How captions work, and why some videos have none.

A caption track is text with times.

A caption track is a list of short lines, each with a start time and a duration. The player shows each line while its time runs. Read in order, the lines are the transcript.

A video can carry several tracks, one per language, plus translations the platform makes on request. The transcript button under a video opens the chosen track as a list.

Human, auto-generated, translated.

• Human captions

  Written or corrected by the creator or a captioner. Punctuated, with names and numbers right. The best transcript a video can have.

• Auto-generated captions

  Made by speech recognition after upload. Lower case, no punctuation, and wrong on names, numbers, and accents. Still a workable transcript for judging what a video covers.

• Translated captions

  Made on request from one of the other two. Good enough to follow, rough enough that a verdict is judged from the original track and reported in English.

Why a video can have no captions.

The creator turned them off. Some channels disable captions on purpose.

The audio is music with no speech. Or it is speech the recognizer could not place: a language it does not support, heavy noise, or several people talking at once.

The video is new. Auto-generated captions take minutes to hours to appear after upload, and longer on long videos.

The video is a live stream that has not ended, or a premiere that has not played yet.

What this means for a verdict.

No captions means no verdict. The panel states that rather than guessing from the title.

Auto-generated captions get a verdict with a note. The points paraphrase rather than quote, because the words may be wrong.

Song lyrics get no verdict, because there is nothing to summarize and no reason to reproduce them.

A video in another language gets a verdict in English, with a line that names the caption language.

Reading a transcript yourself.

Open "Show transcript" under the description. The times at the left are clickable. Search the page for the key word in the title to see whether, and where, the video delivers it.