When writing history, or indeed when doing any sort of analysis, the practice of organizing what is often a glut of information into something digestible necessitates a sort of causality to be invented. To put it another way, since the process by which information becomes accessible is a form of narrativization, all events must have causes and effects, the result of some grand machination or a hand in a preordained endpoint.

While such a view of history gives a sense of orientation, scholars such as Prasenjit Duara in his work Rescuing History from the Nation, or Yongnian Zheng in his monograph The Chinese Communist Party as Organizational Emperor have shown how the construction of time as linear and evolutionary is simply one way to imagine history to the exclusion of other forms, with linear history thereby lacking the power to fully account for why events happen.

Rather, history is something better perceived as open-ended. From the CCP’s victory in 1949 to Donald Trump’s 2024 election, we can only make sense of stunning upsets after they have occurred, and in attempting to explain them, we reduce them to one-dimensional caricatures. What was once flesh is transformed into a hyperreal version of itself; historiography thus exists not as a mapping of reality but a reality unto itself.

Indeed, historical accounts, and historiography more generally, are often no more than a means of tidying what would otherwise be incomprehensible noise, but do little to tell us anything about the world we live in today or where we are going. Such is perhaps why fields like political science are so obsessed with constructing nebulous predictive models, trying to use the guise of scientific rigor to make predictions that are often, as Hedley Bull argued six decades ago, either only useful in extremely limited cases, or are consistently upset by major world events, such as the outbreak of the war in Ukraine or the fall of Afghanistan to the Taliban. Rather, the driving force behind such events lies in their accident.

But how, then, do accidents come to be? We can understand this through analogy with infrastructure. Brian Larkin, in his meditations on infrastructure, argues infrastructural projects often come to be used in variance with their original intent. Ling-I Chu and Jinn-Yuh Hsu elaborate on this further in their article “Accidental Border: Kinma Islands and the Making of Taiwan,” which examines how the Taiwan-China border, as embodied by the Taiwanese islands of Kinmen and Matsu, exists as a form of infrastructure whose accidental activation reconfigured spatial perceptions of what constitutes Taiwan and its relationship to China as Taiwan democratized.

This means that the accident exists as an embedded potential within infrastructure, something waiting to be activated. In physics, when a ball gets stuck in a groove on a hill, it is said to have potential kinetic energy due to gravity that is unexpressed. Instead, it awaits its activation, a happenstance trigger, perhaps by a child dislodging it. But once it is dislodged, its potential energy is released, sending it barreling down the hill. So it is with historical events.

The broader world is perhaps similar to this, with the infrastructures of our society, both physical and representational, awaiting activation. In a January 2025 FT op-ed, Peter Thiel referred to Trump’s 2024 election victory as apokálypsis, an “unveiling.” Trump’s election may have been the activation of a series of latent accidents lying dormant in American society; at the moment of activation, it “unveiled” the hidden pathways that enabled the accident. Likewise, Reform and Opening, while depicted in CCP hagiographies as something preordained by the center, was not the result of careful planning but the activation of a complex interplay between state and society.

It is hard to imagine Deng Xiaoping, as one of Mao’s most loyal enforcers prior to his purge in the Cultural Revolution, would have truly wished to completely overturn the economic direction Mao had set, even if he had wanted reform. Many of my Chinese professors at the Hopkins-Nanjing Center have emphasized to me that none of the Chinese leadership from the 70s could have imagined or predicted China today, or how radically China would change in such a short period. Reform, once initiated, got ahead of them, becoming a logic unto itself.

Indeed, the fact it was Chinese people getting ahead of the party to initiate this economic transformation shows that China’s rise over the past near 50 years was not solely the result of some careful planning, of a Long Game, but instead could be read as the activation of an accident that grew out of the chaos of the Cultural Revolution, producing exactly the opposite of the Cultural Revolution’s intended result. If it had not been for this, perhaps today China would more resemble a Soviet-style planned economy.

This perhaps also explains why historians have traditionally tended to be theory averse; attempting to explain the underlying structure would open a Pandora’s box of complexity that would make the whole affair unmanageable and perhaps tempt people into a nihilistic view of human history, something anathema to the neat worlds which exist in historical monographs, or—even more severely—in popular textbooks.

But are accidents, and the fact that history is primarily driven by them, truly something to fear? Embracing events as contingent helps not only to expose histories as representations but also frees us to realize that the future is truly open. While this means we are not destined for utopia, neither does it mean we are doomed to destroy ourselves. Furthermore, if history is narrative, that means that history exists as an imagined form of concrete events. It is we, through our stories, that give accidents meaning. By assigning them a narrative, we co-create history with them, and in doing so, open and restrict a myriad of future pathways. If this is so, it leaves the possibility open to imagine a better world.