Since the exam is going to focus on integrating information, I'm just leaving some connections here that I'd like to revisit (in class or just here): 1. computation and language, how are they the same and how do they differ (what makes language more than just computation) 2. the relationship between feeling, the other minds problem and the hard problem 3. Homoncular causality & regress 4. Baldwinian and darwinian evolution 5. panpsychism and hard problem 6. Unthinkable thoughts and UG 7. Is there an independence of syntax and semantics 8. pantomiming and mirror neurons 9. connecting categorization, communication and consciousness 10. evolutionary psychology and the easy problem
Good study points, but keep in mind that the core of the course (and of cogsci itself) is the reverse-engineering of categorization and language ("easy"), not "consciousness" ("hard")!
And as you reflect on the benefits of consciousness (weasel-word -- it's feeling) try to think about whether it has adaptive (i.e., evolutionary) advantages (and why it's so hard to say what they are, even though it's so obvious to introspection).
(1) How is language involved — and not involved — in category learning and categorical perception? How is this related to the Whorf-Sapir Hypothesis (Strong and Weak)?
(3) What is the relation of the Cogito (Sentio) to Searle’s Chinese Room Argument? Why does it not follow that solving the Easy Problem requires solving the Hard Problem?
There will be 4 questions on the exam and you can write about 750 words per question. Make connections within and between the questions. Remember that the purpose of the exam is so that I can tell whether you have full grasped the course as a whole. You can brainstorm together but you must write your answer on your own, because if the words overlap, I have to deduct them from everyone. (Also, if you've understood the course as a whole, each of you will have your own way to tell it.)
SUGGESTION: Take the keyword list, and for each entry, try to make as many cross-connections as you can with the other entries. That's one of the best ways to practice for the final exam (and test your understanding of the course. Every entry should have at least a dozen cross connections, and will many have a lot more! You should be able to construct your own final exams out of them...
Just a few clarifying questions before the exam, some of them will be definitions that I want to make sure are correct:
1. Could you clarify psychokenesis and how it's relevant to the course? 2. The symbol grounding problem is about how symbols get their meanings. We know that symbols need to be grounded in order for there to be meaning/understanding (CRA showed this) but grounding does not guarantee meaning. This is also exemplified by the sense/reference distinction where multiple terms could have the same referent but a different 'sense' about them 3. Could you clarify the importance of underdetermination? 4. To expand upon the connection between mental states and felt states: one way to articulate the hard problem is to ask how/why are mental states felt states (I suppose not all mental states are felt states since there is vegetative function but anything conscious is felt). Why does cognition feel like anything at all and how is it that the physical gives rise to feeling?
1. According to those who believe in "psychokinesis," the mind (feeling) has causal power over matter. If so, that would be the solution to the hard problem. Is it?
2. Not sure what the question here is, but meaning is not just grounding, because to mean something, or to understand something, feels like something. If T3 grounding does guarantee meaning, we just need to know why and how. That too would solve the hard problem.
3. Try it, and if you miss something, I'll tell you. Start with Descartes, and causal explanation, and the T-hierarchy, and the other-minds problem (but there's more)...
4. "Mental" and "conscious" are both just weasel words for felt. Unfelt states are unfelt states. The hard problem is explaining how and why (some) internal states are felt states.
2. There wasn't really a question in there, I was just writing a small summary for my understanding. Grounding definitely doesn't guarantee meaning but it does seem to be prerequisite. We're not sure that all things that can ground have understanding but we do know that all things that can understand have grounding.
3. Underdetermination means that there are possible solutions to a problem but there isn't enough information to be sure which one is correct. This relates to the correlation and causation distinction because if we find a correlation between two things, concluding that one causes the other is underdetermined: there are alternate explanations that could explain the correlation. The motivation of Descartes' Cogito was to find something that one could be certain of, and the answer was that you cannot doubt that you feel. This might not be a correct connection, but you could say that Descartes' was looking for something that wasn't underdetermined, something for which there wasn't doubt or alternative explanation for. By definition a causal explanation is one that is not underdetermined (but of course a causal explanation for one thing may still be underdetermined for larger questions). The goal of cognitive science is to causally explain how we do what we do and this is mainly a question of reverse engineering because if we can reverse engineer our doing-capacities we have a causal explanation for them. The Turing Test is a 'test' of reverse engineering but not all of its levels are satisfactory to explain everything we can do, some are underdetermined. T2 is the level of the Turing Test where a successful candidate can do anything a human can do verbally for lifetime in a way that is indistinguishable from a human. The issue is that humans can do a lot more than just speak: we interact with the world, we move, speak etc. and so T2 is underdetermined. T3 and above can do anything and everything a human can do for a lifetime so if we have a candidate that can pass it we have a causal explanation for our doing capacities, I’m not sure if anything above T3 would still have an underdetermination problem? For the OMP we can’t be sure that anyone other than ourselves feels and what they feel (of course we can infer it but in terms of Cartesian certainty we can’t be sure). Is underdetermination relevant here because we haven’t figured out a causal mechanism for feeling and so the ‘evidence’ that someone else feels isn’t enough for us sure? I know that underdetermination is relevant in categorization and other aspects of the course but this is already very long so I think I’ll leave it here. Hope this makes sense!
So just to make sure, what I wrote is correct with the exception of those missing connections (and those connections are definitely important I just wanted to make sure I had the basics down)?
Since the exam is going to focus on integrating information, I'm just leaving some connections here that I'd like to revisit (in class or just here):
ReplyDelete1. computation and language, how are they the same and how do they differ (what makes language more than just computation)
2. the relationship between feeling, the other minds problem and the hard problem
3. Homoncular causality & regress
4. Baldwinian and darwinian evolution
5. panpsychism and hard problem
6. Unthinkable thoughts and UG
7. Is there an independence of syntax and semantics
8. pantomiming and mirror neurons
9. connecting categorization, communication and consciousness
10. evolutionary psychology and the easy problem
11. Benefits of consciousness
DeleteGood study points, but keep in mind that the core of the course (and of cogsci itself) is the reverse-engineering of categorization and language ("easy"), not "consciousness" ("hard")!
DeleteAnd as you reflect on the benefits of consciousness (weasel-word -- it's feeling) try to think about whether it has adaptive (i.e., evolutionary) advantages (and why it's so hard to say what they are, even though it's so obvious to introspection).
DeleteHere are 3 Sample Questions:
ReplyDelete(1) How is language involved — and not involved — in category learning and categorical perception? How is this related to the Whorf-Sapir Hypothesis (Strong and Weak)?
(2) In what way are both imagery and computation homuncular? Is heterophenomenology homuncular? What about T3?
ReplyDelete(3) What is the relation of the Cogito (Sentio) to Searle’s Chinese Room Argument? Why does it not follow that solving the Easy Problem requires solving the Hard Problem?
ReplyDeleteThere will be 4 questions on the exam and you can write about 750 words per question. Make connections within and between the questions. Remember that the purpose of the exam is so that I can tell whether you have full grasped the course as a whole. You can brainstorm together but you must write your answer on your own, because if the words overlap, I have to deduct them from everyone. (Also, if you've understood the course as a whole, each of you will have your own way to tell it.)
ReplyDeleteSUGGESTION: Take the keyword list, and for each entry, try to make as many cross-connections as you can with the other entries. That's one of the best ways to practice for the final exam (and test your understanding of the course. Every entry should have at least a dozen cross connections, and will many have a lot more! You should be able to construct your own final exams out of them...
ReplyDeleteJust a few clarifying questions before the exam, some of them will be definitions that I want to make sure are correct:
ReplyDelete1. Could you clarify psychokenesis and how it's relevant to the course?
2. The symbol grounding problem is about how symbols get their meanings. We know that symbols need to be grounded in order for there to be meaning/understanding (CRA showed this) but grounding does not guarantee meaning. This is also exemplified by the sense/reference distinction where multiple terms could have the same referent but a different 'sense' about them
3. Could you clarify the importance of underdetermination?
4. To expand upon the connection between mental states and felt states: one way to articulate the hard problem is to ask how/why are mental states felt states (I suppose not all mental states are felt states since there is vegetative function but anything conscious is felt). Why does cognition feel like anything at all and how is it that the physical gives rise to feeling?
1. According to those who believe in "psychokinesis," the mind (feeling) has causal power over matter. If so, that would be the solution to the hard problem. Is it?
Delete2. Not sure what the question here is, but meaning is not just grounding, because to mean something, or to understand something, feels like something. If T3 grounding does guarantee meaning, we just need to know why and how. That too would solve the hard problem.
3. Try it, and if you miss something, I'll tell you. Start with Descartes, and causal explanation, and the T-hierarchy, and the other-minds problem (but there's more)...
4. "Mental" and "conscious" are both just weasel words for felt. Unfelt states are unfelt states. The hard problem is explaining how and why (some) internal states are felt states.
Thanks for the reply!
Delete2. There wasn't really a question in there, I was just writing a small summary for my understanding. Grounding definitely doesn't guarantee meaning but it does seem to be prerequisite. We're not sure that all things that can ground have understanding but we do know that all things that can understand have grounding.
3. Underdetermination means that there are possible solutions to a problem but there isn't enough information to be sure which one is correct. This relates to the correlation and causation distinction because if we find a correlation between two things, concluding that one causes the other is underdetermined: there are alternate explanations that could explain the correlation. The motivation of Descartes' Cogito was to find something that one could be certain of, and the answer was that you cannot doubt that you feel. This might not be a correct connection, but you could say that Descartes' was looking for something that wasn't underdetermined, something for which there wasn't doubt or alternative explanation for. By definition a causal explanation is one that is not underdetermined (but of course a causal explanation for one thing may still be underdetermined for larger questions). The goal of cognitive science is to causally explain how we do what we do and this is mainly a question of reverse engineering because if we can reverse engineer our doing-capacities we have a causal explanation for them. The Turing Test is a 'test' of reverse engineering but not all of its levels are satisfactory to explain everything we can do, some are underdetermined. T2 is the level of the Turing Test where a successful candidate can do anything a human can do verbally for lifetime in a way that is indistinguishable from a human. The issue is that humans can do a lot more than just speak: we interact with the world, we move, speak etc. and so T2 is underdetermined. T3 and above can do anything and everything a human can do for a lifetime so if we have a candidate that can pass it we have a causal explanation for our doing capacities, I’m not sure if anything above T3 would still have an underdetermination problem? For the OMP we can’t be sure that anyone other than ourselves feels and what they feel (of course we can infer it but in terms of Cartesian certainty we can’t be sure). Is underdetermination relevant here because we haven’t figured out a causal mechanism for feeling and so the ‘evidence’ that someone else feels isn’t enough for us sure? I know that underdetermination is relevant in categorization and other aspects of the course but this is already very long so I think I’ll leave it here. Hope this makes sense!
A few other links: weak/strong equivalence, degrees of freedom, uncertainty, information, categorization, approximation, explanation...
DeleteSo just to make sure, what I wrote is correct with the exception of those missing connections (and those connections are definitely important I just wanted to make sure I had the basics down)?
Delete