Translation Quality Control at the FBI
Translations serve an essential function at the FBI. They serve not only to aid agents in understanding and creating a criminal case, but also as important evidence in the subsequent trial (Fishman, 2006). Having an accurate translation, an accurate representation of what was said in the source language, is crucial.
Starting in 2008, the University of Maryland Center for Advanced Study of Language (CASL) conducted a study to determine whether providing a human-generated quick transcript in the same language as the audio would allow FBI linguists to create verbatim translations more quickly and/or with greater accuracy (Wayland, et al., 2009). The study was commissioned to help the FBI decide how to assign transcription and translation work tasks and to aid in staffing decisions.
To maximize the relevance of that study to the FBI, we designed our translation scoring methodology to work with their existing Quality Control (QC) process. This document describes the FBI’s method, and provides a brief overview of other translation quality assessment methods, showing the principal strengths of the FBI’s approach, specifically:
- The emphasis on and identification of errors that impact the intelligence value of the translation, and
- The identification and correction of errors for the purpose of mentoring.
We hope that this document will serve as a useful resource for the FBI and any individual or agency interested in learning about the FBI’s QC process.