Gobrey, Smith, Hume

by Henry on October 15, 2014

I wanted to note this disagreement between P.E. Gobry and Noah Smith because it allows me to pull out my favorite underappreciated David Hume quote.

Gobry:

Science is the process through which we derive reliable predictive rules through controlled experimentation. That’s the science that gives us airplanes and flu vaccines and the Internet. But what almost everyone means when he or she says “science” is something different. … Since most people think math and lab coats equal science, people call economics a science, even though almost nothing in economics is actually derived from controlled experiments. Then people get angry at economists when they don’t predict impending financial crises, as if having tenure at a university endowed you with magical powers.

Smith:

One way of systematically understanding the world is just to watch it and write down what happens. “Today I saw this bird eat this fish.” “This year the harvest was destroyed by frost.” “The Mongols conquered the Sung Dynasty.” And so on. All you really need for this is the ability to write things down. This may sound like a weak, inadequate way of understanding the world, but actually it’s incredibly important and powerful, since it allows you to establish precedents. … A second way of systematically understanding the world is repeated observation. This is where you try to make a large number of observations that are in some way similar or the same, and then use statistics to identify relationships between them. … The first big limitation of empirics is omitted variable bias. You can never be sure you haven’t left out something important. The second is the fact that you’re always measuring correlation, but without a natural experiment, you can’t isolate causation. Still, correlation is an incredibly powerful and important thing to know. … Experiments are just like empirics, except you try to control the observational environment in order to eliminate omitted variables and isolate causality. You don’t always succeed, of course. And even when you do succeed, you may lose external validity – in other words, your experiment might find a causal mechanism that always works in the lab, but is just not that important in the real world.

Hume:

Mankind are so much the same, in all times and places, that history informs us of nothing new or strange in this particular. Its chief use is only to discover the constant and universal principles of human nature, by showing men in all varieties of circumstances and situations, and furnishing us with materials from which we may form our observations and become acquainted with the regular springs of human action and behaviour. These records of wars, intrigues, factions, and revolutions, are so many collections of experiments, by which the politician or moral philosopher fixes the principles of his science, in the same manner as the physician or natural philosopher becomes acquainted with the nature of plants, minerals, and other external objects, by the experiments which he forms concerning them.

{ 208 comments }

1

Anarcissie 10.15.14 at 3:18 am

Economics could be a science, were there not always a thumb or two upon the scale.

2

Bruce Wilder 10.15.14 at 3:20 am

It does seem odd that Gobry should feel no need to mention theoretical analysis, and Noah Smith gives it only a sidelong glance. I imagine Euler has something to do with airplane wings, and folks like John von Neumann or Claude Shannon something to do with the internet.

One of Hume’s claims to fame is his observation that one can never simply observe cause-and-effect in action.

Odd business.

3

Bruce Wilder 10.15.14 at 3:24 am

If the economists took their thumbs out of their own eyes, and put them on a scale, that would be a major advance. At least, they would be measuring something.

4

Chris Mealy 10.15.14 at 4:41 am

This is all backwards. The steam engine came before thermodynamics, airplanes came before aerodynamics, and vaccines came before the germ theory of disease. When economists (or more likely, non-economists) have enough useful techniques and technologies they’ll be getting somewhere. Until then they’d be better off just trying to describe, like astronomers and geologists mostly do, rather than pretending to explain.

5

ZM 10.15.14 at 5:09 am

I think Hume is making a great assumption that there are ‘constant and universal principles of human nature’. It would seem he made this great assumption about constancy and universality before he even read all the history and anthropology of the world – this is jumping to conclusions or beginning your journey at your intended destination.

You may or may not think this is a great folly, but I will tell you about a silly academic book I found the other day which I am sure everyone will agree is a folly. If this book is a folly then Hume is also folly-ing.

The book is the Algebra of Conscience – wherein the author uncovers the mathematical foundations of conscience through the translation of english sentences about conscience into algebra. Into the book the author decides he will translate Crime and Punishment and Hamlet from words into algebra — I can’t read all the algebra but it looks to me like he loses such a great deal in the process that even a grade schooler doing paraphrasing compositional exercises would not lose so much.

I recommend everyone to look at this book to see that in order to avoid such great folly one ought always to proceed with caution before jumping to conclusions about human nature being so very constant and universal that it is appropriate to study with the same sort of scientific method as one might use to study the movement of the heavenly bodies.

Of course it depends what you take to mean by science. Science itself just pretty much means knowledge except via latin roots to sound more authoritative than knowledge which comes from Old English and Germanic. You see people doing this all the time – the latinish word has the same content as a plain english word but the speaker feels more authoritative if they choose the latinish.

But then the loathsome torturer Francis Bacon made his Baconian Method of experimental philosophy prominent and over time a popularised form of this experimental philosophy came to be called science. If we go back to using ‘science’ as meaning its proper meaning of ‘knowledge’, then we can use ‘experimental philosophy’ properly too and stop having these arguments.

So then maybe some economics would be experimental philosophy if the economists did experiments. But then someone might say like the melancholy Jacques “All the world’s an experiment, and all the men and women merely experimenters and experimentees…” And another concur “For what else is the life of man but a kind of experiment in which men in various costumes perform until the experimenter motions them out”.

So this replaces the familiar opposition between art/culture and nature – with this experiment and nature opposition (maybe it is a dialectic?).

But experiment is difficult to define – because the word comes from Latin and Old French – from experiri (to try) with the suffix -ment meaning action/result. This is not a very useful meaning to contrast with ‘nature’ because animals try-act all the time and we would not usually oppose them with nature.

So we need to think of experiment as a type of art wherein a powerful subject/actor puts less powerful inert things or sentient beings on trial.

It makes sense Francis Bacon would do this because he used torture on people – conceivably to put them on trial to get them to tell him something he wanted to know.

However – we find this is quite interesting when looked at in relation to nature — because in experiments something or someone’s nature is put on trial to determine what it is. So we can now define experimental philosophy as an art wherein the experimental philosopher as subject will subjects various objects/natural-forms to various sorts of acts by way of a trial in order to get a result as to what the objects/natural-forms are.

Therefore – the idea that ‘all the world’s an experiment’ is no more correct than saying all the world’s a stage — because one would have to first prove that everyone has the subjectivity of the experimenter.

Many people object to having experiments done on them, or on experiments being done on other people or animals or even on genes. Therefore – we know that not everyone shares the subjectivity of the experimenter.

So , whilst experimenters may desire to think the whole world is an experiment wherein they can experiment just as they please — there is no good reason to trust their word for it. And, in fact, that they would assert this just goes to show that they are not very good at accommodating other understandings of the world — this is no great wonder when you stop to consider that the whole method is started by an awful torturer.

So then Hume makes an additional folly to the one engaged in by the author of the Algebra of Conscience — as well as assuming a mechanical view of humans and the social world as consisting of “regular springs” – he also assumes that all history — both in the doing of and the recording of — is in essence experimental, as in the sense of experimental philosophy. Thus has Hume perpetuated a greater folly than to translate Hamlet into algebra.

“These records of wars, intrigues, factions, and revolutions, are so many collections of experiments, by which the politician or moral philosopher fixes the principles of his science, in the same manner as the physician or natural philosopher becomes acquainted with the nature of plants, minerals, and other external objects, by the experiments which he forms concerning them.”

* The main likenesses are the subduing of nature and bending it to their own ends by Prospero and the torturer Francis Bacon, and the the setting of the play in an island in the new world like the setting of Francis Bacon’s science fiction secret society utopia book

6

ZM 10.15.14 at 5:14 am

ooops – I forgot to delete the last two paragraphs. One day I will remember I should proof read.

7

Merian 10.15.14 at 5:47 am

This is the point where I don’t know if I should feel comfortable or marginalized by having grown up with, and ingrained in the way I think, an overarching notion of Wissenschaft, or Wissenschaften, that is, all the disciplines that deal with the systematic and formal construction of knowledge, rather than having to figure out the exact demarcation of science (archaeology, linguistics and sociology — yay? but history — nay?). I don’t know who Smith is, but I go with them.

8

Bruce Wilder 10.15.14 at 6:53 am

Chris Mealy: The steam engine came before thermodynamics, airplanes came before aerodynamics, and vaccines came before the germ theory of disease.

What Papin, Savery and Newcomen understood was probably largely limited to mechanics and pressure, but Watt had a well-developed theory of heat and work. And, the Wright Brothers (and their rivals) certainly had an aerodynamics.

9

Bruce Wilder 10.15.14 at 6:56 am

ZM @ 155

I think Hume, in his remarks concerning human nature, is just chipping in his ante as an Enlightenment philosophe. They all did that.

10

ZM 10.15.14 at 7:18 am

Bruce Wilder,

“I think Hume, in his remarks concerning human nature, is just chipping in his ante as an Enlightenment philosophe. They all did that.”

Well lots of people always complain here about the sad loss of Enlightenment thinking like Hume’s (not myself obviously, it would be better riddance if it had completely gone and then we wouldn’t know how to make climate change like we are currently making with our great learnings from experimental philosophy).

But Henry says this quote is relevant to Smith and Gobrey’s argument.

So far as I can tell Gobrey is saying that most economics is not scientific if by science we are taken to mean experimental philosophy – because most economics is not about experiments and economists are not good at predicting what will happen in economies as they would be able to if either they were better scientists or economies were more regular in their functioning like an airplane (I can tell if he thinks it the fault of economists or economies is what I mean) .

Whereas Smith seems to be saying: you can write down you observations of what happens in the world; or you can do repeated observations of something that occurs very regularly (although he does not say which are regular events and which are events that happen irregularly like harvest being destroyed [if they happened with regularity you would stop farming or farm elsewhere]) ; or you can conduct repeated experiments in controlled conditions and write down your observations.

I do not quite see the disagreement, it would have been helpful if Henry had summed the crux of the matter up. It might be that Smith is asserting economics is more scientific and has greater predictive power than Gobrey asserted. But if this is the case I am unsure if Smith is making such a strong claim as Hume?

11

dsquared 10.15.14 at 8:32 am

Since there is obviously no such thing as a controlled experiment in astronomy, this definition of “science” would suggest that Kepler, Galileo and Copernicus were not scientists and I think that’s a bit daft.

12

dsquared 10.15.14 at 8:35 am

By the way, it cannot be said often enough that this:

Then people get angry at economists when they don’t predict impending financial crises, as if having tenure at a university endowed you with magical powers.

Is self-exculpatory bullshit, of the sort which the profession does itself no favours at all by trying to get away with.

Predicting financial crises with exact enough timing to make money out of doing so? Difficult. Noticing that an economic order is based on increasing debt/income levels? This was not hard. Understanding that real estate bubbles tend not to have soft landings? Not hard.

13

reason 10.15.14 at 8:37 am

I think here I should mention the long series of blog posts that Paul Krugman has made pointing out the important qualitatively correct PREDICTION he and others who understand what a liquidity trap is made that economists who either deny the possibility of a liquidity trap or deny it’s current relevance did not. Economics does too make predictions. The problem is that large parts of the “profession” don’t want economics to be a science (i.e. they deny falsability).

14

reason 10.15.14 at 8:44 am

dsquared
fully agree.

P.S. I don’t always agree with Paul Krugman or John Quiggin (in particularly I’m much more concerned about resource limits and how close we are to them than they are , I am a “national dividend” – i.e. basic income – supporter, and I think MMT have a point about some things) but mostly I think they are right, they are certainly more right than the Chicago school which got badly lost at some point.)

15

Zamfir 10.15.14 at 8:48 am

@ dsquared, that might be an exception that proves the rule. Astronomy is so strongly oresent early science, because its uncontrolled situations were often more ‘controlled’ than anything in a lab in those days. The mechanics of planets are almost perfectly friction-free movement under gravitation, repeating themselves with extreme accuracy. And even their complications are simpler, cleaner than the complications in an earthly lab.

That role continues to the present day, with some physical phenomena showing up in astronomy in a ‘purer’ form than in any lab. Neutrinos, gravity waves, possibly dark matter.

Social scientist would give up a few pinkies to have such a type of ‘uncontrolled’ observations.

16

reason 10.15.14 at 8:49 am

ZM @10
Smith is saying that Gobrey’s definition of science is far too narrow.

17

ZM 10.15.14 at 8:49 am

dsquared,
“Since there is obviously no such thing as a controlled experiment in astronomy, this definition of “science” would suggest that Kepler, Galileo and Copernicus were not scientists and I think that’s a bit daft”

They are not controlled – but because the heavenly bodies are so far away and very big neither astronomers or anyone else interfere with the conditions of their movements. Just the heavenly bodies interact amongst themselves. Also they are very routine in their interactions more than most things unless you could use great telescopes to look back in time (if that is what great telescopes allow you to do ) before all the orbits were regular. They are so non interfered with and routine people used the heavenly bodies to situate themselves in time and then space with astrolabes and the like.

So maybe this gave the idea of if you studied other things in controlled conditions and repeated your experiments you would find all the rules in every field you could imagine and thus understand the universe.

This worked better in biology than it did in sociology and anthropology and economics.

18

Zamfir 10.15.14 at 8:54 am

Addendum: i have no interest at all in casting cootied social sciences from the House of Real Science.

It’s just worth pointing out that astromical observations do behave a lot like lab observations, more than perhaos any other type of uncontrolled observations.

19

Agog 10.15.14 at 10:07 am

That snippet from Hume reads like it could have been one of the triggers for Kant’s Idea for a Universal History – which of course includes the famous phrase “out of wood so crooked and perverse as that which man is made of, nothing absolutely straight can ever be wrought.” (In de Quincey’s translation).

20

Brett Bellmore 10.15.14 at 10:23 am

“Then people get angry at economists when they don’t predict impending financial crises,”

I don’t particularly get mad at economists who fail in that regard, in part because predicting the timing of a financial crisis is like predicting next month’s weather in detail, and in part because economics largely fails to be a science for the reasons Reason identified: Its a discipline highly dependent on funding sources that wouldn’t want to hear what a genuine science might tell them.

I do get amused by economists who, after the crisis starts, make graphs, and the come up with all sorts of explanations for why they should be trusted despite reality not following the graph.

21

ZM 10.15.14 at 11:06 am

Zamfir,

“Addendum: i have no interest at all in casting cootied social sciences from the House of Real Science.”

The problem with this is twofold.

1. First — On The Scientific Subject: Here we have the question of what is a science in the House of Science? And why can’t we put every single thing into the House of Science?

I showed how ‘science’ used correctly just means knowledge – but no-one uses the word science correctly as just a latin synonym for knowledge. Knowledge must come in two parts – the getting of knowledge, and the knowing and telling of knowledge.

On the getting of knowledge — I repeatedly walk down my street — I expect much more repeatedly than any economist repeats their experiment. Therefore have I a great scientist getting knowledge as I walk down my street so repeatedly often noticing the different flowers for the seasons or when the neighbour’s hen is in the front garden? I have not myself noticed I am a scientist, but I will listen to your arguments if you would like to declare me one. If I am sadly not considered a scientist as I walk down my street — then the question becomes – which categories of knowledge-getting are scientific and which are not?

On the knowing and telling as knowledge — there are all sorts of knowledges that no-one usually has as sciences. For instance – is The Life of Johnson a scientific work? Is The Canterbury Tales a scientific work? Is Van Gough’s Sunflowers a scientific botanical illustration? Is Thucydides work scientific? Are all essays scientific? What about sandwiches? Etc Etc

So — then if you think not every way of knowing and every telling of knowledge are scientific — what are the criteria that makes knowledge (or representations of knowledge) scientific or otherwise?

2. Second — On The Scientific Object: Here we have the question of what makes objects suitable for scientific study if only some forms of knowledge are scientific? Are all objects suitable for knowing in a scientific way (as defined by whatever the criteria is agreed upon for 2. above)?

We will go back to my earlier example of the Algebra of Conscience — from the section on translating Crime and Punishment into algebra: “Luzhin is aggressive, but Lebeziatnikov is a coward and inclined to compromise, and neither is sacrificial.
Luzhin corresponds to the expression
(b) (I,2)
and Lebeziatnikov to the expression
(d) (II, 2) A “

Is Crime and Punishment the right object for algebra?
I would say it is not. If you agree it then follows that some objects are appropriate for scientific study and others are not.

The social sciences like sociology, anthropology, and economics I think all came to be when some sort of mechanical structural view of things was dominant. That is why you had all those silly books with the ages of mankind and some tacked on teleologies for the last chapters.

But I have not ever met a living professor who thinks structuralism is true these days. Structuralism died out of prominence around the middle of the twentieth century — but it is sort of like a ghost because a the social sciences don’t really make sense as sciences unless you accept structuralism as the equivalent to gravity. Without structuralism you may as well just go back to having history which is a very old muse and not a social science.

22

Neville Morley 10.15.14 at 11:31 am

I think this is one of the points where Hume is – probably deliberately – channelling Thucydides, and his claim that studying the events of a particular war can offer broader insights into the behaviour of humans and hence future events; “because of the human thing” (a phrase that’s often translated as “because of human nature”), events tend to repeat themselves in much the same way. And this strengthens the link to Kant, who cites Hume’s line about the first page of Thucydides being the first page of true history.

@ZM #21: for what it’s worth, much ink has been split on the question of whether Thucydides’ work should be considered scientific, which of course depends on what one thinks that means, and if so whether that’s a good thing or not. From early C19 to early C20, the idea that Thucydides had invented ‘Geschichte als Wissenschaft’ was an established if not universally accepted proposition. I don’t think that invalidates any of your other arguments.

23

Anderson 10.15.14 at 11:51 am

“History is philosophy learned from examples.” I’ve seen it attributed to Thucydides.

24

ZM 10.15.14 at 12:10 pm

Neville Morley,

“much ink has been split on the question of whether Thucydides’ work should be considered scientific, which of course depends on what one thinks that means….From early C19 to early C20”

I am not overly surprised to learn this, although my knowledge is far from strong in this area. the somewhat dim understanding I gathered as an undergraduate is that a certain split or difference in western historiography is considered to date back to the differing approaches of Thucydides and Herodotus. Where Thucydides is seen as an early example of the universalising impulse Herodotus is seen as an early example of interest in cultural differences. Is that about right?

25

Anderson 10.15.14 at 12:49 pm

“Since there is obviously no such thing as a controlled experiment in astronomy”

Even better: study of the universe outside our solar system is based on data that is years old when it reaches us — sometimes MANY years old. History, in other words?

26

Zamfir 10.15.14 at 1:21 pm

@ZM, i would say that you could indeed study your street in a scientific way. That coukd include bith cataloguing aspects of your street, and looking for deeper patterns and connections. Just walking along and noticing the colors of the flowers is not enough, but we don’t consider simple stargazing astronomy either.

The problem then is really one of importance. People do not have much interest in a rigorous study of your street, and the generalizable aspects of your study are mkst likely already known. So it would be probably be highly uninteresting science, but I am OK with still considering it science, just as I am happy to consider my hobby drawings art even if they are not of much value.

27

Bloix 10.15.14 at 1:35 pm

Gobry:
Science is the process through which we derive reliable predictive rules through controlled experimentation.

Stephen Jay Gould (from Wonderful Life: The Burgess Shale and the Nature of History (1989):
Beyond a platitudinous appeal to open-mindedness, the “scientific method” involves a set of concepts and procedures tailored to the image of a man in a white coat twirling dials in a laboratory–experiment, quantification, repetition, prediction, and restriction of complexity to a few variables that can be controlled and manipulated. These procedures are powerful, but they do not encompass all of nature’s variety. How should scientists operate when they must try to explain the results of history, those inordinately complex events that can occur but once in detailed glory? Many large domains of nature–cosmology, geology, and evolution among them–must be studied with the tools of history. The appropriate methods focus on narrative, not experiment as usually conceived.

The stereotype of the “scientific method” has no place for irreducible history. Nature’s laws are defined by their invariance in space and time. The techniques of controlled experiment, and reduction of natural complexity to a minimal set of general causes, presuppose that all times can be treated alike and adequately simulated in a laboratory…

Historical explanations are distinct from conventional experimental results in many ways. The issue of verification by repetition does not arise because we are trying to account for uniqueness of detail that cannot, both by laws of probability and time’s arrow of irreversibility, occur together again. . . . And the issue of prediction, a central ingredient in the stereotype, does not enter into a historical narrative. . . .

These differences place historical, or narrative, explanations in an unfavorable light when judged by restrictive stereotypes of the “scientific method.” . . .

Thus, historical scientists often import an oversimplified caricature of “hard” science, or simply bow to pronouncements of professions with higher status. Many geologists accepted Lord Kelvin’s last and most restrictive dates for a young earth, though the data of fossils and strata spoke clearly for more time. (Kelvin’s date bore the prestige of mathematical formulae and the weight of physics, though the discovery of radioactivity soon invalidated Kelvin’s premise that heat now rising from the earth’s interior records the cooling of our planet from an initially molten state not long past.) . . .

The firm requirement for all science–whether stereotypical or historical–lies in secure testability, not direct observation. We must be able to determine whether our hypotheses are definitely wrong or probably correct…

28

mattski 10.15.14 at 1:36 pm

sometimes MANY years old.

When I look up at the stars, are they old or young?

29

ZM 10.15.14 at 1:47 pm

Zamfir,

Hmmm, I am not convinced you have happened across the way of determining what is science from what is not science with these two categories:

Not-science: going along and noticing things
Science : going along and cataloguing things

I think you have instead happened upon the way of determining specific cataloguing behaviour from general noticing behaviour of which cataloguing behaviour must be a subset.

People do study streets as it happens, Jan Gehl for example. I don’t see why you think my street is a particularly unimportant street unworthy of study — it’s really a quite nice street I promise you. It would certainly be a very interesting study indeed, a study of my street. In fact — studying a street is actually quite an expansive topic — once I was assigned to do a whole study of just one building , this was a lengthy enough assignment and I didn’t even cover the topic of just the one building exhaustively.

30

Brett Bellmore 10.15.14 at 2:18 pm

Science is the challenging practice of letting the evidence lead you where it wants to go, regardless of where YOU want to go. Probably the most challenging thing we half evolved apes are capable of. If you don’t have procedures in place for reality to prove your favorite theory is wrong, you’re not doing it.

31

Teachable Mo' 10.15.14 at 2:23 pm

“Understanding that real estate bubbles tend not to have soft landings? Not hard.”

Unless you are Robert Lucas.

32

ZM 10.15.14 at 2:24 pm

“Science is the challenging practice of letting the evidence lead you where it wants to go, regardless of where YOU want to go”

Should we be hopefully expecting you to change your mind about climate science then?

33

Anarcissie 10.15.14 at 2:43 pm

Brett Bellmore 10.15.14 at 2:18 pm @ 30:
‘Science is the challenging practice of letting the evidence lead you where it wants to go, regardless of where YOU want to go….’

On the other hand, Uncle Albert said, ‘It is the theory which tells us what we can observe.’ That is, the observer participates in and is part of the observation.

In regard to Science versus knowledge, we generally conceive of Science as specifically social, and organized institutionally, regardless of the Noah-Smithian category a particular instance of it happens to fall into. Its social nature introduces prejudice and politics, the thumb on the scale I mentioned earlier in the case of Economics.

34

Peter Erwin 10.15.14 at 2:45 pm

Even better: study of the universe outside our solar system is based on data that is years old when it reaches us — sometimes MANY years old. History, in other words?

Cosmology and sub-fields of astronomy like galaxy evolution involve direct measurements of things changing over time, at least in a statistical sense. You can’t observe our own galaxy back in time (though you can something like “archeology”, in the sense of looking at our galaxy in detail and trying to figure out how it got that way), but you can observe broadly similar galaxies back in time by looking at things further and further away: galaxies as they were 1 billion years ago, as they were 5 billion years ago, as they were 10 billion years ago…

35

mbw 10.15.14 at 3:00 pm

All the definitions of science given above sound myopic. Trying to put together the developments of General Relativity, evolution by natural selection, routine materials characterization, etc. in one algorithm is a fool’s errand. Even worrying about defining big words like “science” is rather pointless. For those who find such definitions comforting, perhaps Bridgman’s is the best:

“The scientific method, as far as it is a method, is nothing more
than doing one’s damnedness with one’s mind, no holds barred.”

36

Peter Erwin 10.15.14 at 3:06 pm

dsquared @ 11:
Since there is obviously no such thing as a controlled experiment in astronomy, this definition of “science” would suggest that Kepler, Galileo and Copernicus were not scientists and I think that’s a bit daft.

This also applies to quite a few other sciences: most of geology, meteorology, quite a lot of biology (e.g., evolutionary biology, though you can do some experiments there, too), etc.

And, pace Zamfir and ZM, you really aren’t doing “controlled experiments” in astronomy. You can’t try changing the mass of the Sun to see what happens, or swapping Jupiter and Mars in their orbits or changing their velocities; you can’t even (prior to the 20th Century) introduce a small satellite into the system to see what happens to it.

(I’d venture that Medieval and Renaissance alchemy was probably a meaningful precursor to the idea of experiment.)

37

tomsk 10.15.14 at 3:14 pm

“Since most people think math and lab coats equal science, people call economics a science…”

Do economists wear lab coats?

38

Harold 10.15.14 at 3:20 pm

It *is* a very nice quotation.

39

Merian 10.15.14 at 3:23 pm

As a working geoscientist who received my original training in pure physics, I’ve come to the conclusion that for purposes of figuring out the object of science as a whole, the experimental vs. observational debate is vastly overrated. It’s essential in some disciplines that just happen to have a huge body of observation — and started out as such — but also offer important space for experimentation, such as biology or psychology, and where the development of the controlled experiment qualitatively changed the discipline and markedly expanded its scope and force. In geophysics, not so much. Sure, you’ll ran across the rare experimental volcanologist once in a while, but as a field, it’s unapologetically observational. My own work could be described as finding ways, or improving existing ways, of turning data into information and knowledge.

40

TheF79 10.15.14 at 3:27 pm

I find these “is economics a science” discussions to be amusing, since by “economics” people seem to mean a particular sub-branch of macroeconomics, reflecting maybe 1-2% of the discipline (I’m at an R1 school and precisely zero of my dozens of colleagues do the sort of big macro policy, thumb-on-the-scales stuff that shows up in the papers and op-eds). Hell, the number of people working in the field of “experimental economics” and directly doing controlled experimentation is probably equal to the the number of people in this particular sub-branch of macro, not to mention all the folks in development economics and related fields doing Randomized Control Trials (RCTs). Anyone really interested in thinking about the roles of experimentation and empiricism in economics (and not just the one particular sub-branch of macro related to recessions and whatnot) should at least be familiar with Angrist and Pischke’s article on the state of empirics in economics: https://www.aeaweb.org/articles.php?doi=10.1257/jep.24.2.3 as well as the replies in that same issue to A&P (Christopher Sims takes issue with them on their discussion of empirical macro).

41

stevenjohnson 10.15.14 at 3:27 pm

I followed the links.

Gobry wrote “Countless academic disciplines have been wrecked by professors’ urges to look ‘more scientific’ by, like a cargo cult, adopting the externals of Baconian science (math, impenetrable jargon, peer-reviewed journals) without the substance and hoping it will produce better knowledge.” Bacon didn’t do math. Bacon’s prose was quite the opposite of jargon. Gobry should be pleased to write essays as stylish yet accessible as Bacon’s! Peer-reviewed journals in the only likely sense meant came long after Bacon. Everybody makes mistakes but misrepresenting something for its opposite is symptomatic of profound mental incompetence. I’m faintly surprised he can boot up his computer.

Smith doesn’t write anything so flagrantly foolish. His discussion of history omits any discussion of how documents and testimony are analyzed, or how historical facts are generalized into narrative. His second “empirics” cites statistical analysis without discussing the epistemic status and function of mathematics in the natural or social sciences. His comments on experiment fail to note the crucial role of scientific instruments. His notion of “natural experiments” as midway between empirics and experiment makes no sense to me. I don’t think you really derive much statistics in those cases, and it’s just history. His discussion of theory doesn’t say anything specific about generalizations nor about simulations. The collective nature of science is unremarked.

Hume was the great Tory historian, which I daresay is an aspect that Hume proponents overlook.

42

Jerry Vinokurov 10.15.14 at 4:02 pm

You can’t run experiments with a different sun, but you can observe lots of solar systems. Surprise, they behave just like GR predicts they’d behave. What’s more, you can do repeatable experiments in the laboratory to work out, say, how nuclear reactions happen, and then apply that knowledge to explaining how stellar nucleosynthesis works, as Hans Bethe did decades ago.

43

Harold 10.15.14 at 4:02 pm

Hume a Tory historian? That’s a new one on me.

44

Bruce Wilder 10.15.14 at 4:04 pm

mattski: When I look up at the stars, are they old or young?

When I look up at the stars, I am young.

45

Jerry Vinokurov 10.15.14 at 4:05 pm

Also, Gobry gives zero fucks about science as such (and his understanding of it is at such a primitive level as to render engagement with it unnecessary). He only cares about making sure big bad Neil DeGrasse Tyson doesn’t horn in on Catholicism or whatever.

46

Bruce Wilder 10.15.14 at 4:05 pm

Harold: Hume a Tory historian? That’s a new one on me.

Not just “a”, but in some ways for a fairly long time, “the”.

47

mattski 10.15.14 at 4:08 pm

Hume was the great Tory historian, which I daresay is an aspect that Hume proponents overlook.

What do you think is the significance of this? In my limited exposure to Hume I don’t at all feel a strong pull of class interest.

In my model of the universe we have the self at the center surrounded by a vast expanse of “non-self” with a blurry zone of humanity & ambiguity in between. Observation (science!) of the natural world excluding humanity is relatively straightforward. But the closer our observations get to ‘ourselves’–and that includes humanity–the greater the distortion due to self-interest (desire.) So, social sciences like economics will always be tainted by self-interest, but that’s no reason not to keep working at it.

So, stereotypically, science is the observation of things ‘out there’ and spirituality is the observation of things ‘in here’ and the goal in both cases is to try to distinguish between that which is seen vs that which is wanted or desired.

48

mattski 10.15.14 at 4:09 pm

Bruce 44

Right on.

49

Bruce Wilder 10.15.14 at 4:17 pm

tomsk @ 37: Do economists wear lab coats?

Apparently, they do, in a manner of speaking, from the evidence offered by TheF79 @ 40. Their donning of lab coats is called “econometrics” and involves much highly-stylized mucking about. (And, yes, I did click thru to skim the Angrist & Pischke article; and for what it is worth, I was once a student of Edward Leamer.)

50

bob mcmanus 10.15.14 at 4:55 pm

43 and 47 were predictable enough to make me laugh. Ideology works best when invisible or deniable. Yeah, yeah, empiricism, positivism, liberalism, capitalism, imperialism all so amazingly coincidental in period.

Whatever. Pilling on Empiricism at Marxists Org FWIW

51

L.M. Dorsey 10.15.14 at 4:56 pm

in other words, your experiment might find a causal mechanism that always works in the lab, but is just not that important in the real world.

Fwiw. This happens regularly in software development, where the peculiarities of one environment will cause problems of one sort and another not seen in another environment. For instance, transient memory leaks.

The confidence of Enlightenment science (and of popular scientism — both of which need to be distinguished from science as currently practised) flows in part from the belief that induction can be bullet-proof: if you collect enough data, you can arrive effortlessly (and infallibly) at general principles. Et voilà, the world’s your oyster.

Hume put paid to that, of course, and with it threatened to scuttle Enlightenment experimentalism. Karl Popper, taking Hume’s argument, was able to rescue induction, but, as I read it, at the price of its once boundless confidence to know (and to control) the world. Not that anybody paid much mind, power being its own justification.

52

Harold 10.15.14 at 5:28 pm

Bruce Wilder @46
I suppose you could say that Hume was a “Tory” in the same (a-historical) sense that some maintain that Tocqueville was a “reactionary and Paolo Sarpi an “atheist”.

Ernest Campbell Mossner, “Was Hume a Tory Historian? Facts and Reconsiderations”, Journal of the History of Ideas 2:, 2 (Apr., 1941): 225-36:

Excerpt: Hume’s political theory belongs to that large, non-party, Liberal tradition indicated by Professor Price. But the practical lessons of history that he teaches are colored by a cautionary skepticism concerning the likelihood of continuous human progress that belongs to what may with equal justice be called the large, non-party, Conservative tradition. Both of these traditions are timeless; but in so far as the one becomes identified with Whiggism and the other with Toryism, then, to that extent, Hume is in both camps. Such identifications, however, were developments of the nineteenth century.

If, then, a specific answer is to be offered to the original question posed by this paper, “Was Hume a Tory historian?”, that answer must take a plural form.

For the problem is a problem of semasiology. With some confidence now, in light of the facts and reconsiderations already presented, it may be said: (1) That Hume’s “History of England” was deemed Tory by his contemporaries chiefly because it was “inclined to indulgence towards the first James and Charles'”; (2) that Hume as a skeptic repudiating the dogmas of both parties is a Liberal in the large, non-party (and, historically speaking, nineteenth-century) sense; (3) that Hume as a skeptic chary of planned progress is a Conservative in the large, non-party (and, historically speaking, nineteenth- century) sense. It is the nature of the skeptical method to partake of all and to subscribe wholly to none; and the great skeptic fittingly rests content,
While Tories call me Whig, and Whigs Tory.

53

L.M. Dorsey 10.15.14 at 5:42 pm

@51
blech. Should read “Popper…was able to rescue experimentalism”, not “rescue induction”.

54

Kiwanda 10.15.14 at 6:32 pm

dsquared:

“Since there is obviously no such thing as a controlled experiment in astronomy, this definition of “science” would suggest that Kepler, Galileo and Copernicus were not scientists and I think that’s a bit daft”

While there may not be controlled experiments in astronomy (ignoring lunar landings and robots on Mars etc), there are predictions that have been tested: the cosmic background radiation was predicted, then found; black holes were discussed since the 18th century, and a supermassive one found in our galaxy; I believe there are other examples. My crude understanding of “dark energy” and “dark matter” is that they are a workaround to make calculations match observations, but they are a kind of prediction that may be validated or disproven some day.

While evolutionary biology in-the-large has few controlled experiments, I think it’s often happened that intermediate forms are predicted and then found; also, any geological sample that is examined has the implicit prediction that fossils found in it will be consistent with its date (and each other).

Macro-economics and string theory? Dunno.

55

Barry 10.15.14 at 6:48 pm

Jerry Vinokurov 10.15.14 at 4:02 pm
“You can’t run experiments with a different sun, but you can observe lots of solar systems. Surprise, they behave just like GR predicts they’d behave. “

Which is observational.

56

Barry 10.15.14 at 6:51 pm

dsquared 10.15.14 at 8:35 am
“By the way, it cannot be said often enough that this:

Then people get angry at economists when they don’t predict impending financial crises, as if having tenure at a university endowed you with magical powers.

Is self-exculpatory bullshit, of the sort which the profession does itself no favours at all by trying to get away with.

Predicting financial crises with exact enough timing to make money out of doing so? Difficult. Noticing that an economic order is based on increasing debt/income levels? This was not hard. Understanding that real estate bubbles tend not to have soft landings? Not hard.”

Not only was Noah wrong in his post (just because you can predict something doesn’t mean that you can prevent it or even alter it much), the criticism is not about predicting the details of the Great Financial Crisis.

What I heard people saying could be summarized as:

1) The reforms which were to make the system work better had the opposite effect.

2) The actions needed to mitigate the crisis were opposed by many of the people who had supported the reforms (Chicago and all neoliberals, that means you).

57

Jerry Vinokurov 10.15.14 at 6:52 pm

There’s not a really important substantive difference between “observation” and “experiment.”

58

Peter Erwin 10.15.14 at 7:04 pm

Kiwanda @ 54:
You do realize that you’re agreeing with dsquared, right? (And with me and with Meriand and Jerry Vinokurov.) It’s Gobry who is, according the quote that Henry presented, stuck in the “only controlled experiments make for real science” mentality.

59

Jerry Vinokurov 10.15.14 at 7:23 pm

The whole thing about “controlled experiment” is just about knowing what can be folded into some sort of ceteris paribus condition or abstracted away. It’s just that when you control something in a lab, you can often engineer an explicit situation in which external variables are “controlled for.” This is of course very nice when possible, but obviously not the only correct way to do things. How this bears on economics’ status as a science is a somewhat separate question, but it’s not one that Gobry is properly equipped to tackle with his extremely limited view of science.

60

Kiwanda 10.15.14 at 7:42 pm

“You do realize that you’re agreeing with dsquared, right?”

Ah, yes. Not sure why you might think that I would not realize that.

I didn’t actually understand that testability aspect of evolutionary biology until embarrassingly recently, so I like to mention it. (I understand that we see evolution of small organisms, and of human cells into cancer, all the time.)

Although, saying basically that “it’s all observation” makes the notion of “observation” kind of broad. Surely there’s some useful distinction between hind-sight explanation and testing of predictions; there is in a setting where statistical analysis is appropriate.

61

J Thomas 10.15.14 at 7:42 pm

#11 dsquared

Since there is obviously no such thing as a controlled experiment in astronomy, this definition of “science” would suggest that Kepler, Galileo and Copernicus were not scientists and I think that’s a bit daft.

I think you’ve established that the central thing here is not controlled experiments — though they are often important.

Put it this way — there’s an important difference between astrology and astronomy, and there’s an important difference between alchemy and chemistry.

Which side is economics on? Is economics more like astrology or more like astronomy?

I think it’s on both sides, but on the whole, on average, more like astrology.

Part of the difference is that anybody who wants to can do astrology, but astronomers insist it has nothing to do wit them. But we have a bunch of economists including particularly a lot of teachers who do nothing but astrology and I don’t hear any generally-accepted claim that it isn’t economics.

In all fairness, if our solar system had a dozen or more Jupiter-mass planets that interfered in a major way with other planets’ orbits, and if our sun was deep in the galactic core, close to hundreds of other stars that visibly affected each others’ orbits, astronomy might have spend considerably more time floundering about orbits. We lucked out that there were obvious regularities in the observable data that meant simple theories fit the imprecise observations, and then when the observations got more precise they revealed new regularities that new simple theories fit, so it was a sequence of successes building on successes.

62

Peter Erwin 10.15.14 at 9:07 pm

Kiwanda @ 60:
Ah, my apologies, then; I had the impression you were trying to correct or argue with dsquared in some fashion.

63

Dr. Hilarius 10.15.14 at 9:26 pm

Bloix @27: thank you for the Gould quote. Evolutionary biologists are often subjected to the claim that because they can’t do controlled experiments in labs (nonsense, they do so when it’s possible) evolution has no scientific basis.

Predicting the unknown from the known using a falsifiable hypothesis is science. Darwin predicted the existence of a moth with a 35mm long proboscis based on the flower characteristics of the orchid, Angraecum sesquipedale. Twenty-one years after Darwin’s death, the moth was discovered and confirmed as the orchid’s pollinator. No lab but still science.

64

guthrie 10.15.14 at 9:29 pm

Actually controlled experiments were part of geology back in the 19th century, asnwering the quesiton, “what conditions formed this rock”, feeding back into questions like “What are conditions inside the earth like” and so on. They often sealed minerals in iron containers and heated them to see what minerals would turn up.

Astronomy is an observational science; I’m sure economics could be much more so if we didn’t change so many variables over the years.

Finally, there’s the point that doing science, science as a verb, is a process involving feedback from experiments and theories, and often involve a creative leap of imagination to sort out, not a mere mechanical logical thought process.
So in summary, I view economics as a social science that is perhaps at the stage that physical sciences were back in the late 19th century – they have most of the pieces, but aren’t too sure how they all fit together, and they know a lot about the scale, from small to large, but a TOE is lacking and may never be possible. IN the case of economics we’d need a complete theory of human behaviour and something that ties that into large masses of humans.

Ultimately the real issue is that because of what economics is about, it gets involved in politics. A salutary example is climate change; the science is clear enough, was worked out back in the 70’s, based on nearly century old work, and yet because money and influence are involved you can still find people muddying the waters, lying, being deliberately stupid and ideologically blind.
The net result is that most people don’t know how good a handle we have on climate change, why we should act now, or rather 10 years ago, and how it will affect them.

Meanwhile, back in economics land, people find it easier to lie and obfuscate to make a living sucking up to the rich and powerful, rather than treating economics as an actual science. Maybe it’s the abstractions involved, or too many data driven geeky folk lose their heads over hobnobbing with the rich and powerful. Either way, the public perception of economics as not really being a science is formed, no matter that many economists are trying to do scientific work.

65

gianni 10.15.14 at 9:36 pm

All (well, most) of these different notions of ‘science’, conceived as this or that methodological approach, somehow strike me simultaneously as equally plausible and equally wrongheaded.

I find it far more interesting to think about the term ‘science’ sociologically – specifically, as a type of shibboleth demanded by certain gatekeepers prior to one’s admission into that cadre of the elect capable of producing ‘proper’ knowledge. One can only hope that their particular knowledge-community has taught them the right manners and pronunciations, but we all know of course that certain ideas are much easier expressed in one manner of speech than other. Rather than hearing the word ‘science’ as describing a discrete pattern of inquiry, it has always struck me as primarily concerned with patterns of inclusion/exclusion, and deployed as a weapon in the arena of knowledge production.

66

Harold 10.15.14 at 11:33 pm

As everyone here probably already is aware, only in the English language and culture, is the term “science” limited to empirical or experimental physical science. Not that this is a bad thing, necessarily. Still, this usage is rather recent. Even in English, up to the middle of the 19th c., what we know as “science” used to be called “natural philosophy” and in other languages “science” still means knowledge in general. I think it is recognized that the science (or discipline) of philology, at least, owes some of its methodology in assessing evidence to practices that originated in trial arguments in the legal profession.

67

pauldu 10.15.14 at 11:53 pm

Those who look to economics for scientistic predictions are not doing so without encouragement from within economics. Essays on economic imperialism (Lazear, Casson and others) suggest without a hint of irony that economics rightly holds “imperial” sway over other social sciences because of its superior scientific qualities, while Friedman, in his essay on “Positive Economics” defends his positivistic claim on the grounds that “The ultimate goal of a positive science is to develop a theory or hypothesis that yields valid and meaningful predictions about phenomena not yet observed”–and economics, in his mind, is capable of doing that.

As to Hume, it is worth noting his juxtaposition of the moral and the natural philosopher in this passage. Though it is decades since I read it, I seem to remember Passmore arguing in Hume’s Intentions that in challenging causality, Hume sought to not so much to debase natural philosophy, as to elevate moral philosophy, by suggesting that both trod on uneasy ground, each subject as much to subjective passions as to objective reason, allowing us to pursue economics as a “science” if we choose, but rejecting any easy imperialism for either economics or the natural sciences.

Per Agog, above, it wouldn’t be surprising to sense antecedents to Kant in this, for it is of course DH who with this argument awoke IK from his “dogmatic slumber”

68

otpup 10.15.14 at 11:54 pm

When geologists or astronomers or evolutionary do their own brand of science within their disciplines they rely heavily on experimentally based and verified science (physics, chemistry, biochemistry, molecular biology, etc). When economists practice whatever it is they practice they rely on math but appear to be in open warfare with related disciplines (e.g., psychology) which might have some relevance to the foundations of their theoretical concepts. Obviously this applies most relevantly to macroeconomics (though also to some extent to micro).

69

Jim Harrison 10.15.14 at 11:55 pm

As a couple of people pointed out, Hume’s remarks on history echo the line, attributed to Thucydides by Dionysus of Halicarnassus, that history is philosophy teaching by example. Thing is, this take on history was so traditional in Hume’s time, I wonder if anybody would have thought the ideas expressed in the paragraph were particularly Humean. Since the sentiments expressed aren’t part of our common sense, they make a more striking impression on us than than they would have on Hume’s contemporaries. What we have here is the situation explored in Borges’ famous story about rewriting Don Quixote.

About Hume as a Tory: Hume claimed that he was criticized because (quoting from memory) “I dared to shed a tear for king Charles.”

70

Harold 10.16.14 at 12:32 am

Dr. Johnson was also a Tory in this sense.

71

Thornton Hall 10.16.14 at 2:23 am

My head hurts. Gorby and Noah Smith both specialize in the field of misunderstanding things that most people don’t know are things in the first place.

What is it with Smith who has written several posts on how great “empirics” are, when that’s not even a word, let alone an obvious feature of economics as discussed in everything he blogs about when not writing vaguely about “empirics”?

72

ZM 10.16.14 at 3:32 am

Harold,

“Dr. Johnson was also a Tory in this sense.”

I had almost forgotten what a good idea it is to get Samuel Johnson’s opinions on his historical figure contemporaries.

You may or may not be surprised to know just as Samuel Johnson had only bad reports to make of Edmund Burke’s propensity to say whatsoever he thought would profit him with his masters, Samuel Johnson also abhorred David Hume and says he was a Tory merely by reason of being a Scotsman and not out of holding to any great Tory principles. The quote is:

“I observed that Mr. Hume, some of whose writings were very unfavourable to religion, was, however, a Tory. JOHNSON. “Sir, Hume is a Tory by chance, as being a Scotchman; but not upon a principle of duty, for he has no principle. If he is anything, he is a Hobbist.””

Boswell dreamed of the 8-year long departed Hume in 1784, dreaming Hume was pious inside but due to his vanity made outward shows of skepticism:

“Awakened after a very agreeable dream that I had found a Diary kept by David Hume, from which it appeared that though his vanity made him publish treatises of scepticism and infidelity, he was in reality a Christian and a very pious man”

Not only Samuel Johnson, who “holds Mr Hume in abhorrence and left a company one night upon his coming in” had an aversion to Hume — even the knave Edmund Burke “told Boswell that he only spoke to Hume because “the present liberal state of society required it.”” Imagine — Burke may have departed from his knavery and Liberal ideas if he had arrived upon a philosophy allowing him to ignore David Hume in good conscience.

I can tell you are wondering upon Adam Smith’s view of Hume, well, let me tell you, Adam Smith wrote favourably of the departed Hume but this “single, and as I thought, a very harmless sheet of paper which I happened to write concerning the death of our late friend, Mr Hume, brought upon me ten times more abuse than the very violent attack I had made upon the whole commercial system of Great Britain.”

(quotes except the first from http://www.markhannam.com/essays/essay3a.htm)

73

M.Jamison 10.16.14 at 3:43 am

Is economics a science or is it a technology used to design systems of prediction.
Economists look at exchanges and interactions between entities and design systems to explain those interactions and perhaps predict future interactions. Essentially though those explanations and predictions are really suggestions on how to manipulate those exchanges in order to achieve some goal.
While most economists would say that they start with a neutral slate and a purpose of observing and explaining it seems that most economist start with some fundamental values, judgments, ideas about equity and justice. Are they really explaining how markets work or are they offering models and systems that seek to organize markets to work in some way that reflect a set of preferences on systems of equity and justice.
That seems less a science of observation, explanation,and prediction than a tool or technology designed to promote an outcome. Economics can never escape politics or systems of governance because markets are constructions arising from sets of rules prescribed by society and societal institutions.
Economists may seek the mantle of science rather than the technology of systems design because science implies a search for truth whereas technologies of systems often require a certain type of political advocacy.

74

jkay 10.16.14 at 4:18 am

Thucydides used evidence to come to conclusion after conclusion. He came up with the idea of progress from tomb and text evidence. He was first to point the best kind of constitution, a checked and balanced democracy like ours, from life and war experiences. He had Athens’ trouble, democratic oppression, happen to him personally. He was exiled quickly in his service, for being beaten. So, scientist is what he did.

He was one of the first wave of scientists,`along with India. Did you know The Buddha was a major democrat?

75

Harold 10.16.14 at 5:58 am

Johnson’s religious beliefs underlay all his opinions, so it is not surprising he didn’t like Hume, an atheist. I had forgotten that he didn’t like Burke, but it stands to reason also, since Burke supported the American Revolution and Johnson did not.

As for his politics, he famously remarked that “, “a wise Tory and a wise Whig, I believe, will agree. Their principles are the same, though their modes of thinking are different.” More here http://www.samueljohnson.com/jpolitics.html

76

J Thomas 10.16.14 at 7:14 am

As for his politics, he famously remarked that “, “a wise Tory and a wise Whig, I believe, will agree. Their principles are the same, though their modes of thinking are different.”

As in the scriptural quote, “Everybody I know who is right always agrees with ME.”

77

Neville Morley 10.16.14 at 7:52 am

Belated responses to Thucydides-related points above…

@Anderson #23 and @Jim Harrison #69: actually the line is attributed to Thucydides by an anonymous, late text that gets attributed to Dionysius of Halicarnassus; no evidence whatsoever that it tells us anything about Thucydides himself thought, but it’s possible to read it as a response by a fan of his to Aristotle’s notorious put-down of historiography as (unlike philosophy) telling us only about individuals and their individual deeds.

I would argue, however, that Hume is definitely not arguing in this tradition of exemplary history (which was, as Jim observes, very well established as the default for the humanists). Rather, the mid-C18 is precisely when a new awareness of historical difference is emerging, so exemplarity no longer works because the past is nothing like the present, so instead you switch into the claim that there are universal principles or attributes that can be discerned even in this past.

@ZM #24: yes, there’s a long tradition of presenting the development of historiography in terms of Herodotus v. Thucydides – Thucydides is generally seen as the victor, at least until the mid-C20, but the terms in which the conflict is presented change over time. It can be universalising v. cultural difference (which is how Marshall Sahlins takes it in ‘Apologies to Thucydides’ – but still ends up championing Thucydides, oddly enough); more commonly it’s critical analysis and scepticism v. any old story about flying snakes. Really it depends on what you think history should be, so you then pick the intellectual ancestor that suits you and write the history as a teleological progression from that point – or, in the case of some C19 accounts, you see Thucydides as a modern before modernity whose genius was completely unrecognised until von Ranke et al invented Geschichte als Wissenschaft and recognised him as their peer and inspiration.

@jkay #74: well, that’s the conventional account; most of those statements, apart from the exile, have been disputed…

78

Neville Morley 10.16.14 at 7:54 am

Apologies for prolixity, but this is one of my research interests, so I’m prone to witter on for hours about it…

79

Scott Martens 10.16.14 at 10:05 am

I have a favourite example of the problems of defining a scientific method. In the third century, Claudius Ptolemaeus wrote the definitive smackdown of heliocentrism. He said, basically, if the Earth was actually spinning fast enough to cause day and night with its rotation, it would have to be spinning at thousands of miles per hour in the vicinity of Egypt (where he lived). Now, go outside and throw something into the air. Watch how it doesn’t whoosh away at thousands of miles per hour. Ergo, the Earth cannot possibly be spinning, or at least that cannot be the cause of the apparent movement of the sun, moon and stars.

Ptolemaeus was using advanced math, he was a very learned, thinking man with an open mind to the possibilities of the world, and made an argument that passes every test of scientific methodology I am aware of. And he still managed to reject an absolutely true theory about the universe in a fashion so definitive it was more than 1000 years before anyone seriously contested his claim. Smith is right about “omitted variables” posing a problem for empiricism – and you don’t need statistics or complex math to make that point, you just have to understand why Ptolemaeus was wrong despite doing impeccably empirical, rational, modern science.

Gobry is wrong: Controlled experimentation has nothing to do with many branches of science. Consider astronomy. Smith is wrong too: Repeated observation of unelicited events plays very little role in many modern sciences and is positively misinformative in areas as fundamentally important as medicine. And the discovery that the Earth turns, and Ptolemaeus was wrong, did not involve a single controlled experiment whatsoever. Nor did it involve using many observations to devise better predictive models. Copernicus’ actual figures were worse than the state-of-the-art geocentric model at the time. His main line of argument was that he could explain the retrograde motions of the planets in a more intuitive way.

And as for Hume… if he’s right, where is my Asimovian psychohistory? He is proposing that man is always the same, and by observing his behaviour under varying circumstances we can make a predictive science of his behaviour. Contrast this with Marx, for whom man is the constant who is constrained in his behaviour by his varying circumstances, so we can make a predictive science of history by ignoring the individual almost entirely. Both assumed a largely constant man and a constantly changing world, and the sciences they both believed in don’t exist. So somewhere, something is wrong with their postulates.

It’s not clear to me that science is anything other “math and lab coats” and arguing and the systematic replacement of each generation of scientists by their grad students who can choose to agree with them or not depending on how convincing their arguments are. And honestly, I’m not convinced the math is really required.

80

Barry 10.16.14 at 12:32 pm

Jerry Vinokurov 10.15.14 at 6:52 pm
“There’s not a really important substantive difference between “observation” and “experiment.””

Wow. Just – wow.

81

Anarcissie 10.16.14 at 1:55 pm

If we first construct a theory about what we are going to observe, as it seems we must in order to observe anything coherently, then in a sense every observation is an experiment; the difference between a formal experiment and an ordinary observation is in the degree of our participation, and the line drawn between them somewhat arbitrary.

82

stevenjohnson 10.16.14 at 2:37 pm

As far as I can tell, all attempts to define science as a particular method fail. Popper is one who fails the worst. His attempt to define away the possibility of social science by reducing science to predictions has always been acceptable to all the pilgrims to Mont Pelerin, of course. If you must insist on philosophy of science, I believe Mario Bunge’s presentation of a hypothetico-deductive method is much superior, though I think it is still not satisfactory.

I think science is the body of knowledge about reality attained by systematic study, which means correction of individual errors and generalizations from experience and corrections of generalization from further experience. “Science” is not some illusory method that provides logical a priori certitude in the way math or logic, religion or philosophy, are alleged to do. It’s what we’ve found out about the way the world is. And we found it by correcting the errors of individual experience and by surpassing the limits of individual experience by use of instruments. When it comes to finding out how things are, observational science is just as much science as controlled experiments. Science is not a sequence of theorems but a collections of theories.

However, this is very much a minority view. It is very possible that the majority of working scientists, whatever they may do in their own field of research, formally adhere to the notion that science does not describe reality, which can only be experienced, not questioned nor explained. They merely correlate instrumental data. That’s why predictivism becomes so critical to antirealist theories of science I think. There has to be some sort of internal justification for accepting some theories and rejecting others. Naturally by the antirealist standard mere observational science must be a lesser enterprise, being predicated upon the naive realist idea you can actually learn about reality.

Which brings us to the notion that Claudius Ptolemaeus was a good scientist. In the Ptolomaic system the earth is near the center of a system of concentric spheres and is just as it seems, motionless. The outermost sphere is of course the firmament of stars. Originally it was held something turns it, which motion is passed on to the inner spheres. Aristotle postulated a different type of matter, called ether, which did not behave like anything on earth. Epicycles responsible for the changing path of the planets (the Sun and Moon are also “wanderers” in this system) are embedded in each sphere. In his personal contribution to the system, the motion of the epicycles is uniform in relation to a point called the equant, removing the need for an active agent causing planetary variations in speed and brightness. The whole apparatus is kept centered on the earth by the northern and southern poles.

The Ptolemaic system did indeed correctly predict observations, at least much more correctly than any other known. And it was compatible with Greek philosophy (and religion too, I think.) What is not clear is what other properties than circular motion the ether possessed? Or in what sense ether could be matter rather than a prosy version of the Fates’ Spindle of Necessity? If there are holes in the deferents for the epicycles to revolve through, how did they form? Why doesn’t the motion of the spheres transmitted to the poles cause the earth to rotate? Why don’t those gigantic poles loom over the northern skies? Why didn’t Ptolemy drop a rock from a ship’s mast to see if it fell behind, like a ship’s wake? (Yes, I saw Agora, sue me.)

From the antirealist, predictivist viewpoint, Claudius Ptolemaeus is just as good a scientist as any today. From the outmoded naive realist viewpoint, Ptolemy’s failure to take his model seriously as a representation of the the world and try to connect with all the evidence available made it questionable science from the very beginning. There’s no justification for sneering. Most of us still can’t do any science, much less bad science that at least has predictive utility! I don’t think there’s much excuse for refusing to acknowledge that the spectacular failure of the Ptolemaic system is a powerful example of the inadequacy of the antirealist, predictivist notions of science.

As for the scientificity of modern academic economics? I think that it has its own version of the poles, in the sense that if favored economic theories are valid, they should connect to economic history, anthropology, sociology, political science, psychology, geography, demography. Personally I know of some isolated efforts, like development economics, but I can’t say I’m aware of any great successes by them. By and large, it seems to me academic economics appears to be fixated on correlating numerical indices in selected and limited samples, or working out theorems without regard to how they apply to the real world.

83

Ze Kraggash 10.16.14 at 2:42 pm

“If we first construct a theory about what we are going to observe”

Agreeing with gianni, 65 here: it’s not enough to have a theory and observe or experiment; your theory has to be approved by the authorities, otherwise you are a crackpot, not scientist.

84

Jerry Vinokurov 10.16.14 at 3:09 pm

“There’s not a really important substantive difference between “observation” and “experiment.””

Wow. Just – wow.

Barry, I don’t know what you think your incredulity proves. I’m just telling you that unless you don’t think that, say, astronomy is an experimental science, you’re going to have to be prepared to not put that much weight on the difference between “experiment” and “observation.” There just isn’t that much difference between them in actual scientific practice; an experiment is just a setup in which you control for certain factors, nothing more. Conversely, systematic observations just are the things that make up a good experiment.

85

Jerry Vinokurov 10.16.14 at 3:21 pm

@stevenjohnson,

It is very possible that the majority of working scientists, whatever they may do in their own field of research, formally adhere to the notion that science does not describe reality, which can only be experienced, not questioned nor explained. They merely correlate instrumental data.

I can only speak of the scientists I know personally, but none of them adhere to this view. Of course, it’s widely acknowledged that some models get closer to reality than others, but I haven’t met anyone who didn’t think they were trying to get at the underlying truth rather than “correlate instrumental data.” Most scientists that I know are “naive” realists, and a good thing it is too. The use of the word “outmoded” to describe that position is, in my view, unwarranted.

86

stevenjohnson 10.16.14 at 4:54 pm

Jerry Vinokurov @85

I would like to think this is true, but in antirealism, as in Hume (and Berekely before him,) the theoretical annihilation of reality leaves the everyday world untouched. Common sense is all you need for that, and no one need confuse themselves wondering how other scientific results connect together. Near as I can tell, scientists like the common man are encouraged to take things at face value, just so long as they don’t get uppity. Personally, I think antirealism has been flagrantly outmoded since Ernst Mach denied the reality of atoms.

But that’s me. Every time a scientist says all scientific claims are provisional is denying that facts that have been long corrected are still not reliable descriptions of reality, but temporary assertions awaiting replacement by something better, is basically denying that science finds out anything about reality. And it isn’t helped by pretending that they can label some claims as having a negligible probability of refutation. If they really thought that, they would actually neglect the possibility of refutation and forget “provisional!” If they insist that they must say “There is probably no such thing as magic,” instead of “There is no magic,” the only point is to deny that science can achieve a description of reality.

The very widespread fetish for Bayesian statistics seems to mean a commitment to subjective probability, as opposed to outmoded frequentism. I don’t see how that’s realism either. Similarly, I don’t see how Copenhagen is realist and I’m pretty sure that’s the dominant view by far.

87

bianca steele 10.16.14 at 5:10 pm

Steve Johnson @ 82: I think science is the body of knowledge about reality attained by systematic study, which means correction of individual errors and generalizations from experience and corrections of generalization from further experience.

I think this is wrong. When philosophers of science talk about science as an error-correcting enterprise, they don’t mean that science goes out into the real world and corrects ordinary people’s errors. They mean that science is a way of getting better knowledge through correcting previous knowledge (sometimes previous science, sometimes previous earlier knowledge). Civil engineers and inventors, like those mentioned above, don’t go on their merry blundering way, occasionally corrected by science. That’s absurd. Civil engineers and inventors study and apply science itself. Science isn’t optometry, correcting individual errors, or psychoanalysis (unless they’re actually optics or psychoanalysis). There’s no “science” that looks at individuals’ errors and figures out what’s wrong with them so they can be corrected by science.

At least, this way of putting it seems to be likely to confuse.

(Crossposted, posting anyway)

88

Jerry Vinokurov 10.16.14 at 5:29 pm

I would like to think this is true, but in antirealism, as in Hume (and Berekely before him,) the theoretical annihilation of reality leaves the everyday world untouched. Common sense is all you need for that, and no one need confuse themselves wondering how other scientific results connect together.

Ideally, I would hope that our common sense would connect to our scientific picture of reality in some coherent way. Otherwise, all that “common sense” does is punt the question of reliability of evidence.

Near as I can tell, scientists like the common man are encouraged to take things at face value, just so long as they don’t get uppity.

Not sure what you mean here. Encouraged by whom? “Face value” is a tough thing to determine when you’re talking about complicated observations that are mediated by layers of technology. What is a “face value” reading of an fMRI scan or an observation of extragalactic structure? I don’t think there’s a really sensible way to apply that term in many cases.

But that’s me. Every time a scientist says all scientific claims are provisional is denying that facts that have been long corrected are still not reliable descriptions of reality, but temporary assertions awaiting replacement by something better, is basically denying that science finds out anything about reality.

I think this is a most ungenerous interpretation of what scientists mean when they say that. “Provisional” is merely an admission that we could be wrong about something; it’s not an admission that we don’t find out anything about reality. I don’t know for a fact, but am willing to bet that if you talked to most physicists, you would find that they subscribe to just the kind of realist position you’re advocating. They merely acknowledge that the reality might be different from the currently accepted picture, which is healthy and normal. Especially when you get into the weeds of dealing with very complex systems to say that a result is “provisional” is to point out that we might discover that mechanisms whose individual operations we have characterized very well might exhibit collective behavior that is unexpected from a “naive” analysis of individual components.

And it isn’t helped by pretending that they can label some claims as having a negligible probability of refutation. If they really thought that, they would actually neglect the possibility of refutation and forget “provisional!” If they insist that they must say “There is probably no such thing as magic,” instead of “There is no magic,” the only point is to deny that science can achieve a description of reality.

I’m pretty comfortable saying “there’s no magic” and I’m sure most physicists are as well. Again: really ungenerous reading. Most scientists do not deny that science can achieve a description of reality; their work is typically predicated on that assumption.

The very widespread fetish for Bayesian statistics seems to mean a commitment to subjective probability, as opposed to outmoded frequentism. I don’t see how that’s realism either. Similarly, I don’t see how Copenhagen is realist and I’m pretty sure that’s the dominant view by far.

There’s nothing in Bayesian inference that commits anyone to anti-realism. Bayesian inference is useful because it gives you a way of expressing conditional probabilities and a nice update rule that lets you bootstrap yourself to better models. It’s not some nefarious plot to undermine reality. Neither is Copenhagen; in fact, Copenhagen punts on the very question, saying, in Feynman’s words, “shut up and calculate!” I think the most constructive way to view the Copenhagen interpretation is as a kind of minimal baseline: you can have all sorts of divergent views about the ultimate reality, but whatever they are, they’d better accord with experiment at least as well as QM does.

89

J Thomas 10.16.14 at 5:37 pm

If you want to do science you probably want results that will generalize. Otherwise what’s the point? You might as well study history and speculate about the causes of some revolution or another.

One PhD thesis got a stir in the popular media because he did electron microscopy and photographed one single sand grain from all directions. The laymen seemed to think this was worthless, but some scientists thought it would be good for something. I tend to think his methodology might be useful but the data about his sand grain not so much, until somebody gets at least two more sand grains from the same beach and maybe three from another beach.

So you want to be able to abstract out stuff that’s repeatable from your observations.

When I’ve seen science that worked, it started out with observations and ideas about the observations, and then maybe rough experiments that showed something, and then maybe more focused experiments that got precise results given the background of knowledge.

Like, I spent 3 months in a biochemistry lab. They were looking at a membrane enzyme found in hog liver. The enzyme degraded pretty fast, but they had a crude test of its function. They took an unpurified slurry of ground up hog liver and added a radioactive sample of the protein’s known reactant. After a time, they added a solvent and shook it up a lot, then they centrifuged it and got 3 phases — sediment, water, and solvent. They measured how much radioactivity there was in the solvent versus the water, to get a sense how much active enzyme there was. And they added various things that might keep the enzyme from degrading. It turned out that glutathione worked well, so they tried adding a lot of glutathione to the hog liver before they ground it up. Suddenly they had thousands as much enzyme activity, and it was much easier to study. The crude test was perfectly adequate to get that result.

I spent 6 months in a lab that studied yeast. They had a grant to study a cancer drug. It was a carcinogenic cancer drug, it intercalated in DNA some. The yeast very often had mitochondrial mutants that could not use oxygen and so grew slowly. They had found that when they suspended yeast in distilled water that contained high concentrations of the drug, the surviving yeast were 50 times as likely to have those mutations. Since it was easier to study DNA damage in yeast, they then got the grant.

But it turned out that the drug did not create lots of mitochondrial mutants. What happened was that the mutants were much more resistant to being killed by the drug. Once they found that, they grew lots of yeast and made flasks of partly-purified yeast mitochondria. They added the drug to them, and did chromatography to find a variety of new fluorescent compounds that weren’t in the mitochondria or the drug beforehand. Functioning mitochondria changed the drug into something even more toxic.

Simple qualitative studies that had lots of room for error, gave results that might potentially be useful for medicine. The big problem with the drug for treating cancer wasn’t that it made a few mutations, but that cells that metabolized fast used it to kill themselves. That’s why it was so bad for heart muscle and immune-system cells. I never heard that they actually found a way to make the drug more useful, but at least they started looking at the real problem.

Scientists start with simple models and simple vague realities, and they ring complications on them once they have simple models that they don’t just reject. It’s hard to draw the line between vague observation and real science, because it isn’t anything like a firm line.

I think what ought to make it science, is that you look for ways to test your ideas in case they’re wrong. Looking for tests to make it wrong is central, and that’s hard to do in practice. I again want to recommend the game Eleusis which demonstrates this and provides concentrated practice.

http://en.wikipedia.org/wiki/Eleusis_%28card_game%29

90

Barry 10.16.14 at 5:53 pm

Jerry Vinokurov 10.16.14 at 3:09 pm

“Barry, I don’t know what you think your incredulity proves. I’m just telling you that unless you don’t think that, say, astronomy is an experimental science, “

I think that it’s overwhelmingly a non-experimental science. I

“you’re going to have to be prepared to not put that much weight on the difference between “experiment” and “observation.” There just isn’t that much difference between them in actual scientific practice;”

Wrong. An experiment gives much more internal validity.

“… an experiment is just a setup in which you control for certain factors, nothing more. “

A small matter-transmutation even over your city, nothing more.

“Conversely, systematic observations just are the things that make up a good experiment.”

No, they can approach a good experiment (and sometimes carry far more external validity).

91

John David Stanway 10.16.14 at 5:54 pm

So, what’s your second favourite Hume quote?

92

Jerry Vinokurov 10.16.14 at 6:15 pm

I think that it’s overwhelmingly a non-experimental science.

In that case, I suggest you conduct the following experiment. Go down to your local university’s astronomy department and ask whether they consider themselves experimentalists. I predict they will say yes.

How would you account for that? You could argue that astronomers don’t understand the meaning of that word, but it seems pretty unlikely that a bunch of people with advanced degrees would be flummoxed by a word they use all the time. You could say that astronomers are just wrong about whether or not they’re experimentalists, but then you’d have to explain why your meaning of the term should be privileged over the meaning imposed by those who actually do the experiments.

My preferred explanation (as a former astronomer) is that the definition of “experiment” is simply more expansive than you think. Astronomers do experiments, they just don’t do them the same way that biologists or solid state physicists do them.

Wrong. An experiment gives much more internal validity.

Internal validity how? What is the metric of internal validity?

Obviously, I don’t mean that just observing a bunch of random stuff and writing it down constitutes an “experiment” (though it can be useful in itself); there’s a structure to these things and the structure is important. I just don’t see how anything terribly crucial hangs on the distinction. Imagine that I concede your point about astronomy being “observational” rather than “experimental.” What then? Are you all of a sudden going to change your mind about the validity of general relativity or the theory of inflation? I suspect not, so why does it matter so much which word you use?

I think that characterizing an experiment as a structured set of observations provides a wide enough area for all sciences to fit in to. The alternative leads to what are, in my view, pretty absurd conclusions that geology and astrophysics and other similar enterprises are not “experimental,” when clearly experiments are very much at the heart of these disciplines.

93

Barry 10.16.14 at 6:19 pm

Jerry:

“Imagine that I concede your point about astronomy being “observational” rather than “experimental.” What then? Are you all of a sudden going to change your mind about the validity of general relativity or the theory of inflation? I suspect not, so why does it matter so much which word you use?”

You seem to have some odd idea, that I believe that ‘non-experimental’ means ‘non-scientific’ or ‘invalid’, or something like that. Since I’ve been arguing against that, I’m not sure where you got that idea.

“I suspect not, so why does it matter so much which word you use?”

Because ‘experimental’ and ‘observational’ are not the same thing. Note that we have different words for that, even in technical fields.

94

Barry 10.16.14 at 6:23 pm

Jerry: “Internal validity how? What is the metric of internal validity?”

Go educate yourself, since it’s pretty clear that you don’t know what you are talking about:

http://en.wikipedia.org/wiki/Internal_validity

http://www.socialresearchmethods.net/kb/intval.php

95

The Temporary Name 10.16.14 at 6:54 pm

Barry: what are you arguing? Can you boil it down? Is it that “experiment” and “observation” are two different words/ideas both encompassed within “scientific study”? Is it about some kind of difference between “active” and “passive” science?

96

Luke 10.16.14 at 7:50 pm

I feel compelled to further muddy the waters by pointing out that Thucydides is the man who swears that everything he reports is exactly what happened, except for the bits he just made up. So, yes, T. is absolutely the model for modern *Wissenschaft* — a snide, ostensibly neutral tone of authrority doing its best to make us forget that we are reading the opinions and misrepresentations of a grumpy aristocrat.

97

Mdc 10.16.14 at 7:58 pm

“the inadequacy of the antirealist, predictivist notions of science.”

“I feign no hypotheses”
– Isaac Newton

98

Jerry Vinokurov 10.16.14 at 8:00 pm

Internal validity is not a term I’ve encountered in physics; it appears to be a term of art in the social sciences. Of course some of the points translate to physics, but there’s nothing about astronomy that makes it particularly susceptible to violations of internal validity, no more so than any other field of physics.

99

Jim Harrison 10.16.14 at 8:02 pm

@Neville Morely #77.

“the mid-C18 is precisely when a new awareness of historical difference is emerging, so exemplarity no longer works because the past is nothing like the present, so instead you switch into the claim that there are universal principles or attributes that can be discerned even in this past.”

I have a bit of a problem seeing that the assertion that there are universal principles of history is a fundamental challenge to exemplarity. I appreciate that the then prevailing taste preferred principles to cases—it was the Golden Age of capitalized abstractions—but the past remains perfectly legible to a historian like Hume. Forty years ago, I actually read his History of England and don’t recall him making any special effort to recreate the world view of the Anglo-Saxon kings. The past portrayed in those interminable volumes is very like the present since the squabbling princes and barons display the same emotions and confront the same political difficulties reign after reign—that’s what makes ’em interminable, though Hume’s amazingly polished style eases the reader’s journey through more than a thousand years of one damned thing after another. Circumstances change, obviously, but Man does not. Which is why it continues to be possible to learn lessons from the historical record, albeit with the help of theory of human nature to guard against a too literal application of old cases to new situations.

Hume seems very far away from 19th Century historicism in all this. What am I missing?

100

Barry 10.16.14 at 8:02 pm

“Internal validity is not a term I’ve encountered in physics.”

That would explain a lot, both that you’ve not encountered it before, and that you’re a physicist.

There was a Chicago economics professor making a joke, that if you are a good economist, you’ll be reborn as a physicist, but if you are bad, you’ll be reborn as a sociologist.

The point was not what people might assume at first glance.

101

stevenjohnson 10.16.14 at 8:05 pm

Jerry Vinokurov@88
I got the impression that Bayesian inference was committed to subjective probability from Nate Silver and Lubos Motl. Too small a sample, my fault.

I am much reassured that physicists and other working scientists aren’t still tangled up in Humean skepticism. Thanks for the happy thoughts.

bianca steele @87

No, sorry, I really do think the corrected information is the science. I know that science then blends in principle with activities performed by everyday people. But that’s the way it was historically. Today professional scientists compare to ordinary people as NBA players compare to an after work pickup game, so there’ll be no problem distinguishing the scientists. Science as a body of knowledge is obviously used by all sorts of people, engineers and inventors included. The kind of error correction I had in mind were errors in describing reality, such as counting forty eight chromosomes in human beings.
And I’m afraid I still think that trying to define science as a method has produced more confusion than enlightenment.

102

Thornton Hall 10.16.14 at 8:26 pm

@79 Scott Martens 10.16.14 at 10:05 am

Why does everyone forget Tycho Brahe?

Kepler was not building on Copernicus (who was, as you point out, less accurate than Ptolemy), he was building on a massive pile of observations put together by his teacher Brahe. The Copernican revolution didn’t require reams of data, but figuring out that orbits were elliptical, did.

Therefore, everything in physics from Kepler to Newton to Einstein required the massive pile of data created by Brahe. He’s the giant whose shoulders Newton was standing on.

I think it’s interesting that one of the main economics bloggers pointing out the non-empirical nature of people like Smith and the need for massive piles of data (rather than “empirics”, whatever that is) is Lars Syll: maybe it’s something about Scandinavia?

103

The Temporary Name 10.16.14 at 8:35 pm

An interview with a scientist publishing a retraction of his results.

http://www.cbc.ca/player/Radio/As+It+Happens/ID/2555667461/

104

Jerry Vinokurov 10.16.14 at 8:46 pm

Just to clarify: there exist subjective interpretations of Bayesianism, but there’s nothing (that I know of) about Bayesianism per se that commits one to any sort of subjectivity about probabilities. Andrew Gelman has a pretty nice discussion of Bayesian inference in science here (PDF).

105

Sebastian H 10.16.14 at 9:21 pm

I’m not sure I understand the observation/experiment cleavage here.

Generally science is about observing things in the attempt to accurately describe how things work. Experiments are a subset of observations. They are observations where you attempt to limit the inputs so you can prove a theory about why the things you have (hopefully) accurately described act the way they do. Experiments aren’t necessary for science (see astronomy) but they make it easier to isolate things.

The key to science is accurate predictions based either on observed correlations, or later on theories about how things work. Economics (macro) invokes a lot of the jargon of science, but I’m not sure it really is science at least at this point. All of the social ‘sciences’ have the problem that humans have lots of variables that are hard to sort out and isolate. Economics has additional problems because it has to deal with abstract things like ‘money’ that are also changed by out of control human variables.

But macro economics seems to currently lack anything like good predictions. It seems to be a few steps above rain dancers–if you dance long enough the rain eventually comes. But that doesn’t mean you predicted it or caused it in any useful way. All the conflicting opinions on Japan, the EU, the USA and China suggest that macro economics just isn’t there yet.

106

William Berry 10.16.14 at 10:29 pm

Down with JV on the supposed dichotomy of “observation” and “experiment”. The distinction is more linguistic and abstract than practical.

The real divide between scientists, as I understand it, is between math-oriented theoreticians and observer-experimentalists. Even here, the divide is not so pronounnced as it might seem at first glance. E.g., they literally find common ground in the break-rooms and conference rooms at the LHC.

107

William Berry 10.16.14 at 10:52 pm

Also, too: astronomy considered as the old-fashioned business of charting the cosmos might well be primarily “observational” as opposed to “experimental” (not so much, really, when you consider the technological sophistication of some of the instrumentation involved). But that’s a small part of the role of astronomical observation nowadays.

The main business is astro-physics, not “astronomy” per se. The comparison of esperimental results from particle physics with the observation and analysis of cosmological phenomena is as “experimental” as you can get.

Astronomy and particle physics also meet at the LHC.

And kudos to JV for consistently making the most sense in this thread (IMNSHO : ) ).

108

Collin Street 10.16.14 at 11:05 pm

One PhD thesis got a stir in the popular media because he did electron microscopy and photographed one single sand grain from all directions.

But of course.

I remember reading something… how to explain, it said that we used to think that there were different types of mitochondria, laminar and round, but when we took multiple slices through the same cell it turned out that there was only one sort and it all depended on the cutting angle.

109

J Thomas 10.16.14 at 11:46 pm

#98 JV

Internal validity is not a term I’ve encountered in physics; it appears to be a term of art in the social sciences.

It’s about a lot of ways you can fool yourself, and how to avoid them.

Like, I talked to a woman who was doing molecular biology, who hated to work with abstract theory. She wanted to actually see the truth. So she went into electron microscopy. Under some circumstances she could see the paths that DNA was laid out in, and study how it folded back on itself. But she found that when she looked at many thousands of examples, she could find anything she looked for. Interpreting pictures of DNA is sometimes tricky. So she found herself doing complicated statistics. She would look for the pictures she would expect to get from two competing theories, and count them, and she worried that the abstract statistical theory she depended on might somehow fail her.

Physicists mostly don’t have that problem. Their computerized machinery can run through billions of examples easily. And the number of possible outcomes is strictly limited. So for example running the LHC involved throwing away billions of samples that could be assumed uninteresting, and looking only at the ones that might represent a Higgs particle. The standard model had predicted everything that could happen and everything that could not happen. The Higgs particle could not happen, so if it was observed more often than statistical chance would predict that false observations would happen, then it was real. It was the only anomaly that was worth looking for, so any other anomalies could be discarded as uninteresting. Anyway there was no reason to expect any other anomalies. The theory that predicted how many false positives to expect could not be wrong because the situation was so simple that there was no room for error. Even though it was new equipment running in an unprecedented energy range, there could not be calibration errors. The predicted false-positive rate could not possibly be so far off as to get the observed rate.

Physics simply has fewer unknowns and far more precise measurements, to the point that there is simply no room to fool yourself, and you never have to worry about that sort of thing.

Most other sciences deal with more complicated topics, correspondingly more complicated theories, and more difficult measurements. So it’s easy for bias to creep in to them, which is of course impossible in physics.

110

J Thomas 10.17.14 at 12:00 am

#108 Collin Street

I remember reading something… how to explain, it said that we used to think that there were different types of mitochondria, laminar and round, but when we took multiple slices through the same cell it turned out that there was only one sort and it all depended on the cutting angle.

Yes, I think they are right and also I am skeptical.

If mitochondria have one particular shape then slicing them different ways will give you multiple pictures. But there are various other shapes that mitochondria could have that may give the same pictures. Maybe in some real cells the mitochondria form some complex network, that can be sliced to look like round or oval or stringy. And of course many of the things people do to cells while trying to get pictures of them can disrupt the original structure. The older idea was inadequate, but the simplest idea compatible with the slices might not be right either.

http://www.med.unipg.it/imagelab/mitoch.html

111

Bruce Wilder 10.17.14 at 12:08 am

. . . bias . . . which is of course impossible in physics.

LOL

112

Bruce Wilder 10.17.14 at 12:40 am

I do think it a bit odd that this conversation took so long to get from “observation” to measurement. And, it also seems — though perhaps I’m wrong about this — that there’s a reluctance to place theory at the center of science.

113

ZM 10.17.14 at 1:00 am

Perhaps since no one agrees on what is science and what is not science the term science is rather useless. I am not sure why there was such a vogue for using this very undefined term, especially by people in disciplines where they are meant to come up with great elaborate categories of things. You would think such people could categorise what science is , and what is not science easily enough, since they do it with elements and animals and plants etc continually – but they have not.

Therefore I suggest it is best to give away altogether this vogue for calling things science (likely just to lend an air of authority since no one can define science properly) – and go back to natural philosophy, natural history, and experimental philosophy and so on, which are much firmer categories than the word science is.

114

temp 10.17.14 at 1:59 am

Repeated observation of unelicited events plays very little role in many modern sciences and is positively misinformative in areas as fundamentally important as medicine.

I don’t think either of these is true. Case-control studies and longitudinal studies consist of “repeated observation of unelicited events” and both are extremely important to modern medical science.

Repeated observation is the typical way of doing science. Controlled experiment is what you do when you get really lucky and happen to have a convenient study system at hand.

115

phenomenal cat 10.17.14 at 2:18 am

Therefore I suggest it is best to give away altogether this vogue for calling things science (likely just to lend an air of authority since no one can define science properly) – and go back to natural philosophy, natural history, and experimental philosophy and so on, which are much firmer categories than the word science is. –ZM @113

That’s fine, but then you wind up undermining modernity’s faith in itself that it really was on to something historically unique. Capital S science was supposed to be the path that would lead us forevermore out of the land of superstition, dumb tradition, and mere belief.

I do think it a bit odd that this conversation took so long to get from “observation” to measurement. And, it also seems — though perhaps I’m wrong about this — that there’s a reluctance to place theory at the center of science. –BW@112

You’re probably got something here, especially if you replace the word *theory* with with another word that better describes what the reluctance is really about: metaphysics. To speak of the metaphysics at the beating heart of all scientific production is to speak of that which cannot be spoken of.

116

temp 10.17.14 at 2:22 am

Even subjective interpretations of Bayesianism do not imply either anti-realism or skepticism. Anyways, if there’s a “very widespread fetish for Bayesianism”, you wouldn’t know it by reading high-impact scientific journals, where nearly everyone is still using t-tests and chi-squared. Maybe in certain circles of the internet.

117

ZM 10.17.14 at 3:06 am

phenomenal cat,

“That’s fine, but then you wind up undermining modernity’s faith in itself that it really was on to something historically unique. Capital S science was supposed to be the path that would lead us forevermore out of the land of superstition, dumb tradition, and mere belief.”

You say that like modernism is still our period. Post-modernism started before I was born – so modernism is just a past era like the Renaissance now.

I always wondered what made us post-modern. Some said ‘the literature is so reflexive’ this is not correct, lots of literature is not reflexive and Shakespeare and Jane Austen and others were reflexive ages before post-modern reflexive writers.

Others said – ‘we don’t believe in grand narratives’. So I think back – and I don’t find any grand narratives in king Lear or the Canterbury tales either. Moderns were just very easily taken in I feel by the people who made up grand narratives. It is quite funny that they were taken in in this fashion out of their need to feel superior to past people or other cultures . I am sure there must be a saying that epitomizes this sort of gulliblility based or wanting to feel superior, but I can’t think of one.

118

L.M. Dorsey 10.17.14 at 3:08 am

@82 stevenjohnson
Hrm. I don’t recognize the your Popper. As I remember him, tho, falsifiability, not predictability was his preferred device for sifting the science from the not. Which, yes, makes Freud and Marx something else. But that something else is not nothing (which is not to declare it “literature”). Economics, tho. That’s a different kettle of eels altogether.

And maybe it’s the head cold, but to my ear, your second paragraph sounds like it could be a grab from this chestnut, so perhaps your “minority view” really isn’t:

The history of science, like the history of all human ideas, is a history of irresponsible dreams, of obstinacy, and of error. But science is one of the very few human activities — perhaps the only one — in which errors are systematically criticized and fairly often, in time, corrected. This is why we can say that, in science, we often learn from our mistakes, and why we can speak clearly and sensibly about making progress there.

119

phenomenal cat 10.17.14 at 3:30 am

ZM,

I hear what you’re saying, but POMO is a non sequitur in this context. It’s a term that comes out of art history departments in the 60’s, gets picked up by architecture departments in the 70’s, moves into lit. departments in the 80’s and achieves mass consciousness in the 90’s.

The topic under consideration is science and you better believe the issue of a “grand narrative” is a live one here; else how do you explain the inability of all these very intelligent people (on this blog and over the last 4 centuries) to reach a real consensus on what science is and how to do it?

120

ZM 10.17.14 at 4:01 am

phenomenal cat,

“how do you explain the inability of all these very intelligent people (on this blog and over the last 4 centuries) to reach a real consensus on what science is and how to do it?”

that people cannot agree on what science is shows quite well enough that they cannot find a grand narrative about science to grasp at without it slipping through their fingers like a wisp of smoke.

Not being able to clearly define what is and what is not science would be a poor way to demonstrate we are still in the modernism era (unless you are thinking of high modernism’s absurdism writings?)

121

stevenjohnson 10.17.14 at 4:20 am

temp @116

Quantum Bayesianism seems to connect antirealism and subjective probability and Bayes. But as you say probably due to the availability of the internet and the unavailability of scientific journals.

L.M. Dorsey @118

Predictivism, not predictability, the notion that science is making a prediction, then carrying out an experiment to falsify the prediction. It’s not even science if you can’t make a prediction then carry out the experiment. And yes, Popper had trouble with declaring Darwin to be scientific. But it’s always nice to find an admirer of Popper. Perhaps a quote from The Open Society and its Enemies? To really understand a writer, I always find it specially helpful to look at their first book.

122

phenomenal cat 10.17.14 at 5:34 am

that people cannot agree on what science is shows quite well enough that they cannot find a grand narrative about science to grasp at without it slipping through their fingers like a wisp of smoke. –ZM@120

I interpret the significance of the above almost exactly opposite to your intended meaning. Grand narratives are not set, given, or routinely agreed upon. They are grand insofar as they are severely and hotly disputed.

I’ll give you an analogy: Europe’s best minds spent no less than 10 centuries trying to nail down the basic mechanics of just how the Father, the Son, and the Holy Ghost fit together and what the hell that meant to the average schmuck on the street.

That’s a grand narrative; it was the explanation of the universe for learned and not-so learned Europeans for a very long time.

Now, pray tell, what is that we rely upon to explain the universe to us? What are the learned people in this forum discussing and arguing exactly? What exactly was it that kicked the Trinity to the curb and got the modern age underway?

So long as the nature, purpose, and practice of science is a hot topic you can be reasonably sure that modernity has not died off. When science no longer matters very much then we’ll have taken leave of being modern.

123

Harold 10.17.14 at 5:51 am

@122 “I’ll give you an analogy: Europe’s best minds spent no less than 10 centuries trying to nail down the basic mechanics of just how the Father, the Son, and the Holy Ghost fit together and what the hell that meant to the average schmuck on the street.”

No, they didn’t. They figured it all out in 325 AD and have not changed it since. That’s why it is called dogma, not science.

124

ZM 10.17.14 at 6:01 am

phenomenal cat,

But normal various sorts of Christian theologies are not grand narratives. If you want an example of a grand narrative about Christianity an example would be Hegel’s idea of god . Thomas Acquinas’ ideas don’t count as grand narratives because he was medieval in his thinking not modern.

Grand narratives are sorcific to modern ideas like Marx or the man who wrote that book to sum up all folklore The Golden Bough.

Nowadays only disreputable people peddle grand narratives like this, because reputable people have to cite all their sources. Grand narratives take up too many footnotes and it’s too great a scholarly task to know everything about everything – just remember the trouble David Graeber got in.

What people are discussing is what counts as science and what does not count as science. No one agrees. I brought up science only means knowledge , but people want to define it differently to privilege their own areas and exclude others.

We are now in post-modernism, this is in historical time. The other one is the anthropocene – for if you’re talking about geological time.

You say we have “science” still so we are modern – but they had “science” in Ancient Greece I notice and nobody but nobody says modernism includes the classical age.

125

reason 10.17.14 at 8:09 am

Isn’t the problem when people talk about science, that they want it to be a thing, but in fact it is a process. Science never “arrives”, but it “travels”. That is where the Chicago School went wrong, they think they “know” things.

126

Ronan(rf) 10.17.14 at 10:44 am

I’m aghast that people pay any attention to P.E. Gobry, the embodiment of that most peculiar modern creature, following in the inglorious tradition of Marc Andreessen et al, the ‘entrepeneur intellectual’. One wouldnt give the average barroom ranter such creedence, yet the end product is much the same.

127

Ronan(rf) 10.17.14 at 11:08 am

A jeremiad against science by P E Gobry? No thank you very much. I’m flummoxed here. Bewildered and perplexed. What kind of stage in human development are we living through ?

128

Neville Morley 10.17.14 at 11:45 am

@Luke #96: I don’t have much of a problem with that characterisation; I do think there is a lot more to Thucydides than disgruntled aristocratic prejudice, but it’s undoubtedly true that most of the admiration for his critical methodology rests on (i) not paying full attention to what he actually says and (ii) taking it for granted that he actually did what he said he was going to do throughout the rest of his history, despite the absence of any means of testing this.

@Jim #99: I wasn’t presenting a recourse to supposed universal principles as a direct challenge to exemplarity; rather, this is a period in which traditional assumptions about the exemplary nature of the past – Cicero’s historia magistra vitae etc. – are coming under increasing pressure from a growing sense of historical difference and distance, and an explicit shift to identifying universal principles of human behaviour offers a means of continuing to argue that studying the past can be useful. In other words, it’s a new means to the same goal, justifying the idea that history is a source of some sort of wisdom.

@ZM #124: just to be annoying, there actually are various writers in the second half of the nineteenth century who claim that classical Greece was modern, not least because it shows a modern frame of mind in its science and historiography…

129

Barry 10.17.14 at 12:21 pm

JV, tou description is rather wrong on the LHC:

1). There were major statistical problems precisely because a vast amount of data was collected.

2). Running new instruments at new energy levels is not simple, and it’s likely that problems will arise. And when collecting billions of data points, looking for a few anomalies, very rare misbehavior is a major problem. Do you not remember the FTL neutrino?

130

Jerry Vinokurov 10.17.14 at 12:23 pm

JV, tou description is rather wrong on the LHC:

I don’t believe I ever mentioned the LHC? I think you’re responding to J Thomas.

131

Barry 10.17.14 at 12:25 pm

Scott: “Gobry is wrong: Controlled experimentation has nothing to do with many branches of science. Consider astronomy. Smith is wrong too: Repeated observation of unelicited events plays very little role in many modern sciences and is positively misinformative in areas as fundamentally important as medicine. And the discovery that the Earth turns, and Ptolemaeus was wrong, did not involve a single controlled experiment whatsoever. Nor did it involve using many observations to devise better predictive models. Copernicus’ actual figures were worse than the state-of-the-art geocentric model at the time. His main line of argument was that he could explain the retrograde motions of the planets in a more intuitive way.”

As Temporary Name pointed out, this is simply not true. Now, it’s because you can’t get controlled experiments in many situations, but that’s what life – and science – has to deal with.

BTW, go ask experts about how many times controlled experiments have been misleading. That’s why the term ‘external validity’ was coined.

132

Lee A. Arnold 10.17.14 at 12:52 pm

I think science is an extension of our valid perceptions and it simply continues to expand.

Science is also a spectrum of different sciences. The common characteristics are not so easy to describe. Indeed Popper decided that science only includes those statements which are falsifiable — one of those clever hierarchical jumps in epistemology to the absolute negative, which proves, yet again, that there are always hierarchical jumps to the absolute negative.

I like Gregory Bateson’s description of science in his book Mind and Nature: Science is the mapping of a “tautology” onto a “description”, to give an “explanation”. However, it cannot ultimately prove anything, though it can disprove hypotheses (the Popperian jump).

Bateson didn’t put those into quotation marks, but I do here, because “tautology, description, explanation” are not so simple. Each is a category that includes different things, depending upon the specific science, and within that specific science, they take their positions relative to each other.

The tighter the tautologies, and the tighter the mapping onto the descriptions, then the “harder”, the more secure, we feel that the explanation is. Physics uses math, and the mapping is performed by measuring, and we have found that the explanations work for additional descriptions — descriptions that are in forms of both other observations and new experiments. So quantum physics is better than Bode’s Law, because it extends further.

For sciences in general, how is this done? How do we map tautologies onto descriptions to give explanations? Well, we point to what the science is, then we point to how you, the new student, are to perform it. It is all ostensive. In this sense, all of the sciences are a Wittgensteinian language-game.

In some sciences — geology, evolution — the experiments are already done for us, and come as descriptions of other locations, other clades.

Social sciences are the most difficult, because our observations are perceptions of others’ perceptions, or are otherwise close to our hearts, and on the other hand, experimentation needs isolation diachronically and synchonically (and ideally, I would think, without the subjects knowing they are in an experiment), which are often impossible. We look for mathematical anchors for the descriptions — e.g. social statistics, or cliometrics, or quantities of money, supply and demand, etc. — and then promote these to the forefront of the science, even forming a priori tautologies of sometimes dubious merit.

We used to believe that physics is the most basic science (and many of us still do believe it), and we used to believe that social science would be reducible into biology, then biology into chemistry, and chemistry into physics.

I think it was Whitehead who said, “Murder is the prerequisite for the absorption of biology into physics.”

Because appears that new phenomena are emergent at each level, and some sciences remain distinct.

Of course some other sciences are close to each other in the hierarchy of our perceptions. We are able to find some basis for chemistry in physics and so on. If one contradicts the other, we are compelled to think that one of them is wrong, and that a new explanation is forthcoming.

It could be that the whole of science is a moving fan of perception that expands out into a spacetime, a fan that is a spectrum of different sciences which we must find to be coherent, or else our perception is incoherent. But it could be that there is no absolute — that there is no bottom discovery to physics, it just keeps going. And there is no top discovery to social science, it will just move to examination of the social patterns on alien planets. And we will just continue to map tautologies onto descriptions to find more encompassing explanations, but without an absolute end, because the absolute is always a jump to the negation.

133

dax 10.17.14 at 12:57 pm

A science allows one to make accurate predictions, where it is unambiguous today as to what counts as validation of the prediction. This is a necessary condition, though it is not sufficient. Much of economics fails this condition, so it is not a science.

134

J Thomas 10.17.14 at 1:13 pm

#127 Ronan(rf)

What kind of stage in human development are we living through ?

It depends on your timescale, and also on your location.

In the short run we have a period of confusion. The USA used to take the stand that US consumer culture was the best, and that in theory the rest of the world could and should someday consume like the USA. But now we are running out of various resources so that we can’t maintain even US consumers the way we used to, and individual Americans are franticly trying not to get thrown out of the lifeboat. People argue about what to do about the situation but mostly agree that conservation of resources is a hopeless dead end, that instead we must find ways to continue to consume more.

On a slightly longer timescale we are at the end of a generation of hope that science would solve our problems. The scientism dogma held that superstition kept us chained to methods that didn’t work very well, but that science would bring us technology that would transform the world and also transform our society into something new and wonderful. This was proselytized particularly by science fiction. But as it turned out, science was our secret weapon against the communists, so the best science needed to be kept top secret and is reserved for military uses where cost does not matter. Some important people worried that if society changed they may not still be important, and they wanted to restrict new technology until they were sure it would not hurt them. Computers and the internet were allowed out though, and they have had big effects on society. Anyway there is reason to think that technology cannot be restrained, that it always gets re-invented by clever individuals who market it well, if a market exists. The reason most of the new technology has not been marketed, the reason that all the new advances depend on giant industrial processes controlled by giant corporations, is that the rest simply cannot be marketed at an acceptable price. It is the iron laws of economics that prevent us from using new technology, not any conspiracy. Regardless, the dream is dying.

On a longer timescale than that, we are at the end of the Enlightenment. People had been stuck in roles they felt they had to believe in, and then the idea spread that they could observe the real world and create their own roles in a good society, based on reason and logic and their own inner goodness. There was a lot of excitement to that prospect. But in practice we have not been that good, and when we observe the world it does not give us stories that inspire hope and belief. We are of course slow. In 1900 it was still a topic of debate among physicists whether atoms were real or were only a convenient mathematical abstraction. (The consensus was moving toward real.) So OK, we are giant complex mechanisms built out of atoms. Well, but we are conscious! But the best story to date is that consciousness is only a false model you have of yourself. Like a computer program that runs a simplified model of its own operation to help it predict the outcomes of its interactions with other programs. It’s useful, but there’s nothing magic there. We are the mechanisms that have happened to survive out of three billion years of evolution. Mostly nobody gets inspired by that but rich social darwinists.

Enlightenment has had its chance in memespace and it is no longer competitive. Reality sucks and people would rather believe something else.

On a longer timescale, for a long time we have usually managed to increase food production and so our population has increased. By Parkinson’s Law, our organizations have gotten more complex — more people to organize creates more work for them to do. Even 5000 years ago there were places where 10% of the population was not involved in food production. Now in the USA there are about 750,000 agricultural workers (not counting managers who do not actually farm) out of 320 million people. About 75% of them were born in Mexico. Instead of doing what our ancestors did — hunting, gathering, or later farming — we do weird things that society has created for us to do. Work retail. Sit in offices and prepare reports. Adjust insurance claims. etc. Would it be surprising if it takes us a few more generations to adapt?

It depends on your point of view, what stage of human development we are living through.

135

Barry 10.17.14 at 1:14 pm

Jerry: “I don’t believe I ever mentioned the LHC? I think you’re responding to J Thomas.”

Yes, I typoed. Sorry.

BTW, ‘internal validity’ is what physicists do when they conduct runs of an experiment under different conditions, seeking to measure (and ideally control) factors which might make their results inconclusive. For example, calibrating instruments to give the same results (or have known adjustment factors) at different temperatures and over time.

136

Trader Joe 10.17.14 at 1:32 pm

Is there a degree to which, in order to be deemed a scientist, the topic under study needs to be either the natural world (including humans, plants and animals)?

Most investment banks have what we call technical analysts who examine stock patterns using charts (observation), derive algorithms from their observations and then conduct experiments both backword looking and predictive. The goal is simply to use the past as a predictor of the future in making stock trades and/or choosing entry points for investments.

While this profession seems to employ virtually all of the skills the collective wisdom here has applied to “science” and “scientists” (i.e. observation, controlled and external experimentation, data collection, modeling) I doubt anyone, themselves included, would label their work as science nor them as scientists.

137

J Thomas 10.17.14 at 1:49 pm

#128 Barry

JV, tou description is rather wrong on the LHC:

I think you’re responding to me.

1). There were major statistical problems precisely because a vast amount of data was collected.

OK, maybe physicists do need to consider the various ways they can accidentally bias their results.

2). Running new instruments at new energy levels is not simple, and it’s likely that problems will arise. And when collecting billions of data points, looking for a few anomalies, very rare misbehavior is a major problem.

Yes, that certainly makes sense. I am not a physicist so all I have seen is from the popular media and science blogs etc. They tend to say the Higgs boson is definitely real, there are no other anomalies because after all the Standard Model is right apart from not predicting the Higgs Boson, if there were any anomalies it would be big news. And if there were any they would have been found by now. The sort of comments that spectators make. I’m sure the actual physicists do better.

Do you not remember the FTL neutrino?

Yes. They got a result that did not fit theory, so there was a tremendous effort to find their mistake. Eventually something was discovered which could have explained it, so then everybody quit.

Of course it deserved that level of effort because it would be so important if it was true. But something about the process would worry me if it was, say, economists doing it. Like, say a whole school of economics got upset about Piketty, and they looked hard for flaws in his work, and each time they thought they found a flaw they got excited but then it turned out not to be sufficient but finally they found one they could point to, to say he was wrong. And then they all relaxed. It would leave me wondering if they were biased. But luckily physics has only one school, the correct one, so that isn’t an issue.

138

ZM 10.17.14 at 1:50 pm

Ronan(rf),
“What kind of stage in human development are we living through ?”

Erlich wrote an article saying we were in the Endarkenment now. This is not a proper historical era, but it is somewhat amusing in a gallows humour sort of way.

139

Bruce Wilder 10.17.14 at 2:52 pm

reason @ 125

I think the problem is that science is, at its core, successful guessing. People, who want to explain the success, are reticent about the true prerequisite, which is curiosity. Most humans are believers and are impressed by the promise of successful, hence powerful belief — they want science to vindicate explanatory belief with proofs — hence the emphasis placed on Popper’s predictions as tests of theory.

It is much harder to get at the difficult balance of understanding and puzzlement, which is science in motion, even for many would-be scientists, who long for fame as the discoverers of some nameable truth.

I thought what Brett Bellmore said about science being hard, because it required letting the evidence speak, and what Anarcissie said in reply about science as a social activity, were very smart.

People, especially people in social groups, want to know things, to share a positive doctrine and purpose, to have answers, to share a faith. Economics isn’t a science, because economists are too busy being civic priests and augurs, legitimizing the political order, to be curious about how the actual economy works or to notice its salient features. It isn’t that economists are prevented from doing controlled experiments — they can, just as they can walk thru a supermarket with their eyes open, but won’t.

It is hard to trade a theoretical doctrine for a theory of puzzles. And, hard to keep a theory of questions from ossifying into a system of answers.

People had invented the social roles of the knowledgeable authority, the wizard, priest or physician, long before there was much of genuine knowledge for them to know. Kepler’s day job was as a court astrologer, and he struggled to accept the observable reality that that the motion of the planets described elliptical orbits, because it violated his neo-platonist prejudices. Jenner struggled for years to elevate the folk knowledge that infection with cowpox conferred immunity to smallpox, and marry it to the semi-accepted practice of prophylactic inoculation with smallpox, to create vaccination. He had to overcome both the social resistance in the hierarchy of medical science, and his own propensity to get distracted with spurious notions, like a supposed relation with horse pox.

People want science to be identical with propositions of sure knowledge, when it would be accurate to identify it with puzzlement, which often requires shedding the notion one already knows the answer, an answer approved by a high-status social group. Academic science makes its own use of social processes, with sometimes mixed results.

140

Barry 10.17.14 at 3:49 pm

J Thomas:

“OK, maybe physicists do need to consider the various ways they can accidentally bias their results.”

I agree. And remember, these problems can pop up at any time; it’s just that competent physicists have a good handle on how to prevent them, and do the thinking, planning and work to prevent them and detect them. When they fail in any one of those tasks, they to to GIGO land. I’m sure that Jerry, as a practicing physicist, could provide hours of tales from his and his colleagues’ experience.

141

Barry 10.17.14 at 3:50 pm

Trader Joe:

“While this profession seems to employ virtually all of the skills the collective wisdom here has applied to “science” and “scientists” (i.e. observation, controlled and external experimentation, data collection, modeling) I doubt anyone, themselves included, would label their work as science nor them as scientists.”

Read ‘Zen and the Art of Motorcycle Maintenance’.

142

The Temporary Name 10.17.14 at 3:56 pm

As Temporary Name pointed out, this is simply not true.

Misattribution I think: not sure I’ve done any refutation of anything. I would, though, like to defend “science” as a useful two-syllable word, verb and noun. Dictionary definitions work well for people less interested in splitting hairs.

143

Trader Joe 10.17.14 at 4:13 pm

@140 Barry

I’ve actually always meant to read that book and have never quite gotten it to the top of my reading list…thanks for the additional push to do so.

144

Barry 10.17.14 at 5:50 pm

It’s worth reading. He covers the question of ‘what is science?’.

145

Peter Erwin 10.17.14 at 5:58 pm

Scott Martens @79

Nor did it involve using many observations to devise better predictive models. Copernicus’ actual figures were worse than the state-of-the-art geocentric model at the time.

To amplify what Thornton Hall said:

The acceptance of the heliocentric model did involve using a great many observations to test the predictions of different theories. These included, among other things, the careful, higher-precision observations of Tycho Brahe and his assistants, the observations of the phases of Venus by Galileo (and the moons of Jupiter, which demonstrated that at least some celestial objects did not orbit the Earth directly), and the observations of the position of the Sun over time made using projection devices built into 17th Century Italian churches (see Heilbron’s The Sun in the Church). And, of course, Kepler used the observations of Brahe to devise a better predictive model in the form of elliptical orbits and what are now called Kepler’s Laws of planetary motion. The fact that Kepler’s modification of Copernicus’ model significantly outperformed both Copernicus and Ptolemy in terms of observations matching its prediction was actually pretty important.

146

Bruce Wilder 10.17.14 at 6:05 pm

And, led, via the apocryphal falling apple, to Newton’s analytic insight concerning the force of gravity as a mechanism.

147

phenomenal cat 10.17.14 at 7:34 pm

No, they didn’t. They figured it all out in 325 AD and have not changed it since. That’s why it is called dogma, not science. –Harold@123

Oh, really? Did you ever hear of a guy named Luther? Or, you know, the whole of medieval scholastic philosophy? What was it that Descartes was writing against and seeking to overcome with his demonstration of *method*?

ZM@124 Disreputable people, huh? Like whom, for example? I don’t know, I’m kinda stumped as to who these people of ill-repute would be. Anyway, it’s hardly a matter of one hero knowing everything and committing to the herculean task of citing ALL their SOURCES. These narratives are produced socially over very long periods of time by lots of people.

And if science is just knowledge then we can all pack it in and go home. There’s certainly nothing more to talk about. Science is the same as what some indigene in the bush does and knows–is that right? That’s definitely an interesting question.

Anthropocene works as an epoch descriptor, I suppose; POMO not so much. Yes, the Greeks did some science-y stuff. In that way they were kind of modern, don’t you think?

148

Harold 10.17.14 at 8:06 pm

If you are speaking about the doctrine of the trinity, Luther said we cannot hope to understand it, we must have faith. http://pastorstrey.wordpress.com/2010/05/27/luther-john-16-trinity/

The scholastics did discuss the nature of matter and spirit, the heavens being composed of the latter, in their opinion. This is why the spots on the moon, and even more on the sun (observed in 1612 by Galileo ), were so contentious, since spirit, of which light was a manifestation, was supposed to be too pure to have spots. Dante explained the lunar sphere as being the abode of nuns who had imperfectly fulfilled their vows. http://www.foliamagazine.it/guarda-che-luna/

I will concede to you that the scholastics did talk about what the soul was made of, but I am not sure that this was a topic of intense study for ten centuries, but rather characterized the late Medieval and early Renaissance period. http://www.artchive.com/viewer/z.html

149

mdc 10.17.14 at 8:12 pm

Theology is the queen of the sciences.

150

Harold 10.17.14 at 8:22 pm

:)
Here is the link to the dispute over the sacrament: http://upload.wikimedia.org/wikipedia/commons/4/4f/Disputa_06.jpg

151

Harold 10.17.14 at 8:30 pm

They are arguing “from authority” (not observation) — notice all the books!

152

William Berry 10.17.14 at 9:05 pm

What is science?

Dare I suggest: Science is what real scientists are doing while philosophers and would-be philosophers of science, not to mention blog commenters, debate the question “what is science?”.

153

Thornton Hall 10.17.14 at 9:29 pm

@152 to be fair to the philosophers, a fair number of them would agree that science is “what scientists do.”

More generally, the discussion here is not worthless philosophizing. For example, comments like the following are quite instructive:

“You don’t know about internal validity? Moron.”

Imagine if Jane Goodall or Charles Darwin concerned themselves with “internal validity”? Whatever it is, you won’t find it on an island or in the rainforest simply looking at things and writing down what you see.

154

William Berry 10.17.14 at 9:45 pm

Thornton Hall @153: Agreed.

Thinking of something such as, e.g., Plotnik’s Foundations of Experimental Research, just one case of the vast literature on experimental design in the social sciences, esp. in sociology and psychology, it seems clear that obsession with methodological and epistemological considerations is principally the domain of philosophy and the social sciences.

In the hard sciences, it seems to me that, for the most part, only platonists such as the late J. A. Wheeler seem much concerned with these questions.

155

Barry 10.17.14 at 9:50 pm

Thornton Hall:

““You don’t know about internal validity? Moron.””

I’d appreciate it if you’d not put words in my mouth.

I’d also appreciate if you actually read the thread, and see what I ended up saying to Jerry.

156

Thornton Hall 10.17.14 at 9:51 pm

Will do.

157

Peter Erwin 10.17.14 at 10:30 pm

Barry @ 155:
Thornton Hall was overdoing it, but you certainly seemd to be implying that Jerry was an ignoramus for not knowing the term “internal validity.”

I’d agree with Jerry that “internal validity” is not a term of art in (at least some of) the physical sciences, though some of the individual concepts are certainly known. The references you provided are very much from a social sciences and psychology point of view (all but one of the references in the Wikipdia article are to social sciences or psychological texts).

From the point of view of astronomy, some of the concepts in the Wikipedia article are quite well-known and intensively discussed — e.g., selection bias, for which we have terms for specialized forms, like Malmquist bias and the Scott effect — while others — e.g., maturation, repeated testing, diffusion, compensatory rivalry — are completely irrelevant. (No matter how many times I observe a galaxy, it’s not going to know about it, and its properties aren’t going to change because of it.)

So when you say that “An experiment gives much more internal validity”, I get the impression this may be coming partly from ignorance of how observational sciences work (and what kind of problems they do and don’t have to deal with).

(Yes, I am placing a certain emphasis on observation vs experiment as a sometimes meaningful distinction, though fundamentally Jerry and I are in agreement.)

158

phenomenal cat 10.17.14 at 11:01 pm

Thinking of something such as, e.g., Plotnik’s Foundations of Experimental Research, just one case of the vast literature on experimental design in the social sciences, esp. in sociology and psychology, it seems clear that obsession with methodological and epistemological considerations is principally the domain of philosophy and the social sciences.– William Berry@154

There’s a reason for that you know, and it’s intimately related to the fact the those working in the *hard sciences* don’t seemed much concerned with these questions.

(Hard) science IS the epistemology as well as the methodology. What need would those working in fields that can legitimately claim to be *hard* have to concern themselves with epistemic or methodological questions?

159

William Berry 10.17.14 at 11:19 pm

@158:

Well, yes, of course.

The question then might be: What is the utility of meta-physical/ epistemological/ even (perhaps, some kinds of) methodological considerations?

Not meaning to come off as smart-assy here. I am completely serious. I have very reluctantly come to conclude that most philosophical speculations/ analyses are pretty much epi-phenomenal.

That’s a philosophical positition, too, of course, fwiw.

And I love philosophy, having read voraciously since being captivated by Hume’s Enquiry at eighteen or thereabouts (sixty-three now) and then going on to Nietzsche and others (even spending a good deal of time in the garden of Walden II and its environs).

Philosophy, including philosophy of science, is very interesting.

Unfortunately, it might be that, for the most part, it is only interesting.

160

ZM 10.17.14 at 11:24 pm

phenomenal cat,

ZM@124 Disreputable people, huh? Like whom, for example?

How about the book The Secret “The book begins by introducing and explaining the mechanisms of the law of attraction, then goes on to describe its historical applications and the great men and women in history who are claimed to have harnessed its power. The book describes the law as a magnetic power emitted through one’s thoughts…”

“Anyway, it’s hardly a matter of one hero knowing everything and committing to the herculean task of citing ALL their SOURCES. These narratives are produced socially over very long periods of time by lots of people.”

No grand narratives are not made up over time by lots of people – they are Herculean efforts by one person normally a white man who makes grand claims about all of history or all of people’s nature without citations adequate to justify the claims. Eg Marx and Freud. David Hume, our topic, could maybe be considered to be a prototypical grand narrative figure – for I guess inventing grand narratives seemed the normal thing to pass your time doing if you were always reading Hune and Hobbes and the like.

“And if science is just knowledge then we can all pack it in and go home. There’s certainly nothing more to talk about. Science is the same as what some indigene in the bush does and knows–is that right? That’s definitely an interesting question.”

Well no one will agree as to what science is and is not , so it is a more useless word than the word knowledge. I suggested we can go back to natural philosophy et al. Whether indigenous people pre-colonisation could be said to have practices of knowledge finding we could say we’re equivalent to science, would depend on our agreeing upon science having a set meaning .

“Anthropocene works as an epoch descriptor, I suppose; POMO not so much. Yes, the Greeks did some science-y stuff. In that way they were kind of modern, don’t you think?”

Anthropocene refers to geological time not historical time. Usually the Anthropocene is said to start at the beginning of the industrial revolution – so historically in the modernist period. I read an article about how human activities causing climate change might date back 8,000 years to the start of large scale changes in land use occasioned by people beginning agriculture. So you could start the Anthropocene 8,000 years ago if you wanted to be neat – then all of historical time would fit in the Anthropocene – this would be very neat and fitting indeed.

Po-mo works for the period after modernism from mud 20th c. No one has come up with a better name – I think eventually they will because Pomo sounds very transitional – but first the historical acting needs to take place before you can name the period , so whatever the name of the period is it will depend on how people act now.

No, Ancient Greece is the classical age, or age of antiquities. It is most definitely not part of the modern period.

161

guthrie 10.17.14 at 11:34 pm

Phenominal cat #158 – typing as a chemistry graduate who got no training in philosophy of science or how and why you did experiments when at university, some training in metholodogy and philosophy of science would have been useful. Instead I had to learn it all afterwards, but since I haven’t had a career in scientific research you could say it’s been a bit of a waste. Certainly other scientifically educated people have agreed with me, a better knowledge of these issues would improve their scientific capabilities etc. Which of course is not an endorsement of the post-modern guff upthread.

162

Barry 10.17.14 at 11:42 pm

Peter Erwin 10.17.14 at 10:30 pm
“Barry @ 155:
Thornton Hall was overdoing it, but you certainly seemd to be implying that Jerry was an ignoramus for not knowing the term “internal validity.”

I’d agree with Jerry that “internal validity” is not a term of art in (at least some of) the physical sciences, though some of the individual concepts are certainly known. The references you provided are very much from a social sciences and psychology point of view (all but one of the references in the Wikipdia article are to social sciences or psychological texts).”

The term (judging from you and Jerry and others) don’t seem to be in much use in your sciences, but you do deal with it. As I’ve explained above.

What I’m surprised is that Jerry, as a prime example, seemed to not understand the idea, in the sense that a controlled experiment is – well, controlled. He’s undoubtedly spent many long hours dealing with issues of control (including measurement) to make sure that the comparison of A vs. B wasn’t contaminated by a whole host of things.

“From the point of view of astronomy, some of the concepts in the Wikipedia article are quite well-known and intensively discussed — e.g., selection bias, for which we have terms for specialized forms, like Malmquist bias and the Scott effect — while others — e.g., maturation, repeated testing, diffusion, compensatory rivalry — are completely irrelevant. (No matter how many times I observe a galaxy, it’s not going to know about it, and its properties aren’t going to change because of it.)”

In any science, I’ll bet that you will find different subsets off of a list which basically says, ‘what observational effects do we have to deal with, which wouldn’t matter/matter much less in a controlled, randomized, double-blinded experiment’.

“So when you say that “An experiment gives much more internal validity”, I get the impression this may be coming partly from ignorance of how observational sciences work (and what kind of problems they do and don’t have to deal with).”

I have some experience and training in that, actually. And a lot of what I’ve read or been lectured to on the analysis of observational data is trying to recover internal validity.

163

Thornton Hall 10.18.14 at 2:52 am

@162 Ok… I guess… my reading of the links is that internal validity is a quality that only applies to experiments. Saying that experiments have more internal validity than observations seems to be like saying that baseball players have higher batting averages than basketball players.

The reason I think it is an interesting comment in the first place, is that what makes economists so dangerous is not that they are ignorant about how humans behave in the market (and they are) but that they think they know something about other subjects, when they don’t. Apropos of the OP, Noah Smith does this all the time. He expounds on the morality of the Spanish Empire without the slightest bit of knowledge of the role Jesuits played fighting for the interests of indigenous peoples. He calls an Asian blogger important because:

America is in the middle of a huge, titanic, epochal change. It is changing from a nation of mostly white people (with a few black people) to a fully multiracial nation.

This is in the same post where he claims to like Ta-Nehisi Coates’s blog–which he obviously has never read. (The Confederacy was 40% African American, by the by.)

The confusion when a hard science person is ignorant of “internal validity” demonstrates a false confidence in one’s knowledge of how those sciences work, and is of the same kind that makes Smith’s blogging so infuriating.

Perhaps it is just different terms for the same thing? Perhaps Barry is not an economist?

In any case, it’s the prejudice against just watching how things work that I find interesting and think is very important. Another example is recent Krugman, claiming that every single causal claim in economics carries a model, whether implicit or explicit. So simply watching one market transaction cause an outcome and writing about it is sophistry, in his reading, because it’s hiding the model that is in the background.

Was I not an economic researcher in college when I noticed that my dorm with private bathrooms was much less friendly that older dorms where people had to share bathrooms? Did I need a model of markets to be able to claim a causal connection between more stuff and less well being? (This has recently been replicated: http://nymag.com/scienceofus/2014/10/shared-bathroom-creates-college-friendships.html)

So any model has to deal with that observation, right? Or does my observation never get assumed away in the name of “internal validity” that requires experiments?

And yet, so much great science is built on just that: observations in search of a model. To me, it’s obvious that Piketty’s lasting influence will be the fact that he took the effort to compile so much data. The r and g nonsense should have been left in the original French, where I’m sure it makes sounds good over a cigarette and a cafe au lait. In English, like most economic theorizing, it just sounds like bullshit.

Noah Smith link: http://noahpinionblog.blogspot.com/2014/01/heroes-of-blogging.html

I think he was so wrong about Spain, he took this one down: http://noahpinionblog.blogspot.com.au/2014/08/before-nazis-there-were-spanish.html

Krugman on how the model with the right answers is the right model but you have to have a model to get the right answers (don’t you dare come in here without a model): http://krugman.blogs.nytimes.com/2014/10/14/the-state-of-macro-six-years-later/?_php=true&_type=blogs&_r=0

164

Jerry Vinokurov 10.18.14 at 4:14 am

““You don’t know about internal validity? Moron.””

I didn’t read Barry’s comment that way, and didn’t take offense at it.

The term (judging from you and Jerry and others) don’t seem to be in much use in your sciences, but you do deal with it. As I’ve explained above.

I think what you call “internal validity” we typically refer to as “calibration,” as you note. Different terms for a very similar sort of thing.

What I’m surprised is that Jerry, as a prime example, seemed to not understand the idea, in the sense that a controlled experiment is – well, controlled.

I do understand it; it’s just that there’s more than one way to “control” for things. You can do so manually by limiting the values of some of your parameters, or you can do it statistically by filtering out subsets of your data. As I understand it, and correct me if I’m wrong, that’s how large-scale social surveys work; that’s how astronomical surveys work too. Anyway, I don’t think this distinction presents a terribly fruitful line of debate.

165

Neville Morley 10.18.14 at 8:24 am

@ZM #160: if you conceive of history as a unilinear succession of different, clearly defined periods, then clearly classical Greece was not the start of the Modern Era. But if you take a more cyclical view, then it’s possible to see classical Greece as a modern period, analogous to rather than identical with later European modernity. Periodisation isn’t objective; it’s a way of telling different stories about the past.

166

ZM 10.18.14 at 9:23 am

Neville Morley,

But seeing Ancient Greece as modern is not cyclical unless you are about to posit your own grandma relative of cycles of modernity every 1500 years or so for all eternity. I do not think you are going to go out on a limb to support this theory of the eternal return of modernity?

Early modern usually starts around about after the crusades I think. If you were writing a long book then of course you would bring in the influence of classical and Arabic texts on European early modernity and modernity.

Periods are usually linear in western chronologies, with some overlapping – like restoration England is a period within a period etc.

I did not say historical periods are not a matter of people working out the periods , remember I am the one who thinks we had better go back to saying natural philosophy et al since no one seems to know what science is and is not (You were the one who likes Thucydides universalising, whereas I am more fond of Herodotus who you said just accepted everyone’s high tales).

That is how the dispute about historical periods came about – because phenomenal cat said we have to keep the classification of science even if we can’t agree upon its meaning because otherwise the whole modern project would fall apart – so I pointed out the modern project fell apart before I was born and we have been in the post-modern period since around the 1960s – then he/she said we still have science therefore we are still in the modern period – so I said the ancient Greeks had science and they weren’t in the modern period and neither are we now.

Whereupon you said modernity could be cyclical – like the eternal return idea – but I am doubtful that you will make a case for the eternal return of modernity – although it would surely be very fascinating if you would do so?

167

ZM 10.18.14 at 9:24 am

Um, by ‘grandma relative’ I meant to type ‘grand theory’…

168

Peter Erwin 10.18.14 at 9:38 am

Barry @ 162:
In any science, I’ll bet that you will find different subsets off of a list which basically says, ‘what observational effects do we have to deal with, which wouldn’t matter/matter much less in a controlled, randomized, double-blinded experiment’.

As Jerry pointed out, controlling for different effects or parameters can and is done in observational studies, and the same is true for randomization. As for double-blinding: this is something that’s relevant only when the subjects of study are in some fashion sentient, and when the nature of the study (e.g., administering a medical treatment) could impinge on their awareness. For much of science, it’s an incoherent idea: how would you double-blind a study of different classes of volcanos?

(Single-blinding, in the form where some details of the subjects are hidden from the researchers, can certainly be done. If you’re worried that people on your research team might, e.g., unconsciously — or even consciously — bias their classifications of some objects or subsets of the data so as to support a hypothesis they favor, you can arrange it so they don’t know which subsample different objects belong to when they’re classifying them. There are “observational” examples of this in the “Physics” section of the Wikipedia article on Blind Experiments.)

169

J Thomas 10.18.14 at 12:27 pm

#168 Peter Erwin

Single-blinding, in the form where some details of the subjects are hidden from the researchers, can certainly be done. If you’re worried that people on your research team might, e.g., unconsciously — or even consciously — bias their classifications of some objects or subsets of the data so as to support a hypothesis they favor, you can arrange it so they don’t know which subsample different objects belong to when they’re classifying them.

Sometimes you can do that, sometimes you can’t. It’s worth doing when you can.

It’s possible to bias your results by the application of your model. When it’s complicated, how you decide which complicated outcome belongs to which theoretical result, and what proportion of them has statistical meaning, can itself include as much bias as unconscious tendencies to put outcomes into the expected bins.

#164 Jerry Venokurov

I do understand it; it’s just that there’s more than one way to “control” for things. You can do so manually by limiting the values of some of your parameters, or you can do it statistically by filtering out subsets of your data.

Yes. And remember that recently climate deniers (including our own Brett Bellmore) argued that temperature readings were wrong because the researchers filtered some of the data? They claimed that the researchers intentionally removed data that proved there was no temperature change, and artificially made it look like temperatures were rising. The researchers should instead just use the raw data.

But no, wait, they thought the researchers should throw away the data that’s too close to cities because growing cities are growing heat islands that make it look like temperatures are rising when they aren’t. The raw data has to be filtered to remove bias.

The question whether we increase bias or reduce bias when we filter our data is not in general easy to answer. It depends on the model and the goals.

#163 Thornton Hall

Another example is recent Krugman, claiming that every single causal claim in economics carries a model, whether implicit or explicit. So simply watching one market transaction cause an outcome and writing about it is sophistry, in his reading, because it’s hiding the model that is in the background.

“One man’s experimental error is another’s Nobel prize.”
“Some things must be seen to be believed. Most things must be believed to be seen.”

Was I not an economic researcher in college when I noticed that my dorm with private bathrooms was much less friendly that older dorms where people had to share bathrooms?

Did you know that students were randomly assigned to dorms? It could have been that for one reason or another unfriendly people tended to be assigned to your dorm. When you connected the unfriendliness to the private bathrooms, you implicitly had a model. You could have looked at some other aspect of the architecture, or at the location, or whatever. There could be very good reasons you chose the one you did — I don’t claim it was arbitrary or unreasonable — my point is that it didn’t happen in a vacuum. You did have reasons which predisposed you to notice that one.

170

Barry 10.18.14 at 1:39 pm

Jerry: “I do understand it; it’s just that there’s more than one way to “control” for things. You can do so manually by limiting the values of some of your parameters, or you can do it statistically by filtering out subsets of your data. As I understand it, and correct me if I’m wrong, that’s how large-scale social surveys work; that’s how astronomical surveys work too. Anyway, I don’t think this distinction presents a terribly fruitful line of debate.”

I’ll drop it.

171

mattski 10.18.14 at 1:39 pm

Anyway, I don’t think this distinction presents a terribly fruitful line of debate.

Amen to that.

172

Barry 10.18.14 at 1:40 pm

Thanks for dealing with Thorton, J Thomas. It’s interesting that he read Krugman’s statements, and didn’t really understand them. Your description of models inherent of the bathroom observations was good.

173

Thornton Hall 10.18.14 at 1:51 pm

Did you know that students were randomly assigned to dorms? It could have been that for one reason or another unfriendly people tended to be assigned to your dorm. When you connected the unfriendliness to the private bathrooms, you implicitly had a model. You could have looked at some other aspect of the architecture, or at the location, or whatever. There could be very good reasons you chose the one you did — I don’t claim it was arbitrary or unreasonable — my point is that it didn’t happen in a vacuum. You did have reasons which predisposed you to notice that one.

That counts as “dealing with Thornton”? Why and how?

My economic model was what? I have no idea what y0u mean!

Yes, I made the choice to believe that the causal factor was bathrooms when it could have been something else, but claiming that is a model is begging the question. You seem to be saying that a model is defined as “think that attributes causality”, which makes Krugman’s claim a mere tautology.

(By the by, I agree I am not arguing against what Krugman really strictly meant, which is to say I am choosing to not understand him.)
J

174

Thornton Hall 10.18.14 at 2:03 pm

And I know the Communists on the blog won’t appreciate this, but the conflation of “model” and “world view” is just intellectual idiocy. Yes, I have some sort of model of the external world that colors my every observation. But calling that a model is how very smart people become very, very dumb, because once you’re in that rabbit hole there is a ton to talk about and not a single genuine insight into the real world.

175

L.M. Dorsey 10.18.14 at 2:46 pm

@stevenjohnson 121

Thank you for the clarification. Not “an admirer” tout court, but how can you not admirer a man who taught himself ancient Greek to read the fragments of Parmenides (the fragments of Parmenides) — tho the idea that ancient Greek (or any language) is susceptible to autodidacticism…

The quote is, I think, from Conjectures and Refutations. But I agree about first books, at least those of “thinkers”. As a philosophy instructor of mine once remarked, you have to keep a close eye on their opening gambits. It’s where the “slight adjustment” is made that pulls the world inside out.

176

William Timberman 10.18.14 at 3:22 pm

Thornton Hall @ 174

And I know the Communists on the blog won’t appreciate this, but the conflation of “model” and “world view” is just intellectual idiocy. Yes, I have some sort of model of the external world that colors my every observation. But calling that a model is how very smart people become very, very dumb, because once you’re in that rabbit hole there is a ton to talk about and not a single genuine insight into the real world.

To me this sounds uncomfortably reminiscent of Spiro Agnew’s nattering nabobs of negativism. Do you really hope to put anyone — let alone everyone — in their proper place with this sort of rhetoric?

One can indeed infer models from world views, sometimes in great detail. Although admittedly speculative prior to some confirmation in experience, they can prove very helpful. One can also reverse the process, although for your sake I hope no one is uncharitable enough to apply it to the snippet quoted above.

177

Neville Morley 10.18.14 at 4:37 pm

@ZM #166: just to be clear, I’m not advocating a cyclical view of an ever-recurring modernity as a historical fact, but simply noting that one could represent global history in these terms – the successive rise and fall of civilisations, with each of them passing through a modern phase before starting to decline, and that various writers have made precisely this claim, including the idea that classical Greece experienced a form of modernity. Yes, periods are usually linear in Western chronologies, but not invariably; the main thing is that periods aren’t actually real, they’re just ways in which we imagine the past, and it’s always possible to do this differently.

Likewise, while I remain endlessly fascinated by Thucydides, I don’t myself think that he offers universals in the way he’s commonly assumed to, nor that Herodotus offers mere stories; my point was that this is how they are conventionally understood and represented.

178

J Thomas 10.18.14 at 8:53 pm

#173 Thornton Hall

That counts as “dealing with Thornton”? Why and how?

I hoped I was making a useful response to some of your interesting ideas. I didn’t think I was invalidating much if any of what you said.

Yes, I made the choice to believe that the causal factor was bathrooms when it could have been something else, but claiming that is a model is begging the question. You seem to be saying that a model is defined as “think that attributes causality”, which makes Krugman’s claim a mere tautology.

Well, but what we talk about as scientific method is mostly about how we test hypotheses. There’s very little there about how we come up with hypotheses in the first place. If there’s bias in the way we come up with hypotheses in the first place, then that can’t help cause bias in the results. You can’t test a hypothesis that you never think to make.

Is there bias in the way we make hypotheses? Of course there is, but it’s hard to study. Is it a good thing or a bad thing? (It would be good for us to be biased against making stupid hypotheses that are obviously wrong, provided that didn’t also keep us from making some important useful hypotheses.) I don’t know whether it’s a good thing or a bad thing. I haven’t seen any well-done experiments studying how it happens or what practical results it has.

You always start out with ideas about things that fit your experience but aren’t well-tested. I doubt there’s any way to avoid that. It affects your results and that’s probably uavoidable. If you get past some of your prior unconscious biases there will be others that are still hidden. I recommend a certain humility.

Beyond that, awareness of your pre-existing unconscious biases might help, but on the other hand it might not help either. So if you’re doing fine keeping them in place, you risk wasting your time by trying to get past them.

179

Bruce Wilder 10.19.14 at 3:49 am

J Thomas @ 178: Well, but what we talk about as scientific method is mostly about how we test hypotheses. There’s very little there about how we come up with hypotheses in the first place.

My little theme through this whole thread has been how weird I find it that the discussion, from the OP on, latched so determinedly on the observation and experiment, and seemed to neglect the theoretical. Because I would say that what distinguishes science from, say, folk wisdom, is that science does its best to move beyond finding correlations to developing an analysis that says something about what is not observed. It’s that guessing at an underlying mechanism that we call inference that puzzles the philosophers of science, but it is also that characteristic that distinguishes science from more casual ways of knowing things.

It may be that Thornton Hall noticed a correlation between the “friendliness” of dorms and how elaborate the dorm plumbing was. People have been noticing correlations for millenia. A lot of useful human knowledge over the centuries has come down to craft, based on noticing correlations and coincidence.

What may be said to distinguish science is systematically testing correlations and trying to find an analytic model that “explains” the correlations in a powerful way, a way that allows us to manipulate the world and gives us technology, allows us to teach each other that knowledge and leverage that knowledge to learn more.

Having a good theory prepares one to be curious. It’s a strong theory that allows one to develop hypotheses in a formal sense, but even more importantly, it is a strong theory that allows a scientist to figure out out to observe, measure and interpret the results. I’m not saying that a scientist cannot usefully observe and measure in advance of a strong theory; scientists have been systematically accumulating data, classifying things, and looking for measureable constants for centuries. Finding a system of biological classification that seemed to work was a huge advance, for which Carl Linnaeus is duly honored as a great scientist. That system and its apparent relationship to a tree of life tantalized scientists and sent some into rhetorical paeans to evolution. But, the truly great thing was Darwin finding an apparent mechanism, of descent, variation and natural selection. It was finding a theory, which allowed biologists to generate incredibly productive hypothesis, that allowed the discovery of ecological systems and dna and aided the understanding of natural history.

180

Thornton Hall 10.19.14 at 4:02 am

Exactly. Its the skipping a step that drives me crazy. If we never bother to amass all the (possibly spurious) correlations and skip right to the theory, all sorts of problems arise. My anti-Marxist blast was about the endless circles of debate that ensue regarding the epistiological status of those theories. Because it just doesn’t matter. The problem is prior. The problem is the missing steps. And the obsession with models, fueled in part by all the different discourse communities having their say about the models, is that it declares the pre-theory or proto-theory stage as a waste of time, when, given the state of the science, it is actually the only useful way for an economic scientist to spend his time.

181

john c. halasz 10.19.14 at 4:42 am

B.W. @179:

Ida thought you knew this. The problem is not observation vs. experimentation or induction vs. deduction. but the difficult problem of accounting for “abduction” (Peirce). What makes some set of concepts such that they are worthy of “empirical” testing, and how are such ideas generated. Of course, no set of concepts can be completely “operationalized” and completely verified via predictive criteria, else you end up with a Rube Goldberg apparatus. But just why do certain concept-sets emerge as worthy candidates for selection and rejection in empirical terms and others remain hopeless or errant. That’s why the philosophy of science tends to appeal to the history of science. And why the complex inferential structure that we refer to as “science” is rooted in a broader project of human understanding, orientation, and practice.

182

Bruce Wilder 10.19.14 at 7:44 am

@ jch

I hope I do understand it.

I don’t think the problem is Thornton Hall’s “missing steps”. There are no steps, because there’s no bridge between the analytic model and the actual world across which to take such steps. We leap from one to the other. It’s that inferential, intuitive, insightful leap that troubles the philosophy of science, which wants to find or build a bridge where there can be no such thing.

Popper recognized that there was a gap, and distinguished between the a priori nature of analysis and the synthetic nature of auxiliary hypotheses bundled into operational models, which were necessary to organize measurement of the observable world. He still wanted there to be some kind of “falsification” that would reach back with the necessary force of logic from the operational model to the analytic model. It doesn’t work that way. The analytic model is off on its own ideal plane — it is never wrong in a factual way, because it is never factual, never descriptive; qualitative, not quantitative.

Sometimes, as Kuhn surmised, scientists will abandon one analytic model or paradigm for another, but the switch does not rest on any singular test of measurement, because measurement depends delicately on the model, and different paradigms will be mutually incommensurable in the operational models they produce.

I think economics goes off the rails in the first instance, because economists refuse to understand that their analytic theory is NOT descriptive; they think their analytic models are analogues to the actual world, and when they work out even the simplest little model, they are instantly waving their hands at the classroom windows, imagining against all reason that, for example, perfect competition is a reasonable description of at least some industries/markets just beyond the pane of glass.

I don’t know much about physics, but I get the definite idea that the analytic theory of classical, Newtonian physics, even if it is pretty obviously not descriptive, does have a pretty straight-forward correspondence to any operational model, even after adding the the complications of auxiliary hypotheses and quantitative parameter values. It must be reassuring, or at least a guide to intuition.

Economics isn’t like that: the actual economy is topsy turvy from the gestalt image formed by the analytic theory. All the assumptions that allow deductive reasoning to operate to arrive at a solution in the analytic model — all those assumptions are violated by the conditions of the actual world. Economists have been working for years on the logical consequences of violating each assumption in isolation, but the actual world presents the violation of all of them, coincidentally. The theory understands a world of complete information; the world presents radical uncertainty. The theory understands stasis; the world presents path dependent and persistent disequilibrium. The theory assumes away technical problems and strategic behavior; the world is obsessed with technical problems and breeds strategic behavior of the most creative, albeit often degenerate types.

If Popper had been right about falsification — and I don’t think he was — then, economic theory would be falsified in spades, should the discipline ever admit the existence of any, but the most stylized facts.

I do not reject economic theory. I accept it, indeed I endorse it, as a body of concepts and demonstrations of logical reasoning. What troubles me is what seems to trouble Thornton Hall as well: that economists seem to know so little, and care so little, about understanding the actual economy, about collecting and cataloguing facts about the actual economy’s institutions and history. I wouldn’t have them grubbing about with hodgepodge econometrics, collecting spurious correlations; I’d want them doing field work like sociologists or some better class of anthropologists, building operational models to understand how the institutions they encounter operate. So, I’m not far from Thornton Hall.

I’m not on board though with Pierce and his criteria for good guesses (economical “abductive” reasoning). It seems to me that this is the kind of argument Krugman often gives for preferring a model: it is the simplest one that explains the phenomena. Lot of loose ends, there.

I think you need a very good theory, indeed, to do really useful interpretative inference or abductive reasoning. In particular, you need an analytic theory that identifies the necessary and sufficient elements, so that you have some basis for saying whether the model you invoke is sufficiently complex. I think Krugman’s model selection often fails in that regard. His insistence that he can leave money out of macroeconomic models when it suits his purpose seems borderline insane to me; macroeconomics doesn’t happen without money — you cannot ever leave it out willy nilly (though, maybe, sometimes it can be shown to not be the strategic factor).

Not long ago on another thread, I argued that economics should never treat the allocation of resources as a sufficient theory of production — something economists do fairly frequently, when they unthinkingly invoke the production function. If you attempt an operational model of a firm or a production establishment, you see the need to have an analytic theory of production, which spells out the necessary and sufficient elements — maybe something like Stafford Beer’s “viable model, as dsquared suggested.

That’s probably too brief a sketch to explain adequately why I think analysis is necessary or how it is useful, even though it cannot be, strictly speaking, descriptive.

In Krugman’s case, one of the consequences of not having a commitment to an analysis, or a vision, is that his models are too loosely jointed to carry back to his world view, the jolt of confronting inconvenient facts. Thornton Hall has sometimes expressed the hope that Krugman or Noah Smith might wake up and smell the coffee after a sufficient jolt. Things are not that connected. Krugman doesn’t have a vision of the economy as an actual, operational system of systems, so there’s nothing to feedback upon. His choice of models is always convenient; no inconvenient facts need apply.

183

J Thomas 10.19.14 at 9:30 am

I want to once again suggest people play the game Eleusis.

https://en.wikipedia.org/wiki/Eleusis_%28card_game%29

A card game. Players take turns playing cards that either play or fail to play depending on a rule which depends on the cards which have played before. If the card plays, put it beside the previous last card that played. If it fails, put it below the previous card. So every player can see at a glance the entire previous history of the game. The dealer writes down the rule but does not show it to anybody. A player wins the game by correctly guessing the rule.

A rule of thumb is that rules should be set up so that about a quarter of the cards in the deck will play at any one time.

Here’s what I consistently see happen when people start playing the game. Say we start with a very simple rule. Spade Heart Club Diamond and repeat. After a lot of cards have failed to play, the cards that do play will be

AS 2H 3C 4D 5S

People think they know the rule. Part of what’s going on is that they don’t like to play cards that fail, so they choose the ones that are most likely to play. But eventually somebody makes a mistake.

AS 2H 3C 4D 5S 6H 7C 8D 10S

Then everybody gets upset. How can that play? It doesn’t fit the rule! But it does. People are likely to start thinking about more complicated rules.

AS 2H 3C 4D 5S 6H 7C 8D 10S JH KC 2D 3S

They might decide that it’s the first or second number that works. It takes awhile for them to get the point that they want to look for cards that won’t play. Somehow people think of cards that won’t play as failed experiments. But eventually they do get the rule.

Say the next game the rule is black follows red follows black and something about numbers. Not unlikely the beginning cards that play will be

AS 2H 3C 4D 5S

People get surprised and kind of upset to find that 6D plays.

This can be played as a game of perfect information. Everybody knows everything except the rule. And still people tend to guess rules that are far more restrictive than the real rule, and they tend to add complications to their guesses when they find errors.

When they only see part of the data, it’s even easier to assume that the rule is more restrictive than the real rule, and that apparent failures are because of the hidden data. They can predict what the hidden data has to be, to fit their guessed rule.

It’s a good game. It’s worth it to play this game long enough to get good at it. Play with a variety of people so you don’t fall too much into the same habits.

It’s unfortunately easy to make rules that nobody can guess. Occasionally players will feel like they’ve accomplished something by doing that, but it just makes the other players cranky.

The game is unrealistic in that the dealer tells players when their guessed rule is right. That’s part of what makes it a fun game instead of an unsolvable philosophical conundrum.

184

ZM 10.19.14 at 10:07 am

Neville Morley,

“@ZM #166: just to be clear, I’m not advocating a cyclical view of an ever-recurring modernity as a historical fact, but simply noting that one could represent global history in these terms – the successive rise and fall of civilisations, with each of them passing through a modern phase before starting to decline, and that various writers have made precisely this claim, including the idea that classical Greece experienced a form of modernity. Yes, periods are usually linear in Western chronologies, but not invariably; the main thing is that periods aren’t actually real, they’re just ways in which we imagine the past, and it’s always possible to do this differently.”

Who are the writers who have presented history in this way? I would guess that they were structuralist grand narrative sort of writers from 19thC or early 20thC, or maybe people from other more sciencey disciplines who thought they’d try their hands at a grand sort of history, like Jared Diamond or someone. I have not studied classical history formally only some translated literature – but I think not only are the differences too profound between the classical world and European modernity, but the influence of the classical world- both unbroken and carried through and broken but brought back with the crusades – marks a difference .

I think that the seasonal and ritual years (usually interlinked) are cyclical – and some cultures are based more faithfully around following and celebrating these calendars . There are I think religious cosmologies (Indian?) that have a cyclical understanding , but this is not our culture’s cosmology , even where that cosmology is more broadly contested ie. Big Bang and evolution or intelligent design etc. Definitely with climate change we are seeing human influence that will have effects for centuries or thousands of years in a linear way – regular cyclical seasons will be changed.

I do not agree with you that period are entirely relative – history , while I don’t think it is a science in the sense of a natural philosophy , is meant to be anchored in truth. I think an assertion that the historical period of Ancient Greece was the modern period departs too far from the truth. My favourite historian said the main thing in writing history is to try to conjure up the moments in themselves before the subsequent events happened – that there were other possibilities pregnant in the moment, not just what we now know happened. In that sense every moment is modern to itself – but that is not what we mean by the modern historical period in Europe

I think this question you bring up of whether historical periods are simply relative human constructions (you can be for, I will be against) – connects to our topic of what is science, and what relation does science have to truth. It seems that with natural philosophy and experimental philosophy making so much ‘progress’ in measurable fields – there came to be more and more a feeling that only measurable things are able to be spoken of as truths. So we have social sciences, which people want to remain sciences , but their subjects are not really very measurable. But then this does end in post-modernism – because it is what is not measurable that is most important. Climate change science matters less, really, than the question of what are we to do about it.

“We shall not cease from exploration, and the end of all our exploring will be to arrive where we started and to know the place for the first time” T S Eliot

185

William Timberman 10.19.14 at 1:14 pm

Some aspects of this discussion remind me of the cultural relevance debates that took place in American universities during the 60s and 70s. To the extent that you get only what you’re aiming at — never mind, for a moment those individual eureka moments that change the perception of the researcher — it doesn’t seem to me that asking scientists how, and with what intention they go about choosing the objects of their research should be looked upon as a form of lèse-majesté. When sociology and psychology wind up being wholly-owned subsidiaries of the internal security apparatus, and economics winds up being the staunchest defender of the status quo since the Roman Church, questions about their legitimacy as scientific disciplines more or less ask themselves. About physics, I haven’t a clue, but I do wonder why it is that we have H-bombs in our national arsenals, but not fusion power reactors in our backyards. Is it really only because an explosion is easier to conceive of and experimentally confirm than a controlled reaction, and must therefore logically precede the controlled reaction in the chain of research? I don’t know the answer, but I think that the question is entirely legitimate. If that leads to impertinent epistemological doodling in some quarters, why not just accept it as the cost of doing some very important business?

186

john c. halasz 10.19.14 at 3:08 pm

187

Bruce Wilder 10.19.14 at 4:01 pm

William Timberman @ 185

There’s a great scene in the recent film, Boyhood, where the ne’er-do-well father is having his son over for a sleepover and they’re talking late at night. The little boy wonders aloud if there’s really no magic in the world, no elves and such. The father, clearly in sorrowful pain at being present for this moment though he’s been absent for so much, rises to the occasion and says,

What if i tell you a story about how underneath the ocean there was this giant sea mammal that sang songs and it’s so big that it’s heart was the size of a car and you could crawl through the arteries? I mean, you’d think that’s pretty magical, right?

And, the son says, “so no elves?” And, the father says, right, no elves.

There’s a lot of ambivalence about science and its displacement of magic, a wish for science to be magical and meaningful, a nostalgia for a religious understanding of the world, and just a weariness with the deafening crescendo of scientific and technological advance.

A lot of people have a reduced form understanding of scientific method, where it is all just post hoc, propter hoc, searching for rules that fit the sequence of events, as in the card game J Thomas describes. There’s a whole pseudo-controversy underway, led by people, who think technological advance is slowing because we don’t have flying cars, while others wonder why the iPhone 6!?! isn’t cheaper already and whether robots are going to eat everyone’s jobs just after everyone dies of Ebola. Meanwhile, we cannot collectively come to grips with climate change and peak oil, because certain decline is so uncertain. What can you do?

188

Rich Puchalsky 10.19.14 at 4:09 pm

William Timberman: “About physics, I haven’t a clue, but I do wonder why it is that we have H-bombs in our national arsenals, but not fusion power reactors in our backyards. Is it really only because an explosion is easier to conceive of and experimentally confirm than a controlled reaction, and must therefore logically precede the controlled reaction in the chain of research?”

If you think that the question is legitimate, then are you going to accept a simple answer as legitimate?

Without getting into the history of WW II, which might sort of answer the question on its own, it’s a whole lot easier to make something explode than it is to make a controlled reaction that doesn’t break its container and so on. If you’d asked the question about gunpowder, it might have been “why do we have fireworks before we have cannon?” and come up with a whole new (and largely BS) answer about societal priorities.

189

Bruce Wilder 10.19.14 at 5:45 pm

john c. halasz @ 186

I love it.

190

J Thomas 10.19.14 at 6:44 pm

#187 Bruce Wilder

A lot of people have a reduced form understanding of scientific method, where it is all just post hoc, propter hoc, searching for rules that fit the sequence of events, as in the card game J Thomas describes.

Try getting some people together and play the game. I expect you’ll like it. It’s a useful metaphor even if you think of other things as scientific method. One of the things that’s going on is that what you already think has a big effect on what new experiences you go out and try. People have a surprising tendency to make self-fulfilling prophecies — they behave in ways that tend to give them more experiences that fit the ones they already have.

It would be possible to expand the game to something that fit, say, 2 dimensions so it wouldn’t be post hoc, but I’m guessing there’s enough to learn from the simple version that you might want to put that off.

191

William Timberman 10.19.14 at 6:45 pm

It’s not just a matter of why we don’t have this instead of that, which in any case is a question with more relevance to applied science and engineering than to so-called pure science. The influence of funding, of whose dream, or rather whose carefully calibrated and therefore somehow mundane intention governs the process of research is relatively trivial, and therefore easy to see.

The Blackberry arguably arose out of a vision that judged corporate — i.e. hierarchical — control foremost among the needs which it might prove profitable to satisfy, The iPhone arose out of a different vision altogether, that of providing us as individuals with a tool to deepen our interactions with the world, and therefore increase our leverage over it. As it turned out, that vision proved equally profitable, so Steve Jobs is now no longer considered a hippy humanist, but a visionary fit to reside in the corporate pantheon one niche over from Bill Gates and the founders of RIM. A revealing story about complex motivation, this, but not exactly the story I was trying to tell.

We’re encouraged to believe that what motivates pure research is a kind of naive curiosity, even though we’re hip to the fact that motivating and governing are two different things, and freely accept as a limiting condition of modernity that an Oppenheimer without a Groves is simply not on. What do we make, though, of an Einstein who refused give up on his pursuit of a Unified Field Theory? The bit about God not playing dice with the universe tells us, if it tells us anything, that intention, much less a general curiosity and careful observation, were never the whole story about how science gets done. It also tells us that we should be a little more humble in our assumptions, and a little slower in coming to conclusions about something that has a habit of turning itself inside out just as we thought we had it figured out.

192

Bruce Wilder 10.19.14 at 8:57 pm

ZM @ 184: Who are the writers who have presented history in this way?

One of my guilty pleasures is The Archdruid Report.
http://thearchdruidreport.blogspot.com/

He very much posits secular rationalism as a stage in the rise and fall of civilization, drawing on writers like Toynbee and Oswald Spengler. He clearly expects that secular rationalism will age into corruption and meaninglessness as surely as a boy entering manhood transitions imperceptibly but with tragi-comic inevitability from exuberant growth to terminal decay.

193

William Timberman 10.19.14 at 9:28 pm

Thanks, BW. I too collect apocalypsos (as a sort of nervous habit, rather than a full-blown hobby) but never thought to look in the Druidic liturgy. I should get out into the (full) moonlight more often, I guess.

194

William Berry 10.19.14 at 9:56 pm

BW @192

Thnx for that link. Bookmarked on my i-Pad.

Big Toynbee guy in my youth, and one of those oddballs who plowed through A Study of History just for the hell of it.

195

William Berry 10.19.14 at 10:10 pm

“He clearly expects that secular rationalism will age into corruption and meaninglessness as surely as a boy entering manhood transitions imperceptibly but with tragi-comic inevitability from exuberant growth to terminal decay.”

Yes. Toynbee’s view, while not exactly au courant, has a lot going for it. His rep died the death of a thousand cuts from critics picking mostly on his many imprecisions, and from a general turn away from the idea that you can derive regularized rules or laws of history.

I am with Neville Morley on this one. The Hellenistic Period, coincident with the ascendancy of the Alexandrine dynasties, can certainly be seen as a late, “modern” (postmodern, even?!) stage of Greek culture.

Obvious parallels to our own time. Are we living in the end stages of a latter-day “hellenistic” era?

196

ZM 10.19.14 at 11:58 pm

There is quite an interesting article on our topic in today’s Guardian

http://www.theguardian.com/books/2014/oct/19/rebecca-newberger-goldstein-interview-science-philosophy-plato-googleplex

The person interviewed Rebecca Newberger Goldstein wrote a series of platonic dialogues in contemporary times. Apparently there was a great debate about the boundaries of science last year:

“Last year the debate flared up in a much-publicised intellectual spat between the cognitive psychologist Steven Pinker and the cultural critic Leon Wieseltier, who accused Pinker and his fellow scientists of practising “scientism”, a term he defined as “the translation of nonscientific discourse into scientific discourse”.”

Goldstein believes science is better at ontological questions – answering what is?; but philosophy is better at answering what matters? She does largely agree with Pinker over Wieseltier, but she is married to Pinker so that is not too surprising. She thinks philosophy is better than literature because it progresses –

I am unsure of this idea, I’m not convinced there is a great natural progression to goodness inately in the historical trajectory of philosophy as a discipline. Although I do not agree with William Berry’s view of there being an inate decline to decay in the general historical trajectory either. It is quite a silly idea – because there is always growth and decay happening at the same time – that is how forests and the bush or your garden at home and compost keep on being. But perhaps since we have such great troubles with climate change and un-sustainability at the moment all the philosophers will work very earnestly to the good , and so there will be a great upswelling of philosophical goodness as a result.

197

William Berry 10.20.14 at 12:28 am

So, the birth, maturation, life, senility, and decay of organism-like entities—with more-or-less discrete identities in space and time—is a silly idea?

Got it.

I have a pretty good idea of what is silly.

198

William Berry 10.20.14 at 12:29 am

Not to mention that ZM reminds me generally of a certain famous couplet of Pope’s.

199

ZM 10.20.14 at 12:52 am

William Berry,

“So, the birth, maturation, life, senility, and decay of organism-like entities—with more-or-less discrete identities in space and time—is a silly idea?
Got it.”

A tree starts as a seed then grows then depending on its nature and environment at some point starts to die and decay. That is the life time of a tree. But the tree is just one tree in the forest – so while the tree is just a young shoot there are many other towering trees, and different sorts of trees with different life cycles, and there are lots of little grubs in the soil and between layers of bark, and birds flying about and singing and nesting and so on. The forest is the all these things and the interrelationship between them – maybe it will last for thousands upon thousands of years – only to be destroyed by a logging company not by all the trees and grubs and birds somehow timing their decay together.

any way , what I meant was silly is the idea that a forest or human community has to stick to the life cycle of a tree – a human community is intangible and made up of lots of various people with agency and languages and arts and technologies etc. There are constraints, contingencies etc – but there is no natural law like gravity (which is not a law, but you know what I mean) that says human communities must adhere to certain stages like a man who is born, matures, grows old, and dies. Of course it is a nice metaphor and sometimes is fitting – but this does not make it a law like gravity.

Also – It is not polite to say I remind you of a couplet from Pope without mentioning which one, just think of how this would weigh on my mind. Which couplet does William Berry think I remind him of? How can I defend myself if you do not even write what couplet it is?

200

William Berry 10.20.14 at 1:25 am

ZM:

Just want to suggest that you think a little more about the concepts of domains, their boundaries and, in particular, emergents; in short, what constitutes entities of all kinds.

And of course there are no gravity-like “laws” involved when it comes to history. Indeed, I don’t believe in natural laws of any kind; laws govern, and only the forces govern natural phenomena. The term “law” is just (well, not “just”, but whatever) a short-cut for (a very high degree of) consistency in observed patterns.

Just forget the Pope thing. It was uncalled-for snark.

Pollo a la brasa, caesar salad, hot bread and too much Riesling for dinner.

‘Night, all.

201

Val 10.20.14 at 2:17 am

William Berry @ 200
What are you talking about domains and entities for? What is the point you are trying to make? ZM makes some very good points about what I would describe as the difference between an individualistic approach and an ecological approach, and rather than responding to these points it seems you are just chucking some words at her. That doesn’t prove anything.

202

Val 10.20.14 at 2:20 am

William Berry @ 200
My questions above are, of course, for when you wake up and feel able to tackle these questions again.

203

J Thomas 10.20.14 at 1:23 pm

#199 ZM

any way , what I meant was silly is the idea that a forest or human community has to stick to the life cycle of a tree – a human community is intangible and made up of lots of various people with agency and languages and arts and technologies etc. There are constraints, contingencies etc – but there is no natural law like gravity (which is not a law, but you know what I mean) that says human communities must adhere to certain stages like a man who is born, matures, grows old, and dies.

Trees are highly-evolved organisms. What if we were to instead consider sponges? Sponges can do a lot of self-assembly. If you separate a sponge into individual cells and then give them a quiet place, they are likely to build themselves together into a new sponge with a new shape to reflect their new location. Pretty soon they’re back to pumping water and filtering it for food. If you mix cells from two sponges of the same species you can get a mixed sponge. They work together. Do sponges die of old age? Probably not.

For that matter, do trees die of old age? It appears some do not. Apple trees kind of act like they do. An old apple orchard makes little apples and doesn’t grow very well. Is it that they have a lot of dead wood etc that they must live around? Is it that after a long time producing apples they are running out of some trace element? Maybe after a long time in one place, a variety of parasites have found them and are sucking away their vitality? Maybe old age is inevitable for some tree species but certainly not for all.

Maybe what WB was getting at, is that there are things that do some homeostasis. When things change around them, they respond to maintain a sort of identity. And under some circumstances they do that very effectively, while with others they start to get out of tune or fail to respond. Some “advanced” systems make fresh copies of themselves that result in something like a reset key — if their past history has led them into a homeostatic dead end, they can die and be replaced by a new one which lacks that baggage. But less advanced systems must survive as best they can.

For things like ecosystems and human cultures, it’s all fuzzy. How many cultures are in competition, and how fast do they replace each other? If it’s only a few and they compete slowly, that doesn’t give a lot of opportunity for them to evolve. But then, how do you decide where one sponge leaves off and the next begins? Maybe our concepts of evolution are too specific. Maybe we don’t see the whole picture.

If life had somehow developed without cell membranes, so it was harder to be sure where one thing left off and the next began, we might have a more generalizable concept of such things. Or, I mean, our cell-less analogues might.

204

Ronan(rf) 10.20.14 at 1:45 pm

“Who are the writers who have presented history in this way? “

Maybe look at people like Peter Turchin ?

http://cliodynamics.info/

(Although I havent been following this closely enough so don’t know if this is your Q)

205

J Thomas 10.20.14 at 4:15 pm

#187 Bruce Wilder

A lot of people have a reduced form understanding of scientific method, where it is all just post hoc, propter hoc, searching for rules that fit the sequence of events, as in the card game J Thomas describes.

Here’s the thing, and I think this is very important: People tend in practice not to be very good at that. With practice they get better. And this game is one way to get practice with feedback about what works.

We are so good at seeing patterns, how can we be bad at science? I think maybe society has set it up that way. People latch onto random patterns, and believe in them, and over a period of time we see which people with which random beliefs prosper. Society wins in the long run when people believe irrational things at random, and test out how those beliefs affect their survival.

That might be good for society in the long run, but it’s good for *you* *personally* to learn how to hold multiple hypotheses at once and how to test them effectively. You are better off if you play Eleusis.

206

phenomenal cat 10.20.14 at 5:59 pm

No grand narratives are not made up over time by lots of people – they are Herculean efforts by one person normally a white man who makes grand claims about all of history or all of people’s nature without citations adequate to justify the claims. Eg Marx and Freud. David Hume, our topic, could maybe be considered to be a prototypical grand narrative figure – for I guess inventing grand narratives seemed the normal thing to pass your time doing if you were always reading Hune and Hobbes and the like.–ZM@160

It seems you are actually buying into the GREAT MAN theory of (intellectual) history while arguing against it. As if all these individual, hep cats show up on the scene and sui generis create their own *herculean* narratives out of the void–yet down thread you argue for an *ecological* perspective. What gives? You think Freud or Marx transcend their ecology? Do you think they are not expressions, briefly lived assemblages, of myriad and unique historical conditions?

Well no one will agree as to what science is and is not , so it is a more useless word than the word knowledge. I suggested we can go back to natural philosophy et al. Whether indigenous people pre-colonisation could be said to have practices of knowledge finding we could say we’re equivalent to science, would depend on our agreeing upon science having a set meaning .–ZM@160

And our running disputation was started by my acknowledging you probably had something there, but it that might require a collective recalibration of the value we have heretofore placed on science. Except you see this as no biggie b/c we are already in the POMO. That’s you’re argument. The fundamental relationship between modernity and science is a big yawn b/c in 1979 Lyotard wrote a book about POMO. I humbly submit you are missing the forest for the trees.

Po-mo works for the period after modernism from mud 20th c. No one has come up with a better name – I think eventually they will because Pomo sounds very transitional – but first the historical acting needs to take place before you can name the period , so whatever the name of the period is it will depend on how people act now.–ZM@160

Works why exactly? I thinks it sounds transitional for a very good reason b/c it denotes another phase of modernity. So yeah, the *historical acting needs to take place before you can name the period* b/c the *historical acting* that would denote something other than the fundamental extension of the modern period has not taken place– and apropos the thread topic, my point is that the privileged place science occupies in our ongoing history is directly linked to said historical period.

FYI, asserting to me that Ancient Greece is *part of the antiquities* is 5th grade stuff. I’m aware of the general chronology of historical reckoning.

207

ZM 10.21.14 at 8:18 am

J Thomas,

“If life had somehow developed without cell membranes, so it was harder to be sure where one thing left off and the next began, we might have a more generalizable concept of such things. Or, I mean, our cell-less analogues might.”

Very funny ;)

phenomenal cat,

1. It is a very interesting topic the topic of individuals and communities. I always forget which way around the saying goes – is it ‘you can’t see the forest for the trees’ or ‘you can’t see the trees for the forest’ (now I’ve written it down I think its the first?). Marx was the only person who wrote all his books, even his great friend Engels seemed to like to write about similar things differently. Nonetheless, imagine our great surprise if we found Marx’s works were actually dated to make him a contemporary of Chaucer. Marx might have written all his books, but they were left unread and only dusty manuscripts sold by lot after he passed away and thereafter kept in some minor archive somewhere. That was not the case though, Marx and his works became quite famous and influential. Part of writing history is recording famous things or making things famous you think should be famous like E P Thompson’s stockingers. Anyway, in history you can tell about general social things, famous things, unappreciated minor figures etc. all together if you wish to.

2. Oh, I see – you were not so much arguing against what I said but saying that it would be difficult to achieve. I guess I think it is already apparent to some extent because of post-modernism etc but you think people still correlate modernism with science so it would be difficult to go back to saying natural philosophy. Well I am not really going to go out of my way trying to encourage people to replace science with the term natural philosophy , even people who grow succulents in laboratory beakers might find my insistence on the term natural philosophy outlandish . So you’re probably right there, post-modernism has not yet gone backwards far enough for a renaissance of the term natural philosophy and experimental philosophy, unless I am just before my times and next antipodean autumn it will be very fashionable (you must allow me some hope). I see you have written the saying you can’t see the forest for the trees here, so that is probably the right way around for the saying.

3. Fredric Jameson I think has written that he thinks post-modernism is the heightening of modernism. I think this is because he is a Marxist and primarily associates modernism with capitalism. I think it is more of a departure from modernism – I think people realised that what might be called the modern project failed and was doomed to failure/an illusion, as well as being wrongheaded and implicated in a lot of wrongs. On the other hand Bruno Latour looks at scientists and sees they are human like anyone else and wrote his We Have Never Been Modern Book. I don’t think that still having science means the modern project is continuing the way it had – I think the authority of science is greatly diminished and there is not really the same sort of expectation of transformation by some great radiant modern scheme. Also, with modernism there was generally the idea of progress via scientific methods – but we have all sorts of problems , the biggest being climate change. While science is needed to provide information on how we can change to non-climate changing economies and be sustainable – at this point things are not looking especially good. Hopefully things might take a turn for the better, through people acting, but we are really not in a place where we can equate our progress in science with progress in human wellbeing – if anything the opposite has occurred we’re just putting off the awful for some decades if we don’t change.

4. I only asserted Ancient Greece was in the age of antiquities because you said that Ancient Greece could be modern since they had science. I did not think you were unaware of historical reckoning I thought you were claiming normal chronologies were wrong and you had a new sort of chronology. I do not agree ancient Greece is modern.

208

J Thomas 10.21.14 at 8:39 am

I always forget which way around the saying goes – is it ‘you can’t see the forest for the trees’ or ‘you can’t see the trees for the forest’

It’s two different concepts. Use the one that fits.

Comments on this entry are closed.