Cites & Insights: Crawford at Large
ISSN 1534-0937
Libraries · Policy · Technology · Media

Selection from Cites & Insights 5, Number 4: March 2005


The Dangling Conversation

Many of you know about a multifaceted brouhaha regarding one item in last issue’s The Library Stuff. I say “many of you” with some hesitancy: I have no idea what proportion of regular C&I readers read any or all of the three weblogs and one LISNews journal where the whole mess primarily played out, but I believe one of those weblogs, The Shifted Librarian, has a significantly larger readership than Cites & Insights. I’m not going to revisit the controversy itself. If you missed the whole thing, it’s not hard to track down, but why bother?

I am going to use part of the controversy as a springboard. Warning: If you’re looking for black-and-white simplicity, for a clear statement saying “A is good, B is bad”—or even “Walt loves A and hates B”—then maybe you should skip to the next article.

This essay touches on several tools: lists (Listserv™ is a trademark for one brand of list management software), publications (such as the one you’re reading), email, weblogs, RSS, aggregators, wikis, group software (Lotus Notes, Groove and others), and categories of software I haven’t seen wholly satisfactory names for, but that include slashcode-type systems such as LISNews and Kuro5hin as well as systems similar to LiveJournal. I would mention “social software,” but I believe that term includes all of these tools except publications.

All these tools have real, worthwhile uses. I don’t currently participate in wikis or LiveJournal-like systems (call them group journaling systems), but that’s happenstance as much as preference. All these tools also have weaknesses, as is true of most everything in life (and everything that depends on a computer!).

I automatically raise objections when I read that X is “the future of communications” or that Y is “how we will all interact” or that Z will wipe out A, or similar claims of ubiquity, inevitability, or monolithic futures. But if you tell me X is worthwhile (where X is any of these tools)—or, for that matter, that you find Y suits your preferences better than X—then you won’t hear me object.

Sufficiently confused? Good. What I really want to talk about is modes of communication, conversation, community, and claims that one mode is somehow superior to all others. I’m afraid it’s going to be another long one. Sorry about that. (One weblogger commenting on the 6,000-word Wikipedia and Worth [Revisited] perspective managed to boil it down to “we should all just get a grip.” Now that’s concise writing. I’m jealous.)

Jenny Levine’s Comment—and Disclaimer

Here’s a portion of one post from the controversy—with the caveat that I’m using Jenny Levine’s comments as a springboard, not implying that she holds the “This, not That” attitudes I find questionable. (The home page at The Shifted Librarian says “RSS Bigot” in the right-hand column, but I take that as a joke.) I am not accusing Levine of bias here; I just find her comment a good place to begin. Here are the passages (from a February 9 post), eliminating portions that don’t deal with the issues I want to consider:

[W]hat really struck me today was the format of Walt’s responses. In the past, he’s left comments on my posts, and I love him for that. Community is a very cool thing that I never anticipated when I started my blog, and I value every comment I’ve ever gotten and ever will get, especially thoughtful ones like those Walt tends to leave.

Usually, though, when he has more than just a few sentences to say, he saves his commentary for the next issue of C&I. But he didn’t do that this time. Instead, he left a couple of comments and then felt the need to blog his major response. What he wanted to say was so important that it couldn’t wait a month for his normal publication cycle (probably because he felt attacked, which he kind of was, but in the friendly way that Walt and I agree to disagree with each other…) Other than pointers to announcements of new C&I issues, I think Walt gets a lot more of an online community and conversation from his blog and the comments he leaves on other bloggers’ sites. I think it’s a very different audience for him, one that expects a conversation and is frustrated by the lack of interactivity a PDF provides. I’m not knocking the format or C&I, I’m just noting how different a monthly PDF feels from blogging.

I know Walt isn’t against blogging; instead, I want to use this example to illustrate the essential elements blogs can bring to libraries: conversation, dissemination, and community. We’re having a conversation that others are joining in on, we’re both disseminating our thoughts easily and efficiently, and we both have communities built up around our writing. Obviously Walt felt the need to make use of that interactivity and immediacy for this one.

Your library’s monthly newsletter—it has the same problems as C&I in this case. Yes, it may have its place and I’m not saying you should get rid of it, but blogging gives you something very different…

Early response

Here’s an excerpt from my comment (after I managed to get it down to the 2,500-character comment limit in four editing passes):

My LISNews journal post wasn’t a major response and had nothing to do with importance—and everything to do with feeling singed, wronged, and angered. And believing most people would only see the attacks, not the commentary they were attacking.

I know “conversation” is one of the claims for why blogs are so wonderful. I even believe it—partly. Except that (a) many blogs don’t allow comments—and many that do have character limits, including yours, (b) some blogs make it difficult to comment unless you know the secret handshake, (c) some blogs only display comments after the blog owner has a chance to review the comments, (d) people who read blogs via RSS don’t see the comments at all in most cases, and (e) I’m guessing that most blog readers don’t bother to click through to comments. It’s an unusual sort of conversation, and that may be the subject of a future essay in C&I.

Along with a generous offer to post a longer response from me on her main page—which I didn’t choose to do, because I was already tired of the controversy—Levine responded, in part:

[While] this is an “unusual” type of conversation, it’s more than I get with C&I. Certainly, letting others chime in has added value to the discussion, as well. I prefer this to the one-way flow of a monthly PDF, but haven’t we already agreed that you should do what works for you and I’ll do what works for me?

And I responded, in part:

Yes, I absolutely agree that (some) weblogs provide better and more immediate feedback mechanisms than C&I. You may note that I never have advocated that everybody—or anybody, for that matter—should emulate what I’m doing in C&I as The Way to Communicate. I began it as an experiment and it seems to work, for me, for some functions. I doubt that it would work for very many people, and I may find that some functions work better in a more formal blog than my LISNews “blog lite.”

That’s enough from the discussion. The other primary participant in the controversy never did post my comment, or at least hadn’t for a week. My charitable interpretation would be that the comment got lost in the infosphere somehow.

Dissemination, Immediacy, Community

Publications, lists, email, weblogs, RSS, aggregators, wikis, group software, threaded bulletin boards, group journaling systems (those latter terms serving as names-of-convenience for slashcode-style systems and LiveJournal-style systems). If you want to be up with the latest innovations, I’ll call podcasting a form of audio publication (but I haven’t used it: Is there a “push to talk back” function?).

Each tool supports dissemination. Otherwise, they wouldn’t work at all.

I’m going to suggest that immediacy of response is not a direct characteristic of the tool being used. Nothing except common sense stopped me from issuing a new Cites & Insights the day after Levine’s initial post. It had been a week since the original publication, I had enough material on hand for a 12-page issue, and I’ve never held fast to a monthly schedule. It would have been stupid and petulant for me to do so, but not impossible. Other publishing methods are immediate—radio and television, for example, with newspapers only a few hours behind.

On the other hand, while many lists offer immediate feedback possibilities, many others do not, with moderation requirements imposing delays of an hour to a day or more. Similarly, while some blogs allow immediate comments, others require owner approval or don’t allow comments at all. (I don’t understand wikis well enough, but assume that they either can be entirely open to contributions or can involve a moderation layer as well.)

So it’s really about community and conversation. Or maybe it’s really about conversation.


An interesting, focused weblog can build a community or enhance an existing community. That’s absolutely true. So can a wiki—for either a self-defining open community or a closed, invitational community. Group software assumes a community of interest, almost always a closed group. Group journaling systems only work well when there’s a common nexus of interest, one definition of a community.

I believe every tool in the list above can foster and enhance communities. Lists (open or closed) directly support communities and are defined by the community they serve. Those communities can be as small and specialized as six people working on a project or as large and open-ended as Web4Lib or LITA-L. Threaded bulletin boards work better in well-defined communities than in vague open-ended communities: LISNews is less prone to flamewars than Kuro5hin, which in turn is less self-destructive than slashdot, to name three examples.

Good publications also build and enhance communities. There’s a community of core Cites & Insights readers. I’m fairly certain there’s a Library Juice community. The communities get vaguer as the publications become larger. The nation’s largest-circulation periodical certainly serves a community—people over 50—but that’s about as vague a community as you can get: It’s too diverse to be very significant.

So let’s look at conversation. How well does each tool work as a conversational tool?


No question: Publications suck when it comes to conversation. A letters column, my occasional Feedback section: Those don’t really constitute conversation. Publications aren’t interactive (even if they include user-controlled pseudo-interactivity); that’s why they’re publications.

Group journals

Group journaling systems appear to be great for conversation, as long as the group is closed or specialized enough so that it doesn’t grow too large. They have all the elements I consider crucial for true conversation:

Ø    The tool does not privilege any voice over any other voice. Only the clarity, meaning, and power of a voice lend it extra weight—excluding, of course, the weight provided by personal awareness. (Knowing who’s saying something is always important.)

Ø    Everyone can see everyone else’s comments immediately and on an equal level.

Ø    The tool encourages informality while maintaining a record.

Ø    The tool discourages anonymity and allows for (but does not require) the use of true names.

I dislike “avatars,” screen names and other attempts to separate people within a conversation from their real-world identities. I see the reasons for such devices, and they’re better than total anonymity, but I believe they interfere with conversation. You don’t really know who you’re conversing with—and that, to me, is a significant aspect of conversation.

Lists and group software

Lists can be as effective as group journaling systems; so can group software. A note on argues that lists can offer more sense of community than weblogs, because they’re more of a shared experience—but only if you’re really subscribed to a list, not getting it as a digest. The post goes on to note the magic of spontaneous conversation and how it can shift in unforeseeable ways. The blogger thinks this is more prevalent in lists. “Of course, such spontaneous conversation isn’t always a good thing. It is very easy for a [list] conversation to take a distinct turn for the worse. But sometimes these unexpected shifts can be amazing.” Earlier, the post notes that one of the most valuable things about a list is also one of the most annoying: “It’s not very easy to unsubscribe from hearing a particular person’s views.” That means listening to people you find disagreeable—but also tends to mean a greater diversity of views.

Steven Cohen blogged about this entry, suggesting that list “banter” isn’t spontaneous: “These are thought-out posts, just as in a weblog post.” He also agrees that lists encourage listening to people you disagree with and asserts (jokingly, I assume), “I only read blogs from writers whose views I agree with.” Then he suggests that “[lists] still have more content that ha[s] no bearing on my opinions at all; probably 85% of it (a semi-arbitrary number).”

To which I say, “That depends on the list.” It’s quite possible that no list is mostly relevant to Cohen’s interests and opinions, but if so that’s a little surprising. As for the spontaneity of list postings, that absolutely depends on the list and its participants. Many lists (probably most) are small, highly focused, and only open by invitation or because you’re part of some organization. I can say with absolute assurance that some such lists support extremely spontaneous comments, can go back and forth almost as rapidly as a conference call, and consist almost entirely of comments relevant to most people on the list.

At the other extreme, some lists are so big and diffuse that they can’t survive without moderation and wander into irrelevance. Others, Web4Lib one astonishing example, manage to carry on a series of conversations involving several thousand people with neither moderation nor much in the way of flame wars. Many lists require the use of real-world identifiers. Few lists favor “screen names” or avatars and most well behaved lists don’t allow for anonymous posts. Those are good attributes, in my opinion.

The disadvantages of lists are that they require some effort to establish and maintain (including hosting in somewhat “non-weblike” ways, although there are exceptions) and, for some people, that lists work via email. (There are also one-way lists; those are just another form of mass email or publication, so have no role in this discussion.)

Email. wikis, threaded bulletin boards

Email itself is great for one-on-one conversation, a little less great for small-group conversation, and not that useful for community building (except as a carrier for lists). Email is the great conversational medium of the internet. Unfortunately, that means it’s been damaged by a flood of people who want to butt in on the conversation in ways that wouldn’t be possible in the real world. (But I got served an ad when I tried to look at comments on a blog recently, and Boing Boing’s RSS feed now inserts text ads among the posts. The flood of unwanted commercial messages creeps into every medium. Have you changed your Huntington Bank security settings yet?)

As regards wikis, I’m going to take a pass. They’re clearly worthwhile tools for collaboration and can be ways to define and build a community. I haven’t used them enough to know whether they’re particularly good tools for conversation. I suspect “it depends” is a good answer.

Threaded bulletin boards are all about conversation—or are they? You’ve guessed by now that I regard slashdot with a mixture of horror and fascination: If that’s a community conversation, call me a hermit. On the other hand, I continue to participate in LISNews, which uses a variant of the same software. These boards can certainly build or assist a community (or disrupt it, depending on how they’re run and how they work). However, I’m not sure they are truly conversational tools, at least not in the way that lists can be. Why not?

Ø    The standard paradigm for such a board is story-and-comment, with comments responding to other comments and comments responding to those responses and so on, as many levels deep as is needed. That automatically gives greater voice to the person creating a story (and greater power to the editors, those empowered to approve stories).

Ø    In some configurations, you only see the stories unless you specifically ask for the comments. That methodology (used at LISNews) makes the bulletin board a better medium for posting stories, since they’re easier to browse through—but it gives even greater advantage to the person originating the story. You have to go looking for the responses. That’s not how conversations work.

Ø    The point-moderation functions typical of these systems may be essential to keep them from being entirely buried in flame wars, spam, and huge doses of irrelevant chatter. But they also create various levels of advantage and disadvantage that may have little to do with the quality of the actual messages. (Metamoderation helps somewhat.)

Ø    While threading works to make individual subtopics more coherent within a busy discussion, it violates the rules of a group conversation. That may be a good thing—but it’s not conversation. I habitually view sets of comments in straight chronological order at LISNews, setting a threshold to see all comments. To do so at Kuro5hin or slashdot would be madness.

Ø    Threaded bulletin boards seem to favor screen names and (in some cases) anonymous posting over real-world identification. Maybe that’s just my sense, but that’s how it seems to work out. I regard that as a slight disadvantage for effective conversation.

Weblogs and supporting tools

Which brings us back to weblogs—and RSS and aggregators, both tools which (in the “conversation and community” space) serve to support weblogs. (Yes, I know, RSS and aggregators have lots of other functions, but those aren’t typically conversational in nature. Do you consider newsfeeds conversational?)

You already know what I’m going to say here, because it’s in the second paragraph under “Early response” above. But there’s more—although it’s all variations on the rest of this essay:

Ø    Every weblog gives a considerably larger voice to the owner(s) of the weblog than to anyone else wishing to “join in the conversation.” It’s not a conversation. It’s a statement that may be followed by responses (and responses to those responses), but one person (or a small group) always gets to make the initial statement—and usually the final one as well.

Ø    Almost all weblogs I’ve seen give an even larger voice to the owner because you have to specifically ask to look at comments. (Note that typically they’re called “comments” or “interjections”—not “the rest of the conversation.”) I’ve seen weblogs that incorporate comments into the main body of the post as soon as they’re made or approved, but they’re rare, at least in the library-related, copyright-related and other weblogs that I follow.

Ø    Only the most “conversational” weblogs support immediate comments, one click away from the post itself, making it easy to enter a comment and to use your real name. The Shifted Librarian falls into that category—and even in Levine’s case, in order to minimize spam, there’s a 2500-character limit to each comment. Given the length of some posts (entirely appropriate to their subject), that fundamentally biases the “conversation”: If you need to reply at similar length, you’ll have to post several separate comments.

Ø    Some weblogs require two steps to write a comment (true of most Blogger weblogs), discouraging the conversation. Many weblogs don’t show you the post while you’re writing a comment, making the conversation more cumbersome (although multiple tabs and windows make this a minor problem). Many weblogs want you to have an account with the particular software—and won’t let you sign a comment unless you have such an account. (That problem’s declining over time.) These all discourage conversation.

Ø    Going one step further, some weblogs won’t accept “anonymous” comments (which I applaud, given that I don’t care for anonymous conversations)—but, in a few cases, that means they will only accept comments from people with accounts for specific software. That’s a fairly sizable bar to spontaneous or even well thought out conversation.

Ø    Then there are the real problems: Weblogs that only post comments after they’ve been approved by the owners—and weblogs that don’t support comments at all. In the latter case, “conversation” is a complete misnomer—and in the former case, I’ll argue that the conversation is so fundamentally biased that it barely deserves the name.

Just for fun, I looked at the situation on some of the library blogs in my Bloglines list. I only looked at the first 40 alphabetically, skipping a few very stale or wholly atypical blogs (I’m not saying how many are in the total list). Here’s what I found: Sixteen—40%—do not accept comments (or comments just don’t work). Eleven more require two steps in order to enter a comment. Thirteen (there’s overlap with the eleven) favor those with accounts for certain software. Several—my notes are muddy—require a name and an email address; prefer that. None of these 40 shows the comments within the main body of the blog. Since I didn’t actually enter comments on all 40, I don’t know how many of these have length limits or require owner approval before a comment actually appear. (One of the 18 appears to use slashcode or some variant; comments appear as threaded lists.)

Interactive? Yes, in 60% of the cases. Conversations? In a manner of speaking—but with nowhere near the equity and full conversational power of some other tools.

Which is not an argument against weblogs. They’re easy to establish and easy to use. They seem to encourage a level of informality that I like. They have a whole infrastructure (RSS, aggregators, “blogrolls,” linkbacks, and all those sites like Technorati that play with blogs) that makes them more interesting. I have a “blog lite” at LISNews; you can’t get lower overhead than a LISNews journal—and those journals have a range of “conversation-friendliness,” from barring comments to allowing them but not showing whether there are any, to showing a count. I chose the most conversation-friendly option.

What about RSS and aggregators? To my mind, they typically work against the conversational role of conversation-friendly weblogs, and slightly against the community-building role. That’s particularly true with full-text feeds, by far the most reader-friendly form. When I follow weblogs via Bloglines, there’s no real difference between a weblog and a news site: I’m reading a group of publications. Sure, I can click through to the weblog, then click through again to see the comments—but that’s two extra steps and negates much of the convenience of weblogs. And if I’m using the aggregator partially to avoid yellow-on-black text and other design abominations, I’m really unlikely to click through. (There are exceptions: I’ve seen feeds that incorporate comments. But they’re rare.) On the other hand, feeds and aggregators make it feasible for people to have weblogs who really don’t have something to say every day: I track at least five times as many library blogs via Bloglines as I would if I had to visit each site once a day.

That last may be the other reason why, for people who use email fairly steadily (as many of us need to do in our jobs), lists make much better conversational tools than blogs. Blogs, either directly or via aggregator, are pull media: I don’t see posts until I visit, and then I won’t see other comments until I visit again. With blogs, that would never be more than once a day; with an aggregator, it might be twice. Even if it’s more often, Bloglines only checks for content once an hour—a limitation that’s vital to avoid overloading weblog servers. That makes for a very “dangling” conversation. On the other hand, list posts—for unmoderated lists—just show up in email, and the email client I use checks for new mail every ten minutes.

Multiplicity and Preferences

Instead of my old “And, not or” theme, maybe I should quote the name of a weblog run by three OCLC staff: “It’s all good.” All of these tools can build community. All can—to one degree or another—allow conversation. All can be tailored to improve participation or make it more difficult. At this point, different people have different preferences—not just awareness, but preferences. For some of us, time and mental energy are both too precious to bother picking up a new tool unless and until we’re convinced it’s better for us.

A discussion of library blogs on Web4Lib touched on this point when Bill Drew, who’s starting a library blog, asked whether it was possible to have a feed of some sort that would show up in people’s email. Michael Sauers seemed shocked by the idea: “Why fall back on e-mail when you’ve got the perfect solution (RSS feeds) already in place and working?” The fact that users may not be “RSS capable” “gives you the perfect opportunity to teach them. Help them become more technologically [savvy] and at the same time show them that it’s not difficult.”

Drew responded, “I am looking to get information to them in the way they prefer or are familiar with, not what I might prefer. RSS being considered better than e-mail is a preference not a fact.” Kevin Broun at the National Cancer Institute checked his server logs for 2005 through February 8, finding that there were “a few dozen” hits on the library’s RSS feeds—and “a couple thousand emails” notifying several hundred clients about content. “Sure, we can do more to educate our users about the feeds—but for the most part, they aren’t interested or ready for it yet.” Another poster noted that she has an RSS reader at home—but she still prefers email, as does Bill Drew. Fortunately, there are applications to cross the bridge: Peter Scott mentioned one, RssFwd. ( You can also go the other way: Bloglines makes it plausible to have lists and other email come into your aggregator.

Despite the seeming incongruence, it makes sense for such services to be available—so that people can work they way they prefer. Maybe librarians can and should also do some education on the virtues of the newer tools; that’s a separate issue.

After I began this, I tried an experiment related to Cites & Insights that accidentally offered another piece of evidence about the relative conversational merits of blogs, RSS, and lists—although I wouldn’t put too much weight on the item. I was considering adding HTML versions of some stories to facilitate inbound links (from blogs and elsewhere) and to encourage readership outside the library community, by people who really wouldn’t get much out of C&I as a whole. I put up some samples (all the stories from C&I 5:3) and did three announcements inviting comments, staggered over two days:

Ø    Early Monday, February 15, I posted a notice on the C&I Updates blog, which I’m guessing reaches about 180 to 200 people via aggregators (based on the number of Bloglines subscriptions). By the end of the week there were two responses.

Ø    Later that day, I posted a similar entry in my LISNews journal—my “blog lite,” if you like. Blake Carver estimates that I have around 1,200 readers. By the end of the week, five people had commented on that entry.

Ø    Tuesday evening, I posted a similar entry to the Topica CICAL Alert mailing list, which has 395 subscribers. The entry was significantly less convenient: Unlike the other two posts, it didn’t have live links to the C&I contents page from which the HTML examples could be viewed. There were 18 responses within 12 hours, 22 or 23 by the end of the week. Note that these pepple didn’t even have the ease of list commenting: They had to create new emails to me, since CICAL Alert is an announcement-only list that doesn’t accept replies.

I’m not sure what that means, except that lists and email are potent means to encourage conversation.

I’ll close with comments from Steven Cohen’s Library Stuff sidenote during the controversy that triggered this Perspective. Cohen hates email: he’s said so in so many words. (“I hate e-mail.”) Here’s what he says about the Proper Tools to use when getting library information out into the community. “So, if you want to use e-mail, then fine, use e-mail… And, if you want to have e-mail notifications on your blog, then do it… Just get the content out to the readers as soon as you possibly can after posting to the blog… Use RSS, e-mail, IM, Morse Code, smoke signal, whatever. Just get it out to your patrons.” He also says that the controversy itself indicates that RSS is a hot topic, one that librarians can’t afford to ignore. I agree fully.


I know that’s not a word. It’s also not a mistake. I don’t believe this final paragraph really comes to conclusions. (Hmm. Maybe I have a local neologism here: A lot of my essays come to inconclusions.) If I wanted to start up a new conversational community on a new topic on my own time, I probably wouldn’t start a list—even though I think it’s currently the most “conversational” of the tools that don’t require special apps (i.e., Groove). I’d probably start a group weblog or, if I understood them better, a group wiki—not because they’re better at conversation or community, but because they’re more “weblike,” which tends to make them lower overhead for new services. But I can’t imagine converting an operational list to those other tools. The overhead’s already been covered, and the tools work very well. There’s room for all these tools—and more that I don’t know about yet.



Cites & Insights: Crawford at Large, Volume 5, Number 4, Whole Issue 60, ISSN 1534-0937, a journal of libraries, policy, technology and media, is written and produced by Walt Crawford, a senior analyst at RLG.

Cites & Insights is sponsored by YBP Library Services,

Hosting provided by Boise State University Libraries.

Opinions herein may not represent those of RLG, YBP Library Services, or Boise State University Libraries.

Comments should be sent to Cites & Insights: Crawford at Large is copyright © 2005 by Walt Crawford: Some rights reserved.

All original material in this work is licensed under the Creative Commons Attribution-NonCommercial License. To view a copy of this license, visit or send a letter to Creative Commons, 559 Nathan Abbott Way, Stanford, California 94305, USA.