About this site

This site curates and interprets 1940 U.S. Census entries for people with the last name Singh, most of whom were either immigrant Punjab Sikh men, or their family members.

The dataset is unique in that it includes many Punjabi-Mexican families, and offers a snapshot of the community at a time during those decades when legal immigration slowed down between 1917 and 1946.

This site and underlying dataset was developed by Anirvan Chatterjee of the Berkeley South Asian Radical History Walking Tour, as part of his ongoing research on early South Asian migration to the United States.

Census categories

Many entries include household structure (head, spouse, child), age, gender, birthplace, and the race label assigned by census officials. Categories like race and birthplace were recorded inconsistently by census takers, sometimes reflecting the racial and political climate of the era. This site does not include all census categories; you can look up individual records on websites like FamilySearch.org or Ancestry.com to see more.

Interpretive stance

The census is both a source and a technology of power. For South Asian American studies, it offers traces of migration, labor, and family formation, but it also encodes the state's racial logics. This project keeps both realities visible: it shows the data and the patterns, and it foregrounds how those patterns are shaped by the census itself.

• Race labels are shown as recorded, with inferred groupings used only to help explore the data.
• Birthplaces are mapped with historical ambiguity in mind (for example, we map potential references to "Indiana" to "India," while still showing the original label in display text).
• Inferred family links are conservative: typically, only reciprocal matches and shared census locations are used.

Digital humanities lens

This site is a light-touch DH intervention: it makes the dataset navigable at multiple scales (individuals, households, and aggregate patterns), and keeps interpretive claims close to the data. The maps and charts are not arguments by themselves, but prompts for further reading, contextualization, and archival work.

The design encourages slow looking and comparative reading. For example, the group cards let you pivot between categories, while the household ordering keeps family relations legible without over-asserting them.

How to cite and reuse

While the main dataset is in the public domain, you are encouraged to cite the source. The GitHub repository provides the underlying JSON, documentation, and provenance notes.

For example:

Chatterjee, Anirvan. (2026). 1940 Singh Census. Retrieved from https://1940-singh-census.berkeleysouthasian.org/ and https://github.com/anirvan/1940-singh-census .

If you are using the OpenStreetMap geodata embedded in the JSON dataset, please include a credit to OpenStreetMap contributors:

Map data from OpenStreetMap (openstreetmap.org/copyright).