Accident Society: Fiction, Collectivity, and the Production of Chance

Stanford University Press

Stanford, California

© 2012 by the Board of Trustees of the Leland Stanford Junior University.

All rights reserved.

This book has been published with the assistance of University of Wisconsin–Milwaukee, Department of English, College of Letters and Science.

No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying and recording, or in any information storage or retrieval system without the prior written permission of Stanford University Press.

Printed in the United States of America on acid-free, archival-quality paper

Library of Congress Cataloging-in-Publication Data

Puskar, Jason Robert, author.

Accident society : fiction, collectivity, and the production of chance / Jason Puskar.

pages cm

Includes bibliographical references and index.

ISBN 978-0-8047-7535-9 (cloth : alk. paper)

1.American fiction—19th century—History and criticism.2.American fiction—20th century—History and criticism.3.Chance in literature. 4.Realism in literature.5.Literature and society—United States—History—19th century.6.Literature and society—United States—History—20th century. I. Title.

PS374.C39P87     2012

813.009—dc22

2011011483

E-book ISBN: 978-0-8047-7845-9

ACCIDENT SOCIETY

Fiction, Collectivity, and the Production of Chance

Jason Puskar

STANFORD UNIVERSITY PRESS

STANFORD, CALIFORNIA

ACCIDENT SOCIETY

To my parents

They gave me books.

Contents

Cover

Copyright

Acknowledgments

Introduction: Writing the Accident

1   The Insurance of the Real: William Dean Howells

2   Aimless Battles: Stephen Crane

3   Detecting Absolute Chance: Charles Peirce and Anna Katharine Green

4   The Feminization of Chance: Edith Wharton and Crystal Eastman

5   Performing the Accident on Purpose: Theodore Dreiser and James Cain

Notes

Index

Acknowledgments

I am grateful to the many people who supported me during the writing of this book. My wife, Erin O’Donnell, has been a gracious reader, a skillful editor, and an unfailing source of practical and moral support. I couldn’t have written this without her, and I wouldn’t have wanted to. Lawrence Buell, Philip Fisher, and Elisa New encouraged and guided this project from its earliest stages. They have been my most trusted models of scholarly curiosity and commitment. My graduate research assistant, Chad Jorgensen, was tireless in his work, especially for Chapter 2, and a marvel at sifting archives from afar. More people than I can remember have read generously, offered sage counsel, and otherwise supported this project in whole or in part, including Derek Barnett, Erica Bornstein, Chip Colwell-Chanthaphonh, Jenny Davidson, Andrew DuBois, Jonathan Eaker, Elizabeth Freeman, Debra Gettelman, Joyce Goggin, Christoph Irmscher, Oren Izenberg, Mark Jerng, Robin Kelsey, Robert Koelzer, Maurice Lee, Elizabeth Lyman, Jennifer Marshall, Katie Peterson, Jennifer Ratner-Rosenhagen, Matthew Rubery, Laura Thiemann Scales, Jeff Severs, Carl Smith, Sarah Song, John Stauffer, James von der Heydt, Elana Weiner, Sharon Weiner, and Michael Ziser. David Aldous and Benoit Mandelbrot graciously engaged some of these ideas from the field of mathematics, despite my inability to respond in the language of numbers with anything like equal fluency. My colleagues in the Department of English at the University of Wisconsin–Milwaukee have given me a rich intellectual home and have been valued interlocutors through the final stages of this project. I’m especially grateful to Sukanya Banerjee, Liam Callanan, Jane Gallop, Kristie Hamilton, Gregory Jay, Barrett Kalter, Gwynne Kennedy, Andrew Kincaid, José Lanters, Andrew Martin, Mark Netzloff, Patrice Petro, and Peter Sands, who have all, in different ways, shared their time and ideas with me. My sincerest thanks also to Emily-Jane Cohen at Stanford University Press and to the anonymous readers of the manuscript.

This project received generous financial and material support from the Whiting Foundation, the American Academy of Arts and Sciences Visiting Scholars Program, the UWM Research Growth Initiative, the UWM Graduate School Research Committee, and The Center for 21st Century Studies at UWM. Part of the first chapter appeared in American Literary History in 2006, and I am grateful for permission to republish it here.

For my smallest helpers, Owen and Jonas, the last word: While I was writing a book about chance, they reminded me of the special pleasures of chaos and of the rewards of having life turned upside down.

Introduction

WRITING THE ACCIDENT

Chance is made out of words. No mystical force, no scientific property, no real thing, chance is a conceptual category crafted through language and put to use in a wide range of institutional and political contexts. This book’s simple claim is that American writers in the late nineteenth and early twentieth centuries produced chance in new and specifically modern forms through narratives of spontaneous and blameless violence and that those narratives in turn supported emergent modes of social organization. In writing the accident, many novelists purported to detect and describe a radical instability at the heart of modern life, but in fact they were producing that instability by modeling it in narratives of causeless and blameless catastrophe.

By configuring industrial injuries, gun violence, transportation collisions, falls, financial losses, happy discoveries, fires, social coincidences, and even interpersonal conflicts specifically as matters of chance, novelists helped persuade popular audiences that a wide range of events happen for no reason at all. To consider just one example, what Americans routinely call industrial accidents do not come straight from nature as accidents. Over time lawyers, judges, journalists, sociologists, insurers, managers, safety reformers, novelists, and even workers themselves began to narrate those events specifically as matters of inscrutable chance. The 25,000 Americans who died from industrial injuries in 1913, or the 36,140 injured or killed on the railroads, became accident victims largely because Americans stopped describing their injuries in terms of fault and blame and started describing them instead with the language of chance.¹ By narrating violence in new and often experimental ways, American writers actively configured injury specifically as accident, both in particular, in the wake of a specific event, and in general, by classifying some kinds of injuries as accidents before they occurred.

The body of critical work on the literature and culture of chance is extensive, but to a surprising degree it has focused on the limited cultural context of gambling. This comes as little surprise given that gambling was so prominent in early works of classical probability theory from the late seventeenth and early eighteenth centuries. According to Gerd Gigerenzer, Gambling was the paradigmatic aleatory contract, and the very first problems solved by the mathematicians were of this sort.² But as a result, gambling has come to seem not just one way of thinking about chance but as the paradigmatic way. From historical work by Gerda Reith, Jackson Lears, and Ann Fabian, to literary criticism by Thomas Kavanagh, Walter Benn Michaels, and Jeffrey Franklin, a large and influential body of scholarship has concentrated on the important role of card games, casinos, lotteries, and many other kinds of gambling in Western culture.³ When Reith opens her history of gambling with the statement, We are all gamblers, she makes a claim with which many other historians, literary critics, and cultural theorists would readily agree.⁴

That argument is especially appealing in the period considered here, as the discourse of gambling frequently betrays a deeper concern with the changing conditions of liberal capitalism at the end of the nineteenth century. As the traditions of civic republicanism yielded to an even more individualistic and competitive ethos, gambling proved a ready metaphor for an entire financial system that seemed to be more atomizing, aleatory, and speculative than ever. Indeed, as Ann Fabian has argued, the outlawing of most forms of gambling by the 1890s actually indicates the extent to which the broader economy had absorbed gambling’s speculative ethic. As a result, Fabian says, the ‘new’ gamblers, who profited from the operations of stock and commodities exchanges, presented themselves as virtuous, rational citizens in a culture that was turning older values upside down.⁵ By translating old vices into new virtues, Americans made the gambler’s encounter with chance into a model for a revised set of liberal-capitalist values, which reflected a greater tolerance for risk taking, a corresponding decline in the valuation of steady labor and prudent saving, and a more pervasively isolating individualism than ever.

The following pages will not contest this view in any sweeping way but rather characterize the association of chance with gambling as only one part of a more complex history. From the very beginning, classical probability theory made use of another set of practical examples very different from gambling situations: mortality tables, first for annuities, and later for systems of insurance. When we think about chance not in terms of voluntary and highly individualistic acts of gambling but in terms of insurance and its allied forms of social and economic interdependence, we can see that chance has a far more varied character within popular culture than we have yet recognized, and far more salutary political consequences than we have tended to suspect. This story of chance might be traced back not to an early treatise on gambling such as Christiaan Huygens’s 1657 De Ratiociniis in Ludo Aleae but to something like Johann De Witt’s 1671 mortality tables for Dutch annuities.⁶ If gambling is a way of producing liberal individuals willing to assume disproportionate amounts of risk and responsibility, then other cultures of chance were capable of doing just the opposite, by opening up new possibilities for emergent modes of social and collective organization.

Chance collectivity, as I call it here, thus resists the individualizing functions of gambling by affording American culture new opportunities for fashioning systems of social and material interdependence. At the very moment when the values of civic republicanism were yielding to the values of Gilded Age competition, doctrines of social Darwinism, and the business protections of the Lochner Era, a wide range of writers and reformers recognized that public consciousness of chance might exert an oppositional force, consolidating, rather than eroding, effective structures of mutual affiliation. From the private risk pools of the nineteenth-century life insurance industry, to models of collective scientific inquiry, to the new public workers’ compensation programs of the 1910s, chance collectives mobilized Americans to join together against the lurking threat of chance. This proved especially useful in a fractious nation like the United States, riven as it was by racial and ethnic differences, divided into hostile political factions, and troubled by lingering regional antagonisms after the Civil War. In a climate rife with such hostility and distrust, the abstract menace of chance functioned as a common enemy against which the diverse constituency of the chance-afflicted might unite.

Each of the following chapters is a case study in the intersection of American fiction with various cultures of chance from Reconstruction to the New Deal. Novels and fiction by William Dean Howells, Stephen Crane, Anna Katharine Green, Edith Wharton, Theodore Dreiser, and James Cain directly engage, and so help us understand, key cultural sites of chance production, including the insurance industry, American gun culture, pragmatist philosophy, and industrial accident reform efforts. None of the novels considered here are specifically about gambling; instead, they focus on cultures of chance more closely related to socializing and collectivizing endeavors. While communists further to the left dreamed of more orderly utopias, these largely progressive writers and reformers pressed chance into the service of evolutionary, rather than revolutionary, social change. By reading their works in the contexts of emergent cultures of chance collectivity, we can better understand how writers—not just statisticians and mathematicians—helped shape and maintain certain modern concepts of chance, apply those concepts to definite political ends, and in doing so, advance some of the early twentieth century’s most meaningful social democratic reforms. As a result, chance collectivity must not be seen as opposed to liberal capitalism in any sweeping way, because it is instead an expression of a different kind of liberal capitalism, one that meant to facilitate more egalitarian and interdependent social arrangements and that led, in some cases, to the implementation of significant parts of the American welfare state. The New Deal, however, marks the end of this project, not the beginning, for chance collectivity was most appealing during the decades leading up to the passage of large-scale public insurance programs. In the absence of well-funded public provisions, and in a legal climate hostile to such measures, the threat of chance was sometimes the only thing powerful enough to justify the systematic and cooperative sharing of resources.

Before going any further, we had better pause to clarify just what we mean by the term chance, a word bound to ring with dozens of competing and contradictory associations. The term descends from the Latin casus, the root of which is cadere, to fall or to happen. That term is closely related to accidere, to fall down, which is in turn the direct source of Modern English accident. Both words thus trace their roots to the same reference to falling, as my parallel use of the two terms is meant to reflect. For these purposes, however, the term accident will typically refer to a specific event, while the term chance will have a far less stable and determinate meaning, for reasons that we will shortly see. Because chance and accident are so devoted to the case, they necessarily resist generalization, and partly for that reason Aristotle regarded accident as no proper object of philosophical reflection. The problem with accident, according to Aristotle, is precisely that one cannot generalize about it, for the accidental is by definition nonessential, insubstantial, and so bound to vary from one case to the next. That, however, is the best reason of all to study chance historically, for if chance has no permanent and essential nature, it must unfold within history and in relation to particular cultures. Michael Witmore and Ross Hamilton have recently charted the legacy of Aristotelian thinking about accident in precisely this way, revealing it as all the more interesting because of its permutations across cultures and through time.⁷ But an important corollary to all of this is that one simply cannot define chance once and for all, for chance is nothing other than a moveable category of action demarcated and applied in different ways in different contexts. Definitions of chance thus must be historical definitions, particularized to a time and place, and always attentive to the term’s inner tensions, residual meanings, and constantly shifting boundaries.

Without belaboring a history that has been told more fully elsewhere, we should acknowledge that the modern history of chance involves, to a significant degree, the denial of chance’s very existence. In the early modern period, religious thinkers such as John Calvin had seen the accident as a divine disruption of the routine state of earthly affairs, which is to say, as special Providence.⁸ By the eighteenth century, new statistical and probabilistic methods partly secularized and rationalized a related version of that view. For Enlightenment thinkers, what seemed to be chance was not the mark of Providence after all but rather an indication that unapprehended laws and causes were yet at work in an entirely rational universe. Abraham De Moivre, author of The Doctrine of Chances (1718), wrote that chance was a mere word that can neither be defined nor understood and that imparts no determination to any mode of existence.⁹ David Hume later concluded, "Though there be no such thing as Chance in the world; our ignorance of the real cause of any event has the same influence on the understanding."¹⁰ And Pierre-Simon Laplace, one of the greatest eighteenth-century probabilists, opened his Philosophical Essay on Probabilities (1795) with this: All events, even those that on account of their rarity seem not to obey the great laws of nature, are as necessary a consequence of these laws as the revolutions of the sun. As a result, according to Laplace, what we call chance is only the expression of our ignorance.¹¹

This too rapid review of just a few important episodes in the early history of chance and accident is meant to indicate just the outlines of some extremely influential ideas, against which many later theorists of chance will rebel. Even so, these shifting conceptions of chance are not as easy to periodize as this brief account might suggest. Even in the United States in the period considered here, many different conceptions of chance circulated together, competed with each other, and interpenetrated at their borders, including the religious and scientific views of the seventeenth and eighteenth centuries. The minister and the statistician, the safety expert and the philosopher, might all use the word chance with perfect confidence that they knew what they were saying, but they would not necessarily be talking about the same thing. Accordingly, lingering theological and deterministic conceptions of chance certainly are evident throughout the twentieth century, and they can show up where we least expect them. To consider just one cautionary example, Walter Benjamin regarded gambling as an outgrowth of capitalism, but he also saw gambling as a source of wonder, and even of mystical knowledge. Benjamin wrote that gambling generates by way of experiment the lightning-quick process of stimulation at the moment of danger, the marginal case in which presence of mind becomes divination—that is to say, one of the highest, rarest moments of life.¹² Many avant-garde surrealists, dadaists, and futurists also sought escape from dreary routine by romanticizing chance similarly, as we will see in the final chapter, but when Benjamin associates the moment of the gamble with divination, it is impossible not to hear this great voice of Western Marxism and Messianic Judaism oddly invoking a very traditionally Calvinist view.

Chance is thus not just historically contingent but also heavily sedimented with the accumulated residue of its own past. Attempts to refashion chance anew are thus often laden with remnants of earlier views. Accordingly, the mostly realist novelists considered here attempted to carve out a conception of chance that was neither theological, deterministic, nor purely mathematical and statistical, but that still rejects the traditional Aristotelian dispensation. In accounts that still vary considerably, these writers began to configure chance as a radical indeterminacy principle, and so, as far more than simply a name for human ignorance of divine intentions or operative laws. In that way, these writers are important sources of what Gerda Reith has identified as the most distinguishing feature of chance in the twentieth century, its purported ontological status. According to Reith, chance came to seem real after all, an indeterminacy principle actually resident within the nature of the universe, and not a name for some higher law misapprehended nor for a mere method of probabilistic assessment. There will be more to say about the intellectual sources of the ontological view in the third chapter, but for the moment we need only note that the ontological view makes chance more ideologically potent than ever, for by naturalizing chance, it threatens to exempt it entirely from cultural analysis. We would do better to see all conceptualizations of chance as historically contingent and politically implicated, especially those most familiar to us, precisely because they do not yet appear to the naked eye in the sepia tones of the distant past. Over time, perhaps the ontological view will prove more durable than those it has only partly displaced, but I have no great optimism that this will turn out to be the case, and suspect instead that ontological chance eventually will appear only as a peculiar habit of thinking from the twentieth century, and no better or worse than those that came before. At that point, it may be clearer than it is now that ontological chance reflects deep and unarticulated cultural assumptions and serves profound political purposes, some of which these pages will bring to light.

In the broader view, then, Abraham De Moivre had it exactly right when he called chance a mere word, for chance is nothing other than a fluid category of action and eventfulness, generated to a considerable degree through language and applied to a wide range of social, economic, and political contexts. Rather than ask what chance is, then, it seems more fruitful to ask what chance does, and what chance does in the works considered here is forestall investigation into causes and culprits that otherwise would render some injury amenable to the prevailing order of reason or morality. Chance creates dynamic instabilities at many different levels—knowledge, ethics, identity, economics, power—that are not just disruptive but finally useful to a loosely allied set of progressive reforms. Narratives of chance are thus prescriptive and not just descriptive, for they obscure agency and responsibility to precisely the same degree that other kinds of narratives clarify causation and establish blame. In using the term chance, then, I do not mean to imply that it is in any way an essential property of the natural world. Its ontology is a historical ontology, a mode of being contingent upon changing cultural and historical circumstances.¹³ Obviously, many of the writers who use these terms are committed to other definitions, but my goal is to show how their attempts to naturalize chance by affirming its ontological status are better understood as expressions of particular cultural investments, and better studied in terms of their social and political consequences.

This is not a scientific definition of chance, then, but a cultural definition that attempts to describe how chance functions as a rhetorical strategy most evident, and most powerful, in literary narratives. Mathematicians, statisticians, economists, philosophers, and representatives of many other disciplines might object that chance has been, and should be, defined much more narrowly and technically, but my concern lies with the term’s function outside these professional cultures and with its consequences for the kinds of public debates that lay readers could access and that novels could facilitate. Others might object that chance does have a materialist or realist basis, in the Brownian motion of molecules, the fluctuating prices of cotton futures, or the quantum dynamics of particle physics, and that chance should be understood as the name for some specific property of the real world. Although this project cannot address all such claims directly, it assumes that trying to find a material or scientific basis for chance is no more advisable than trying to find a material or scientific basis for race, for what matters is that chance names a category of action that consists entirely of its cultural ascriptions. As such, and in all the ways that matter, chance derives its meaning not from nature but from the cultural work to which it has been applied. Chance is real, to be sure, but only in the sense that it has powerful social and cultural functions, which are in turn the products—not the objects—of our methods of classification.

Literary works thus have an extremely important role to play in the production of chance, because they help determine what does and does not count as chance in the first place, especially for popular audiences. Accordingly, the narrative production of chance turns out to function rather differently than the taming of chance that Ian Hacking has associated with the avalanche of printed numbers that descended on the West in the early nineteenth century and led to new technologies for classifying and enumerating, and new bureaucracies with the authority and continuity to deploy the technology. Paradoxically, as Hacking claims, the increased attention to chance from statistics and probability theory actually reassured everyone that the world could be regulated after all. Through probability theory, Hacking says, the more the indeterminism, the more the control.¹⁴ Drawing on Hacking’s work, literary scholars have sometimes tended to associate qualitative narratives with quantitative numbers, especially in the eighteenth and nineteenth centuries. Thomas Kavanagh has influentially cast literary representations of chance as directly equivalent to probability theory. Like probability theory, the novel promised a greater understanding and mastery of life’s apparently random events,¹⁵ Kavanagh claims, because both the novel and probability theory worked toward a domestication of chance, toward the elimination of its threat to the Enlightenment’s faith in a rational and knowable world.¹⁶ Narrative theorists also have been quick to see literary and historical writing as similarly rationalizing. As Hayden White has argued, events in a narrative must be not only registered within the chronological framework of their original occurrence but narrated as well, that is to say, revealed as possessing a structure, an order of meaning, that they do not possess as mere sequence.¹⁷ Because narrative presumes that all events have such an inner logic, the disclosure of which is narrative’s primary function, to narrate is always to moralize, according to White’s well-known formulation. Accordingly, if the work of historical writing is to restore rational and moral clarity, accidents become especially enticing to those determined to narrate history, because their very blankness causes the dowsing rod of historical curiosity to dip tellingly. At the site of the accident historians read: dig here. In that sense, chance may be absolutely essential to the work of history, but mainly in the sense that it is the very thing that narrative history works to eliminate.

These arguments hold great weight for the literature of earlier periods, but they do not tell the entire story for the novels of the period considered here. When we define chance not as a radical antilaw that defies representation but as the narrative practice of concealing or obscuring causal relations within broader cultural contexts, we can see that narrative has far more resources at its disposal. To think about narrative as producing chance—rather than simply making it amenable to reason—is to recognize that chance is a name for the active decoupling of actions from agents, and of events from overarching moralizing and rationalizing schemes. Narrative’s capacity to demoralize and derationalize turns out to be no less pronounced than its capacity to do the opposite. As a result, we find a very different state of affairs in the American literature of this later period than in French fiction of the eighteenth century and, indeed, than in many other naturalist and modernist American novels. These primarily realist novels are powerful instruments for refashioning the social order, to be sure, but rather than assuage fears of chance, they tend to inflame them. If, as Hacking says, the more the indeterminism, the more the control, then the qualitative, literary production of chance could be a powerful first step in justifying new and potentially salutary kinds of social organization.¹⁸ By insisting that the future is largely out of control and that connections between causes and effects are nonexistent or entirely obscure, these novels do not reinforce liberal fantasies of mastery but undermine them. And as a result, by fostering chance consciousness, novels helped new kinds of liberal affiliation flourish, precisely because they challenged powerful fantasies of the liberal individual’s alleged independence, rational prowess, and might.

The City That a Cow Kicked Over

At this stage it may be useful to consider the historical situation more specifically, along with an example that can make some of these introductory claims more concrete. The question of why Americans preferred to represent injuries as accidents and, further, how such accidents came to function as instruments of social collectivity, involves the cultural transformations under way in the United States after the Civil War. In the wake of that long and traumatizing conflict, accidental violence may have seemed a positive relief from deliberate and divisive slaughter. As Louis Menand has argued, The Civil War swept away the slave civilization of the South, but it swept away almost the whole intellectual culture of the North along with it, revealing in the wake of abolition’s triumph a corresponding failure of the American experiment: people who live in democratic societies are not supposed to settle their disagreements by killing one another.¹⁹ Menand condenses the lesson learned from the Civil War into a pithy mission statement for the relativistic and pluralistic philosophy of pragmatism that flourished at the end of the century, certitude leads to violence.²⁰ If that seemed to be the case to many Americans after the Civil War, chance offered a surprising remedy. Precisely because accidental violence is not rationalized or moralized, it cannot be laid at the feet of one’s enemy, nor does it foster the kinds of certainty that might justify violent retaliation. If certainty divides, the uncertainty of chance, paradoxically, might unite. The rational and moral confusion surrounding accidental violence thus could function as an antidote to the divisive certainties of moral conviction and, accordingly, could prove a valuable instrument of social solidarity, which Americans sorely needed in the 1870s.

There will be more to say about pragmatism later, which has its own complex entanglements with the cultural history of chance, but for the moment we can see these dynamics in play as early as 1871, and well outside the scholarly confines of philosophy. In the midst of Reconstruction, the United States endured what must have seemed an uncanny replay of the burning of the cities of the South, but this time in the fastest-growing metropolis in the North, the city of Chicago. The Great Chicago Fire of 1871 displaced one hundred thousand people from eighteen thousand destroyed buildings, a number that equaled the entire population of London during its fire of 1666. The fire burned a footprint almost four square miles in size, which included almost all of Chicago’s prosperous downtown and much of its immigrant outskirts.²¹ Whereas the London fire moved so slowly that only six people died, a northwest wind fanned Chicago’s blaze into a blast furnace that consumed the city in less than forty-eight hours, chasing a mob of refugees north of the city and catching more than three hundred of them. Uncounted others may have been immolated in a blaze so hot it buckled the metal frames of the first buildings termed fire-proof.²² Only the Lisbon earthquake of 1755 destroyed a city of comparable size and killed more, so it was to Lisbon that Chicagoans looked to console themselves that it could have been even worse.

Urban fires on this scale have generated telling stories of causation. A God incensed by sexual vice burned Gomorrah. The mad emperor Nero kindled the great fire of Rome in 64 CE, then fiddled while it burned. In 1666, the fates of that annus mirabilis burned much of the city of London to the ground. But after the greatest fire of them all in Chicago in 1871, the American popular imagination embraced the absurd legend of a dairy cow accidentally kicking over a lantern in an Irish immigrant’s barn. Even though that tale has acquired very little evidence in its favor in the years since, and was refuted in Chicago almost from the moment it first appeared, the story of Mrs. O’Leary’s cow rapidly became the symbolic narrative for the most sensational urban catastrophe the modern world had yet endured, and the first New World disaster of international note. But by the standards that prevailed in the aftermath of the world’s other great fires, it is difficult to understand exactly why Americans would attribute the utter destruction of an entire metropolis to an accident, let alone to such an absurd and apparently trivializing mishap as this. Yet if the stories about the fire of Rome reflected anxieties about imperial caprice, and if the stories about the fire of London channeled Puritanical fears of an angry God, the story of Mrs. O’Leary’s cow activated, for the first time on a national scale, the dynamics of chance collectivity that would shape American social life through the coming decades.

Unfortunately, the O’Leary story has frequently been overinterpreted as a crude attempt to foist blame on an impoverished Irish-Catholic immigrant woman, a sinister and little known ‘other’ that might in fact bring down an entire city, as Karen Sawislak has argued.²³ But that seems to be true only in the immediate wake of the fire, as baffled and outraged survivors turned their anger against the usual suspects, including communists, anarchists, immigrants, and Jews.²⁴ Some early visual representations of Mrs. O’Leary thus depict her as a hag or a witch, and the few narratives of the fire’s cause configure her, variously, as a vindictive welfare cheat or a stereotyped Irish drunk. But none of these pejorative characterizations survived in popular culture, and those accounts that blame the fire