What is the best approach for closed captioning a video that changes narrative languages in a sporadic and interspersed manner? Keep reading to find out.
I was recently asked a closed captioning question that initially stumped me a bit. It was something I personally hadn’t encountered, so an immediate answer didn’t come to mind. Confession: These somewhere in between / grey zone / "it depends" questions and answers, are my favourite. It’s the problem solving part of accessible design one comes to love.
Anyways—being intrigued by the question, I started digging around in search for some answers and found there wasn’t much out there?
I had some answers and suppositions, but I wanted to reinforce them before responding to my colleague. The information out there was either scattered, ambiguous, or missing altogether. Hence why I decided to write this. Something that answers the question and provides solutions in a more direct, layman, and practical sort of way.
The question asked was >
"What is the best approach for closed captioning a video that changes narrative languages in a sporadic and interspersed manner?"
﹡﹡
As in—closed captioning for TWO languages, on *one* closed caption track. A video with narration in one language that at certain points momentarily CHANGES to another language and back again. In my specific case, the origin language was Inuktitut and the subbed in words English (example below, subbed language highlighted).
[...] sananiaqtuksauvut. Suurlu isumagittiarlugu fashion industry there is no name for that one yet in Inuktitut. So, the fashion industry taassuminga qimirrualaurluni [...]
Why does the language change you ask? As expressed from our client and alluded to in the example, some words are simply not available (missing) from the origin language vocabulary. Put another way, there aren’t words to describe what the speaker is trying to communicate *in that language*. The speaker therefore switches to another language to "fill in the blanks".
To summarize, example 1 would be:
1) Language vocabulary is missing words and speaker speaks a second language.
A similar instance where you might see this sort of thing is when the speaker is less affluent in the spoken language, and as a result—switches between two languages. You see this when a speaker is fluent in one language but less in another. If the speaker can’t conjure up certain words, or articulate themselves cohesively, the will switch (especially if they know the person spoken to also speaks the second language).
To summarize, example 2 would be:
2) Speaker is bilingual and variably changes from one language to another.
Solutions
Ok—enough with the theory already! Let’s look at some concrete examples.
Starting with example one illustrated above, there are a couple of things we can do:
1) We can say exactly what’s going on via descriptive information syntax. I don’t speak Inuktitut, so for the sake of this example let’s *pretend* English was the primary language with French mixed in as the second.
00:10–00:16 [speaker speaking in French, words unknown and missing from French vocabulary]
2) Another option, is to include the untranslated alternate language words, prefixed by a descriptive word that identifies the alternate text language.
00:10—00:13 > [French] ça m'a fait mal, c'était incroyablement 00:13—00:16 > frustrant, je me sentais trahi.
Notice how the language demarcation is *only* included at the time marker where the alternate language STARTS. Anything more is redundant and unnecessary.
Most of the time, people "play" videos through (e.g. start to finish or from 00:00 to somewhere). The only time the solution above is flawed is if someone STARTED the video at 00:13. It’s an imperfect solution, yes—and that’s OK. We should aim to intentionally design for the norm, and not the exception.
The last thing you can consider doing (for option 2), is re-identifying the origin language to signal the end of the alternate one.
00:10—00:13 > [French] ça m'a fait mal, c'était incroyablement
00:13—00:16 > frustrant, je me sentais trahi.
00:16—00:19 > [English] Anyways—it doesn’t matter that
00:19—00:21 > all happened a long time ago.
I say "consider" here to illustrate an important point. Like a lot of accessibility and text comprehension things, there isn’t always a hard rule or standard for something. You’ll at times, have to use good judgement and take into consideration your specific context and audience.
For instance, *having* or *not having* another captioned version of the video in that alternate language, can influence the decision above.
For instance, if the origin content language matches that of your core audience, then perhaps you don’t need to explicitly identify when the text changes back to that language?
That said, like all accessibility questions and considerations, it’s best *not* to make assumptions. Assuming that the core audience speaks English *fluently* would be wrong. There are ample Canadians, permanent and temporary residents, that understand English yes but not as their first language. If that’s the case, then perhaps the explicit language demarcation option is the best solution—it’s more inclusive and less presumptious.
﹡﹡
Change your shoes, man
Always put yourself in the shoes of someone who might interact with what it is you’re designing. First ask *yourself*, would you understand this? If you answer yes, don’t stop there. Ask for input and test your theories and assumptions with real people (colleagues, strangers, or better yet—the wearers of the shoes). Do they find it comprehensible? Get them to weigh in on the problem and then recalibrate your solution.