For a few years, I’ve become fascinated by counterfactuals, especially why they have so much rhetorical power, even when they are very, very wobbly. What makes them fascinating it that, despite their being logically problematic, they often serve as an incantation then ends deliberation. To take simply one example, in late July of 1965, LBJ was meeting with Congressional leaders about Vietnam. LBJ presented five options: first, “blow them up;” second, “get out;” third, maintain the status quo of 80-90k troops; fourth, “we could declare an emergency, call in the reserves, thousands of men and billions of dollars—tell the country that our best guess was Y billion dollars and X thousand men, and ask for it;” and fifth, “supply Westmoreland with the men and materials he had requested, increase draft calls […], extend enlistment of sailors and marines, and go into the Senate for an additional appropriation of $1 or $2 billion and defer a full presentation of requirements until January, 1966.” And then he said “the choice was between course four and course five”— different variations on escalation. Ambassador Henry Cabot Lodge endorsed LBJ’s view; he “knocked down choice one, two and three,” partially by invoking Hitler. Speaker of the House John McCormack “said that he did not see any alternatives […] He was impressed by the analogy to Hitler. The road to appeasement was a road to war.” And that was the end of considering other alternatives.
But why?
The analogy between options in Vietnam and appeasing Hitler is vexed in several ways, especially the sense that appeasement has necessarily policy implications for all foreign policy decisions—yet that’s the role it’s often played. What makes it such a puzzling role is that what matters is not what actually happened in Munich, but continually morphing or vague narratives about choices and events that never happened: counterfactuals.
There are many forms of counterfactual statements—lies, speculative fiction, metaphors (including metaphorical frames), thought experiments, predictions about the future, war games, alternative histories. Although all those forms of counterfactual statements can affect policy deliberation, it is the last that is most interesting to me—because it’s the one that so often shuts down deliberation and silences dissent. So, that’s the kind I mean in this post.
Alternative histories are used in much more than just policy deliberation. They have an important role in the stories we tell ourselves about our own lives—“If I hadn’t taken the train that day, I would never have met Terry.” Psychologists characterize counterfactuals as either downward (if something else had happened there would have been a worse outcome) or upward (if something else had happened there would have been a better outcome. So, “If I hadn’t taken the train that day, I would never have met Terry, and I wouldn’t be so happily married,” or, “If I hadn’t taken the train day, I would never have met that asshole Terry, and I wouldn’t be so miserable today.” Counterfactuals about Hitler are always upward (“If we hadn’t appeased Hitler, WWII wouldn’t have happened/been shorter/less destructive…”).
Counterfactuals can serve a variety of purposes. The three most common are: managing feelings, solving problems, and advocating policy (not necessarily in that order). And the various purposes can be at odds with each other.
The feelings we’re trying to manage might be our own, or those of an interlocutor. The feelings might be grief, despair, shame, pride, anxiety (especially about control), fear/threat/anxiety, animosity toward an out-group. So, for instance, in reddit subs that show dashcam footage of car accidents, there is always at least one person who comments that the accident could have been prevented if the dashcam car had done [something or other] even in cases when the counterfactual course of action was dangerous, unwise, impossible, or wouldn’t have prevented the accident. The comment is in service of the commenter’s anxiety that they might someday get in a car accident—they want to believe that they are always in perfect control of the situation, and can prevent bad things from happening to them. (The Just World Model) My crank theory is that the more someone believes in the Just World Model the more likely they are to engage in counterfactual narratives about how that bad thing would not have happened to them. Counterfactuals restore a sense of control.
On the other hand, they can be used to disavow ourselves of responsibility—if my mother hadn’t kept me, I would have caught the earlier train, and then I wouldn’t have met that asshole Terry. So, sometimes the affective power of a counterfactual is to regain agency, and sometimes to deflect responsibility.
The counterfactual(s) we create to make ourselves feel better can undermine their ability to help us solve problems. We can’t learn from mistakes unless we admit that we made a mistake. Perhaps the more useful factor in the counterfactual about Terry is that I need to have better taste in partners, or I need to stop falling for narcissists, or something else about me that needs to change. While the narrative that blames my mother might make me feel better and save some pride, it doesn’t help me identify, let alone ameliorate, the problem.
I think the unreasonable and unmerited power that counterfactuals have in policy advocacy is related to what a particular counterfactual does for our feelings. There are many counterfactual decisions and actions that might have prevented WWII, shortened its duration, or reduced its devastation—had WWI not happened (which creates its own set of counterfactual narratives), had the Versailles Treaty not been so punitive, had the German demilitarization aspects of it been enforced, had the worldwide depression of 1929 not happened, had all the participants in the Evian conference opened their doors to refugees, had anti-fascist parties formed an effective coalition, had von Papen not been so confident about controlling Hitler, had Stalin believed the warnings of his advisors, had Hitler been admitted to art school. And, again, I’m not condemning counterfactuals—they are inevitable at least in implication in causal arguments, and have usefully informed policy decisions. The counterfactual about the Versailles Treaty being too punitive and too weakly enforced strongly influenced post-WWII policy toward Germany and Japan, for instance, with beneficial consequences. But counterfactuals don’t necessarily imply policies we must adopt. While Hitler getting into art school would have prevented WWII, I don’t think we should conclude that we must enact a policy that everyone is guaranteed admission to art school.
Because counterfactuals never happened, they are by definition untrue. They are narratives of possibility and probability, and there are a potentially infinite number because very few involve simply one step. If France and Britain had responded with aggression to Hitler’s sending troops into the Rhineland would have triggered a coup, which would have been successful, which would have put nice Nazis in place instead of Hitler (?). Or, it would have been popular with French and British voters, who would not have agreed with the plan to hand the Rhineland back to Germany in a couple of years, and have thereby given Allied troops easier access to Czechoslovakia, so that a military response in the Sudeten Crisis would have been successful. Both of those counterfactual narrative have a lot of forks.
Others have narrated equally plausible counterfactuals about the Sudeten Crisis that would not have had better outcomes. Some have argued that Hitler wanted war over the Sudetenland. Or, he might have moved up the pact with the USSR and included Czechoslovakia in that agreement, so that threatening a military response would have only delayed war—the same outcome that the Munich agreement achieved. I should say that I think Chamberlain made a mistake, but even I don’t think there is only one reasonable reading of the situation, let alone that it has undeniable implications for all foreign policy decisions. (For a longer discussion of the problems with the Munich analogy, see here.)
The upward counterfactuals we use to make ourselves feel better or to advocate a policy are particularly prone to motivated reasoning—for instance, we assume that missing the train would have meant we never met Terry, or some other asshole, and we would have stopped being drawn to narcissists. As mentioned above, at each fork in the path of consequences of refusing to appease Hitler, advocates of that counterfactual assume best case outcomes, ignoring all the butterflies. That doesn’t make them wrong, or useless. But it means that they are arguments, in the realm of plausibility and probability, They are not in the world of certainty, but in the world of perhaps.
Counterfactuals are inevitable, necessary, and potentially very useful in policy deliberations. We should try to learn from history when advocating a policy—I often ask people who are advocating a particular policy to name a time when that kind of policy worked—but there are some questions to consider. If it’s never worked, why would it work now? If it has worked, was it under conditions much like the current context? If it didn’t work, was that for reasons that don’t apply? How analogous are the situations?
The magical deliberation-stopping Munich analogy, despite being almost compulsively common in Cold War foreign policy rhetoric, had weak relevance to the discussion at hand. Although people were worried about a repeat of the Korean conflict, in which American actions had caused China to intervene directly, no one believed that Mao was determined to start WWIII, whereas scholars of Hitler are very clear that he was determined to get a second world war started. So, there was at least one very good reason that the Hitler analogy might not be the most relevant and useful analogy for Vietnam. Why not France’s recent experience in Indochina or Algeria, the Boer Wars, the occupation of Germany, Korea?
Perhaps more important, McCormack’s statement is odd. While appeasing Hitler in regard to Czechoslovakia did not prevent war, the other option was going to war immediately. Both roads led to war because Hitler wanted war. Chamberlain’s appeasing Hitler is not the only time nations have relied on appeasement, and many of those instances did prevent or end war (since many treaties involve compromises).
Because something didn’t happen, it can’t be proven what would have happened—that’s the nature of a counterfactual. By definition, counterfactuals are not factual; they are not true. They are, at best, more or less plausibly argued.
They are stories about what people might have done, and what might have happened as a consequence, not what would have happened. Counterfactuals are arguments, and I wish we could treat them that way.