By Request: Counterfactuals

JoeJoeJoe asks:

You must have had a subscription to What if...? comics when you were a kid. You sure do love counterfactuals. Maybe British colonialism would have done for North America what it did for Africa? Maybe the United States in the latter part of the 20th century more like South Africa than Canada. Maybe if the American revolution happens in 1915 instead of 1775 we find ourselves allied with German nationalists in the late 1930s. Maybe in the 25th century Isiah Thomas will be viewed as the best Knicks GM in history.

Assignment desk: Explain the value of counterfactuals and your affection for them.

My affection for counterfactuals and my sense of their value derives from when I took Richard Heck's class on "Realism and Anti-Realism" and we did a unit on counterfactuals. For that segment of the class we were assigned David Lewis' On the Plurality of Worlds where I was first exposed to the argument (much less controversial than Lewis' conclusions about the metaphysical status of modal claims) that there's an intimate link between talk about causation and talk about counterfactuals. My thoughts on this matter were further influenced by when I took the final course Robert Nozick ever taught which was on the philosophy of history (some of Nozick's thoughts on the matter are reflected in Invariances).

At any rate, among historians talk of counterfactuals is in a bad air. But philosophers generally find it pretty uncontroversial to say that (modulo certain complications) causal claims can generally be translated into causal claims. So you might say that the Casey v. Planned Parenthood opinion came out the way it did because David Souter turned out to be a moderate, rather than a conservative, replacement for William Brennan. Alternatively, you might say "if Souter had turned out to be as conservative as Ted Kennedy feared, the Casey decisions would have gutted constitutional protection of abortion rights." Now you can't say these things are precisely the same, because maybe Souter would have been a wingnut who got run over by a bus the day after his confirmation etc. etc. etc. but in a commonsense way the "What If?" question is just a vivid way of thinking about causal claims.

Comments

Contrafactuals? Who gives a shit.

What a waste of time and brain power.

And if my grandmother had balls she'd be my grandfather.

There's a contrafactual.

Posted by calipygian | July 6, 2008 11:34 AM

The reason historians don't like counterfactuals is because on most interesting issues, it's already impossibly hard to explain the actual causes of historical events.

You're really stacking the deck by talking about the Supreme Court instead of the questions posed by joe^3.

You have isolated a case where it was possible that Souter would be a moderate on abortion, and possible he would be a conservative, and we can see an actual fork in the historical road.

But large- and medium- and, basically, all non-micro scale historical events, this doesn't work. Think of the wide variety of causes of the American revolution - from economic to ideological to cultural to particular histories of the individuals involved - and precisely which factors would have to change in what ways in order for there not to have been a revolution. I don't have any idea how to sort those things out - again, it's impossibly hard to isolate actual causes of things that actually happened, let alone work out the calculus of

In general, I think this is why historians find much analytic philosophy to be, at best, irrelevant to our field. Most all of the interesting questions have to deal with the irreducible complexity of human events, and you've assumed it away to make a point about a tiny subset of human events.

Posted by DivGuy | July 6, 2008 11:43 AM

Yeah, an inference and its contrapositive are formally equivalent-- but does the observation 'This, arguably, could have been not-white and therefore would not have been a swan' really say anything useful about -that- particular swan?

Posted by MattF | July 6, 2008 11:47 AM

I don't see that counter-factuals are totally out of style in history. Any number of historians are still working on the biggest counter-factual of American history: "Why no socialism in the U.S."

Counter-factuals can, and are, often used by historians as way of interogating the historical record to find why things happened the way they did rather than some other way. The idea is not to come up with some alternative history, but to identify which elements were critical, the sine qua non of an event, and which elements incidental.

Posted by Cap and Gown | July 6, 2008 11:55 AM

Counterfactuals are to political discourse, what Taylor series calculus is to engineering... mental gymnastics with little practical use. It's exercise for the brain.

JoeJoeJoe's frustration is very well placed. What if... we hadn't revolted in 1776? and waited til 1820? 1865? What if we hadn't killed off the indigenous Americans by 1876?

What if Europeans HAD killed off all the indigenous Africans by 1876, similar to what had happened in America?

What if we HAD found WMD's in Iraq? Would that have justified a trillion dollar war? What if George W. Bush HAD read the daily briefing Aug 6, 2001 and averted 9-11? And on and on.

Silliness. The value is only in understanding that hypotheticals are endless and pointless to pursue outside of the ivory tower.

Posted by Tony | July 6, 2008 12:01 PM

DivGuy,

Does that mean that historians never talk about causes either?

Posted by Pithlord | July 6, 2008 12:02 PM

The study of counterfactuals would only be interesting if you could, somehow, exhaustively catalog ALL the possible alternative outcomes, and then examine patterns in their distribution.

Sort of like in quantum mechanics, where you can calculate the path of a particle by first tracing all possible paths, then eliminating those that cancel out.

As it is, investigators of counterfactuals only consider one or at best two different alternative worlds; and the way this selection is done stacks the deck with regard to whatever argument they want to make about the actual course of events.

It's really not much better than astrology, then, since it's just another way of using scientific-sounding arguments to disguise a heaping pile of guesswork.

Posted by lampwick | July 6, 2008 12:03 PM

Counterfactuals are to political discourse, what Taylor series calculus is to engineering... mental gymnastics with little practical use. It's exercise for the brain.

JoeJoeJoe's frustration is very well placed. What if... we hadn't revolted in 1776? and waited til 1820? 1865? What if we hadn't killed off the indigenous Americans by 1876?

What if Europeans HAD killed off all the indigenous Africans by 1876, similar to what had happened in America?

What if we HAD found WMD's in Iraq? Would that have justified a trillion dollar war? What if George W. Bush HAD read the daily briefing Aug 6, 2001 and averted 9-11? And on and on.

Silliness. The value is only in understanding that hypotheticals are endless and pointless to pursue outside of the ivory tower.

Posted by Tony | July 6, 2008 12:04 PM

Counterfactual reasoning is irrevocably linked to formulating causal statements. The definitive and most formal statement of this, at least in my field, is this:

http://www.jstor.org/pss/2010470

To say that A is truly causal, you must both hypothesize and attempt to estimate, what the outcome would have been in the absence of A. This is typically unproblematic for historians, at least in my experience, since historical precedent explains 100% of the variance of future outcomes and problems of selection bias are not discussed. Among social scientists, this logic is anathema.

Posted by Marshall | July 6, 2008 12:14 PM

Does that mean that historians never talk about causes either?

No, it means that we have limited time, and we spend it on actual causes of actual events, because those are more than enough, in their irreducible complexity, to fill up the time we have to work on history.

For example, the question "why no socialism in the US?" needs to be answered by an in-depth archival study of what precise forms the political economy has taken in the US - "not-socialism" is an exceptionally poor descriptor - and an investigation into the myriad causes of this system and its perpetuation and change over time. There's little time for "what would a socialist US look like?", and that seems a waste of time to me, except perhaps for some doodling around the edges to add a little color to a real historical investigation.

Posted by DivGuy | July 6, 2008 12:16 PM

Okay Mr. Y but remember: popular sentiment as well as a significant swath of the British elites were on the side of the Confederacy in the Civil War.

Would the British have treated the indian fellows in the same bad way they treated the Indian fellows after the colonists began to spread west over the Appalachians or left that to the French and the Spanish and Mexicans?

Posted by Linus | July 6, 2008 12:19 PM

To say that A is truly causal, you must both hypothesize and attempt to estimate, what the outcome would have been in the absence of A.

But "A" is never causal on its own, but only in concert with B through 4F6, all the interrelated factors that make up human events. You can never just theorize A's absence.

Perhaps another way of putting it is that the type of highly rigorous determination of causality that you're discussing isn't meaningfully possible or useful to most historical discussions, because of the complexity involved. We're much more involved in writing plausible stories within particular historical contexts, and I think that's the way it ought to be.

Posted by DivGuy | July 6, 2008 12:27 PM

On the other hand they give you like - what? - a 30 minute tea break in England don't they?

And even their pop charts (not just the rock and r&b charts) have decent songs on them - historically, at least.

I like that soda that tastes like bubble gum too.

There was kid named Mark who moved from the UK in grade school and never lost his accent (intentionally, one guesses). He had those salt and vinegar chips before anyone.

Posted by Linus | July 6, 2008 12:32 PM

My two cents on this issue, speaking as a philosopher: It seems unhelpful for historians to express whatever worry they are trying to express by saying that they have a problem with counterfactuals. Some counterfactuals are perfectly straightforward and easy to evaluate. We use them all the time. For instance: if we had been here 10 minutes earlier, we would have caught the bus, if we had cooked the chicken longer, it would not have been pink in the middle, if I had terminated my contract with t-mobile yesterday, it would have cost me $100, and so on.

What historians seem to have a problem with are highly speculative counterfactuals, such as the one's joejoejoe mentions. I think they are right to regard discussions of such counterfactuals as not really serious history. But such discussions do not count as serious history for the same reason that a discussion of, say, exactly how many dinosaurs existed at such and such a time at such and such a place does not count as serious history: it is a debate about something which is impossible to verify in practice.

What makes problematic counterfactuals problematic is not that they are counterfactuals. It is that they are pure speculation.

Posted by Ari | July 6, 2008 12:34 PM

What makes problematic counterfactuals problematic is not that they are counterfactuals. It is that they are pure speculation.

Sure. I haven't argued that counterfactuals, in and of themsevles, are bad.

What I'm saying is that in the actually-existing field of history, the study of human events, there are a vanishingly small number of questions which are both interesting and amenable to useful counterfactual speculation.

Arguing on behalf of the counterfactual is precisely the kind of technically correct but useless to actual people and the actual world argument that makes much analytic philosophy irrelevant to people working in other fields of study.

Posted by DivGuy | July 6, 2008 12:38 PM

"...Richard Heck's class on "Realism and Anti-Realism" and we did a unit on counterfactuals. ...My thoughts on this matter were further influenced by when I took the final course Robert Nozick ever taught which was on the philosophy of history (some of Nozick's thoughts on the matter are reflected in Invariances)."

This must have been a part of the Harvard education that taught Matt the art of:

"a lot of arguing, a lot of playing with words, and a lot of learning about how to make a contribution to a discussion without a lot of factual background on the subject at hand."

And is the proposition that P is necessary itself necessary?

Posted by kafka | July 6, 2008 12:41 PM

Tony,

I dunno about in engineering maybe, but in science Taylor series are a big deal. After all, Taylor series not only allow you to numerically calculate certain functions (which I bet would be a big deal in engineering no less than science) -- although oftentimes better series do exist -- but they also allow you to propagate uncertainty ... which may seem like mental gymnastics until you actually see the results of thinking you only need to build a system with 1% tolerance when in fact the required tolerance is 2%.

Posted by DAS | July 6, 2008 12:42 PM

I got a counterfactual for y'all: what would have happened if, instead of being visciously anti-Semitic, the Nazis would have viewed the Jews of Eastern Europe as one of the many German-speaking minorities of Eastern Europe to be brought back into the German fold (at times, IIRC, the Austro-Hungarian Empire did treat the Eastern European Jews in its domain with this mindset ... so there is historical precident)?

Of course, bringing people into the German fold who didn't want to be brought into that fold did cost the Nazis resources (e.g. the "resettling" of Gottschee Germans), but probably not so much as the Death Camps did.

Posted by DAS | July 6, 2008 12:47 PM

Pretending that history is a science experiment that has interchangeable parts and is thus amenable to "thought experiments" like counterfactuals is probably the worst approach to history that exists. And that is how you get history a la Marx and Hegel. The abuse of history by philosophers is obvious to everyone outside a philosophy department. Philosophers look for patterns and structures, whereas historians generally look at the contingent, particular circumstances.

When two historical situations look similar to you, it is usually a sign that you don't know enough about them. Most philosophers can't bear this approach because they have a psychological need to fit everything into some sort of system or structure.

Posted by jim | July 6, 2008 12:53 PM

I think the status of counterfactuals in history is more complicated than DivGuy is suggesting. It's true that counterfactuals have in general had a bad rep in history, but there are counter-currents, and recently there has been some real attempt to argue for their usefulness -- IMS Neil Ferguson was in this camp.

Posted by Stephen Frug | July 6, 2008 12:58 PM

Sure. I haven't argued that counterfactuals, in and of themsevles, are bad.

I did not think that that is what you were arguing, but when someone says, "The reason historians don't like counterfactuals is..." he certainly leaves himself open to that interpretation. You can blame those who interpret you this way, and those who point out that you might be interpreted this way, and accuse them of being useless pedants who insist on making sure that every statement is "technically correct"; or you could just be a little more careful with how you state your view, and thereby avoid any misunderstandings which might make real progress in this debate more difficult.

Posted by Ari | July 6, 2008 12:59 PM

"Perhaps another way of putting it is that the type of highly rigorous determination of causality that you're discussing isn't meaningfully possible or useful to most historical discussions, because of the complexity involved. We're much more involved in writing plausible stories within particular historical contexts, and I think that's the way it ought to be."

DivGuy:

This was an intentionally dramatic over-simplification of causal processes, but the point, independent of causal interdependencies and interactions, is still the same. And, as a social scientist, I strenuously object to the epistemologically-defeatist notion that human behavior is too complex to predict and explain. Creationists/Intelligent Designers make similar claims.

Oh, and the reason that there was no socialism in the United States is due to the absence of permissive political institutions that elevate the interests of smaller, less cohesive societal groups. The counterfactual is, in the absence of majoritarian political institutions, would the US have had a thriving socialist/left-leaning party. The answer is probably yes. The British example, and the move away from a more permissive system to a majoritarian one, to stave off rising populist sentiment, is instructive here. See the work of Carles Boix.

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=159213

Posted by Marshall | July 6, 2008 1:00 PM

And, as a social scientist, I strenuously object to the epistemologically-defeatist notion that human behavior is too complex to predict and explain.

Well, it is and it isn't. For all interesting questions, we're never going to know in the way we know A=A what the various causes were. But we can definitely make informed claims and engage in careful argument as to which factors had what sorts of causal force.

That sort of argumentation and engagement, though, requires massive amounts of historical work to best describe different political economies and cultural norms and their perpetuation and change over time, and descriptions of causality must acknowledge the irreducible complexity of the questions that are being engaged. I think, in this case, a more modest model of history in which competing plausible narratives are constructed, defended, combined, engaged, and so on, is much more reasonable and useful than a model of history in which A is isolated and removed from the playing field in a thought experiment.

I am troubled by your presumption that if I think that history is irreducibly complex, I must think that human behavior can't be explained. I think that indeterminacy is absolutely central to history and human events, and thus a full and complete explanation will always be beyond our grasp, but that in no way means that there can't be better and worse explanations, and that real discussion and debate of these explanations isn't precisely what historians ought to be doing.

Posted by DivGuy | July 6, 2008 1:10 PM

So here's the state of play:

1. DivGuy, presumably a historian, gives the party line on how counter-facutals are bad, but causal explanations are good.

2. Some people with background in analytic philosophy make the non-obvious-but-true point that counterfactuals and causal explanations are logically equivalent.

3. DivGuy argues for a bit, gives up, and parts with a rude, philistine objection to the whole project of analytic philosophy on the grounds it is "useless."

No one says law, as such, is useless. Anglo-American tort law is obsessed with the problem of legal causation and, in particular, when it is necessary to show that the plaintiff's injury would not have happened but for the defendant's fault. Lewis's insights don't resolve these issues, but lawyers who understand those insights are less likely to make certain kind of conceptual mistakes. They can concentrate on the real policy disputes.

From my amateur take on historian's methodological debates, it would be truly awesome if they didn't involve errors any undergraduate philosophy student could see through.

Posted by Pithlord | July 6, 2008 1:11 PM

When I was in engineering school we used to refer to subjects like philosophy as “gut majors”. Much of this was based on observation. The hardest working and most ambitious students, often from places like China and India, would get up early to attend their chemistry, physics, math, etc., classes. At the same time the lazy students, often from places like Scarsdale and Alexandria, would drag themselves out of bed just before noon to attend their sociology, women’s studies, philosophy, etc., classes.

And now it’s time once again to examine Moore’s conundrum: is the proposition that P is necessary itself necessary?

Posted by kafka | July 6, 2008 1:11 PM

"...exactly how many dinosaurs existed at such and such a time at such and such a place does not count as serious history..."

Wait a minute, the existence of some number of dinosaurs at a specific place in time is an actual occurrence that is well within our ability to estimate to some degree of accuracy with standard scientific practices. On the other hand, asking "what if dinosaurs were still around?" would require enough speculation as to yield an answer with little value.

Posted by charles | July 6, 2008 1:15 PM

"Counterfactuals are to political discourse, what Taylor series calculus is to engineering... mental gymnastics with little practical use. It's exercise for the brain."

I would compare them more with Monte Carlo simulations, which are actually very useful.

As other have said, the problem with most historical counterfactuals is that it is surprisingly difficult with history to just determine what happened, let alone why. You have to be very careful with counterfactuals to keep them from turning into just speculation.

Take the colonies not separating from Britain. OK, how does this happen? Do the British and Americans handle the crisises in the 1760s and 1770s differently and there is no war for independence? Which crisises? Do the British destroy Washington's army? Does the revolutionary cause just fall apart because the colonies can't agree on the form of the federal government (something like this happened four years after the Treaty of Paris?)? Do the British decide that it is worth more to keep the North American colonies than to prop up the East India company or the Caribbean plantations? Based on which point of divergence you pick, you get tons of alternates.

And that is not getting into the butterfly effect. Which each alternate, you get different people having different children, with different mates, than in our timeline. So you really can't chart an alternate with any accuracy beyond one generation.

I disagree that counterfactuals are useful. There is some value of taking some alternate course of events, say no United States, then figure out the minimum that would have needed to change to produce that. This gives some insight in what caused the actual historical event and how likely it was. Its also fun to take a specific point of divergence and ask what would have happened if it has gone some other way, but you can only chart the alternate path with any confidence for a few years, so while fun this is just not that useful.

Posted by Ed | July 6, 2008 1:19 PM

1. DivGuy, presumably a historian, gives the party line on how counter-facutals are bad, but causal explanations are good.

2. Some people with background in analytic philosophy make the non-obvious-but-true point that counterfactuals and causal explanations are logically equivalent.

3. DivGuy argues for a bit, gives up, and parts with a rude, philistine objection to the whole project of analytic philosophy on the grounds it is "useless."

What I've been arguing is that the structure and causes of human events are so complex that different observers necessarily produce different stories, and that as such we shouldn't be in the business of attempting to produce the sort of causal explanations in which factors are isolated and imagined as absent in order to demonstrate their causal effects.

Pithlord, here, has produced a plausible narrative of our debate that is quite different from what I would put forward, and which presumes a set of motivations on my part that I did not consciously experience. This is sort of why historians don't engage in the sort of imaginative games that Matt discussed - because isolating actual causes and descriptions of actual events already involves such massive work that we're not going to be able to have a useful discussion of counterfactuals because of all the inevitable dispute about prior issues.

Posted by DivGuy | July 6, 2008 1:24 PM

It is also important to remember that the only manifestation of counterfactuals is not through historical reasoning and argumentation. James Heckman won a Nobel Prize, in part, for his econometric work formalizing counterfactual reasoning. To argue that the insights his work has generated is "useless" or trivial, or the domain of "ivory tower" elitists, is philistinian.

Posted by Marshall | July 6, 2008 1:28 PM

Some historical facts, though, are very hard to dispute.

I would suggest that one such is that I did not "part" with my 12:38 post, but rather had a quick lunch and set my fantasy baseball lineup, then re-opened the page and responded the further course of debate at 1:10. It wasn't possible for Pithlord to know that at 1:11, but he was nonetheless incorrect.

Posted by DivGuy | July 6, 2008 1:30 PM

No one says law, as such, is useless. - Pithlord

Paul comes pretty close in his Epistle to the Romans.

Posted by DAS | July 6, 2008 1:40 PM

Kafka - as someone who took a lot of both styles of class at a fairly demanding school (about 70 credits in natural sciences, and about 60 in humanities, with a light sprinkling of social sciences), I think you oversimplify. Yes, the humanities are usually easier than the sciences, especially in introductory classes, but you'd probably be shocked at the workload in a high level English or Philosophy class at any decent university. It's easier to *pass* in humanities and social science classes, but the A isn't always as easy as you think, and grad school in English and Philosophy can be *harsh*.

By the way, I know quite a few hard-working immigrants in the humanities.

Posted by Harvey Lobster | July 6, 2008 1:43 PM

Paul comes pretty close in his Epistle to the Romans.

No way!

7.12-14: So then, the law is holy, and the commandment is holy, righteous and good. Did that which is good, then, become death to me? By no means!

Paul's argument, as explained by Stanley Stowers in his Rereading of Romans, is best understood as arguing to a Gentile audience that they are not to follow certain specific portions of the Jewish law (the "works of the law" that were generally understood as circumcision, sabbath, and food laws - see 4QMMT in the Dead Sea Scrolls, for example), but that they also ought not imagine that they have surpassed Israel, of which all people will be saved (11.1), for just as these Roman Gentiles were grafted on to God's people, they can be lopped off. They should recognize that God's plan for Israel and for the Gentiles is a mystery, in which the law holds different force for different people.

Paul never opposes "law" or "works" as a philosophical comment. He is quite clear in Rom 2 that the false teacher is shown a hypocrite precisely by his failure to follow the law, by his works. He argues at the conclusion of Rom 3 "Do we, then, nullify the law by this faith? Not at all! Rather, we uphold the law." (3.31)

Certain works of the law are prescribed to Jews but not to Gentiles, this is his primary point as the Apostle to the Gentiles. It's not an attack on the Jewish law or on law conceptually.

Posted by DivGuy | July 6, 2008 1:54 PM

DivGuy,

Paul doesn't completely oppose the law, but in Chapter 7 of Romans especially he flirts with the idea. Perhaps I am mistaken, but it seems to me the argument (which does have a grain of truth in it -- sometimes people have the spirit of pervisity and contrariness in them) is that "once you give someone a law by which to live, they'll then be tempted to break that law just to break the law, where without telling them that law, they would have not have even thought to commit the sin of breaking it" (Rom. 7:7-8). Certainly Paul explicitly says "Is the law sin? God forbid." but he does then issue a critique of legalism that raises some valid points about human nature (even if, as a Jew, I find that critique ultimately wrong-headed and defeatist).

Posted by DAS | July 6, 2008 2:35 PM

I don't think that's Paul speaking in Rom 7. He's using "speech in character" a very common trope of Hellenistic literature, which would have been as easy for ancient audiences to grasp as blog commenters grasp snark.

See, in particular, 7.9: Once I was alive apart from law; but when the commandment came, sin sprang to life and I died.

Paul was never alive apart from the law - this verse tips the audience off that this is someone else speaking. I think it's most likely that this is a speech-in-character of a Gentile who has attempted and failed to follow the law of Moses - the vacillation between the desire to do good and the inability to follow through draws deeply on the story of Euripedes' Medea, suggesting a Gentile speaker using Greek literature to work out an ethical point.

As such, the questions in that chapter about the value of the law are specific to Gentiles, rather than universal for all people.

I think that Krister Stendhal's insight that we should read Paul through the name he gives himself - Apostle to the Gentiles - is the key to reading Paul, and in particular the most difficult passages of Romans. He is speaking to Gentiles about Gentile salvation.

I don't love Paul - I think he's profoundly anti-feminist, and I think he appears to be a bit of an ethnic absolutist (though a very moderate one by ancient standards) - but I think that the argument that he was anti-Jewish and opposed to the law can be reasonably rebutted.

Posted by DivGuy | July 6, 2008 3:04 PM

isolating actual causes and descriptions of actual events already involves such massive work that we're not going to be able to have a useful discussion of counterfactuals because of all the inevitable dispute about prior issues.

Contemplating counterfactuals might call into question assumptions about the 'actual' causes of events, which is why many historians who are happily and rather uncritically weaving causal narratives tend not to like the whole idea.

Posted by novakant | July 6, 2008 4:39 PM

It's kind of hard to figure out why people -- as opposed to, say, scholars working within various formal traditions -- would care much about what happened in history and the effects of various actions if it's considered out of bounds to consider what else might have happened (as opposed to weirdly rigid declarations of what would have happened).

We can't predict the future, either, with any sort of scholarly rigor, yet we seem to speak quite often about the likely consequences of one or another choice.

Posted by El Cid | July 6, 2008 4:52 PM

Remember, this isn't a history journal we're talking about, but a blog.

For me, what-if's a hobby. Bwahahaha! For some reason, I truly enjoy thinking about this kind of thing. Is it really worse than scuba diving or stamp collecting, though?

I like to think I understand people and society better by this game, because each time I read a what-if, it makes me think about whether that really seems plausible based on my own life experiences and history I've read, which in turn gives me insight into what kinds of things are easy or hard to accomplish and how they can be done. But I mostly do it because I enjoy it, and I wouldn't dream of writing academic papers on them. I can see why a historian might not enjoy the game, as it'd seem too much like work.

Posted by Jon Kay | July 6, 2008 4:57 PM

Counterfactuals should be discussed because counterfactuals are FUN!
What if Lees' orders to the Army of Northern Virgionia weren't captured by McClellan and Lee had routed McClellan ? James McPherson posed this counterfactual and theorized that it might have led to a southern victory. Read Harry Turtledove's take on it

http://en.wikipedia.org/wiki/Timeline-191

What If England stayed out of WWI?
What if the Germans won at the first Battle of Marne?
What if Halifax rather than Churchill became PM in May 1940?

Hey, it's a fun game for history buffs :-).
All you people grumbling about counterfactuals being useless have no sense of fun. Lighten up already.

Posted by stonetools | July 6, 2008 4:58 PM

Counterfactuals are only useful when they suggest ways that later problems could have been avoided had certain decisions or events gone differently - and why.

Clearly if Iraq had not been invaded, very little that was bad would have taken place and the US would be ahead by at least a trillion and probably 3-5 trillion dollars, 4,000 less dead troops, a million less dead Iraqis, and probably a lower oil price, and perhaps not be in a recession as a result.

That's counterfactual worth considering.

Asking somebody what might have happened if the United States had never existed - not so much.

Science fiction writers produce a lot of stories on "alternative histories". They're good reads, sometimes, but not terribly useful as guides to the future. Real actual history is better. Any lessons you need to learn can be learned from actual history.

Posted by Richard Steven Hack | July 6, 2008 6:01 PM

I think counterfactuals are a fun way to reexamine assumptions. I guess my problem with counterfactuals is there are so many people out there making arguments with no facts that counterfactuals have lost some of their appeal. "What if we found WMD in Iraq in April 2003?" isn't a productive exercise if the response is "We DID find WMD in Iraq!"

Here are a few of my favorite 'What if...?' comic titles -- with answers.

"What if Sgt. Fury had fought World War 2 in outer space?"

There would have been fewer German rocket science emigres.

"What if Shang-Chi, Master of Kung Fu, fought on the side of Fu Manchu?"

The side of Fu Manchu would be totally kick ass.

What if Captain America had been elected president?

He'd govern effectively and take his message straight to the people. Candidate America was forced to develop a strong communications team in order to overcome the harsh unfair coverage he received in the press when facing GOP nominee Sen. John McCain.

What if the Hulk had the brain of Bruce Banner?

The senior Senator from Alaska would be known for his Wolverine ties.

Note: I'm recommending this blog to all my friends. The service is great. Try the veal!

Posted by joejoejoe | July 6, 2008 6:48 PM

Shouldn't our confidence in historian's ability to answer "what if?" style questions about the past be equal to our confidence in their ability to predict the future (mutatis mutandis; given the amount of information they possess about the time period they are posing counterfactual questions about and the amount they possess about the present)? Aren't those logically symmetrical projects?

The history of science somewhat exempted, since we have a pretty good idea of what they are eventually going to discover, but don't know what future scientists are going to discover.

Posted by Bryan | July 6, 2008 9:03 PM

Shouldn't our confidence in historian's ability to answer "what if?" style questions about the past be equal to our confidence in their ability to predict the future (mutatis mutandis; given the amount of information they possess about the time period they are posing counterfactual questions about and the amount they possess about the present)? Aren't those logically symmetrical projects?

The history of science somewhat exempted, since we have a pretty good idea of what they are eventually going to discover, but don't know what future scientists are going to discover.

Posted by Bryan | July 6, 2008 9:06 PM

I think another reason that historians don't like counterfactuals is that they usually turn on one single event happening differently - for example, the case of Lee's orders being captured. Most historians would argue that almost any interesting historical phenomenon was the product of larger trends and forces (for example, the North won because they had all the industry). Additionally, such counterfactuals often involve great men doing particular things (what if Hitler was a good strategist) and are often focused on military history, both of which are approaches that are generally out of favor in the historical profession (speaking generally).

Posted by Sam TH | July 6, 2008 9:50 PM

Not all historians despise counterfactuals.
About half the contributors to the What If books are fairly well regarded historians. Military historians, Princeton profs and Englishmen appear to be big fans of the genre.

Posted by Robert Cowley muses on what if Yglesias was never born | July 6, 2008 11:43 PM

I’m a history teacher, with a BA in History and a MA in Education. I don’t know if it qualifies me as a history expert but I must be more qualified than most.

Yes, historians don’t like answering counter-factuals, at least as working professionals. This is because there is no definitive way to prove a counter factual. Historians search, to the best of their ability, for the truth. It’s like asking a pro basketball player to do some of that Harlem Globetrotter stuff, what is the point? Of course, as the mentioning of the “What If” series of books show, it’s a different story if the speculation is all in fun.

In my opinion, the biggest problem with most “What If” scenarios is not that you can’t make a good case as to what would have happened if “X” happened differently than it did in reality. In many situations, the variables don’t matter that much, no matter the fact that there are so many different ones. The problem is, the most likely scenario is not all that sexy. And everyone buying the “What If” books is looking for a sexy scenario as their answer. They ask the question with a pre-conceived fantasy in their heads.

For example; what would have happened if that taxi cab that hit Winston Churchill in the 1920’s had killed him, instead of just dirtying up his suit? The implicit question laying in waiting is “How would have WW2 turned out, could the Nazis have won”? The most likely answer is…”not that much different”. Yes, more people, particularly on the Allied side would have died. But Hitler was not going to win fighting against Britain, the US and the USSR. Putting Stanley Baldwin, Neville Chamberlain or Lord Halifax in the PM’s wouldn’t have changed that.

Some single micro events are more likely to change history of course but let’s delve into a question using the macro structure mentioned. Another example; what if America didn’t win it’s independence from England in the Revolutionary War? Well, as it as Ed mentioned, Concord and Lexington never happening produces a different result than Washington getting caught crossing the Delaware, which would produce a different result than Yorktown going the other way. But again, the implicit question here is, “Would we have been having tea time and talking in cockney accents”? The most likely answer is “no”. America, at worst, would have been in the Commonwealth system like Canada and Australia. This would have happened quicker for the US and full independence is still a more probable answer.

Of course, if Concord and Lexington never happened, you could make a case that the French Revolution would never have happened, as France would have been richer in funds and poorer in revolutionary spirit. But the person reading the “What If” book’s section on the American Revolution doesn’t care about France.

To get into a situation where Hitler is victorious over the world or the US has a trans-Atlantic parliament with the UK, you have to change so many recorded facts, (this is where as those variables come into play), you end up writing a pure fiction novel.

There are a few exceptions to this dynamic though. Keep Tsarist Russia out of WW1 and you would most likely end up with a constitutional monarchy that would have been able to survive until this day. Russia had bankrupted itself playing off the nobles for confiscated land given to the surfs once they were freed. But by 1914, things were actually well on their way to recovery. WW1 just destroyed everything. No, the government would not have been as functional as England, as the tsar would have ended up with too much power. But no Russia in WW1 equals no Russian Revolution, which equals no Soviet Union. That’s a pretty significant change and very probable chain of events.

Posted by Raindog | July 7, 2008 1:48 AM

Sam, good point, but then we have to ask why historians, as a profession, favor one approach over another. I sometimes think that it's as simple as wanting to believe that what you're studying is meaningful; if the larger social factors don't explain why things happen, then why have you been studying them?

Posted by AlanC9 | July 7, 2008 2:14 AM

Ahem. It's Planned Parenthood v. Casey. Thank you.

Posted by Anon21 | July 7, 2008 1:21 PM