The Women Behind Texas Czech Preservation: Dr. Svatava Pírková Jakobson and Dr. Lida Cope

When people ask about my work applying AI to Texas Czech preservation, they often want to know about the technology. But understanding the AI requires understanding the foundation: decades of meticulous archival work by two scholars, women, who recognized the value of a disappearing dialect.

March is Women’s History Month, and here is a bit about each of them.


Dr. Svatava Pírková Jakobson: Folklorist and Archive Builder

Svatava Pírková Jakobson came to the University of Texas after a remarkable career studying folk music and culture. In the 1930s, she conducted sociological-aesthetic research on Moravian folk songs in Czechoslovakia. Then came the war.

She later wrote about years of displacement: “A life of uncertainty inside and out… fleeing Norway on foot with Germans right behind us, our friends there arrested and killed… our ship stopped and searched by Germans in the open Atlantic, and finally that incredible feeling of freedom and peace.”

After reaching America, she taught Czech at two universities in New York and continued her work on how Czech folk music and culture survived in emigration. From 1944-1945, she ran the Hudební Dotazník (Musical Questionnaire) project, collecting readers’ letters and hand-written songbooks.

While in New York, she wrote: “Sometimes I dream that I’m walking through a crowd of people speaking Czech. When will it become reality?”

Finding Czech Communities in Texas

At the University of Texas, Jakobson encountered thriving Czech-speaking communities across rural Texas. She began systematic documentation:

  • 189 hours of audio recordings (now digitized)
  • 800 reel-to-reel recordings
  • Over 25 boxes of songbooks (earliest from 1917), photographs, manuscripts
  • 20 reels of film — approximately 1,000 minutes of raw footage shot in Moravia, TX in summer 1984
  • Film script and proposal for a documentary “Czechs in Texas”
  • 33 linear feet of archival material

In 1984, Jakobson spent three months filming in Moravia, Texas, documenting Czech fluently spoken in rural communities and taught at the high school level. Tragically, the film’s sound editor (Goran Milutinovic) died in a car accident shortly after filming ended. The sound reels were lost. Unfortunately, the film remained unfinished.

Jakobson received the Texas Czech Heritage Award in 1993. When she died in 2000, she left this extensive archive for future scholars.


Dr. Lida Cope: Building the Texas Czech Legacy Project

In 1997, Dr. Lida Cope, a native Czech speaker and Professor of Applied Linguistics, came to Texas for fieldwork. Roger Kolar, a former student of Dr. Jakobson and prominent activist for Czech educational and cultural preservation in Texas, invited her to review Jakobson’s audiotapes at the Briscoe Center for American History at UT Austin. She met Jakobson briefly in a retirement home in Taylor before her death in 2000.

Dr. Cope recognized what the archive represented: documentation of Texas Czech at Stage 8a/b on Fishman’s Graded Intergenerational Disruption Scale—endangered and approaching extinction. Natural intergenerational language transmission had ended decades ago.

27 Years of Coordinated Preservation Work

As a full-time Professor at East Carolina University, Dr. Cope has maintained this archival work as an External Research Associate with UT Austin. The Texas Czech Legacy Project coordinates:

  • International transcription team at Charles University in Prague
  • Archival standards ensuring research reliability
  • Community relationships with Texas Czech families
  • Researcher training across three states and two countries
  • Film project coordination — working with former Jakobson students and community members to complete the 1984 documentary (film digitized in May 2024)

The Linguistic Significance

Texas Czech blends:

  • Archaic Northeastern Moravian dialects preserving 19th-century forms
  • Standard Czech from the Austro-Hungarian Empire era
  • Systematic English integration developed over 150+ years
  • Regional variations unique to Texas

Examples: Muka/můka (not mouka – ‘flour’), Odkad’ ses’? (not Odkud jsi? – ‘Where are you from?’), Rozumja, ale nemluvja (not rozumí ale nemluví – ‘they understand but don’t speak it’).

Dr. Cope’s 2010 presentation to the American Association of Applied Linguistics argued that Texas Czech should be viewed “as a language variety in its own right”—not dismissed as “broken” Czech. She noted: “Language revitalization ultimately is a community’s choice. But there is enough interest to be harnessed while the vernacular (a resource!) still exists.”


Current State and AI Acceleration Potential

The archive contains approximately 450 hours of untranscribed audio. At current transcription rates with human experts maintaining quality standards, completion could take 15+ years. Most remaining speakers are elderly.

This is where AI becomes relevant—not as replacement for expert transcription, but as acceleration:

AI capabilities:

  • Process hundreds of hours in days instead of years
  • Generate initial transcriptions for expert review

What AI cannot do:

  • Distinguish dialect variations from transcription errors
  • Understand cultural context
  • Make judgment calls about ambiguous phrases
  • Match voices in 1984 film footage to audio recordings
  • Ensure archival quality standards

Expert work remains essential for quality review, contextual annotation, dialect analysis, and archival preparation—including completing Jakobson’s documentary project.


Why This Matters

Jakobson spent decades documenting Texas Czech communities, creating an archive that captures both audio and visual records of 1970s-1980s Texas Czech culture. Dr. Cope has spent 27 years building the infrastructure—international teams, archival standards, community relationships—to make that archive accessible and complete the unfinished work.

My contribution focuses on one specific technical challenge: using AI to accelerate initial transcription while maintaining the quality standards Dr. Cope’s team has established. This continues a family connection—my great-grandfather Jan Štěpán served Czech immigrants from 1889-1948, teaching over 3,000 people English and citizenship while preserving Czech culture.

The technology is a tool. The vision, standards, relationships, and decades of scholarly work belong to Jakobson and Cope.

The speakers are elderly. The timeline is urgent. But the foundation exists to preserve these voices—and complete a 40-year-old film documenting Texas Czech community life—before it’s too late.


Texas Czech Legacy Project: https://txczechproject.laits.utexas.edu

About this series: Documenting the application of AI expertise to accelerate preservation work within existing scholarly frameworks and quality standards.

We are on LinkedIn:

Anne: https://www.linkedin.com/in/anne-stepan-ph-d-502005224

Lida: https://www.linkedin.com/in/lida-cope-0b498548?

TCLP: https://www.linkedin.com/company/texas-czech-legacy-project

Published by AC Stepan WordWorks

Language as a System. Risk as a Design Problem. Senior advisory at the intersection of language, AI, and regulation.

What do you think? I'd love to hear from you!