Expecting a revolutionary theoretical work about narrative, I instead found Franco Moretti’s Graphs, Maps, Trees to be a compelling, albeit disarming, argument for the analytic toolkit of a social historian. I’m not well-versed in the current state of the literary history field, but his book left me wondering if, in 2005, this argument was really that groundbreaking.
I suppose that Morelli’s most controversial point is the assertion that deep textual analysis of single, exemplary works or a single author’s career creates a false canon and obscures structural change. To this point, Morelli contrasts traditional philology with the intellectual foundation of the social history movement. Fernand Braudel, Lucien Febvre, and the Annales school they nurtured argued that studying monumental events in history was futile and and that traditional historicism had missed the woods for the trees. They argued that revolutionary movements, events, moments, and people are not the most important factors in the history of civilization – rather, it is the long-standing, impersonal structures and customs of a civilization which determine the progress of events.
In this vein, Morelli believes that scholars of literature have been wrong to focus on a canon of exemplary works, to study the lives and works of a few excellent authors. These scholars should instead turn their attention to quantitative data, and above all, the forms the underlie literature. He presents three forms of representation as models for this type of analysis: graphs, maps, and trees.
He uses these representative forms to argue that cycles of literary history “constitute temporary structures within the historical flow” – periods of time between events (all flow and no structure) and the longue duree (all structure and no flow) – and that these are the most useful units of time to think about change and development in a field. For instance, these “temporary structures” are useful middle ground to examine genres, which he defines as “morphological arrangements that last in time, but always only for some time” (Franco Morelli, Graphs, Maps, Trees [London: Verso, 2005], 14). As novel forms outlived their “artistic usefulness” in English literary history, they were replaced by another genre form that followed the general rise and fall in popularity and hegemony of its predecessor – a cycle of genre forms.
Here’s where Morelli’s argument for a largely quantitative, structural analysis begins to crack. His observation about novel genres outlasting their “artistic usefulness” is that “a genre exhausts its potentialities – and the time comes to give a competitor a chance – when its inner form is no longer capable of representing the most significant aspects of contemporary reality. At which point, either the genre loses its form under the impact of reality, thereby disintegrating, or it turns its back to reality in the name of form, becoming a ‘dull epigone’ indeed” (Morelli, 17).
This thesis is certainly enhanced by Morelli’s quantitative analysis of specific genres’ popularity in the English novel market, but could just as easily find support in specific works of literature; needs qualitative and textual analysis, in fact, to have any merit. One novel from the observably-waning days of the Gothic novel, for instance, might merit deep textual study in order to prove the point that the author had become more interested in form than in storytelling. Is Morelli rejecting that category of analysis?
To give one example of the fruitful potential of blending data analysis and traditional textual history, when historians of the colonial Chesapeake began adopting the Social Sciences’ analytical tools and graphical forms of representation in the 1970s, they made solid discoveries about settlement patterns, social shifts, and cultural forms. I think that Morelli proves that graphs are useful tools for revealing temporal change from quantitative data, maps for spatial understanding of a text, and trees for morphological analysis of literature – based on divergence of traits. Each of these forms of representation can lend compelling evidence to support a thesis about literary history or any structures over time.
But the quantitative facts and trends that Morelli isolates always need qualitative answers, based in deep textual and contextual reads of a particular book or set of works, in order to make a valid point about the broader history of literature. The representations of data are, in the end, useful illustrations of important information, but cannot stand alone.
Carlo Ginzburg, in his article “Morelli, Freud and Sherlock Holmes: Clues and Scientific Method,” (History Workshop, No. 9 [Spring, 1980], pp. 5-36) comes closer to my understanding of history, which requires specific examples as well as broad, quantitative structural analysis to chart changes over time:
“Historians cannot help sometimes referring back (explicitly or by implication) to comparable series of phenomena; but their strategy for finding things out, like the volumes in which they present their work, is basically about particular cases, whether concerning individuals or social groups, or whole societies. In this way history is like medicine, which uses disease classifications to analyse the specific illness of a particular patient. And the historian’s knowledge, like the doctor’s, is indirect, based on signs and scraps of evidence, conjectural. “(36)
Sharing History 