Can we stream digital video?

I had not even had a chance to open my daily e-mail from Inside Higher Ed yesterday before four colleagues had sent me a link to this story about an educational video trade association forcing UCLA to halt its practice of streaming digitized video on course Web sites.  Several suggested that I would surely want to blog about the story, and they were, of course, correct.

The story contains some chilling rhetoric from the representative of the Association for Information and Media Equipment – intentional, I am sure – about their plans to investigate and threaten other colleges and universities that are doing the same thing.  Many schools, of course, have explored these options because the pressure from faculty and students to provide greater digital access to our film collections is intense.  Some have concluded that the legal risk is too great and are resisting that pressure, at least for now.  Others have tried various justifications, clearly hoping to “fly under the radar.”  This story will certainly strike fear into many, and will give more ammunition to faculty members who complain that copyright law prevents them from teaching effectively in the media-saturated world of 21st Century America.

In response to the story, I want to suggest here what the major alternatives for legal streaming of digital video might be and the problems inherent in each alternative.  I know from conversations with colleagues that each of these strategies is being tried somewhere.

The first, and most obvious, possibility is to rely on the TEACH Act, which amended one of the Section 110 exceptions to the public performance right in copyright in order to allow “transmissions” of certain performances for distance education.  TEACH (or Section 110(2)) has a lot of specific requirements that must be met (see this TEACH ACT toolkit from NC State University), although many of those requirements would appear to be satisfied when digital video is streamed through a closed-access course management system.  The real problem with relying on TEACH is the portion limits it imposes; it permits transmission of entire “non-dramatic musical and literary works” and “reasonable and limited portions” of other audio-visual works. This second provision seems to apply to films and to disallow the transmission of entire films.  Some institutions would argue, I think, that an entire film is often the only “reasonable” portion to use for a particular teaching purpose, but that argument ignores the word “limited.” The point about a reasonable portion is well-taken, in my opinion, but only proves that TEACH was never an adequate solution to this problem.

Other institutions could assert fair use as the justification for streaming digital video.  These schools would point out, I imagine, that courts have often held that the use of an entire work can be a fair use, based on the overall balancing of the fair use factors and the totality of circumstances.  The trade group clearly disagrees, although the comments about fair use in the article are not really on target.  It is correct that password-protection alone is not enough to guarantee fair use, but it does strengthen the university’s position in the complete analysis of the factors.  Simply to say that a password does not make something fair use is as incomplete as asserting that an educational purpose always means a use is fair; both assertions miss the need for a complete examination and balancing of the factors.

The problem I see for a fair use justification is that courts would be likely, in my opinion, to look at the portion limits in the TEACH Act and say that that legislation was Congress’ opportunity to provide guidance on educational transmissions, and it selected a limited standard.  A court that took that approach would be unlikely to let a school “shoehorn” the entire film in under fair use, simply in order to avoid the inconvenient limits imposed by TEACH.  But I have to add that there is no agreement on this point — even the intern in my office this year, who is also a lawyer, disagrees with me — and some universities clearly have decided to rely on fair use to stream entire videos.

Perhaps the most interesting argument, however, is the one that UCLA seems to be making, according to the article, based on the performance exception that proceeds TEACH in Section 110 and permits performances in face-to-face teaching situations.  Section 110(1) is clearly meant to have some “give” in it, since it refers to “teaching situations” rather than classes and to “classroom[s] or similar place[s] devoted to instruction [emphasis mine].”  UCLA seems to want to stretch these terms to include the course Web site as part of the face-to-face instruction.  I know of other institutions, less bold than UCLA, perhaps, but still unwilling to accept unworkable limitations, that read 110(1) to permit streaming to designated sites like language labs, but not to course sites that can be accessed from anywhere.  These efforts to clarify the fuzzy boundaries of 110(1) are fascinating and seem to invite a court to step in and clarify; it is just that no one wants to be the defendant in that case if they can help it.  While I admit to lingering doubts, this last approach seems to me to be the most surprising, yet most promising, of the three.

There is still another obstacle, however, posed by the anti-circumvention rules added to copyright law by the Digital Millennium Copyright Act.  This provision prevents the circumvention of technological protection measures even, in some cases, when the purpose of the use would be permitted.  So ripping a DVD protected with CSS (Content Scramble System) may violate these rules even if it is otherwise legal.  The DMCA specifically stated that these rules should not inhibit fair use, but courts have been inconsistent about that provision in circumvention cases.

Also, the Library of Congress is charged with declaring categories of exceptions to anti-circumvention in a rulemaking process every three years.  New rules, which are desperately needed, were due in October 2009 but the Library punted, extending the old rules while giving itself unlimited time to try and craft new ones.  The situation is getting worse on university campuses and we have to ask when the Library of Congress is going to clarify the situation.

In the end, I agree with Tracy Mitrano from Cornell, quoted in the article, that this is one more place where copyright law is not up to the technological challenges posed in higher education today.  The need for revision “before [the law] does any more damage” is clear.  We can only hope that the educational media industry will eventually come to understand this (they are supposed to be educational, after all) and move away from threats and towards real dialogue.

15 thoughts on “Can we stream digital video?”

  1. Hey, Kevin. Do you take the availability of services like Netflix into account when considering the fair use argument for streaming readily available (ie, available from Netflix as either a download or by mail) films to students? One can subscribe for about $9 mo. (ie, about $27/semester) and get access to a considerable library of film. Not all films are available this easily or cheaply, and for those, do you think fair use would be a strong contender, even for entire films?

  2. It is not legal to show a film to a wide audience without clearing it with the copyrights holder. Renting it through Netflix, your video store, etc., is for family or a few friends use, not a whole classroom.

  3. Thank you both for the comments.

    Mary, I think your assertion about classroom showings of rented movies is incorrect. The Section 110(1) exception for performances during face-to-face teaching says the copy must be lawfully made, and a borrowed or rented copy usually meets that standard. These videos sometimes carry draconian warning notices, but an early 20th century Supreme Court decision told us that simply putting a passive warning on copyrighted works cannot take away rights that are granted by statute.

    Georgia, I do believe that the availability of Netflix streaming impacts this analysis in two ways. First, in some of the cases where faculty members want to stream video, it may make more sense to have each student obtain and view a copy, whether through Netflix or some other means. Video is both more central to teaching and easier to obtain than it was when I was a student, and we need to stop thinking of media as “enrichment” and recognize that sometimes it is core course content.

    Second, the failure to make some video available in a usable format that is becoming a standard in the industry — and Netflix is busy changing what is standard — could strengthen the fair use argument by indicating a lack of market harm. If we have bought the video in the only form it is available and need a streamed copy that is not available, where is the harm in making the stream? This could be especially persuasive in the absence of a well-developed licensing market (which raises another argument about the relevance of such markets in fair use analysis). But as you well know, that is only one important part of the fair use argument, it does not settle the question.

  4. Thanks for the thoughful comments, Kevin.

    I think one important factor that seems to be completely neglected in recent discussions re the UCLA affair is the fact that the commercial video marketplace is incredibly diffuse in terms of sellers, their primary audiences, and their products. One very small sector–the sector primarily represented by AIME–is the educational/independent documentary distribution community. Many of the distributors in this sector are moving into streamed video licensing models specifically geared toward both K-8 and higher education. If one of key tests of fair use is economic impact on the market, it’s hard to see how streaming such materials in their entirety without license can even vaguely be justified as fair use (or under the provisions of TEACH).

  5. Kevin,
    I agree with your comments and also Tracy’s exasperation with copyright law’s inability to keep up with technology. I have a real problem, however, with attempting to apply 110(1) to this fact scenario. Focusing on expanding the definition of the term “classroom” as a means for shoehorning (good analogy) the streaming of entire copyrighted videos (either in an online setting or to a room filled with live students) into 110(1) ignores the primary difference between 110(1) and 110(2). That difference is “transmission”, not “classroom”.

    As soon as the performance or display of a copyrighted work is transmitted, one is out of 110(1) and into 110(2).

    To ‘transmit’ a performance or display is “to communicate it by any device or process whereby images or sounds are received beyond the place from which they are sent.” 101 Definitions. As I have understood it, you cannot transmit and be in 110(1).

    110(2) already acknowledges that the online environment IS the students “classroom’. That change was made with the TEACH Act – the old 110(2) required the transmission “be made to a classroom or other place normally devoted to instruction.” See Lolly Gasaway’s TEACH Act Comparison Chart. http://www.unc.edu/~unclng/TEACH.htm. I recall making the same expanded classroom argument before TEACH removed the placebound restriction.

    TEACH also expanded the performance transmission to include the “reasonable and limited portions of works.” Before that, transmissions of performances had been restricted to nondramatic literary and musical works- period. (Keep in mind that the copyright act definition of “literary” specifically excludes AV works – so the old 110(2) had nothing for performances of “other works”).

    Even the Senate Report accompanying S. 483, TEACH Act, indicates that reasonable and limited portions are “less than the entire work”. (usually).

    While my sympathies as well as rational thought lie with the premise that if an entire film is worth showing to the F2F section of a class (110(1)), it is also worth showing, in its entirety, to the online students, the Copyright Act does not, as currently written and defined, permit that under 110(2). Would that it did. As Kenny Crews once said, “When you enter 110, you check your logic at the door.”

  6. Peggy,
    Thank you for the comment.
    I have one question to try and press on the rigid distinction between F2F classroom and transmission that you suggest. Do you think that a system that played films from a central location and streamed them into classrooms where a live class viewed them would be permitted under 110(1) or would it be subject to TEACH Act limitations? I have heard lawyers I respect suggest that this scenario would be acceptable under 110(1), since that provision does not specify how the performance must be accomplished. It was from this perspective, suggesting that 110(1) had some “give” in it in this regard as well as it the definition of teaching activities and classroom, that I was expressing some sympathy with this the 110(1) argument.

  7. Kevin,

    You’re welcome and I certainly did not mean to seem argumentative or rigid. Lord knows, I push the envelope as hard as I can in many areas. And I understand your need to explore the 110(1) argument, since this is not the first time it’s come up and it will be interesting to see where it goes. (opinion from ARL and Jonathan Band) However, I also think that at the time 110(1) and 110(2) were authored, that Congress intended 110(1) to cover the scenario where the faculty, the students, and the work were in the same location. Here is the relevant legislative history for Section110(1) of the 1976 Copyright Act:
    “There appears to be no need for a statutory definition of “face-to-face” teaching activities to clarify the scope of the provision. “Face-to-face teaching activities” under clause (1) embrace instructional performances and displays that are not “transmitted.” The concept does not require that the teacher and students be able to see each other, although it does require their simultaneous presence in the same general place. Use of the phrase “in the course of face-to-face teaching activities” is intended to exclude broadcasting or other transmissions from an outside location into classrooms, whether radio or television and whether open or closed circuit.”

    So, according to that legislative history, 110(1) does not apply to performances that are “transmitted”. And the last sentence appears to address the other question about transmitting from an outside location into a classroom.

    My question is -If “transmission” isn’t the difference between 110(2) and 110(1), what is? And why would you even need 110(2) – wouldn’t you always chose 110(1) and avoid all those nasty TEACH requirements?

    And again, as to your question regarding streaming videos from a central location to a physical place where live students were gathered (like a classroom), I think the legislative history addressed that as well. Do I like it? No, of course not. In fact, that scenario sounds a lot like the opinion that came out from ARL and Jonathan Band recently that condoned that activity. I doubt very much whether streaming a video into a F2F classroom from a central location on campus would ever give rise to a lawsuit, in ‘real’ life, although it could. (never say never) It is another example of being potentially punished for doing something more efficiently.

    Some areas of copyright law are more gray than others. Whether or not an activity fits within 110(1) or 110(2) isn’t all that gray unless the threshold question ceases to be “transmission” and instead becomes something more like the social good of the proposed activity.

    Best, Peggy

  8. Peggy,

    I never meant to suggest that you were being rigid, just wanted to test how rigid the distinction you were making is. Thanks very much for the detailed explanation.

    Kevin

  9. Coming late to this party, but would like to add some thoughts.

    First, to quote Charles Dickens, “Now, what I want is, Facts.” It’s really not entirely clear from the press reports I’ve seen exactly what it is UCLA is doing and for what purpose, and what the objections of AIME to it are. As we all know, the actual facts are critical to legal outcomes, particularly when it comes to fair use questions.

    That said, I will hazard some guesses.

    From the reports, it *appears* that UCLA is streaming videos in connection with classes of some sort, rather than setting up a more general video reserve that is available on line. If so, the primary objection is probably that the TEACH Act allows the performance of only “reasonable and limited portions” of video in the context of a TEACH class, unlike Section 110(1), which imposes no similar limitation on a face-to-face class.

    Note, however, that the definition of “reasonable and limited” is itself not entirely clear, and may perhaps actually authorize the performance of an entire video under *some* (albeit limited) circumstances. According to the legislative history, it’s a case-by-case determination: “What constitutes a ‘reasonable and limited’ portion should take into account both the nature of the market for that type of work and the pedagogical purposes of the performance.” (http://thomas.loc.gov/cgi-bin/cpquery/T?&report=sr031&dbname=107&amp😉 And a more recent Congressional Research Service report about TEACH read that language fairly broadly: “Although what constitutes a ‘reasonable and limited portion’ of a work is not defined in the statute, the legislative history of the Act suggests that determining what amount is permissible should take into account the nature of the market for that type of work and the instructional purposes of the performance. For example, the exhibition of an entire film may possibly constitute a ‘reasonable and limited’ demonstration if the film’s entire viewing is exceedingly relevant toward achieving a educational goal; however, the likelihood of an entire film portrayal being ‘reasonable and limited’ may be rare. (http://assets.opencrs.com/rpts/RL33516_20060706.pdf) (footnotes omitted)

    But, even if turns out (as no court to my knowledge has yet held) that TEACH doesn’t extend that far, TEACH is only a sort of “safe harbor” in the broader, though murky, sea of fair use, not the outer limits of what is permissible. Indeed, the legislative history also specifically notes that showing an entire film may qualify for fair use even if it doesn’t qualify for TEACH: “Fair use is a critical part of the distance education landscape. Not only instructional performances and displays, but also other educational uses of works, such as the provision of supplementary materials or student downloading of course materials, will continue to be subject to the fair use doctrine. Fair use could apply as well to instructional transmissions not covered by the changes to section 110(2) recommended above. Thus, for example, the performance of more than a limited portion of a dramatic work in a distance education program might qualify as fair use in appropriate circumstances.” (http://thomas.loc.gov/cgi-bin/cpquery/T?&report=sr031&dbname=107&amp😉

    It is here where I think the analogy to 110(1) could be useful, and perhaps why (when it otherwise appears to be completely irrelevant) it is being raised in this discussion. If the *only* part of TEACH that doesn’t “fit” is the portion limit — that is, if you incorporate all of the technological restrictions that are designed to prevent the material from “escaping” and being used for other, noneducational uses — you’re really doing no more than what is authorized by 110(1), at least in terms of impact on the copyright owner. It is true, technically, that you’re making and distributing a copy, but that’s also already allowed under TEACH — as well as under the “time shifting” and “space shifting” theories of fair use. In many ways, this is really just a variant of those latter theories — a sort of “place shifting” as it were. (I think you could even plausibly make a similar argument for a broader video reserve, though it would strongly bolster such an argument in that context if you limited simultaneous access to only the number of physical copies you have acquired.)

  10. Hello all

    I think the issue regarding whether or not the display of streamed video within the confines of a password protected LMS constitutes a logical and valid extension of F2F teaching is a red herring in this case. The much bigger issue, in my book at least, is the potential encroachment of UCLA (and others engaged in streaming without license) on evolving and currently available markets and services. (See my earlier comment). Again, I think it is extremely important to understand that the wide scope of video materials potentially used in the classroom come from a hugely fragmented marketplace. Netflix, amazon, et al. will undoubtedly continue to move aggressively into the home/mass market video on demand arena. Other sectors–educational and independent documentary films, for eg–are developing substantially different models for marketing VOD. Again, as I commented earlier, it is difficult to see how displaying an entire work that is currently or potentially being commercially licensed for the very same type of use meets the requirements of fair use (or TEACH).

  11. Gary —

    While I understand your point as a policy position, it is not the policy position that Congress has chosen. Impact on the market is an issue *only* under fair use (and even there, the fact that you *could* pay someone is not in itself sufficient to defeat fair use; otherwise, nothing would ever be fair use, because you always *could* pay someone.) Under both Section 110(1) (F2F) and Section 110(2) (TEACH), that factor is completely irrelevant. If you have a legitimate copy of a copyrighted work, you are completely free to perform or display it (subject to certain technical and portion limitations) without asking permission and without paying anyone anything more. Congress has decided that that limited incursion on the copyright owner’s rights serves a broader social purpose.

  12. Steve McDonald, many of whose comments I’m in sympathy with, said he’s coming late to the party. I’m even later here… yet hoping you are still reading on this streaming issue.

    I don’t quite have a set of thoughts fully fleshed and connected into an argument here. But I’ll proceed as if that’s close to the case, in numbered points below.

    1. I’m not convinced that any interpretation of 110(1 and/or 2)–conventional or creative– will make the case for CMS streaming of entire video titles. 110(1) provides guidance by analogy, in both rationalizing and carefully applying a Fair Use defense under 107. 110(2) simply stops you short, unless you read this section very narrowly. When I did that this past week–read it without my standard interpretation of how it applies to distance education, and without referencing the interpretive gloss in HR Report No. 94-1476–I arrived at a new spin: You can actually read 110(2)’s language as applying–by analogous extension from the physical classroom to that online–to synchronous transmissions comparable to broadcast, rather than asynchronous ones. Thus, consequential thinking goes, you’re free to deal with asynchronous streaming under Fair Use. This would be a creative dodge of the “rules” worthy of our financial services industry. But then you have to argue that the House Report’s interpretation of 110(2) can be ignored–because it ain’t written explicitly into the law itself. A hard sell.

    2. Exactly how ARE courts likely to balance the indicators–the law itself (in discrete and competing sections), official interpretive documents, and even practice-in-context over time? Sorry Kevin, Georgia, Tracy and everyone who really IS a lawyer: you may have a better take than yours truly. I just take random observations that seem to resonate. In Eldred, the SCOTUS read our Constitutional copyright clause so narrowly as to avoid an unorthodox practical conclusion about “limited times” and punted it to Congress (in effect allowing, like the late Jack Valenti, that the qualification COULD mean eternity-minus-one-day). Then again, Fair Use defenses have been upheld when appropriation of an entire work was at issue. I don’t know about you, but it’s always bothered me that among the 107 criteria is “amount of the portion of the work”. It’s true to insist that often the whole thing is what’s needed for the teaching goal, but it’s just clever to say that it’s also the portion required. Amount of the work, yes; portion, no; and unfortunately the second terms qualifies the first. I don’t know of any English language dictionary that does NOT define “portion” as being somewhat less than an entire thing. But there it is, in decision-making if not the wording of the law itself.

    3. So where are we? In my mind, still with enough uncertainty–still with enough speculative maneuverability–to say that just because 110(2) deals with distance ed without naming it in the statute itself, doesn’t mean it has to circumscribe 107’s Fair Use. I’m not inclined to buy that, short of a real court opinion on the matter. Not even an out-of-court settlement will do.

    4. Having said that, another observation: if a defensible act of Fair Use involves copying, then it produces by fact of that act a “lawful copy.” Further, another hidden-in-plain-sight oddity that struck me recently: Fair Use deals with works. It doesn’t deal with copies in which works are embodied. The upshot of this–if I haven’t failed to connected dots from some other parts of the statute–is that the source of the work treated to a (putative) Fair Use doesn’t have to be a lawful one itself. Though an educator would have to be in either an ethically challenged or an absolutely unavoidable situation to use one that is not. But the latter situation is defensible, I think. I’ll give you one here that’s closest at hand because it will involve me. I have an interest in Latin American movie posters from the classic period of the 1930s-50s. My university library mounted an exhibit of posters from Mexico recently (Spanish Civil War exile artists). We’re moving soon to an online version that includes Argentine posters as well. My conception for the exhibit will include clips from the movies portrayed, that help give significance to the artistic elements stressed in the posters. All images in the exhibit will be from real posters and other graphics (lobby cards, magazines) in my possession, or from publicly accessible (and linked) Internet archives. Photography of the posters has been specially done by library staff–not sourced from sellers’ images that could conceivably be copyrighted as such (and which, I assure you, look nowhere near as good as the ones I can put up on the web now). Film clips, by and large, will be from Mexican films currently or recently in release and print. (of which I also own lawful copies on DVD). But Argentine films from this period are a different matter. Vintage films that match the posters are currently off the cultural radar–unavailable in what can obviously be taken as fully “lawful” copies. Yet they receive the same sort of open, probably-suspect distribution as many American films do, assuming they’re not escaping through legal coverage gaps that make them public domain. The major source I know of is: http://www.artevideo.com.ar. I’ve done business with them, and some of their films, at least, are obviously transfers (often barely watchable on a small screen) from private collection prints. But they are the only game in (or out of) town, unless you are Argentine and really know unusual market sources. One can’t be sure of their status–but the company itself has been in business for years and does reputable transactions. Why insist on more? I’d say that given my project as one that would fit the defense of Fair Use anyway, the source of the film clips hardly matters. The clips are to be considered instead in relation to the works–not the copies. The DMCA–and the courts, in the DeCSS case–indicate that FU can’t be a reason to access a work unlawfully… but that’s not the issue here.I didn’t have to break any laws to get my copies unlawfully. I’m just gonna use what I got. Defensibly.

    5. To get back on track: If we dump 110, take Fair Use as the strategy–and entertain the interpretation that FU can sometimes involve an entire work, under the right circumstances–then what else would make the case stronger? I propose several factors:
    a. Not only use of a “lawful copy,” but also retiring that copy from other loan circulation while it’s being used in the CMS. Have a problem with equitable access if you do that (as it appears UCLA’s service policy might have been concerned with)? Then get a second copy for circulation, or make it available streaming in-library/media center on an “as needed” basis. When you place a physical tape or DVD or book on “reserves,” it’s no more accessible to random, non-assigned users than that–and it’s just as unavailable for interlibrary loan.
    b. If the work does have a license-able version, consider getting it. At least (if notice of need is short) after the first semester of use. A serendipitous effect of instituting this measure as a policy would be to weed the wheat from the chaff in streaming requests received, especially if the academic dept. may have to support license funding in the future. Should a request and its use be a matter of trivial convenience, it probably won’t be insisted upon again down the road.
    c. I said “consider” a license, because: what if the license is really expensive, relatively speaking? And this is where some more leeway perhaps comes in, along with the second FU criterion: nature of the work. The USCO’s own Rob Kasunic wrote an article promoting a more careful look at sound FU assessment that can be gained from considering a work’s intended function and markets, primary and ancillary, among other issues. (Kevin references it here: http://library.duke.edu/blogs/scholcomm/2009/08/13/choosing-between-reform-and-revolution/). For films, I’d give a particular spin on the market issue (with the second factor becoming a check on the fourth): in effect, look at the nature of the audiovisual work’s primary market. If it’s educational, the field is starting to sort out with the typical range of pricing we can expect for streaming licenses (I’ve seen roughly $200/3yrs to $3-350/5yrs per program, sometimes better in bigger group deals). What if you’re dealing with a work outside of our “natural” market and outside of our expected pricing range? How about feature film performance licensing? Sure, the MPLC and others can license features for educational use, but how do they compare rate-wise with Netflix streaming where the same material can be sourced? How about all those human resources training videos? Sure, they’ve got “special educational” pricing–but is that the last word on the issue, when comparing them with true educational-market titles? (I’ll admit it is the last word, though, in the case where you purchased a title with specific licensing restrictions that would disallow any FU maneuvers.)

    There is more to be said… but perhaps I’ve been provocative enough for now. I’d welcome either concurrence or a reality check or both!

Comments are closed.