post hoc ergo propter hoc

[warning: very long post. but hopefully interesting]

I’ve been thinking more about explanation & causation in history recently. It seems to me that there are structural elements in both film and narrative that are quite similar to the Humean model of causation: that is to say, images or phrases are placed one after the other, and the viewer or reader has a “habit-driven” inclination to infer the (causal) relationship between the two. Examples:

From film: (1) A camera shot shows one person pointing a gun; (2) the camera cuts to another man falling to the ground, accompanied by a gunshot. Post hoc, ergo propter hoc. (2) happened because of (1).

Or in fiction, we might have this passage:

(3) She refused to go to work today. (4) Her head throbbed with almost military vigour.

Reading the two in succession, we tend to assume a causal relationship, even in the absence of any explicit “because”. (Post hoc, ergo propter hoc). By the end of it, the second sentence has become an explanation of the first, Interestingly, in this case we do assume causation, whereas in the following case, we would probably, unless context demanded it, not:

She refused to go to work today. Her cat ate some cheese.

The causal relationships are inferred by us; there is no need for an explicit “because”. (Furthermore, inserting an explicit “because” between (3) and (4) seems to injure something stylistic about the passage. And is history-writing not, after all, in part a stylistic enterprise?) Narrative in general seems to be a series of individual statements about what-is-the-case, which, taken collectively, may be joined up by habit-driven inference on the reader’s part, as above. In historical narrative, then, it seems to me that a series of individual statements about e.g. what happened in the years leading up to 1939 can become, by the last sentence, an explanation of it. And so, the common distinction between a chronicle and a history — that the latter explains while the former does not — isn’t quite so clear cut, if it seems possible that a chronological series of well-selected statements can, in fact, yield an explanation by the end of it. Which is, I think, the entire value of narrative history.

‘Well-selected’ here is the key, of course. My dear, long-suffering boyfriend Adam, who puts up heroically with my torturous ruminations (and indeed subjects me to his own), wrote me the following to make the point that an unselective chronicle might, in fact, look nothing like a narrative:

Events of the year X:
The commencement of King Cuthbert the Flatulent’s reign. (A)
A rain of frogs in Wessex.
A monkey is hanged as a presumed foreign combatant by a group of Yorkshire villagers.
Pope Pius XXXVII is defenestrated. (B)

Events of the year X+1:
The weeping visage of the Virgin Mary appears in the sky as Carcasonne falls to the Cathars.
It is revealed that Pope Pius XXXVIII is a woman. (C)

Events of the year X+2:
King Cuthbert spontaneously combusts during a battle with King Canute of Denmark. (D)
Lisbon is engulfed by a tidal wave. (E)
The population of Rome is decimated by a terrible pestilence. (F)

Events of the year X+3:
Pope Pius XXXIX is appointed. (G)
Yellow devils on horseback descend on Poland and Hungary, burning Pest and Lvov.
A rain of communist pamphlets falls on Paris. (H)

Events of the year X+4:
The Irish potato crop fails. (I)
Barricades appear on the streets of the French capital. (J)

Events of the year X+5:
Parliament passes legislation permitting Irishmen to eat their babies. (K)
Dostoevsky publishes Crime and Punishment.

Further points about our habits of causal inference, Adam adds: “There are several items in this chronicle that are difficult for us not to link causally: (A) and (D), for instance, or (H) and (J). Why? We perhaps associate (for no very good reason) flatulence (A) with spontaneous combustion (D); we have grounds to suppose that a city’s first exposure to communist ideas (H) could lead to a workers’ uprising and thus to barricades (J). Many people in the past would have been tempted to infer a causal relationship between (C), (E) and (F). We know better.”

And though we might infer an unhappy causal relationship between (I) and (K), the inclusion of those two statements, related or not, seem to cast doubt on the veracity of the whole document. And rightly so, it being a Chronicle of Unfettered Though Amusing Fabrication. But that we have in the past inferred such causal relationship between events that we now know better not to (C, E and F) is a little worrisome: it’s where, for example, the issue of standards of justification come into play in history, and it’s why I think it’s important to wonder now and then: what are historians really doing when they explain things?


11 responses to “post hoc ergo propter hoc

  • Gavin Robinson

    You can also see/hear newsreaders doing this, especially on regional news. eg “A man was taken to hospital after an accident on the A16”. They obviously mean “because of” but they very rarely say it. I’m not sure how significant that is, as the way newsreaders talk is really, really bizarre.

    You’re right that this issue needs more attention, otherwise careless historians might accidentally imply things they didn’t mean to imply, and dishonest historians might use this effect to insinuate things without having to argue their points.

  • Rachel

    Yes, I’m very interested in, well two things: firstly, how a lot of our language is implicitly causal; secondly, how a passage that is seemingly “just narrative” can conceal an explanation. And they’re connected. On the first point, a lot of verbs are explicitly causal: maim, push, break; but a lot of words are just implicitly causal: father, victim, author. Expand this to narratives, and you get explicitly causal narrative, like “Because…Therefore…For this reason”, as well as implicitly causal narrative, like “Inevitably…Naturally…Conditions favoured…It brought about…This influenced”. So it’s not just newsreaders — though I do agree that they are a rather bizarre bunch … 🙂 I think the mealy-mouthedness has to do with lawsuit-phobia. Historians maybe have less excuse.

  • Oli

    From what I can gather from a few reviews, Pinker’s ‘The Stuff of Thought’ is quite good on how the way in which we implicitly organise and associate information is revealed through language. Perhaps that’s relevant to your interests. Personally i’m waiting for the paperback to come out in the Summer – the hardback’s a little expensive.

  • Profane

    To pick nits, I would contend that Adam’s Chronicle does not quite reach that level; my instinct as a student of Irish sources would be to label it a collection of Annals. A Chronicle, in my view, must have lengthier narrative entries, or a thematic unity which is not even implitcitly present in the Newspaper Headlines above, and in the sources which I too often find myself relying on!


  • Geschichte Grad

    “what are historians really doing when they explain things?”

    Not to be banal, but doesn’t this point at the political influences and implications of our work? We choose not only causal sequence, but also the story itself, believing it to be important. And “importance”, of course, is determined outside the story and causal chain itself. So: what are we doing? Making a political statement, perhaps.

  • Jeremy Young

    But what you’re describing IS our job: to make connections that we do not know that future historians will smile upon. We are, for the most part, interpreters of information. I don’t think we should be concerned about whether our interpretations hold up over time, so long as we made them with the best of our judgment at the time.

    As an example, Charles Beard wrote An Economic Interpretation of the Constitution of the United States in 1913, in which he argued that the Framers created the Constitution in order to preserve their personal economic property. The book created a furor when it was published, and it’s been generally discredited since. So what? Beard tried something. The fact that it didn’t work is, really, of little consequence, I think. He looked at something from a new angle and shone new light on it, and in doing so he contributed to our knowledge, even his his interpretation was ultimately wrong.

  • Gavin Robinson

    Evolutionary biologists have even bigger problems here. It’s almost impossible to talk comprehensibly about evolution without using teleological language, even though evolution is obviously not teleological.

  • Mike L

    Your reference to film makes me think of this instance where those expectations of causality are disrupted. In ‘The Silence of the Lambs’ there’s a sequence where Federal Agents are moving in on the killer’s house. (If you’ve not seen it, turn away now. In fact, turn off your computer and head straight to the video store.) The film cross-cuts between an undercover agent ringing the doorbell and the killer grumpily walking from the basement to answer the doorbell. Agent (with armed backup) rings doorbell. Killer’s house reverberates with doorbell. Etc. Killer finally opens the door, and standing there is not the armed and ready posse of agents, but an alone and vulnerable Clarice Starling, who’s just making enquiries and has no idea that she’s stumbled on the lair, or that her fellow agents are at the wrong house.

    The sequence depends both on there being no necessary causal relationship between the shots and also on the convention that there is such a relationship.

    And, on history rather than film, as you and your readers note, it’s worth being careful about accidental and disingenuous implications. I think also that what historians must do sometimes is to draw out those implications when the language or chronicle or narrative doesn’t offer them. That is, to urge caution where apparent connections might not be true, but also to make explicit those connections that might not be apparent. I’m thinking, for example of Raul Hilberg’s use of mundane evidence like cargo manifests and train records to demonstrate the scale of the Holocaust and processes by which it happened.

  • Tim

    There’s a well-known experiment in film testing what’s called the “Kuleshov Effect,” after Russian film theorist Lev Kuleshov. Kuleshov intercut a shot of an actor’s face with different alternate images — a plate of soup, a young girl playing, an old woman’s coffin — and viewers attributed hunger, happiness, and sorrow to the actor accordingly, depending on which image was intercut.

    Film causality and continuity work the same way. So long as the intercut images are reasonably coherent, our brains form a coherent narrative of events to explain them.

    Arguably, we do the same thing with historical narrative. By juxtaposing a string of events that are apparently connected — by time, place, relevance, biography, etc. — historians create a convincing appearance of a unified story or event, one that suggests a causal or teleological chain or even a witty observation or analogy (a la Plutarch’s Parallel Lives). And the reader participates in the process by assenting to or even forming the coherence suggested by the author. This is especially easy if the form and content correspond to habitual modes of thought, presentation, and reading.

    Of course, at some point, a critical reader, armed with logical fallacies or theoretical objections or a better grasp of the facts or even just different habits of reading/standards of explanation starts muttering to themselves, “That doesn’t explain anything!”

    What’s remarkable is how historically malleable our expectations for historical explanations are. We’ve pretty much trained ourselves out of the “historical figure as exemplar” mode that guided classical writers like Plutarch, or biblical-typical exegesis. But that was a pretty convincing form of explanation for thousands of years; and you can still find it in residual forms, like the tired “ad Hitleram” pseudo-arguments you find thrown around by American conservatives (and to be fair, nonconservatives).

    Maybe one question that could be asked is: How do emerging forms of explanations become legitimate? For example, when you read classical histories, some of them really read like little Marxian parables: the clashes and battles are all about money, and power, and grain, and colonialism, etc. Yet Marx’s historiography was still radical in the 19th-century, because it seemed to go quite past what his contemporaries had been doing.

    If there are newer kinds of critical lenses and narrative/argumentative forms emerging as legitimate explanations — say, after Michel Foucault, or Walter Benjamin — then what is our attitude towards them? How/why do we accept them as a legitimate new way forward, or resist them as an untenable stretching of the plausible?

  • Rachel

    @ Mike L & Tim: I like both your thoughts on film. That the deliberate subversion of assumed causal links arises out of primarily stylistic considerations is interesting for me because, as I said, history seems to be in part a stylistic enterprise. Rhetorical flourishes can disguise a hidden (or absent) argument.

    On the juxtaposition of images, there’s a film that just came out on DVD today called Our Daily Bread, which, from what I’ve seen of it, is going to exploit those implicit causal links between a series of successive and interposed images to depict the disturbing mechanization of food manufacture. there’s no guiding voice-overs. Many people have written of that film that it “lets the images speak for themselves”. I wonder, given what we’re thinking here about hidden implications of cause and effect, to what extent that’s true. Does the fact that the film is almost universally viewed as “disturbing” attest to the actual nature of those processes, or to the filmmakers’ explanatory narrative powers, all the more potent because they have been achieved wordlessly (implicitly)?

    Though I’m also told it’s not a film to be missed, so while you’re on your way to that video store …

  • Rachel

    PS: also, Tim’s question of accepted or untenable critical lenses says something, I think, to Jeremy’s comment that even bad and incorrect interpretations can contribute to our knowledge. Did we really need Holocaust Denial? (ah, I knew Godwin’s Law would kick in eventually :)) Though I suppose, in historiographical resistance to bad interpretations, we glimpse these boundaries of the acceptable and plausible — perhaps, Jeremy, that is what you meant by “contributing to our knowledge”?

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: