Conversational intensity: Claiming a term

Here goes nothing:

I believe that I coined the term “conversational intensity” as applied to blogs, when I first used it in “Perspective: Investigating the Biblioblogosphere“–which was published on August 10, 2005 as part of the September 2008 Cites & Insights. (That essay, by the way, appears to be the second-most-frequently-viewed essay in the history of C&I, with just under 20,000 full-issue downloads and single-article pageviews to date.)

No, I’m not claiming credit for the term itself. It’s been used in social sciences for decades, but it’s related to things like the loudness of conversations within a party and the like.

Conversational intensity, if you’re new to this game, is a relatively simple metric for a blog: Over a period of time (I typically use a three-month period), it is:

Number of comments divided by number of posts.

That’s it. The higher the number, the greater the conversational intensity.

U’r doin’ it wrong

The earliest use of the term related to blogs that I can find, other than my own, is from Mitch Ratcliffe’s Ratiuonal Rants blog at ZDNet, as the title of a February 5, 2006 post. That’s half a year later (although I’m not suggesting Ratcliffe picked it up from me–that would assume that Big-Deal Bloggers pay attention to anyone else other than other Big-Deal Bloggers, which seems unlikely)–and his usage is, in my humble opinion, wrong wrong wrong.

To wit, he’s using this term as a synonym for Conversational Index, a metric proposed by Stowe Boyd. The problem with this metric is that it’s inverted (and includes something that I regard as peripheral to conversation, namely the number of trackbacks, which are frequently just spam). CI is

Number of posts divided by number of (comments plus trackbacks)

That’s counterintuitive: Lower is better. How many metrics do you know (other than error metrics) where lower is better, higher is worse?

Otherwise, if you take out the trackbacks, it’s exactly the same metric, but inverted: A Conversational Index of 0.5 is the same as a conversational intensity of 2.0.

So what?

So not much of anything. I explicitly disclaim any intention of trademarking the term, claiming proprietary rights, or hounding anyone who uses it without giving me credit. Period. It’s just a bit of recent history.

Now, if someone was silly enough to write a Wikipedia entry on conversational intensity, I would expect credit (and the September 2005 C&I would be citable as a source). But why would anyone do that? (And no, please Gaia, don’t go the next step and try to create such an entry for the unnotable Walt Crawford.)

If anyone has an earlier citation of this term used in relationship to blogging, let me know.

4 Responses to “Conversational intensity: Claiming a term”

  1. Thanks for bringing the question of intensity up, again. I didn’t use the term as a synonym for Stowe Boyd’s conversational index, but in contrast to his usage, stressing that the ratio you cite is too easily gamed and not particularly useful.

    Conversational Intensity, which is a term I developed while working on BuzzLogic’s influence algorithm, reaches across blogs rather than focuses on one blog in order to determine how postings in one place influence others. Nobody owns the term, the idea, because they only take on utility through usage.

  2. walt says:

    Mitch, thanks for the clarification; that wasn’t at all obvious from your post. I don’t know that I’d agree that conversational intensity in the easily-measured sense is “not particularly useful”–within the context of a single blog, it’s one of several metrics worth noting.

  3. Well, at the time I was busy working on a system to measure influence and my business partner always cringed when I started talking about details. There were a couple other postings around that time that fill in details, including this one:

    Interestingly, I heard exactly this conversational intensity metric come up with a company just this week as one of several metrics that are worth noting. My reply, as to Stowe’s posting and, by extension, yours, is that the in/out volumetric measure is a very rough, easily gamed one. And it can play games with one’s own perception of one’s influence because it can be an overly broad indicator.

    Influence is reputational, topical and variable based on the frequency with which someone writes on a topic. Using number of comments alone, especially over all postings on a blog during a timeframe, when examining a blog’s influence on a particular topic can be very misleading if the blog is popular because it is focused on, say, digital cameras, but the author writes one posting about French cheeses. If they didn’t get any comments on that posting, but average 3 comments per posting overall, would they be influential on the topic of cheese?

  4. walt says:

    Mitch, I would absolutely agree that if you’re trying to measure influence, conversational intensity is not a particularly useful measure. Fortunately, that’s not what I was trying to do–I’ve only used it in the context of looking at a field of blogs (in this case, liblogs: blogs by library people) and offering a few basic metrics on how they compare to one another. “Influence” was not one of the things I was looking at; as you say, it requires far more sophisticated measures, ones I wouldn’t begin to attempt.

    Basically, for my purposes, conversational intensity measures conversational intensity: Nothing more.