Coptic Quarter, Cairo, 2013

Collaborative Knowledge Production

Collaborative production of any knowledge artefact could be called ‘Collaborative Knowledge Production’ or CKP. Understanding the right strategy for CKP is reliant on an understanding of the content to be produced, the greater context, and the resources available. From these, a process can be designed.

When Paper Fails

When all the activities and practices that we now call “publishing” exist in a networked environment, something radical changes – affecting creators, content, ownership, and trust. That might sound like the end of publishing as it is now, but it also sounds like the beginning of something exciting. And of course, it is argued that this future is already here, but, to paraphrase William Gibson, perhaps a little unevenly distributed.
Responses to these new challenges are already partly in motion inside the industry (e.g., the work Safari Books is doing with bibliographies connected to their ‘cloud library’) and outside (too many to mention but one example is the very interesting Open Oil book project) and as we move forward I firmly believe these futures will become increasingly present and their economics more mature.
Where does that leave the publishing business? Well, it might be better to ask yourself, ‘where does that leave business?’ Forget capital P, ‘Publishing’, for a moment. What are the skills necessary to survive here, what will you be doing, and what is the economy?
People are going to continue to require services that deliver and produce information. Finding ways to create information and finding someone to pay for it is the heart of the matter. That is not going to change anytime soon. The need for information won’t change, but how information is produced and delivered will change. In fact, I believe the demand for content is going to rise (it is already rising rapidly), and the demand will increasingly be for more individualised, customised content and it will need to be delivered faster, much faster than today.
So, what would the world look like when the walls that contain the publishing industry fail and spill their innards onto the web? Or to see the same question through the lens of Eric Raymond, what is the essential difference between the cathedral and the bazaar?
Let’s quickly look at the environment of this particular kind of “bazaar” – the web – for a few clues. The most important issue at play is that the web always appears to find a way to route around arbitrary constraints. People, processes, and information route their way around unnecessary blockages looking for and finding the most efficient and least resistant paths. So what would happen if publishing was immersed in that environment? What are the containers, the constructs of the publishing industry, that would be routed around and may break down and fail? Here, I want to explore four main issues pertinent to this discussion – Books, Ownership, Authors and Authority.
Books: In this environment content containers, like books, lose the definitiveness of their boundaries. What is separating an EPUB, which is made of HTML, from the web? As Hugh McGuire has said many times, this differentiation is arbitrary.  Arbitrary containers like zip files (EPUBs, which we might call portable websites) might assist in the transport of curated content, but, in the long run, they will be under a lot of pressure to remain contained and will increasingly become unbound.
Ownership: Another “container” that will come under increased pressure from these forces. If the mere fact of copyright ownership protected their content then publishers wouldn’t be looking to DRM (digital rights management). We know that the way in which book content is owned and licensed will change dramatically. Protecting ownership will increasingly become an impediment to business as it decreases the utility of information (something that O’Reilly was smart to recognise early on and perhaps reflects Samuel Johnsons’ famous quip regarding his writings that “I have been paid for them, and have no right to enquire about them”).
Authors: Also an arbitrary construct, in as much as both the realities of book authorship, and its production, are more collaborative and iterative than commonly perceived. This is another dimension that will be radically transformed by the new collaborative possibilities opened up by digital technologies. Indeed, perhaps the cultural construct of isolated genius will remain only as a brand. In reality, people are less and less isolated on the web and there is more genius out there than you can imagine. I would argue here that the concept of the author will also become more “porous”. We will be looking at a world of “networked genius” rather than the traditional standalone kind. This has been discussed in fascinating detail by Martha Woodmansee and Jack Stillinger.
Authority: The web doesn’t seem to allow anyone to merely assert ‘authority’ – such notions are subject to the ebb and flow of public web “opinion”. Publishing as an authority will certainly come under immense pressure and one possibility is the move to “distributed opinion networks” built and mediated by technologists. We have already seen this on the web, and the question of authority in these networks is well articulated in commentary surrounding Wikipedia vs Encyclopedia Britannica, for example.
When paper fails, it affects creators, content, ownership, and trust in radically transformative ways. Production processes change, content is both harvested and produced, contributors are corralled and facilitated, books become individualised ‘outcomes’. We might say the engine this “new” publishing economy revolves around includes two critical factors:
(1) content production, harvesting and curation for increasingly individualised contexts; and (2) speed of delivery.

Helping people to get what they want, their outcome, is going to be the bread and butter of this economy. This is a move from selling the artifact, to developing and selling a service, or towards providing services to help others produce and distribute content. The faster you can deliver it the more competitive you will become. People, businesses, governments, schools, etc., are all going to be very happy to pay for that.

*Please note the Open Oil project is a small project and not provided here to illustrate this kind of model at scale but to point at a very interesting and important emerging model.

Many thanks to David Berry for improving this post. This post and all others by Adam Hyde are CC-BY-SA

Originally posted on 13 Feb 2013 on O’Reillys Tools of Change site: http://toc.oreilly.com/2013/02/when-paper-fails.html

Visualising Book Production

Data visualisation is one of the hot topics of the last year or two. So what does this offer publishing and book production?

Open data activists, in particular, have been lobbying governments for access to databases which they use to create infographics and visualisations for campaigns. It’s not a new science, of course, it was here long before the internet (for some background on contemporary practice see the wonderful books by Edward Tufte), but the net is made of data and is a good mechanism for transporting it. The internet is a good medium for scraping and re-presenting data in more palatable forms.
Prior to the net, visualising book production was tricky since the information either wasn’t recorded or was embedded in ungainly ‘record changes’ data in word processing files. So the few examples were preceded with some forensic historical research. One wonderful pre-Word example is The Preservation of Favoured Traces by processing inventor Ben Fry. He visualises the production of The Origin of Species over time to illustrate Darwin’s evolving thesis. Rather appropriately, his is a nice visualisation of the evolution of a book.

This is where the web steps in and changes the game. Online book production platforms enable you to store and retrieve historical data and use it as you like. You can record and access information quite easily. Information such as who is actively working on a book and when, how much they changed, what they changed, who else was online, word counts over time, is all available. If we could access and process this information in chunks, it could perhaps help us to make books better.

Just to show you where this might go, the following are simple prototypes Juan Barquero and I put together using real data from the online book production platform Booktype. The nice thing about Booktype is that it already has all this data recorded in the history for each book, so we could write a visualisation application and then look back over the history of many books. So we made a simple API (Application Programming Interface) and Juan put together a few demos using the JavaScript visualisation library D3JS. The following are some images taken from these trials:

channel1 Contributions to chapters (x) over time (y)

circles5Proportional contributions per contributor

These are just a few simple and raw examples that are of course very far away from being production-ready. However, they serve as interesting prototypes for thinking about how this might look and what we might learn from such techniques in a production environment.

Originally published on O’Reilly, 4 Feb 2013 http://toc.oreilly.com/2013/02/visualizing-book-production.html

Zero to Book in 3 Days

One of the burdens of book creation is the enormous time periods involved. Ask any publisher for a timeline for producing a book and you will be surprised if you get back an answer this side of 12 months. In this day however that timeline is looking increasingly glacial. How can we accelerate book production? How fast could it get? How does three days sound? Enter Book Sprints.sprint_booksThese three books were created in a three-day Book Sprint and output to paper, MOBI and EPUB on the third day.

Book Sprints bring together 4-12 people to work in an intensely collaborative way, going from zero to book in 3-5 days. There is no pre-production and the group is guided by a facilitator from zero-to-published book in the time available. The books produced are made available immediately at the end of the sprint in print (using print-on-demand services) and ebook formats. Books Sprints produce great books and they are a great learning environment and team-building process.

This kind of spectacular efficiency can only occur because of intense collaboration, facilitation and synchronous shared production environments. Forget mailing MS Word files around and recording changes. This is a different process entirely. Think contributors and facilitators, not authors and editors.

There are five main parts of a Book Sprint (thanks to Dr D. Berry and M. Dieter for articulating the following so succinctly):

  • Concept Mapping: development of themes, concepts, ideas, developing ownership, and so on.
  • Structuring: creating chapter headings, dividing the work, scoping the book (in Booktype, for example).
  • Writing: distributing sections/chapters, writing and discussion, but mostly writing (into Booktype, for example).
  • Composition: iterative process of re-structure, checking, discussing, copy editing, and proofing.
  • Publication

The emphasis is on ‘here and now’ production and the facilitator’s role is to manage interpersonal dynamics and production requirements through these phases (illustration and creation of other content types can take place along this timeline and following similar phases).

Since founding the Book Sprints method four years ago, I have refined the methodology greatly and facilitated more than 50 Book Sprints – each wildly different from the other. There have been sprints about software, activism, oil contract transparency, collaboration, workspaces, marketing, training materials, open spending data, notation systems, Internet security, making fonts, OER, art theory and many other topics.

People love participating in Book Sprints, partly because at the end of a fixed time they have been part of something special – making a book – but they are also amazed at the quality of the books made and proud of their achievement. The Book Sprints process releases them from the extended timelines (and burdens of guilt) required to produce single-authored works.

Here are some interestingwrite-upss that provide more detail on the process:

http://techblog.safaribooksonline.com/2012/12/13/0-to-book-in-3-days/

http://google-opensource.blogspot.com/2013/01/google-document-sprint-2012-3-more.html

http://www.booksprints.net/2012/09/everything-you-wanted-to-know/

Originally published on O’Reilly, 28 Jan 2013 http://toc.oreilly.com/2013/01/zero-to-book-in-three-days.html

Fork the Book

As one of the first mass-produced industrial artifacts, the book remains a solid signifier of stability. That aura is pretty strong and makes it difficult to think about books as being anything other than static. It appears to be part of their DNA.

While we continue to refer to ebooks as ‘books,’ this genetic legacy seems to be inherited along with the name. Books are stable. Websites are not. That’s the lore.

However, this distinction is more than a little arbitrary and it’s interesting to consider what advantage there might to destabilising books. What would we gain and lose if books became unstable and staticness was viewed as a property of a particular era intimately linked to the days of the printing press and paper books?

The technology is right here to make books unstable. EPUBs, for example, are easy to re-edit and ‘re-publish’. In early 2012, I made a demo with Juan Barquero to demonstrate this reality (the ebook itself is no longer online but Phil Schatz took up the idea and you can see something similar here.

The demo uses EPUB as a storage mechanism and the editor enables you to edit the ebook directly. You can edit the pages in the ebook itself. While this is in itself interesting, it doesn’t offer much more than what you can do with word processing now. It is true that the book can be generated as an EPUB on the fly, but this is optimisation of current processes, it in itself doesn’t offer anything new. You could do the same with MS Word files, it would just take a little longer.

What is more interesting is that the demo uses GIT as a back end (Adam Witwer and the O’Reilly crew have also been working on using GIT with Atlas). GIT is a technology programmers use to collaborate on code. It allows programmers to copy (fork) code, work on it and then re-combine (merge) the changes with the work of others. The git demo opened the door to cloning, editing, forking and merging the book into infinite versions. Clicking the download button gives you a new version of the book in its current state, immediately, but there could be 5 different versions of the same book in production. Or 50, or 100. Creating a new version is simply a matter of clicking a button and the book will be forked to its own repository. Further, it is possible that each fork can inherit the improvements in other forks by merging the content of several versions into one.

What does this point to? It points towards a particular kind of book instability. However, while it is technically feasible, the question really is – does this kind of book instability have any value? I believe it does. I would go as far to say that this kind of book, which I would prefer to call a ‘forkable’ book, is more valuable in many situations than stable books.

As a very small case in point, let’s have a quick peak at the life of a forkable book. The Cryptoparty Handbook.

The CryptoParty Handbook was created in Berlin during a 3 day Book Sprint last October. It consists of over 440 pages of information for those wanting to be safe online. The speed with  which the book was created was as fast because it reused content from two existing books (licensed under Creative Commons): How to Bypass Internet Censorship and Basic Internet Security. Both of these publications were also created with Book Sprints the previous year.

Creating The CryptoParty Handbook was simply a matter of forking (copying) each of the other books and merging them into a new container. Easily done. The team, under my facilitation, then structured a new table of contents, removed chapters that were not necessary, identified content that needed to be created, and then started writing and illustrating. It didn’t take them much time to produce a book which was immensely useful for their audience and which could also be easily remixed and translated.

The handbook has now been forked quite a bit. The first version hit 30,000 downloads in the first few weeks. There have also been some interesting forks, including one by the Liberation Tech list hosted by Stanford University from where it has been forked again another 50 or 60 times. The book is now being used by CryptoParties all over the world to train people in small informal workshops. As if this isn’t enough forking, the foundation books, from which the Crytoparty handbook was based, have also been forked into many many other books.

This is just one example highlighting what can happen when we allow books to be forkable. They become extremely powerful bodies of content that can be re-purposed infinitely for whatever context is necessary. That’s pretty valuable.

As a final note. This is not some kind of hippie content love-in. There are economies in action here. It takes skill to curate and corral content, shape it, get it to meet the needs of a specific audience, and find experts to fill in the gaps. It takes experts in facilitation and curation — and that points towards new sets of skills required in the emerging publishing practice.

Originally posted on 22 Jan 2013 on O’Reillys Tools of Change site: http://toc.oreilly.com/2013/01/forking-the-book.html