Over the past several months, I have written many times about "social architecture" (see here). I recently invited a group of thought leaders to join me in developing a one-day Corante symposium on the topic, and got a great response; but I also got one email (from Ross Mayfield) that said "Sure, sounds fun. What's Social Architecture?" For the sake of my co-conspirators on the event, and anyone else, I am writing this post to clarify what I think the term denotes, and set a loose collection of questions to start a dialogue about the event.

[Note: I should be formally announcing venue (Boston, provisionally) and date (early November, provisionally) in the next few days.]

Social Architecture Dynamics

The following diagram is an attempt to charcterize the interactions of three sorts of "social agents" in the blogosphere -- the human creators (or authors) of blog writings, the human readers of blog writing, and the social software applications (or "machines") that search and analyze the blogosphere based on the social "gestures" that human writers and readers leave behind. Note that human authors and readers are collapsed into one category -- they are almost identical from the viewpoint of social architecture, since they both are reading and then leaving a gestural history behind.

Authors and readers both leave social traces behind (or "gestures"), as a result of their activities. Authors point to other blogs in their posts - either by link or by name - and create ageless links like blogrolls: these represent an implicit social network relationship between the parties, not just a topical pointer, like a search engine provides. And the actions of readers (which includes all authors) create similar gestural information: explicit, shared evidence of reading like comments and bookmarks, and implicit value indications, like the frequency of return to a specific blog, or the number of comments left.

Authors and readers can make assertions about blog posts, based on various capabilities that are basic to the current Web, like HTML keywords, or relying on specific capabilities supported by various software implementations, like rating services, blogging tools (Movable Type categories, for example), or tags. Tags in particular are an area of intense interest, to a large measure as a result of the premise of a distributed, decentralized, and bottom-up approach to making sense of the exploding volume of the blogosphere. For example, we browse through the tagspace of our Deli.icio.us network of friends or all Del.isio.us users as a whole to discover web pages of possible interest: a social search mechanism.

Machines -- software applications, like Google or Technorati -- "read" the blogosphere, too, although not in the way that people do. These apps are plowing through the blogs, indexing the text, and, on the social side, algorithmically evaluating the value of various blogs or blog posts based on the social cues that readers and writers have left behind, as well as less social analysis, like keyword incidence.

The analysis that machines provide serves the general needs of readers, and specialized reader constituencies, like advertisers. We use the analysis of Google and other search tools to provide us the most relevant and most highly valued results based on our search terms. We use Technorati's tag-based analysis to help us find the most recent or most relevant and highly rated posts associated with given tags, or sets of tags. They provide, therefore, and very useful service necessary for us to make sense of the expanding blogosphere.

On The Road To Get There

In essense, what people are doing is an endless search for more stuff to read.

In a real sense, what we do on the Web can be reduced to the graph above: we are somewhere -- looking at some page, a search result, the New York Times -- and then we read what's there, we make comments, capture bookmarks, or write blog posts. These are all -- including the micro details of how we read the page -- gestures that represent, implicitly or explicitly a value judgment about the material we are looking at. Sooner or later we leave the page, perhaps following a local link: one embedded in the post, a blogroll link, or a tag. Alternatively, we might jump from the local context not using local, hard coded links, but just typing in specific terms or tags at Google or Technorati, that are related in some way to what we were reading.

Clicking on any link is a vote -- clicking on an embedded link leads to overall link counts for the target page, while clicking on a tag is an endorsement of the relevance of the tag, itself, given the context where it occurs. All these gestures are ways that we extend ourselves in the world, and thereby make it our own, and socializing it.

[Note: This is why graffitti is a creative act. What is considered defacement is in fact an innate socializing impulse -- to leave our mark on what we behold, and thereby denote our liaison with the greater world.]

But we are always moving from Somewhere to Elsewhere, and everything we do on the way is potentially a gesture that could, if it were captured, lead to a richer understanding of the relevance and value of the pages -- and by extension, the authors -- involved.

Toward an Ecology of Social Architecture

The elements of social architecture are appearing at a bewildering rate, and there are a number of very complex societal and economic issues emerging along with the explosion of social artifacts:

1/ Ethics and Economics of Social Gestures -- Who owns the traces of social architecture? If authors create public tags -- for example -- can companies accumulate them, and sell the resulting information gleaned without consideration for the authors? Do we need to tag all tags with creative commons-like agreements? The same considerations arise relative to other public gesture spaces -- comments, links, and so on.

2/ Open Architecture -- How open is enough? How should various sorts of gestures be implemented: for example, there has been a lot of discussion recently about making tags more open (see here). If a few major companies (Google, Yahoo, and Microsoft, for example) come out with competitive, closed Technorati-like solutions, we could rapidly find ourselves in a fragmented world, with three non-interoperating, partially overlapping tagspaces. It is clearly not in the public interest to go down this path, like what has happened in the instant messaging world.

3/ Privacy and Identity -- What measures for privacy should be contemplated? Is there some way to make gestures only sharable with known others? What does anonymity mean in a socialized Web? Is it possible at all? Are we defined as the sum of our gestures? Will we be declaring our willingness to be advertised to by a tag-based profile? What is the aggregate complement of the history of our meandering around the Web, writing, comments, and tagging?

4/ Better Social Elements -- Blogrolls and other explicit links are very coarse-grained mechanisms to represent social relationships between people, but explicit mechanisms to denote degrees or depth of relationships have not emerged. Is there a solution here, buried in the countless gestures we make in the world, including closed spaces like your email and instant messaging, or explicit social networks?

5/ The Personal and Global 100 -- The recent spate of criticism about the various top 100 lists suggests that new ways of analyzing social architecture are needed so that the oft-quoted notion -- "everyone can have their own top 100" -- might be more than just an handwave. How do can we manage our own lists, really? Explicit blogrolls (embodied in blog readers, on on our blogs) is not at all the same as determining who are the most relevant top 100 writers on a topic of interest, based on personal preferences and inclinations.

Close

The continued growth of the Blogosphere will make its social architecture even more of an global asset that it has already proven to be. We will continue to witness enormous technological innovation, with dozens of new Flickrs, Technoratis, and De.licio.uses appearing in the next year. As more writing (and other media, like audio, video, and photographs) is generated on an ever widening range of topics, more and more machine-generated analysis of human social gestures, and the gestures themselves, will play an increasinglt important role in making sense of the Web. Without these techniques, the explosion of the Blogosphere will overwhelm our traditional information-based approaches.

The criticality of these activities will cause friction on technological, societal, and economic levels, and as so those of us who are most interested and involved in these discussions may have a significant impact on the future direction of the socialized Web. The planned Symposium is intended to bring together thought leaders, practitioners, and entrepreneurs in the arena and to explore the various threads making up the discussion about social architecture.