Bibs & Blather
But Still They Blog:
The Liblog Landscape 2007-2009
But Still They Blog: The Liblog Landscape 2007-2009 is now available–at a special early-bird price through the end of the ALA 2010 Midwinter Meeting (January 19, 2010 or thereabouts). Order your copy at www.lulu.com/content/7952668; until January 19, 2010, it’s available for only $29.50 trade paperback, $20 PDF download. (After Midwinter, those prices will go up $5.50 and $5 respectively.)
This 319-page trade paperback provides a sweeping look at liblogs (blogs created by library people but, generally, not blogs that are official library publications), with trends, facts, figures, graphs, and profiles for each of 521 liblogs. It continues what I believe to be the most extensive survey of the blogs within a given field.
The liblogs included here (you’ll find the list at walt.lishost.org/blogs-in-the-liblog-landscape-2007-2009) appear because:
· They’re in English.
· They began in December 2008 or earlier.
· They have at least some relevance to libraries and librarianship, although that point gets stretched in a few cases.
· They had at least three posts during March-May 2007, 2008 or 2009.
· They were available on the web in the summer of 2009 (even if they’d ceased).
They were known to me–either because they were listed in the LISWiki list of blogs or the LISZen list of blogs or because they showed up in one of a hundred or so blogrolls I checked.
· They were “visible”–in this case, having a Google Page Rank of at least 4 in either early fall 2008 or early summer 2009.
That final criterion was used deliberately to narrow this study’s focus slightly from the 2007-2008 study (which continues to be available, The Liblog Landscape 2007-2008: A Lateral Look.). I’d hoped to get down to 400-450 blogs, making analysis easier and the book shorter. I didn’t do quite that well, although the list of 607 blogs from the earlier study did come down to 480 (there are 41 new blogs).
If you’re wondering: Only 50 liblogs were eliminated because of their low visibility. The others were either non-English , defunct (that is, no longer viewable in August 2009 and with no clear trail to a new URL or blogname) [15, plus three that now require passwords], or didn’t have at least three posts in March-May 2007 or March-May 2008 …or, in three cases, really didn’t have any posts that had anything at all to do with libraries.
If you’ve read the series of posts on Walt at Random—starting December 4, 2009 and ending December 14, 2009—or if you’ve already ordered the book, you can skip this section: It essentially repeats the posts, without the lists of liblog profiles. (If you’ve already ordered the book: Thanks.)
The first chapter considers what might be happening with liblogs, changes in methodology and inclusion since The Liblog Landscape 2007-2008, changes in metrics this time around, and a few general comments on the liblogs:
· Their age
· Blogging platform used
· Currency as of September 30, 2009–that is, the most recent post as of that date.
The chapter ends with profiles for “pioneers”—liblogs that began in 2003 or earlier, often under different names. Later chapters include profiles for liblogs first mentioned (in 2009 or year-to-year changes) in that chapter and not already profiled.
Some blogs are rivers of posts—and if you subscribe to several, you may come to think of them as firehoses. Others, including most liblogs, are streams or rivulets: Writers and groups of writers letting you know when they have something to say that works best as a blog post.
How often do posts appear on a blog?
Until feeds and aggregators became common, that was an important question. If you didn’t provide a reasonably steady stream of posts, people wouldn’t have reason to come back to your blog or bookmark it. Few posts, few readers. Some people advised trying to do at least one post a day. Others offered less strenuous advice.
These days, when most readers see posts indirectly, a steady stream of posts is only important for certain kinds of blogs. Indeed, too many posts can work against readership, particularly if posts appear to be for the sake of posting.
This chapter considers frequency of posts among the 521 liblogs for 2007, 2008 and 2009—and changes in the overall picture. The next chapter considers changes on a blog-by-blog basis, a somewhat different consideration.
In all, 449 blogs had countable posts in March-May 2007, ranging from one post to 1,161, with a median of 25 posts (roughly two per week). 486 blogs had countable posts in March-May 2008, ranging from one post to 919, with a median of 20 posts. 434 blogs had countable posts in March-May 2009, ranging from one post to 909–with a median of 13, exactly one per week.
There’s lots more in the chapter, of course.
It’s clear from Chapter 2 that, on the whole, visible liblogs had considerably fewer posts in 2009 than in 2007, with fewer liblogs having any posts and fewer posts per blog.
But blogs don’t all change in the same way. This chapter considers changes in posting frequency on a blog-by-blog basis…
Quite a few libloggers did significantly more blogging in 2008 than in 2007—all of [the top 20%] and part of [the next 20%]. The median blog in Quintile 1 [the top 20%] had 75% more posts. The next year, the median increase was only 50% and, while the entire first quintile included more posts, the change ranged down to barely noticeable (8%). Over the two-year period, the top quintile includes a number of blogs with slightly fewer posts in 2009 than in 2007. Still, there were dozens of blogs with more posts in each successive year.
The second quintile, representing blogs with somewhat better year-to-year records than average, almost exactly matches my “relatively unchanged” definition (+20% to -20%) for 2007-2008, but ranges from tiny increases to losing a quarter of posts for 2008-2009—and, for the two-year period, includes blogs dropping four out of ten posts over two years.
Last year, it seemed reasonable to suppose that, on the whole, liblogs would have fewer posts but longer posts, as Twitter, FriendFeed, Facebook and others replaced many of the uses for very short posts.
If anything, that’s even more true in 2009, even as a number of bloggers simply stopped blogging. One new liblog is an extreme case: In the Library with the Lead Pipe, a group blog that’s essentially an essay magazine done in blog form, with each (reviewed and edited) entry the length of a typical magazine or journal article.
While more of the remaining libloggers seem likely to write essays rather than quick posts, there are still blogs for which a single sentence or two is the norm, including link blogs and some others.
Chapter 4 begins with metrics on overall blog length and how they’ve changed. The longest blogs seem to get longer every year: While March-May 2007 tops out at 186,467 words, March-May 2008 jumps past the 200K mark (204,517 words) and March-May 2009 finds one blog all the way up to 238,351…noting that it wasn’t feasible to measure total length of some blogs. At the same time, the median length declined each year–from 6,216 words in 2007 to 5,536 in 2008 and 3,621 in 2009.
More interesting, however, is post length, even if it’s only practical to measure average post length. (It would be interesting to measure length distribution within each blog, but also incredibly time-consuming…) Most of this very long chapter is devoted to discussions and tables relating to average words per post and how post length in blogs has changed over the years–and to the largest set of blog profiles in the book, partly because terse blogs (those averaging less than 100 words per post) are profiled along with the essayists.
Is blogging publication or conversation? Yes and sometimes. Blogging is always a form of publishing—but some posts on some blogs become conversations. The conversational function varies heavily from blog to blog, and newer tools—particularly FriendFeed and FaceBook—may have weakened blog conversations, with the odd result that some extended FriendFeed conversations are based on blog posts and might otherwise take place on the blogs.
Some blogs don’t have comments, either because the blogger doesn’t allow them or because the posts don’t attract comments. There are some blogs where I couldn’t determine the number of comments—although there are also blogs where I couldn’t track length but could count comments.
This chapter considers overall comments for each blog during the three-month study periods (March-May 2007, 2008, and 2009)—but also the more interesting metric: conversational intensity or average comments per post. There’s an anomalous change in the highest overall comments (dropping from 1,689 in 2007 and 1,219 in 2008 to 581 in 2009), almost certainly the result of one particular blog moving onto the inscrutable (or at least unmeasurable) LJ/SLJ blog platform–I’d call it “blowing a fuse,” but that would be a cheap joke. In fact, highest conversational intensity went up sharply in 2008 (from 28.9 to 53.0) and stayed up in 2009 (51.0), although the gap between the highest CI and the second highest CI was huge (second highest: 13.8 comments per post, with four others over 10).
The chapter also includes three-year patterns for changes in conversational intensity. It’s hard to draw any overall conclusions, since over the 2007-2009 period, roughly 40% of blogs increased significantly (more than 20%) in conversational intensity while another 40% decreased significantly!
Before considering patterns of change (how blogs change across multiple metrics), let’s look at some standouts and standards: Blogs that are within the same quintile either across all three key metrics (frequency, post length and conversational intensity) or across all three years within a given metric, and are also within the top three quintiles for the metrics in which they show consistency.
This chapter is about consistency—falling into the same general population across several metrics. It’s not about quality, and no larger conclusions can be drawn. Think of this as a break in the narrative. You’ll discover early on that no blog is in the first quintile throughout—although two come close, with consistently top rankings in two of the three years.
So far, the book has looked at one metric at a time (except for chapter 6) but a blog is more than its individual metrics. This chapter and the next look at patterns—patterns of change from one year to the next. Three elements make up the change pattern for a blog:
· Change in number of posts: Were there more posts in 2008 than in 2007, fewer, or about the same number?
· Change in post length: Was the average post in a given blog longer in 2008 than in 2007, shorter, or about the same length?
· Changes in comments per post: Was the blog more conversational in 2008 than in 2007 (that is, did the average post have more comments), less conversational, or about the same?
Table 7.1 offers a simplified view of these three changes—“simplified” because it breaks blogs down into “More” or “Less” (where no change at all is counted as “More”)—and that overstates the significance of small changes.
For those who read last year’s study, note that there’s one significant change this time around, for both the simplified table and the triplets: I’m leaving out blogs that lack length metrics in either of the two years being compared. That’s never more than 10% of the blogs, and it means the tables can be considerably shorter (24 lines rather than 36 in the case of Tables 7.1 and 8.1) and easier to understand. Since every blog with a length metric has a valid comment metric (even if the comment count is zero), that further simplifies the process. Blogs are omitted if they have no posts in 2007 as usual—but not if they have posts and no comments. (Note that a blog with zero posts in both years would be counted as having “more” conversational intensity in the second year—an example of the problems with straight up-down comparisons.)
That’s the start of the chapter. Most of the chapter deals with triplets—blogs that have increased or decreased more than 20%, and those that haven’t changed all that much. It’s a rich discussion; I won’t attempt to provide a summary here.
There’s one peculiarity for 2008-2009 that wasn’t present in 2007-2008: Half a dozen blogs that went from no posts to some posts—and are included because they also had posts in 2007. Since moving from nothing to something is an infinite increase, these show up as having significantly more and longer posts with significantly more conversational intensity…
If patterns of change across the landscape were completely random, each of the fully-indented rows (combinations of three metrics changes) would have roughly 56 blogs and 1,660 posts and show 13% in each percentage column.
None of the eight patterns is close to those figures.
Three outliers are interesting:
· The most common pattern by far is the “discouraged” pattern: Fewer, shorter posts with less conversation. That pattern represents 125 blogs (28%) but only 11% of the posts.
· The next most common patterns are two with fewer posts and more conversation—77 blogs with longer posts and 71 with shorter posts. Combined, those represent a third of the blogs and 28% of the posts.
· The pattern with fewest blogs is the same as for 2007-2008: More posts, but shorter and less conversational. That has 20 of the blogs (4%) and 564 posts (4%).
It’s also interesting that two-thirds of blogs had (slightly) longer posts—and that a solid majority had more conversation.
The chapter also goes through the “better model” of triplets. You’d need to read it and study the tables to gather much meaning.
A short chapter, and I can’t claim to have found anything startling. It’s much less graphically interesting than the corresponding chapter in the earlier book, as I chose not to prepare scatterplots (they’re fun to do, but I didn’t find them meaningful in these cases). Let’s just say that there are no startling correlations between pairs of metrics.
This study includes 521 blogs. What they have in common is that each involves one or more “library people” as defined very loosely—people who have some connection to the library field and write, at least part of the time, about library-related issues.
How do these people blog, and how is that changing? That’s largely what this book is about, on an objective, quantifiable basis. I discuss qualitative areas in Cites & Insights from time to time.
Why do these people blog—and how is that changing? There are many reasons for blogging, some more sensible than others. Here’s my quick take on plausible and implausible reasons for starting and maintaining liblogs, followed by some comments from bloggers themselves…
The chapter begins with some reasons I believe people blog–just a few of the many–and continues with material from the July 2009 Cites & Insights, followed by new material (some of which appear, in different form, in this issue).
Why does a blogger pause (which I’ll define as not blogging for at least four months) or stop altogether? I’m certain the most common reason is premature blogging, that is, starting a blog before you really know whether you have much to say. I suspect other reasons are all over the map, with the second largest probably running out of steam or losing interest (or, these days, finding that saying what you have to say is easier and faster on Twitter, FaceBook or FriendFeed).
A number of libloggers stopped between mid-2007 and mid-2008, or at least paused for so long that they don’t have any posts—at least 13% of those with enough impact to make it into But Still They Blog and probably more than that among the broader liblog population. Some returned; many didn’t.
What follows is a sampling of posts on why people have stopped or paused blogging—or, in some cases, the fateful final posts that don’t appear intended to be final. Included are some “haven’t been blogging much lately” posts.
This is the point in the book at which I should find profound meaning from these metrics. It’s the perfect opportunity for sweeping conclusions—if there were any.
You’ve seen smaller conclusions throughout the chapters. Yes, a fair number of bloggers have stopped (when has that not been true?). Yes, there seem to be a lot fewer new fairly-high-profile liblogs in 2008 than in previous years. Yes, most bloggers are blogging somewhat less (and very slightly longer).
And yes, some of that can probably be traced to FaceBook, Twitter and FriendFeed, along with the usual reasons—fatigue, changes in life and work, balance, boredom.
Underlying all that, liblogs still offer a broad, varied landscape of people with interesting and worthwhile things to say. Blogging may be dead (if you believe some pundits)…but still they blog.
The remaining liblogs—those that didn’t turn up in a previous chapter—aren’t “leftovers” by any means. A few of these are among my personal favorites, one or two are among those I choose not to comment on so as to avoid snark, several have gone by the wayside—and many just don’t have quite enough frequency, long enough posts or enough comments to stand out in a metric (or had metrics problems).
Again: metrics only measure quantity, not quality. You need to judge quality for yourself.
That’s it. The book is textually richer than The Liblog Landscape 2007-2008 and, I believe, continues an important study that should be of interest to every library school and many bloggers. It’s about as transparent as research can be. The price, particularly through January 18, 2010, is (I believe) unusually reasonable for a 300+-page trade paperback, on proper book stock, representing original research in the field.
Cites & Insights 9 (2009) is now available as a 434-page, 8.5×11, trade paperback, exclusively from Lulu at www.lulu.com/content/7903887/
The volume includes all 13 issues exactly as published (typos and all), except that two book covers in the January issue are in grayscale, not color.
It also includes a contents list showing the articles and pages in each issue, and a volume index.
The price is $35 for the paperback or $25 for a PDF download.
The book is printed on bright-white 50lb. paper (my copy looks great!).
As to the cover (a wraparound color photo–you’re only seeing the front part here):
Taken by my wife on Molokai, years ago, on the Kaluakakoi golf course running alongside our room at what was then, I believe, a Sheraton at the Ke Nani Kai resort on Molokai’s isolated west coast. (The hotel’s been closed for some time…tourism on Molokai is an iffy thing.) The only manipulation done to the picture (scanned from a 3×4 print) was to flip it horizontally, so most of the tree would be on the front cover rather than the back. Crappy type position is entirely my responsibility.
C&I 9 was originally priced at $50 paperback or download, the price set for other book volumes of Cites & Insights, set so that sales would actually help support Cites & Insights itself. While that was an interesting notion, it hasn’t been successful—and I believe the trade-paperback volumes are worth having, at least for library schools and quite possibly for a few readers. So I’ve reduced the price on Volume 9 and on the other three volumes available as trade paperbacks to $35 paperback, $25 PDF.
Admittedly, I produce the books in the first place because I need bound print volumes for later reference, and doing them through Lulu yields a much nicer finished product than doing them at FedEx Kinko’s, for not a lot more money. (Let’s ignore for the moment the fact that, the next-to-last time I did a Velobound annual at Kinko’s, bringing them a ream of bright-white paper to use for the copying, they apparently screwed up and did 80% of the book on the usual cheapest-available copy paper, without telling me.)
Still, they’re great ways to have access to the C&I archives. Volumes 6 & 7 even have special features only available in the book version: For Volume 6, a special preface; for Volume 7, the briefly-available Cites on a Plane. For Volume 9, I did something I should have done previously—included a table of contents showing the essays in each issue, reformatted from the online contents page.
I’ve also reduced the prices on the other two 6x9 trade paperbacks currently available from C&I books (from Lulu and CreateSpace/Amazon; the C&I annuals may also be available via Amazon, although that’s less certain to continue). Balanced Libraries: Thoughts and Continuity and Change and The Liblog Landscape 2007-2008: A Lateral Look are each available for $25 (plus shipping) trade paperback or $16 as PDF downloads. You’ll find them all at stores.lulu.com/waltcrawford and some of them on Amazon.
Since no new sponsor has appeared for Cites & Insights, I’ve restored a PayPal Donate button on the C&I home page. If you find C&I worth enough that you’re comfortable paying for it—that is, you think it’s worthwhile and paying a little wouldn’t deprive you of anything else—you might choose to donate, either via PayPal or credit card (the button opens a secure link).
There’s no set amount. I’ve suggested $7 or $8 if you find one particular issue worth paying for, or $50 (or $25) if you’d like to support an entire volume of C&I. (You could also buy the PDF or hardcopy versions of the annual volumes, but they’ve been lowered in price such that they’re not really a support mechanism.)
There’s also no guilt, threat or pressure. C&I is free. Donations are entirely voluntary.
Opinions herein may not represent those of LYRASIS.
Comments should be sent to firstname.lastname@example.org. Cites & Insights: Crawford at Large is copyright © 2010 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 http://creativecommons.org/ licenses/by-nc/1.0 or send a letter to Creative Commons, 559 Nathan Abbott Way, Stanford, California 94305, USA.