Purpose.

This poster introduces a digital humanities project that maps how fires were reported—and imagined—in Manila’s colonial newspapers between 1898 and 1941. Beyond recording disasters, fire reports framed vernacular neighborhoods as risky and combustible, shaping both public perception and urban planning. The project asks: how did newspapers construct fire as both a material hazard and a justification for modernization?

Methods.

The study assembles a corpus of English-language newspapers (Cablenews-American, The Tribune, Freedom, Manila Times) and extracts articles containing keywords such as “fire,” “conflagration,” “burn,” and “blaze.” Python-based text mining is applied to track term frequencies, collocation clusters, and sentiment patterns around descriptors like “nipa,” “shanty,” “crowded,” “unsanitary,” and “criminal.” Named-entity recognition identifies place names, which are then geocoded and mapped in QGIS (spatial analysis) and StoryMapJS (interactive presentation) against colonial zoning boundaries and planning proposals.

Results.

Preliminary analysis reveals two consistent patterns. First, fire reporting was spatially uneven: working-class districts such as Tondo and Binondo dominated the press, while elite residential zones appeared mainly as threatened by encroaching flames. Second, collocation networks show that “nipa house” (light, stilted dwelling made of palm and bamboo—was reframed as "shanty") was persistently linked to “hazard,” “danger,” and “problem,” recasting a vernacular form valued for affordability and resilience as evidence of backwardness. A prototype fire map highlights incident clusters in peripheral neighborhoods, aligning closely with areas targeted for drainage works, firebreaks, and street widening.

Conclusions.

By combining distant reading and mapping, this project recovers a missing layer of Manila’s urban history: the geography of fire as constructed in colonial discourse. No comprehensive historical atlas exists for Manila; only scattered archival maps survive, and modern studies begin only in the late twentieth century. Mining the colonial press provided an alternative path for reconstructing fire risk geographies. In doing so, the project shows that fire coverage was less about neutral reporting than about encoding bias and legitimizing intervention.

Impact.

The poster will present three key outputs:

1. Frequency timelines of fire coverage across decades.
2. Collocation clusters linking fire to material and social descriptors.
3. A prototype map of reported fire sites overlaid with planning zones.

These visualizations offer not only a new tool for understanding Manila’s built environment but also a replicable workflow for other cities without historical fire-insurance atlases. Looking ahead, the project will expand by linking articles to cadastral maps and incorporating postcolonial fire data, creating a long-view “Atlas of Fires in Manila.”