Humanizing nameless speakers

Kyle talking with Roger in a screenshot from an Allstate commercial

The Speaker ID is a powerful tool in the captioner’s arsenal. Care must be taken with it. When used carelessly or applied rigidly, it can do some damage to the integrity of the narrative.

A while back I shared a clip from the TV show Dollhouse that featured a Speaker ID that gave away the identity of the speaker lurking in the shadows:

When the star of the show (Eliza Dushku) asks “Who’s there?,” the caption viewer already knows the answer to her question. The Speaker ID told us it is Nicolas. That dramatic, suspenseful moment is compromised by a caption that reveals the shadowy figure’s identity before the narrative is ready to reveal it.

This example features what I call captioned irony. The caption viewer knows the identity of the shadowy man before the non-caption viewer. Just as dramatic irony defines those moments when the audience knows more than the characters, captioned irony defines those moments when the captioned viewing experience gives the audience advance or more knowledge than the audience watching without the additional benefit of captions (and sometimes more knowledge than the characters as well).

The benefit of more or prescient knowledge may come with a price, as we’ve just seen. Captioners are supposed to identify speakers who are off-screen or can’t be clearly identified by viewers. But just because a speaker is off-screen and can’t be identified doesn’t mean the Speaker ID rule should be invoked. In this Dollhouse scene, the need to nurture suspense trumps the need to identify unknown speakers. The mystery of his identity is the point of the scene; captions should support that mystery until the narrative is ready to lift the veil and bring Nicolas into the light.

Humanizing Man 1 and Man 2

The Speaker ID can serve other purposes beyond identifying off-screen speakers or giving away plot points too soon. Recently, I became intrigued by a TV commercial for Allstate Insurance in which the captioner (presumably drawing on info from the producers) ascribes names to the nameless characters, instead of opting for the typical “Man 1” and “Man 2” approach to identifying nameless speakers:

Notwithstanding the demands that this commercial places on caption viewers who must read quickly while trying to process the meaning of “with voice of Dennis,” the Speaker IDs for “Kyle” and “Roger” humanize what are otherwise nameless characters. Only Dennis Haysbert is a known entity. The other two guys have no names in the uncaptioned version. Only caption viewers know their names. That’s captioned irony. Speaker IDs can humanize nameless characters and increase the audience’s sense of identification with them. Kyle and Roger are more real because they have names. Using names is a more effective strategy in certain situations than “Man 1” and “Man 2.”

It’s a shame the captioner didn’t follow through on this approach by identifying “Dennis” when he appears at the end of the commercial. The “voice of Dennis” captions only make sense if we know who Dennis is.

In another captioned version of this same commercial, the function of the “Dennis” captions become even less clear as “with voice of Dennis” is replaced with just “Dennis”:

[Fair use notice: The videos on this site are transformative works used in good faith, in keeping with Section 107 of U.S. copyright law, and as such constitute fair use of copyrighted material. Read this site’s full fair use notice.]

S. Zdenek

Dr. Sean Zdenek is an associate professor of technical and professional writing at the University of Delaware. He is the author of Reading Sounds: Closed-Captioned Media and Popular Culture (University of Chicago Press, 2015).


2 Responses

  1. I’m not familiar with the scene in the first clip, but one thing that a hearing listener has that a deaf/hh person doesn’t is the ability to recognize and identify a person by voice. If the character has appeared before, the hearing listener might recognize his voice. Sometimes that’s an issue that a speaker id compensates for. (as in “voice of Dennis”).

    • That’s an excellent point, Bill. That’s why it’s important to attend to context when dealing with Speaker IDs. If the voice gives away the identity, then a Speaker ID (with a specific name) may be warranted.

      I prefer more flexible guidelines over rigid rules.