Weblogs (“blogs”), frequently modified webpages containing individual entries displayed in reverse chronological sequence, are the latest mode of computer-mediated communication (CMC) to attain widespread popularity. As with other new CMC technologies, blogs have been hailed as democratizing—any literate person can self-publish content in a blog, and reach an audience of potentially millions, for little or no cost. Moreover, the success of individual blogs in attracting readers and influencing opinion depends less on their formal credentials than on the quality of their ideas and their writing (what Winer, 2003, calls their "voice"). Certainly blog authors are numerous: In the five years since the introduction of the first free web-based blogging tools (Pitas and Blogger; Blood, 2002b), the number of people creating and maintaining blogs has grown exponentially, from fewer than 100 to over four million (Henning, 2003).
Blogging Technology Effects on Common people in 2004
Anecdotal accounts also suggest that they are diverse: the mainstream media have reported on popular blogs authored by individuals as varied as university adjuncts, dark horse candidates for political office, and a gay Iraqi dissident (McCarthy, 2003). As yet, however, there has been little empirical examination of the claim that blogs are “democratic,” or that blog authors represent diverse demographic groups.
Fifteen years ago, a similar claim was advanced with respect to Internet discussion forums and chat rooms. Text-based CMC was purported to be inherently democratizing, enabling anyone with access to participate, liberated from traditional biases associated with gender, age, race, social class, (dis)ability, and physical attractiveness (Graddol & Swann, 1989). Subsequent research revealed, however, that the demographics of actual forum participants were strongly skewed towards adult, white, English-speaking, technically-savvy males (Herring, 1992, 1993; Kramarae & Taylor, 1993). As recently as 1992, Lee Sproull (quoted in Kramarae & Taylor, 1993) estimated that only 5% of participants in Usenet newsgroups were female. It was not until the rise in popularity of Internet service providers and the introduction of the World Wide Web in the mid-1990s that Internet access became available to a broader demographic spectrum, and that women started going online in numbers similar to men (Herring, 2003a). The history of online discussion forums thus shows that a “democratizing” technology does not automatically result in social equality, and points to the importance of social and cultural factors surrounding technology adoption and use.
What, then, of weblogs? An initial consideration of the demographics of blog authors reveals an apparent paradox. Quantitative studies report as many (or more, depending on what one counts as a blog) female as male blog authors, and as many (or more) young people as adults (Henning, 2003; Orlowski, 2003), suggesting a diverse population of bloggers as regards gender and age representation. At the same time, as will be shown, contemporary discourses about weblogs, such as those propagated through the mainstream media, in scholarly communication, and in weblogs themselves, tend to disproportionately feature adult, male bloggers. This inconsistency led us to ask: what are the actual demographics of blog authors, determined according to what criteria? If significant numbers of female and teen bloggers exist, how can their relative absence from public discourses about weblogs be explained?
In this essay, we draw on methods of content analysis to establish both sides of the paradox, and advance an explanation for it. Specifically, we propose that the apparent gender and age bias in contemporary discourses about weblogs arises in part as a result of focus on a particular blog type, the so-called “filter” blog, which is produced mostly by adult males. We argue that by privileging filter blogs and thereby implicitly evaluating the activities of adult males as more interesting, important and/or newsworthy than those of other blog authors, public discourses about weblogs marginalize the activities of women and teen bloggers, thereby indirectly reproducing societal sexism and ageism, and misrepresenting the fundamental nature of the weblog phenomenon. We conclude by advocating a broader characterization of weblogs that takes into account the activities of a majority of blog authors, and more research on weblogs produced by women and teens.
The remainder of the essay is organized as follows. The next section presents quantitative evidence concerning the gender and age breakdown of contemporary blog authors. Based on this evidence, an interpretive argument is advanced and illustrated with observations from public discourses about blogs in multiple domains. Weblogs produced by women and teens are then considered in their own terms, followed by a discussion and conclusions that explore the implications of the observations presented.
Gender and Age of Blog Authors
Guernsey (2002) claims, on the basis of informal observation, that 40-50% of bloggers are women. At least one report (Orlowski, 2003) goes further, asserting that a majority of bloggers are teenage girls. What percentage of blog authors are females and teens? To address this question, we conducted a gender- and age-focused content analysis of a random sample of 357 blogs collected from the largest available blog tracking site, blo.gs. The site tracks blogs hourly from four sources: antville.org, blogger.com, pitas.com, and weblogs.com (the last of which itself draws from multiple sources). We collected blogs twice, six months apart, as part of a larger ongoing longitudinal analysis of the weblog genre. At the times of our data collection, in March 2003 and in September 2003, the blo.gs site was tracking a total of roughly 350,00 and 700,000 blogs, respectively. We used the site's “random” selection feature to collect two samples from these totals: the first containing 203 blogs, and the second containing 154 blogs.
Our goal in selecting these 357 blogs was to represent clear exemplars of the weblog genre. First, we did not sample from hosting sites such as LiveJournal or DiaryLand, in as much as they self-identified at the time more as journals or diaries than as weblogs.We also excluded blogs with no text in the first entry, blogs that had not been updated within two weeks, and blog software used for non-blog purposes, since relatively few such blogs were identified by the blo.gs random selection feature, and could be assumed to be less prototypical. This resulted in a sample comprised exclusively of active, text-based weblogs.
Gender of blog authors was determined by names, graphical representations (if present), and the content of the blog entries (e.g., reference to “my husband” resulted in a “female” gender classification, assuming other indicators were consistent). Age of blog authors was determined by information explicitly provided by the authors (e.g., in profiles) or inferred from the content of the blog entries (e.g., reference to attending high school resulted in a “teen” age classification). The gender of the blog author was evident in 94%, and the age of the author in 90%, of the blogs in the combined samples.
The results of the analysis of gender and age indicators reveal that the numbers of males and females, and of adults and teens, are roughly equal, especially in the later sub-sample. This is summarized in Table 1 (for gender) and Table 2 (for age).
Table 1. Gender of Blog Authors
Age was coded into two categories for the earlier sample (adult and teen, operationalized as less than 20 years of age). For the later sample, we added an “emerging adult” category for authors between the ages of 20 and 25 (cf. Arnett, 2000), based on our impression after coding the first sample that many “adult” blog authors were in their early 20's.
Table 2. Age of Blog Authors
Males and females are distributed unequally across the age categories, as shown in Figure 1 (for the earlier sample) and Figure 2 (for the later sample). That is, there are more female than male “teens,” and more male than female “adults.” Participation by gender is equal only in the “emerging adult” category in the later sample.
Figure 1. Gender and Age of Blog Authors in March Sample (single-authored blogs)
Figure 2. Gender and Age of Blog Authors in September Sample (single-authored blogs)
There is also a skewed distribution of the gender and age of blog authors in relation to blog type. In a recent study, Herring, Scheidt, Bonus and Wright (2004) found evidence of three basic types of weblogs: the content of filters is external to the blogger (links to world events, online happenings, etc.), while the content of personal journals is internal (the blogger's thoughts and internal workings), and k(nowledge)-logs are repositories of information and observations with a typically technological focus. In the present study, we coded each blog in the sample as journal, filter, k-log, or mixed (a combination of two or all of the first three types). The results for the two sub-samples combined, broken down by age and gender of blog author, are presented in Figure 3.
Figure 3. Blog Type by Gender and Age of Authors (single-authored blogs)
Figure 3 shows that gender and age vary in the sample according to blog type. The journal type is dominated by teen females (and is favored by females in general), whereas adult males predominate in the creation of filter-type (e.g., news and politics-oriented) blogs and k-logs, as well as in the “mixed” category, which necessarily includes either filter or k-log content. At the same time, Figure 3 reveals an overwhelmingly greater frequency of personal journal-type blogs than of any other blog type. At 71% of the total number of blogs in the sample, the personal journal is the most popular type in every demographic category.
The preponderance of personal journals, and the large number of blogs maintained by teenage girls, in particular, are striking given that our sample did not include popular online journal hosting sites such as LiveJournal. Including such sites more than doubles the number of “blogs” available, and increases the number of female and young bloggers. A study released in October 2003 by the Perseus Development Corporation of blogs created on the services Blog-City, BlogSpot, Diaryland, LiveJournal, Pitas, TypePad, Weblogger and Xanga estimated that of 4.12 million hosted blogs, 56% were created by females and 52.8% by people under the age of 20, with an additional 39.6% being created by young adults between the ages of 20 and 29 (Henning, 2003). These data provide further evidence of a correlation between female gender, youth, and the personal journal blog type.
The Discursive Construction of Weblogs
There is thus a relationship between blog type and author demographics. We propose that this relationship sheds light on how weblogs have been discursively constructed--that is, how meanings and values have been assigned to the emergent weblog phenomenon through its invocation in public discourses--and why such constructions favor men. A selective focus on filter-style blogs, and to a lesser extent, k-logs, characterizes mass media reports, scholarship about weblogs, definitions and historical accounts of the weblog phenomenon produced by blog authors (including by women), and patterns of linking and referring within the blogosphere itself, as described below. Since men are more likely to create filter blogs than are women or teens, this selective focus effectively privileges adult male bloggers. In each case, this outcome is mediated by other motivations that are arguably not sexist or ageist in and of themselves, but that reproduce societal sexism and ageism around weblogs as a cultural artifact.
Mass Media Reports
Media reportage about weblogs, even when ostensibly concerned with the phenomenon of blogging in general, tends to focus on adult male weblog authors. To quantify this impression, we conducted an informal content analysis of 16 articles about blogs from mainstream news sources that happened to come across our desks between November 2002 and July 2003. These articles had been collected by or forwarded to the us by colleagues as being of general interest about the weblog phenomenon, before we decided to study gender and age of bloggers, and thus would not be expected to contain any particular gender or age bias. (A list of the articles is included in the Appendix.) The results reveal that:
more males (88%) are mentioned in the articles than females (12%);
males are mentioned multiple times in the same article more often than females;
males are mentioned earlier in the articles than females;
males are more likely to be mentioned by name than females; and
all 94 males mentioned are adults, except for one adolescent male blogger.
The preference to mention adult males is consistent across the articles, regardless of their topical focus. The one exception is an article focused on female weblog authors (Guernsey, 2002), published in the New York Times, which mentions 7 females and 6 males, although all of the bloggers named are adults. With the exception of the New York Times article, none of the articles in the sample mentions the gender or age of the blog authors—rather, adult male bloggers are presented as if they are “typical.” While this sample is admittedly small, informal observation suggests that articles such as these were common around the time we conducted our random blog analysis.
Although they constitute a minority (13%) of blogs, as noted above, filters and k-logs receive the majority of media attention in this sample. Two phenomena that figure repeatedly in the 16 articles are political filters that comment on U.S. aggression in Iraq (so-called “warblogs,” e.g., Ostrom, 2003; Webb, 2003; cf. Cavanaugh, 2002), and Dave Winer's efforts to establish k-logs at Harvard University (e.g., Festa, 2003; Hastings, 2003). It may be that journalists deem filters and k-logs more “newsworthy” in that their content is information in the external world (events, technology developments, etc.; i.e., “hard news”), rather than internal to the blogger (cf. human interest stories and “soft news”; ben-Aaron, 2003). An unintended effect of this practice, however, is to define blogging in terms of the behavior of a minority elite (educated, adult males), while overlooking the reality of the majority of blogs, and in the process, marginalizing the contributions of women and young people—and many men—to the weblog phenomenon.
Scholarship on weblogs is still in its infancy, so there is little published literature as yet. However, some scholarly activities associated with weblogs already show evidence of an adult male bias. Conferences to discuss weblogs have thus far tended to attract more male than female participants. A seminar on blogs organized in the spring of 2003 by Dave Winer at the Harvard Berkman Center was heavily male dominated, judging by photos of the event posted on Dave Winer's blog. At the Association of Internet Researchers (AoIR) conference held in Toronto in October, 2003, twelve out of sixteen papers (75%) presented in the four sessions dedicated to weblogs were authored by men, and males made up roughly 70% of the audience who attended those sessions as well, at a conference that was otherwise more than 50% female. Tellingly, two papers by female scholars analyzing LiveJournal communication (Kendall, 2003; Raynes-Goldie, 2003) were relegated to a separate session, the name of which did not include the word “(web)log.”
The papers about blogs presented at the AoIR conference that are based on empirical observation have tended to focus uncritically on what are, in effect, filter-style blogs. Krishnamurthy (2002) studied discussion on the popular “community blog,” Metafilter, of the terrorist events of September 11, 2001, as an example of online democracy. Delwiche (2003) tracked news stories linked to by blog authors in support of the claim that the blogging “community” is interested in political issues. Park (2003) focused on the four most popular blog authors, assessed by the number of incoming links their blogs receive from other bloggers, whom he characterized as “public intellectuals”; these included Glenn Reynolds of InstaPundit fame along with other authors of political filter-type blogs. Similarly, the 25 “scholars who blog” described by Glenn (2003)—many of them aspiring public intellectuals (and all but two of them men)—produce filter blogs focused on political issues outside and inside academe.
In choosing to focus on filter blogs, Internet scholars are not necessarily intending to privilege adult male blog authors. Rather, such blogs are deemed interesting for their “democratizing,” “socially transformative” potential as alternative news sources (Delwiche, 2003; Krishnamurthy, 2002; cf. Lasica, 2001), whereby individuals with something to say can attract and potentially influence a mass audience. Sometimes, as in the case of the blogs studied by Park (2003), their sheer popularity makes them interesting. In this sense, scholars, like journalists, are mirroring what they observe within the blogosphere itself.
Blog authors themselves contribute unwittingly to creating a hierarchy within the blogosphere with adult males at the top. They do this by linking to “A-list” blogs, which tend overwhelmingly to be filter-type blogs created by men, thereby contributing to these blogs’ perceived popularity and status. The “A-list” blogs, in turn, link mostly to other men’s blogs: in a count of links from the blogrolls of the top ten blogs (as determined by number of incoming links), Ratliff (2003) found that only 16% led to sites of female bloggers. As we have seen, men are more likely than women or teens to comment in their own blogs on political issues. They are also more likely to post entries to public-access group sites such as Metafilter (cf. Krishnamurthy, 2002). Thus male blogs are more likely to be very popular (where popularity is defined in terms of number of incoming links), and males are more likely to frequent popular blogs. To the extent that those who write about blogs focus on those that are most popular or otherwise have the highest public profile, the tendency for men to be featured is partially explained.
Some blog authors also write about blogs, defining and narrating the history of the weblog genre. Defining and historicizing are powerful discursive means of constructing reality, and of de facto exclusion. The filter type plays a central role in definitions and historical accounts of weblogs produced by influential blog authors. Notable among these is Dave Winer, a software developer often credited with creating the first weblog circa 1996: a newspage containing links to information related to his software products. Winer (2002) himself credits Web inventor Tim Berners-Lee with having the first functional weblog—a regularly-updated list of links to new sites on the Web—thereby effectively defining the weblog as link-centered (the definition of the modern filter) from before the time the weblog as such was known. No females are mentioned in Winer's (2002) account of the history of blogging. Rebecca Blood (2002a), a blogger since 1999 and a published authority on weblogs, largely echoes Winer's history, adding an observation about the rise of journal-style blogs, which she suggests were already more numerous than filters by late 1999. However, Blood's account focuses on filter blogs, as more representative of the genre as a whole: “I would go so far as to say that if you are not linking to your primary material when you refer to it—especially when in disagreement—no matter what the format or update frequency of your website, you are not keeping a weblog.” According to this definition, personal journal blogs, many of which contain no links (Herring et al., 2004), are not “weblogs”!
Bloggers such as Winer and Blood (who is female) are presumably not intending to exclude women and youth from the definition of blogging. Rather, they are defining the weblog based on their own activities and those of the people they know, and extrapolating back in time to the antecedents of those activities. In so doing, however, they overlook an important phenomenon that predates Winer's first filter, and in which women and teens play a central role: the online journal.
Online journals, known as such since 1995, are the precursor of the personal journal blog (Herring, 2003b) as well as of journal hosting sites such as LiveJournal. Like journal blogs, they contain self-revealing content, are updated frequently, and tend to present messages in reverse chronological sequence. A number of people who maintained online journals in the mid-1990s have since switched to using blog software, further blurring the distinction between the two. From the outset, online journals, like the tradition of hand-written diaries they draw from, have been associated with women (McNeil, 2003). Flynn (2003) describes the rise of online communities of women journaling about weight loss, illness, pregnancy, child rearing, and other topics of special concern. Women (and men) also journal about events in their everyday lives. This is illustrated for three different journal formats in Figures 4-6.
Figure 4 is a screen shot of the home page of an early online journal created by a female science fiction writer. The first entry introduces the journal, then describes the author's recent activities (which include giving her boss a hat for her baby for Christmas, selling a novella to a publisher, and visiting her parents), followed by a poem “to an old lover.”
Figure 4. An early online journal
Figure 5 shows a journal-style blog from our random blog corpus created by a young married woman. With the exception of the modern two-column format and blog-specific features such as archives and a blogroll on the left, it is functionally and stylistically similar to the early online journal example: the author relates, in chronological order, events from her personal life, including getting her hair cut short, shopping, and socializing with her husband and father-in-law.
Figure 5. A personal journal blog
Figure 6 is the home page of a LiveJournal created and maintained by a teenage girl. It contains features characteristic of the LiveJournal format—a mood indicator, an indicator of the music the author is currently listening to, and a profile of the author with her username (“flickering star”) and an image (in this case, of a night scene). As in the other examples, however, the content is current events in the author's personal life—learning how to play bridge, socializing with friends, plans for future entertainment, and complaints about schoolwork.
Figure 6. A LiveJournal
These three examples, although illustrating different time periods and formats of online journals, are similar in their content; women and girls have led in the creation and use of all three formats. A historical account of weblogs that accorded a central place to personal journals—as their prevalence merits—would thus identify females as the creators, early adopters, and most characteristic current users of weblogs.
While it is beyond the scope of this essay to compare the content and style of journal blogs created by females with those created by males, it is our impression that many similarities exist. Male journalers, who comprise about 40% of journal writers, also write about their personal lives, friends, family, and school or work activities, often in self-revealing ways. In an interview with New York Times reporter Emily Nussbaum (2004), one 15-year-old boy described his online journal as "better than therapy," a way to get out the emotions he thought might get him in trouble if he expressed them in school or at home. This constitutes another reason why a comprehensive analysis of blogging should take online journaling into account. Excluding personal journals—defining them as less important or “not weblogs”—not only minimizes women’s and teens’ contributions to the evolution of blogging, but overlooks broader human motivations underlying the weblog phenomenon.
Women and young people are key actors in the history and present use of weblogs, yet that reality is masked by public discourses about blogging that privilege the activities of a subset of adult male bloggers. In engaging in the practices described in this essay, participants in such discourses do not appear to be seeking consciously to marginalize females and youth. Rather, journalists are following “newsworthy” events, scholars are orienting to the practices of the communities under investigation, bloggers are linking to popular sites, and blog historians are recounting what they know from first-hand experience. At the same time, by privileging filter blogs, public discourses about blogs implicitly evaluate the activities of adult males as more interesting, important and/or newsworthy than those of other blog authors.
Many of these participants (including most of the journalists) are themselves female. Nonetheless, it is hardly a coincidence that all of these practices reinscribe a public valuing of behaviors associated with educated adult (white) males, and render less visible behaviors associated with members of other demographic groups. This outcome is consistent with cultural associations between men and technology, on the one hand (Wajcman, 1991), and between what men do and what is valued by society (the “Androcentric Rule”; Coates, 1993). As Wajcman (p.11) notes, “qualities associated with manliness are almost everywhere more highly regarded than those thought of as womanly.” In this case, discourse practices that construct weblogs as externally-focused, substantive, intellectual, authoritative, and potent (in the sense of both “influential” and “socially transformative”) map readily on to Western cultural notions of white collar masculinity (Connell, 1995), in contrast to the personal, trivial, emotional, and ultimately less important communicative activities associated with women (cf. “gossip”). Such practices work to relegate the participation of women and other groups to a lower status in the technologically-mediated communication environment that is the blogosphere, and more generally, to reinforce the societal status quo.
It remains to explain why weblogs, but not other forms of CMC, have been discursively constructed so as to exclude women and young people from the realm of active participants. In the early days of the Internet, participation by diverse groups was exaggerated, if anything, to show the “democratic” nature of the medium (cf. Herring, 1993). With weblogs, the opposite is the case; actual diversity (and hence evidence of the democratic nature of weblogs) is discursively minimized. Two reasons for this suggest themselves. The first is that the larger context has changed; gender dynamics online now broadly reproduce the offline status quo (Herring, 2003a), making gender equity less of an issue in discourse about the Internet. This may explain why participation in blogging by females and members of other demographic groups merits relatively little comment. The second is that unlike in more interactive forms of CMC, the individual author is central in weblogs, as in traditional forms of print authorship. In keeping with the Androcentric Rule, male authors historically have been more highly valued than female authors (Spender, 1989). Moreover, personal journal-writing, traditionally associated with women, is generally not considered “serious” writing (Culley, 1985; McNeill, 2003). This may explain why weblogs are being discursively constructed so as to exclude women and young people (also assumed to be incapable of “serious” writing), and why journal-style blogs receive little attention despite being the most popular form of blogging for all demographic groups.
We began this essay with an apparent paradox: Why, given that there are many female and teen bloggers, do public discourses about weblogs focus predominantly on adult males? The observation that men are more likely than women and teens to create filter blogs provides a key: It is filter blogs that are privileged, consistent with the notion that the activities of educated, adult males are viewed by society as more interesting and important than those of other demographic groups. However, the blogs featured in contemporary public discourses about blogging are the exception, rather than the rule: all the available evidence suggests that blogs are more commonly a vehicle of personal expression than a means of filtering content on the Web, for all demographic groups including adult males. It follows that more attention needs to be paid to “typical” blogs and the people who create them in order to understand the real motivations, gratifications, and societal effects of this growing practice. This would require advancing a broader conception of weblogs that takes into account the activities of diverse blog authors, considering personal journaling as a human, rather than exclusively a gendered or age-related activity, and conducting research on weblogs produced by women and teens, both for their inherent interest and to determine what differences, if any, exist among groups of bloggers.
Are weblogs inherently “democratizing,” in the sense of giving voice to diverse populations of users? The empirical findings reported for gender and age at the beginning of this essay suggest that they are. Yet public commentators on weblogs, including many bloggers themselves, collude in reproducing gender and age-based hierarchy in the blogosphere, demonstrating once again that even an open access technology—and high hopes for its use—cannot guarantee socially equitable outcomes in a society that continues to embrace hierarchical values.