Anonymity and Self-Disclosure on Weblogs
Bloggers are typically cautious about engaging in self-disclosure because of concerns that what they post may have negative consequences. This article examines the relationship between anonymity (both visual and discursive) and self-disclosure on weblogs through an online survey. The results suggest that increased visual anonymity is not associated with greater self-disclosure, and the findings about the role of discursive anonymity are mixed. Bloggers whose target audience does not include people they know offline report a higher degree of anonymity than those whose audience does. Future studies need to explore the reasons why bloggers visually and discursively identify themselves in particular ways.
Weblogs, commonly known as blogs, have become hugely popular in recent years. They are “frequently modified web pages in which dated entries are listed in reverse chronological sequence” (Herring, Scheidt, Bonus, & Wright, 2004, p. 1; see also Blood, 2002; Nardi, Schiano, & Gumbrecht, 2004; Trammell & Keshelashvili, 2005). As a relatively new form of computer-mediated communication (CMC), blogging serves a variety of purposes. It has been effectively employed within the political arena (Drezner & Farrell, 2004; Lawson-Borders & Kirk, 2005; Trammell, Williams, Postelnicu, & Landreville, 2006), adopted for various educational purposes (Deitering & Huston, 2004; Dron, 2003; Schroeder, 2003; Trammell & Ferdig, 2004), and used for marketing promotion and business development (Dearstyne, 2005; Kelleher & Miller, 2006; Seltzer, 2005). In addition, there are thousands upon thousands of personal blogs through which people share their experiences, thoughts, and feelings. These diary-like, personal blogs (see Blood, 2002; Miura & Yamashita, 2004) are sites of self-disclosure where individuals share observations and thoughts about their online and offline lives. Among various kinds of blogs, personal journals account for 70% of all blogs (Herring et al., 2004). In a recent longitudinal study (Herring, Scheidt, Kouper, & Wright, 2006), most blogs were found to be single-authored personal diaries.
As Serfaty (2004) explains, personal blogs are self-representational writing and are essentially online diaries. Because such writing is posted in a public space and open to potentially thousands of people online, blogging is by no means constrained within one’s personal domain and should be characterized as a social activity (Nardi, Schiano, & Gumbrecht, 2004). However, making one’s experiences and private thoughts publicly accessible necessarily involves some risks and may have real-life consequences (Viégas, 2005). For this reason, many blog services provide users options of anonymity and pseudonymity as a cloak of protection.
This article examines how anonymity influences self-disclosure on blogs and related concerns people have in blogging. Do people who write anonymously reveal more intimate details about their lives than those who use their real names? Do people who have their photos posted online hold back more personal stories than those who do not? Do people have a different sense of anonymity when they target a different audience with their blogs? Do people have any particular concerns in terms of self-disclosure, and if so, what do they do to address such concerns? The findings of an online survey suggest that a higher degree of discursive anonymity may be related to more self-disclosure, and that the target audience determines the amount of anonymity that bloggers perceive themselves to have.
According to Marx (1999), anonymity refers to a state where a person is not identifiable. It is essentially social, requiring “an audience of at least one person” (p. 100). Not surprisingly, there has been much discussion as to what role it plays in communication. For example, anonymity has long been associated with a deindividuating effect, which may lead to unruly behavior (Jessup, Connolly, & Galegher, 1990; Zimbardo, 1969). It is also related to certain pro-social behavior (Johnson & Downing, 1979; Zimbardo, 1969). Although CMC is not the only mode of communication where anonymity is possible—in such traditional communication forms as the letter or even the telephone, it is also possible to avoid identification—computer technology has greatly facilitated anonymity by providing many channels for communication between people separated in time and space.
The issue of anonymity is often privileged in CMC scholarship (e.g., Etzioni & Etzioni, 1999; McKenna & Bargh, 2000; Postmes, Spears, Sakhel, & de Groot, 2001; Turkle, 1995). Some studies have attributed anti-social online behaviors to anonymity (Davis, 2002; Suler & Philips, 1998). Others have shown that it may foster group norm violations (Jessup, Connolly, & Galegher, 1990; Postmes & Spears, 2000). Communication online is also characterized as “hyperpersonal” due in part to anonymity (Nowak, Watt, & Walther, 2005; Walther, 1996). That is, the lack of visual cues allows people to selectively self-present for better impression management.
In both online and offline environments, anonymity can be either visual or discursive (Scott, 2004). Visual anonymity refers to the condition where the physical presence of a message source cannot be detected; discursive anonymity, on the other hand, refers to the condition where verbal communication cannot be attributed to a particular source. In CMC, visual anonymity typically refers to the lack of any visual representation of a person, such as pictures or video clips (Barreto & Ellemers, 2002; Lea, Spears, & de Groot, 2001; Postmes et al., 2001). Discursive anonymity is more complicated. Although the writing itself might reveal to a certain degree something about the message source, in an online environment people usually feel anonymous when their personal information (name, email, gender, location, etc.) is withheld.
Conceptually, anonymity is not dichotomous—it varies in degrees (Anonymous, 1998; Nissenbaum, 1999; Scott, 2004). Certain identity knowledge, for example, can be used to identify a person uniquely (e.g., a legal name and an address), while some other identity knowledge may not be as effectively used to trace a message source (e.g., information about social categorization or a pseudonym). By the same token, a picture, typically coupled with some other identity knowledge, may be enough for complete identification, whereas a photo with a blurred face may provide limited information about the subject. Anonymity is also shaped by the features and affordances of the technology. A tool that allows for anonymous input is different from one that requires a user name; similarly, an option to add a photo or other image is different from tools where such additions are not technologically feasible. Thus, anonymity (visual and discursive) concerns both objective aspects of the medium and subjective perceptions about the degree of anonymity afforded; both are usefully conceived as points along a complex continuum.
Blog services typically offer users a number of options in terms of anonymity. People can choose to be totally anonymous, pseudonymous, or identifiable. For example, when a user starts to set up a new blog on Blogger, s/he is only required to provide a user name, email address, and a display name. Only the display name will be shown on the blog, because it is used to sign blog posts. However, users can choose an easily ignored meaningless sign to avoid providing a name. A pseudonym or a real name is equally acceptable. Provision of any other personal profile information, such as one’s real name, gender, date of birth, location, job, homepage, and interests, is optional. Users can select to share or withhold their profile information as they like. To start a new blog on LiveJournal involves a very similar application process; the only difference is that the user must provide his or her birthday. However, this is part of one’s profile information and by default is not displayed.
Many blogs feature no pictures at all, whereas some blogs use only mug shots of the owners and yet others provide fairly revealing photos about the owners’ offline behavior. Some people who blog about sensitive issues may choose to hide any personal profile information. Others do not care as much and readily post personal information online. There are also those who take a somewhat cautious position and are not willing to volunteer anything more than just a name, which can be their legal name or some chosen pseudonym (Herring et al., 2006). The concepts of visual and discursive anonymities, therefore, are both relevant in the examination of how different degrees of anonymity relate to blogging behavior.
Self-disclosure refers to communication of personal information, thoughts, and feelings to other people (Archer, 1980; Derlega, Metts, Petronio, & Margulis, 1993)—especially in interpersonal relationships (Berg & Archer, 1983; Jourard, 1971; see Laurenceau, Barrett, & Peitromonaco, 1998). However, self-disclosure can also be risky because it may invite ridicule or even rejection, thereby placing the discloser in a socially awkward or vulnerable position (Pennebaker, 1989). Consequently, people are more likely to disclose to a stranger (Derlega & Chaikin, 1977; Rubin, 1975) because they feel secure in that whatever is shared under such circumstances is unlikely to be shared with actual friends and acquaintances who may have some material impact on the discloser’s life.
Although research has shown that self-disclosure plays an important role in intimacy development in interpersonal exchanges (Derlega et al., 1993; Perlman & Fehr, 1987), people also tend to be heavily invested in the enterprise of impression management. Disclosing one’s inner world, where typically there are socially embarrassing or unspeakable facts and morally suspicious or unjustifiable emotions, represents a grave risk of jeopardizing others’ impression of the discloser, thereby diminishing the likelihood of reward, increasing the chance of punishment, and lowering the level of self-esteem (see Leary & Kowalski, 1990; Schlenker, 1980).
Research has shown that online communication lends itself to self-disclosure (Joinson, 2001; McKenna & Bargh, 2000; Tidwell & Walther, 2002; Wallace, 1999). Because of the relative anonymity that online interactions may offer, the risks of self-disclosure may be greatly reduced, and disclosers should be much less fearful of potential condemnation or rejection (McKenna & Bargh, 1998, 2000). In particular, self-disclosure appears to be prevalent in blogs. Blogging about one’s personal life has often been looked upon as keeping a diary in public space (McNeill, 2003; Nardi, Schiano, & Gumbrecht, 2004; Sorapure, 2003).
Miller and Shepherd (2004) contextualize self-disclosure on blogs within a social milieu. Citing Calvert (2000), they argue that the rise of blogs parallels people’s increasing expectation of more information as they progressively lose control of their own personal information. Such an expectation feeds into an interest in other people’s stories. At the same time, many people willingly seize the opportunity for mediated exhibitionism through the use of self-disclosure on blogs, which serves some important purposes: providing better understanding of self, confirming one’s beliefs, offering rewards in social interactions, and manipulating others’ opinions.
Papacharissi (2004), after content-analyzing a random sample of 260 blogs, confirms that blogs serve the purpose of personal expression well and represent an ideal medium for self-disclosure. As in interpersonal relationships, self-disclosure has been found to be an effective tool for self-presentation management and relationship construction on blogs (e.g., Bortree, 2005; Huffaker & Calvert, 2005; Trammell and Keshelashvili, 2005). Van House (2004) proposes that blogs constitute a communicative genre in which self-disclosure has become part of the norms, as blogging is strongly related to individuality, self-representation, and personal relationships.
Nevertheless, not all bloggers self-disclose to the same degree. Some share more private thoughts and intimate details about their lives than others. Such differences may be partly explained by individual differences in personality or temperament (e.g., some bloggers are more open and forthcoming than others). This article, however, explores an alternate explanation through the examination of people’s self-disclosure as it relates to their perceived anonymity online.
In general, people who do not use any personally identifying visual elements on a blog should have a stronger sense of anonymity than those using some sort of identifying image. A doctored photo probably gives a blogger a limited sense of anonymity, whereas a photo that includes some revealing details about one’s life (e.g., a photo taken in one’s home or including one’s family or friends) is likely to give a blogger less anonymity than a mug shot. Nonetheless, it is not clear if visual anonymity is connected to increased self-disclosure on blogs. Thus, we begin with the following question:
RQ1: What is the relationship between visual anonymity and self-disclosure on blogs? Is more visual anonymity related to more self-disclosure?
Discursive anonymity is easy to achieve in an online environment (McKenna & Bargh, 2000), where people may interact with little of their identification information given. Some newsgroups even encourage people to participate anonymously (Donath, 1999). This proves particularly valuable for people with socially stigmatized identities, because they can safely self-disclose and share their emotions with others (McKenna & Bargh, 1998). Bloggers can easily control how much of their identity information to reveal, so that they can be anonymous, partially anonymous, or identifiable. This leads to a second research question:
RQ2: What is the relationship between discursive anonymity and self-disclosure on blogs? Is more discursive anonymity related to more self-disclosure?
Since bloggers have the option to have visual anonymity, discursive anonymity, or both at the same time, it becomes relevant to examine if there is any interaction between different types of anonymity. In a study designed to extend the SIDE model, Merola and Hancock (2005) identified significant interaction between the two types of anonymity. Specifically, the effect of discursive anonymity was observable only when visual anonymity was provided. Thus, we ask a third research question:
RQ3: How do discursive and visual anonymities interact to potentially influence self-disclosure on blogs?
Blogs serve different purposes (Hartelius, 2005; Herring et al., 2004) and therefore have different target audiences. Audience plays an important role related to anonymity; specifically, anonymity is only achieved with the presence of an audience. It is reasonable to assume that a blogger tends to have a stronger sense of anonymity if the blog is designed for an online audience that does not know the author offline; in contrast, such a sense may be greatly diminished or even non-existent when the audience mainly consists of one’s family or friends. Such changes in anonymity may ultimately affect the degree of self-disclosure as well, leading to our next research question:
RQ4: How are different target audiences related to the types of anonymity and range of self-disclosure on blogs?
Finally, there have been some reports about bloggers running into various kinds of trouble in their lives offline because of self-disclosure online: What they write has offended others or revealed information that should not have been shared (Viégas, 2005). Self-disclosure can be a risky enterprise on blogs. Trouble of this nature typically occurs as a result of identification, and anonymity obviously may offer some protection. However, there are other solutions for bloggers as well. For example, on Blogger or LiveJournal people can control access to their blogs and may decide to share access with a few online friends only. In an extreme case, a blogger may grant access to nobody but himself/herself, which effectively turns a blog into an electronic personal diary in a strict sense. This leads to a final research question:
RQ5: To what extent do bloggers worry about negative consequences of their online posts? Is anonymity perceived as a viable solution to such concerns?
Procedure and Participants
An online survey of bloggers was conducted in late 2005. Because defining such a population is difficult given the rapid growth in users, a convenience sample was used. Participants of this survey were recruited mainly via a promotional flyer posted on a number of major blogger forums such as BloggerTalk, Blogger Forum, and Bloggeries. Additionally, announcements were sent out to a number of classes in a large southern U. S. university to encourage student participation. In both cases, the survey was intended only for people who keep a personal diary/journal blog where they write about their experiences, observations, thoughts, and feelings. The survey relied on a snowball sample for additional responses.
A total of 242 people filled out the questionnaire, of which 76 were university students and the remaining 166 were respondents recruited from blogger forums. After initial inspection, 35 responses were discarded because of incomplete answers, leaving 63 valid responses from university students, and 144 from the general public. There are a number of similarities between these two groups. In both, females slightly outnumber males, and over 90% of the subjects have at least some college education. Nearly all the university students were between 18 and 25, and almost half of the general public respondents (48.6%) are also within this age bracket (with most others one age category older). Previous studies have shown that bloggers overall are fairly young. Herring et al. (2004) report that about 40% of the bloggers in their random sample are teenagers, and 60% are adults. Two surveys conducted by the Pew Internet & American Life Project reveal that 48% of the bloggers are under age 30 (Rainie, 2005). Furthermore, these samples are generally experienced Internet users (78.3% online for 6+ years), which is consistent with the Pew studies. Consequently, a combination of the data from these two groups makes sense in that they are similar and help to capture a range of bloggers that appears similar to other studies of bloggers on key demographics.
In the questionnaire (see Appendix), discursive anonymity is assessed by asking what personal profile information survey participants give on their blogs. Six options are provided: offering no personal profile information, using an obvious pseudonym (e.g., “graveyard”), using a non-obvious pseudonym (pen name that looks like it could be someone’s actual name), using a partial real name, using a full real name, or giving one’s full real name plus additional personal profile information. Similarly, visual anonymity is assessed by asking bloggers what type of photos they use. Again, six options are provided: no photos, obviously fake photos (e.g., of an animal or known celebrity), non-obviously fake photos (one that readers might assume is the blogger but is not), distorted actual photos (blurred or altered in some way to hide identity), actual photos such as mug shots, or revealing actual photos about one’s life and even family members and friends. For both types of anonymity, bloggers are asked to consider their primary personal blog. The questionnaire also includes one seven-point Likert-type question about the overall perceived anonymity of the blogger, ranging from total anonymity to total identifiability.
The measure of self-disclosure consists of nine seven-point Likert-type questions, adapted from part of the Interpersonal Competence Questionnaire (ICQ) (Buhrmester, Furman, Wittenberg, & Reis, 1988). Cronbach’s alpha reliability for these items is .833, indicating an acceptable internal consistency. The sum of the scores on these questions is used to represent the degree of self-disclosure. Respondents are also asked in an open-ended question if they have any concerns about blogging, and if they do, what those entail.
Table 1 displays the range of responses related to the type of discursive anonymity used by bloggers, revealing that the largest number of respondents (30%) uses a partial real name.
Table 1. Use of various types of discursive anonymity on blogs
|No identification information||6.3% (13)|
|Obvious pseudonym||27.1% (56)|
|Non-obvious pseudonym||5.3% (11)|
|Partial real name||30.0% (62)|
|Real name||12.6% (26)|
|More than real name||18.8% (39)|
Table 2 displays a similarly diverse set of options related to visual anonymity, revealing that the largest number of bloggers (43.5%) posts revealing photos of themselves and their lives.
Table 2. Use of various types of visual anonymity on blogs
|No photo||25.1% (52)|
|Obviously fake photos||5.3% (11)|
|Non-obviously fake photos||0.0% (0)|
|Partial actual photos||3.9% (8)|
|Actual photos||22.2% (46)|
|Revealing actual photos||43.5% (90)|
Among all the respondents, 19.8% (n = 41) believe they are totally anonymous, and 3.4% (n = 7) think they are totally identifiable. The average score of the overall perceived anonymity is 2.99 (s.d. = 1.63) on a scale between one (total anonymity) and seven (total identifiability), indicating that the respondents consider themselves to be somewhat more anonymous than identifiable on their blogs. The average score for overall self-disclosure is 3.26 (s.d. = 1.17) on a scale ranging from one (no disclosure) to seven (total disclosure).
Anonymities and Self-Disclosure
The first two research questions ask about the relationships between the two types of anonymity and self-disclosure. Two one-way ANOVAs were performed (see Tables 3 and 4). The first looked at effects of visual anonymity on self-disclosure. No significant main effect was observed, F(4, 202) = 1.081, p = .367. A second ANOVA was conducted to explore the effects of discursive anonymity on self-disclosure. There was no significant main effect observed, F(5, 201) = 1.947, p = .088. However, the p value of the findings (.088) is fairly close to the significance threshold of .05. This suggests that, unlike visual anonymity, discursive anonymity may have some influence on how bloggers self-disclose.
Table 3. ANOVA: Effects of visual anonymity on self-disclosure
|Within Groups||22330.51||202||110.547|| |
Table 4. ANOVA: Effects of discursive anonymity on self-disclosure
|Within Groups||21754.91||201||108.233|| |
Given the exploratory nature of this research, the self-disclosure scores were examined more closely. These scores (see Table 5) seem to suggest—with the exception of the extreme group of bloggers who disclose more than their real names—that generally the more identification information given on one’s blog, the less self-disclosive people seem to be. Another one-way ANOVA confirmed significant effects of discursive anonymity on self-disclosure when this extreme group (people who provide names and other identification information) was removed, F(4, 163) = 2.500, p = .045. In this case, an even clearer relationship was noted, with increasing discursive identification related to increasingly less self-disclosure.
Table 5. Means of self-disclosure for bloggers using different types of discursive anonymity
|Partial real name||3.26||1.05|
|Real name and further identification information||3.36||1.31|
Interactions between Visual and Discursive Anonymities
In order to examine possible interactions between visual and discursive anonymities (RQ3), we first collapsed the six categories in each into two groups. Respondents who give no personal information and those who use either an obvious or non-obvious pseudonym were recoded as “discursively anonymous,” and the remaining categories were labeled “discursively identified.” Such a division is justified in that people in the first category essentially provide no profile information, whereas those in the second volunteer at least some identifying information. Similarly, people who do not use any pictures, and who use obviously or non-obviously fake photos, were recoded as “visually anonymous” and all others were labeled “visually identified.” A univariate GLM test indicated that there was no interaction between discursive and visual anonymities on self-disclosure, F(1, 206) = 1.233, p = .268 (see Table 6).
Table 6. GLM: Interaction effects between visual and discursive anonymity on self-disclosure
|Corrected Total||22808.329||206|| |
RQ4 asks about the intended audiences of one’s blog and how it may influence anonymity and self-disclosure. Only 3.4% (n = 7) of the respondents in the survey report that they write for an online audience who does not know him or her offline. In sharp contrast, 51.2% (n = 106) claim that their blogs are for people they know offline. Independent samples t-tests found that there was a significant difference in perceived anonymity between bloggers who target an online audience (M = 1.71, SD = .76) and bloggers whose audience is mainly people they know offline (M = 2.89, SD = 1.44), t(7.106) = −3.699, p = .005. Similarly, there was a difference between those who target an online audience and those whose audience includes people they know both online and offline (M = 3.12, SD = 1.83), t(7.78) = −3.970, p = .001. This implies that perceived anonymity is related to the type of a target audience. When it is an online audience that a blogger does not know offline, his or her sense of anonymity tends to be much stronger (see Table 7).
Table 7. Independent samples t-test: Effect of target audience on perceived anonymity
|Only people known online||7||1.71||0.756|| |
|Only people known offline||106||2.89||1.436|| |
|Only people known online||7||1.71||0.756|| |
|Both people known online and offline||78||3.12||1.830|| |
Next, the relationships between the target audience and the chosen levels of discursive and visual anonymities were explored. Two Chi-Square tests of independence were conducted (see Tables 8 and 9). The target audience seems to have an effect approaching significance for discursive anonymity (χ2 = 3.712, p = .054). In comparison, the audience does not influence the level of visual anonymity that a blogger chooses to have (χ2 = .079, p = .400). A closer look at Tables 8 and 9 reveals that when the audience includes someone they know offline, bloggers use more identifying information (e.g., names); when the audience does not include anyone they know offline, they use greater discursive anonymity.
Table 8. Chi-square test of independence: Effects of target audience on discursive anonymity/identifiability
|Online Audience Only||5||2||7|
Table 9. Chi-square test of independence: Effects of target audience on visual anonymity/identifiability
|Online Audience Only||3||4||7|
Further examination was carried out to explore if there is any difference in self-disclosure among people with a different target audience. There was no significant difference detected between bloggers with an online target audience (M = 24.29, SD = 8.38) and those whose audience includes people they know offline (M = 29.23, SD = 10.40), t(7.188) = −1.241, p = .216 (two-tailed). Therefore, bloggers with an exclusively online target audience do not engage in self-disclosure more than other bloggers.
The final research question is designed to examine the extent that people worry about negative consequences of blogging. Among the 207 survey respondents, 87 expressed concerns that what they blogged could negatively impact their lives. Among these respondents, 23.00% (n = 20) explicitly reported they were afraid that their family members might read their blogs. For example, one respondent wrote: “[I’m afraid] that my parents will eventually discover its existence. They might find some of its content surprising.” Another blogger shared a similar sentiment: “I have concerns about relatives reading stories about them that might not be complimentary.” One respondent reported even having to do something fairly dramatic: “My concern is that my family will get a hold of it. This has already happened once, so I created a new blog and made it totally private, meaning that no one but me can see the entries.” About 16.09% (n = 14) worried that what they wrote might hurt their friends’ or acquaintances’ feelings or they might become socially vulnerable through self-disclosure. For instance, one respondent stated: “The only concerns I have are that I may say something that none of my friends are supposed to know and that they might see it.” Another survey participant confessed: “I worry that someone I write about might read it and find out I like them before I‘m ready for them to know.”
About 8.04% (n = 7) of the respondents mentioned potential damage to their careers if their employers were to be able to associate their blogs with them. One blogger worried about “revealing too much personal info thus facing repercussions about something I write negatively about, in particular my employer.” Another respondent had similar concerns, because “[his] name appears in the blog from time to time.” Many other concerns were presented in more general terms, including that it would be a problem for someone disliked, unwelcome, or with an evil intention to have access to a blog. Table 10 reveals the respondents’ major causes of concern about their blog.
Table 10. Cause for concerns about one’s blog (N = 87)
|Family members may read the blog||23.00% (20)|
|May hurt others’ feelings or become socially vulnerable||16.09% (14)|
|May damage one’s career||8.04% (7)|
Among the respondents who have concerns about potential negative consequences as a result of what they blog, 42.53% (n = 37) choose to censor themselves. That is, they are very selective about what is made available online. “I would like to write more critically of people I know. But I often chicken out,” admitted one respondent. Many bloggers deliberately stay with safe topics. For instance, one blogger reported that he does not share his “deepest thoughts or emotions,” and another intentionally keeps his blog “middle of the road” to avoid any trouble. About 17.24% (n = 15) reported that they use filters for access control.1 Only their trusted associates, oftentimes their closest friends, have access privileges. A few block all access by others and make themselves the sole reader. About 11.49% of respondents (n = 10) seem to have some confidence in anonymity on blogs. Some are intentionally vague in their blog postings. Others did not mention how they cope with the potential problems. Table 11 lists the major ways the respondents address concerns related to what they blog.
Table 11. Ways to address concerns about one’s blog (N = 87)
|Self-imposed censorship||42.53% (37)|
|Access control||17.24% (15)|
|Confidence in anonymity||11.49% (10)|
The survey results reveal that bloggers with more visual anonymity do not self-disclose more. However, level of discursive anonymity is somewhat related to self-disclosure. People who refrain from giving identification information are more likely to self-disclose. This may be due to the fact that name and personal identification information given will more likely place one’s blog within the reach of ever more powerful online search engines, and a blog may easily show up in a search with the blogger’s name as a query. In contrast, bloggers with only their pictures posted may not perceive similar risks, and thus their level of self-disclosure is not as affected (even though images can also increasingly be searched and cross-referenced online).
Interestingly, the effects of discursive anonymity on self-disclosure become particularly obvious when one extreme group is excluded (people with name and other identification information provided). This suggests that people within this group may hold different views about including such information, which may have little to do with self-disclosure or needs for anonymity. They may be more interested in being identified so others know exactly who they are and so they get credit for their ideas; thus, they are willing to share very personal thoughts and stories. In other groups, however, the use of various degrees of anonymity and identifiability may be more closely tied to self-disclosure, as bloggers make more conscious choices about what to reveal or conceal. These people may have a greater need for anonymity and seem more aware of the subtle distinctions in anonymity. Their attitude about blogs protecting their identity appears to be more cautious, and they very likely understand that the choice of their names on their blogs is related to the risks of self-disclosure.
The target audience is related to how much anonymity bloggers perceive themselves to have. Specifically, the bloggers in our study feel more identifiable if the audience includes people they know offline. At the same time, target audience also influences the way posts are written and what information is made available. When a blog is for people one knows offline (e.g., family/friends), the goal may be to identify oneself for them and to gain recognition for one’s ideas from others whose opinions matter to the blogger. Indeed, Nardi, Schiano, and Gumbrecht (2004) conclude that the audience drives various dimensions of blogging behavior.
When a blog is targeted at an audience its author does not know offline, the level of discursive anonymity tends to be stronger and the blogger is less likely to provide identifying information. When a blogger has a target audience of online others, the blog can serve as an emotional outlet; thus, it can be important to keep one’s blog or one’s identity hidden from one’s offline family/friends. One blogger made this very clear: “I wanted to have a secret way of voicing my thoughts to others. It acts as an outlet.” Another offered a more detailed explanation: “I really enjoy keeping a journal. In the past I have kept a journal since I was in 6th grade. It really helps me control my feelings and get my emotions out. Sometimes when I write I read back over it and it seems like the perfect therapy.” In any case, a purely online target audience bears some resemblance to a stranger one sits next to on an airplane, providing a sense of anonymity and protection.
This explanation is further strengthened by the fact that nearly half of the respondents are firmly convinced that their blogs definitely reach their target audience, and an additional one third think that their blogs probably reach their target audience. This is consistent with Viégas’ findings (2005) that bloggers believe they know their core audiences.
The survey also reveals that respondents who feel more anonymous do not necessarily engage in more self-disclosure. Those who want to blow off steam on their blogs would likely seek anonymity for self-disclosure. However, for others in search of something other than catharsis, anonymity is not intentionally sought, and the goal of blogging may not be self-disclosure in the first place. Thus, in addition to the intended audience, the goals of the blog itself are important in understanding possible relationships between anonymity and self-disclosure.
Personal diary/journal blogs are highly social spaces, reserved largely for families and friends offline. Nearly 90% of the respondents identified people they know offline as their main audience. Such blogs represent a new form of interpersonal interaction. The loop of communication may be completed by the posting of responses by readers following an entry on a blog, by face-to-face interactions, or via the use of other media, such as telephone and email. Although online communication offers the possibility of anonymity, people do not necessarily take advantage of it. As an extension of interpersonal communication from the offline to the online world, people may seek relatively little anonymity.
It does not appear that bloggers will post absolutely anything in their blogs. Almost half of the survey respondents were concerned that unlimited self-disclosure might land them in trouble offline. Many expressed their misgivings about the possibility of remaining truly anonymous. One respondent observed, “[I]t is difficult to share the same intimacy as one does in a diary.” Another commented philosophically, “The Internet is a small, small world, and nothing goes unnoticed. If you think you can get away with something, it will almost certainly be a temporary hiding spot.” Bloggers worry that their families may find out what they are doing in their lives, or that people within their social circle may discover certain feelings that they harbor. Such incidents may lead to serious consequences. One respondent stated that, despite the anonymity blogs appear to provide, he would never reveal any true secrets. Another respondent pointed out that a personal diary that can be locked up in one’s drawer is where the real secrets belong. Indeed, anonymity seemed to provide relatively few of the bloggers with a perceived solution for their privacy concerns.
Arguably, the way a blog is organized—with archived entries making it one-stop shopping for anyone interested in what a blogger has written so far—suggests that blogging may be less anonymous by nature than online gaming, newsgroup posting, or anonymous emails. Moreover, gaining access to a message source’s profile information or photos is by no means the only way to identify a message source (Anonymous, 1998; Marx, 2001); people may be able to identify the author of a blog through his or her content or writing style. It is no wonder that self-imposed censorship has become a social norm on blogs (Viégas, 2005). This may explain why even those respondents who provide no profile information and use no photos on their blogs do not report high degrees of perceived anonymity.
This study also suggests that the name “personal journal” is inaccurate and calls for revision, as many bloggers do not share their true feelings and thoughts. The typical approach to blogging, at least among the respondents of this study, may more appropriately be described as lighthearted: Bloggers simply intend to share a little fun and stay in touch. When it comes to self-disclosure, it seems that they are cautious so as to avoid potential backlash.
The categories and degrees of visual and discursive anonymity were found to be generally valid based on the survey. Although no one used any non-obviously fake photos, all other possibilities were represented in the sample by at least 4% of the respondents. This suggests that anonymity and identifiability on blogs (and in similar forms of CMC) are not dichotomous choices where one’s posts are either “identified” or “attributed.” Instead, there is a continuum, with a range of possible options that are all utilized.
This study found that discursive, but not visual, anonymity is related to the amount of self-disclosure, especially for blogs intended for certain audiences. The target audience plays an important role in determining not only how anonymous a blogger feels but also how much personal identification information s/he will provide. Having a whole archive of past posts sitting on the Internet seems to be a source of misgivings for many people about the anonymity that blogs can offer. As a result, most bloggers think twice about disclosing highly personal details and private emotions.
This study is limited in several key respects. Although we asked about anonymity and identification, the research did not take into account why bloggers chose a particular level of discursive or visual anonymity/identifiability. A fake name, for example, may have less to do with anonymity and more to do with appearing cool or following what other bloggers have done. Similarly, a blogger may have no picture posted because it lends a sense of mystique, seems a more mature approach, or a photo is unavailable—none of which necessarily relates to a desire to be anonymous.
Additionally, this study treats bloggers as one homogeneous group. A larger and more diverse subject pool would allow categorization of bloggers along several dimensions (e.g., age, education, gender, motivation). Some findings in the current study already point to the value in grouping blogs based on target audience. Moreover, although anonymity behaviors and perceptions are easily assessed by people and do not necessarily require multiple items, the lack of several multi-item measures on the questionnaire makes it difficult to determine reliability.
In addition to addressing these limitations, several other lines of future work are suggested by this research. An online target audience is related to a higher perceived sense of anonymity; however, what exactly an online audience entails remains unclear. It may include people whom a blogger has no awareness of, or people whom s/he gets to know fairly well online over time but has never met offline, or both. Self-disclosure in these two cases may be very different, and a more careful examination is accordingly required.
Anonymity on blogs calls for a more nuanced perspective. Withholding one’s personal information per se is not a definitive index to anonymity. A blogger may, on the one hand, choose to use no name or pictures, and on the other hand, expend much effort in plugging the blog among friends. Such a situation is not unlike a fairly familiar scenario where one exclusively singles out acquaintances to talk to within a large group. Only bloggers who take measures against identification by all people alike can be characterized as tapping the full potential of anonymity offered by this mode of CMC. These people may only be a relatively small group, but they are where true anonymity resides. To what extent these people self-disclose and use blogs for catharsis remains an interesting question.
Finally, the issue of blog categorization needs to be carefully re-evaluated. How anonymity and self-disclosure operate and interact on blogs likely needs to be part of any such mapping efforts. Although Blood (2002), Herring et al. (2004), and Hartelius (2005) have made serious effort to map out blog types from different perspectives, a more refined taxonomy is needed. Such categorization should consider the ways in which bloggers are identified and anonymous, both visually and discursively.
The authors wish to thank the two anonymous reviewers and the editor of JCMC for their insightful suggestions and comments.
Not all commercial blog services provide access control.
Survey on Blogging Behavior
The past few years have witnessed the dramatic development of blogs. For many people, blogs have already become an important aspect of their lives. This study is designed to examine blogging as a new phenomenon. If you keep at least one personal diary/journal type of blog, please kindly respond to each of the questions below.
1. How long have you been using the Internet?
less than a year
11 years or more
2. How many hours do you spend online every day on average?
less than an hour
between 1 and 2 hours
between 2 and 3 hours
between 3 and 4 hours
between 4 and 5 hours
more than 5 hours
3. How long have you been blogging?
less than 3 months
3 to 6 months
6 months to one year
one to two years
more than two years
4. How many personal diary/journal type of blogs do you maintain?
five or more
If you have more than one blog, the following questions only apply to your primary personal diary/journal blog, that is, the one that you spend most time on.
5. When you post entries in your blog, what name do you use for yourself? (Please only pick one that you use primarily.)
I remain totally anonymous (no name, no personal information at all)
I use an obvious pseudonym (e.g., graveyard or catlover)
I use a non-obvious pseudonym (e.g., John Philips, which sounds like a real name but is not your real name)
I use a partial real name (like your real first name, or last name, or initials only)
I use my full real name
I use my full real name, and also reveal further personal info (like age, location, job etc.)
6. What type of photos do you primarily use when you post entries in your blog? (Please only pick one that you use primarily.)
I do not use any photos.
I use obviously fake photos (e.g., a borrowed picture of celebrities)
I use non-obviously fake photos (readers may mistake them for real pictures of mine)
I use partial actual photos (e.g., my real pictures but with my face doctored or hidden in the shadow)
I use actual photos (real pictures but not quite revealing about my life, e.g., mug shots)
I use revealing actual photos (real pictures about you in your real life, even with my family or friends included)
7. What is the primary target audience of your blog?
people that I know offline (in real life)
people that I don’t know offline (in real life)
both people I know and people you don’t know offline (in real life)
other. Please specify: ___________________
8. Do you believe that your blog reaches your target audience?
Answer the following questions using the scale provided:
9. To what extent are you afraid that you blog may be read by people you know offline (in real life).
Not afraid at all 1 2 3 4 5 6 7 Extremely afraid
10. To what extent do you think you are anonymous on your blogs?
Totally anonymous 1 2 3 4 5 6 7 Totally identifiable
11. To what extent do you write something intimate about yourself in your blog?
Not at all intimate 1 2 3 4 5 6 7 Extremely intimate
12. To what extent do you show your softer, more sensitive side in your blog?
Never 1 2 3 4 5 6 7 Always
13. To what extent do you reveal things about yourself that you are ashamed of in your blog?
Never 1 2 3 4 5 6 7 Always
14. To what extent do you let down your protective “outer shell” in your blog?
Never 1 2 3 4 5 6 7 Always
15. To what extent do you write things that secretly make you feel anxious or afraid in your blog?
Never 1 2 3 4 5 6 7 Always
16. To what extent are you willing to reveal that you dislike someone you know in your blog?
Never 1 2 3 4 5 6 7 Always
17. To what extent are you willing to reveal that you hate someone you know in your blog?
Never 1 2 3 4 5 6 7 Always
18. To what extent are you willing to reveal that you like someone you know in your blog?
Not willing at all 1 2 3 4 5 6 7 Totally identifiable
19. To what extent are you willing to reveal that you love someone you know in your blog?
Not willing at all 1 2 3 4 5 6 7 Totally identifiable
20. Imagine you had kept a personal diary or journal that is exactly the same as your blog, to what extent were you willing to show it to people you know?
Not willing at all 1 2 3 4 5 6 7 Totally identifiable
21. Please tell us about your motivation(s) to start this blog:
22. Do you have any concerns about what you write on your blog? If you do, what are they?
high school or GED
26. Please provide the URL of your blog: ≤http://___≥
About the Authors
Hua Qian is a Ph.D. candidate in the Department of Radio-Television-Film at the University of Texas. His research interests include online interpersonal communication, human-computer interaction, and usability.Address: Department of Radio-Television-Film, University of Texas, 1 University Station A0800, Austin, TX 78712 USA
Craig R. Scott is an associate professor in the School of Communication, Information and Library Studies at Rutgers University. His research examines issues of anonymity, identification, and communication technology use primarily in the workplace and in various collaborating groups.Address: Department of Communication, Rutgers University, New Brunswick, NJ 08901 USA