What is a Blog? A Wiki?
Despite years of debate, constructive discussion and an occasional flame war as well as scores of wikipedia edits, there remains ambiguity and disagreement on "what is a blog" and "what is a wiki." In a series talks at KMWorld, Burton Group’s Catalyst Conference, IQPC’s IntranetWeek and others over the last year, I've offered my own definition. So, here goes my attempt at a baseline set of definitions, with a bit of historical context.
The “What is a Blog” Debate Begins...
People have argued about “what is a blog” for a long time, expressing differing opinions about whether a blog is for one or many individuals, or whether comments are required, among other technical details.
I can trace the debate back to about June 2003. In this month, the definition of Blog found its way into Wikipedia, and Dave Winer debated the point from the back of the room at Jupiter's Business of Blogging conference. During a conference session led by Tony Perkins (AlwaysOn), Dave Weinberger reports on a "kerfuffle" that broke out when Winer asked "How will what you do be like a weblog?" The A-List bloggers in the room batted around ideas about when a blog stops being a blog and differing opinions seemed to focus on whether there is an editor, whether it’s a group blog, or whether it takes on a “participatory journalism” format as was the case at AlwaysOn.
MarketingVox reflects on the same session, saying "The whole question of what is a ‘blog’ degenerated into an excruciating holy war and we never got anywhere near actually settling on an answer that people could agree on."
In an entry about an expert panel at the conference, Weinberger said “Tony is misappropriating our work for his own purposes. I have nothing against his purposes — I hope AlwaysOn succeeds — but having him misuse and abuse the term ‘blog’ makes it harder for us to explain what is special about the world we have built together. It harms the growth of blogging. IMO.”
As Weinberger also reports, Perkins retreated from his claim that AlwaysOn is a blog. I think Perkins did this to call off the wolves, rather than a change of heart. Four years later, there is still a “Blog” tab at AlwaysOn.
AlwaysOn and Corante were two of the first public facing group style blogs. Given that we were at a Business Blogging conference, the concept of group blogs and the necessity for various manifestations of the technology (with or without editorial oversight, with or without comments, and so forth) ought to have been an assumption. When Paul Perry (Verizon Communications, speaking about a Traction deployment) and Rock Regan (then CIO for the State of Connecticut and a Traction customer 9 months later) took the stage for the Blogs in Enterprise IT panel, there was no challenge to their concept of group blogs behind the firewall. Nevertheless, there was (and may still be) confusion and disagreement about when or if a blog stops being a blog and becomes something else.
What about Wiki?
In true collaborative spirit, the conversation over "What is a wiki?" reads like a collaborative discussion rather than a debate. Nathan Matias (Sitepoint), Luis Suarez (Elsua), Mark Choate (CMS Watch), and a host of others reflect and build on the idea of what a wiki is.
Traditional definitions such as the one offered by wikipedia converge on the idea that "a wiki is a website that allows the visitors themselves to easily add, remove, and otherwise edit and change available content, typically without the need for registration." Over at Wiki.Org the base definition is "a piece of server software that allows users to freely create and edit Web page content using any Web browser" and makes no mention of access controls.
On the point of access controls, folks like Melanie Turek (Principal Analyst at Frost & Sullivan, blogging at Collaboration Loop) recommends that, in the enterprise context, you "make sure authorization for access is strictly controlled, and that permissions are revocable at a minute’s notice." In Best Practice and the Wikipedia Big Brain I point out how even Wikipedia, the perennially referenced archetype of a wiki, has access controls that undercut the typical perception of wiki as a place where "anyone can edit and anyone can read.".
Back to the Future
On 31 December 2006, Arrington wrote that "a blog is a conversation" and argued that the definition of blog should "exclude journals that do not allow reader comments." Arrington said that Google’s corporate blog, which does not allow comments, is not a blog.
In an apparent response the next day, Winer says a blog is the "unedited voice of a person" and notes that comments are optional.
I agree with Arrington, that a blog is a conversation, but there is no reason the conversation must be conducted in a manner that requires comments. The “public” conversation, of which blogs are a part, is conducted across blog entries, news articles, conference sessions and other outlets. In an “Enterprise” context, the conversation extends across ERPs, CRMs, collaboration applications, and document systems. Links, comments and trackbacks on blogs and wikis are all ways to connect that conversation as it happens over time.
In 2003, Winer points us to the first weblog, also the first website, built by Tim Berners-Lee at CERN. There are no comments, and no “opinion” or “advocacy” – terms often conflated with "personal voice". Berners-Lee simply used the journal style site to report factually on key events information and new sites. In my experience, status reporting, an inherently time ordered activity, is a leading use case for blogs in the enterprise.
The idea of a blog as “personal voice” limited to personal opinion and advocacy and a wiki as "group voice" or encyclopedic type site where anyone can edit sounds more like a manifestation of a particular use case for technology, not a definition of the technology itself.
Based on my experience with hundreds of deployments, that's not the whole story, or even the most important part for enterprise use. Traction supports both scenarios, to allow open or closed readership and authoring, as well as “edit only what you posted” alternatives. It also supports forward, reverse, alphabetical, reverse alphabetical, user defined and even random ordering of any cross-section of content. Accordingly, the content display and permissions form around the need or use case, rather than being bound to a particular approach as a result of a technology limitation.
Why are technologies cast in terms of their predominant use case? Maybe the approach to date has been a way to provide an easy and useful simplification that talks to the most common use cases for which there are ample examples on the public Internet. This approach helps collective understanding of the technology concepts, to a point.
In Wikis in the Enterprise: Democratizing and Directing Collaboration and Content Management (client login required) and in a March 2006 briefing, Peter O'Kelly of the Burton Group says "While Wikipedia is indeed a wiki, it’s not representative of what most collaborative workgroups are doing with wikis today."
So, to the extent that we stick to our common conception of Wiki and the encyclopedic use case that goes with that conception, the true nature of the technology and the immensely valuable applications of it will be missed.
In his predictions for Enterprise 2.0 in 2007, Rod Boothby (Innovation Creators) says: the terms "Activity Centric Blogs and Wikiflow are going to take-off." He continues, "OK. Maybe those specific terms are not going to make it. However, we collectively need a term to describe the difference between the way that blogging and wikis are used by consumers and the way that they will be used within a large organization." Others including Ward Cunningham have described a combination of wiki and blog with terms such as WikiLog or Bliki
Boothby makes a case for escaping traditional definitions, and possibly coining new words or phrases to reflect the various adaptations of blog and wiki technologies to different use cases.
For example, in the enterprise context, a product management and engineering team's control over requirements documentation written in wiki style is vital. The chronological stream of issues, questions and feedback pertaining to changing a requirement or the milestone to which it is assigned is also vital as a matter of record and as a means for arriving at and explaining the context for the decision. By way of example, read the story about ShoreBank's IT group’s use of Traction for IT project management. If a group puts a Read or Edit access control on the wiki or enable features for comments and blog-like time ordered views of new pages and comments, does it cease to be a wiki? I say no.
Swinging back to Winer. In a May 2003 blog post, he wrote one of the best explanations of what makes a weblog a weblog.
A weblog is a hierarchy of text, images, media objects and data, arranged chronologically, that can be viewed in an HTML browser.
There's a little more to say. The center of the hierarchy, in some sense, is a sequence of weblog "posts" -- explained below -- that forms the index of the weblog, that link to all the content in sequence.
And that, finally, brings me to a baseline definition for both blogs and wikis:
A system for posting, editing, and managing a collection of hypertext pages (generally pertaining to a certain topic or purpose)...
Blog: ...displayed as a set of pages in time order...
Wiki: ...displayed by page as a set of linked pages...
...and optionally including comments, tags or categories or labels, permalinks, and RSS (or other notification mechanisms) among other features.
Both "blog" and "wiki" style presentations can make pages editable by a single individual or editable by a group (where group can include the general public, people who register, or a selected group). In the enterprise context, more advanced version control, audit trail, display flexibility, search, permission controls, and IT integration hooks may also be present.
This, hopefully non-controversial, pair of definitions is a basis for explaining all forms and adaptations of blogs and wikis. An outward facing weblog run by a company is a Corporate Blog, an inward facing blog run by an organization for many individuals or groups is an Enterprise Blog, the "unedited voice of an individual" is a Personal Blog, a time ordered series of posts by a group of individuals, each writing in their own voice is a "Group Blog," an enterprise team using a wiki for unedited meeting notes and edited requirements may have a Project Wiki, and on and on.
At their core, all these permutations are single user or collaborative hypertext systems. In the enterprise context, they allow for a combination of posting notes and comments in personal voice while also providing for group editing of work product like requirements or procedures. In my synopsis of the Burton Group report on Hypertext and Compound Interactive Document models, I dig deeper into the importance of authoring in hypertext and the impact it will have on collaboration and the generation of dynamic and large documents. Simple hypertext systems like blogs, wikis and all their hybrid permuations promise to bring the power of the Intranet to the enterprise Intranet, and change, for the better, the process and format by which we communicate and we create documentation.
An Engelbart Tangent
A history of terms is not complete without reference to Doug Engelbart who created the first system for collaborative hypertext journaling in 1969.
Engelbart created what he called a "journal" to capture the voice of each person in a group as well as the work product. He used "journal" in the sense of both a time ordered diary and publication of record, as in a scientific journal. He clearly focuses on the use case of capturing external intelligence, supporting internal dialog, and producing knowledge product. He uses the word Journal in a sense that encompasses both "blog style" (time ordered) and "wiki style" (collaboration in place) communication. Engelbart’s work is core to our own roots, as a basis for Traction’s own journaling technology.
In the 1968 “Mother of All Demos” Engelbart showed how he could publish a set of content in hypertext and then display in a way that suited a given task:
His shopping list was organized hierarchically by category and he had the ability to easily expand or elide sections of the heirarchy or move items between sections. Also, documentation for the system and other similar text was displayed and manipulated as text with inline links to other places, in a mode evocative of the WWW hypertext style.
This need for a mixed-mode set of interactions within a hypertext system is likely why Boothby, Cunningham and others seek to escape “Blog” and “Wiki” constraints by combining terms or adding adjectives.
Engelbart called his system NLS (oN-Line-System) and later renamed it Augment. As the first hypertext journal, NLS was arguably the first web-log. He called it NLS.
I suppose you can call it what you want.