Bill McCoy, the General Manager of ePublishing at Adobe, wrote an influential blog posting that catapulted the term “social DRM” into wide use. He said

“For eBooks, I really like the ‘social DRM’ approach of The Pragmatic Programmers, who ‘stamp’ PDF eBooks with a ‘For the Exclusive Use of …’ and the name of the purchaser.”

Traditional Digital Rights Management (DRM) requires implementing technological obstacles that prevent the purchaser of a digital object from copying, displaying, and accessing the object except in limited ways. These obstacles can cause endless aggravation to the consumer. For example a Kindle format e-book cannot be read on an iPhone or iPod touch even though the hardware sales of the latter Apple devices dwarf the sales of the Amazon device. (Update: Kindle e-books are now readable on the iPhone and iPod Touch but are still unreadable on PCs and Macs.)

I do not know if McCoy invented the term “social DRM”, but his blog post certainly helped to popularize the term. The article facilitated an important dialogue about e-book security, and this post is not meant to be discourteous. However at this stage of the conversation I suggest that the term “social DRM” should be replaced by “digital watermark” or simply “watermark”.

Physical watermarks are well known for paper stationery and the idea has been expanded to apply to digital pictures, music, video and now e-books. For additional background there is a useful Wikipedia entry on the term digital watermarking. Here are some reasons for the switch:

  • The term “social DRM” is confusing because it does not really refer to an “access control” technology. Instead it refers to a technology for “tracking” and “display”.
  • The term “social DRM” verges on the oxymoronic since “social” is nearly the opposite of “DRM” in the context of e-book “security”.
  • Agitating against DRM while simultaneously being open to “social DRM” is terminologically confusing.
  • The existing term “digital watermark” or “watermark” can typically be substituted for “social DRM” and the meaning can be preserved.
  • There is a rich preexisting vocabulary for watermark description and classification such as “perceptible” and “imperceptible”; “fragile”, “semi-fragile”, and “robust”.

I was originally planning to provide a cluster of suggestions to replace or supplement the term “social DRM” such as: customize, tailor, imprint, stamp, inscribe, personalize, endorse, bookplate, dedication page, insert page, fingerprint, hash, and signature. But the easiest approach appears to be adopting the expression watermark.

Here is an example of how to use the term “watermark”: I wish Amazon would use a standard open format without DRM for its Kindle e-books. If Amazon deems some security measure necessary then why not try watermarks. With watermarks and an open convertible format I could still read my Kindle format e-book on my cell phone, computer, or dedicated e-book hardware (with conversion if needed). The catalog of e-books for the Kindle is extensive with Amazon claiming “more than 190,000 books available, including more than 109 of 112 current New York Times Best Sellers.” Please do not lock up this catalog by coercively tethering e-books to the Kindle hardware using DRM.

The landscape image above is a fragment of a picture in the Flickr photostream of Shiny Things. I have superimposed a watermark image of the word watermark. Some rights reserved.


  1. Good points.

    I have been consistently (I hope…) trying to differentiate Digital Rights Management (DRM) — as expression and management of rights — from Technical Protection Measures (TPM), without any success. So I’m afraid ‘DRM’ is now too deeply associated with TPMs to be of any avail, even qualified with the positive adjective ‘social’ — which will probably be considered as positive only as long as the 2.0 buzz works…

    Now, watermarking is not the only solution to fulfill the kind of rights indications: the stamp Bill McCoy considers is in fact a kind of ‘ex-libris’ (or ‘bookplate’, according to the American Heritage Dictionary).

    And one can also hide a “digital watermark” using ‘steganography’…

  2. I was actually unaware of what popularized the term “social DRM”, so I came to my own conclusions as to why the label was in use. Many people established within the book publishing industry have a deep-seated belief that many of their customers would be criminals given the chance. (The other, less-spoken-of side, is that some view book lending as theft from the company.) The term DRM has become associated with the technique of preventing their customers from doing that.

    We, as customers, do not wish to accept the arbitrary and very limiting restrictions being pushed by the publishers. Our response to their fear can’t be a simple “no, don’t do that”, however. We need an idea we can market, and social DRM is the marketing label. It maintains the association with a protective technology (because it is), thus getting the foot in the door for discussing the merits.

    From a consumer perspective, it is watermarking. However, labels implying a weak, non-restrictive approach aren’t going to change the industry. Teleread feels like a community of e-book enthusiasts with a target audience of the book publishers, and I think it should continue to fill that niche.

  3. Thank you for your responses. One positive development occurred earlier this year when a major publisher, Random House, announced “we will no longer require that our retail partners use DRM when selling audiobooks via digital download”. Their decision was based on an experiment using watermarks instead of DRM on audiobooks sold through eMusic. Here is an excerpt from a letter sent from Random House to its partners:

    For tracking purposes, we watermarked all of the eMusic files and then hired a piracy watchdog service to monitor and report back to us if any of our titles appeared on the major filesharing networks. We tracked a mix of popular titles, including some that were not available through eMusic. Because piracy is already a fact of life in the digital world, what we were interested in finding out was not whether piracy exists, but rather whether there is any correlation between DRM-free distribution and an increased incidence of piracy.

    The results: we have not yet found a single instance of the eMusic watermarked titles being distributed illegally. We did find many copies of audiobook files available for free, but they did not originate from the eMusic test, but rather from copied CDs or from files whose DRM was hacked. It is worth noting that these results are entirely consistent with what the music industry has found in the last six months. After conducting their own tests with Amazon, and others, the major labels have reached the conclusion that MP3 distribution does not in itself lead to increased piracy, they are now moving their entire catalogs to this approach.

    The text of the Random House letter is here in PDF format. BoingBoing reported on the story here.

    Alain Pierrot: The example given by Bill McCoy is similar to a “bookplate” and “bookplate” was one of the terms included in the list of suggestions given near the end of my article. Yet I think that the umbrella term watermark is appropriate. A watermark on a digital picture is usually deliberately perceptible, and it gives ownership or licensing information. This is somewhat similar to a bookplate. Watermarks can contain creator, author, licensing, transaction and other data. They can be perceptible or imperceptible.

    Logan Kennelly: It is plausible that some publishers would be more comfortable with the term “social DRM” because it contains the embedded acronym DRM and that provides consolation to them. Yet there are instances of knowledgeable publishers such as Random House cited above that are using watermarks instead of DRM and are also properly using the term watermark.

    Lastly I should note that there is opposition to watermarks too. An article at Wired ”DRM Is Dead, But Watermarks Rise From Its Ashes” mentions privacy and tracking concerns.

  4. I agree with you that watermark seems the best umbrella term.

    “They can be perceptible or imperceptible.”

    Or a combination of both: a visible mark of ownership, and hidden details or link for the licensing information when the text is displayed as an image, or in an appendix to the file, when the text is reflowable (just as the copyright and colophon are printed on specific pages).

    As for the privacy and tracking issues, in the terms mentionned in Wired, they do not seem to be very different from the statistics and tracking possibility social networks impose on their users (or credit card payments). Regulation, third party management of identity and self discipline could provide an acceptable mix.

    And bibliophiles do appreciate numbered editions…

The TeleRead community values your civil and thoughtful comments. We use a cache, so expect a delay. Problems? E-mail