How Photobloggers are Framing a New Computerization Movement

Eric T. Meyer Howard Rosenbaum Noriko Hara School of Library and Information Science Indiana University, Bloomington, Indiana

Paper presented at: Internet Generations 6.0 Annual Conference of the Association of Internet Researchers (AoIR) Chicago, Illinois, USA Oct. 6-9, 2005 Please request current citable version from corresponding author at [email protected]

Address correspondence to: [email protected], mail: E. Meyer, 791 Union Dr., Indianapolis, IN 46202

looking at any photo in a photoblog, you can generally 1 Introduction navigate directly to the next older or next newer photo. Photoblogs also tend to have archives that allow you to Photoblogging, photo-sharing, and other internet view all the photos sorted by date” activities geared toward enabling photography as a central (wiki.photoblogs.org, 2005). purpose have been growing in number and popularity in The definition also distinguishes between photoblogs the last several years. Photographs have been part of and galleries, which are sorted into albums and categories, websites since the beginning of the graphical internet era rather than viewed as a chronological stream. This ushered in by the Netscape browser. However, only definition, however, ignores gray areas. .com, for recently have sites appeared that are geared toward sharing instance, is a popular photo-sharing site that in some ways photographs as a central concern rather than just one operates like a gallery, with photos organized into albums element of a website really started to spring up. and tagged into categories, and in some ways like a Flickr.com, for instance, was founded in 2004 and photoblog, with photos viewable as a chronological photoblogs.org was founded in 2002. Most other photo- “photostream” that allows comments both from the sharing websites were started in the last 2-3 years (see account owner and from visitors. In addition, flickr Table 1 below for details). supports storing photographs on its site, displaying them Due to their novelty, relatively little research has been on an external and distributing them as feeds of a done into the nature of photo-centric websites. In photostream. Flickr.com and other sites that blur the particular, there has not yet been 1) a systematic boundaries between pure photoblogs and other photo- accounting of the landscape of photo-centric websites and sharing sites point to a methodological decision for a description of how they are being used, 2) an analysis researchers. Is it better to follow the definition offered by comparing photoblogging and more generic photo-sharing photoblogging purists, or to more broadly look at photo- sites, nor 3) a discussion of whether these represent a sharing behaviors on the internet? potential long-term computerization movement in Since the study of photography on the internet and photography or a short-lived fad. This paper is an early digital photography in general is still a relatively untapped attempt to address these issues by presenting some area, for this research we will use a looser definition so preliminary research data on photoblogger’s behaviors. that we 1) do not exclude other photo-sharing behaviors and 2) so that we can hopefully discover differences 2 Photoblogs as between photoblogging in the purest sense and more general practices of photo-sharing on the web. Analytically, we think that the apparently clear distinction A photoblog is defined on the photoblogs.org wiki as: between photoblogging and photo-sharing made by the “A chronological log of photography. Each photo in wiki definition above is actually a gray area. One of the the log has a date associated with it. When you're 1 purposes of this research is to find out just how much collective), relationships (both creating and maintaining), photoblogging and photo-sharing behaviors overlap. and self-expression” (p. 6). Van House et. al. included a Herring et. al. in their recent work characterizing preliminary review of online photo-sharing, and concluded blogging genres, argue that “photo blogs and audio blogs, that “people use online sites for many of the same which were excluded from the present corpus, also deserve purposes (friends and family, vacations, events), but with a to be characterized in their own terms” (Herring, Scheidt, preponderance of fun or art pictures which are likely to be Bonus, & Wright, 2005:164). Their exclusion of meaningful to strangers on public photo sites” (p. 5). They photoblogs is understandable both from a practical and a also noted that while traditional photo-sharing (e.g., theoretical perspective. First, including photoblogs which showing other people your photo albums or slides) are primarily composed of visual images rather than primarily helps to strengthen existing relationships, textual elements would present a practical analytic “apples photobloggers also use photo-sharing to create new social and oranges” problem when analyzing blogs based mainly relationships. These categories are a useful starting point on their textual content as Herring et. al. did. Second, when analyzing how people are using photoblogs. As Van since photoblogs are very different in initial appearance House et. al. acknowledge, “it is possible, even likely, that from text blogs, it is probable that they have some very with changes in technology people will use personal different characteristics from their more textual cousins. photos for purposes other than the current ones” (Van That is one of the ideas that impelled this current research: House et al., 2004:3) such as those mentioned above: to map the general landscape of photoblogs as they now taking snapshots of family, friends, and vacations, and exist to begin to determine what features and behaviors, if commemorating important personal and public events. any, make them unique. In Cohen’s (2005, Forthcoming) article on photoblogger motivations, he interviewed 30 photobloggers and found that photoblogging enabled those already interested in photography as a hobby or avocation to use the technology as a tool for self-motivation and self- accountability. According to Cohen, “photobloggers almost all state that they want, or have wanted to take more photographs, that they like having photographs as depictions of their experiences, but that in the past, this desire has often failed to produce the necessary action of making a lot of photographs…One function of the photoblog, its practitioners say, is to provide motivation for taking photos” (p. 19). Counts & Fellheimer (2004) claim that sharing Figure 1. Types of blogs (Krishnamurthy, 2002) photographs on the web is still somewhat labor intensive, and have begun to examine “lightweight photo sharing”, Krishnamurthy (2002) suggested a typology of blogs which includes cell phone to cell phone sharing and the along two dimensions, a ‘personal versus topical’ axis and experimental PocketPC system described in the article. an ‘individual versus community’ axis (see figure 1). They argue that simplifying photo-sharing tools will help Using Krishnamurthy’s typology, it would appear that people enhance their social presence and visual photoblogs would fall primarily into quadrant I. However, storytelling habits. some of the photo-centric sites, such a Flickr.com, also Other scholars (e.g., Ballagas, Yu, & Hibino, 2004; encourage community participation though the creation of Debaty et al., 2004; Frohlich, Kuchinsky, Pering, Don, & community albums and searching by shared tags. Thus, it Arris, 2002; Wiley, 2005) are also beginning to explore is possible that some photoblogs are located somewhere in other aspects of photoblogs and photo-sharing on the quadrant IV or near the boundary between quadrants IV internet. and II, a quadrant that Herring et. al. (2005) found was rarely represented in their sample of textual blogs. If it is the case that amateur photographers are using community 3 Computerization Movements photo-sharing sites to create virtual camera clubs online, this would indicate that photoblogs of this type are quite The main and deepest analyses of computerization different from textual blogs. movements are found in Kling and Iacono (1994), Iacono Understanding the social uses of photography, and Kling (1996), and Iacono and Kling (2001). A close including photoblogs, is a central concern of Van House reading reveals that their conception is grounded in a et. al. (2004). This study identifies three major social uses sociotechnical concept of computerization and a classic for personal photography: “memory (both personal and sociological concept of a social movement.

2 Kling and Iacono’s conception of a computerization categories. Photoblogs, in their typology, would be movement (CM) has six main characteristics. A CM is a external (to a formal organizational structure), non-market social movement that develops around core ICTs and driven, operating in a narrow domain, stand-alone depends on a “socially constructed process of societal (operates without complex interaction with other mobilization” (Iacono & Kling, 2001:97); its key technologies which would be ‘bundled’), and positive in components include a core ICT; organizational structures nature. (computer movement organizations); a historical trajectory; organized opposition (computerization 4 Methods countermovements); two main types (general and specific); and collective action. In this paper, the This research involved two steps. The first involved collective action framework drawn from social movement understanding the public frame being built about theories is most relevant to the discussion of photoblogs, as discussed above (Iacono & Kling, 2001). photoblogging as a computerization movement. The Using the Nexis database, we searched for all English- collective action framework includes technological action language news articles on photoblogs, which yielded 141 frames and public discourse, ideology and myths, and articles. To get a rough estimate of the types of concepts organizational practice. being discussed in the media, the articles were saved to a One of the fundamental forms of collective social text file that was analyzed using a word frequency action that drives CMs is the shaping of public discourse counting program. The sample contained a total of 3028 about its core ICTs by those who write and speak about terms, of which 51 were relevant to photography and them and those who study them and publish their work. photoblogging. The results are presented and discussed in Central to this shaping is the concept of “framing,” a Tables 2 and 3 below. To augment our understanding of process by which social meanings are constructed, this public media frame, we then did concept analysis of disseminated, and stabilized in discourse. CMs are built the threads being discussed on the forums at around technological action frames (Bijker, 1995) that photoblogs.org and flickr.com. The photoblogs forum describe the socially constructed meanings ascribed to (http://groups.google.com/groups/ photoblogs) has fewer specific technologies. This is a useful idea because discussion threads with fewer replies than the comparable framing “describes the actions and interactions of actors, flickr group (http://www.flickr.com/groups/central/ explaining how they socially construct a technology” discuss), so we used different cutoffs or determining (Bijker, 2001:15526). In this case how the use of digital popular conversations on the two forums. For photography in a form of blogging is portrayed in public photoblogs.org, any thread with greater than 25 replies was discourses, such as news paper articles and trade coded, which resulted in 20 threads and a total of 696 magazines, shapes photoblogging as a CM. individual messages out of the 305 discussion topics in the The ideology embedded in a CM’s technological forum. For flickr, any thread with greater than 100 replies frame can be one of either revolution or reform. What was coded, which resulted in 83 threads and a total of these share is a deeply held belief that the CM’s core ICTs 16,621 individual messages out of the total 2033 can cause fundamental positive social change. discussion threads. The threads were coded for their Consequently, technological utopianism is a central overall content. Messages were not coded individually, element of CMs and provides a set of longer-term goals but counts of replies provide an indication of the relative that people can use to identify with the movement (Iacono popularity of any given thread. The results are reported & Kling, 1996:92). Hara and Rosenbaum (2005) indicate and discussed in Table 4 below. that some CMs are perceived in the public discourse as Second, pilot data on the scope and general behaviors negative or destructive in their intent, such as those on photo-sharing sites were gathered using content involving spammers and virus writers. Although the analytic techniques based on Herring et al. (2005) but majority of CMs are oriented towards positive social modified to more accurately code the specific elements in change, the expansion would allow a more complete photoblogs (see appendix I for details of coding scheme). picture of the conception of CMs. The pilot data consists of a sample of 70 photo-sharing Further, Hara and Rosenbaum (2005) underscore the users collected in September 2005. The sample includes limitation in the conception of computerization movements four sub-samples as follows: 25 randomly selected (CMs) that is a bias towards organizational CMs. For photoblogs selected using the randomizer on example, Iacono and Kling (2001) state that CMs are photoblogs.org1, 10 photoblogs listed as “top” on the developed around “organizational practice.” Certainly photoblogs.org site, 25 randomly selected flickr accounts2, photoblogging is a practice enacted outside of and 10 flickr accounts with at least one photograph with organizations. As a result, Hara and Rosenbaum expand the list of top favorites over 3503. Fewer “top” sites were the conception of CMs and create a typology for coded at both sites because the top ones quickly showed computerization movements involving several binary 3 greater similarity to each other within the type of site and wallpaper on their Windows desktops, is the highest had less variability in the data. ranking and the oldest of the sites. Flickr.com, which is relatively new, is one of the more talked about photo- 5 Data sharing sites, and is one of the sites used for this research. Photoblogs.org, which is often cited by those doing research on photoblogs, isn’t actually a terribly popular Before proceeding it would be useful to discuss briefly 4 the scope and scale of photo-sharing sites on the internet. site based on traffic rankings by Alexa . Nevertheless, it is Table 1 shows characteristics of some popular photo- the best way to find sites that can be more purely defined sharing websites. The sites in this table represent a number as photoblogs rather than as either private or public of popular photography sharing or related types of galleries, such as flickr or fotolog, or more commercially websites with available statistics in alexa.com’s traffic oriented photo printing sites such as Kodak’s rankings. It is clear that webshots, which is primarily a tool kodakgallery.com (formerly Ofoto). for people to have randomly displayed background

Table 1. Some Popular Photo-Sharing Websites

Personal Alexa Photo Public Galleries / Rank Founded Members Photos Blogs Galleries Printing Other Webshots.com 72 1996 — 234 M X Flickr.com 226 2004 775,000 19.5 M X X Ofoto/kodakgallery.com 338 2000 — — X Fotolog.net 499 2002 1,809,991 59 M X X Smugmug.com 2106 2002 51,000 17 M X X Fotopages.com 9174 2003 — — X X Fotothing.com 10438 2004 24,860 130,763 X X Photoblogs.org 47292 2002 11,495 n/a X X Sources: Alexa.com, web.archive.org (Wayback machine), and company information pages of the websites.

4 Photoblogs.org is the largest site dedicated to 5.1 Scope of photoblogs providing links to photoblog sites. It does not provide photo storage like many of the photo-sharing sites do, but As is reflected by the relatively recent founding of instead provides links and rankings for photoblogs that are many popular photo-sharing and photoblogging sites in set up using many of the same tools used by regular text Table 1, photoblogging is one of the more recent activity bloggers, such as MovableType, Blogspot, and so forth. trends on the internet, arising mainly in the last 3-4 years. Figure 2 shows the number of photoblogs registered at As a result, it is not clear yet whether this type of blog is photoblogs.org during the time since its inception in likely to become a (comparatively) long-term web October 2002. Figure 3 shows a screen shot of the top- behavior, or another fad like webcams or begging sites that ranked photoblog on photoblogs.org at the time of this will experience a boom and subsequent bust. This sample. research seeks to describe the current landscape of At one level, the rate of increase in photoblogs is fairly photoblogging and photo-sharing on the internet. steady, and there were 11,495 registered sites as of Regardless of the eventual success of this potential September 16, 2005. By comparison, however, computerization movement, having this early snapshot will Pubsub.com had evidence of over 16 million weblogs as of hopefully prove useful as a basis for further research. the same date, and earlier in the year reported that about 14000 57% of the blogs tracked were considered active blogs (Wyman, 2005). blo.gs claims to track about 20 million 12000 11495 blogs as of September 2005 (blo.gs, 2005). 10000 8432

8000

N 6426

6000 5410

4611 4004 4000 3264 2703 2247 1834 2000 1534 1166 607 287373

0 Jul-03 Jul-04 Jul-05 Apr-03 Oct-03 Apr-04 Oct-04 Apr-05 Jan-03 Jun-03 Jan-04 Jun-04 Jan-05 Jun-05 Feb-03 Mar-03 Feb-04 Mar-04 Feb-05 Mar-05 Nov-02 Dec-02 Aug-03 Sep-03 Nov-03 Dec-03 Aug-04 Sep-04 Nov-04 Dec-04 Aug-05 Sep-05 May-03 May-04 May-05 Month and Year Figure 2. Number of photoblogs registered with photoblogs.org, November 2002-September 2005. Source of data: Wayback Machine5 (The Internet Archive, 2005)

Figure 4. Typical popular flickr account

Expanding beyond pure photoblogs to the more hybrid sites such as flickr.com and fotolog.net reveals a much larger number of users. While flickr does not report membership numbers, fotolog.net which bills itself as a Figure 3. Typical popular photoblog way to “create a free fotolog photo-blog” on the home page reports over 1.8 million members with over 59

5 million photographs loaded into their service. Figure 4 shows a typical flickr photo-sharing account. Table 2. Descriptive photo/blogging related terms in news If these numbers are accurate, photoblogs represent articles only a small fraction of one percent of the , Photography / photoblogging General blogging / internet which in turn only represents a relatively minor type of Word Rank Freq Word Rank Freq photo 29 388 blog 6 971 activity on the internet. Pew/Internet, for instance, reports photos 40 328 web 28 394 that only 7% of internet users in their sample have ever pictures 49 258 site 41 328 created a blog (Pew Internet and American Life Project, camera 57 228 online 48 268 photoblog 60 214 51 255 2005), although when asked simply whether they have digital 69 195 post 71 187 ever posted a photograph to any website, 21% of internet images 133 127 internet 79 173 picture 139 119 blogging 83 169 users respond affirmatively (Lenhart, Horrigan, & Fallows, cameras 183 92 technology 125 130 2004), suggesting that photo-sharing may actually be a photographer 201 83 software 166 98 more widespread trend than the more narrowly defined image 225 74 group 167 97 photograph 235 71 member 169 97 photoblogging. This research will begin to develop and photography 253 66 weblog 202 83 understanding of the measurable ways in which features 257 65 photoblogging and photo-sharing may be similar or website 334 51 different types of online activity. Table 2 shows the rank order (among all words used in 5.2 The framing of photoblogs the sample of news articles) of descriptive words about photography, photoblogging and blogging in general.

30 These words are what one would expect to see in an article Total N = 141 28 in this sample. Table 3, on the other hand, is more

25 tantalizing. The concepts reported in table 3 were selected

21 from the top 500 terms and represent more emotive and 20 potentially meaningful concepts. Note that in general, the terminology is positive or neutral. Words like good, 15 15 14 friend, create, interesting, and creating all have positive 12

NumberArticles of 11 10 connotations. It is also interesting to note that certain 10

7 7 7 popular sites also appear frequently in the media: flickr, 5 Yahoo (which has purchased flickr), and fotolog. 5 4

0 Table 3. Conceptual photo/blogging related terms in news 4th 1st 2nd 3rd 4th 1st 2nd 3rd 4th 1st 2nd 3rd 2002 2003 2004 2005 articles Calendar Quarter Potentially meaningful specific concepts 6 Figure 5: Number of articles in the press on photoblogs Word Rank Freq since first appearance in Nov. 2002 share 131 128 good 132 127 friend 134 124 Kling & Iacono (1994) argue that key ideological work 135 123 beliefs about technologies are communication through life 136 121 free 141 117 public discourses. This language—written and spoken create 180 93 public communications—that is used to construct a art 194 85 technological action frame is disseminated to the target flickr 212 78 yahoo 214 78 audiences of technology users. The media is a key creator allow 260 64 and reporter of these public discourses. Thus, looking at growth 265 63 American 285 58 the media discussion of photoblogs can give us a rough fotolog 290 57 understanding of the public understanding of sharing 303 55 photoblogging, which we can then refine with further diary 307 54 interesting 308 54 research. Figure 5 shows the number of articles published album 312 53 on photoblogs since their first appearance in the sample in graphic 318 52 2002. The selection of relevant words from the top 500 professional 330 51 visual 423 41 terms in the sample, yielded the following 51 terms in amateur 464 37 tables 2 and 3. creating 465 37

6 Table 4. Discussion topics on photo-sharing forums Photoblogs.org Flickr.com Threads Replies Threads Replies Topic Category (samples of types of discussion topics) n % n % n % n % Ethics/values issues 5 25.0 181 26.0 10 12.0 1495 9.0 (censorship, nudity, privacy, copyright, image theft, photo modification, offensive content) Social networks — — — — 10 12.0 1244 7.5 (links via favorites, tags leading to each other, reciprocal commenting behavior, discerning personalities from photostreams, demographics) Technical aspects of the medium 7 35.0 224 32.2 8 9.6 1954 11.8 (terminology, perks, jargon, rules, interface issues, important features) Popularity 2 10.0 88 12.6 7 8.4 1417 8.5 (number of hits/visits to photostream, feedback/lack of feedback, popularity rankings, awards, generosity of rankings) Cameras 2 10.0 74 10.6 3 3.6 402 2.4 (types of cameras, how ‘fancy’ of a camera is needed, camera carrying habits) Social impact — — — — 2 2.4 667 4.0 (addiction, overload and competitiveness) Careers 1 5.0 27 3.9 — — — — (how to profit from hobby) Other: Service issues — — — — 21 25.3 4529 27.2 (Yahoo buying flickr, gifts of ‘pro’ accounts, etc…) Other: Off topic discussions 3 15.0 102 14.7 12 14.5 1896 11.4 (ASL (age, sex, location), political topics, current events (bombings, hurricanes, etc…) Other: Photostream threads — — — — 10 12.0 3017 18.2 (what is your 3rd/7th/42nd most viewed/most favorites photo?, pictures of food, blue pictures) Total 20 100.0 696 100.0 83 100.0 16621 100.0

If this is how the media is developing the frame about features on the site. Flickr allows people to mark their photobloggers, what are photobloggers themselves talking favorite photographs on another user’s account, to about? Both photoblogs.org and flickr have discussion comment on photographs, to build collections of favorite groups (http://www.flickr.com/groups/central/discuss/ and flickr accounts, and to participate in groups that share http://groups.google.com/group/photoblogs, respectively). interests. They also have tags, that allow searching for Not surprisingly, there is much more discussion of topics photographs with the same , or the contribution of such as features of flickr or various blogging software sites photographs to groups that are constituted around those than there is reflexive discussion of “why do we blog”? tags. These features allow members to participate in Nevertheless, there are some interesting topics being groups with people with whom they share interests. discussed on these forums, as reported in table 4. For Discussions of the popularity of photobloggers or of instance, the high frequency of discussion about issues the popularity of specific images make up a related set of related to ethics or value indicates that people are threads. It is somewhat self-evident that people posting interested in these topics, and also potentially worried photographs in public forums on the internet want other about the power of the new technology to have negative people to look at their work. Why do it otherwise? This social impacts. Discussions in this category centered concern is reflected in the discussions of popularity about around the posting of nude photos and whether they numbers of hits/visits, feedback, and popularity ranking. constituted art or pornography, on issues of censorship and It was somewhat surprising to see so little discussion surveillance both by site moderators and by government of cameras and their technical features on these sites. One agents, and concerns about copyright and the theft of an of the discussion groups monitored by one of the authors individual’s images. In some cases, the theft was not of this paper is rec.photo.digital, which is a highly popular theoretical: two discussions began when someone found and very busy group discussing technical issues their photos posted on another person’s site and being about photography and cameras. On photoblogs.org and passed off as the artwork of the alleged image thief. flickr, however, the lack of discussion of similar topics Discussions of social networking are also interesting to gives at least some indication that these photo sharing sites note. On photoblogs.org, where there are fewer tools may be frequented by a different sort of photo hobbyist designed expressly for social networking, there was no than those attracted to other more traditional camera clubs discussion of social networking topics. On flickr, or online discussion groups may be. however, 7.5% of messages in the sample were on this topic, due mainly to the prominent social networking

7 Table 5. General Characteristics of the Sample (n=70)

Photoblogs.org Sample Flicker Sample Combined Sample Random Top All Random Top All Random Top All (n=25) (n=10) (n=35) (n=25) (n=10) (n=35) (n=50) (n=20) (n=70) Characteristic n % n % n % n % n % n % n % n % n % Blog Type Snapshots 7 28 — — 7 20.0 16 64 — — 16 45.7 23 46 — — 23 32.9 Hobby 17 68 10 100 27 77.1 9 36 10 100 19 54.3 26 52 20 100 46 65.7 Filter 1 4 — — 1 2.9 — — — — — — 1 2 — — 1 1.4 Gender Male 15 32 7 70 22 62.9 12 48 9 90 21 60.0 27 54 16 80 43 61.4 Female 2 8 3 30 5 14.3 8 32 1 10 9 25.7 10 20 4 20 14 20.0 Unknown 8 32 — — 8 22.9 5 20 — — 5 14.3 13 26 — — 13 18.6 Age < 21 — — — — — — — — — — — — — — — — 0 0.0 21 – 30 5 20 3 30 8 22.9 4 16 5 50 9 25.7 9 18 8 40 17 24.3 31 – 40 1 4 — — 1 2.9 7 28 1 10 8 22.9 8 16 1 5 9 12.9 40 + — — — — — — 1 4 — — 1 2.9 1 2 — — 1 1.4 Unknown 19 76 7 70 26 74.3 13 52 4 40 17 48.6 32 64 11 55 43 61.4 Region of residence North America* 4 16 8 80 12 34.3 11 44 2 20 13 37.1 15 30 10 50 25 35.7 Europe† 12 48 1 10 13 37.1 2 8 6 60 8 32.9 14 28 7 35 21 30.0 Asia‡ 1 4 — — 1 2.9 1 4 — — 1 2.9 2 4 — — 2 2.9 Australia — — 1 10 1 2.9 — — — — — — — — 1 5 1 1.4 South America§ — — — — — — — — 1 10 1 2.9 — — 1 5 1 1.4 Unknown 8 32 — — 8 22.9 11 44 1 10 12 34.3 19 38 1 5 20 28.6 Name Type Full name 9 36 7 70 16 45.7 5 20 6 60 11 31.4 14 28 13 65 27 38.6 Partial name 5 20 2 20 7 20.0 1 4 2 20 3 8.6 6 12 4 20 10 14.3 Pseudonym 3 12 — — 3 8.6 17 68 2 20 19 54.3 20 40 2 10 22 31.4 None 8 32 1 10 9 25.7 2 8 — — 2 5.7 10 20 1 5 11 15.7 Sells photos from site 1 4 4 40 5 14.3 — — 1 10 1 2.9 1 2 5 25 6 8.6 E-mail address listed on site 7 28 7 70 14 40.0 — — — — — — Allows comments about photos# 16 64 4 40 20 57.1 25 100 10 100 35 100.0 Archives of site’s photos# 17 68 10 100 27 77.1 25 100 10 100 35 100.0 Advertisements shown on site 1 4 1 10 2 5.7 — — — — — — n/a^ Links to other photoblogs 13 52 5 50 18 51.4 — — — — — — Links to other photo information 6 24 5 50 11 31.4 — — — — — — Links to other non-photo content 9 36 4 40 13 37.1 — — — — — — # Feature on Flickr, so Flickr users don’t have a choice ^ Total cells not computed when entire category present or missing on Flickr.

* North America includes USA (n=20) and Canada (n=5). † Europe includes Germany (n=5), the United Kingdom (n=5), Sweden (n=2), and one case each from France, Hungary, Iceland, Italy, Poland, Portugal, Romania, Spain and Switzerland. ‡ Asia includes one case each from Malaysia and an unspecified Asian country. § South America includes one case from Argentina. 8 to be even more internationally diverse, to see if different 5.3 Activity on photo-sharing sites services are being used more heavily in other non-Western countries. Next, we turn to the data from the main portion of this The language of this sample, on the other hand, was study, which involved visiting and doing content analysis primarily English (table 6). This is despite the fact that of 70 photo-sharing sites, 35 each from photoblogs.org and non-English sites were not excluded from the sample. flickr.com divided into 25 randomly selected sites at each Since we were coding mainly for structural features and and the 10 top-rated sites at each. visual information, we were able to get everything we Several demographic points of interest arise in the needed from most sites using a crude translation such as data. While the dataset is predominately male (61.4% to that available through Google.com. This is a strength of 20.0% female), 18.6% do not identify their gender in any this study over blog studies that have coded only English way. If this is due to concerns about security, it is possible textual content, since other languages were not a rule out. that many of these unidentifiable gendered people are female, which would make the gender ration much less Table 6. Language of Blog lop-sided. Another privacy issue involves e-mail Language n % English 61 87.1 addresses: even though the media perception of German 3 4.3 photobloggers is that they are in some ways attention Italian 1 1.4 hungry, only 40% of the photographers in the Polish 1 1.4 Portuguese 1 1.4 photoblogs.org sample provided an e-mail address as a Spanish 1 1.4 means of contacting them, while 60% did not, and no Swedish 1 1.4 flickr users had an e-mail as a way of contacting them (due Unknown 1 1.4 at least partly to the design of the interface). Nevertheless, the idea that bloggers seek feedback is somewhat in One other demographic piece of information gathered question if there is no easy way to contact them. was the occupation of the blog’s author if provided. Of the Additionally, only 57% of photoblogs in the sample of 70, only 24 (34.3%) had an identifiable photoblogs.org sample even provided a way to comment occupation shown on their site. Of these, none had more on photos, further limiting the ability to get or give than two specific instances, although some could be feedback. grouped into small categories. The sample included six Also, names were much more commonly shown on the artists (art director, artist, designer, photographer), six websites of the top photobloggers than the randomly computer professionals (engineer, IT manager, information selected ones. The top-rated portion of the sample had full security, web designer), and a variety of other names available 65% of the time, compared to only 28% miscellaneous occupations (secretary, architect, laborer, of the randomly selected sample. professor, musician, screenwriter, homemaker, student, Age is also of interest in the table. None of the sample tourism, trader). Thus, while half of the sample was that had ages listed on their sites were under 21 or over 50. engaged with either art or technology as a profession, the While the majority did not list their age, of those who did other half was not working in either art or technology, and the overwhelming number were in their 20s and 30s. instead was engaged in photo-sharing purely as an In addition, the region of residence was not always avocation. able to be ascertained from these site, with about one-third When we compare posting behavior between identifiably North American in origin, and nearly as many randomly selected blogs/flickr accounts to the top from Europe and from the U.S. and Canada. Future work blogs/flickr accounts, we see the trends reported in Table with these data includes coding fotolog.net, which appears 7.

9 Table 7. Average and median rates of posting

Photoblogs.org Flickr Source comparison Sample comparison Random Top Random Top All PBlogs flickr Random Top (n=25) (n=10) (n=25) (n=10) (n=70) (n=35) (n=35) (n=50) (n=20) Posts per month overall Mean 37.0 25.7 156.8 71.0 89.5 34.1 * 132.3 * 103.6 52.4 Median 12.5 27.0 67.0 41.5 26.5 Range 1-394 14-31 0-1234 9-259 0-1234 Number of months posting Mean 10.8 21.9 6.0 7.5 9.6 13.7 ** 6.5 ** 8.1 * 13.4 * Median 8.0 21.0 4.0 8.0 8.0 Range 1-42 14-27 1-16 5-10 1-42 Posts in most recent month Mean 18.9 18.0 114.0 70.3 60.1 18.6 * 101.5 * 66.4 44.2 Median 0.0 26.5 7.0 36.0 9.5 Range 0-226 0-31 0-976 14-363 0-976 Posts in first month of posting Mean 24.7 28.6 139.0 26.1 73.0 25.8 106.7 91.1 27.1 Median 13.0 30.0 53.0 14.5 17.5 Range 1-226 6-74 1-1982 1-74 1-1982 Maximum number of posts in a month Mean 84.2 41.6 317.0 172.1 187.4 73.1 * 275.6 * 213.5 29.7 Median 19.5 32.0 100.0 76.0 54.0 Range 5-713 31-74 1-1982 15-675 1-1982 Days between most recent posts Mean 10.2 1.4 2.1 1.7 4.8 7.8 2.0 6.0 1.6 Median 2.0 1.0 0.0 1.0 1.0 Range 0-165 0-3 0-20 0-7 0-165 Significance: ** p < .01, * p < .05 In table 7, the means and median rates of posting for than comparing a long-experienced poster to a newbie. In the different subsets of the sample are reported. Several this figure, it is clear that the top sites are more consistent data instances are worth pointing out. First, in general the and their average posting rates slightly increase over time. top sites, particularly the top pure photoblogging sites, are The random sites, on the other hand, are very inconsistent much more consistent in their posting patterns over time. and their average rates decline over time. Also, on the Note that when looking at posts in the most recent month bottom right chart, the drop-off of random accounts is before their site was sampled, the median for top much steeper over time than the rate among top accounts. photobloggers was 26.5 with a range of 0-31, nearly at the This pattern of people adopting, bingeing, and then “ideal” rate of one fresh photograph per day in the wiki abandoning their photoblogs or photostreams is worth definition of photoblogs. The median random additional study in the future. photoblogger, on the other hand, did not post any photos in One trend that became evident during coding include the previous month, even when they did not have an that although the “official” definition of a photoblog abandoned site, and the range for all random sites was implies daily posting, very few blogs or flickr accounts from 0-363, which is very large. actually follow that pattern. In fact, only two blogs in the Statistically significant differences in the above table entire sample, both from the top-ranked photoblogs.org include the difference in overall posts per month, number group, regularly posted one photo per day, month in and of months posting, the number of posts in the most recent month out for an extended period. Several reasons are month, and the maximum number of posts in a month possible, all of which will require additional followup when comparing photoblogs.org sites and flickr accounts. interviews with bloggers to gain more insight: perhaps the In general, the photobloggers post fewer photos, have top bloggers get enough positive feedback to make the posted for a longer period of time, and are much less likely effort worthwhile while less popular bloggers drop off in to have a posting ‘binge’ in a month with a large number their posting rates after receiving little feedback; of maximum postings. For comparison, note that the range alternatively, certain photographers may be using for the maximum posts in a month on random flickr sites is photoblogging in ways suggested by Cohen (2005, as high as 1982 photos. Forthcoming)—as an incentive to create photographs more These data are displayed somewhat differently in regularly, but that these types are in the minority. It is also figure 6, where they are normalized by looking at the possible that it is just too much work for most people. number of posts per month as an average of that photoblogger’s maximum monthly posts. This gives some indication of consistency over time, and compares people in their first month of posting to others in their first month of posting, and so on, which allows a better comparison 10 Figure 6: Posts per month as average percentage of maximum monthly posts

0.8 30

0.7 25

0.6

20 0.5

0.4 15

0.3 10 Number of sites posting by month . month by posting of sites Number 0.2

5

0.1 Average posting rate as percentage of maximum monthly posting .

0 0 M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 Month Month Photoblogs - Random Photoblogs - Top Photoblogs - Random Photoblogs - Top Flickr - Random Flickr-Top Flickr - Random Flickr - Top

0.8 50

45 0.7

40 0.6 35

0.5 30

0.4 25

20 0.3

15 Number of sites posting by month . 0.2 10

0.1 5 Average posting Average as percentagerate of maximum monthly posting .

0 0 M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 M1 M2 M3 M4 M5 M6 M7 M8 M9 M10 M11 M12 Length of time posting to account in months Top sites Length of time posting to account in months Top sites Random sites Random sites

11 Table 7: First images from each site at the time of the sample

12

Table 7 provides a visual impression of the ratio of but digital cameras blur that line, and it is possible that snapshot-type sites in the sample to those made by artistic uses for the internet will blur the line more. hobbyists that appear to be striving to be artistic. Certainly, dividing the sites into snapshot type sites versus 6 Conclusion artistic ones is somewhat subjective, but when coding we had very little trouble separating the two from each other. In this paper, we have attempted to lay out the general It would be interesting for future research to contact site frames being built around photoblogging, and to determine owners to see if their perceptions of their own sites match the general landscape of some photo-sharing sites on the those of the coders. internet. In doing so, several themes with regard to Regardless of the subjectivity, it is interesting to note photoblogging and photo-sharing as a potential that 100% of the top sites in the sample contained mainly computerization movement are now clearer. First, photographs of an artistic quality, while a large percentage photobloggers, based on this data, appear to fit with the of randomly selected sites were simply snapshot sites. types of computerization movements suggested by Hara & That 64% of flickr sites contain mainly snapshots while Robenbaum (2005). However, it appears that rather than a only 28% of photoblogs.org sites is probably at least partly large scale CM which will affect society broadly, it is quite due to ease of entry into the medium: getting started on likely based on current adoption numbers that flickr takes little technical skill and is quite quick, whereas photoblogging will become a narrow CM, affecting a setting up blogging software is still somewhat less intuitive smaller segment of society since its current trajectory does for most users. Furthermore, photoblogs.org requires that not suggest a mass movement. Furthermore, the the photoblogger actually register their external blogs on differential posting rates over time between top-rated the site, which is one more step that may limit the number photobloggers and randomly selected ones suggest that of people who primarily have snapshots from registering. photoblogging may become a specialized community of It is interesting that photoblogs are heavily “art” practice over time. Photoblogging may end up having photography sites. Although the popular press recently high utility and long-term meaning for this community of has been discussing such topics as “citizen journalist” sites practice, but have much less impact on general lay in the wake of the New Orleans 2005 hurricane Katrina, photographers. these sorts of sites did not appear in this sample. There Future work will include expanding this sample to were some examples of pages showing essentially “here include larger n’s, a larger variety of sites as sources of are my friends partying” on flickr. Still, the fact that sites samples, and doing personal ethnographic interviews with striving to be artistic are so well represented is interesting. photobloggers to get their perspectives on the patterns in People generally think of internet as technology, not art, this data. 7 References Cited

Abernathy, K. (2003). Global Diffusion of the Internet II: National Differences in Web Site Connectivity. Communications of the Association for Information Systems, 12, Article 49. Ballagas, R., Yu, K., & Hibino, S. (2004, Dec. 7). Taking and Sharing Pictures with Phonecams: An Ethnographic Study. Retrieved Sept. 25, 2005, from http://kathrynyu.com/portfolio/247b/phonecams-final-with-cc.pdf Bijker, W. E. (1995). Of Bicycles, Bakelites, and Bulbs: Toward a Theory of Sociotechnical Change. Cambridge, MA: The MIT Press. Bijker, W. E. (2001). Social Construction of Technology. In N. J. Smelser & P. B. Baltes (Eds.), International Encyclopedia of the Social & Behavioral Sciences (Vol. 23, pp. 15522-15527). Oxford: Elsevier Science Ltd. blo.gs. (2005, September 17). blo.gs homepage. Retrieved September 17, 2005, from http://blo.gs/ Cohen, K. (2005, Forthcoming). What Does the Photoblog Want? Media, Culture and Society. Counts, S., & Fellheimer, E. (2004, April 24-29). Supporting Social Presence through Lightweight Photo Sharing On and Off the Desktop. Paper presented at the CHI 2004, Vienna, Austria. Debaty, P., Goddi, P., Gossweiler, R., Rajani, R., Vorbau, A., & Tyler, J. (2004, Oct. 18). Enabling Informal Communication of Digital Stories. Retrieved Sept. 30, 2005, from http://www.hpl.hp.com/techreports/2004/HPL- 2004-180.pdf Frohlich, D., Kuchinsky, A., Pering, C., Don, A., & Arris, S. (2002, Nov. 16-20). Requirements for Photoware. Paper presented at the CSCW'02, New Orleans, LA.

13 Hara, N., & Rosenbaum, H. (2005, March 11-12). On the success and failure of computerization movements. Paper presented at the workshop "Extending the Contributions of Professor Rob Kling to the Analysis of Computerization Movements", Irvine, CA. Herring, S. C., Scheidt, L. A., Bonus, S., & Wright, E. (2005). Weblogs as a bridging genre. Information, Technology and People, 18(2), 142-171. Iacono, S., & Kling, R. (1996). Computerization Movements and Tales of Technological Utopianism. In R. Kling (Ed.), Computerization and Controversy: Value Conflicts and Social Change (2nd ed., pp. 85-105). San Diego, CA: Academic Press. Iacono, S., & Kling, R. (2001). Computerization Movements: The Rise of the Internet and Distant Forms of Work. In J. A. Yates & J. Van Maanen (Eds.), Information Technology and Organizational Transformation: History, Rhetoric and Practice (pp. 93-136). Thousand Oaks, CA: Sage Publications. Kling, R., & Iacono, S. (1994). Computerization Movements and the Mobilization of Support for Computerization. In L. Starr (Ed.), Ecologies of Knowledge: SUNY Press. Krishnamurthy, S. (2002, October). The Multidimensionality of Blog Conversations: The Virtual Enactment of September 11. Paper presented at the Internet Research 3.0, Association of Internet Researchers (AoIR) Annual Meeting, Maastricht, The Netherlands. Lenhart, A., Horrigan, J., & Fallows, D. (2004, Feb. 29). Pew Internet & American Life Project: Content Creation Online. Retrieved Sept. 30, 2005, from http://www.pewinternet.org/PPF/r/113/report_display.asp Pew Internet and American Life Project. (2005, June). Pew/Internet Latest Trends: Usage Over Time Spreadsheet. Retrieved Sept. 30, 2005, from http://www.pewinternet.org/trends/UsageOverTime.xls The Internet Archive. (2005). Internet Archive Wayback Machine for Photoblogs.org. Retrieved September 16, 2005, from http://web.archive.org/web/*/www.photoblogs.org/ Van House, N., Davis, M., Takhteyev, Y., Good, N., Wilhelm, A., & Finn, M. (2004, November 6-10). From “What?” to “Why?”: The Social Uses of Personal Photos. Paper presented at the CSCW'04, Chicago, Illinois. wiki.photoblogs.org. (2005). Photoblog vs. Gallery. Retrieved Feb. 15, 2005, from http://wiki.photoblogs.org/wiki/Photoblog_vs_Gallery Wiley, D. (2005, Aug. 4). Understanding the CC License Selection Behavior of Flickr Users. Retrieved Sept. 30, 2005, from http://wiley.ed.usu.edu/docs/flickr_and_cc.html Wyman, B. (2005, January 18). Over 8 million blogs served. Retrieved September 16, 2005, from http://bobwyman.pubsub.com/main/2005/01/over_8_million_.html

14 8 Appendix: Coding Categories

Variable Name Description/Values BlogID Automatically generated unique key Source Source of blog: (1) photoblogs.org random site, (2) photoblogs.org top site, (3) flickr.com random site, (4) flickr.com top site BlogType Overall type of blog: (0) unknown, (1) snapshots, (2) personal hobby, (3) professional hobby, (4) filter, (5) other. AcquisitionDate Date of access AcquisitionTime Time of access BlogHomeURL URL of blog home page BlogTitle Title of blog BlogSoftware Software used for blog, if known NumberofPeopleWhoThinkFavorite Number of people who call this photo site a favorite BloggerNameType Type of name for blogger: (0) none, (1) pseudonym, (2) first name or nickname, (3) full name, (4) other, (5) first name + initial, (6) initial + last name AuthorGender Gender of blog author: (0) unknown, (1) male, (2) female AuthorAge Age of author: (0) unknown, or code age AuthorOccupation Text field of author occupation, if known AuthorCountry Text field of author’s home country, if known BlogLanguage Language of blog AuthorState State or region within home country, if known PersonalInfoLoc Where the personal info was located: (0) none, (1) homepage, (2) one page away, (3) elsewhere DateofCurrentEntry Date of the most recent photograph posting on the home page TimeofCurrentEntry Time of the most recent photograph posting on the home page DateofPreviousEntry Date of the second most photograph posting entry on the home page TimeofPreviousEntry Time of the second most photograph posting on the home page DateofOldestEntry Date of the oldest photograph posting in archives TimeofOldestEntry Time of the oldest photograph posting in archives NumberofPhotosWithoutDate Total number of photographs not classified by date CameraUsed If identified, the type and model of camera used IncludeExposureInfo Yes/no: Exposure/EXIF information included with photographs TypeofPhotoonHomePage Text description of photo on home page FirstPhoto Image capture of the most recent photograph posted CommentsonEntries Yes/no: Are comments allowed? Search Yes/no: Does the site have a search feature? Calendar Yes/no: Is there a calendar to allow moving to a date? Archives Yes/no: Are there archives of old postings? EmailAddress Yes/no: Is there an e-mail address on the pages? GuestBook Yes/no: Is there a guestbook on the site? MailingList Yes/no: Is there a mailing list signup? OtherCMC Yes/no: Are there other computer-mediated communication channels identified (IM for instance)? SellsPhotos Yes/no: Is there a way to purchase photographs from site? Ads Yes/no: Are their ads on the site? Webring Yes/no: Is there a link to a webring? LinkstoOtherPBlogs Yes/no: Are there links to other photoblogs? ReferrerLinks Yes/no: Are there links to referring sites? LinkstoOtherNonPBlogPhotoContent Yes/no: Are there links to other photography content with aren’t blogs? LinkstoOtherNonPBlogContent Yes/no: Are there links to other non photography content, blogs or otherwise? Note Any additional text notes

PostsPerMonth Subtable MID Unique generated key Month Month of postings (January = 1, February = 2, etc…) Year Year of postings MonthNum The number of the month (e.g, first month of postings in archive = 1, etc…) 15 NumberPosts Number of posts in that month

Interface for coding Database including coded data and / or database interface for coding photoblogs available upon request to academic researchers by contacting corresponding author at [email protected].

Figure 7. Coding interface.

9 Endnotes

1 http://www.photoblogs.org/randomizer/ 2 http://www.krazydad.com/gustavog/FlickRandom.pl 3 http://www.flickr.com/groups/top-f/ 4 Alexa has limits due to collecting their data through the Alexa toolbar, which results in biases towards users with Internet Explorer running on Windows. See Abernathy (2003) for a good discussion of Alexa’s use and limitations for research. 5 While most data points were available in approximately 2 month intervals, the Wayback Machine does not archive recent pages until 6-12 months after they have been accessed to comply with their agreements with Alexa (which is the source of the data) and to allow for processing time. 6 Nexis search: 09/21/2005 search of News, All (English, Full Text) for “photoblog* OR (photo pre/1 blog*). 283 initial results were reduced to 141 by elimination of duplicate articles reprinted in multiple sources, repeated weekly articles from one newspaper asking readers to visit their photoblog, and articles erroneously meeting search criteria by having the word photograph and blog next to each other without referring to photoblogs in any way.

16