How to Search Polish Names in American Newspaper OCR

US Records · Newspaper OCR search

Digitized newspaper search depends on OCR, software that guesses the words printed on a scanned page. Small type, damaged paper and Polish diacritics produce errors, so a perfect surname search can miss the exact notice you need.

Front page of the Polish-language newspaper Czas Baltimorski, August 15, 1940
Front page of Czas Baltimorski, a Polish-language weekly of Baltimore, August 15, 1940, digitized in Chronicling America. Dense type and Polish diacritics are exactly what trips up OCR. Library of Congress

Why Polish names fail in OCR

Columns may be crooked, letters may touch and older typefaces make combinations such as rn, m and ni difficult to distinguish. Diacritics can disappear or become punctuation. The same person may also appear under a Polish spelling, an American spelling and a printer’s mistake.

Linotype operators at work in the New York Times composing room, 1942
Linotype operators in the composing room of the New York Times, September 1942. Hot-metal type set this way can print with touching or broken letters that confuse OCR. Marjory Collins, FSA/OWI, Library of Congress

Search from broad to narrow

  1. Search the complete surname without diacritics
  2. Try stable fragments from the beginning and end of the name
  3. Substitute likely letter confusions such as cz, sz, rz, w and v
  4. Combine a short name fragment with a street, spouse, parish or lodge
  5. Limit the date only after the search produces too many results

Use the event, not only the name

For a death notice, search the address, funeral home, cemetery, parish and names of surviving children. For a wedding, search the church and the bride’s address. Society notices may identify a lodge number that appears clearly even when the surname does not.

Front page of the Polish weekly Gwiazda of Holyoke, Massachusetts, April 14, 1945
Front page of Gwiazda, the Polish weekly of Holyoke, Massachusetts, April 14, 1945, reporting the death of President Roosevelt. Search both Polish and English wording in papers like this. Library of Congress holdings

Search English and Polish wording

A Polish-language death notice may use terms such as zmarł, pogrzeb or świętej pamięci, while an English paper may use died, funeral or services. Search both languages and use the newspaper’s date format. Browse the expected issue when OCR remains unsuccessful.

Keep an image and citation

Save the whole notice and enough of the page to identify the newspaper, date, edition and page number. OCR text is a search aid, not a reliable transcription. Read names and places directly from the scan.

A newspaper notice was supplied by a family, society or reporter and can contain errors. Confirm dates and relationships with civil, parish and cemetery records.

Sources and research tools

How we can help

We search Polish and English newspapers with name variants, addresses and organization clues, then translate and document the original notice.

Keep reading

Polish-American newspapers

Funeral home records

Americanized Polish names