September 13, 2004

Blogs and Scandal

The CBS presentation of badly-forged memos as evidence to smear President Bush has been pulled in a lot of directions. On the part of the libertarian and Republican-aligned blogs, as well as more than a few sane Democrat-aligned blogs, the issue has been the veracity of the documents: if the documents are unreliable, the charges are unreliable and must be discarded. This is leading many of these blogs to question the reliability of all mainstream media reporting. CBS, and the far-Left bloggers, seem to want to ignore the evidence and get right to the discussion of how much of a liar George Bush is - never mind about "evidence."

I reject (as would any reasonable person) the idea that accusations are sufficient proof of wrongdoing, and evidence is immaterial. Facts are requisite to truth, if not to Truth (which is really just another word for religion). While I have quite a bit of interest in the media reliability angle (indeed, my previous posts on this issue have mostly centered on that), I'm actually more struck by the undiscussed ramification: the effect of blogging on creating and sustaining a scandal.

Blogs are new, coming into their own only after 9/11, and this is the first event with national implications in which blogs have played a key role. But why? After all, the precursors of blogs existed during, say, the Clinton impeachment. USENET, personal web pages and email all fulfilled the functions of collaboration (USENET), communication (email) and rich content provision (personal web pages), and certainly a lot of people were online by the late 1990s. But there are some key differences, some ways in which blogs enhance communications and opinion formation well beyond what earlier technologies could provide.

USENET provides a meeting place, ostensibly divided by topic, for a large number of people. This architecture has two problems (even ignoring the fact that USENET is only text-based) with forming opinion and filtering good information out of the stream, both of which are overcome by blogs: noise levels and specialization. USENET has so many people talking in one place that the useful information (signal) is drowned in useless information (noise). The high volume of useless information hides the useful information one might be trying to find, and filtering and searching mechanisms are not particularly well-developed. In addition, any given newsgroup addresses only a narrow topic, at least in theory, so it is unusual to find a typographic expert, for example, reading a newsgroup dedicated to discussing military memorandum formats, and vice versa.

Blogs are not structured around topics, by and large, but around personalities. The fact that Rand Simberg blogs primarily about space access doesn't mean that he won't indulge his other interests. Because of this, there is an immense cross-disciplinary polinization of ideas on blogs that doesn't occur in other formats. Experts in one field often read blogs written by experts in another, and find items that intrigue them. This sparks off a chain of discovery and allows far more broad and deep information to come together than, say, USENET does.

Further, blogs have both active and passive filtering mechanisms built into them. A blogger who proves untrustworthy loses his audience. A blogger who doesn't provide useful information does not gain, or cannot keep, his audience. InstaPundit and others act as a filter by pointing to interesting information on other blogs. Most blogs maintain blogrolls, directing users to related information. Bloggers link to each other (as well as to non-blog sources), and trackbacks further enhance the connectivity on related topics. The comment mechanisms on many blogs further enhance the discussion. Finally, the use of blogrolls tends to result in the formation of communities with related interests or worldviews, which enhances the information flow. In each of these mechanisms, the key point is that the blogger has to have something useful to say, or he gets filtered out of each of these mechanisms over time.

But USENET has another problem as well: it's text-based. Being unable to easily provide rich content, it is difficult to make a point which has visual elements. Consider trying to do this on USENET. Personal web pages, of course, can and still do provide this kind of rich content, but it is provided in an isolated medium. Yes, Google provides a way of finding this information, but it's not self-selecting in the way that blog links are. So while USENET provides great connectivity of information, but no filtering and no rich content, personal web pages provide very highly-filtered rich content, with no connectivity.

Email provides directed, highly-filtered connectivity, but is non-public, so only the sender and recipients ever see the content. And, again, this content is not easily made rich: it's primarily text-based. Trackbacks and links provide the openly-available two-way discussion path that email lacks. (And mailing lists, while more publically-available, frequently suffer from most of the drawbacks of USENET.)

Blogs, by providing all of these mechanisms, can do something that until now only television, newspapers and magazines could do: blogs provide rich content, publicly available, filtered and analyzed and readily found. It is this that makes blogs such a threat to mainstream media: blogs can do everything journalists can, but generally blogs bring a higher level of subject matter expertise to the topics they cover than can mainstream journalists.

Given this, what impact are blogs having, and will they have in the future, on political scandals and for that matter on campaigns in general? It is pretty clear that blogs are at the forefront of the CBS document forgeries. The history of the beginning of that story is largely told at the New York Post. Basically, after "Buckhead" at Free Republic raised a question, bloggers took off with it, calling in experts and making their own tests. This signals something critical: blogs are capable of killing a scandal if the sources are not absolutely accurate. There is simply too much expertise available when you start playing six degrees of separation for a scandal to get away with unnamed sources and innuendo any more.

This will make negative campaigning much more difficult in the future. The filtering mechanisms on blogs will drive partisans naturally into blogs with similar affinities, and these groups will eagerly pounce on anything provided by the other side. If the negative charges are factually false, they will be disproved in short order. On the other hand, if the charges are true, the scandal could grow more quickly than it has in the past. Remember, it was Matt Drudge who broke the Lewinsky scandal, and that was what amounts to a personal web page at the time, without a network of blogs and the experts they can bring to bear.

UPDATE: Andrew Olmsted is following another line of reasoning: using blog-like decentralized methods for intelligence analysis. It's certainly worth thinking about.

UPDATE (9/14): Read this post by Bill Quick at Daily Pundit, which has a well-reasoned take on the incentives of the mainstream media and the blogs. (hat tip: Kevin Murphy)

Posted by Jeff at September 13, 2004 12:26 PM | Link Cosmos

Unfortunately, blogs are not about truth but about choice as I wrote in a recent post

leading not to consensus but to faction. The filter function you ascribe to Glenn Reynolds is quickly extinguished in favor of a switch function. There are several such almost completely discrete switches in the blogosphere e.g. Eschaton.

Posted by: Dave Schuler on September 13, 2004 01:49 PM

I thought of including Eschaton in the article alongside InstaPundit. While I despise Atrios on many levels, it is true that Eschaton serves a similar purpose, and there are other blogs from various political viewpoints which fit into similar "gatekeeper" roles.

However, I don't buy the Glittering Eye's analysis of factionalism. While it is true that bloggers have viewpoints, it is not true that they simply break down into red/blue. Glenn Reynolds, for example, is hard to pigeonhole into the same category as LaShawn Barber. While factions tend to seek their own, and there's a strong echo chamber effect at times, it's also true that in the center and at the particularly large blogs there is a lot of crossover.

To some extent, this doesn't matter, because we are living in a world where one side argues that facts and Truth are disconnected, and another argues that Truth isn't valid if established facts contradict it. This is hardly limited to blogs. Dan Rather, for example, was busy last week arguing that it doesn't matter if the memos are forged, because the memos raise questions that need to be answered. Ummm, ok. So if I type up a doc that says Dan Rather is a transvestite, does that raise questions that need to be answered? Hardly.

Until this fundamental schism in views of the nature of reality is healed, there will be issues not only in blogs, but in all media, where people are divided from each other at a fundamental level. However, the vast majority of people believe their eyes, and respond to logic over pseudo-logic.

As a result of all this, there is a definite filtering process that occurs. Even though there is not always an agreement in interpretation, the best evidence does become available.

Posted by: Jeff on September 13, 2004 04:04 PM

I agree that the red/blue is an over-simplification: I wrote as much in my post. But I think the middle (of which I'm a member) is actually rather small and the fervent partisans far outnumber them. Look at the numbers. The clearly partisan (although not necessarily doctrinally pure) sites get the hits. Moderate sites just don't get the number of hits.

I read both Glenn Reynold's blog and Eschaton (it's right at the level of my tolerance which stands to reason since I'm slightly right of center). I can't face either Daily Kos or Free Republic.

After an initial rough-sort by faction I don't think a lot of filtering goes on. What I see is honing of arguments. Dropping the unessential weak bits. Fine-tuning. As I wrote in my post: they don't even agree on each other's facts.

Posted by: Dave Schuler on September 13, 2004 05:14 PM

Great post. I will link to it in my "Blogs: mandatory marketing tools" article in

Your comments, in IE 6, the lines are squeezed together, making them virtually unreadable. Is it my browser's fault? The effect is maddening. Too bad. Hope you can fix this problem. Have someone else check it on IE 6 browser.

Posted by: Steven Streight on September 18, 2004 01:50 AM
Post a comment