Isn't there an aphorism to the effect of 'Don't confuse the model with what it represents?'

Albatross is correct, you need simplifying models otherwise the complexity of the world will blind and deafen you, but the model is not the world.

The devil, it has been said, is in the details, and, contrary to what almost all ideologues say, they do matter. Shi'ite and Sunni Muslims began as contingent products of what were, in essence, leadership disputes in early Islam, but they became distinct streams within Islam very quickly, and they are still harping on 1,1400-year-old quarrels. Unlike some 'ancient' hostilities that hardly go back to the 19th century, the differences are real -- and to have missed them is not only to have ideological blinders on, it is to be completely incompetent.

Models are not the world, just maps; and the map is not the territory.

Albatross is correct, you need simplifying models otherwise the complexity of the world will blind and deafen you, but the model is not the world.

The devil, it has been said, is in the details, and, contrary to what almost all ideologues say, they do matter. Shi'ite and Sunni Muslims began as contingent products of what were, in essence, leadership disputes in early Islam, but they became distinct streams within Islam very quickly, and they are still harping on 1,1400-year-old quarrels. Unlike some 'ancient' hostilities that hardly go back to the 19th century, the differences are real -- and to have missed them is not only to have ideological blinders on, it is to be completely incompetent.

Models are not the world, just maps; and the map is not the territory.

Another useful observation about models:

All models are wrong. Some models are useful.
George E.P. Box

Fragano @ 2-3... Models are not the world, just maps; and the map is not the territory

That way way lies disaster, especially when the guy who makes the decisions does do nuances.

(correction to #5)

...especially when the guy who makes the decisions does NOT do nuances...

My opinion on this is best summed up by Clive James, from The Crystal Bucket (a book of TV criticism from the 1981):

The advantage of being in possession of an all-embracing political theory is that you need never be at a loss either to explain events or to propose their remedy...

My own all-embracing political theory, for what it is worth, is that an inordinate proportion of the world's misery is brought into being by all-embracing political theories. These might tend either to the Right or to the Left, but what they have in common is the unwavering conviction that ends justify means. In this respect any attempt to choose between the two sides is pointless. Nor should anyone who finds himself in the middle feel weak on that account. Powerless yes, but weak no. if history is with anybody, it is with those who are not sure where it is heading...

[Vanya] Kewley elegantly embodies the principle that the truth is absolute, even if our grasp of it is relative. Q. Baebius Herennius [James' strawman for the essay, but a useful one] believes that the truth is relative and his grasp of it absolute. She can understand him, but he will never be able to understand her.

Yes, simplifying models are necessary.

But you've always got to pay attention to the circumstances under which your model is valid, and when it breaks down. Be very clear about what assumptions you're making in order to simplify your model. And you've got to check your model predictions against experiment or at least the known laws of physics.

*has math modeling course notes in her lap*

And yes, the idea very much exists that you should be able to manage things you don't really understand. Hence, management consulting.

The SF writer Jerry Pournelle, who's absolutely against the Iraq war, refers to the current ideology for the war as Jacobinism.

The French revolutionaries that overthrew the monarchy thought that they could quickly convert France into a utopian democracy. They were wrong.

Failure to understand the model is not reality, or being overly enamored with rhetorically-powerful but inaccurate models, explains why people seriously advocate libertarianism and communism. Both systems fail in predictable ways, but they sound good (for certain values of good.)

The tension in model-building is between keeping out the detail that doesn't matter, because it makes the model harder to understand and use, and including the detail that does, because the model doesn't work correctly without it. There's always a temptation to strip the model down both because that makes it easier to build and use and because it's more "elegant"; 19th and 20th Century physics, with its terse mathematical formulae, taught people who aspired to scientific precision to value brevity; unfortunately many of them didn't realize that, for instance, "e=mc²" isn't the whole

of the model.

The other common mistake with models is to take a model that works in one domain, and apply it to a domain where it doesn't work. "Democracy leads to peace" worked (sort of) in Europe; it isn't clear that it works well in other parts of the world possibly because "democracy" has different meanings to different people.

On the other hand, it's also notable that these points are frequently harped on as a way of encouraging political quietism and discouraging critical thought. "Ideology" isn't evil and it certainly isn't stupid. Oversimplification is foolish and can lead to evil results. It's perfectly possible to have an ideology (for instance, to believe that some resources ought to be held by the duly-constituted state for the good of all, rather than concentrated in private hands) without being oversimple and hence stupid about it.

There *are* some models that do a VERY good job at relieving the requirement that one understand the details. Thermodynamics is probably the best such example, but in analyzing many situations one can use logic to find terms that cancel out and can then ignore those terms (as in all objects falling equally fast in vacuum because mass is on both sides of the equation so it cancels out).
Neoclassical economic models *do* show some of this strength, but only if used *very* carefully, e.g. if maximum attention is paid to the exact meaning of the assumptions, those are recognized to be assumptions, and in general the process of turning symbols back into concepts is a faithful reversal of the initial. process of turning concepts into symbols.

Certainly there's "a widespread idea in management that you should be able to manage things whose details you don’t understand all that well." The Dilbert cartoons show the effects of that idea, and many of us will have enountered it in 'real' life.

The trouble is, it's also a necessary idea. Say you're running an outfit, such as a university or an aircraft carrier, that includes a wide range of specialisms. You can understand them all superficially, but you can't have in-depth knowledge of everything, yet somehow you have to keep the show on the road. It takes a good manager to do that well.

Good points about checking models against reality.

This article from National Geographic goes into more detail on how our models can lure us into danger.

http://www.nationalgeographic.com/adventure/survival/skills/index.html

Patrick @ 11... It's perfectly possible to have an ideology (for instance, to believe that some resources ought to be held by the duly-constituted state for the good of all, rather than concentrated in private hands) without being oversimple and hence stupid about it.

That's what good models do. Look at the Constitution. It's a good model for Democracy, but it doesn't spell out all the details because circumstances can change.

Maps and models are both kinds of analogies. No analogy is perfect—and that's by definition, because when it becomes perfect it ceases to be an analogy and becomes a description of reality, or even reality itself (because descriptions, however detailed, are models as well).

In fact, your understanding of anything is a model that exists in your brain, and there absolutely will be discrepancies between it and reality. The important thing is to keep that in mind, and fix the discrepancies when they become serious. And by "serious" I mean that they impair the purpose of the model.

For example, my model of the world I live in, like most people's, is essentially Newtonian. I know that if I pull my leg in I spin faster, if I knock a ball into a row of balls the one on the other end may move, and that if I throw something straight up it will probably come down on my head.

That's suitable for everyday use, and pretty consistent. It is not, however, how the universe works on a larger scale.

But that's OK, because the purpose of my model is to allow me to function in the world, to avoid getting hurt whenever possible, and to put off dying as long as I can. Relativity and quantum mechanics don't really affect my daily world very much. But if I want to write about interstellar travel, I have to know something about relativity. (I have yet to think of a situation in which I need to know about quantum mechanics, but it could come up.)

So it's not that the model is discrepant from reality. All models are, and in fact that's what makes them useful. It's whether the discrepancy is relevant to the purpose of the model that matters. The difference between Sunni and Shi'a is not only relevant to the model of outcomes in Iraq, it's critical, and they either missed that, or didn't care.

John S @ 13

You can understand them all superficially, but you can't have in-depth knowledge of everything, yet somehow you have to keep the show on the road.

The problem is when the necessity of that superficial understanding isn't understood either. That produces the idea that the MBA (or whatever) is in itself sufficient background to allow running any business or other organization, without knowing anything about the business or organization being run. (Usually resulting in said business or organization being run right into the ground.)

The whole Barney and Betty Hill saga, which Jim wrote about so compellingly a few days ago, seems to me a classic example of fitting unusual data (the light in the sky) to the wrong model (an alien craft), and then continuing in such a way that that wrong model got further and further wedged into place.

We all need to model-fit to get through the day, but some models are less useful, and harder to displace, than others.

I think the screwups in Iraq have largely been caused by very smart people whose ideology led them to think that they had grasped the essentials of the situation there, despite scary stuff like not knowing the difference between Shia and Sunni.

I disagree. I think the screwups in Iraq were caused by people whose ideology led them not to give a tenth of a shit if they grasped one iota of the situation there. They understood perfectly well what they were doing; successful management just didn't figure into it.

The rest of it's spot-on though. Can you say "Show, don't tell?"

Patrick #11:

Yeah, I'm not saying all ideologies are evil. In fact, I think there are adherents to most widespread ideologies who do see the difference between map and territory, who do understand that there are places where their models will and won't work. But most ideologies I can identify in the world also have their blind adherents, and it often seems like the guys who get into power and apply the ideology to governing are the ones who confuse map and territory.

Perhaps this has to do with the inherent complexity of their jobs; it's a little scary to think of how much you'd need to know to be a decent president or secretary of state or secretary of defense, and I can imagine a deep desire for simplifying models that helped you make sense of the mass of data facing you.

As an aside, I was a little surprised when I checked Making Light and saw my own post at the top. That was cool!

As the pioneering linguist Edward Sapir put it, "All grammars leak."

(For some reason, I've always found this observation, with its acknowledgement of the unwillingness of reality to make a perfect fit with anything except itself, to be a comforting one. I suppose that for someone with a different cast of mind, it might be an inducement to despair.)

Fragano (#2 and 3):

Not only is the Devil in the Details, but, as any Connie Willis fan is aware, God, too, is in the Details.

Samuel Tinianow @ 19 - whether the Bush administration believed their rhetoric or not, they had to sell the concept to others. The model of we'll bring democracy to Iraq was a fairly easy sell to Americans.

Serge #6: The guys at the top rarely do nuance, this causes them all manner of surprises.

I think there's a useful distinction to be made between principles and ideology. Principles act as side constraints - they rule out certain actions a priori, leaving you to solve your practical political problems within the remaining possibility space. This is a good thing: it means you rule out, say, tackling a problem of street littering by introducing armed street wardens to shoot dead anyone who drops any rubbish on the pavement.

Ideology is different: it tells you the answer - often the One True Answer - to your problems, whatever they may be. This can lead you to do ineffective, counterproductive or damaging things because you haven't chosen the most reasonable possibility, but rather the one that most closely fits your ideology. So, for example, you might be driven to solve the littering problem by instituting a complex set of private property rights over the street, and enabling complex legal battles over punitive damages for dropped litter.

If it turns out that the most efficient way to deal with this particular problem is to install a set of litter bins in the street, the principled person would be happy to do this: the ideologue would remain fixated on their One True Way.

(Or could this just be another of those irregular verbs? I am principled, you are an ideologue, he/she is a raving nutter.)

Adam Lipkin #23: Oh, no doubt!

On other parts of this discussion: I agree with Patrick that ideology (in the sense of a coherent explanatory system)is useful, and necessary. The problem is always not mistaking the map for the territory -- the map will get you where you want to go but it won't show you the kind of detail you may need to get there.

Fragano Ledgister #3: the map is not the territory

That is one of my favorite quotes from the movie "Ronin", spoken by Sam (played by Robert DeNiro).

"Things should be made as simple as possible, but no simpler." -- A. Einstein

Adam Lipkin @ 23... Not only is the Devil in the Details

Combine that with the well known fact that the Road to Hell is paved with Good Intentions is paved with tiny stepping stones?

Debra Doyle @22 reminded me of Joel Spolsky's Law of Leaky Abstractions, which is another way of looking at this entire subject: from the perspective of an engineer who has to work with the models because the details are hidden.

I think Bruce Cohen's point about the elegance (or perhaps tractability/computability might be better nowadays) of models being a crucial ingredient in their acceptance can be extended: elegance and tractability become increasingly important as the testability of models declines.

Which means that the less you know about a subject, or the more difficult it is to tease meaning out of the facts on the grounds, the more attractive a simplified picture of the world will look.

Then apply positive feedback.

Adam Lipkin (#23): God, too, is in the Details.
Very true, unless the "Details" are arbitrary and imposed by an institution -- then, they revert to the Devil again.

Chris #24: - whether the Bush administration believed their rhetoric or not, they had to sell the concept to others. The model of we'll bring democracy to Iraq was a fairly easy sell to Americans.

It was an even easier sell than that, considering that wasn't the justification they used to start the war.

#32 paul: I think you're right about this. Thomas Sowell* talked about this a lot in a couple of his books, particularly _Knowledge and Decisions_; when your decisions cause some kind of immediate feedback, they get judged on how good they are. When there's no feedback, your decisions get judged on how good they look.

* It's important to distinguish Thomas Sowell the Republican hack who writes a newspaper column from Thomas Sowell the deep and insightful thinker who writes some really fascinating books. Same guy, different hats, but radically different quality of thought.

Earl Cooley III #28: The saying's been around longer than that. A quick search shows that it was originally uttered by Korzybski.

Iain Coleman, #26: I think your last sentence is what nails it.

Ah ha! Fragano just beat me to it regarding Korzybski. Maybe I am the only one here who's actually read "Science and Sanity"?

(Bearing in mind I'm just making this up as I go along:)
The funny thing about models is that they are abstractions of "reality". Therefore they leave information out. Hence, as has been said above, they are always wrong, but sometimes useful, i.e. their symbols and structure accord closer to reality.

At the moment the biggest topic I see this on is climate change. I've lost count of the number of denialists and simply ill-informed people I have talked to who are so dismissive of models. Yet models are used to do everything from design cars to airplanes to help with new medicine research.

My current place of employment is run by bad managers, and they seem to like making great big complex models on whiteboards, and then fouling things up bigtime. This is because in making their models, they have not checked out the small stuff adequately and thus abstract the wrong information to the model itself. They do this partly because they do think they are competent is various areas, so of course they can be competent in this....

It's reached the stage where all 3 members of the technical department, who know quite a bit about how things work and how they should be modelled, are just leaving them to it.

For more about models and survival, see Deep Survival by Gonsalez. This is a fascinating book, and I think it's beyond cool that it's online for free.

For models and predicting the human world, see Expert Political Judgement by Tetlock. From what I've read about it, it's about a cruel experiment of asking pundits to make predictions, then checking to see if anyone got things right. IIRC, pundits with less comprehensive theories are more likely to be accurate. I had no idea it had been put up for free.

re Nancy @39,

Tetlock's talk at the Long Now Foundation's ongoing lecture series* is summarized here and available for listening here (scroll to Why Foxes Are Better Forecasters Than Hedgehogs. Also, yay for LN for providing an Ogg version**). It was one of the best lectures this year, and that's competing with people like Vinge.

What he'd done is get several hundred experts and asked them to both give forecasts and say how confident they were in their forecasts. He also classified the experts as Hedgehogs (one theory explains all) or Foxes (borrow from multiple theories).

Foxes were much more accurate, but foxes- with all their "Yes, but also consider that's" don't get onto Fox.

He mentioned that several of his experts are currently in government positions. Are they hedgehogs? How accurate and confident were they previously? He can't say due to his study's confidentiality agreement.

---------
* If you live in the San Francisco Bay Area, I highly recommend the LN's lecture series.

** an open source and higher quality audio type.

guthrie @ 38... Maybe I am the only one here who's actually read "Science and Sanity"?

I wouldn't know, but I recognized the line from A.E. van Vogt's use of it in his null-A books.

Richard Brandt @ 34 - I hate to quibble, but "bring democracy to Iraq" was in fact one of the stated reasons for the war. There were a number of reasons stated.

(Please don't assume I support Bush - the last Bush I voted for was in 1988)

Re #40: I agree in principle with use of the Ogg format, but unfortunately it won't work in my iPod... really that's a failing of Apple rather than of Ogg, but it's still a nuisance.

One of the really serious mistakes that's commonly made in using models is based on "physics envy"*: where the model-builder fails to realize that not all models can or should be made mathematically precise. Psychologists and economists fall for this one all the time. And anyone who bases sociological and political models on those psychological and economic models is going to be still farther from a working model. Even where you can build a precise numeric model, you have to remember the error bars on your data: precisely inaccurate data lead to precisely incorrect conclusions.

* Where a professional in a less rigorous field than physics thinks that the rigor comes from the use of numerical models, so throwing numbers into fuzzy** models will make them rigorous.

** Fuzzy isn't a negative term, it just means that your models don't map well to models with mathematical metrics and infinitely precise category boundaries. Just don't try to use fuzzy models and descriptions to get out results more precise than they can really give.

paul @ 32

And that's another serious misuse of models: choosing one model over another because it's computable or easily worked with, even if it's not as good a model in the areas you need.

Stephen Hawking, who isn't a dumb guy, once astonished everyone by solving a wavefunction for the entire universe that showed that time went forward from the Big Bang to some median time, then reversed back to the BB, which became a Big Crunch to any observer because their arrow of time was reversed with respect to the BB. Many people oohed and ahed until one less reverent soul pointed out that the model he was using was matter and energy free, isotropic, and homogeneous, removing several classes of physical action that overwhelmed the effects of the metric all by itself. The model did have the advantage that it could be solved with the mathematical tools that Hawking had available at the time. Yea, even the best of us go down the rabbit hole on occasion.

Dan @40,

They also have MP3 versions. (LN originally only had MP3, iirc) For a speech it might not make a difference- Ogg was optimized for music, and so for speech it might not be better than mp3. But then some of LN's lectures include music.

And what was Apple's model that they think that no one cares that much about better quality codecs? Given the existence of people who buy $10k stereo systems (at the very low end of high end) and gold-coated CDs, did Apple simply decide to ignore them? Or does Apple include a better but proprietary music codec?

Fragano Ledgister #36: The saying's been around longer than that. A quick search shows that it was originally uttered by Korzybski.

I knew that. I just didn't figure out a clever way to slop that gobbet of information into my post. It was just to show that there are more reasons to like "Ronin" than the bang-bang car chases.

Bruce Cohen (SpeakerToManagers) #44: A lot of political scientists suffer from 'economics envy' (a quick glance through the APSR or the AJPs will demonstrate that -- I might even go so far as to suggest that there's a degree of 'sigma competition' [i.e., mine's bigger than your's] going on).

I think, though that rather than 'physics envy' economists suffer from 'mathematics envy'. Formal theory (or positive political theory as not-so-closet econometrists like Neal Beck or Jonathan Katz are wont to call it) in both political science and economics is mathematics, as is game theory which is now occupying a large chunk of political science.

The problem, of course, is that while a formal model can produce a good rough guide to reality, it's never going to be an exact guide to human reality because human beings are not the two-dimensional creatures they have to be for formal modelling to work (that is, we have more motivations than the micro-economic ones). And we're also not always rational actors (I seem to recall that surveys done in the South in the 1940s showed that whites realised that segregation was a drag on economic growth, and liked it that way).

One problem with models as information filters: they usually decide early on as to which effects are first-order effects (that will make a difference to the outcome).

That makes implicit assumptions to ignore second/third order effects as being too small to affect the big picture
(and that they're independent, so that their cumulative effect is not all weighted in the same direction, and sum to something that is neglibile compared to a first-order effect).

Once those assumptions are set in stone, they are rarely re-examined, even if the model is being used in very different circumstances (which might be as simple as used sometime later than the original...) -- and they need to be.

Fragano Ledgister @ 48

That's even worse. Physics models systems in the real world using mathematics. It seems reasonable to describe economics in the same way. Mathematics doesn't model anything; it's sui generis and self-describing. For economics to be like mathematics we would have to say that it described some set of Platonic ideals rather than things in the real world. That strikes me as either useless, or if you ignore the ontological dissonance, quite dangerous to anyone trying to use the results.

As a dyed-in-the-wool constructivist (since last Tuesday), I have to quibble with ``the model isn't the world''. That's true , but pointless (and to my mind, meaningless). We can never know ``the world''; all we ever have are models. We can never know if our models correspond to ``the world'', all we ever know is that this model works (or, more commonly, not) in our experience, and if it is consonant with our collection of other models. However, it is perfectly possible to place the model above the experience, neatly inverting the chain of knowledge.

That's not always wrong -- at least to some degree. If I have a passing jaundice, it makes perfect sense for me to refuse to believe that everything is yellow tinged. However, if my model of the universe is that God made it all in six days and nights, resting on the seventh, then subordinating experience to that isn't a great idea.

Models don't get in the way of our knowing objective reality; models can get in the way of our understanding subjective experience.

Kathryn from Sunnyvale (#46): Apple supplies several additional codecs.

AAC (Advanced Audio Coding, part of the MPEG-4 spec) is the better-than-MP3 lossy codec which is the current default for iTunes when ripping CDs, and (combined with Apple's DRM) for songs sold on the iTunes Store. (Other gear also supports AAC, including Sony Ericsson cell phones and the Zune.)

iTunes and the iPod also support "Apple Lossless" (roughly 50% of the size of an uncompressed audio file), WAV, and AIFF.

guthrie @ 38; Yep - read it in college as a freshman. I was *so* stoked to see it on the shelf after having read van Vogt. Much less stoked after grinding through all of it, but I've never lost "the map is not the territory, the word is not the thing described".

(slightly off-topic, but on-topic to the MP3 player discussion)

For those wishing that their iPod could use other common audio formats, check out Rockbox, an open source replacement firmware for most models of iPod and some other portable MP3 players. It installs such that your player can boot into either the original firmware or Rockbox firmware.

You will need to keep the original firmware around because Rockbox doesn't do video (yet) or DRM-protected audio (probably never will), and doesn't do non-DRM'd AAC as well as Apple's firmware.

It does take a modest amount of computer literacy to install and configure at the moment, and installing it may void your warranty. I have it installed in my non-apple MP3 player (Cowon iAudio X5) and have been pretty happy with it (it's better than the original Cowon firmware). But then again I really despise the overall iPod and iTunes interface and way of doing things, one of the reasons why I didn't buy an iPod in the first place.

linnen: There is, "the map is not the territory"

The crap is not the lavatory.

The pad is not the upholstery.

The joke is not the humour.

Especially that one.

Yet another kicker is 45 and 51 together:

Not only does the model condition your understanding of reality, but in many social-science or economic situations it creates your reality. If you have enough power to treat people according to your model (without too much immediate reality-check feedback), then by golly people will start acting in the ways your model allows for. Small examples include pretty much every spectacularly ill- or well-managed place of employment.

Large examples: the interlocking mobilizations that started ww1 and on several occasions almost started ww3. Or the Bush administration, which has created pretty much out of thin air an America standing alone, beseiged by the unshriven hordes...

(About 20 years ago I wrote a paper about the dangers inherent in easily-computable models of human behavior, but my initial audience got so hung up on the idea that such things could even exist that it didn't go anywhere.)

The LDAP is not the directory

Bruce Cohen #50: What econometrists seem to aspire to is the purity and elegance of mathematical proofs. Some people in both economics and political science get stuck in the beauty of mathematical models and forget that human beings are not the same sorts of things as physical particles or pure numbers. That's not much of a problem if you're talking game theory, but gets to be if you're talking about formal theories that are more concerned at the elegance of the solution than what people actually do.

guthrie @ 38... Maybe I am the only one here who's actually read "Science and Sanity"?

Not quite, though (age 16-ish, c. 1945) I didn't understand it nearly as well as I had Hayakawa ("Language in Thought and Action"), and probably should (but probably won't *sigh*) try re-reading.
Both of them, come to think on't.

But I've thought of that Basic Concept every time I've seen the phrase "Flag-desecration amendment". Treatment of a symbol/map does nothing to affect any reality it might represent.

Me; (#63)

There may well have been better works, later, but (IMHO) the Hayakawa one still should be Required Reading for all teen-agers who aspire to become reasonably-rational human beings. Or even just more rational than their parents, which might be a more common attitude.

#60 paul:

Further, if your model becomes widespread among decisionmakers, then the model really does become part of the territory instead of the map. This affects stuff like how the Fed tries to be unpredictable and inscrutable, or what gambling games a casino ought to offer. (Once card counting strategies are widely known, the casino has to change its behavior or lose money on blackjack.)

Come to think of it, I bet there have been times when a literal flawed map held by many different sides of some fight has basically shaped the fight.

Frangano(#62) beat me to it in part. It's one of the great failures of many modern economists and the politicians who believe them that emotional, partly informed human beings will always make the most rational decision possible, at least if they're encouraged to do so by economic incentives or disincentives. This is reflected in the health care debate. Those who believe that the market is virtually infallible think that if you just penalize those whose use of the medical system isn't optimal financially by making them pay more in one way or the other that they will shop around. They'll evaluate doctors and hospitals just like they would televisions and cars. In their dreams. Everyone who knows they're going to need a cardiologist tomorrow raise your hand. Very often we don't know what treatment or kind of care we are going to need. And when we do find out what we need we are often not in the most rational state of mind nor do we have the time to just carefully evaluate different doctors or hospitals. We take the advice of the doctor who diagnoses our problem and move on. But the models (Yeah, I got back to it.) that these economists and politicians choose to believe in insist that every decision that involves money will have an economic answer that counts on the rational actor model working. And they aren't going to abandon that model for anything.

The crap is not the suppository.

Not only does the model condition your understanding of reality, but in many social-science or economic situations it creates your reality.

''That's not the way the world really works anymore,'' he continued. ''We're an empire now, and when we act, we create our own reality. And while you're studying that reality -- judiciously, as you will -- we'll act again, creating other new realities, which you can study too, and that's how things will sort out. We're history's actors . . . and you, all of you, will be left to just study what we do.''

An unnamed Bush Administration aide said that to Ron Suskind in 2004. Subsequent events have demonstrated - at least to the reality-based community - the hazards of acting on reality without first studying it.

albatross observes: "...the widespread idea in management that you should be able to manage things whose details you don’t understand all that well."

That sounds to me like the fundamental animating principle of the Harvard School of Business. Everything and everyone in the enterprise reduces to line items in the financial model, so a good manager needs only to be a specialist in finance, c.f. the recent particle about ERS packages. Everything else can be outsourced to interchangeable peasants.

One idea I think is related (though I have trouble describing the relation I think I see) is that much of what goes wrong in management, and in particular with managers blinded by an ideology like the ones albatross is talking about, is a failure to recognize that management is about organizing things, while leadership is about organizing people. The other side of "I don't need to know the details" is "You only need to know what I tell you." That's a basically dehumanizing posture, I think. It says, "You're not a person requiring leadership. You're a thing to be managed. Don't bother me with details. Shut up and do what I tell you, or I'll replace you with someone who will."

Pick any of your favorite 'isms from the set in the original post. Scratch the surface and you're likely to find that the phenomenon of relying on willful ignorance is basically founded in efforts to turn thinking feeling human people into cogs in an elaborate machine.

Jim Satterfield, Fragano Ledgister,

OK, so why haven't those people heard of Herbert Simon's theory of bounded rationality? I know, he's pretty obscure, he only won a Nobel Prize in Economics. Simon explicitly stated that humans are not rational, optimizing actors, and they are always making decisions with insufficient information, and can't optimize the results of their actions to get the best results; they best they can do is get satisfactory results.

Of course, I know of him primarily because of his work on the foundations of complexity theory, and on artificial intelligence. He was a one man scientific revolution.

paul @#60, albatross @ #65: Specifically, it's social contexts where the model can feed back to the reality. Presumably that's because social interaction depends heavily on people negotiating agreement (or boundaries) for their world-views.

That actually might be why the map/territory confusion is so common -- so much of our attention does go to social matters....

paul: If you have enough power to treat people according to your model ... people will start acting in the ways your model allows for.

Note that in this context, even an economic or political model is manifested through, essentially, communication backed by power.

Thanks to several people for the source references! I'd heard and appreciated a variation of the phrase ("the map is not the terrain"), but never knew where it came from.

Serge: The snowclone is not the meme. ;-}

j h woodyatt @ 69

Any model that includes people provides a great temptation to consider people as components of the model, plug-replaceable and fungible. That is not an attitude that's compatible with the attributes necessary to make a good leader.

One of my very favorite quote is something I saw in an article about the effect of the Vietnam War on the US military. The article made the point that the McNamara regime in DOD changed the nature of the hierarchy in the military from leadership to management, and then stated that it failed to create a viable military command structure that way because "You cannot manage psople to their deaths."

Back in the 70s or so whenever the health insurance thing about making a profit with hospitals a bit of my thoughts went, Ut-oh, something is farked. I was reassured that patients would get care no matter what so "don't worry my little head."

The hospital I candy striped/transportation orderly'd at was closed recently because it wasn't profitable enough for HCA. It does still have an ER, but that leaves me queasy because if you have something bad, you'll have to be transported again once they stabilize you.

Paula Leiberman,

We have a severe shortage all over the rural areas of the western United States of both practicing doctors and hospital beds, because you just can't make a lot of money at it; not enough people, not steady enough a patient stream, too much capital outlay for the return, etc. Even offering bountys to doctors to spend a year or two in the boonies isn't working.

Oh, and if you think that's bad, I was talking to my son the other day about the mental health situation in New Orleans. They have half the population pre-Katrina, twice the suicide rate, and less than 10% of the trained mental health workers of all kinds. Nobody seems to want the workload even for the bounty money they're offering. My son is trying to set up a link between his lab in Baton Rouge and a clinic in NO, so he can get 6 or 7 grad students to do basic screening and therapy under a clinic license umbrella. The clinic he's been working with has one doctor, who's retiring in two months.

The profit motive is much overrated as a means for solving problems. And the public sector not doing very well either.

Oops, I'm sorry. That last post should have been addressed to Paula Helm Murray. Really, I know that Paulas aren't fungible; I was remiss in not double checking that comment.

Details quotation (Adam Lipkin @23): — according to this source

Bartlett's Familiar Quotations says: "God is in the details: A popular aphorism with the architect Ludwig Mies van der Rohe and the art historian Aby Warburg; attributed [Le bon Dieu est dan le detail] to Gustave Flaubert but without verification."
William Safire mentioned it in a New York Times article on July 30, 1989. He spoke with the editor of Bartlett's, who said: "We've had little success with 'God (or 'the Devil') is in the details,'" says Mr. Kaplan. "We know that Mies van der Rohe used it in discussing architecture; Flaubert has been suggested, but nobody can find it in his writings. I think it may come from John Ruskin, because it sounds like him on the subject of workmanship, but we need the specific citation."

albatross @ 65: "if your model becomes widespread among decisionmakers, then the model really does become part of the territory instead of the map.". Yes, there can be feedback in these things. It's the reason I didn't completely dismiss the whole of the "we aren't reality-based" idea from a few years back. You can get self-fulfilling aspects of life if everyone (or enough of the people with power & influence) believes and behaves as if certain things are true or untrue, but then there's the time when it all does run into a brute physical fact, or up against a different group which has different beliefs and behaviour.

Bruce Cohen @ 70:

I've recently become aware that there are economists who know about Simon, and of the limits to rationality, and who are even approaching the field as an exploratory science: i.e., "People clearly aren't always the simple, all-knowing/all-rational selfish automotons of neoclassical theory; let's see if we can find out how they actually behave." And so they are talking to and working with experimental psychologists and sociologists, and apparently making some progress. Here's an example of sorts in the form of a discussion about people's attitudes towards welfare. (I discovered this via Cosma Shalizi's long but fascinating post on what's wrong with "econophysics".)

One of the interesting approaches seems to be using game theory as an exploratory tool. That is, rather than assume participants are all selfish, rational maximizers and using game theory to predict how they'll behave, you set up game theory scenarios, invite real people to play them, and then look at what they do and try to work out what underlying strategies or motivations they might be using.

Building a good simplifying sociopolitical system without understanding the underlying details. That would be like expecting, say, an electrical engineer to design a chemical reactor.

The devil is always, always, ALWAYS in the details. You have to know what you can safely ignore.

Fragano @ 48:
I think, though that rather than 'physics envy' economists suffer from 'mathematics envy'. Formal theory (or positive political theory as not-so-closet econometrists like Neal Beck or Jonathan Katz are wont to call it) in both political science and economics is mathematics, as is game theory which is now occupying a large chunk of political science.

I remember reading a review in Nature of an economics book (back sometime in the 90s) where the reviewer suggested that the problem with economics (or at least the kind of economics exemplified by the book in question) was that it was fundamentally a form of "applied mathematics" rather than anything like a science -- mathematical rigor trumping any sense of empirical fidelity.

"Physics envy" doesn't just take the form of using mathematical models, though. It also involves the hope (or delusion) that one can find clear, simple principles which are easy to understand and universal within the field under study: "iron laws" of [insert your field here], to match things like Newton's Laws of Motion, conservation of energy, etc.

Epacris #76:

Keynes talked about this kind of very unstable situation, where everyone's model incorporates estimated models of everyone else. (This describes financial markets pretty well.) The interesting thing is, these models can globally change very quickly (making a speculative bubble burst) because the whole market is dependent on this recursive set of models (I model you modeling me modeling you...).

I think something similar happens in the transition from peaceable change of government to civil war. Part of your decision about whether or not to try a coup is based on your evaluation of whether anyone else will go along, and their evaluations of whether they should go along has to do with whether they think others will go along. A coup won't work here because too many people will absolutely not go along, partly because most everyone expects that many others will also refuse to go along. And yet, let that belief change, and a coup becomes possible, as the announcement from the new Presidente Para Vida is greeted by people staying home, waiting to see what happens, rather than by people taking to the streets or potshotting soldiers attempting to impose martial law.

#72 Bruce and others:

The problem is, if you're going to teach management in a generalized way (rather than specifically management of engineers building suspension bridges in Oregon, say) you need to be able to abstract out the parts of the management tasks that are generalizeable. But some part of being a good manager is knowing your employees and the business you're in, and how those vary from the "default" model you'd carry.

Bruce Cohen @ 72: "You cannot manage people to their deaths."

"The trick is to take a break as soon as you see a bright light and hear dead relatives beckon." -- PHB

albatross @ 81

The trick to teaching management is the same as for any other subject which is based on abstract principles but has a strong empirical component; engineering, for instance, or political science. There has to be a practical portion of the curriculum in which students become in effect apprentices, learning how the principles are applied in the wild. My sense, without doing any research to back it up, is that less of this is done in most fields than has been done in the past, at least in the US.

Joel Polowin @ 82

Just be sure not to go towards that smoky red ... oh, well.

I am surprised that no one has mentioned Plato's 'Allegory of the Cave' yet.

Terry Karney @55 Thanks.

Joel Polowin #82: Sacrifice: Your Role may be Thankless, but if You're Willing to Give it Your All, You Just might Bring Success to Those Who Outlast You.

Grumble. Marxists, for all our failures, are always trying to fit their theories to new facts, even if some have a tendency to quote Marx or Trotsky or Lenin in the same way as Christians quote the Bible.

Iain, #26: absence of ideology is an ideology too; the great swindle of the last thirty years has been making people belief that you can have politics without ideology.

Martin Wisse @ 87... Marxists, for all our failures, are always trying to fit their theories to new facts, even if some have a tendency to quote Marx

"Outside of a dog, a book is man's best friend. Inside of a dog it's too dark to read."

Bruce Cohen #70: To be fair, rational choice (public choice, social choice) theory isn't a single simple thing.

Some rational choice theorists, such as Amartya Sen, make the point that human beings aren't simple utility maximisers but have more complex motivations not all of which are rational. Still, it's much easier to create a mathematical model if you assume that human beings are all driven by microeconomic motivations (i.e., that we all want to do well in the market) providing you define almost every kind of interaction as a market. Human beings, of course, have many more motivations -- and they can't always be caught by a mathematical model.

The thing is, there are interactions that can be modelled (how deals get made in Congress, how peasants make choices about cultivation, how businesses respond when customers go bankrupt). They just aren't the full range of political or economic interactions. And they can't explain the full range of human motivations.

Nonetheless, because models (for example, the Prisoners' Dilemma) do have some predictive power, the temptation is to apply them to everything. Now, I'm a state-and-society type, so I want to look at interactions in terms of things like values, beliefs, uses of power and so on that aren't always (or even often) quantifiable or reducible to simple microeconomic motivations. Human beings are perverse (in an economic sense) because our motives may often run contrary to our economic good.

Peter Erwin #79: Interesting.

Economists do have real laws (supply and demand in microeconomics) that explain real-world events (how prices are created). They can't explain, for example, choices that have little to do with price (ask most Americans if they'd like a nice bunny stew, and step back quickly).

There are some statements in political science that qualify as laws (or at any rate, we call them that) such as Duverger's Law on the relationship of the single-member plurality constituency system and the two-party system.

God is in the details:

"I don't see why my belief in a God you can't accept is any more rarefied than Mike's vision of the atom as a hole-inside-a-hole-through-a-hole. I expect that in the long run, when we get right down to the fundamental stuff of the universe, we'll find that there's nothing there at all – just no-things moving no-place through no-time. On the day that happens, I'll have God and you will not – otherwise there'll be no difference between us."
(James Blish, A Case of Conscience)

Serge #88: "All social life is essentially practical. All mysteries which lead theory to mysticism find their rational solution in human practice and in the comprehension of this practice."

Fragano @ 92... Harpo?

Serge @ 93: Honk!

To me, the greatest lesson of game theory, and especially of the Prisoner's Dilemma, is that it's really easy to lock the solution out of your problem box.

(The biological solution to the PD can be summed up as "figure out who you can trust". That's exactly what's forbidden in the classical form of the dilemma....)

Good to see some more people have read "Science and Sanity". General Semantics was the name of teh orientation method (I don't think it counts as a philosophy) and I think the basics of it could indeed be tought to people at school.

Being somewhat cynical, I thought you could manage people to death, I'm sure it happens all the time in companies with a poor health and safety record (Telling you all about the place I work would take 5,000 words at least), or else in underfunded hospitals.

Serge #93: Karl.

David Harmon #95: The problem, certainly as we learn it in political science, is how to cooperate without communicating (that is to say, creating trust). That requires, at least according to Axelrod, iteration.

ask most Americans if they'd like a nice bunny stew

Damn you Fragano @ 90! Damn you! Now you've got me craving the rabbit pies our local baker used to do back when I was a wee peerie lad, and that I've never found the like of since. Gaaaah!

Disclaimer: Not an American

guthrie @ 96

I probably misquoted the original, but the meaning was, you can't manage people into sacrificing their lives for others. You can blackmail them or extort them into it, but you can't make them want to do it. Leaders can; it's not necessarily a good thing, but sometimes very useful for the leader. A manager who does that typically will get fragged: a grenade will be thrown into his bed while he's asleep or in the bathroom.*

* I've never heard of a female officer getting fragged, though it might have happened, hence "he".

SKapusniak #99: Rabbit pie sounds tasty to me (but I'm not an American either, although I do live in the states).

It is. Jugged hare is good too.

SKapusniak @ 99

Yeah, my mouth is watering too, and I am an American.

Bruce #100:

ISTM that most really productive, creative people are not primarily by money or even by other non-monetary rewards like reputation. They like what they're doing. I think good managers in general find a way to hook into your internal motivations, rather than your fear of punishment and desire for reward. One reason is that punishment/reward schemes are often gamed by people whose primary motivation is avoiding the punishment and gaining the reward. (Man, that looks *way* more convoluted in print than it did in my head.)

Fragano #98: Yes! Axelrod's work is cool. Though he's often talking about iterated prisoner's dilemma setups where you have to deal with many people to succeed.

One of the creepier parts of that was the discussion of stable discrimination in _The Evolution of Cooperation_. Skin color can be a form of communication, which can make a strategy of screwing minorities stable in some circumstances.

I keep thinking a lot of the discussions of rationality really involve what people have experienced. For interactions that occur a lot, the iterated prisoners dilemma sort of setup is probably a pretty good model for how your strategies adapt over time, as being too nice or too mean in the wrong circumstances just doesn't pay off. People learn from experience even when they can't explain what they know, so if they try a strategy that doesn't work, they're likely to realize it sooner or later.

Fragano @ 90:
ask most Americans if they'd like a nice bunny stew, and step back quickly

I suspect the reaction might depend a little on where in the US you asked.

For what it's worth, my main US-produced cookbook (Mark Bittman's modestly titled How to Cook Everything) has a recipe for "Marinated and Stewed Rabbit," and includes in its discussion of "Game Meats" this observation: "Thus rabbit, which is raised in a manner similar to chicken and can be found in the freezer compartment of most supermarkets (and the fresh meat department of many), is lumped into the same category as rattlesnake or camel..."

It's certainly a bit unusual, but I don't think most Americans consider it a true non-food, the way they would dog or rat. (E.g., Bittman's book, despite its title, does not have recipes for dog or rat.)

A bit OT, but since we are talking rabbit stew the follow exchange was brought to mind. Ahem;

Rabbit hunting season!

Duck hunting season!

Rabbit season!

... Rabbit season!

Duck season! Fire!

BLAM!!

You're despicable!

albatross @ 104

You won't get any argument from me on any of that. I will say that, based on the criterion that a good manager does that sort of thing, that there aren't many good managers. Also, reward schemes (aka "compensation policies") frequently reward the wrong things, hilarity and unintended consequences ensue.

linnen @ 107:

Curse you... made me go dig up my Looney Tunes DVD and watch the cartoon again!

(But now I'm wondering: why does Bugs Bunny have a cookbook entitled 1000 Ways to Cook a Rabbit in his hole?)

A "hobby" of mine is to cook for friends things they've not had before. From different groups at different times, everyone wanted to try pigeon, most of them were okay with rabbit or crab, but most of them did't like the idea of eel.

(I get the idea that those who don't want to eat rabbit do so because it's cute; crabs or eels are non-food because they're rugose and sqaumous)

why does Bugs Bunny have a cookbook entitled 1000 Ways to Cook a Rabbit in his hole?

For the same reason I keep a copy of To Serve Man handy in the kitchen?

I've had skunk and mountain oysters (just about anything is palatable if your common sense is reinforced with Everclear strawberry punch) but I've never had a chance to try rattlesnake.

Albatross #105: Well, you're dealing with complex interactions, and those involve many actors. Axelrod's a good summary of the Prisoners' Dilemma, with some interesting examples.

Neil @ 110: From different groups at different times, everyone wanted to try pigeon, most of them were okay with rabbit or crab, but most of them did't like the idea of eel.

On my last trip to the UK, I was determined to have eel in Ely, and I did--as an appetizer for an entree of rabbit stew.

Yum.

Neil Wilcox @ 110

Interesting; I don't think I've ever lived anywhere for any length of time in the US where crab wasn't considered a delicacy. I grew up catching the little buggers (well, they were little there) and cooking them on the beach. And the crabs where I live now are much better tasting (and much bigger). Eels, on the other hand, not so much. I've eaten them and I'm not enthused. Most places they're not verboten, but they're not high on the menu.

On the gripping hand, everywhere I've lived but where I grew up (and maybe for 50 miles around at the most) the very thought of scrapple makes people ill. For me, at least when I can find it and hide it from my wife (she hates the idea), it's what's for breakfast.