Dubbed Documentaries
As tools for language learning
The ‘parallel texts’ technique, which self-directed learners have used for centuries, breaks down into three steps. Primus, the autodidact acquires two versions of a given piece of writing, one in an idiom he finds easy to read and another in a vernacular he has yet to master. Secondus, he reads the first text, focusing on the story that it tells. Tertius, the student uses his new-found knowledge of the substance of work to explore the peculiars of the target tongue, and, in particular, to figure out the likely meanings of words and expressions that are new to him.
The addition of alternate audio tracks to video programs makes possible a viva voce variant of this time-honored technique. That is, rather than reading a pair of parallel texts, the learner listens, one after the other, to two versions of the same audio program.
The programs that translate the original words of a video into a second language have yet to master the art of providing each member of an ensemble cast with a distinctive voice. Likewise, they do work better with formal speech than with slang, cant, or jargon. For that reason, learners who wish to use the ‘parallel tracks’ technique should employ programs in which a single, well-spoken person, whether lecturer, narrator, or presenter, speaks from a well-composed script.
Of late, two programs of this sort enabled me to practice several of the languages that interest me. One of these took the form of a single-speaker podcast about a historical event, one in which an actual human being spoke from a libretto that he had created in the old fashioned way.1 The other monologue owed much more than translation to various forms of artificial intelligence. That is, it had been read by an AI-generated avatar from a script written by another creature of its kind.2
The first program employed a single visual, a painting about the central incident of the event in question. Thus, there was no need to worry about any mismatch between the voice I heard and the face I saw. When, however, I listened to the alloglot versions of the second program, I avoided the ‘old kung fu movie effect’ by covering the image of the humanoid on my screen.
The synchronizing alternate audio tracks with each other led to a great deal of fast talking. Thus, when I listened to translations, I often reduced the playback speed of the video.
For Further Reading
Barry Crytzer Lord Dunmore’s War: 1774 (YouTube)
Julia McCoy The Truth Behind GPT-5’s Failure (YouTube)