The copyright mistake at Concordia – a poetry centre scanned several books and posted them on the Internet without permission – has attracted considerable attention in the press and social media. Kate Taylor wrote a Globe and Mail column placing much of the blame at the feet of fair dealing, while I responded with a post yesterday that noted that no one claimed that the posting of the full-text books was permissible and that Concordia was an ill-advised target for fair dealing criticism given that it has a copyright collective licence with Copibec that compensates for copying on campus.
While the focus of the Taylor column and my response was on fair dealing and collective licensing, the Taylor column also included several references to the use of a scanner to digitize books. In particular, it concludes by stating that “Ottawa needs to plug that education loophole good before somebody tries to drive a $10,000 book scanner right through it.”
The inclusion of education as a fair dealing purpose is clearly not a loophole – fair dealing is a user’s right and adding education as a purpose in 2012 only marginally expanded it – but the inference that scanning books is an activity to be stopped requires a response. As I wrote last year, Canada has unveiled a modest digitization goal that calls for the digitization of 90 per cent of all published heritage dating from before 1917 along with 50 per cent of all monographs published before 1940 within ten years. It also hopes to cover all scientific journals published by Canadian universities before 2000, selected sound recordings, and all historical maps. The Canadian targets pale by comparison with other countries. For example, the Netherlands plans to digitize 90 per cent of all books published in that country by 2018 along with many newspapers and magazines that pre-date 1940.
Digitization initiatives often raise legal fears, but the experience in other countries demonstrate that the legal challenges are frequently overstated. For example, U.S. courts have ruled that massive digitization programs such as those undertaken by Google qualify as fair use. The Google Book Search initiative rightly distinguishes between the act of digitization or scanning (a data mining and non-consumptive use of the work that qualifies as fair use) and the making available of those works to the public. This means that millions of books can be freely digitized without fear of copyright infringement, though full access to the text is limited to public domain works (where the copyright has expired) and licensed materials where the copyright owner has granted permission. Partial access (often referred to as snippets) may be granted consistent with fair use.
Canadian law features fair dealing rather than fair use, but a similar approach could be adopted. It goes without saying that all public domain works – which could reasonably be estimated to include anything published before 1940 – could be digitized and made immediately accessible in full text. This appears to be part of what Concordia had in mind with respect to its poetry digitization plans. In addition to the materials that are almost certainly in the public domain, educational institutions, archivists, and the government could launch a crowdsourcing initiative where Canadians help identify additional public domain works of authors who died more than 50 years ago. This would include many books published in the 1940s, ’50s, and ’60s. Those works could also be digitized and made accessible in full text without the need for permission or further licensing payments. Such efforts should be welcomed by supporters of Canadian culture as a way to breath new life into long-forgotten Canadian works and as a means of raising new awareness of Canadian history.
The situation for in-copyright works is more complicated and requires separate legal analysis for digitization of the works and the manner in which they are made available. The first part – digitizing or scanning the works by educational or research institutions – likely qualifies as fair dealing as it meets the research purpose (and therefore the 2012 reforms to fair dealing have little impact on the issue) and the six-factor fair dealing analysis would likely treat the act of digitizing for full text search purposes as more fair. It may involve a large amount of copying, but the other factors – including alternatives and effect on the work – would tend toward fairness. Enabling full-text searching for Canadian books would increase their “discoverability” without negatively affecting their commercial value. Indeed, it would likely increase their value.
The more challenging legal issue, however, would involve the public availability of the text of scanned or digitized books. Much like the Google Book Search approach, fair dealing and/or de minimis would likely permit a snippet of the work be made available without the need for further permission. For full text, authors would be entitled to specify how, if at all, their works should be accessible. This could involve licensing arrangements or strict limits on the amount of the scanned text that is publicly available. The digitization would create new opportunities for publishers and authors, while opening the door to heightened interest in Canadian works.
In the Concordia case, full-text of poetry books were briefly made available without permission, which would not qualify as fair dealing. However, the act of digitizing works already purchased by Canadian libraries for the purposes of full-text search is consistent with current law and if combined with the availability of full-text of public domain works and snippets of other books (or full-text with permission), would provide an enormous boost to Canadian culture where discoverability has emerged as one of the biggest barriers to commercial and cultural success.