It’s all relative

Share on FacebookShare on Google+Email this to someoneTweet about this on Twitter

I was talking the other day with a friend about why economists think they’re so smart to the point of ignoring easily accessible data from other fields. Will points me to this Greg Clark review of a new book from the field of economic sociology, Adam Smith in Beijing. Clark doesn’t think much of the reasoning, and my own first reaction was shock at what seemed to be a childishly naive argument worthy of a simpleton if the summary is accurate. Whatever the merits of the book, the arrogance of economists made a lot more sense if these were their adjacent fields whose scholarship they were aware of. Especially considering the relatively large number of economists who switched from physics or mathematics. Here are average GRE scores by intended major for graduate school.

42 Comments

  1. I am somewhat amused to find that they are essentially on par with would-be business students on the quantitative scale. 
     
    OT, the average quant. GRE score for would-be math grad students seems a little low, considering the huge disparity between undergrad math and GRE math…

  2. banking and finance are bracketed out from other business. the latter is 100 pts lower.

  3. It’s an interesting chart. Probably it is skewed by non-native speakers in engineering. The more applied fields tend to rank low, even including medicine. (Seemingly RN and MD are lumped, but in my experience science PhDs who teach med students tend to be unimpressed). I’m pretty sure that “Humanities & Arts – other”, the highest Humanities quantitatively, is mostly linguistics.  
     
    The chart was obviously prepared by a philosopher. Philosophers suffer worse tunnel vision even than economists, so I’d say that the GRE misses something.

  4. In US Economics PhD classes of good universities there are tons of foreign students from non-english speaking countries percentage wise because the US has the best schools. 
    This lowers the verbal and analytical writing score of the GRE. 
    I know of an econ professor who bothered to explain the use of “its” and “it’s” and other things to his econ graduate students. 
    This might be the case for math and a few science related subjects at many universities.

  5. “The chart was obviously prepared by a philosopher” 
     
    Absolutely. I dont find philosophers that analytical. Sure there are a lot of therefores, and statements which purport to follow previous statements, as in mathematics , but they tend not to actually follow; or they argue – like medieval theologians about trivialities. In any case the English language, of itself ( or to more exact any human language) cannot prove anything since it is a highly informal system. The veracity of a sentence cannot be proven, or disproved, by it’s formulation, or grammar. I could imagine it possible to test the grammar of a sentence by a software program, but not the truth of the statement. But good analytical writing probably – as defined in this case – means good writing. So the statement “If a man were to walk off a cliff he would fly to the moon” is better writing than “No. The man, he walk over cliff, he fall off, he die, see?” but the former is nevertheless wrong, and the latter is right.  
     
    Even worse lots of words change meaning depending on context. I read Julian Baggini’s ( British Philosopher) book recently and he gave some time to the “paradox” about the guy who is told he will be hanged next month but it will be a surprise, and works out that it can’t be the last day because he would not be surprised, and thus it could not be the second last day as he would not be surprised, and so on… So he’s safe. 
     
    This kind of thing bugs me because the word surprise in this context merely means not pre-announced. I know there are even mathematical models on this, but seriously – a surprise visit from the environmental health agency to your restaurant means in this context that they are not giving you the exact time. There is to be no real surprise on the behalf of the recipient, and if the visit is to be on the last day, it will be on the last day, at an un-annouced time. If it is the last hour of the last day you will not be surprised ( unless they dont turn up) but the visit will still have been unannounced, you will not have been told the exact hour previous to them turning up, which is all that surprise means in this case. 
     
    But philosophers are still scratching their heads on that one. 
     
    There are plenty of other examples where words mean something only in certain context, and thus analytical writing cannot hope to be truly analytical. Human language is not rigorous enough.

  6. It’s not just “adjacent fields of scholarship” – economists have to deal with everyone around them being stupid about economics, in ways that have have real-life consequences.

  7. Human language is used broadly enough, eoin, that rigorous formulations are permitted within it. We need merely restrict our use of language to accomplish this. 
     
    I would have more sympathy for your argument if the fact that you’d presented it in natural language didn’t imply that it was incorrect. 
     
    Oh, and they came for him on the third day. When he protested, they told him “Surprise!”

  8. “I would have more sympathy for your argument if the fact that you’d presented it in natural language didn’t imply that it was incorrect.” 
     
    I didnt say that. It is, however, unproven.

  9. My problem with philosophers is that they’re plenty analytical, but not very empirical, and that they tend to solve problems by narrowing them, so that they get a good answer within their way of formulating the question that doesn’t necessarily apply to any actuality. But I’ve been assured by people in AI and related areas of linguistics and psychology that philosophers are players in those fields, and I started out thinking that they were just tagalongs and copycats.  
     
    Quibbling about subtleties and abstruse questions in logic, etc. has been very productive in philosophy / logic / systems theory / information theory / game theory etc. It’s often or usually possible to find a commonsense normal-language solution, as with the Baggini problem, but that doesn’t mean that working on these problems is pointless. I do think that philosophers spend too much time on weird hypotheticals, but you shouldn’t be sure that arguments are pointless. (I remember that when I first read about Boolean algebra 46 years ago, my immediate reactions was “Who cares?”) 
     
    the falling of a cliff example makes no sense.

  10. “the falling of a cliff example makes no sense.” 
     
    Sure it does. I am pointing out that I could write a parser to test the grammatical integrity of a an argument but not it’s veracity. Yet, I am pretty sure than analytical writing means, first and foremost, good writing, and the inarticulate defense of the truth will get an F, while the articulate defense of the untrue, or obscure, an A.  
     
    i am really talking about high level philosophy. I believe that some lower level philosophers deal with the complexity of language, even for simple statements. But Marx, to take an instance, did not prove a thing by writing his thesis about capitalism. He wrote well, however ( the communist manifesto is a cracking read), and that is what counts in Grand Theory.

  11. I didnt say that.You didn’t use those words. But you made a definitive philosophical argument in natural language. That is precisely what you conveyed. 
     
    Natural language is perfectly capable of supporting good arguments. It’s just not automatic. Slagging on language because it isn’t necessarily as rigorous as you need it to be is stupid.

  12. There’s a lot more to making a good verbal argument than following grammar or writing with an OK style. That’s the bare minimum for getting anyone to read your stuff, and it’s not a hard standard to reach. Verbal arguments are still criticized on various other grounds, and the example you gave would be quickly rejected as empirically false.  
     
    Yet, I am pretty sure than analytical writing means, first and foremost, good writing, 
     
    No, it doesn’t. 
     
    The inarticulate defense of the truth will get an F, while the articulate defense of the untrue, or obscure, an A. 
     
    You’re right about “inarticulate”, because if you just assert a truth without giving reasons why it is true, you’re not doing anything. It’s the same in science — right answers without scientific backing don’t count. 
     
    I am really talking about high level philosophy. 
     
    Oh, come on.  
     
    As for Marx, he’s hardly a typical philosopher. Some of the things he said were true, and economists since then (including capitalist philosophers) have had to deal with them. Some were not true. The same is true of scientists. None of them are right all the time.

  13. I think it is strange that everyone tries to explain these results away. Where you expecting results which would prove that “the harder it is, the more cash you will earn”. 
     
    I know that many engineering and science students don’t have that great verbal skills. Economics is something strange as well, but different from those. Of course it depends on the class, micro/macro courses are harder. But: I don’t think I have ever had so “airy” courses that the ones I had in the Economics University. The thinking there was neither scientific nor critical: the strange applied approach might come from the non-academic background of higher education in economics.

  14. I read another review of that book Clarke pans so vigorously and would tend to agree with Clarke, if the thumbnail explanation of the book’s contents is correct, that this book is silly. 
     
    Saying that though, ‘economists ignore other disciplines’ is I guess a true statement, but all men are sinners too I’d guess. Clarke’s book. much praised here, has at least one howler about the decrease in interest rates from say the year 1200 to 1850 being caused by some sort of biological change in the ability to defer gratification, when if he understood fractional reserve banking, which as a functioning system first appears in human history in Holland in the 1600′s, he would never had said. If I read the book, maybe I’d find more howlers, though I don’t think his thesis is in any general sort of way silly. 
     
    Also, concerning this Clarke guy, though I don’t want to say that I think that where he’s coming from is generally bad and stupid, since I don’t think so, it seems to me that everything in that book is probably a recapitulation of something Herbert Spencer wrote in 1880 or so, and there is very little ground breaking or new in it.

  15. Physicists have traditionally been preposterously arrogant towards their neighbouring disciplines. The only sign I’ve seen that possibly points to the development of such an attitude among bioscientists is a reluctance to explain their vocabulary. I suspect, though, that that reluctance comes from a fear of outsiders shouting “Is that all it means; why didn’t you say so?”

  16. Physicists have traditionally been preposterously arrogant towards their neighbouring disciplines. 
     
    with great acheivments….

  17. Per language: 
     
    Arabic numerals are really cool, thought I think they were invented by Indians, I guess those Indian guys are really smart! Westerners had numbers long before Arabic numerals though, the great thing about them is that they are very efficient symbol set with rules for manipulation, as anyone whovever tried to divide MMCCLXXVI by CCDVII. 
     
    But to say they are a symbol set, is to say that the symbols themselves, and the rules for manipulation, are completely definable in natural language. In fact, all mathematical symbols are so defineable, any symbol that wasn’t would be every bit as meaningless as ‘trjkjre’. There is no mathematical symbol, like an integral sign, or a square root sign that isn’t. All mathematical theorems can be expressed in ordinary language, in that there is no string of mathematical symbols, including those of formal logics like Fregian predicate logic, that cannot be translated into natural language. 
     
    Which is to say that all of math is a natural language thing, the symbols are useful in the sense that they are tremendously convenient, but they aren’t necessary. 
     
    So, though one can of course be unclear and in confusing in natural language, that natural language is always unclear and confusing would make math impossible. Is math impossible? 
     
    I’d agree that a certain set of philosophers might not think this, but the ones that do are generally ‘analytical’ philosophers. The word ‘analytical’ comes from ‘logical analysis’, coined by Bertrand Russell to describe what he was up to in the first decade of the 20th century. A better description of what Russell was up to would be that ‘analysis’ is Bertrand Russell deciding what English sentences had to mean if one were to try to translate them into a form where they could be written in a convenient system of stenography (symbol set) known as Fregian predicate logic, no matter how bleeding obvious it is that not all meaningful to English speakers, English sentences can be translated into Fregian predicate logic. This notion leads to some fantastically hackneyed discussions on what a man means when he says ‘Carnivorous cows do not exist’ and the like, and fantastically bad discussions about the degree of follicle challengedness the present king of France is suffering under. This doesn’t mean that one cannot ‘analyze’ such statements intelligently, it means that Russell proved that they can be ‘analyzed’ stupidly. 
     
    That analytical philosophers have trouble with language isn’t surprising, the whole school itself is founded on being stupid about language.

  18. Just re that a man who knows what’s coming being surprised, that looks like one of the gazillions of statements that map onto the liars paradox from ancient greece about a cretan, a bunch that always lie, saying that he is lying. Other forms are ‘One can never be certain’, or the old time favorite about the clean shaven barber of seville, who shaves all the men who don’t shave themselves. That doesn’t look confusing at all.

  19. Well, it is not the liar’s paradox at all. I claim there is no logical fallacy to be answered here, and that in common usage of the term “surprise visit” merely means a visit whose exact date is unannounced. The common theoretical argument assumes that both sides assume “surprise visit” to mean “We will definitely surprise you on the day we turn up”. The subject then logically deduces that they cannot turn up any day since he would not be surprised on that day( logically working back from the last day where he wouldn’t be surprised, then the day before that etc.). Because of this logic they can turn up any day and surprise him!  
     
    I claim that philosophers are treating language here like no-one else is, and that language has to mean what it means in real life – for how else would we communicate – and if they want a technical language they need to stop using natural language as a proxy ,and get their own; and if they use human language they should use it as used in normal situations, but given the inexactness of natural language they have very shaky foundations for logic. 
     
    In short a surprise visit means not that the subject will be surprised but that he is not given the actual day and the hour of the visitation. The inexactess of language is a greater philosophical problem then the relatively trivial original “solution” to this non-problem.

  20. Mrs Webster walks into the room to find Mr Webster in the arms of his lover. “Mr Webster I am surprised” she says. “No, my dear” he says, “you are astonished, I am surprised.”

  21. j mct, I know Clark knows about the history of financial systems. He’s written about them. 
     
    There’s a couple hypotheses one can draw here: 
    1. Clark is being disingenuous (purposefully ignoring evidence about financial institutions) 
    2. Clark finds this evidence wanting and finds more explanatory power in the evolving preferences theory 
     
    That having outrageous views sells books is evidence for the 1st hypothesis, but that Clark has been studying economic history and these particular topics for over 20 years suggests the latter is more likely. 
     
    Also, Clark makes no claims to originality. People, of course, have said the difference between nations is accounted by differences in the “quality” of labor (with various meanings of the word quality). Clark’s contribution is a get us closer to the true meaning of “labor quality” and to document evidence in favor of this hypothesis. This is generally, to my understanding, how science progresses.

  22. I do not like today’s analytic philosophy, but the things that Frege, Russell, the logical positivists, Tarski, et al were doing around 1900 — 1950 were necessary precursors of IT and AI and made enormous contributions. And a lot of it as word-game type stuff. 
     
    Eoin, that doesn’t fly. You solved the verbal problem without solving the logical problem it was trying to express. That means either that the logical problem was not expressed well, or that you missed the point of what was going on. In fact, the philosophers who use that kind of story problem are often the same ones who do develop the kind of formal ways of making unambiguous statements that you want.  
     
    I’m sorry that you did poorly in your philosophy class. But I can tell you that it’s very unusual for me to say anything good about analytic philosophy, but you made me do it.

  23. Razib, I think that the takeover of economics by physicists around 1940 or 1950 was a very mixed success. For one thing, they weren’t very good physicists. They’re just now finding out about “long tails”, etc., which (Cosma Shalizi and others tell me) physicists have know about since 1900 or 1930. The systems they described were unnaturally regular.

  24. “I’m sorry that you did poorly in your philosophy class” 
     
    dude, my degree and post grad is mathemathical. I read popular philosophy, and other soft science texts and I have yet to be impressed. 
     
    “you missed the point of what was going on.” 
     
    no i didn’t, the philosophers who think this problem a problem need to learn the actual meaning of “surprise visit” as generally used. The term surprise here, but nowhere else, means unannounced, not “astonishment at the unexpected”. It is precisely because words can change in context that any attempt to use natural language for logical deduction is fatally flawed. ( You are really not even engaging this argument). Another fundamental problem (out of many) is the natural language usage of or. In English or can be either a logical, or exclusive-or. This again depends on context.  
     
    What we cannot speak of we must pass over in silence

  25. Well, I’ll quit after one more try. The philosophers you are talking about, unless they’re incompetent, know all the things you are saying. Exactly the issues you are raising have been a standard area of philosophy for about a century and are part of almost everyone’s early training, but you’re speaking as though this was a great discovery of yours.  
     
    I don’t know what the source of your unhappy experience was — presumably an undergrad class or a book you read — but the conclusions you have drawn from your data point are worthless. There may have been a real problem with the particular expression of the example you gave — I’m not sure even of that — but it doesn’t impact philosophy in general. You have chosen to attack philosophy at one of its strong points, and you would not have done that if you had known what you were talking about.

  26. “presumably an undergrad class” 
     
    Clearly there is a language divide even on this blog. I Thought I explained this: the university system I engaged in specialized in it’s subjects, so I did mathematics only. There was no unhappy experience with philosophy I just think it generally bunk. I also think creationism bunk, and this is not due to bad childhood memories of creationists. Sometimes we think badly of things for logical reasons unrelated to bad childhood, or undergraduate experiences. I do however admit to being a bit queasy about such argumentation, it seems close to ad homimen, no?  
     
    I notice that you seem to indicate that the problems I mention have been noticed before, and that I am not really discovering anything. Never said I was, although it is interesting that I seemto have noticed these issues independently. Kudos to me. Maybe I should take this philosophy up, huh?  
     
    In any case your argument is an appeal to (unaccredited) authority.I take it that these issues have been discussed. Great! Have they been solved. Even better!! Can you explain how, who, when and what?  
     
    See saying that the informality of language issue ( to coin an unwieldily term) has been discussed in the literature is not to explain the solution.

  27. Take a phil class featuring Frege, Russell, the early Wittgenstein, Tarski, Carnap, Reichenbach, and the other logical positivists; or classes in formal / symbolic logic; or philosophy of mind classes at the intersection with AI.

  28. Interesting data- but it’s the low quantitive score for sociologists that concerns me. It adds credibility to the stereotype that the field is just a soap box for activists whose ideas are so unbelievable they need a scientific facade to be able to look someone in the eye.. do we really want these people in our bureaucracy? why else would they be so poor at it? 
     
    Also, the table reinforces my suspicion that some fields need more ‘verbal’ training on the curriculum. For example, one of the biggest problems in the software industry is failing to correctly translate human-language user requirements in to a formal language that computers understand better. There’s much we can learn from the philosophers J Emerson is talking about.

  29. The problem with economics has not been the takeover by physics but the takeover by mathematics. At least those naively enamored of mechanics worried about experimental reality. In contrast, leading economic theorists are more focused on proof and formalization than on experimentation. The worst that can be said of blind worship of Newton is naivete. The worship of Bourbaki (think Debreu or much of the journal Econometrica) produces angels-on-pins scholarship that makes the worst of analytic philosophy seem like good classical physics. 
     
    This is not just a minority opinion. Such distinguished practitioners as Leontief, Friedman, Fisher, Baumol, Vernon Smith, Mankiw, and Levitt have all said as much. 
     
    Indeed, for the most part the leading young empiricists — from Oster to Krueger to Levitt himself — do exciting work which would exist fruitfully even in the absence of the leading formal research of the last thirty years. Theorists often solve problems that only they care about and browbeat economists who disagree. 
     
    Insiders will tell you that it is this split — not ideology — which tends to cause the greatest dissent within departments at all ranks.

  30. crabby, your comment is hopelessly outdated. You seem to admit this yourself when you mention the critics of “mathematicization” of economics, most being at least a generation or more ago. 
     
    Today, the most intense mathematically based theorists in economics are associated with behavioral economics. This is also, as should be expected, a very empirical branch of the science. 
     
    Its a fight, in every discipline, to get the theorists to be more relevant AND for the empiricists to care more about the theory. My experience in economics, and admittedly its been short, has been much more with the latter struggle than the former. Freekonomics (nee public/labor economics) is famously a theory-free zone and what theories they have are ad-hoc and not taken very seriously. For example, Levitt recently wrote on his blog that his theory about the connection between abortion and crime doesn’t hold anymore. Apparently, it only held in the time period he was studying and in the data set he examined. God, let’s hope the man wins a Nobel someday.

  31. To echo John Emerson: 
     
    j mct and eoin seem to think that philosophy ended with Bertrand Russell. Modern analytic philosophers are highly aware of the problems with natural language that you’ve raised. There’s a lot of work being done on them that you’ve clearly not read about in the popular philosophy books you mention. Criticizing analytic philosophy on the basis of its presentation in popular philosophy is like criticizing string theory after having read only The Elegant Universe. To start you off on how philosophers approach contextualism, here’s a summary from the Stanford Encyclopedia of Philosophy.

  32. pushmedia 1 
    I think the world is changing, but less than you think. I still see the struggles inside departments. Furthermore, macro is still dominated by theorists producing weird models that don’t go anywhere. So I don’t see my critique as hopelessly outdated. Furthermore, the best work using instrumental variables and the like generally does not require formal theory. Put it another way, you could mathematize almost any empirical work today using techniques developed mostly before 1975. That means most of the formal work published between 1975 and 2000 has been mostly useless. 
     
    Secondly, it is still the case that young professors in top departments get publications by formalizing an insight developed verbally by great minds of an earlier generation. Usually without improving testability or creating new predictions. 
     
    My experience, unlike yours, has been long and while I see some hope, it’s too early to tell if this is a “false dawn.” I’ll know it’s not when pointless theory isn’t rewarded with publication and tenure, especially at the strong mid-level departments.

  33. In any case the English language, of itself ( or to more exact any human language) cannot prove anything since it is a highly informal system.  
     
    Of course you can prove things in informal language. In fact, the language of choice for mathematicians in their published proofs is usually English. Take a look at any math journal where proofs are published. Prose is used almost throughout. Here’s a real simple proof (lucas’ theorem) Logical and mathematical symbols are sprinkled in between but the main work is done by natural language. There’s a good reason for this. If you were to try and prove things in a purely formal proof system (for example, natural deduction or a tableaux system) you’d have a proof that’s literally, thousands of pages long even for very simple proof.

  34. This passage from wiki is correct. Proofs published in math jorunal are almost always informal (done using natural language). [emphaisis mine] 
     
    The informal proofs of everyday mathematical practice are unlike the formal proofs of proof theory. They are rather like high-level sketches that would allow an expert to reconstruct a formal proof at least in principle, given enough time and patience. For most mathematicians, writing a fully formal proof would have all the drawbacks of programming in machine code. 
     
    Formal proofs are constructed with the help of computers in interactive theorem proving. Significantly, these proofs can be checked automatically, also by computer. (Checking formal proofs is usually trivial, whereas finding proofs (automated theorem proving) is typically quite hard.) An informal proof in the mathematics literature, by contrast, requires weeks of peer review to be checked, and may still contain errors.

  35. Hey, I like the fact that words change colors according to context, but then I’m an arts-and-cutlure kinda guy. 
     
    I suspect that the bunch of you who like thinking about these things would enjoy Stephen Toulmin. He trained as a physicist, turned to philosophy, and spent most of his career asserting the glories of informal reasoning, which he didn’t see as a contemptible stew of superstitions and common sense but as a to-be-respected way of thinking in its own right. (Incidentally I remember that John Emerson is a fan of Toulmin’s too.) I wrote an intro to his work here
     
    Michael Oakeshott, another fave of mine, is also someone to be savored where these questions go. Start with “Rationalism in Politics.”

  36. I thought all of Toulmin’s degrees were in philosophy?  
     
    Anyway, philosophers have delt with the contextual nuances of language. There’s a mountain of literature on this subject. Check out the seminal works by Richard Montague, David Kaplan, Robert Stalnaker and David Lewis to name just a few. They have worked to develope very rich contextual formal semantics for everyday speech. A good introductory to this body of work is Lewis’ paper, “Scorekeeping In a Language Game” published in the Journal of Philosophical Logic, vol. 8, Jan. 1979.

  37. I think a lot of valid points have been raised. 
     
    But I also find it terribly ironic that the parent story talks about how economists willfully ignore relevant neighboring fields before they theorize– and here we are, a bunch of (apparently) willfully ignorant non-philosophers, theorizing about how dumb Philosophy is. 
     
    On a more serious note,  
    1. There’s cruft in any discipline. 
    2. Philosophy serves several roles in the ecology of academia. 
    3. Philosophy is not as fact-driven as it should be. But it’s gotten a lot better in recent times.

  38. I don’t mean to come down too harshly on eoin. A lot of Philosophy is cruft. But I also feel Philosophy is often judged by unfair standards: e.g., this thread’s criticism of Marx, or first-generation logical positivism. The discipline moves on; I suppose an analogue would be if I criticized Biology on the grounds that the theory of Vitalism doesn’t make sense.

  39. Mike, I can take it. Come down harshly. My point is this: 
     
    I have a degree in mathematics and work in IT. I have written lexical parsers for computer languages. This is basically syntax, or grammar checking. Upgrading these parser for natural languages is a not impossible task. Clearly there are grammar editors in the field, most of limited success. Though I have not written any. 
     
    What I need – as an applied scientist – from you philosophers is a set of algorithms for judging the veracity, or not, or a sentence. If there is a more formal way of expressing language to make philosophical texts more easily parsed for truth then I should know that too. I dont see anything in a google search, but I may be looking in the wrong areas. Show me the beef. Then I’ll code it up. 
     
    By the way I dont take the general argument that I have not read enough foo as valid criticism – I mean I haven’t read all astrogical texts either, am I expected to? A defender of astrology once told me that I was being naive in thinking that a certain type of astrology represented the real “thing”, Dawkins faces similar criticisms as he hasn’t read all theological texts.

  40. Sure, natural language is actually a pretty good example. 
     
    First, though it may be obscured now, the whole field of NLP was once, and was created by, Philosophy. Depending on how far you want to trace things back, you can trace it to Chomsky, or further back to a cross-fertilization between Philosophy and Linguistics (itself an outgrowth of logical positivism). 
     
    You can find this in many other fields– Mathematics, for one, with people who considered themselves philosophers- Russell and Whitehead- systematizing modern mathematics. Which is to say, one role Philosophy has traditionally played in academia is to generate new disciplines. If you look at Philosophy and all you see is people either, 1. trying to talk very clearly about things that seem pointlessly simple, and/or 2. talking nonsense, my point is that this is part of the process of making new disciplines– shaking down the ontology enough such that one *can* start to pose meaningful questions. Then the discipline gets a real name, and starts to do useful things. The side-effect of this is that content-wise, Philosophy is pretty rubbish. But process-wise, it’s invaluable, and specifically has formed the foundation for your work in NLP. 
     
    See e.g., this interview with Dennett where he says as much: http://www.theatlantic.com/unbound/digicult/dennett.htm 
     
    And this interview with Searle (though I hate to link to him as his Chinese Room argument stinks so much). 
    http://globetrotter.berkeley.edu/people/Searle/searle-con2.html 
     
    That’s a big role that Philosophy has played in the past. Maybe this process has changed, now that academia has gotten larger and, in general, hungrier. 
     
    (Another metaphor I’ve heard for philosophy is that of an intellectual plumber. Philosophy is a pretty diverse field.)

  41. eoin said: 
    What I need – as an applied scientist – from you philosophers is a set of algorithms for judging the veracity, or not, or a sentence. If there is a more formal way of expressing language to make philosophical texts more easily parsed for truth then I should know that too. I dont see anything in a google search, but I may be looking in the wrong areas. Show me the beef. Then I’ll code it up. 
     
    I don’t know exactly what you mean by “algorithm” for judging “veracity.” Even in math we don’t have that and we never will. (Consequence of Godel’s incompleteness theorem which is an inherent nature of all non-trivial formalizable axiomatic systems)  
     
    In mathematics, we can construct set-theoretic models and use precise language in theory if we had the time, effort and patience to do so. But as I’ve shown, in practice, that is nearly impossible. No self-respecting mathematician or philosopher-logician will even attempt to publish such a proof because it is in practice impossible. Besides, that type of formal drudgery is really meant for computer programers (no offense). There’s a lot more art in math than most people realize.  
     
    In theory we can also express much of philosophy in set-theoretic terms and have a precise language and functions that map our language to our model and have a formal truth theory. See the work of Tarski for example on truth. 
     
    However, another inherent limitation of that would be that this type of formalism would be in practice also impossible to construct because it would be so complicated (much like the usual proofs in mathematics that are published). The more precise your language is (less ambiguous for example) the longer your proof will be due to its syntactical complexity. You can’t have a proof that is both precise and relatively short. That’s like trying to have your proof-theoretic cake and eating it too. If you could in practice construct a non-trivial mathematical proof using strictly formal language and proof structure, most mathematicians would probably do it for it then becomes machine checkable. However, in real life, we cannot escape the peer review process with all its “art,” amibiguity, human error, etc. A famous example of this is Wiles’ proof of Fermat’s last theorem. It took months for other mathematicians to discover an error. Wiles had to prove an independent theorem (modularity theorem) and plug it back into his main proof before resubmitting it.  
     
    Interestingly, there is a branch of recursion theory (AKA, computability logic) that works on machine proof, trying to get automated machine proving algorithms. Some philosophers actually work in this field. Branden Fitelson, Vincent F. Hendricks, have published in this area among others.

  42. But then again, what do I know, I am just a philosophy major.

a