NOTE: This story ran in the Economist in 1994 (May 7) in rather a different form. One does not really write for the Economist; it’s more like a process of language translation, in which you submit native English copy that is later translated into Economistese.
Though television is often blamed for a decline in literacy in the industrialized world, a hidden feature of the television medium actually allows viewers to watch TV and read at the same time. Closed-captioning, originally intended for deaf and hard-of-hearing viewers, is increasingly seen as a medium capable of everything from improving the reading skills of children, adult illiterates, and second-language learners to setting videocassette recorders by remote control. But as this formerly marginal technology is coming increasingly to the fore, so too are its political and technological intrigues.
In captioning, the dialogue of a television program (or a film on home video, a commercial, a music video, or any other TV signal) is transcribed, divided into chunks (as subtitles in a foreign film would be), and edited to a comfortable reading rate. (For live programs, a stenographic or real-time captioning system is often used in which shorthand signals from a court reporter’s keyboard are translated into English words instantaneously.) The caption data are added to the television signal in the form of a code residing in the vertical blanking interval. The VBI is the Swiss bank account of television, a hidden reservoir in which data can be banked. An engineer would describe the VBI as a concealed space reserved for the electron beam’s return trip from the bottom of the screen to the top at the beginning of each frame of the moving picture. There are 30 frames per second in North American television and 25 per second in British TV.
The VBI occupies about 20 of the 525 lines of a North American TV picture and a similar number in a 625-line British picture. Since the VBI does not contain picture information that most consumers would ever want to see, television sets are adjusted at the factory so that the VBI is not visible. But if a set is equipped with a vertical-hold control, it is possible to fiddle with the control so that the picture rolls, displaying the VBI as a black band sandwiched between the top and bottom of the usual visible picture.
With caption data encoded in the vertical blanking interval, a home decoder is needed to turn the codes back into visible words. The captions are thus “closed” – i.e., they stay out of sight for viewers who don’t want to see them. By law, caption-decoding chips have been built into most U.S. televisions – and computers like Apple’s Macintosh TV that display full-screen video – since July 1993. With this newfound ubiquity, captioning is increasing its profile among hearing viewers, and PBS is planning to use the VBI to broadcast signals that will automatically set the clocks of Sony VCRs. For now, though, deaf and hard-of-hearing viewers remain the primary audience.
Captioning is superficially similar to subtitling of foreign films in that both involve adding words to a picture, but there are differences. Captions are intended for deaf viewers, are usually broadcast in closed form, move about onscreen to denote the position of the speaker, notate sound effects (“telephone rings,” “knock on door”), and are in the same language as the audio. Subtitles, on the other hand, are meant for hearing viewers, are usually positioned at bottom centre, are “open” (always visible to all viewers), ignore sound effects, and are a translation of the audio. In addition, subtitles do not always translate words the (hearing) viewer is expected to recognize, like “nein” or “merci.” Subtitling and captioning differ enough that a captioner and a subtitler meeting at a cocktail party would find very little in common to talk about.
The VBI contains enough lines to accommodate multiple streams of encoded information. Different countries have different standards for captioning – and for a related technology, teletext, which typically fills the entire screen with “pages” of text but can be scaled down to display only a caption-like block of text at screen bottom. The North American system, developed in the late ’70s and in wide use since 1980, places caption codes on a single line of the VBI, line 21. In Europe and Australia, captions are coded as just another “page” of the general teletext service, which has more of an appetite for data and uses several lines of the VBI.
While teletext is a technically sophisticated medium, with high data speeds and various fonts and colours at one’s disposal, teletext codes cannot be recorded on home video. That is because signals in the VBI have a visual component; teletext signals look like little dots, and the resolution of typical VHS recorders is too low to record those little dots. (There are exceptions: A VCR with a built-in teletext decoder can record captions in open form – that is, permanently burned into the image on the tape – and a Super-VHS VCR can record teletext captions in closed form, but neither technology is in wide use.)
By contrast, captions in the simpler line 21 system, which are broadcast at the slow speed of two characters per frame, can be recorded in open or closed form on any species of videotape. Line 21 caption codes are white lines about a centimetre wide that flit across line 21; at that size, consumer VCRs are happy to record the signals. Thus while a program captioned using line 21 can be recorded in captioned form, a videotape of a teletext-captioned program will not include the caption data. Movies on home video cannot be distributed to rental stores or consumers in teletext-captioned form, but they can be distributed with line 21 closed-captions.
Indeed, closed-captioning of home videos is a booming industry, with some 3,000 titles available in North America. The lack of captioned home videos in the U.K. prompted the grandly-named National Captioning Institute, Inc., the largest captioning body in the U.S., to expand to the United Kingdom in November 1992. NCI uses the line 21 system – in the U.K. television standard, it’s actually line 22 – on a line of home videos. Viewers require the Videocaption Reader, a £100 decoder that attaches to a television, to display the captions. Thus a person who wants to watch broadcast television with teletext captions must own a TV with a teletext decoder; to watch line 22 closed-captions on home videos, a second decoder is required. Neither decoder interferes with the other’s signals. The VideoCaption Reader serves in part as a market for a multi-standard caption-decoding chip that NCI paid ITT $1 million to design.
NCI, a non-profit corporation, receives millions in U.S. government grants annually, charges fees for its captioning services, and even takes the ethically questionable step of asking deaf and hard-of-hearing people to donate to its Caption Club, which tops up funding for captioning various programs. NCI controls about half of the American captioning market and has until this year held a monopoly on captioning of programs for the ABC network. Market dominance of that sort has led to NCI’s being dubbed “the evil empire” by some in the television industry.
Philip Bravin, NCI’s new president and the first deaf person to hold that post, reports that the $500,000 NCI spent on converting the U.S. line 21 system to the U.K. television standard did not come from U.S. government grants or Caption Club money, but NCI will not disclose where it did come from. Nor will NCI discuss ongoing operating costs of the U.K. venture.
Though NCI maintains an office in Peterborough, England, actual captioning of home videos for the U.K. market takes place at NCI’s headquarters in Falls Church, Virginia, and usually consists of “reformatting” the computer files of the captioned American version of the film. In the process, spellings are changed (neighbor and organize become neighbour and organise, and so on) and the all-uppercase captions used in America are changed to upper- and lower-case. More ominous is the fact that captions, which for reasons of reading speed are rarely verbatim in the first place, are edited further to accommodate the slower data-transmission rate of the U.K. system.
Captioning occupies an unusual and troubling position in the production process. Almost always done as the last step in producing a program, captioning is an artistic process that evades the direct supervision of a program’s director and producer. Since by definition most deaf and hard-of-hearing viewers rely on captioning to understand dialogue, captioners have de facto power over how viewers comprehend a program. Directors, screenwriters, and other creative types rarely approve the edited captions. This is not altogether bad; captioning has its own motifs and customs that an outsider might not understand, and in many cases there are good reasons to edit dialogue. NCI refers to employees who do the actual captioning as editors, while the rival Caption Center, a unit of the Boston PBS station WGBH, calls them writers – a telling difference.
Industry lore has it that Woody Allen used to require a printout of NCI’s caption files for approval. Allen grew so displeased with NCI’s editing that captioning of his films was eventually given over to the Caption Center. That kind of oversight is generally unheard of in captioning, and while the entire anecdote may be apocryphal, Hollywood is generally unaware of the fact that captions and audio can differ enough to damage the integrity of a program. Nor does it understand that standards vary across captioning companies; the highest standards are arguably found at the Caption Center, with NCI’s standards somewhere in the middle and smaller captioners like Captions, Inc. of Los Angeles and the various Canadian companies at the very bottom.
NCI has contracts to caption home videos released by the British branch plants of major American film studios – and by the BBC. The irony of an American company using American techology to caption BBC videos in the United States for the British market has not escaped the notice of U.K. firms doing teletext captioning. “Which nation does the National Captioning Institute really serve?” is perhaps an unfair question. But in a real sense both NCI and its British rivals are locked into their respective positions by incompatible technologies. Perhaps only a world-standard HDTV, if it ever arrives, will have the power to break down the iron curtain between the captioning fiefdoms.
Originally published 1994 ¶ Updated 2001.07.15