"Using the scientific method" would be my definition.

And Reed, your description of the split within political science between "science-y" and "non-sciency" is precisely what I had in mind.

That would be my definition as well; which is why essentially none of history qualifies.

I mean, there is the possibility of a relatively rigorous case study method in history. For instance, the famous problem in economics about how Ghana and South Korea had the same gdp/capita in 1958 but have since diverged sharply could be submitted to a historical analysis (which necessarily would involve a bunch of economics and politics, as well, but could still fundamentally be historical in orientation).

But that's about as close as you're going to get. And it ain't science.

The overall argument would not be science, but it would include some stuff that is.

Human geography or anthropology (I'm not sure where one begins and the other ends) is like that too. If you're going to argue that climate change is causing a certain pattern of migration, you need to A) prove that the pattern is actually happening B) that the climate is actually changing or has changed.

And lastly, that A causes B.

One key difference in what you're describing though is that presumably if it the migration were happening now, geographers and anthropologists could go ask the migrants why they were moving, and use various social science methods to come up with some convincing scientific answers. In history, obviously, you can't do that.

If it's historical, the best you can do is come up with good evidence that the people moved and that the climate changed and that one came shortly after the other. The "why" would be harder to prove without written records of people explaining why they left.

The point is, or ought to be, that just because something isn't a "hard science" amenible to prospective experiments, doesn't mean it's ok to wheel out just any old bullshit. I've encountered a few people studying various humanities and social sciences who use that old meaningless chesnut "It's not a science, it's an ART!" and think that covers all sins.

My question was actually quite narrowly targeted. It was "How can we judge the truth or falsehood of causal hypotheses within history?" I asked it because I think to deal with causality, you need ultimately to deal with counterfactuals. That doesn't necessarily mean you have to be able to set up counterfactual situations ("experiments"), but if you can't do that, I reckon you do need a robust explanatory model of some kind. History lacks both (a) the capacity to set up counterfactuals and (b) a robust theoretical framework for causal claims, I think, however you gloss the term "robust" (and thereby hangs a long, long discussion).

But, look, it's one thing to say that causal hypotheses in history are necessarily rather speculative, and quite another to say that history is in some sense "just narrative". That's a hell of a leap. It brushes aside the whole area of factual hypotheses, for example, which are at least some of the time can be fairly straightforwardly judged true or false.

The class of historical problems which can "straightforwardly" be judged as true or false is pretty narrow, I think. We know that events occurred on particular dates. If we are dealing with individual people, we can ascribe rises and falls pretty easily (he came into a lot of money at age 18, he died of syphilis/a gunshot wound/starvation, etc.). Once a second person enters the scene, you're having to start to guess at motives in interactions. Sometimes the person in question might have left some papers that will help determine motive, but even then there's the possibility that they are censoring themselves in the writing. All you can really do is arrange events into a sequence and provide some plausible hypotheses, buttress it with evidence and tell a story about why they acted the way they did. If the result strikes a lot of people as being especially plausible, then you're a good historian.

However, what counts as "plausible" changes over time, which is why there will always be a market for revisionist history. And why we'll still be writing books on the Origins of the First World War for another couple of hundred years.

Well, anything that qualifies as a "problem" is presumably, by its very nature, not straightforward. Something gets to be a "problem" only if it's in some way problematic.

But the class of historical statements that can be straightforwardly judged true or false is quite large, and this class serves to constrain legitimate historiography in a way that other types of "narrative" are not constrained. In fact, one could argue that this set of constraints is defining: what doesn't respect it isn't history, but something else.

That's not to reduce history to the recyling of uncontroversial truisms, but it is to say that a thoroughgoing epistemic relativism is unsustainable here. You can't claim that the Armistice was signed in Bognor Regis in 1923 and still be doing history. Merely by making such a claim you become the author of fiction, or comedy.

There is, as you say, no controvery about which treaties ended which wars, whre they were signed, who signed them, etc. But the questions historians care about are: how did the final outline of the treaty come to look the way it did? Who gained from the treaty? Who lost? How did the treaty affect subsequent events? etc.

And these, fundamentally, are answerable only through the construction of narratives. Anchored by "true historical statements", as you say, but narrative nonetheless.

I'm not sure what view you're defending here: the view that historians construct narratives, or the view that that's all they do. The former seems to me to be uncontroversial, and the latter seems to me to be false.

I'm saying history is primarily about narrative.

Which I don't think is controversial.

Well, maybe and maybe not: I'm not sure what you mean by "primarily about". Historians create narratives, yes, but what you seem to regard as incidental (the relatively uncontested factual claims that constrain those narratives) seem to me to be precisely what makes those narratives historical as opposed to some other kind of thing. That's pretty primary, I'd have said. But I may be misunderstanding you.

QUOTE:
but like most of Gladwell's pieces, I got kind of bored and thought he grasped onto one clever idea/trend and kept on repeating it over and over again to talk about how interesting it was.

I agree it tends to flog the same thing over and over. Fortunately for me, I read it in an airport and then on the plane, so I was in a position to be much more patient with the repetition.

OK, Wyatt, but for any given subject, there are thousands or tens of thousands of possible "uncontested factual claims" (let's call 'em UFCs for ease of discussion) that might have a bearing on it. Historians don't simply knit together a bunch of these UFCs (which is how I'm reading your position, possibly wrongly); they also select and order these UFCs in a way that tells a story. For every UFC in a historian's thesis, there are thousands of others he/she chooses not to include. And these acts of inclusion and omission - of deciding which bits of a story to tell and which to not tell - are the very essence of narrative, no?