Mississippi newspapers · 1830–1969

Mississippi’s Exclusion Curve

Press attention peaked during federal exclusion, decades before the Mississippi Delta Chinese population reached its height.

314,289 digitized pages 254 newspaper titles 39 archives
Archive attention

The peak came before the community’s growth

Pages with a Chinese identity term, per 1,000
The archive peaks in the 1880s; the published Delta population observations rise later. The units differ, and the comparison is descriptive.
View data
Local record · 1920–1927

The Delta enters the record

Eight verified sources document Chinese Mississippians organizing during the school-segregation fight.

Archive gaps

The case years are missing

1924–1926

Zero candidate pages

No candidate pages survive in the assembled corpus.

Rosedale · 1924–1927

Hometown paper unavailable

The Bolivar County Democrat is digitized only through 1923.

Before 1933

No Black press in this corpus

The corpus cannot support a Black/white press comparison.

These are limits of the assembled digital corpus, not evidence of historical silence.

The peak is not a general archive effect

Each term’s 1880s rate is indexed to 100.

Chinese alone peaks in the federal exclusion era. Italian also spikes in the 1850s, often in dry-goods advertising.
View data
More dataKeyword contexts and reprint clustering.

What surrounded Chinese identity terms?

Six keyword families; passages may match more than one.

Before 1870, many Chinese hits are product labels.

How language circulated among Mississippi papers

Repeated passages show circulation, not origin.

Equal sample sizes; OCR loss makes each share a lower bound.
Select a cluster to inspect its newspaper pages.

Data & methods

Download population context Download term audit
Methods and limits
  • Denominator. All ocr.txt pages in 39 verified msar_* archives: 314,289 pages, 254 titles.
  • Terms. The plotted Chinese series includes Chinese, period identity terms, and Mongolian(s), but excludes standalone Celestial(s). The exclusion is rebuilt from saved passages and recorded in the downloadable audit. Controls are Irish, Italian, and cotton.
  • Population. The plotted Delta points are published historical summaries, not a newly executed IPUMS extract. Statewide and Delta values remain separately labeled; they are not one continuous series.
  • Reprints. ±300 characters; letter-only 5-grams; MinHash LSH; Jaccard ≥ 0.35; core-window recheck. 98.4% of clusters pass. The reprint browser retains the original high-recall term set, so it is an exploration aid rather than a revised prevalence measure.
  • Limits. These are archive rates, not statewide press prevalence. The filtered 1930s context rests on 254 passages. Corpus gaps are rendered as missing coverage, never as zero historical activity.

Full keyword lexicon