences of a true statement (has both defining and characteristic), a technically speaking statement (defining only), and a loosely speaking statement (characteristic only). To try to come up with my own example is a little difficult, but I’ll give it a try. True statement= A trout is a fish. A technically speaking statement= A dolphin is a fish. A loosely speaking statement= A duck is fish. We search for validation and we look at the two lexical units and if there is major overlap we find it true, if there is some overlap we look at them in further detail, and if there is no overlap then we find it false.
The second thing I would like to discuss is the spreading activation model. Here, as the text explains, you map out the concept(s) and the shorter the lines the stronger the association and the longer the lines, clearly, the weaker the association. For example, if you were to map out (by words and categories you associate with) the word book would have a shorter line, and thus a stronger association to library then to less associated and longer line to tree. This theory could explain priming.
Lastly, I want to address connectionism. The text states that this is a theory of the mind that has a lot of smile units that are connected in a parallel distributed network or a PDP. This theory believes that patterns are not stored but the connection strength between them is. This allows recreation. Instead of rules that govern us the connection between the units does.
Like I previously stated I had some difficulty with this chapter. In the future I would like to understand it more. The ideas in this chapter that I did not present here are very unclear to me and because the class is all but over I will have to learn more on this subject on my own or in additional classes.
Terms: Semantic Feature-Comparison Model, Lexical Unit, Defining Feature, Characteristic Features, Validation, True Statement, Technically Speaking Statement, Loosely Speaking Statement, Spreading Activation Model, Association, Priming, Connectionism, Parallel Distributed Network (PDP)
In memory theories many agree that organization of information is more important to the brain than the actual content of the information. Bower proposed a sort of hierarchy model indicating that our brain sorts information in a kind of phylogenic ladder. According to this model our brain starts with a broad concept and becomes more specific discriminating between stimuli and classifying them into different groups according to how far the attributes of the stimuli deviate from the broad hierarchal concept. Bower's work describes memory only in terms of semantic meaning and does not account for sensory memory etc. His model potrays memory as a library storing information until it is accessed like a filing cabinet. But memory is an active system working, for the most part, outside of our conscious control.
The set-theoretical model of memory is interesting because it describes memory not only as a store for information, but as a sort of fact checker. It describes how our brain confirms assertions by checking attributes of the current situation with the attributes stored in memory. If the brain is able to make the association with the available information the fact is confirmed and the brain accepts it. Kind of like rejecting a null hypothesis in research. The set-theoretical model postulates that the brain organized information is a sort of mental venn diagram and can make logical deductions from this organization. In the case of a universal affirmative all of one group fall under the categorization of another group, in a particular affirmative only a portion of a group falls under another. For example all poodles are dogs is a universal affirmative our brain comes to, the particular affirmative is not all dogs are poodles.
My problem with this theory is that if our brain could organize information in a nice neat venn diagram and make judgments based on that then what would be the need for venn diagrams on paper. I know many people who use these diagrams to consciously help them make sense of information, if the brain did that automatically wouldn't people for the most part act in at least a slightly more logical manner? I believe that theory lacks because it assumes that semantic meaning is the most important thing to our brain, and it's my opinion that it isn't.
Collin's and Quillian's Network models, Smith's semantic feature-comparison model, all these semantic models try to create a systematic way in which our brain organizes just pure information. And that is the flaw in all of them, our brain does not care about pure semantic information. We know that fire is red but our brain doesn't spend energy classifying both the words fire and red into several categories based on attributes of them. Our brain exerts energy into the memory "YOU GOT BURNED BY THAT RED THING LAST TIME STAY AWAY THIS TIME!" Collin's and Loftus proposed a spreading activation theory which gets closer to how our brain creates associations between information but all the results described from the experiments they conducted I believe is not evidence for their web like model. In the experiments they observed how their subjects began to associate information presented but this was an a synthetic setting. The only claim the data provides evidence for is that with a little attention our brain can remember some words. The book uses the words canary and bird a lot and describes how bird is a cognitive category that canary falls under. But when someone says canary the first place my mind goes isn't "OK, BIRD, SMALL, YELLOW etc.." I think of the canary islands and how they are all named after dogs. I think that the closest anyone can get to memory through semantic information is deducing that our brain makes associations, the end. Because everyone makes different associations based on events in their particular lives so their is not a universal systematic information organization process present in our mind. We as conscious thinking beings put value on semantic meaning, but our brain puts value on very different things. Sure there is an aspect that is based on semantics but that is a small part of the whole spectrum of information our brain experiences.
First and foremost, the differentiation between declarative and procedural knowledge is key to understanding knowledge as it was discussed in the chapter I read. Procedural knowledge is what it sounds like. I know how to do something. For example, I know how to make pancakes. But that's not the only dimension of knowing involved, because I also know that Jalapeno peppers do not go well in pancakes (I am just making this up). This second part isn't knowing how to do something, it is knowing a fact about the world; it is declarative knowledge.
This distinction is very much like the difference between implicit and explicit (episodic) memory. The only difference between the former two and the latter two that I understand is that the latter two refer to memory and the former two refer to knowledge.
I looked up some information on declarative memory and semantic memory and found that there are other, small differences between them as well. Declarative knowledge is either known or not known but you can sort of know procedural knowledge. For example, I either know or don't know that Mexico is south of the United States. But I can sort of know how to make cup cakes (I can know that it involves blending eggs and sugar and flour until it has a specific texture and can also involve adding some sort of flavoring element and baking powder. Notice I don't know how much of each I need). Another difference between the two is that declarative knowledge is passed on verbally and all at once. My best friend calls me and I all of a sudden know she's going on a date with the guy she's been crushing on forever. But procedural knowledge involves demonstration. This is why math teachers sometimes work out exactly the same exercises as in the book. Unless we see it done, we don't really know how to do even if it is spelled out to us letter by letter in the book.
I really appreciate the "Network models" section in my text book, partly because it is really conducive to a biological explanation. The only thing missing is a researcher saying that these processes happen in the XYZ gyrus of the cortex. Stunningly, though, all this is applicable to the computer as well. But this ought not to surprise us since a computer works on on/off circuits just like the human neuron.
But the idea is that knowledge is organized in a huge, hierarchical network. This is the Semantic network Model. It is really nice and neat. The items that go lower on the hierarchy "inherit" all the properties of the items above them. Sometime there can be "multiple inheritance" where one item descends from two different networks--I can't think of an example right now. Multiple inheritance can sometimes lead to contradictory inheritence. Again, I can't think of an example. This model is frequently tested by doing reaction time experiments. The idea is, the more closely two items are on the hierarchy, the more quickly we identify them together. For example, saying 'yes' to 'is a mammal an animal' will take less time (by a fraction of a second) than saying 'yes' to 'is a dog an animal.' Interestingly, when I looked more information up about semantic network models one criticism of this model was that although the above scenario is true (we say a mammal is an animal more quickly than a dog is an animal) things change the other way around. So we say a dog is an animal a lot more quickly than we say a dog is a mammal. This is completely contradictory to what is expected because 'mammal' is closer to dog than 'animal'. What this theory is missing is emotions. I think an activation of the amygdala when learning something new might make them 'high priority'. But this is my opinion.
Spreading Activation Model works very much like the Semantic Network model. Basically, you have the same network system but you have a lot of crossing over. The Semantic Network model was a series of decision making nodes going up or down the hierarchical network of items with the connections defining the relationship between the items. What's happening in Spreading Activation Model is that, the decisions are not made in their entirety. So when you get to a node, you activate all the different types of relationships involved and you can end up at a spot that is completely different from where you started. Basically you start with a Pig then you go like this: Pig=>pork=>hot dogs=>football stadium=>Super Bowl Game. I suppose that is the only way a non-human animal could be related to the super bowl game.
I don't really understand the Adaptive Control of Thought propositional network (propositional networks are confusing in general). But this was a motive to look it up. There are things I like very much about ACT. I like the tricode representation of knowledge, the idea that there are multiple ways to represent knowledge. For example, if someone asks 'who's face is on the dime' there are three different things that happen in my brain. I pull up a time when I got a good look at a dime, and it was probably at a grocery store. I have a clear image of a small, gray coin between my index finger and my thumb. Even if I have a picturesque memory of it (I don't) I don't recognize the fellow because I don't know him, but then I remember learning in Fifth grade that it was Franklin D. Roosevelt (I probably should but I don't. This information was conveniently provided by Google). But three things are involved: A temporal string of event (I'm pulling out the coin), A spatial image (though I'm looking at the dime, I'm aware of the cashier in front of me, a bunch of junk food to my left and a woman talking behind me), and an abstract proposition (what I learned in fifth grade). I really like the spatial image part. At first I thought of it like a camera, but that is not the case. You're aware 3-D of everything that should be happening around you when you view that specific item.
terms: Semantic Network Model, Spreading activation model, spatial image, temporal string, proposition, declarative knowledge, procedural knowledge, implicit memory, explicit memory, inheritence, multiple inheritence
First and foremost, the differentiation between declarative and procedural knowledge is key to understanding knowledge as it was discussed in the chapter I read. Procedural knowledge is what it sounds like. I know how to do something. For example, I know how to make pancakes. But that's not the only dimension of knowing involved, because I also know that Jalapeno peppers do not go well in pancakes (I am just making this up). This second part isn't knowing how to do something, it is knowing a fact about the world; it is declarative knowledge.
This distinction is very much like the difference between implicit and explicit (episodic) memory. The only difference between the former two and the latter two that I understand is that the latter two refer to memory and the former two refer to knowledge.
I looked up some information on declarative memory and semantic memory and found that there are other, small differences between them as well. Declarative knowledge is either known or not known but you can sort of know procedural knowledge. For example, I either know or don't know that Mexico is south of the United States. But I can sort of know how to make cup cakes (I can know that it involves blending eggs and sugar and flour until it has a specific texture and can also involve adding some sort of flavoring element and baking powder. Notice I don't know how much of each I need). Another difference between the two is that declarative knowledge is passed on verbally and all at once. My best friend calls me and I all of a sudden know she's going on a date with the guy she's been crushing on forever. But procedural knowledge involves demonstration. This is why math teachers sometimes work out exactly the same exercises as in the book. Unless we see it done, we don't really know how to do even if it is spelled out to us letter by letter in the book.
I really appreciate the "Network models" section in my text book, partly because it is really conducive to a biological explanation. The only thing missing is a researcher saying that these processes happen in the XYZ gyrus of the cortex. Stunningly, though, all this is applicable to the computer as well. But this ought not to surprise us since a computer works on on/off circuits just like the human neuron.
But the idea is that knowledge is organized in a huge, hierarchical network. This is the Semantic network Model. It is really nice and neat. The items that go lower on the hierarchy "inherit" all the properties of the items above them. Sometime there can be "multiple inheritance" where one item descends from two different networks--I can't think of an example right now. Multiple inheritance can sometimes lead to contradictory inheritence. Again, I can't think of an example. This model is frequently tested by doing reaction time experiments. The idea is, the more closely two items are on the hierarchy, the more quickly we identify them together. For example, saying 'yes' to 'is a mammal an animal' will take less time (by a fraction of a second) than saying 'yes' to 'is a dog an animal.' Interestingly, when I looked more information up about semantic network models one criticism of this model was that although the above scenario is true (we say a mammal is an animal more quickly than a dog is an animal) things change the other way around. So we say a dog is an animal a lot more quickly than we say a dog is a mammal. This is completely contradictory to what is expected because 'mammal' is closer to dog than 'animal'. What this theory is missing is emotions. I think an activation of the amygdala when learning something new might make them 'high priority'. But this is my opinion.
Spreading Activation Model works very much like the Semantic Network model. Basically, you have the same network system but you have a lot of crossing over. The Semantic Network model was a series of decision making nodes going up or down the hierarchical network of items with the connections defining the relationship between the items. What's happening in Spreading Activation Model is that, the decisions are not made in their entirety. So when you get to a node, you activate all the different types of relationships involved and you can end up at a spot that is completely different from where you started. Basically you start with a Pig then you go like this: Pig=>pork=>hot dogs=>football stadium=>Super Bowl Game. I suppose that is the only way a non-human animal could be related to the super bowl game.
I don't really understand the Adaptive Control of Thought propositional network (propositional networks are confusing in general). But this was a motive to look it up. There are things I like very much about ACT. I like the tricode representation of knowledge, the idea that there are multiple ways to represent knowledge. For example, if someone asks 'who's face is on the dime' there are three different things that happen in my brain. I pull up a time when I got a good look at a dime, and it was probably at a grocery store. I have a clear image of a small, gray coin between my index finger and my thumb. Even if I have a picturesque memory of it (I don't) I don't recognize the fellow because I don't know him, but then I remember learning in Fifth grade that it was Franklin D. Roosevelt (I probably should but I don't. This information was conveniently provided by Google). But three things are involved: A temporal string of event (I'm pulling out the coin), A spatial image (though I'm looking at the dime, I'm aware of the cashier in front of me, a bunch of junk food to my left and a woman talking behind me), and an abstract proposition (what I learned in fifth grade). I really like the spatial image part. At first I thought of it like a camera, but that is not the case. You're aware 3-D of everything that should be happening around you when you view that specific item.
terms: Semantic Network Model, Spreading activation model, spatial image, temporal string, proposition, declarative knowledge, procedural knowledge, implicit memory, explicit memory, inheritence, multiple inheritence