Response to CBC letters (Joe Clark, 2006.05.04)

Since I’ve learned the hard way that one must be careful what terms one uses with the Canadian Human Rights Commission, the Commission and CBC may consider the following my response to CBC’s letters.

In summary, CBC concedes or cannot disprove all the contentions in the information I originally submitted to the Commission. It repeatedly relies on statements that amount to “This is the way we do it” or “A certain document permits us to do things this way.” In response, I provide substantiated reasons backed up with research citations. I say again what I said at the outset: If it comes down to siding with employees and managers new to captioning (who have, moreover, an interest in denying wrongdoing) or someone who has followed and worked in the field for over 25 years, the choice is a no-brainer.

Personal history with CBC

It is now relevant to discuss my history with CBC captioning.

Let’s start with why I care. It simply isn’t the case that CBC’s captioning is the worst in Canada. It isn’t even in the bottom quartile. Really, as Canadian broadcasters go, I’d like to say CBC is doing a reasonable job – and I’d be in a position to actually say that if CBC Television and Newsworld were not subject to higher legal requirements than other broadcasters, higher even than the Corporation’s other TV networks. But it is subject to those requirements, which raise the bar considerably. There is next to no wiggle room whatsoever for absent captioning and barely any more wiggle room for substandard captioning.

I support the CBC and public broadcasting, with the same reservations anyone with a love–hate relationship might have. It’s easy to find broadcasters with significantly less captioning than CBC, and it’s ridiculously easy to find broadcasters with worse captioning. Most Canadian broadcasters have worse captioning. I pick away at the edges of that problem, but what I decided to pay attention to was CBC because I give a damn.

By monitoring CBC captioning errors and omissions, I put in three years of occasional consulting work, free of charge, for CBC and the Commission. You both got a freebie. I wouldn’t even consider doing that for a private broadcaster, who would have to pay full freight all the way. I did not file my documentation with the Commission to get CBC, or to cause them grief or make things worse. I did it because I care about captioning (and audio description, and other forms of accessibility) and because CBC was not living up to its requirements. I get annoyed by CBC’s captioning mistakes, but annoyance is not enough to drive a three-year information-gathering process. If any broadcaster in Canada is to have high-quality, wall-to-wall captioning, it should be a public broadcaster.

That is how I feel and that is why I carried out my project. CBC managers are so feudalistic and are so motivated by pique, resentment, and outright contempt that they locked out their own employees. So of course CBC isn’t going to believe the foregoing. That’s one of the problems: As you’ll see below, I’ve been trying to assist CBC with accessibility issues (with seriously mixed success) for four long years. If you aren’t willing to place credence in my statements, place credence in my actions.

In-house captioners

I met the two captioners who were at work in the CBC Toronto Broadcasting Centre before the Vlug decision. One obviously hated her job, was a lousy captioner, and acted defensive and embittered; she has since left, thankfully. The other captioner welcomed me into her office, where I gave her some advice on a program she was working on (a historical documentary on the Royal Family). She seemed very interested in captioning and, moreover, interested in the advice I was giving her.

The captioner and I kept in touch via E-mail – convivially, as you can see from this message from 2002 (excerpted, but sense is unaltered):

In case you forgot, I was, apparently, the only one thrilled with your unannounced visit to the CBC. I also won’t easily forget you.... I attempted to wade through your Web site (dictionary in hand). I didn’t find what I was looking for. We will be expanding our department this fall[;] it would be nice to have a comprehensive style guide so that everyone would be on the same page. Whatever happened to the one you were doing? Any idea where I can get one that comes with your approval?

Also, I have a number of questions on how to handle a particular captioning concern. Any chance I can hit on you for the answers?

Here’s hoping I can call on you anytime...

Indeed she could, as far as I was concerned. I’ve been giving out free advice for many years alongside the paid kind and I had no objection at all to helping her.

We later had a meeting, on 12 June 2002. Before that, the captioner wrote in an E-mail, “Again, just so we’re clear. This visit is off the books. We will meet in the atrium on a mutual-interest basis.” And we did meet in the Barbara Frum Atrium, the most conspicuous and public place in the building. I am aware that CBC may not be willing to accept the following, but my captioner friend proceeded to pick a fight with me, quickly and efficiently getting me riled up. (I didn’t start the argument. Again, CBC may react with disbelief, but that’s what happened.) Frankly, I didn’t know what hit me.

I followed up with a phone call on 2002.10.21. (I took notes.) The captioner told me they were looking for “upgrades” and “fixes.” She would not say anything about improved training for standardization, except that I wouldn’t be the one she’d choose “not because of what you know but how you deliver it.” Oddly, I seemed to have been viewed as pleasant and credible enough while seated in her office. Her revisionism is, at least, consistent with the grudge-driven environment of CBC management.

Paid work

As I previously explained, I have done paid work for CBC on accessibility, usually at far below my book rate. My projects were as follows:

I seem to be qualified and competent enough for other CBC departments to actually hire. I assume my advice to CBC on captioning is just as credible even though I give it for free.

Correspondence with executive

In 2002, I started corresponding with Ian Alexander, Chief of Staff, CBC Television. I introduced myself and asked if he could pencil me in for a meeting “so that I might drop by and say hello, just to get to know each other? We might be able to figure out a way to work together, or at least we can add each other to our contact lists and keep in touch from there.”

We talked on the phone once or twice in 2002 and I continued to send occasional E-mails. I never got the meeting I was promised.

After I filed my information with the Commission, I wrote:

I am requesting the meeting you promised in 2002 to discuss CBC captioning and, if we feel like it, audio description. This would be an inopportune time to refuse to meet me merely because I published evidence about CBC captioning. The Corporation hired me three times before to work on accessibility and I assume I am able to meet you and your staff without prejudice.

And the response was (excerpted);

Since your complaint is now the subject of correspondence between legal counsel for the Canadian Human Rights Commission and legal counsel for the CBC, I am advised that it would, in fact, be inappropriate for me to meet with you while this matter is under review.

It’s unfortunate that you didn’t reach out to us directly before filing your complaint with the Commission.

However, I would like to arrange such a meeting with you and our closed-captioning people as soon as possible after that proceeding is concluded.

Needless to say, we disagree strenuously with much of your assessment and characterization of our closed-captioning performance. Nevertheless, we would sincerely like to identify and act on any areas for potential improvement.

I wrote back explaining that I had been trying to “reach out” to them since 2002, but he ignored me. (Actually, feel free to ask Ian Alexander what my exact words were.) Note that here the chief of staff yet again promised me a meeting and yet again conveyed aggression in defending CBC’s position (“disagree strenuously” – Cf. “vehemence” below).

Ian Alexander and the captioner effectively isolated themselves from my further information and advice even while other departments of the CBC were hiring me to assist with accessibility issues.

Subsequent captioning errors and omissions

Since filing my documents with the Commission, I have kept up my record-keeping of CBC captioning errors and omissions. This is exactly the sort of thing the Commission could have been doing for 3½ years with scarcely any effort.

Data from 2005.10.13 to 2006.04.07 are as follows:

2005.10.13 12:53
No captions on Washington segment on Newsworld or the entire following business segment. First commercial does, however, have captions.
2005.10.20 17:21
Two promos on Newsworld uncaptioned.
2005.10.24 16:51
Captioning appears and disappears on breaking-news item. Susan Bonner also cannot always hear Christopher Thomas at her studio in Ottawa. The item ends with the following:
  • “Thanks, Christopher. Sorry about the audio problems.”
  • “Hey, it’s live.”
  • “It happens.”
And of course that entire interchange had no captions.
2005.10.24 18:46
Toronto news does not render the URL in the line “You can get more information by visiting their Web site,” even though the URL is simple. It’s left uncaptioned.
2005.10.26 23:33
Strawberry and Chocolate uncaptioned (and “hardsubbed,” i.e., with burned-in film subtitles).
2005.11.04 18:41
Making the Grade promo inside Toronto edition of Canada Now not captioned.
2005.11.09 19:25
Cinema Real Docfest/Whistler Contest promo uncaptioned on Newsworld.
2005.11.15 12:20
Promo for Politics uncaptioned (and seen while I was sending out the press release on uncaptioned programs on CBC).
2005.11.21 14:00
Promo on Newsworld for Secret Mulroney Tapes (previously seen captioned at least twice) uncaptioned.
2005.12.01 00:34
Our Lady of the Assassins has (crappy Arial) subtitles but no captions, including in extended segment in which a jukebox plays an unsubtitled song. Later: “Kio! Turn the music down a bit!” What music? I didn’t read any music.
2005.12.07 18:26
Bumper for The National uncaptioned. There was an unusual pause before the bumper ran; no doubt the stenocaptioners were caught by surprise.
2005.12.09 18:26
Promo for “Taxi Chat” segment on The National uncaptioned until the last seconds, where the only caption was Announcer: and a few garbled characters.
2005.12.13 2212
Discover Wild Canada promo uncaptioned.
2005.12.14 2325
Intro to In the Mood for Love uncaptioned (as are all intros and extros to late-night movies).
2006.01.10 17:50
Bumper uncaptioned on Canada Now on CBNT. Earlier, a streeter that aired before Hanomansing came back on was uncaptioned, no doubt because it took the captioner by surprise.
2006.01.12 00:28
Bumper to International Movie Night uncaptioned. Neither was the movie, Blue.
2006.01.19 11:59
Paul Martin stump speech interpreted from French on Newsworld uncaptioned.
2006.01.20 17:55
Promo for Canada Votes on Newsworld uncaptioned.
2006.01.28 18:32
Entire opening for Hockey Night in Canada uncaptioned, including the first two interviews.
2006.01.31 00:21
La vie, la vie has captioned bumper explaining it’s the last episode, which goes on to run the entire Trudeau “Just watch me” speech uncaptioned, among a host of other captionable effects.
2006.02.02 17:20
Promo for The Hour on Newsworld uncaptioned.
2006.02.02 17:42
Two consecutive promos, including one for The Lens, uncaptioned.
2006.02.02 17:55
Passionate Eye promo uncaptioned. That makes three commercial breaks in a row with no captioning (also not on any commercials I saw, though I didn’t see all of them). But, three minutes later, a Fifth Estate and a generic Newsworld promo were both captioned within this same break.
2006.02.09 18:53
Identical promo for Neil Young interview runs twice within three minutes, once with scrollup captions, once with none.
2006.02.13 11:26
Promo for Olympic news updates uncaptioned on Newsworld.
2006.02.13 11:29
Promo for Ross Rebagliati special uncaptioned. Presumably encoder pass-through in these two cases.
2006.02.13 11:29
Opening by newscaster captioned; intro to Olympic recap from previous years uncaptioned; actual recap captioned. Took the stenotypist by surprise, apparently.
2006.02.18 2100
No captions on Don Cherry.
2006.02.22 ≈11:30
Olympic men’s hockey game between USA and Finland. The stenocaptioner apparently has no player names, not even Chelios, in the his or her dictionary (exception: Hatcher), misspells Finns as “Fins” consistently, and never captions play-by-play, only colour. Caption left margin is also more than halfway across the screen even though the score bug, which would need to be avoided, is at most two tab stops wide.
2006.02.26 15:09
CBC promo for video-on-demand service uncaptioned.
2006.02.26 15:10
Nonsensical and contradictory captioning “sponsorship” from XM Satellite Radio airs with captions.
2006.03.11 18:18
Tommy Douglas promo on CBC uncaptioned.
2006.03.14 17:41
Two consecutive Newsworld promos uncaptioned.
2006.03.14 17:56
Another Newsworld promo uncaptioned.
2006.03.21 16:45
Generic Newsworld promo uncaptioned.
2006.03.31 11:58
Address by Vicente Fox uncaptioned. However, the English interpreter was low in the mix and had a strange voice, being nearly incomprehensible. Captions come on at 2006.03.31 12:01.
2006.04.07 01:10
Subtitled movie Québec–Montréal has scrollup captions for music and NSI, and the captions do not clear from the screen; they just scroll up. [rattling cap of tank] as we see a man obviously twisting off the cap. ALLO? and >> HEIN? sit there superimposed over subtitles.

Response to first letter from CBC (2005.12.08)

  1. I caution the CBC that it is unwise to endlessly repeat the internal catchphrase “CBC/Radio-Canada” whenever discussing the Canadian Broadcasting Corporation. Henry Vlug’s initial complaint, the ultimate settlement, and my submission all concerned CBC Television and Newsworld. Other CBC networks, including Country Canada and the Documentary Channel, were excluded, as was the entirety of CBC’s French-language service.

    Radio-Canada’s and RDI’s captioning are already subject to another investigation based on a complaint by Jean-Robert Gauthier; the Commission’s own report on CBC French captioning is seriously flawed, as I have documented elsewhere (Clark 2006). If the Canadian Broadcasting Corporation insists on using the byword “CBC/Radio-Canada” whenever referring to itself, then I will have to submit my debunking of the Commission’s report on CBC French captioning as part of this current proceeding. Be careful what you invoke via careless repetition of corporate buzzwords.

  2. The Corporation hereby reconfirms unequivocally its commitment to 100% captioning at the highest level of performance reasonably possible.

    That certainly sounds acceptable at first blush, but since the legal threshold is accommodation short of “undue hardship,” CBC’s declaration that the standard is “the highest level of performance reasonably possible” has no legal merit. It is, moreover, not up to CBC to decide what’s reasonable, as it is human nature and is in corporate self-interest to deem as “reasonable” whatever the corporation is already doing or plans to do.

    The Canadian Human Rights Commission itself dismissed a quantitative argument of this sort in its report No Answer. In that report, the topic was TTY usage; in the current case, it is captioning complaints.

    During this study, some federal organizations indicated that demand for TTY service was low. This is an interesting observation, but it is not a justification for not providing equitable service access. Advocacy groups indicated that TTY users have become so frustrated with trying to make TTY calls that they have simply given up. Instead, they ask the advocacy groups, or a hearing friend or family member, to make calls for them.

    Even if true demand [were] low, and there is no proof of this, that would not be justification for providing inadequate services. The law and jurisprudence on this issue are clear: The number of people requiring accommodation does not determine whether accommodation should be provided. Only undue hardship is recognized as a justification for not accommodating a legitimate need.

    Additionally, the settlement in Vlug v. CBC required 100% captioning short of “glitches” which must be “the exceptions. The rule should be full captioning.” As such, the original settlement already sets the threshold for undue hardship (full captioning is not undue hardship) and sets an exemption solely limited to exceptional glitches. The CBC again has no latitude to lay out the term “reasonably possible” as though it had legal bearing on its responsibilities to accommodate the needs of viewers with hearing disabilities.

  3. In 2002, CBC/Radio-Canada made a capital investment of $627,000 in support of the commitment to 100% closed captioning. The project involved the purchase of closed-captioning equipment for Toronto and regional CBC/Radio-Canada stations. The associated incremental operating costs that arising from this are approximately $1.1 [million] annually.

    Again, this seems impressive on initial reading. However, the legal threshold remains accommodation short of “undue hardship.” With 2004–2005 revenues of $564 million and government allocations of $936 million ($1,500 million total; CBC 2005), an expenditure of $1.1 million or 0.07% per year on legally-mandated services hardly meets any definition of undue hardship.

    It is acknowledged that CBC may have spent $627,000 on “captioning equipment,” but it was under no obligation to license the most expensive captioning software on the planet (Swift by Softel, at up to US$12,000 per seat) and associated hardware. CBC presents cost as a defence in this context; if cost were really an issue, CBC would not have bought Tiffany software on its implied Wal-Mart budget, a legally irrelevant position.

  4. Most recently, CBC/Radio-Canada has added an alarm system that alerts operators to technical failures in captioning, and we are reviewing our processes on a daily basis.

    In fact, alert hardware was removed from Country Canada and installed on CBC Television after I filed my submission. There is no technical, operational, or financial reason why alerts could not have been in place on every one of CBC’s (and, yes, Radio-Canada’s) television services all along.

  5. CBC/Radio-Canada established an in-house captioning group, staffed by fully qualified and trained captioners to assure the provision of captions on promos and other telecast material for which, for a variety of reasons, captioning by external providers is impractical or inefficient.

    CBC had in-house captioners in Toronto and Montreal before Vlug’s complaint was settled and before I made my submission. Full qualification and training are significantly in doubt given that there is no such thing as training in captioning, save for court reporting and stenography. Qualification and training were in enough doubt for one of CBC Toronto’s original captioners to ask me for help.

  6. External captioning service providers, which supply the majority of CBC/Radio-Canada’s captioning, are carefully monitored for quality of service. They provide weekly reports detailing issues associated with faults and issues are addressed immediately.

    It is in dispute how “careful” the “monitor[ing] for quality of service” might be when quality of service was an issue in the original Vlug case and is an issue in the current discussion. If CBC had been doing all that monitoring, I wouldn’t have had any documentation to submit.

    Since CBC has low or improper standards for quality of service, then it is hardly impressive if outside suppliers meet them, nor does it necessarily assure compliance with the Vlug ruling or actual provision of accessibility for deaf and hard-of-hearing viewers. CBC, as fox, should not be guarding the henhouse of its outside suppliers, who have a financial imperative to do whatever the CBC wants.

    I dispute the claim that, in the CBC’s unparsable phrase, “reports detailing issues associated with faults and issues are addressed immediately.” The same faults kept happening over and over again in the three years I took notes.

  7. Mr. Clark alleges 130 instances of absent captions. Without agreeing with this assessment

    CBC does not dispute the assessment. Its own record-keeping is inadequate to disprove the assessment or such records would already have been tendered. Its “monitor[ing]” cannot be all that “careful.”

  8. CBC Newsworld broadcasts 24 hours a day and CBC/Radio-Canada’s [sic] main television service is on air an average of 135 hours per week. In total, the two networks telecast approximately 47,000 hours over three years, and that does not include the impact of multiple time zone releases for the main network. Mr. Clark states that the captioning absences are “typically remedied within a few minutes.” If we estimate each of his documented absences as 5 minutes in duration, then at 130 faults, the total time equals 10.8 hours or less than... [0].023% of the total telecast time in question. Surely this represents substantially complete compliance.

    No, it doesn’t, even by published industry standards or by comparison with services provided to nondisabled viewers.

    1. CBC has not demonstrated that hearing viewers must endure dropouts, audio failures, scratchy audio, or other disturbances in soundtrack at least 0.023% of the time. I know from watching both networks extensively that dead audio is exceedingly rare. (The only notable occurrences are dead earpieces worn by guests in remote interviews, and even then the audio is not absent per se.) Nor has CBC demonstrated the analogous case that video drops out or is unfit for broadcast at least 0.023% of the time. Such absences are slightly more common in my experience, but are still so rare as to be unmeasurable by the average viewer.

      In short, CBC implies that deaf and hard-of-hearing viewers must accept a service that is unavailable at least 0.023% of the time in an environment where the standard is 100% save for glitches. Nondisabled viewers are not expected to put up with that level of service interruption. The inescapable conclusion is that CBC’s estimated captioning downtime constitutes adverse differentiation on the basis of disability, contrary to §5(b) of the Canadian Human Rights Act.

    2. A true and viable standard of performance is five nines, or 99.999%.
      1. If we view captioning as a utility on which disabled viewers rely, we can draw comparisons with uptime standards used by other utilities, such as telephone companies, where five-nines reliability, or 99.999% uptime, is a common standard. CBC’s own estimates show it would fail such a standard.
      2. We can draw comparisons with other broadcasters. The Federal Communications Commission (FCC 2005) engaged in a consultation concerning increased captioning requirements in the U.S. (the FCC proceeding). In its filing, HBO (2005) stated (emphasis added):

      HBO’s quality-control program monitors 16 categories of potential technical issues associated with video, audio and closed captioning, the occurrence of any one of which is considered a disruption to the service. HBO’s goal is to have each of the 30 linear programming feeds it originates experience no more than 5.5 minutes of programming disruption per year – a reliability factor of more than 99.999%. Over the past two and a half years, more than 25 of HBO’s programming feeds have met this reliability goal consistently. Those linear feeds that fell short of the goal missed it by an insignificant amount on an annual basis....

      HBO has found that closed-captioning errors on its feeds account for less than 10% of all disruption events (i.e., less than 30 seconds per year). In fact, in HBO’s experience, the errors in closed captioning are fewer than the minuscule amount of audio discrepancies. [...]

      The basic technical infrastructure to keep a set of pay channels up and running is almost exactly the same as that used to keep a public broadcaster up and running. In fact, CBC and HBO use a lot of the same equipment, including Swift. HBO’s assertion of a standard of five-nines reliability is thus a plausible proxy for an actual industry standard. A five-nines standard would permit CBC Television to present uncaptioned programming for 4 minutes 13 seconds a year on average. (135 average hours per week × 52 weeks a year = 7,020 hours/year × 0.99999 uptime = 7,019.9298 hours’ uptime or 0.0702 hours’ downtime = 4′13″.)
  9. CBC/Radio-Canada shares Mr. Clark’s desire to deliver the highest possible quality of closed captioning to Canadians; however, we believe that quality captioning, as Mr. Clark defines it, is impossible to achieve within the real-world broadcast environment and with finite resources.

    Essentially, CBC contends that quality standards achievable by other broadcasters and captioners are impossible in-house. I’ll address these issues in detail below as they are brought up. And we have already established that the finiteness of resources is not at issue, since the CBC has not established and cannot establish that improving its operations will amount to undue hardship.

  10. There appears to be a significant measure of satisfaction among the users of closed captioning. The CBC/Radio-Canada Audience Relations department has received only 26 complaints in relation to closed captioning between November 2002 and November 2005.

    An absence or a paucity of complaints does not indicate anything other than a paucity or absence of complaints.

    1. It is unreasonably difficult for viewers to lodge a complaint. At the highest end of the spectrum, viewers must demonstrably avail themselves of the Canadian Human Rights Commission and Tribunal in order to achieve accommodation of their legally guaranteed needs. (At that stage, the Commission will abet the respondent in using pretexts to attempt to derail any such proceeding.)

      At a more pragmatic level, what exactly is a deaf or hard-of-hearing person supposed to do to complain about captioning?

      1. Get up off the couch at 10:30 at night
      2. Jot down details of the program and the error (if they have a vocabulary for such issues, and thereby missing more of the program, which they must watch, not just listen to)
      3. Wait till the next day
      4. Call the difficult-to-find CBC Audience Relations TTY line during office hours (even while one may have to be busy at the workplace)


      1. Get up off the couch (again)
      2. Head over to the computer
      3. Find the CBC Web site
      4. Submit an online complaint

      In either case, the viewer must stop watching TV entirely just to file a complaint. The viewer must cease to use the service just to complain about the service.

    2. The FCC proceeding received evidence from a deaf organization and captioners that complaint procedures are onerous and unrealistic for captioning viewers. There is further evidence that complaints are insufficient to catch captioning errors.
      Remarks from Global Translation/TranslateTV (2005)

      [A]t the level of local and regional programming of live content, market forces are insufficient to assure quality closed captioning. TranslateTV’s logs of local stations’ newscasts indicate that, on average, between 25% and 60% of all captioned sentences contain errors that substantially impede understandability. TranslateTV classifies these errors by type and severity. This high error rate is true at stations using both live captioning and [electronic newsroom captioning].

      The stations and caption providers do not systematically monitor caption quality. Instead, their chief feedback mechanism is consumer complaints. When a complaint is received, most stations and providers will respond individually to the viewer; however, there is no proactive monitoring system in place to quantify and analyze the overall rate of errors in captioning. The current feedback system is both reactive and narrow in its impact.

      Remarks from WGBH (2005)

      [T]he feedback loop between caption consumers and program providers and producers is very weak. Communication between caption consumers and program distributors requires clear points of contact, widely-published voice and TTY numbers and knowledge of relay services, staffing into the prime-time and nighttime hours, and knowledge and understanding of caption quality issues by program providers, local cable operators and local TV stations. Without those means of communication and points of contact, it’s no wonder that complaints rarely reach the providers. Without such a feedback loop, many providers may assume that a low-cost and low-quality captioning service is adequate, not hearing otherwise from consumers.

      Complaints from consumers more frequently arrive on the doorstep of the captioning agency, which may or may not be responsible for faulty caption services. That caption agency is reliant on the good will of their clients and has an inherent conflict in bringing complaints to the large corporations which are funding them.

      Remarks from Telecommunications for the Deaf, Inc. (TDI 2005)

      In order to bring a complaint, a consumer needs to (1) know to whom a complaint should be directed, and (2) have the means of transmitting the complaint to that person.... Consumers often have difficulty determining where they need to file their complaints. Because of the complexities of television programming distribution, the average consumer often does not know who is responsible for compliance with the captioning obligations – most consumers are at a loss as to whether a complaint needs to go to the to the local station, a national network provider, a cable network or the local cable franchise. While it may be advantageous for consumers who are savvy enough to know how to bring their complaints to the appropriate entity in the video industry to do so before going to the FCC, all consumers should have the option of bringing their complaints to the FCC, wherein the complaint can be redirected to the appropriate distributor for response.

      Complaints are not a popularity contest. This isn’t a vote for The Greatest Canadian. Quantity of complaints does not indicate severity of issue. And in any event, my submission complained about noncompliance. Other complaints are beside the point in the current discussion.

  11. CBC/Radio-Canada’s manager responsible for closed captioning takes immediate action to source the problem, if she is not already aware of it, and frequently responds to the complainants personally, in the interest of assuring our viewers that closed captioning is a CBC/Radio-Canada priority.

    If this is a reference to Brigitte Ouellet, I can cast doubt on the assertions. As part of a paid contract with CBC Zed to improve accessibility of its Web site, expand captioning online, and improve its TV captioning, I was, after several requests, eventually granted an audience with Ms Ouellet on 2004.07.09. I explained that captioning was inadequate in many respects on Zed – a complex prerecorded series with interviews, live music performances, and art videos of many types that was, inexplicably, captioned real-time. (The only point to which she responded substantively was song lyrics. If captioners aren’t being provided with song lyrics, she told me, she’d have to look into that.)

    I told her the program was so complicated that, if real-time captioning really had to be used (which it didn’t), then only the first airing, as in Atlantic Canada, should use real-time captioning, with those captions saved, corrected, and live-displayed or popped on during subsequent airings. This would require saving the first broadcast without captions, airing and recording the cleaned-up second broadcast, and then simply running that recording for subsequent broadcasts. Zed has a 35-minute runtime, meaning that CBC would have to store 35 minutes of video for an hour at a time four nights a week. Ms Ouellet rejected that proposition outright because “disc space” was “tight.” (Even if true, does maxing out a hard drive for just over half an hour a few times a week constitute undue hardship?)

  12. A number of the specific issues that Mr. Clark identifies are judgement calls, about which there are various views within the captioning community itself.

    There isn’t any such thing as a “captioning community,” and disagreement on specifics does not imply that each party doing the disagreeing is equally well informed. I back up what I say.

  13. With regard to the recent CBC/Radio-Canada labour disruption, the operational challenges we faced were not limited to closed captioning.... CBC/Radio-Canada managers made Herculean efforts to maintain our commitment to 100% captioning at the best quality possible under this unique and challenging circumstance. We are proud of the level of performance achieved.

    The Vlug settlement made no room for labour disruptions, including CBC’s unnecessary and (typically) antagonistic employee lockouts. In such an event, it was incumbent upon CBC to acknowledge the legal necessity of compliance with captioning rules and then prioritize. The lockout was not and is not a legally credible rationale for poor performance.

  14. Mr. Clark alleges that CBC/Radio-Canada resents the requirement to caption and does not take it seriously. CBC/Radio-Canada disputes this claim most vehemently.

    Vehemence is a defect to be remedied, not a negotiating position; vehemence partly explains why CBC faced the original complaint by Vlug and the current proceeding. It is typical, and typically unhelpful and self-sabotaging, for CBC to come out with guns blaring and vociferously declare the vehemence with which it disagrees. We don’t need that kind of hyperbole or histrionics.

Second letter from CBC (2006.01.20)

  1. Some [cases of missing captions] relate to technological issues or human error, incidences [sic] which are realistically to be expected and which we attempt to mitigate.

    It is not in question, least of all by me, that the Vlug decision permits an absence of captioning due to “glitches.” Of course nothing has 100% uptime, but “the rule should be full captioning.” Here CBC provides as an explanation a fact that is not in dispute.

  2. Others we can and will mitigate with operational changes.

    In other words, there are cases where CBC will fix the problems I have identified even if CBC, in its vehemence, cannot bring itself to admit them.

  3. We dispute Mr. Clark’s contention that captions should be present in some of the instances he cites. Our environmental scan may produce a rationale for revisiting these.

    Captions have to be present on all programming save for outside commercials under the Vlug decision. CBC specifically and explicitly does not have the right to pick and choose which programming or programming genres need not be captioned. This “environmental scan” is nonsense, since CBC will use it as a rationale for replicating the worst habits of the industry, many of which it already carries out anyway.

  4. As Mr. Clark points out, some missing captions are due to the failure of captioners to leave the CBC encoder in pass-through mode. This is clearly a human-error issue. Both CBC and the captioning services management remind captioners to be vigilant in this regard, and the incidences [sic – the word is incidents] are relatively rare. VTR operators are now expected to monitor all main network captioning sessions, in order to catch disconnects or garbled captions.

    We cannot take this promise seriously. It is humanly impossible for a single person to sit with eyes glued to a television without interruption for an entire shift. It borders on impossible even for several people to do that. Anyway, the kind of malfunctions we’re talking about tend to happen for a few minutes at a time and are remedied, in the case of encoder pass-through, when the newscast comes back on. They can monitor it all they want; they won’t be able to fix it most of the time.

  5. Many of Mr. Clark’s noted absences occur at the top of the clock on Newsworld. This is coincident with changing shifts among captioners. As it is impossible for one captioner to cover an entire 24-hour period, there is a period during which one captioner must kick out and the next dial in. We are investigating whether a solution to this problem is possible.

    A solution is plainly evident and is available to any competent engineer: Switch the shift changes from the top of the hour, when new programming always starts, to a commercial break. You can make all the mistakes you want during outside commercials; CBC is under no obligation to present those with captions.

  6. Unfortunately, power outages and encoder disconnects are technological failures that are bound to occur on occasion. Mr. Clark makes specific reference to the Pan Am Games in 2003. CBC fault records indicate failure on August 10, 2003 for 15 minutes due to a captioner’s inability to establish connection with CBC/Radio-Canada encoders.

    This could be reasonably considered a glitch, albeit an unusually lengthy one, under the Vlug decision. To say the same thing an umpteenth time, of course equipment will break down occasionally and staff and captioners will make mistakes occasionally.

  7. Line-ups and news promos for The National and Canada Now, which may have been uncaptioned formerly, are now captioned routinely. Mr. Clark’s examples arose primarily in 2002 and early 2003, and were a result of the captioning learning curve in individual program units. Captioning is still absent from late-breaking news on rare occasions. We are seeking means to further ensure that captioning is not missed on these segments.

    Bumpers, billboards, sponsorships, and special presentation voice-overs are added in different areas, at various stages of programs in preparation for broadcast. In some cases, captioned program material is edited for time and uncaptioned bumpers are added. These added elements then require captioning, prior to ingest for presentation. Other programs require editing due to the addition of “hosted prime” segments. While in most cases the process has been fail-safe, we have uncovered several opportunities for the inadvertent omission of closed captioning. We have augmented the coordination process to avoid such problems in future.

    Here CBC concedes that my contention is correct and further volunteers an additional area in which captioning is absent.

  8. Sub-titled [sic] programming is not further captioned in normal circumstances. While some individuals may prefer more text to cover the video images, it is CBC’s view that, on balance, the marginal gain from the addition of captioning for the hearing impaired is outweighed by the additional video lost.

    This is the kind of nonsense that an “environmental scan” might back up, and it is the kind of nonsense that the Canadian Human Rights Commission might believe, since the Commission is staffed by Francophones who may have trouble differentiating captioning (sous-titrage) from subtitling (sous-titrage). Let me unpack this hogwash one step at a time.

    CBC seems to think that a captioned subtitled program will reproduce the original language in text. This is, of course, an absurdity and speaks to CBC’s unfamiliarity with the true nature of captioning and subtitling. Captions in a subtitled program do all the usual things they do, like identify speakers and render non-speech information. They also occasionally fill in unsubtitled utterances. Captions are added to the subtitles and render what the subtitles don’t. No utterances are duplicated in simultaneous captions and subtitles.

    This practice dates back at least 24 years (Blatt 1982) and the differences between captioning and subtitling have been documented for a generation.

    A standard list of utterances rendered differently by captions and subtitles is as follows:

    1. Captions are intended for deaf and hard-of-hearing audiences. The assumed audience for subtitling is hearing people who do not understand the language of dialogue.
    2. Captions move to denote who is speaking; subtitles are almost always set at bottom centre.
    3. Captions can explicitly state the speaker’s name:
      1. [MARTIN]
      2. >> Announcer:
    4. Captions notate sound effects and other dramatically significant audio. Subtitles assume you can hear the phone ringing, the footsteps outside the door, or a thunderclap.
    5. Subtitles also translate onscreen type in another language, e.g., a sign tacked to a door, a computer monitor display, a newspaper headline, or opening credits.
    6. Subtitles never mention the source language. A film with dialogue in multiple languages will feature continuous subtitles that never indicate that the source language has changed. (Or only dialogue in one language will be subtitled, or the third-language passage will be italicized.)
    7. Captions tend to actually transcribe and render utterances in a foreign language, or transliterate that dialogue if a different writing system is used, or state the name of the language being spoken.
    8. Captioning aims to render all utterances. Subtitles are selective and do not bother to duplicate some verbal forms, e.g., proper names uttered in isolation (“Jacques!”), words repeated (“Help! Help! Help!”), song lyrics, phrases or utterances in the target language, or phrases the worldly hearing audience is expected to know (“Danke schön”).
    9. Captions render tone and manner of voice where necessary.

    I asked leading researchers in subtitling – Henrik Gottlieb, Jan Ivarsson, Jorge Díaz-Cintas, Fotios Karamitroglou – to double-check my list in case I had left something out, and Díaz-Cintas (2006, personal communication) added the following:

    1. Silences are never stated.
    2. Ironic statements are never marked in the subtitles because one is supposed to get that information from the soundtrack.
    3. Swear words and taboo language tend to be deleted in the translations.

    In case these examples seem minor, they aren’t. Answers to simple yes/no questions in commonly-understood languages like French aren’t subtitled because it is assumed that the hearing viewer knows what oui and non mean. But the deaf viewer never hears the response and has no written rendering to go by. The deaf viewer cannot follow that segment of the program.

    And, again, if you think that specific example is rare, it isn’t. I direct CBC’s and the Commission’s attention to combined captions and subtitles in Traffic (Steven Soderbergh, 2000; photos), including a case in which an answer to a question was not subtitled and had to be captioned.

    I assume it is not CBC’s real contention that songs should not be captioned, nor should any sound effects that would otherwise be captioned in an English-language show. Otherwise why did that woman know she needed to pick up the phone? How did those firefighters know they had to get in their firetrucks and race out of the station? How did the child know his pet cat was scratching at the door?

    If CBC holds to its errant philosophy that subtitling is captioning or the next best thing, we need to assess how many programming hours per year will go uncaptioned. Looking at a reasonably current schedule, I see one hour a week of dull Quebec French-language programming (like Fortier or Music Hall but never Les Bougon). Let’s give them the benefit of the doubt and assume those shows run for half the year and have 60-minute runtimes with commercials. That’s 26 weeks at 0.5 hours or 13 hours a year.

    We also note occasional International Movie Nights that are presented with subtitling. There’s at least one of those a month, with a two-hour runtime. That’s 24 hours a year, bringing our total to 37 hours a year. It seems small until you realize it’s equivalent to two full average broadcast days on CBC Television (135 average hours ÷ 7 days = 19 hours/day × 2 days ≈ 37 hours).

    If CBC Television is on air an average of 7,020 hours a year, then deliberately refusing to caption 37 hours a year (0.5%) means that CBC’s maximum captioning is 99.5%, not 100%. Given that the CBC already admits that the caption absences I have identified often take place (and admits other absences I did not list), CBC Television alone is never going to reach five-nines reliability, let alone the mandated rule of full captioning save for glitches.

    This admitted corporate policy of refusing to caption subtitled programming is sufficient to demonstrate that CBC is in noncompliance with the Vlug settlement.

    An additional point can be made. I have seen two subtitled films on CBC that also had captions – 2 secondes and The Man Without a Past (2006.03.01). I had no significant complaints with the former save for the absence of speaker IDs, which is significant. The latter preposterously used scrollup captioning, leaving stray captions superimposed on top of subtitles.

  9. Mr. Clark alleges that CBC in-house captioners are unqualified. This is untrue. CBC hires the best qualified persons available, usually experienced captioners, or university graduates with a major in English.

    Again, let’s unpack. There is no such thing as a certifiably qualified captioner because there is no captioning certification outside of court reporting or stenography. Hiring a person with experience doing offline captioning in Canada is a recipe for reproducing all the usual errors of Canadian captioning. An English degree is “useful” in writing university papers on Dickens or DeLillo, but has nothing at all to do with the actual task of transcribing wide varieties of spoken English (a task English-lit students never do), rendering non-speech information, operating specialized software, and doing all the things necessary to place and time captions correctly.

    English literature is the kind of degree a person with a superficial and faulty understanding of captioning would think is important. Captioning is a job that English-lit majors end up in, not one they’re qualified for. (It is also a job they might feel superior to. They spent all that time writing about Dickens or DeLillo and here they are stuck at work on a Sunday captioning a kids’ show.) The Corporation’s jobs are, however, desirable for these candidates, who otherwise will never be employed in their field and who earn at least $38,250.20 a year doing captioning at CBC (CMG 2006).

    In fact, we can draw a comparison with the Caption Center at WGBH, which conducted a survey of its Boston office on my request (WGBH 1987) and listed its staff qualifications as follows:

    • Male, Oberlin ’81 sociology/anthropology with emphasis on sociolinguistics, at WGBH three years, experienced typesetter, working knowledge of several languages
    • Male, U. Nebraska English, two years at WGBH, worked in a bike shop
    • Female, Brown U. ’81 comparative literature, fluent in German and French, background in translation and clerical, two years at WGBH
    • Female, Boston College ’78 Romance languages and English, one year at WGBH, systems analysis for hotel chain, taught ESL
    • Female, three years at WGBH, B.A. English, varied background, artist, chef, writing/reporting
    • Female, one month at WGBH, Middlebury ’83 European history cum laude, year at U. of Paris, editor, writer for market research/publishing firms
    • Female, one month at WGBH, Harvard ’84 comparative literature and literary theory, sales assistant for radio station, proposal writing, word processing

    A more realistic qualification is experience in proofreading or copy-editing, which WGBH has also used in the past.

  10. They are subject to testing designed specifically to evaluate captioning abilities.

    In fact, published evidence (on the blog entitled Dignan and Anthony) documented that the “testing designed specifically to evaluate captioning abilities” amounted to a rudimentary proofreading test.


    I got an interesting phone call this morning from the CBC. Over the past few months, I have been applying for jobs at the Corporation and I finally got a call asking for an interview. The job is for a closed captioner, which certainly isn’t an ideal position for me but one that I think wouldn’t be too difficult to handle.... I’m hoping that this job will lead to something better, and might be a stepping stone to a full-time position doing something a little more my style and interest.


    I headed down to the CBC building today to have what I thought was an interview. It turns out that I only had to take a test. I had to proofread a two-page essay, and to correct 15 misspelled words. It was quite easy and I finished in 14 minutes. I just hope that I didn’t make any stupid mistakes. The essay was such a complete piece of garbage that wanted to completely rewrite it rather than simply correct the grammatical errors.


    I hope that the CBC gives me this job so I can stop this hideous search.... I went in today, and it was one of those interviews that you know isn’t really an interview at all, but is an explanation of the job that they have, essentially, already given me. I’m not sure exactly why I was chosen over the other candidates, and it makes me a little nervous that I might be getting into something unpleasant. I mean I couldn’t have been the only person who spelled “occasionally” right on the spelling test. I couldn’t have been the only one who noticed the you’re/your error on the grammar test.

    There is no such thing as a “test... designed specifically to evaluate captioning abilities” because there is no industry standardization whatsoever. The actual day-to-day captioning process will be unfamiliar to most or all applicants, even those who have watched captioning in the past. CBC has not tendered evidence that it, unique among all captioners, has come up with a means of testing aptitude for captioning by proxy. It has not demonstrated that it has a unique test that can assess, on a sheet of paper, the ability to caption using computer software and live audio and video.

  11. CBC uses an in-house guide to maximize consistency in style. Interpretation and judgement may differ slightly among captioners.

    CBC admits that it has an in-house style guide (a copy of which was requested as part of this proceeding), but CBC does not explain the research basis for its contents. For all we know, it could be just like every other Canadian captioner’s in-house style guide, which consists of writing down what they already do. The purpose of the style guide is vitiated by CBC’s own admission that different captioners caption differently. Even if the style guide were rock-solid, which CBC has not demonstrated and arguably cannot demonstrate, it is of scarcely any value whatsoever if day-to-day captioning is subject to interpretation and judgement.

  12. Mr. Clark cites the use of upper case. CBC policy for use of upper case is consistent with the correct usage as stipulated in the CAB closed-captioning guidelines.

    I find it odd that CBC cites the faulty and discredited CAB captioning manual (2004) in this one and only case. (I thought CBC’s “in-house guide... maximize[d] consistency in style.”)

    The CAB manual states, or, more accurately, peremptorily declares without justification:

    Upper-case lettering is the standard in Canadian English language closed captioning. This will remain so until new standards can be developed for the proper use of mixed case, at which time it will likely replace upper case as the standard.

    In other words, we’ll keep doing it the old way until we figure out how to do it correctly. The manual does not claim that all-upper-case captions are better, merely that the authors hadn’t gotten their acts together enough to write a manual that requires mixed case.

    Captioned text is traditionally in upper case because all television encoders and decoders display it clearly. New encoders and decoders treat mixed case more legibly than in the past, and with this capability caption providers will be better able to follow standard English orthography in the future.

    The manual at least concedes that near-invariant upper case constitutes misspelling (nonstandard orthography). It does. What it manages not to express adequately, presumably because the authors do not understand the issue, is that upper case is “clear” because the lower case on early decoders did not include descenders. (Encoders have nothing to do with visible captions, which are locally generated in the viewer’s television or device.) For “clear,” we can instead read “less illegible,” as had been documented 15 years before the CAB manual was written (Clark 1989). Other researchers have treated the topic in previous decades.

    The authors of the CAB manual ignored the actual facts of caption typography, and ignored published research dating back up to 24 years, choosing instead to declare upper case a standard without justifying it.

    Captioning manuals from other countries and organizations, including Captioning Key by the Captioned Media Program, require mixed case (“The CMP requires pop-on captions in upper- and lower-case letters with descenders”). It simply would not occur to captioners in other countries to caption in upper case only.

    Unlike the CAB manual’s authors or CBC staff, circa 1979 I talked to a PBS engineer who worked on the initial decoder project and was told that engineers bough an off-the-shelf chipset with minor modifications (a redrawn question mark and an added staffnote ). When the CAB manual offers the clumsy and imprecise blandishment that “[n]ew... decoders treat mixed case more legibly than in the past,” they are telling us that more and more decoders have descenders on their lower-case letters. They do. So let’s use them.

    It is worth pointing out that the CAB manual; essentially everything the CBC produces outside of captioning; CBC’s and the Commission’s letters in this proceeding; and the majority of text we read in the run of a day are written in mixed case because that is how we write the English language. The CBC seems committed to following inane and backward advice when it reinforces errors in its own processes, but in fact industry practice and available research militate against all upper case.

    I assume CBC and the Commission are familiar with research in psychology of reading that demonstrates slower reading speeds with all-upper-case text. If not, they can consult Larson 2004 for citations and summary.

    Quite simply, the CBC is producing 1979-standard captions in 2006. This alone merits immediate remedy.

  13. Any departures from that standard are rare. Misuse of character encoding and misjudgement in positioning are also rare.

    CBC concedes they happen.

  14. There is no disregard for quality, but the high-stress, time-sensitive broadcast environment may result in some captioning errors.

    There is no understanding of quality that CBC could disregard.

  15. CBC will schedule a review of protocols with our staff, as well as our outside service providers, to further reinforce the importance of consistency and quality in the captions created.

    What CBC needs to schedule is a meeting with me, and I want to be present at the meeting with staff and “outside service providers.” I have all the research and have the most expertise in captioning.

  16. Mr. Clark cites CBC’s use of scrollup captions on fictional narrative programming. This is relatively rare as most prime time programming comes to CBC complete with pop-on captions. We calculate that 63% of a typical broadcast day on the CBC main network and 80% of non-live programming are captioned with pop-on captions.

    That sounds great, but in reality CBC has a policy of captioning entire series (e.g., Coronation Street and Doctor Who) in scrollup. If you like those shows, then it is 100% of the time, not 37% or 20% (already high figures), that you must struggle to keep up with a dramatic series using scrollup captioning.

    In fact, I have viewed the following genres with inappropriate scrollup captioning:

    Exceptions may occur when CBC must apply captioning on a short timeline prior to broadcast, or when CBC must edit previously-pop-on-captioned films for time.

    In the latter case, the correct course of action is to have the original captioner re-edit the file, i.e., carry out a reformat. It is a simple and commonplace procedure and much less of a drain on staff time than recaptioning in any method. Essentially, CBC admits that it destroys the integrity of an original program’s captions, and does most of the work of recaptioning from scratch, for no viable reason.

    Certain editing equipment strips captions which must be recreated.

    Such equipment must be repaired or replaced.

    Occasionally a distributor delivers uncaptioned programming less than 48 hours prior to broadcast. In these cases, the program or film may be re-captioned using the scrollup method.

    Again, that sounds almost convincing, but the fact is that leading U.S. captioners can produce pop-on captioning in that timeframe. I know because I specifically asked.

    I sent questions to the top five captioners by volume in the U.S., and to nearly every member of the Accessible Media Industry Coalition that carries out offline captioning. Most of the top five responded, as did a couple of the AMIC members. I asked the following:

    1. If I’ve got a one- or two-hour program and I get it to you “less than 48 hours prior to broadcast,” can you turn around a captioned and encoded tape, or can you produce a file only? (Cost penalty, if any?)
    2. If I’ve got a live program that is real-timed and it is scheduled for repeat several hours or days later, what’s the extra cost to clean up the caption text and retransmit as live display?
    Captioner 48-hour turnaround? Cost penalty? Cleanup?
    AYes; file or tape“Typically” not done
    BYes; file onlyYes; +½ to +⅓ the cost
    CNoYes; +100% the cost
    DYes, file or tapeYes; cost varies
    EOnly for a good client[No answer]

    Another vendor replied to all questions with what amounted to a maybe.

    The conclusion, based on attestations as opposed to CBC supposition, is that industry-standard practices at experienced captioners can turn around a pop-on-captioned show in 48 hours and can clean up and re-feed a real-time-captioned program.

  17. Mr. Clark is correct in his assertion that CBC chooses the use of scrollup over pop-on captions in many instances due to time and resource constraints. It is worth noting that the CAB guidelines state, “off-line roll-up captions are being used more and more frequently as an acceptable alternative to off-line pop-on captions” and “when deadlines are extremely tight, roll-up captions can be prepared more quickly than pop-on.”

    Here’s what the CAB manual actually says (p. 6):

    Off-line pop-on captions... are the only type of captions well suited to dramas, sitcoms, movies, and music videos, and they are therefore recommended for these types of programming.... Off-line roll-up captions are not well suited to dramas, sitcoms, movies, children’s programs, or music videos, and their use for these types of programming is discouraged.

    The practice is, furthermore, banned or discouraged by other published captioning manuals:

    1. The Caption Center Manual of Style limits use of scrollup captions to “programs produced very close to airtime or produced live.” (A later clarification states “The rules we have are unwritten and usually involve a discussion between marketing and the client. We have tried to hold a line that says comedies and dramas should be pop-on; news material, documentaries, talk shows, etc. can be rollup. The gray area in-between frequently comes down to turnaround time and cost.”)
    2. Even the outdated and misguided CCDA style guide (submitted in a letter of 1995.01.04 from Beverley Ostafichuk, Canada Caption Inc., under Public Notice CRTC 1994-18) declares: “Traditionally, roll-up captions are used only for live captioning, and offline work is done with pop-on captions.”
    3. Captioning Key: Preferred Styles and Standards (Captioned Media Program, National Association of the Deaf, 1995) flatly states: “The CFV program requires pop-on captions.”
    4. The ITC (now Ofcom) Guidance on Standards for Subtitling limits scrollup captions to real-time stenography:

      The use of scrolling word-by-word captions for live news can present particular difficulties, especially in the area of retention of information. It is not just a problem of speed, but that the text placed upon moving lines is working against the readers’ natural reading strategy.... During an unscripted live broadcast... the subtitles must be composed, entered, formatted and transmitted in a single pass through the program.

    5. The Australian Television Broadcasting Services (Digital Conversion) Act 1998 Draft Captioning Standards don’t even mention scrollup captioning, presumably because it is axiomatic that the method would be used purely for live events and programs completed close to airdate.
    6. The only book currently in print on subtitling (Ivarsson and Carroll 1998) states that neither scrolling nor crawling

      systems can be recommended (except possibly where a live subtitling is being broadcast for the hard-of-hearing). The problem with moving text is that the eye concentrates on the text while waiting for the word or phrase to finish and is thus prevented from seeing what is happening in the picture. This does not matter very much if the picture is only showing the news presenter, but if the images are conveying important information, such fixation on the text is disastrous.

      People who are used to this system may say that in time you learn to move the eye upwards, take in the picture for a moment and then look down again to read some more of the text. But in that case, what is the advantage of a scrolled system over the “quieter” traditional system?

    Based purely on the existing evidence, the use of scrollup captions for fictional narrative programs should be prohibited in all cases save for emergencies. Of course some captioning is better than none in emergencies, but CBC scarcely ever faces such emergencies. Even when necessary, scrollup-captioned fictional narrative programming should be recaptioned in pop-on for later rebroadcast. (Yes, caption it twice. The initial captions must be rectified.)

  18. Captioning with pop-on captions takes four times as long as captioning with scroll up captions.

    Correct captioning always takes longer than incorrect captioning. By its own admission, not more than 33% of CBC Television’s programming is captioned in scrollup, not all of which will be fictional narrative programming. CBC has not demonstrated that using correct pop-on captions for all fictional narrative programming save for emergencies constitutes undue hardship. CBC would never countenance the use of electronic newsroom captioning for live news broadcasts, since it is unequivocally inappropriate and is a failure as an accessibility method for deaf viewers. (It’s also banned by the CAB manual at p. 8.) CBC cannot seriously be advocating the use of another failed captioning method for a different genre of broadcast programming.

    In any case, CBC’s expensive captioning software permits several captioners to work on the same program. For all but emergencies, I don’t see why multiple captioners cannot caption different segments of a program simultaneously or nearly so. It’s done all the time in U.S. captioning houses.

    Correct usage of pop-on captions for fictional narrative programming may require hiring more staff. While this means that the ranks of unemployed English-lit grads might decrease by one or two, and also means that incorrect CBC captioning practices will be transmitted to gullible new staff who have no other point of reference, it might also mean that CBC actually will carry out its responsibilities adequately and provide correct captioning on its programming.

  19. Mr. Clark cites the example of The Greatest Canadian, and points out that there was time for [audio description] to be produced for this series. In fact, the production of [audio description] for this series was extremely challenging for both the network presentation areas and the description service as the last minute delivery of the finished programs left inadequate time for pop-on captioning.

    I’m just trying to understand how hard it might be to re-stripe a third audio track onto a tape before telecast. Has CBC considered that its description service, AudioVision Canada, may be seriously outdated and has proven itself barely capable of real-world jobs with quick turnarounds? If CBC didn’t have to transfer file formats from DigiBeta to VHS for AudioVision’s use, might the turnaround have been shortened? Could a competitor such as Galaviz & Hauber have turned the job around faster? (It is known that the latter firm can handle digital tape and can handle same-day jobs on occasion, and of more complex material than a nonfiction program like The Greatest Canadian.)

    At any rate, even if we were to accept CBC’s explanation for The Greatest Canadian, how do they explain the three CBC Sports specials on the 2006 Turin Winter Paralympic Games, aired on CBC Television on April 1, 8, and 15, 2006? The Paralympics ended 2006.03.19, giving CBC between 12 and 27 days to edit and post the three programs. And all three were presented with audio description. (The first episode displayed a crawl at the top of the screen informing blind viewers, who could not see the crawl, that the program was described.) There was, quite frankly, tons of time to caption these shows in advance, but they weren’t: Real-time captioning was used, complete with serious errors in and omissions of proper names. And if you need further evidence, note that the episode broadcast April 15 was preceded by a profile of skier Beckie Scott, captioned in advance.

    How does CBC Television explain its usage, on 2006.04.22, of scrollup captioning for a repeat of the Turin Olympics women’s gold-medal hockey game, which took place on 2006.02.20? I think a two-hour telecast can be reasonably captioned in two months’ time. Instead, we suffered through the usual nonsense of CBC Sports Saturday captioning, in which stenographers are not given the full list of proper names, utterances are uncaptioned or paraphrased, and captioning lags behind the action.

    It’s pretty simple: Programs that aren’t live shouldn’t be real-time-captioned. Yes, in highly unusual circumstances an emergency will happen and there really is no other way to caption a program – for that telecast. All other shows, and repeat broadcasts of those emergencies, can be and must be captioned in advance.

  20. CBC currently leaves in place the original real-time captions when programs that have been captioned in real time are repeated.

    In other words, once the program has aired and is no longer live, we pretend it still is live. We just rerun the tape, with its choppy, piecemeal captions that run three to ten seconds behind the audio and misrender or leave out entire words, including proper names.

    To recaption all this programming would be prohibitively expensive without significantly increasing the value of the captioning.

    It isn’t “prohibitively expensive.” You’ve already got a transcribed program (with errors and omissions, but transcribed nonetheless) in scrollup format. It is a question of rerunning the program while reading the transcript and making corrections. The total time outlay might be 25% longer than the program runtime, and the result is two captioned telecasts – one where captioning is unavoidably imperfect and another where the captions have been corrected.

    If this method is “prohibitively expensive,” why has a major cheapskate like CHUM Limited done exactly that with live MuchMusic performances and repeats for many years? (I have recently seen an occasional rebroadcast that keeps real-time captions intact, but CHUM has tended to recapture the captions and clean them up for scrollup.)

    In any event, the legal standard is not prohibitive expense but undue hardship. CBC is already spending $1.1 million a year on captioning. We’ve got about four hours a week (chiefly CBC Sports Saturday broadcasts) that need this kind of cleaning up. The outside captioner can do it, and, in the case of CBC Sports Saturday, they’ve got more than 24 hours’ turnaround time to clean up the captioning. This is a drop in the bucket compared to the effort involved in simply real-time-captioning all the local and national newscasts on CBC Television.

    If CBC seriously contends that real-time captions are OK, then why doesn’t the network use it for flagship dramatic series?

  21. CBC Newsworld may recaption real-time programming that has previously been captioned on the main CBC network. While these programs may look identical to the viewer when broadcast on Newsworld, they have been repackaged with new intros and other elements.

    Do the same as above: Clean up the real-time captions and present them anew.

    Mr. Clark points out that the real-time captioner simply stops if there are apparent pop-on captions. We have identified a possible operational adjustment that could improve the presentation of such programming.

    CBC concedes that previously-captioned prerecorded programming is needlessly recaptioned using an inferior method.

I will not be shut out of this process

We will continue to review procedures and associated resource requirements in accordance with industry standards and responses from consumer groups and will provide you with further results of our investigation when they are available.

Industry standards are nonexistent or discredited. I have demonstrable expertise in captioning dating back more than 25 years, nearly two decades longer than the most experienced of CBC’s in-house captioners. I have already done paid work for CBC on captioning and have volunteered this information for the betterment of CBC, CBC’s captioning, and CBC captioning viewers. I will not sit by while CBC conducts one or two perfunctory meetings with cherry-picked deaf groups. CBC’s entire captioning staff is composed of hearing people like me, and we all have a role to play in improving CBC captioning. I provided this information.

While this is the kind of spite we expect from a corporation that reacts most strenuously and with vehemence whenever anyone dares to suggest they are imperfect, I will not be shut out of any future or ongoing consultation process. I’ve already offered repeatedly to meet with CBC on this topic. The eventual response dismissed three years of labour an “informal complaint” for which a meeting “would not serve a useful purpose at this juncture... We will be meeting with groups directly affected by our captioning to discuss captioning issues.” I am directly affected by captioning.


  1. Armon, Carl, Dan Glisson, and Larry Goldberg, 1992. “How closed captioning in the U.S. today can become the advanced television captioning system of tomorrow.” SMPTE Journal, July.
  2. Blatt, Joseph, 1982. Personal communication (“[O]ur captions included some speaker identifications that we added to the BBC titles”).
  3. CAB (Candian Association of Broadcasters), 2004. Closed Captioning Standards and Protocol for Canadian English-Language Broadcasters.
  4. CBC, 2005:51. CBC/Radio-Canada Annnual Report 2004–2005.
  5. Clark, Joe, 1989. “Typography and TV Captioning.” Print, January/February.
  6. Clark, Joe, 2006. Response to report on captioning on French CBC channels.
  7. CMG (Canadian Media Guild), 2006. CMG Salary Rates.
  8. FCC (Federal Communications Commission), 2005. “Notice of proposed rulemaking” (CG Docket Nº 05-23 1).
  9. Global Translation Inc./Translate TV, 2005:3. “Comments of Global Translation, Inc.” In FCC (2005).
  10. HBO, 2005:8. “Comments of Home Box Office, Inc. ” In FCC (2005).
  11. Ivarsson, Jan, and Mary Carroll, 1998: Subtitling. Simrishamn, Sweden: Trans-Edit.
  12. King, C., LaSasso, C. and Short, D., 1994. “Digital captioning: Effects of color coding and placement in synchronized text-audio presentations.” In H. Maurer (Ed), Educational Multimedia and Hypermedia. Charlottesville, VA: Association for the Advancement of Computing in Education.
  13. Kirkland, C. Eric, 1999. “Evaluation of captioning features to inform development of digital television captioning capabilities.” American Annals of the Deaf 144(3):250–260.
  14. Larson, Kevin, 2004. “The science of word recognition.”
  15. TDI (Telecommunications for the Deaf, Inc.), 2005:14. “Reply comments.” In FCC (2005).
  16. Thibault, Marie-Josée, François Lapointe, and Philippe Mabileau, 1998. “Standard for French close-captions: A RQST and TESCO Inc. joint project.” TESCO Inc.
  17. Wells, Daniel R., 1980. “Captioning for the deaf: A PBS progress report.” SMPTE Journal 89.
  18. WGBH, 1987. Personal communication.
  19. WGBH (Navoy, Judith), 1997. Advanced Television Closed Captioning Features.
  20. WGBH, 2005:6. “Comments of the WGBH National Center for Accessible Media (NCAM).” In FCC (2005).