There’s a famous (among academics) essay by Hayden White I often ask suffering students to read, called “the Value of Narrativity in the Representation of Reality,” a ponderous title indeed. White reproduces excerpts from the Annals of St. Gall, a medieval manuscript with multiple authors. It looks like this
709. Hard winter. Duke Gottfried died.
710. Hard year and deficient in crops.
7 12. Flood everywhere.
714. Pippin, Mayor of the Palace, died.
715. 716. 717.
718. Charles devastated the Saxon with great destruction.
720. Charles fought against the Saxons.
721. Theudo drove the Saracens out of Aquitaine.
722. Great crops.
725. Saracens came for the first time.
731. Blessed Bede, the presbyter, died.
732. Charles fought against the Saracens at Poitiers on Saturday.
Those are years on the left: many of them are blank in the original
The Annals continue this way, till:
1045. 1046. 1047. 1048. 1049. 1050. 1051. 1052.
1053. 1054. 1055.
1056. The Emperor Henry died; and his son Henry succeeded to the rule.
1057. 1058. 1059. 1060. 1061. 1062. 1063. 1064.
1065. 1066. 1067. 1068. 1069. 1070. 1071. 1072.
White asks, assuming you have no other sources available, “what would it take to make this into history?” The answer, he says, is you have to invent some idea which can’t be proven to actually exist. For example, the easiest way to make this history would be to say “imagine a god.” Then all these events could be ascribed to that god, who I shall call “Azoth.” Azoth is displeased, and smites the people; Azoth is pleased, and rewards the people. Individual actors come and go because of their relationship to Azoth, who cannot be proven to exist but who, if we accept as existing, can fill the role of “cause.” Floods, bumper crops, invasions: all ascribed to Azoth. The idea of Azoth makes these random events sensible. Pippin, mayor of the palace, should have prayed harder.
White elsewhere calls this a “metanarrative,” something which is outside of your actual narrative but nevertheless crucial to it making any sense at all. Instead of Azoth, for example, you could substitute “progress,” beloved of freshmen: “progress was occurring.” Or “the rising middle class.” No one can define middle class” with any precision; like my hypothetical god Azoth you can’t interview “the middle class” or look it in the eye, but if you assume it exists, and is rising, you can do all sorts of explanatory work.
None of these are good history, but they could be. I’ve invented a frame, a metanarrative, that can take the chaotic, random events the past leaves us with and make them seem orderly.
I could do the same thing with “industrialization.” We all know what we mean when we say this: we get images of factories and smokestacks and striking workers. “Industrialization” is a well accepted and frequently used metanarrative, but it can’t be proven to exist any more than Azoth. Factories exist: “industrialization” is a concept that frames them. It’s a way of making sense of a set of facts, linking them together in a larger story. “You should understand those smokestacks as part of something called ‘industrialization’.”
White’s argument implies that the world has no inherent order, that it’s chaotic and random, and that all systems of order are arbitrary and imposed. History presents you with a bunch of stuff that happened, and ordering it, making it sensible, requires inventing something outside of history, like Azoth or progress or industrialization. There’s a vast list of these metanarratives, things which don’t actually exist concretely but can be imagined as existing: they include “the anglo saxon race,” “man’s thirst for freedom,” “self-interest,” “negroes,” “the American mind” “childhood,” etc etc.
I’m not a religious man, or even what’s vaguely called a “spiritual” person. But the interesting conclusion I get from White’s article is that it’s impossible to do historical thinking or writing without making a “leap of faith,” and choosing to believe in something that can’t be proven to exist. If I were a religious man I’d be making an argument about the necessity of faith.
There’s an equally famous essay by Carlo Ginzburg that makes an opposite claim. In “Morelli, Freud and Sherlock Holmes: Clues and Scientific Method,” Carlo Ginzburg and Anna Davin claim that there are real, objectively verifiable narratives in nature, and we see them in diagnostic reasoning and experimentation, what they call “recipe knowledge. “If I do this, that will predictably result.” If I see tracks that look like bear tracks leading towards a tree with bark worn away, then towards a stream, then into a cave, I can construct an objectively true narrative: “there was a bear, it scratched its back on a tree, drank from the stream, and then went into that cave.” If the marks on the tree are old, I can assume the bear scratches regularly; if the stream is muddy and the the bottom disturbed, I can assume the bear was hungry.
Ginsburg and Davin are interested in forensics and detection. They see a connection between the religious tradition of “divination,” reading the signs to determine God’s intent, and a doctor’s ability to diagnose an illness from its external symptoms. In other words, there is pre-existing order, a pre-existing narrative, in events, and we can uncover it.
This is entirely opposite what White claims. And interestingly, if there is order in events already, a real, objective and verifiable chain of cause and effect–then faith is not necessary at all.
Though I’m not a religious man, I find White’s account much more plausible. On a small level, I can deduce the presence of bears or count on baking soda as a rise. But I can’t write a history of either baking or hunting without some kind of metanarrative that links the countless instances of baking and hunting, the many motives and practices, into a coherent story.
Earlier I posted about the value of history: it seems to me both these essays point to the centrality of doubt and skepticism to any practice of history.
Wonderful piece. But in some ways, it’s not about choosing BETWEEN a belief in the primacy of empirically verifiable evidence and a highly abstract meta-narrative: It’s about making the level of generalizations in the narrative appropriate to the kind of evidence adduced. And while you are doing it, keeping in mind the purpose for which you are constructing the narrative. A religious narrative has a very different purpose and audience from either a forensic investigation or a historical analysis of the past.