What if Wittgenstein had been a macroeconomist?

The commenter Jason sent me a great Wittgenstein quotation, and I immediately knew I had to use it somewhere.  It took me 10 seconds to decide where:

“Tell me,” the great twentieth-century philosopher Ludwig Wittgenstein once asked a friend, “why do people always say it was natural for man to assume that the sun went around the Earth rather than that the Earth was rotating?” His friend replied, “Well, obviously because it just looks as though the Sun is going around the Earth.” Wittgenstein responded, “Well, what would it have looked like if it had looked as though the Earth was rotating?”

It’s quotations like this that make life worth living.  So I wondered what Wittgenstein would have thought of the current crisis:

Wittgenstein:  Tell me, why do people always say it’s natural to assume the Great Recession was caused by the financial crisis of 2008?

Friend:  Well, obviously because it looks as though the Great Recession was caused by the financial crisis of 2008.

Wittgenstein:  Well, what would it have looked like if it had been caused by Fed policy errors, which allowed nominal GDP to fall at the sharpest rate since 1938, especially during a time when banks were already stressed by the subprime fiasco, and when the resources for repaying nominal debts come from nominal income?

OK, not nearly as elegant as Wittgenstein’s example.  But you get the point. 

Jason also wonders what future generations will think of the Keynesian/monetarist split.  Which model will seem like the Ptolemaic system?  I won’t answer that, but will take a stab at a related question.  The Great Depression was originally thought to be due to the inherent instability of capitalism.  Later Friedman and Schwartz blamed it on a big drop in M2.  Their view is now more popular, because it has more appealing policy implications.  It’s a lot easier to prevent M2 from falling, than to repair the inherent instability of capitalism.  Where there are simple policy implications, a failure to do those policies eventually becomes seen as the “cause” of the problem, even if at a deeper philosophical level “cause” is one of those slippery terms that can never be pinned down.

In 50 years (when we are targeting NGDP futures contracts) the Great Recession will be seen as being caused by the Fed’s failure to prevent NGDP from falling.  Not through futures contracts (which didn’t exist then) but through a failure to engage in the sort of “level targeting” that Bernanke recommended the Japanese try during their similar travails.

PS.  W. Peden thinks the quotation is apocryphal, and notes that it’s used in Tom Stoppard’s play “Jumper.”  For some reason I prefer it be Wittgenstein.


Tags:

 
 
 

41 Responses to “What if Wittgenstein had been a macroeconomist?”

  1. Gravatar of W. Peden W. Peden
    16. December 2011 at 13:30

    On the slipperiness of ’cause’, I’ve been reading James Woodward’s work on the concept recently and it seems to apply well to the social sciences in particular: roughly, X causes Y if there is an intervention on the value of X such that it changes the value of Y.

    (Note that is therefore possible for X to cause Y and Y to cause X, since the fact that an intervention on Y changes the value of X doesn’t change the fact that an intervention on X will change the value of Y.)

    An “intervention”, for purely causal inquiries (“What caused the Great Depression?”) just means a hypothetical change. It doesn’t have to be technologically or practicably feasible for human beings to do the intervention; it could be a natural intervention, like a meteor hitting the Earth and causing the extinction of the dinosaurs.

    Interesting causal relations are those that we can use to explain phenomena. Explanatory causal relations are those that are unusual and/or which can be changed by us in order to cause/prevent an outcome.

    So you are right that Friedman’s M2 explanation is interesting because of its practical implications. That’s why arguments that the Fed couldn’t change M2 for institutional reasons (say some law limiting monetary base expansion) are interesting, EVEN IF there is still a sense in which the fall in M2 caused the Great Depression, because we can imagine a hypothetical intervention by the Fed (or God or whatever) whereby M2 fell and caused the Great Depression.

    With NGDP, one can say that bad monetary policy caused the Great Recession because the Fed could have intervened but didn’t. Under a true hypothetical scenario where the Fed intervened, there would have been no Great Recession (in the US and much of the world, though some countries like Ireland would still have had nasty contractions in 2009-2011).

    Woodward’s account is interesting because it links up the philosophy of the concept of causation rather nicely with developmental/evolutionary psychology explanations of why we have a concept of causality at all and why it develops at such an early stage in a child’s maturation: our concept of causality develops out of our deliberate manipulations of the world, which are obviously a very important to our survival!

  2. Gravatar of Vincent Vincent
    16. December 2011 at 13:36

    great quote.

    scott, assuming ngdp targeting, what would have happened in your opinion? no recession at all, or just a much milder one? what about all those failing banks, people who saw their home equity plumet and debt rise, etc?

    also, if keynesianism is the ptoleamic system, what would be the austrian school? greek mythology?

  3. Gravatar of Donald A. Coffin Donald A. Coffin
    16. December 2011 at 13:56

    Yeah, but what if the Gread Depression really *was caused* by the inherent instability of capitalism, and that the inherent instability of capitalism *also caused* the decline in M2? Why should we assume that the “correct” explanation is the explanation with the easiest-to-implement policy implications? Occam’s Razor, and all that, but aren’t we really interested in the correct explanation (for whatever meaning of “correct” floats your boat, I guess), not the explanation with the easiest-to-implement policy implications?

  4. Gravatar of Benjamin Daniels Benjamin Daniels
    16. December 2011 at 14:00

    Scott,

    You say, “the Great Recession will be seen as being caused by the Fed’s failure to prevent NGDP from falling.” But this is tautological if you take the definition of recession as falling NGDP. You basically say “the recession is caused by the Fed’s failure to prevent it” which is meaningless.

    You have to suggest a mechanism that caused NGDP to start falling in the first place. “Falling NGDP” is not that mechanism. The financial crisis is, likely as a transmission/amplification mechanism that transferred the nominal-debt pressure from the subprime/CDS crisis to the real economy.

  5. Gravatar of gwern gwern
    16. December 2011 at 14:00

    It seems highly unlikely that this is apocryphal, contra Peden, and the real quote is, if anything, better because it includes the ending:

    “The general method that Wittgenstein does suggest is that of ‘shewing that a man has supplied no meaning for certain signs in his sentences’.

    I can illustrate the method from Wittgenstein’s later way of discussing problems. He once greeted me with the question: ‘Why do people say that it was natural to think that the sun went round the earth rather than that the earth turned on its axis? I replied: ‘I suppose, because it looked as if the sun went round the earth.’ ‘Well,’ he asked, ‘what would it have looked like if it had looked as if the earth turned on its axis?’

    This question brought it out that I had hitherto given no relevant meaning to ‘it looks as if’ in ‘it looks as if the sun goes round the earth’.

    My reply was to hold out my hands with the palms upward, and raise them from my knees in a circular sweep, at the same time leaning backwards and assuming a dizzy expression. ‘Exactly!’ he said.”

    –Elizabeth Anscombe, An Introduction To Wittgenstein’s Tractatus (1959) http://www.archive.org/details/introductiontowi009827mbp

  6. Gravatar of ssumner ssumner
    16. December 2011 at 14:13

    W. Peden, Nice post. Just one quibble. I wouldn’t say “the Fed could have intervened.” They intervene every second of every day. They control monetary policy and hence NGDP.

    As an analogy, I wouldn’t say “Microsoft could have intervened and prevented the price of Windows from rising,” I’d say Microsoft raised the price of Windows.

    Vincent, Maybe a mild recession, maybe just a long period of stagflation (4% inflation and 1% RGDP growth.) Depends on other factors like UI benefits, etc.

    Donald, There is no “really” it’s just a question of which explanations we find useful.

    Benjamin, Yeah it would be tautological if that was my definition of recessions. Fortunately that’s not my definition. Otherwise Zimbabwe would have the best economy in the world, their NGDP rose by trillions of percent.

    NGDP started falling in late 1929 because the US, France, and the UK all started running tight money policies at the same time. Very tight.

    The crisis of 2008 was caused by tight money at the Fed. They deviated from their normal dual mandate in September 2008, and it was all downhill from there.

  7. Gravatar of ssumner ssumner
    16. December 2011 at 14:14

    Thanks Gwern.

  8. Gravatar of Luis H Arroyo Luis H Arroyo
    16. December 2011 at 15:26

    well, as you know, I don´t participate of your vision that the “financial crisis” didn´t matter. Simply, I see that assets prices change at much more velocity than goods prices, and the influence goes from assets markets toward goods markets, and not the opposite.
    For me that doesn´t exclude the necessity of a monetary reaction. I agree in that Bernanke was too late in QE.
    But I cannot understand the negation of the evidence: the accute fall of assets prices could not be caused by the lower fall in NGDP. It was much more important the systemic fall of assets markets.
    I think that markets monetarism would be much complete if it integrate assets markets in the model.
    In other words I think that without “financial box”, the real world fluctuations would be less accute.

  9. Gravatar of c8to c8to
    16. December 2011 at 15:47

    the quote is idiotic like most of philosophy. the natural assumption is natural because we don’t feel the rotation. (but normally rotations being accelerations are felt) its not just the look: its that we feel a stable 1g environment AND see the sun move relative to the earth.

    your quote is much nicer.

  10. Gravatar of Benjamin Cole Benjamin Cole
    16. December 2011 at 16:04

    Love this quote.

    BTW, I am the guest blogger at Lars Christensen today. Please go see!

  11. Gravatar of Gene Callahan Gene Callahan
    16. December 2011 at 16:17

    c8to, if I was too dull to grasp some subject, I have often supposed a nice way of maintaining my self-image would be to deem it idiotic.

    Does it work?

  12. Gravatar of flow5 flow5
    16. December 2011 at 16:56

    “They deviated from their normal dual mandate in September 2008″

    That’s not exactly what happened. Bernanke tightened MVt for 29 consecutive months. The Case Schiller housing index peaked @189.93 when Bernanke initiated his tight money policy. Based upon the FED’s technical criteria (interest rates), the 4th quarter contraction in 2008 was already “set in stone” beginning in Jan of that year. I.e., the Fed’s failure to prevent NGDP from falling started long before the economy collapsed.

  13. Gravatar of Morgan Warstler Morgan Warstler
    16. December 2011 at 19:02

    Matty comes one step clser to adopting my Guaranteed Income:

    http://www.slate.com/blogs/moneybox/2011/12/16/marginal_tax_rates_and_the_poor.html

    Bruce Goldberg put together Wittgenstein and Libertarian theory for me.

    I don’t think Wittgenstein would think you make good choices of words.

  14. Gravatar of W. Peden W. Peden
    16. December 2011 at 19:28

    Gwern,

    Thanks for the tip! I had no idea that G. E. M. Anscombe was the friend in question (I had always imagined it as Frege, simply because he is usually the person to which Wittgenstein is reacting in one way or another in most of his work).

    Scott,

    Yes, you’re right: the Fed is never passive and so the right phrase would be “if the Fed had intervened DIFFERENTLY”.

    c8to,

    Perhaps, but that doesn’t get round the “looks like” issue. As far as I understand astrophysics (i.e. not very far) the Sun is itself moving relative to celestial objects outside the Solar System, so even if we perceived the Earth as in motion and the Sun as stationary then it would be possible to reform the dialogue on the basis that we didn’t perceive the Sun’s constant motion.

  15. Gravatar of Why this time was a little bit different? | Historinhas Why this time was a little bit different? | Historinhas
    16. December 2011 at 19:58

    [...] this see Scott Sumner´s post on “Wittgenstein as a macroeconomist”: Wittgenstein:  Tell me, why do people always say it’s natural to assume the Great Recession [...]

  16. Gravatar of Steve Steve
    16. December 2011 at 20:24

    Scott,

    I agree that the Fed is never truly passive, and they tightened sharply in the summer of 2008 by doing nothing.

    Still, I’m reminded that even by one traditional measure, the size of the balance sheet, the Fed was also exceptionally tight in 2008. They had been growing the balance sheet about 1% per annum from 2006 to 2008 when AIG collapsed. All the inflationistas who complain about the balance sheet today should keep that in mind. And people who say the Fed was easing because they cut interest rates a couple of times should keep in mind the slowest balance sheet growth in generations leading up to the crisis. The pre-crisis balance sheet is proof of their hard-money intent.

  17. Gravatar of Liberal Roman Liberal Roman
    17. December 2011 at 00:03

    I hope you are right about that 50 years prediction. Because we had 70 years to look at, analyze and figure out what went wrong in the Great Depression. We even basically figured it out (or at least Friedman did). But we still ignored everything when the moment came to put what we learned into practice because it was “unorthodox”.

  18. Gravatar of David Pearson David Pearson
    17. December 2011 at 08:01

    “what would it have looked like if it had been caused by Fed policy errors, which allowed nominal GDP to fall at the sharpest rate since 1938?” I suggest we would have seen:

    -the largest number of AAA-rated bond downgrades in history AFTER that crash in nominal NGDP, not before.
    -the first global, generalized runs on financial institution liabilities after that crash in NGDP, not before.
    -the first major bank fail, as did the the Bank of the United States, after…
    -the failure of our two largest financial institutions (Fannie and Freddie) after…
    -the near-failure of multiple European financial institutions after…

    “Subprime fiasco” is a misnomer. By July of 2008, the crisis had spread from Subprime mezzanine tranches to AAA-rated subprime, to Alt-A and jumbo, to asset-backed commercial paper, to commercial real estate, to corporate junk, and, finally, to Agency-guaranteed AAA-rated RMBS. Concerns over much of the collateral underlying bank repo’s led to a systemic crisis that began in 2007 and culminated in the Lehman failure.

    Saying banks were, “stressed,” by this is like saying a traffic victim felt feeling anxious just after being run over by a truck.

  19. Gravatar of ssumner ssumner
    17. December 2011 at 08:07

    Luis, You said;

    “Simply, I see that assets prices change at much more velocity than goods prices, and the influence goes from assets markets toward goods markets, and not the opposite.”

    We are closer than you think. I see the connection as follows:

    1. Future expected monetary policy determines future expected NGDP.

    2. Future expected NGDP determines future expected asset prices.

    3. Future expected asset prices determines current asset prices.

    4. Both future expected NGDP and current asset prices determine current NGDP, current AD.

    Point 4 is similar to your argument.

    And the biggest fall in NGDP since the Great Depression most certainly could cause a large fall in asset prices.

    c8to, You are much too harsh. It is an aesthetically beautiful quotation. Yes, it’s not logical (gravity wasn’t discovered until after the Copernican revolution, and thus rotation would have seemed to produce high winds before Newton.) But it’s very pleasing.

    Ben, That’s great news, I’ll take a look right after these comments.

    flow5, I think mild tightening was appropriate in 2006, but not 2008.

    Morgan, You said;

    “I don’t think Wittgenstein would think you make good choices of words.”

    Yeah, he’d look down on both of us.

    Steve, Yes, and there was a distinct slowdown in late 2007 and early 2008, if my memory is correct.

    Liberal Roman, Yes, perhaps I’m being optimistic in assuming we’ve absorbed Friedman’s insights.

  20. Gravatar of ssumner ssumner
    17. December 2011 at 08:27

    Ben, Great post! I’ll link to it later.

    David, I don’t agree with your timeline. The crash in NGDP occurred between June and December 2008. The failures of Fannie, freddie, AIG, Lehman, and the subsequent banking crisis occurred DURING that collapse, not before.

    No one disputes that the financial system was stressed in mid-2008, I said as much. But most economists were still predicting steady growth in 2009 because they knew those problems were not the sort of thing that would cause a severe recession. And they were right. When the severe recession occurred these same economists changed their minds and decided it was in fact the sort of thing that could cause a severe recession. After all, what else could it have been? After all money was easy, wasn’t it?

    BTW, My views are shared by the IMF, whose estimates of total losses to US banking from the crisis nearly tripled between April 2008 and April 2009.

  21. Gravatar of David Pearson David Pearson
    17. December 2011 at 09:05

    Scott,
    As I said, “stressed” does not describe the banking system from August. 2007 through July of 2008. “Systemic banking crisis” does. I think you focus to much on discrete announcements. The GSE’s were in trouble long before they failed in August, 2008. Hank Paulson warned financial executives in a meeting in July that they might have to be taken over. If Hank Paulson knew this in July, believe me, he was not the first person to figure it out: Fannie’s stock (FNM) went from $70 in 2007 to $10 by June, 2008. MBIA, the bond insurer, also went from $70 to $10 by the beginning of 2008. The broader S&P Financial Sector index (XLF) had fallen from $38 to $18 by June, 2008. It actually ROSE from July to September 15, during the time you argue NGDP expectations were “crashing”. I think, “experiencing a brief bounce after a crash,” might be more applicable to financial sector expectations during that time.

  22. Gravatar of Luis H Arroyo Luis H Arroyo
    17. December 2011 at 11:01

    David, The problem is that cannot understand the order of facts that you propose. But perhaps the differences are not so high.
    I see, first, the increase in demand for money because the violent change of expectation in financial assets. that began in 2007, August.
    As I see in the graph of saint Louis Fed,
    http://research.stlouisfed.org/fredgraph.png?g=3Uj
    NGDP was growing until well advanced 2008, whereas Stock Market began to fall quite in advance.
    So I suppose that the change in financial expectation was firstly to change in NGDP expectation.
    I think that that is monetarism also, but for me Bernanke was much more later than for you…

  23. Gravatar of Luis H Arroyo Luis H Arroyo
    17. December 2011 at 11:58

    Sorry, I meant Scott…

  24. Gravatar of Integral Integral
    17. December 2011 at 12:18

    Scott,

    “NGDP started falling in late 1929 because the US, France, and the UK all started running tight money policies at the same time. Very tight.

    The crisis of 2008 was caused by tight money at the Fed. They deviated from their normal dual mandate in September 2008, and it was all downhill from there.”

    I’d like to dig at this a little more, because to be honest your transmission mechanism for the key June-December 2008 period has never been altogether clear to me.

    Start with the Depression. You could argue that the Depression was caused by Fed errors of commission: by hoarding gold and sterilizing, the quantity of money fell, NGDP fell, and we experienced a Depression.

    Now turn to 2008. Here we have a case of errors of *omission*. BmV = PQ where B is base money, m the money multiplier, and V the velocity multiplier. The Fed only controls B perfectly. B rose quite a bit during mid- and late-2008, but “not enough to offset the shocks to V and m”, is the story I see you telling. Correct me if I’m wrong on this point!

    The question that Benjamin is asking, and one that I wonder, is “what were the underlying shocks to m and V that the Fed didn’t cushion sufficiently well against”?

    I see a two big shocks:
    1) Oil prices rose sharply in the first half of 2008. Call that a “supply shock”.
    2) Expectations of future income growth plunged suddenly in mid-2008 (source: Michigan Survey of Consumers). Call that a “demand shock” to future income. (which still leaves the question: why did expectations hold up through mid-2008 but collapse suddenly, pre-Lehman and pre-financial crisis?)

    What’s your “story”, if you will, for the underlying shocks that caused AD to plunge? Or is it simply that there were no “underlying factors”?

    You may already have a post on this; if so, merely point me to it.

    And if any point I’m being incoherent, please let me know and I’ll try to rephrase.

    :)

  25. Gravatar of What Was That? Sumner and the Great Depression | Economic Thought What Was That? Sumner and the Great Depression | Economic Thought
    17. December 2011 at 12:28

    [...] Scott Sumner employs a little bit of Wittgensteinian logic to show us why the monetarist (Friedman’s and & Schwartz’, that is) explanation of the Great Depression is more accurate than the Keynesian one (Sumner’s summation: “inherent instability of capitalism”).  Writes Sumner, The Great Depression was originally thought to be due to the inherent instability of capitalism.  Later Friedman and Schwartz blamed it on a big drop in M2.  Their view is now more popular, because it has more appealing policy implications.  It’s a lot easier to prevent M2 from falling, than to repair the inherent instability of capitalism.  Where there are simple policy implications, a failure to do those policies eventually becomes seen as the “cause” of the problem, even if at a deeper philosophical level “cause” is one of those slippery terms that can never be pinned down. [...]

  26. Gravatar of AWK AWK
    17. December 2011 at 21:31

    After reading many of these blogs I think this Witty quote is far more apropos:
    “Whereof one cannot speak, thereof one must be silent.”

    Much less waste of server space… Or think of it as data compression.

  27. Gravatar of cato cato
    18. December 2011 at 02:59

    gene, what don’t you understand about philosophy? =)

    ss, intelligence and beauty are ofter uncorrelated.

    better: why are blue and green more “natural” than bleen and grue? (please reference kolmogorov in your answer and then i’ll show you why this is also mistaken =)

  28. Gravatar of cato cato
    18. December 2011 at 03:03

    gene, what don’t you understand about philosophy? =)

    ss, intelligence and beauty are ofter uncorrelated.

    better: why are blue and green more “natural” than bleen and grue? (please reference kolmogorov in your answer and then i’ll show you why this is also mistaken =)

    WP – yes.

  29. Gravatar of A tale of two monetary policies | Historinhas A tale of two monetary policies | Historinhas
    18. December 2011 at 05:39

    [...] a recent comment in Scott Sumner´s post, Luis Arroyo (responsible for the Spanish Blog Illusíon Monetaria) says: Simply, I see that assets [...]

  30. Gravatar of ssumner ssumner
    18. December 2011 at 08:10

    David , Maybe we are arguing about semantics. The IMF estimates of total US banking losses were nearly a trillion in April 2008. That’s a huge number, so I don’t disagree with you there. By April 2009 the estimated losses had risen to $2.7 trillion, a much bigger number, that’s my point. As of April 2008 the commercial RE market had not yet crashed. By April 2009 it had. Those are very important facts.

    Luis, There was no sudden change in the demand for base money in late 2007 and early 2008–base velocity was pretty stable.

    Integral, I mostly agree. In earlier posts I’ve discussed the fact that the Fed tends to target the nominal fed funds rate. During the key period of April to October 2008 they kept it constant. Thus in that sense policy was unchanged. So any tightening could be viewed as unintentional, due to a fall in the Wicksellian equilibrium rate. Why did that fall?

    1. The housing slump had already sharply reduced credit demand by early 2008, sharply reducing the Wicksellian equilibrium rate. The Fed handled that reasonably well.

    2. Then in the first half of 2008 oil prices skyrocketed, killing off the auto/truck/RV/boat market. Further drop in credit demand. Further drop in Wicksellian equilibrium rate. This time the Fed didn’t accommodate. Why not? Fear of headline inflation, which was high at the time (should have focused on NGDP.)

    3. Right after Lehman failed the Fed still refused to cut rates below 2%, citing fear of high inflation. Yet 5 year TIPS spreads had fallen to 1.23% by then, so they were steering the economy by looking in the rear view mirror.

    4. But the overarching error is growth rate targeting. They need to do level targeting. The markets realized this problem in the fall of 2008, and that’s why the asset markets crashed. They realized the Fed would allow NGDP to fall sharply, and make no attempt to get back up to the previous trend line. The Great Moderation was over–quite a shock!

    Why is level targeting so important? Because some short term fluctuations will occur before the Fed recognizes the problem–such as the GDP fall in 2008:3. With level targeting the markets expect a rebound, not further declines–this helps stabilize the asset markets, and the broader economy.

    So ultimately it’s a flawed procedure, as was true in 1929 (when the real bills doctrine still was influential.)

    AWK, That’s a good one. Fits this blog well. :)

    Cato, Philosophy is “useful” as literature, and for expanding the mind so one can think about problems from a different perspective.

  31. Gravatar of Edwin Herdman Edwin Herdman
    19. December 2011 at 17:30

    Having just spent a good amount of time with the Ptolemaic and Copernican systems, I have some general thoughts that might be of assistance. First, don’t forget Tycho Brahe’s compromise system (between Ptolemaic and Copernican systems; Brahe puts the Earth at the center of the universe but has the other planets, minus the moon, orbit the sun), or the early Pythagoreans and their moving, rotating Earth, or Kepler and his eliptical orbits (and a lot of patent nonsense, as well; such is the fate of the expansive mind who truly advances science, as Kepler was, as much as the parsimonious type, as Galileo was.

    Consider the Pythagoreans. In the second book of his work On The Heavens,* Aristotle criticizes the Pythagorean view that the Earth rotates, as the Pythagoreans had to “balance” the system with a counter-Earth. Being 180 degrees across from the Earth in its orbit, on the other side of the sun, counter-Earth would never be seen from Earth. The necessity of its existence, in that model, starts to get at the realization that some kind of force had to account for planetary motion.

    Here, though, the Pythagoreans had two problems: One, they were limited by the lack of scientific instruments and data to shore up their theory that the Earth moves, and so seeming absurdities fill the gap. This is the second problem: Aristotle criticizes the Pythagoreans for “forcing the phenomena” to fit the theory, rather than altering a theory to fit the evidence.

    From this we can see where the commentator “cato” has erred.

    For there are useful measures of “beauty” – parsimony and other “theoretical virtues” – which are a basis like that Aristotle uses to argue against the Pythagorean view of a turning Earth, and which are the basis Johannes Kepler evidently uses (in the introduction to his New Astronomy from 1609) to dismiss his late, former employer (and sometime antagonist) Tycho Brahe’s compromise system, which provides phenomena that appear identical to the human observer as they would under the Copernican system, but which uselessly proliferates coincidences to fit the Aristotelian system of gravity: “[...] motions are still multiplied to no purpose by Brahe, as they were before by Ptolemy.” i.e., theories requiring a certain order have been removed, but “he nevertheless left in the world the effects produced by those theories.”.

    We now take leave of the ancients, and consider a modern puzzle: Is the history of science one of “revolutions,” as depicted by the Copenhagen School-era quantum scientist Thomas Kuhn’s On the Structure of Scientific Revolutions, where all you have to do is be in the right place at the right time to make an observation? Or do the “theoretical virtues” that seem to crop up again, and again, guiding the best scientific minds to new theories and to dismiss exploration of avenues lined with trees that bear no scientific fruit? Even Kepler – who investigated odd musical theories and found that you could closely approximate the positions of planetary orbits by fitting the Platonic solids one inside the other in a certain order, and of whom was said that he would’ve been more highly regarded if he had burned two-thirds of his output – was guided by theoretical virtues; for him, the evidence of a mechanism was of chief importance; an empty point in space (i.e. an epicycle or its deferent) was of no use in describing the orbit of a physical object.

    Economic science seems fundamentally different, however. While the Earth continues to turn according to a more or less independent reality, it is obvious that there is not only an underlying economic theory, which has some roots in science, but the tendencies of people as well as the way in which they understand their situation do complicate any attempt to find a single over-arching theory. I think economic science has tools to root out theories more than it has theories that suggest the shape of tools.

    Who was it I read recently saying that it takes a more discerning mind to understand economics than to understand physics? This is manifestly true, considering the layers of abstraction economics must necessarily be built on starting from science (and in this case, a good deal of statistics or quantum theory, not the familiar predictability and regularity of Newtonian mechanics, which approximate a special case), it is clear that economics must take up “[theories] multiplied to to no purpose,” including the means by which every person resolves seeming contradictions between the various theories and guiding principles they hold to be true or useful.

    The careful reader will observe that I note two problems, as elucidated by Aristotle, with the Pythagorean view. The first problem was simply that they did not have (and Aristotle did not have) enough data. Today, so much of the economic debate is driven not by a lack of data, but by theorists who seem loath to admit that data exists that readily disproves their theories. To be kind, this is still a data problem, and it may be the case also that there are hard limits on certainty (a favorite topic of von Mises, I am told). Even if knowledge remains an unapproachable object, within the sphere of human affairs we still need working theories and we still need to respect the data.

    * Ref. note: my quotations are from “Philosophy of Science: An Historical Anthology; McGrew, Alspector-Kelly, Allhoff, eds.; produced locally at Western Michigan University and an indispensable anthology to the non-specialist or undergraduate of chief papers in the history of science from its earliest times to the modern 20th century; I recommend it heartily to anyone who wants some good historical sources immediately at hand.

  32. Gravatar of Edwin Herdman Edwin Herdman
    19. December 2011 at 17:36

    Also, to answer Wittgenstein: It would have looked as if we were on the Moon (where it does, indeed, appear as if the Earth was runing)

    …Alternatively, it could appear as if the Earth moved under our feet when we jumped. The problem of frames of reference was a chief problem understood well by many Medieval philosophers, and Galileo discusses its consequence, “combined motions,” pretty thoroughly in his Dialogue Concerning the Two Chief World Systems. A favorite argument of those who believe in a flat or motionless Earth remains “birds should be struggling against a 1,000 mph wind! (at the equator)”

  33. Gravatar of ssumner ssumner
    20. December 2011 at 10:09

    Edwin, You said;

    “Who was it I read recently saying that it takes a more discerning mind to understand economics than to understand physics? This is manifestly true, considering the layers of abstraction economics must necessarily be built on starting from science (and in this case, a good deal of statistics or quantum theory, not the familiar predictability and regularity of Newtonian mechanics, which approximate a special case), it is clear that economics must take up “[theories] multiplied to to no purpose,” including the means by which every person resolves seeming contradictions between the various theories and guiding principles they hold to be true or useful.”

    I agree that economics requires a more discerning mind, and would add that I think physics requires a higher IQ.

    I take a pragmatic perspective on economic models. All that really matters is whether they are useful or not. No model will exactly describe reality.

    Regarding you second comment, I made a similar point about wind above. But I don’t know about jumping. Surely the ancient Romans who rode in chariots realized that if you jumped in place on a fast moving chariot, you landed in roughly the same spot.

  34. Gravatar of Edwin Herdman Edwin Herdman
    21. December 2011 at 18:45

    I agree wholeheartedly that an imperfect model is much better than one which shrugs off the burden.

    The argument for jumping is as follows:

    According to Aristotelian physics (I apologize for the lengthy overview in advance, especially to all who are already familiar with this), the universe has no vacuums and objects are all oriented towards a “center of the universe” – and they spontaneously orient themselves to it, by sorting themselves out by contact with other bodies. Some bodies are “absolutely” heavy, and others have a property of “absolute levity.” These bodies are composed of the four elements. There is also the sphere of the stars and other celestial elements – these are composed of a “fifth element.”

    The four elements move either up or down, in contact with one another; only things made from a fifth element (type) can move in circles.

    Now, some new stuff: This system requires a more complicated view of other motions (including those Galileo calls “combined” motions, i.e. falling to the Earth at the same time as being moved along with the Earth). For Aristotle, this is explained by the mechanism of antiperistasis: An object moving somehow displaces other objects (i.e. a flying arrow displaces the atoms of air) and is in turn pushed along by atoms rushing in from the back (as Aristotle’s detractors, i.e. the early c.490-570 Christian philosopher John Philoponus,* point out, this seems to require the atoms of air impinge directly on the notch at the back of the arrow, which is absurd).

    It seems theoretically possible to somehow press antiperistasis into service to account for “shared motion” in a rotating Earth, and the phenomena of smoke moving off to the West (if I have that right) could suggest that the antiperistatic effect of mountains on great bodies of air was not in play at high altitudes. But medieval philosophers (like Jean Buridan and his student Nicole Oresme) expand on Philoponus by pointing out that a smooth blacksmith’s wheel can rotate in place for a long while without disturbing the air nearby, and so there is no apparent mechanism whereby antiperistasis keeps pushing the wheel around in its circular path.

    The Aristotelians (and their latter-day adherents, the peripatetics) seem to understand this intuitively in the way they argue against a turning Earth. Presuming (correctly) that the blockage of mountains and the rough surface of the Earth is not enough to move the air along with it, and facing the fact that relative to us objects on Earth seem more or less still, they intuit that antiperistasis does not allow for a turning Earth. Unfortunately for them, it is because antiperistasis is wrong that its effects are not evident. If it was right, when you left contact with the Earth, spinning under your feet, you would come down (much) farther to the West than you started out.

    You can see this thinking in relatively modern times in documents like this 1893 “Map of the Square and Stationary Earth” – the globe to the right with two men hanging desperately on illustrates the tendency to think as if any motion that seems large to us would somehow overcome the forces of gravity and other attractive forces (you would need to travel much, much faster than 1000 miles per hour, or be subject to very strong solar winds, to get to the point where unbound atoms, like atmospheres, would start to be loosened in travel; I can’t even imagine what circumstances of speed could possibly split apart quarks, which require more than TeV (trillion electronvolts) of energy, I believe.

    It is worth pointing out that Aristotle understood that the Earth was not flat.

    * ref. cited in a previous post :)

  35. Gravatar of Edwin Herdman Edwin Herdman
    21. December 2011 at 19:17

    I just realized this could be made simpler (assuming the reader understands the history from my previous post).

    Yes, the ancients noted that your facing and movement did not seem to alter the course of motion, but they inappropriately took this as positive evidence for a stationary (but spherical) Earth, rather than evidence that would be rendered almost inconsequential by relativistic effects. To quote Oresme, “local motion can be perceived only if we can see that one body assumes a different position relative to another body.” Since Newton, we realize that acceleration provides the differentiating factor.

    With regards to the chariot jumping: If you believe the Earth is not turning, then it seems easier to explain the fact that you can shoot a cannonball East (this is Galileo’s example from his Dialogue Concerning the Two Chief World Systems, as is his ship and tower example), and shoot another West, there will be hardly any difference in the distance of the two shots. The Coriolis Effect should make an appearance at some point, but not at a level detectable to any but the most modern observers (I think it really only appears when observing weather phenomena from the platform of a satellite in space – could be wrong, though the common wisdom about the coriolis effect determining the direction of a toilet flush is wrong).

    If you assume the Earth is turning, then the Peripatetic would expect a cannonball shot to the West to go less distance (measured against the ground) than a cannonball shot to the East, since the cannonball is no longer in contact with the Earth’s surface and antiperistasis is not going to make a dent in its motion. The speed of the cannon, imparted by the sphere turning under it, would impart more motion (globally) to the cannonball fired West than the one fired East, and the peripatetics assume that motion can be distinguished locally.

    Galileo demonstrates that these sorts of arguments, measured locally, don’t prove or disprove a turning or stationary Earth model, and for this he utilizes the example of the ship and the tower. If you drop a stone (perfectly round, perfectly balanced, if you like) from a tower, it goes grazing (almost, not quite) precisely downward along the tower.

    Now drop that same stone from the bridge right next to the top of the mast of a ship passing under the bridge: The stone will not fall to the base of the mast, but to a point further back on the ship.

    Here the Peripatetics (at least as portrayed by Galileo) err fatally: Since they believe that the distance the stone falls from the straight downward path is not altered by the motion of an object it was carried along with, peripatetics believe that, if somebody in the crow’s nest of the ship drops a stone next to the stone the person on the bridge (i.e. in a tower) drops, that they will both fall to the same place. Antiperistasis is not going to change their trajectories, since the wind and their motions are the same.

    We know that dropping the stone from the moving ship will actually cause it to land at the base of the mast, just as if it had been dropped from the stationary tower, even while the stone dropped from the bridge lands further back on the ship.

    The problem, as Galileo puts it, is that the peripatetic philosophers took as granted that previous philosophers had done the ship and tower experiment (or, at least, an experiment with a universal enough result).

    Obviously, they had not – the result is the classic appeal to authorities – in this case, an unbroken chain of authorities all mouthing the assumptions of their predecessors up to the point where the first person assumed something to be true, or extrapolated inappropriately from an experiment that corroborating evidence was not needed.

    It would be stretching to say that this now has any but a most vague application to the original topic. I will say, however, that the observation of effects locally vs. globally is important to consider, and is probably the answer to Wittgenstein’s suggestion – just get enough distance from the subject! I’m not sure how that would work out in the case of Fed. policy versus a financial crisis alone, however.

  36. Gravatar of Scott Sumner Scott Sumner
    22. December 2011 at 08:14

    Edwin, Thanks for that information.

  37. Gravatar of もしヴィトゲンシュタインがマクロ経済学者だったら by Scott Sumner – 道草 もしヴィトゲンシュタインがマクロ経済学者だったら by Scott Sumner – 道草
    23. December 2011 at 14:56

    [...] by Scott Sumner // サムナーのブログから “What if Wittgenstein had been a macroeconomist?“(16. December [...]

  38. Gravatar of gwern gwern
    31. December 2011 at 17:21

    Peden: Frege is actually kind of unlikely, as he died in 1925 and Wittgenstein only really made his mark in 1918 or so with the early discussions of the Tractatus. And Wittgenstein wasn’t too interested in philosophy until he went into England, so there’s a spatial separation too.

  39. Gravatar of Bandanbladderstiddle Bandanbladderstiddle
    22. January 2013 at 05:36

    The Wittgenstein-quote is not at all apocryphal, you can find the original in a book about meetings with Wittgenstein, the “friend” was G. E. M. Anscombe.

  40. Gravatar of Saturos Saturos
    22. January 2013 at 08:01

    Which book?

  41. Gravatar of Bandanbladderstiddle Bandanbladderstiddle
    23. January 2013 at 02:20

    Ludwig Wittgenstein: Personal Recollections (edited by Rush Rhees)

Leave a Reply