For most of the twentieth century, historians operated with a seemingly uncontroversial assumption: language was a window onto the past. Words pointed to things that happened. The historian's task was to find the right words—the accurate, objective, well-sourced words—to represent reality as it truly was.

Then, beginning in the 1970s, this assumption came under sustained philosophical attack. Drawing on developments in linguistics, philosophy of language, and literary theory, scholars began arguing something far more radical: language doesn't simply reflect reality. It constitutes it. The categories we use to describe the past don't neutrally capture what happened—they actively shape what we can see, think, and know about it.

This challenge, often called the 'linguistic turn,' sent shockwaves through the historical profession. Some historians dismissed it as dangerous relativism. Others embraced it as liberation from naive empiricism. The debates that followed fundamentally transformed how we understand the relationship between historical writing and historical truth. Understanding this transformation remains essential for anyone grappling with what historical knowledge actually is—and what it can never be.

The linguistic turn drew primarily on the work of Swiss linguist Ferdinand de Saussure and French philosophers like Jacques Derrida and Michel Foucault. Their central insight was deceptively simple: the relationship between words and things is arbitrary and conventional, not natural or necessary.

Saussure demonstrated that linguistic signs derive meaning not from their connection to external reality but from their difference from other signs within a system. The word 'revolution' doesn't mean what it means because it corresponds to some essence of revolutionary activity. It means what it means because it differs from 'reform,' 'rebellion,' 'coup,' and other terms within our conceptual vocabulary.

For historians, the implications were profound. If meaning is generated through linguistic systems rather than reference to reality, then historical narratives don't simply describe the past—they construct it according to the conceptual categories available to the historian. The French Revolution wasn't inherently a 'revolution' until that category existed and was applied to certain events.

Hayden White pushed this argument furthest in his landmark work Metahistory (1973). White argued that historical narratives share deep structural features with literary fiction. Historians must 'emplot' their material—organizing events into recognizable story forms like tragedy, comedy, romance, or satire. This emplotment isn't discovered in the evidence; it's imposed by the historian's narrative choices.

The scandal of White's position was clear: if historical writing is fundamentally a literary act, what distinguishes it from fiction? White never denied the reality of past events, but he insisted that any meaning we derive from those events comes from the narrative structures we bring to them. The past itself is silent about what it means.

Takeaway

Language doesn't neutrally describe reality—it organizes what counts as real in the first place. The categories historians use aren't innocent tools; they're constitutive frameworks that make certain histories visible while obscuring others.

If language constitutes rather than reflects reality, then the history of language itself becomes crucial. This insight generated an entirely new mode of historical practice: discourse analysis. Rather than using sources as windows onto past events, historians began examining how language itself shaped historical experience.

Michel Foucault's work proved particularly influential. Foucault argued that 'discourse'—systematic ways of speaking and thinking about particular topics—doesn't merely describe pre-existing objects but brings them into being as objects of knowledge. 'Madness,' 'sexuality,' 'criminality'—these weren't natural categories that discourse simply named. They were products of particular discursive formations that emerged at specific historical moments.

The methodological shift was dramatic. Traditional historians asked: What actually happened to mad people in the eighteenth century? Foucauldian historians asked: How did 'madness' become a coherent category of experience in the first place? What discursive practices—medical treatises, legal documents, architectural designs, institutional regulations—constituted 'the mad' as a recognizable type of person?

This approach transformed the study of identity categories. Joan Scott's influential essay 'Gender: A Useful Category of Historical Analysis' (1986) argued that historians shouldn't treat gender as a natural fact to be documented but as a discursive construction to be analyzed. How have different societies constituted gender difference through language? What work do gender categories do in organizing social relations and power?

The emphasis on discourse also highlighted what earlier historians had largely ignored: the power embedded in language. Dominant discourses don't just describe reality—they authorize certain speakers, marginalize others, and determine what can be legitimately said and thought. History becomes, in part, the study of these linguistic power struggles.

Takeaway

Discourse analysis reveals that the categories historians study—madness, sexuality, race, nation—aren't natural facts but historical constructions. The power to define categories is itself a form of historical power worth examining.

By the 1990s, the linguistic turn had become orthodoxy in many corners of the historical profession—and orthodoxies inevitably generate reactions. Critics increasingly argued that the emphasis on language had gone too far, obscuring the material, embodied, and affective dimensions of historical experience.

The most pointed criticism concerned the relationship between discourse and reality. Did the linguistic turn imply that nothing existed outside language? Critics charged that radical textualism couldn't account for stubborn material facts—bodies that suffer, environments that constrain, technologies that transform. The Holocaust didn't become a catastrophe because of how it was narrated; it was a catastrophe because millions of people were systematically murdered.

Historians began developing what some call 'post-linguistic' approaches. The 'material turn' renewed attention to objects, bodies, and physical environments as historical actors that exceed their discursive representations. Bruno Latour's actor-network theory, for instance, insisted that non-human entities—microbes, machines, documents—possess agency that discourse analysis alone cannot capture.

The 'affective turn' similarly challenged linguistic reductionism. Emotions and bodily sensations, scholars argued, are not simply constructed through discourse but have their own irreducible reality. The terror of combat, the grief of loss, the joy of liberation—these experiences resist complete capture in language.

Yet these developments don't simply reject the linguistic turn; they build upon it. Few historians today would return to naive assumptions about language as transparent medium. The lasting legacy of the linguistic turn is a permanent sensitivity to the constitutive power of categories, concepts, and narrative forms—even as historians pursue dimensions of experience that language alone cannot exhaust.

Takeaway

The linguistic turn's legacy isn't the claim that everything is language but the permanent recognition that our conceptual categories are neither innocent nor inevitable. Historical understanding requires attending to how language shapes reality while remaining alert to what exceeds it.

The linguistic turn didn't destroy historical knowledge—it deepened our understanding of what historical knowledge is. We now recognize that historians don't simply recover the past; they construct representations of it using inherited narrative forms and conceptual categories. This isn't a defect to be overcome but a condition to be acknowledged.

What remains after the linguistic turn is a more philosophically sophisticated discipline. Historians can no longer claim naive objectivity, but they can pursue something more honest: reflexive awareness of how their own interpretive frameworks shape what they see. The best historical work today combines rigorous archival research with critical attention to the categories that organize that research.

The question the linguistic turn posed—what words do to the past—has no final answer. But asking it persistently keeps historical practice intellectually alive, alert to its own assumptions, and honest about its limitations.