The MU SASHA Blog

The official blog of University of Missouri Skeptics, Atheists, Secular Humanists, & Agnostics

Analytic philosophy: the shit, or just shit?

Welcome to the official MU SASHA daily blog!
First time here? Read this.

Click here to Like our Page on Facebook (or use the sidebar if you’re logged in).
Local to Columbia? Join the Facebook Group, too!

_________________________________________________________________________

Seth Kurtenbach, here.  Over at LessWrong, there is a discussion concerning the merits, or lackthereof, of analytic philosophy.  The discussion is linked here.  The original poster, named Potato, makes many claims to the effect that the community of LessWrong’ers is better at addressing philosophical issues than is the average analytic philosopher.  I am quite unfamiliar with LessWrong, but having perused through the responses to Potato, I am pretty impressed with the skepticism and hesitation to join in on the intellectual masturbation party.  Other LWers point out several flaws in Potato’s reasoning.  I will also say that, as an analytic philosopher, I am committed to the idea that almost all analytic philosophers are wrong and stupid.  So I don’t disagree entirely with bold Potato.  I will briefly address a few of his claims that have not yet been tackled in the comment section.

Skeptical Potato

Potato’s initial claim:

By and large, I would bet money that the devoted, experienced, and properly sequenced LWer, is a better philosopher than the average current philosophy majors concentrating in the analytic tradition. I say this because I have regular philosophical conversations with both populations, and notice many philosophical desiderata lacking in my conversations with my class mates, from my school and others, that I find abundantly on this website.

Analytic philosophy trains one to reason carefully, make fine distinctions, ask unobvious questions, and search for counterexamples to any assertion.  We hold in high regard clarity of argument.  Thus, the discipline’s philosophical desiderata include (but are not limited to) careful, precise, thorough, critical reasoning that is presented clearly.  If an analytic philosopher lacks these desiderata in conversation, it is not because analytic philosophy eschews them or discourages them.  However, Potato is not clear about what the desiderata are, in his mind.  It may be that the analytic philosophers with whom he converses exhibit these properties, but that he does not consider them to be philosophical desiderata.  In that case, it may be that the philosophical desiderata held in high regard by the LessWrong community are not the same as the philosophical desiderata held in high regard by the analytic philosophy community.  Without knowing the philosophical desiderata of the LessWrong community, I cannot say how the sets of desiderata add up.  One might wonder why the philosophical desiderata of an Internet community would be superior to those of an academic discipline with dedicated professionals…

Potato Explains

Potato says,

If they really honed their skills in crushing their opponents arguments, and could transmit this skill to other successfully, then we wouldn’t have so many open questions in philosophy, and we would notice the sort of exponential growth of the power of our methods, like we see in molecular bio.

Now, I’ve been doing analytic philosophy for seven years.  Through all of these years of practice, I’ve noticed my process of reading has changed.  Rather than merely reading for understanding, I also constantly search for counterexamples and objections to every claim I read.  It has made me a slow reader, because the additional computation takes time.  This sentence took me a while to read.  As I read a claim, my brain alerts me to objections and counterexamples with something like red flags and alarms.  Imagine the sentence to be a neighborhood street, lined with mailboxes.  As I read the sentence, it is like walking down the mailbox-lined street, and each time I encounter a weak claim, susceptible to an objection or counterexample, a flag goes up.  I stop and make a note of each flag along the way.  The above claim has the general form of a conditional, if-then.  Conditionals are false when the antecedent, the ‘if’ claim, is true, and the consequent, the ‘then’ claim, is false.  In my earlier days of philosophical training, I would have generated a counterexample to this claim, if this be possible, and gone out drinking.  Now, though, I simply make note of this possibility, and search for other potential weaknesses and more charitable interpretations of his claim.  Rather than reading it as a conditional of some sort, we can interpret it as a hypothesis and its prediction.

The hypothesis is: Analytic philosophers hone their skills in crushing their opponents’ arguments, and can transmit this skill to others successfully.

The prediction generated by this hypothesis is:  There are not so many open questions in philosophy, and the power of the analytic method grows exponentially, like in molecular biology.

Thus construed, his claim amounts to a falsified hypothesis, because he alleges that the prediction is false.

But why think that the hypothesis really generates that prediction?  Recall that the methods of analytic philosophy encourage thorough, precise criticism.  This means that there is a focus on finding flaws in existing claims, and on finding new logically possible claims with flaws of their own.  This process results in an expanding of our conceptual landscape, which results in more questions.  So, I don’t think that the hypothesis actually generates a prediction that there be relatively few open questions in analytic philosophy.  Quite the opposite, I would expect the number of open questions to increase.  Analytic philosophy concerns itself with examining questions, criticizing responses to those questions, identifying new questions, and identifying new ways of understanding questions.  This process need not culminate in an answer to questions.  Sometimes, analytic philosophy advances by giving us a better understanding of a question, or by asking a question in such a way that it can be investigated empirically.  Then, a new science is born, and questions can be answered empirically (with no credit given to philosophy!)

So, if this reading captures the essentials of Potato’s claim, then I think his claim misses its mark.  We should not expect relatively few open questions.  But what about the other aspect of the prediction, concerning the exponential growth of the analytic method?  Well, given what I’ve said about the broad and thorough criticism and the conceptual expansion, it seems like we should expect a growth, perhaps not exponential, of the analytic method.  The analytic method, roughly construed, is the analysis of concepts and arguments with the use of formal logic.  Consider this in terms analogous to the scientific method: the systematic observation of nature with the use of experiments (likewise roughly construed).  In what sense is there growth?  For science, the body of facts grows, and our understanding of nature both broadens and deepens.  Our ability to manipulate nature increases as a result.  Analytic philosophy does not aim at amassing empirical facts, so its failure to exponentially grow in this regard is not a mark against it.  As far as understanding goes, analytic philosophy does contribute to an overall increase, by examining questions that are left unexamined, or unexaminable, by science.  But I think the best defense of analytic philosophy’s methodological growth comes from the aggressive surge of development in logic that the 20th and early 21st centuries have seen.  Our ability to analyze arguments, and reasoning in general, has increased at a rate close to exponential.  The field of modal logic is only one small fragment of philosophical logic, and it is already vast beyond all measure.  I intend to spend the rest of my life studying modal logic, and I will never catch up with all the advances made in the 20th century, let alone all those piling on top in the 21st century.  Look at this. And this.  Simply put, our analytic methods have been expanding.  Fucking rapidly.

But we can also interpret his claim as a deductively valid argument.

1. If analytic philosophers hone their skills and can transmit these skills to others, then there are relatively few open questions in philosophy and an exponential growth of the power of the analytic method.

2. There are not relatively few open questions in philosophy.

3. There is not an exponential growth of the power of the analytic method.

__________________

4.  Therefore, analytic philosophers don’t hone their skills, or cannot transmit these skills to others.

My above discussion should show that premise (3) is false, and that premise (1) is doubtful.  So, the most charitable interpretations of his claim fail under scrutiny.  I invite the reader to leave comments with other interpretations that might be more charitable.

Finally, Potato compares analytic philosophy to self-help books:

But let us not forget, that comparing molecular biology and philosophy, is like comparing self-help and physics. We should not be as surprised if a bunch of clever enthusiasts make better self-help than professionals, as if a bunch of clever enthusiasts made better physics than physicists. This is because physicists are better at physics than self-help writers are at self-help, the same is true of biologists and philosophers respectively.

I leave this one to the reader.

_________________________________________________________________________

Seth Kurtenbach is a philosophy PhD student at the University of Missouri.  His research focuses on applications of formal logic and game theory to questions about knowledge and rationality.  He is growing a mighty beard, in order to increase his philosophical powers.  Feel free to contact Seth at SJK7v7@mail.missouri.edu with inquiries about philosophy, logic, guest blogging, or visiting to give a presentation!

Helpful resources:

Godisimaginary.com
Iron Chariots Wiki
Skeptics’ Annotated Bible / Skeptics’ Annotated Qur’an
AtheismResource.com
TalkOrigins.org

YouTubers: Evid3nc3Thunderf00tTheAmazingAtheistThe Atheist ExperienceEdward Current, NonStampCollectorMr. DeityRichard DawkinsQualiaSoup

Blogs: Greta ChristinaPZ MyersThe Friendly AtheistWWJTD?Debunking ChristianitySkepChick, Rationally Speaking.

Advertisements

About Seth Kurtenbach

Philosophy grad student who wandered into a computer science PhD program with a backpack full of modal logic and decision theory.

20 comments on “Analytic philosophy: the shit, or just shit?

  1. gwern
    November 29, 2011

    > But why think that the hypothesis really generates that prediction? Recall that the methods of analytic philosophy encourage thorough, precise criticism. This means that there is a focus on finding flaws in existing claims, and on finding new logically possible claims with flaws of their own. This process results in an expanding of our conceptual landscape, which results in more questions. So, I don’t think that the hypothesis actually generates a prediction that there be relatively few open questions in analytic philosophy. Quite the opposite, I would expect the number of open questions to increase. Analytic philosophy concerns itself with examining questions, criticizing responses to those questions, identifying new questions, and identifying new ways of understanding questions. This process need not culminate in an answer to questions. Sometimes, analytic philosophy advances by giving us a better understanding of a question, or by asking a question in such a way that it can be investigated empirically. Then, a new science is born, and questions can be answered empirically (with no credit given to philosophy!)

    I don’t actually follow this. How does skill at finding flaws in existing claims lead to ever increasing *open* questions? If a claim is still ‘open’ after finding a ‘flaw’, then what is this ‘flaw’? It’s obviously not a proof or disproof. What is ‘openness’?

    If you are actually good at examining a theory and finding errors in it, then the most obvious thing you would expect is to see an ever-increasing heap of discarded or disproven theories; I don’t see this in philosophy, analytic or otherwise, (everything said by Aristotle or Plato or philosopher X is still said by somebody today), while I do see this with the sciences like physics, where the discarding of previous theories is one of their hallmarks – no one still holds alchemy or Ptolemaic astronomy, etc., and in the other direction, the survival of relicts like Freudian psychoanalytics is frequently taken as a criticism of psychiatry or psychology.

    • Seth Kurtenbach
      November 29, 2011

      Thanks for the comment! You said you don’t follow my reasoning when I explain why we should expect analytic philosophy to lead to ever increasing open questions. You say, “If a claim is still ‘open’ after finding a ‘flaw’, then what is this ‘flaw?'” First, it is not the claim that is still open, but the question. Claims are offered as solutions to open questions, and analytic philosophy is very good at finding flaws in those claims, the proposed solutions, and so the questions remain open, i.e., unanswered.

      You say, “If you are actually good at examining a theory and finding errors in it, then the most obvious thing you would expect is to see an ever-increasing heap of discarded or disproven theories;” The questions that remain open are often ones that we have not figured out how to test empirically, or ones that are not amenable to empirical testing. Without empirical testing, it is more difficult to decisively discard or disprove a theory. What happens is that a theory or claim is proposed as a solution to an open question, it is criticized with rational methods, and perhaps whatever empirical data may be brought to bear, and it fails to achieve a consensus. There may still be adherents who seek out new arguments for the theory, or perhaps seek to adjust the theory so as to avoid the criticism, but other philosophers will abandon it and seek new approaches to the question. You shouldn’t expect to see an ever-increasing heap of discarded or disproven theories, but an ever-increasing heap of imperfect attempts to grapple with the questions.

      One striking example of discarded theory occurred in analytic epistemology in the 60’s, when Edmund Gettier provided counterexamples to what was the accepted and traditional theory of knowledge dating back to Plato. His counterexamples achieved consensus agreement among epistemologists, who agreed that the traditional theory was flawed. There were still a few epistemologists who clung to the traditional view, and there are still some today who seek to explain away Gettier’s counterexamples, but this is the nature of analytic philosophy.

      You say, “everything said by Aristotle or Plato or philosopher X is still said by somebody today.” I’m not sure if that is true. I don’t think there are any analytic philosophers who adopt a scholastic metaphysics, for example. The claim is a bit too strong, I think.

      You say that the discarding of previous theories is one of the hallmarks of the sciences, and contrast this with analytic philosophy. Empirical data is responsible for this. Every science began as a sub-branch of philosophy, and once it was discovered how to bring empirical data to bear on the questions, it became easier to discard theories and hypotheses in that specific discipline. Analytic philosophy is not an empirical science, so it doesn’t make sense to evaluate it by scientific standards.

      Although, I would be remiss if I did not mention the Duhem-Quine thesis here. Most scientists and advocates of science hold a roughly Popperian view of science, where theories or hypotheses are tested and either disproven/falsified or tentatively supported. Well, this just isn’t so. It is not possible to test a theory or hypothesis in isolation. One can only ever test the conjunction of a theory/hypothesis and background assumptions. Thus, if a theory/hypothesis predicts an observation, and that observation turns out to be false, one can always continue to accept the theory/hypothesis and reject one of the background assumptions. We saw this recently with the observation that seemed to falsify the theory of general relativity. The scientists who conducted the experiments were very confident in their procedures and results, and those results indicated that the theory was false. However, no one immediately discarded relativity, as Popper says they should have. Instead, everyone immediately began searching for some mistake in the background assumptions. I am not anti-science, I am just pointing out that science is not as clear-cut and dry as many of the more anti-philosophical of its proponents may believe. It is a very messy process, with lots of judgments and decisions to be made about how to interpret data.

      • Brian
        November 30, 2011

        “Edmund Gettier provided counterexamples to what was the accepted and traditional theory of knowledge”

        What is so hard about “beliefs are probabilistic”? If a human’s belief is certain, it’s not justified; “I beseech you, in the bowels of Christ, think it possible that you may be mistaken.”

        “Most scientists and advocates of science hold a roughly Popperian view of science, where theories or hypotheses are tested and either disproven/falsified or tentatively supported…I am not anti-science, I am just pointing out that science is not as clear-cut and dry as many of the more anti-philosophical of its proponents may believe.”

        Beliefs are probabilistic. Do a little experiment, get a little more evidence, update on the evidence, get down tonight!

      • Seth Kurtenbach
        November 30, 2011

        Thanks Brian. You ask what is so hard about “beliefs are probabilistic”? Nothing. I get it. The Gettier problem is not about psychological certainty, or about a belief about which one is 100% confident. I’ll assume that beliefs are probabilistic. Now you still have all the work ahead of you in showing how that addresses the Gettier problem.

  2. potato
    November 29, 2011

    Hi, I am potato. Thanks for taking an interest.

    So maybe we shouldn’t expect there to be less open question, that was unclear phrasing. But we should expect questions to get solved or dissolved quickly after they are proposed, and we should expect that the average time taken to dissolve a question by analytic philosophy decreases as they solve more questions. There are however questions in analytic philosophy which are over 2000 years old, and have not been solved though admittedly often elucidated into a less overwhelming form.

    Less wrong’s philosophical desiderata are roughly those you mentioned, check out the twelve virtues of rationality, those are really the desiderata I am talking about.

    • Seth Kurtenbach
      November 29, 2011

      Hi there Potato! You say, “we should expect questions to get solved or dissolved quickly after they are proposed, and we should expect that the average time taken to dissolve a question by analytic philosophy decreases as they solve more questions.”

      Can you explain why we should expect this? I don’t necessarily disagree, but I would like to discuss the issue on your terms.

      • potato
        November 30, 2011

        I say this because it was we observe in every other successful scientific field, if we aren’t getting better at it, we aren’t doing it right.

      • Seth Kurtenbach
        November 30, 2011

        I see. So, because in the successful sciences we observe questions getting solved or dissolved quickly, and the average time taken to dissolve a question decreases as more questions get solved, we should expect analytic philosophy to do the same, if it is successful.

        First, I’m not sure that the claim about the sciences is entirely true. I’m not sure science ever solves a problem with certainty, such that it closes the book on it. I am under the impression that scientific solutions are tentative and subject to continual scrutiny. If this is correct, then solving/dissolving problems quickly and an average decrease in problem-solving time are not things that we see happening in science. The process of inquiry is continuous, at least in principle, without an endpoint designated as a final solution.

        Second, how can one ever tell when a problem has been solved, assuming that problems can be solved. I am curious what you think about this. It is one thing to solve a problem. It is another entirely to know that you’ve solved it. It seems by your own views that at best we can only ever say that we’ve probably, or tentatively, solved some problem. It’s not like we have an answer key against which to check our solutions.

        Perhaps the standard by which we evaluate the empirical sciences is as follows: we get better at a science when we become more confident in its results. Then it looks like your argument is:

        1. If we are doing it right, then we become more confident in its results.
        2. If we become more confident in its results, then the average time problems take to become tentatively solved decreases as we tentatively solve more problems.
        3. The average time problems take to become tentatively solved in analytic philosophy does not decrease over time as we tentatively solve more problems.
        4. Therefore, by a modus tollens train of pain, we are not doing analytic philosophy right.

        I don’t think 2 is true. Why should our confidence in the general discipline and its various results have any bearing on how quickly we tentatively solve distinct problems? And again, does it make sense to say that tentative solutions are ever complete, such that a unique time of solution can be identified?

        I don’t think analytic philosophy has been around long enough for us to really evaluate premise 3. Analytic philosophy as a distinct method of doing philosophy arose roughly with Bertrand Russell in the late 1800s and early 1900s. Bertie himself said that insofar as philosophy is capable of solving problems, it is with the analytic methods.

        Having said that, let me count the tentative solutions that we do have.
        1. There is a general consensus that the tri-omni god does not exist.
        2. There is a general consensus that the free will required for moral responsibility is compatible with determinism.
        3. There is a general consensus that scientific realism is true.
        4. There is a general consensus that one should flip the switch in the Trolley Problem, which amounts to a consensus that consequences can sometimes override rules.
        5. There is a rough consensus that the mind is simply physical stuff.
        6. There is a general consensus among decision theorists that one should two-box Newcomb’s problem, which amounts to a general consensus that when evidential decision theory and causal decision theory conflict, one should side with causal, and dominance.
        7. There is a general consensus that knowledge is more than justified true belief.

        All of these can be construed as tentative solutions, subject to continued revision. Without empirical tests to help confirm our solutions, we rarely achieve consensus over 80-90%, but in some subsets we do. These tentative solutions have developed through the analytic method over the past century or so, and they are tentative solutions to mostly ancient problems. So, if we are to evaluate premise 3, I should think there is good reason to think it is false.

        I have my doubts about 1. I think we can do skepticism right without becoming more confident in its results. We can do skepticism right and as a consequence see a decrease in our confidence. But then, skepticism is not really a scientific discipline. This brings me to my last point.

        Analytic philosophy is not an empirical science, so it is a mistake to evaluate it with the standards unique to the empirical sciences, whatever they be.

      • potato
        December 2, 2011

        You say, “First, I’m not sure that the claim about the sciences is entirely true.” I wouldn’t say this either as a bayesian.

        But my probability that the world is not flat is close to one, probably as close to one as it can get. We don’t have to find a solution with 100% probability. We can find solutions with pretty good probabilities.

        Well, honestly I was being ambiguous, “solved” is not here being used as binary. Some questions are more solved than others. So let me rephrase, I predict that LW will be able to solve questions more and faster than analytics. THat sounds awkward in english, so I normally just treat “solved” as binary.

        Some of the consensus you offered are powerfully argued against in LW, such as the newcomb stuff.

        Analytic philosophy mght be influenced by decscion theory, but decscion theorists are not specialist analytic philosophers. Don’t take credit for their mistakes.

        In so far as any practice is not an empirical science, or a deductive field, like mathematics, logic, or programing, it is not a field of knowledge. If philosophy was not an empirical field, it is slowly becoming one, and should. Philosophy which cannot be solved with empirical science or deduction, cannot be solved, and it would not be knowledge if you that you had.

        It is not just successful science that gets better as it progresses. Martial arts, music, painting, engineering, technology, transportation, hunting, cooking, dancing, carpentry, farming, etc. are all a bunch of competences/actions which have advanced as time goes on. Alchemy, astrology, religion, self-help, psycho-analysis,etc. are all human competences/actions which never got better as they kept being practiced. Not being testably better than your predecessors is a sign that your training sucks, not knowing how to test it, is a sign that you should seriously think about the assumptions your practice commits you to.

      • Seth Kurtenbach
        December 2, 2011

        Hi Potato. You say, “Some of the consensus you offered are powerfully argued against in LW, such as the newcomb stuff.

        Analytic philosophy mght be influenced by decscion theory, but decscion theorists are not specialist analytic philosophers. Don’t take credit for their mistakes.”

        I will check out the LW arguments against two-boxing, or against causal decision theory, whichever they may be. When you say that decision theorists are not specialist analytic philosophers, do you mean that no decision theorists are specialist analytic philosophers, or that not all decision theorists are specialist analytic philosophers?

        This claim is very interesting to me because my advisors, Paul Weirich and Zachary Ernst, are both decision theorists. Paul sees decision theory as a sub-field of the study of practical reasoning, and so a sub-field of philosophical logic. I haven’t asked Zac how he views the relationship between decision theory and analytic philosophy, but he is an analytic philosopher and a decision theorist. I will ask him.

        The Philosophical Gourmet Report evaluates analytic philosophy departments, and includes sub-categories based on specialized focus. Among the specialties is decision theory. http://www.philosophicalgourmet.com/breakdown.asp

        The PhilPapers survey has subcategories of philosophers by which one can break down the survey. One of these sub-categories is decision theorists. Click the link, and in the top drop-down bar ‘AOS’, select Decision Theory. http://philpapers.org/surveys/results.pl

        This might show only that some decision theorists are analytic philosophers, and that analytic philosophers generally view decision theory, at least in part, as a sub-discipline of analytic philosophy. It makes sense for there to be some interdisciplinary bleeding in a growing and improving discipline. 🙂

        As a side note, in my experience analytic philosophy is quite welcoming of Bayesian applications. I am a TA for a logic and reasoning course this semester, and we do a section on probability theory and Bayes Theorem. Both of my advisors are advocates of Bayes Theorem in many contexts. They differ only in the extent to which they think it is applied.

        For example, Paul thinks that Bayes Theorem and conditional updates are distinct, because Bayes Theorem addresses probabilities at a moment, while updating occurs inter-temporally. Zac thinks Bayes Theorem pretty well captures how scientific reasoning works. The folks over at LessWrong seem to think that Bayes Theorem alone governs how we reason. This is a substantive claim, in need of argument, and it is reasonable to be skeptical of it. But, in a show of good faith, I am [edited to fix typo] reading through the sequences as time permits.

  3. potato
    November 29, 2011

    Basically I’m saying that analyticism doesn’t get you much further than pre-positivism, LW does on the other hand.

    • potato
      November 29, 2011

      LW can solve problems in ways I think analytics would eventually agree with, which are strictly philosophical, and have not been solved by analytics.

  4. Brian
    November 30, 2011

    “then we wouldn’t have so many open questions in philosophy, and we would notice the sort of exponential growth of the power of our methods, like we see in molecular bio.”

    I think the most charitable interpretation sets aside the literal interpretation of the specific claim (number of open questions) and compares philosophy to molecular biology with the thought that the strong aspect of biology intended resembles quantity of open questions.

    This reading sees the “number of open questions” replaced with “number of unclosed questions.” This contrasts philosophy to molecular biology with its many closed – and open -questions. Notice that the claim you attacked, that there are/should be few open questions in philosophy, would also be questionable if directed at molecular biology.

    • Seth Kurtenbach
      November 30, 2011

      Could you explain your proposed charitable interpretation in more detail? I don’t follow.

      You say, “Notice that the claim you attacked, that there are/should be few open questions in philosophy, would also be questionable if directed at molecular biology.” You may be right. So, if I judge molecular biology based on its number of open questions, I am maybe working under a flawed notion of what makes a science legitimate. I wasn’t making this claim though. I think Potato made that claim, assuming molecular biology passes the test. I only argued that the test was inappropriate for analytic philosophy.

      • Brian
        November 30, 2011

        Seth Kurtenbach :
        …if I judge molecular biology based on its number of open questions, I am maybe working under a flawed notion of what makes a science legitimate. I wasn’t making this claim though. I think Potato made that claim, assuming molecular biology passes the test. I only argued that the test was inappropriate for analytic philosophy.

        For the sake of argument assume Potato had an accurate conception of molecular biology and of philosophy, and assume the criticism of philosophy was based on a relative strength of molecular biology.

        It wouldn’t do to, from scratch, construct our own argument highlighting a superiority of molecular biology such as less reliance on intuitions or (suppose it’s true) a more exacting process of peer review. That is of course necessary if we want to believe things most in accord with the truth, we have to consider the best arguments for positions rather than the worst (of which there are an infinity). But it’s not reading charitably.

        The reason it is a legitimate application of the principle of charity to reinterpret “(large) number of open questions” as “(small) number of closed questions” (or “small ratio of questions closed to questions asked,” etc.) is that those claims are similar and a human might mistakenly phrase one as the other, with that being his or her only mistake in an argument.

        You are of course correct that Potato literally made a claim about open questions.

  5. Brian
    November 30, 2011

    Seth Kurtenbach :
    Now you still have all the work ahead of you in showing how that addresses the Gettier problem.

    What problem? I know what robots are, something of how vision works and the plausibility of invisibility, the size of things descended from wolves with their last common ancestor thousands of years ago, the prevalence and efficacy of facades and Potemkin structures, how many people are (and some people aren’t) confused about the correspondence of things with traits to categories, the various more and less justifiable ways people carve reality, their illusions about how others carve it, about human labels being coherent categories etc.

    If I see a thing with a certain set of “dog-like” behaviors, the physical facts in my brain, the physics of which I am a part, lead me to a certain degree of belief (and beliefs about my beliefs if I am pressed on them (and sometimes even if I am not)) etc. I may be 99.999336001% confident it is a “dog” according to this folk definition, and .0000002% more confident it is a “dog” according to some scientific definition. I will be even more confident that there is a dog within a few yards of the thing I see that is probably a dog (by my lights) – unless the question posed to me is about the thing that I see acting like a dog, that might be a recording, and the question of whether there is or is not a dog is about the distance from the place I perceive rather than where the thing actually is.

    My conclusions, the probabilities I assign, resemble an ideal inference machine’s, if it were working with the same information, to a greater or lesser extent. My processes of reaching those conclusions are also measurable, and at each step too.

    The result will be labeled “knowledge” under various folk and philosophical definitions, and not under others.

    Inherent in my belief that there is a 99.99whatever percent chance of there being a dog near the thing I see is the possibility that what I see is an illusion and there is a dog in a fake rock or behind a bush or something.

    If, with my information, I believe an apparent barn on a movie set is 95% likely to be real, and 3% likely to be a facade, and 2% other, and a better thinker would think it 95% likely to b a facade, and 3% likely to be real, and 2% other, our agreement that it is 98% likely to be either real or a facade is not at all confusing. And if the apparent barn on the set turns out to be a facade, that’s not confusing, nor would it be if it turned out to be real, with all the attendant consequences for our conjunctive and disjunctive predictions.

    It hasn’t been shown that “knowledge” carves reality near its joints. The question “but is it really ‘(label)’ assumes that the label describes a coherent natural category.

    At this point I’m rambling because in attempting to show how the Gettier problems are not problems I have begun to describe the entire world and how at no point is there any feature of it for which the problems reveal inconsistencies in my understanding. I have to stop somewhere, so I’ll do it here. I don’t think I face a problem, or that my epistemology is challenged by the supposed problems.

    • Seth Kurtenbach
      November 30, 2011

      It seems like you are essentially dismissing the entire field of analytic epistemology. I guess this shouldn’t surprise me, if you think all of analytic philosophy is bunk.

      In your post, you present many claims. I am familiar with the basic tenets of bayesian epistemology, and its basic claims. I rarely see those tenets and claims defended or supported with an argument. Your comment does not present an argument.

      You claim that you have knowledge sometimes, and present some examples of knowledge that you have. The goal of analytic epistemology is to complete this sentence: A subject knows that p if and only if … The Gettier problems do not deny that knowledge exists. They provide counterexamples to a particular analysis of knowledge, and there are modified Gettier examples that address other analyses of knowledge in similar fashion. So, when you say that your epistemology doesn’t have to worry about Gettier problems, you need to back that claim up by providing your analysis of knowledge, the necessary and jointly sufficient conditions, and explaining how it avoids a Gettier-style counterexample. All you have done, as far as I can tell, is made bare assertions and told hypothetical stories illustrating your idea of how knowledge works.

      • Brian
        December 1, 2011

        Seth Kurtenbach :
        It seems like you are essentially dismissing the entire field of analytic epistemology. I guess this shouldn’t surprise me, if you think all of analytic philosophy is bunk.

        It would surprise me if a similar position to mine wasn’t represented in analytic philosophy. So your statement has unfair implications. I also implied philosophy is worthwhile as a psychological endeavor to explore intuitions and their logical consequences, and it is worthwhile as a study of logics (mostly ones that humans don’t use much/well), etc.

        Seth Kurtenbach :
        In your post, you present many claims…Your comment does not present an argument…So, when you say that your epistemology doesn’t have to worry about Gettier problems, you need to back that claim up by providing your analysis of knowledge, the necessary and jointly sufficient conditions, and explaining how it avoids a Gettier-style counterexample. All you have done, as far as I can tell, is made bare assertions and told hypothetical stories illustrating your idea of how knowledge works.

        I tried to present as broad a picture as I could, rather than a deep one, to invite a specific attack from the Gettier problems. I tried to list implications of my beliefs as a self-consistent system, to see how you might challenge them, and to elicit either the crushing Gettier argument against them that I’m unable to think of, or a justification for using “knowledge” the way you do.

        The category of knowledge needn’t be built on necessary and sufficient conditions of deductive logic, do you believe human words generally are? If you’re seeking a compromise between the various meanings people intend and context provides for the word on the one hand, and a natural-ish category on the other, I still don’t see why you would expect a single thin line separating always true definitions from always false ones, with such a thing applying for every use of the word by every person.

        Seth Kurtenbach :
        You claim that you have knowledge sometimes, and present some examples of knowledge that you have.

        I’m not sure how you interpreted “I know what robots are,” but I meant to convey I have strong but less than certain beliefs about technology and how humans use the word “robot.” I used the word “know” because of my model of your model of how I would use that word when addressing you, likewise for the other words. I’m not sure what sort of justification of my beliefs you are looking for; I have had a great deal of sensory experience for many years and am frequently able to predict what sensory experiences I will get.

        Seth Kurtenbach :
        The goal of analytic epistemology is to complete this sentence: A subject knows that p if and only if …

        There’s no way analytic philosophy is so misguided as to identify itself, as a field, into a misconceived question like that.

        Surely it’s concerned with careful language and clarity of thought and meaning with regard to justification, and not with composing the most complicated, lengthy, and variable ridden sentences ever written, one or more per person? I.e. completing the sentence or explaining why it isn’t practical (or possibly, I suppose, logically possible) to complete. Your framing of the question prejudges the answer, as if one should expect a simple answer to it, as if it were impossible that in each instance the answer depend on the speaker’s theory of knowledge, and his or her idea of how the other thinks of language, and the strength of the proposition intended, etc.

        I admit I haven’t offered much in the way of support for my views. But it’s clearly (?) immune to the Gettier problem, and you haven’t argued it’s inconsistent, in conflict with any known fact, or wrong, so that will at least be as far as this already lengthy individual comment goes. It’s been respectable to not expect human words to refer to magical Platonic categories since at least Plato’s man-chicken, and I’ll stick with that for “knowledge” as well, though I am open to persuasion.

        I personally used to believe in a Platonic realism even for things such as tables, much less abstract concepts, and I understand the intuitive appeal. I’m not under any illusions about my personal ability to have done better than the field had I lived then. Nonetheless, if you remember where this began, I’m not too impressed that someone belatedly found inconsistencies in theories that treat man-made categories as divine, followed by many philosophers recanting the theories. That’s a relatively unimpressive example of Philosophy’s facility in discarding wrong theories.

    • Seth Kurtenbach
      December 1, 2011

      You say, “Inherent in my belief that there is a 99.99whatever percent chance of there being a dog near the thing I see is the possibility that what I see is an illusion and there is a dog in a fake rock or behind a bush or something.”

      This seems to approach something like a Gettier counterexample. Without a precise analysis of knowledge to work with, I suppose I’ll have to work with a sufficient condition you seem to give. You say, “If I see a thing with a certain set of “dog-like” behaviors… [then] I may be 99.999336001% confident it is a ‘dog’ “. Then, you say, “The result will be labeled “knowledge” under various folk and philosophical definitions, and not under others.” Since I am trying to identify your concept of knowledge, I will assume that it is appropriately labelled knowledge. So your reasoning looks to be this:

      1. If I see a thing with certain appropriate characteristics in certain appropriate conditions, then I believe p with an appropriately high probability.
      2. If I believe p with an appropriately high probability, then I know that p.

      So, appropriately high probability of belief is sufficient for knowledge. We can make this a little less susceptible to counterexamples by reformulating it like this: If I have appropriate evidence for p in appropriate conditions and I believe p with an appropriately high probability, then I know that p. I’ll point out that this incomplete analysis does not require the belief to be true.

      Now, the Gettier-style counterexample you hint at might go like this: I seem to see a dog a few yards from me, next to a big rock. I believe that there is a dog near me, with a probability of .9999874. Based on the above partial analysis, this is sufficient for knowledge. So, I know that there is a dog near me. However, the image you see is not a dog, but is merely a convincing high-tech illusion. Behind the rock, unseen by you, a real dog sleeps. So, your belief that there is a dog near you is true, but it is not true in virtue of the sense perception upon which you based your belief. This does not seem like knowledge, because it is too epistemically lucky. There is no connection between the feature of the world that makes your belief true, and the belief. The evidence supports the hypothesis quite well, but nothing about the evidence makes the hypothesis true; it is something else, not accounted for by the evidence.

      In the first quote above, you say that your probability assignment accounts for such possibilities as the dog hiding behind the rock. This indicates that you think the agent in the dog-illusion case does know that there is a dog near him/her. I don’t see how this works. Let me try to put it in Bayesian terms.

      e = the sensory image of a dog a few yards away near a rock
      h = there is a dog nearby

      Pr(h|e) = [Pr(h) x Pr(e|h)] / ([Pr(h) x Pr(e|h)] + [Pr(~h) x Pr(e|~h)])

      This says, the probability that there is a dog nearby, given the sensory image of a dog a few yards away near a rock, equals the probability that there is a dog nearby and a sensory image of a dog a few yards away near a rock (given that the dog is there), divided by that same probability in the numerator plus the probability that there’s not a dog there and there’s a sensory image of a dog a few yards away near a rock (given that the dog is not there).

      Nowhere in this equation is the possibility of an illusion with an unseen dog behind the rock explicitly addressed. The numerator is just the probability of the conjunction that there is a dog nearby and you seem to see a dog near a rock. The denominator is that same number, plus the probability of the conjunction that there is not a dog nearby and you seem to see a dog near a rock. The Gettier case is one in which there is a dog nearby, and you seem to see a dog near a rock, but the image of the dog is not the dog that is nearby. I’m not sure how this is accounted for. We could hold fixed the image and the hypothesis, but remove the dog from behind the rock, and the agent would be entirely misled. He is just lucky that there is an unseen dog behind the rock, making his belief true.

  6. potato
    December 2, 2011

    I updated my Post and integrated some of your suggestions. Thanks

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Information

This entry was posted on November 29, 2011 by in Author: Seth Kurtenbach, philosophy and tagged , , , .
%d bloggers like this: