Hub and terminal: Developing a method for textual analysis on the World Wide Web
First Monday

Hub and terminal: Developing a method for textual analysis on the World Wide Web by Christopher Paul



Abstract
Although the World Wide Web offers a wealth of texts for analysis, the design of the Web opens questions about where potential research objects begin and end. This paper focuses on how the links underlying the Web can both facilitate and deny connections to other sites, which, when viewing the texts based on their interconnectedness, gives scholars a means by which to constitute a text to analyze while preserving the dynamism of the links within a Web site. Two terms are developed to address the dynamics of different sites, with “hub” and “terminal” design becoming metaphors for the ways in which three online sites engage in the process of (dis)connecting themselves to the Web at large. Using the metaphor of hub and terminal design for analysis of Web texts offers scholars a means by which to stabilize a Web site and offer a platform for textual analysis.

Contents

From text to link
Hubs and terminal sites defined
Online news of record: NYTimes.com
NYTimes.com as an emergent hub
A central site of conservatism: Townhall.com
Spreading a message through a hub
Loyalty through humor: Penny Arcade
A hub of trust
Determining textuality

 


 

Hypertexts change boundaries among texts, creating connections and blurring lines between seemingly discrete text chunks. Hypertexts, both online and off, are designed to privilege links among texts, in a manner that calls another text into the body of the original (cf. Morgan, 1999; Carter, 2000; Miles, 2001; Shields, 2000). Building on the ideas of early systems of linking demonstrated in ideas like Vannevar Bush’s Memex and Ted Nelson’s Xanadu, the World Wide Web exists as a massive applied hypertextual system with links that can seem to endlessly span from “text” to “text.” However, hypertexts also offer the opportunity to build a closed system (cf. Aarseth, 1997); and this ability to have both highly connected and isolated texts side by side can be seen in practice on the Web (cf. Pang, 1998). This dichotomy, between openness and closed texts, makes textual analysis of Web based texts quite difficult because links can work in so many different ways.

Although quantitative research has thrived in new media environments, qualitative research, particularly analysis of texts, faces several obstacles, particularly when analyzing the World Wide Web. Textual analysis typically depends on the analysis of a text of some sort, whether it is the speeches and arguments of traditional rhetorical analyses or a social group as defined in an ethnographic study (cf. Warnick, 2002; Hine, 2007). Because questioning the construction of traditional analytical norms, like authorship, the boundaries of texts, and audience, are central to how the Web is constructed (cf. Moulthrop, 1991; Miles, 2001; Pang, 1998; Morgan, 1999; Ricardo, 1998), qualitative methodologies designed for offline use simply should not be applied to the hypertexts of the Web without adapting the tools to properly represent the inherent technological dynamics of the Web.

To this end, textual critics need to be able to develop a reasonable and systematic means by which they can generate a text in order to conduct textual analysis. Initial hypertexts, like those discussed in George Landow’s landmark work Hypertext, largely existed in discrete and identifiable groupings with clearly denotated names and structures. Landow’s hypertexts are set up as “‘the object one reads’” and as an “entrance, the magic doorway, into the docuverse” [1]. Questioning the stability of a given text certainly predates Landow and hypertext though, with works like Roland Barthes’ S/Z establishing the permeability of textual boundaries through the analysis of Sarrasine.

However, both Landow and Barthes had something relatively stable from which to start their analysis, whether it is the hypertext systems addressed by Landow or Balzac’s short story [2]. For both, undercutting the stability of the text comes after being able to point out what the text “is.”

In this vein, and with the understanding that instability and challenges to conventional ideas of textuality are built into the structure of hypertext, it is important to establish a means by which textual analysts can identify the initial framework of texts on the Web. If there is a reliably fair means by which to effectively define the bounds of Web sites, textual analysts can then adapt theories established for analyzing other means of communication to the Web, much like quantitative scholars are already doing (cf. Paolillo, 1999).

I think a reasonable means by which to define the bounds of a site is by visualizing a continuum, with heavily connected sites, or hubs, on one end and isolated, terminal sites, on the other. Identifying the relative connectedness of a site requires looking at what an individual site privileges, movement away from the initial site in the case of hubs, or movement restricted to a given site in the case of a terminus. This distinction among sites based on movement and circulation addresses a central difference between Web–based hypertexts and more traditional analytical objects by focusing on the ways in which sites utilizes a key facet of Web–based hypertexts, links.

There are many keys to how hypertexts work, but dynamic movement among texts facilitated by links is a key characteristic of Web sites (cf. Paul, 2005; Shields, 2000). Conceptualizing hubs and terminal sites is not intended to develop an imposed linearity for online reading, but instead to emphasize that it is movement, rather than the presence of multi–linear ways of reading that characterize hypertexts on the Web. As multiple scholars have noted, multi–linearity exists in some print texts (cf. Dalgaard, 2001; DeRose, 1989), and Landow’s work, which largely addresses hypertexts independent of the Web, surely establishes that his conception of non–Web hypertexts encourages multi–linear readings.

Concerns about textuality and how it is constituted in specific technologies are general topics of interest for new media scholars, but a form of categorization organized around a text’s relative connectedness has not been used as a basis for critique [3]. As the Web has developed, clear examples of both hubs and terminal sites exist, and understanding how their links function to move readers on and off a site offers a means by which to identify the bounds of a site and subsequently constitute texts for analysis.

Demonstrating the analytical benefits of using the notion of hub and terminal sites is best seen by engaging in a descriptive analysis of sites on the Web, with a particular eye to how they use hypertext links to create movement. To demonstrate how these metaphors can constitute a text for analysis, I will provide an overview of what hub and terminal sites can look like and then move into three case studies to illustrate how these categories provide insight into sites, while offering ground for critique. However, the advantages of this approach are more clearly seen after a discussion of the traditional foci of qualitative analysis of online texts.

 

++++++++++

From text to link

Computer–mediated communications (CMC) literature has a lengthy history and is an almost overbroad name for a wide range of ways in which people can communicate via computers. Since the early 1990s the World Wide Web has become an increasingly large part of computer–mediated communication, necessitating the development of methods with which to analyze Web sites.

Ananda Mitra and Elisia Cohen offer two perspectives on Web research, arguing that studies either focus on people using the Web, which they summarize in a paragraph, or a second body of research based on “the text exchanged by users of the Web” that is the focus of the bulk of their essay [4]. This latter perspective is predicated on the idea that “the glue that holds the Internet together is the text exchanged by different users of the Internet” [5]. The focus on analysis of the text exchanged or read online is continued is typical of most qualitative analysis in CMC (cf. Baym, 1997; Warnick, 2002; Warnick, et al., 2005), yet changes in Internet technologies offer a new “glue” for qualitative scholars to analyze, the link.

A focus on text exchange in CMC literature is certainly understandable. Early CMC technologies, like USENET and IRC, were almost exclusively text exchanges and analysis of a given text is something that is well grounded in academic research. Nancy Baym’s essay about a soap opera news group is a perfect example of how texts exchanged by users can be the “glue that holds the Internet together.” However, the development of the Web presents a markedly different technology, which problematizes any overarching generalizations about communication on the Internet at large.

Instead of being able to track a clear set of text exchanges among users, as Baym could in her study, scholars are faced with vast numbers of Web pages that may or may not be connected to other sites. As a result, many scholarly studies have pointed out the problems in treating a Web text like a traditional text, then subsequently performing fairly traditional textual analyses of Web sites (cf. Warnick, 2002; Mitra and Cohen, 1999).

An alternate perspective on how to conduct analyses of Web sites is offered by Steven Schneider and Kirsten Foot. Noting that “most content–focused studies of the Web tend to reflect and perpetuate what we believe is an inadequate construction of the Web as merely a collection of texts” [6], they proceed to outline an analytical approach based on identifying “Web spheres.”

They define Web spheres “as not simply a collection of Web sites, but as a hyperlinked set of dynamically–defined digital resources that span multiple Web sites and are deemed relevant, or related, to a central theme or object” [7]. On face, this idea is similar to that of identifying a hub of sites on the Web, but, in practice, I believe the two approaches are markedly different. The central commonality between the two approaches lies in the belief that the links and connections among sites are vitally important in qualitative analyses of the Web.

A Web sphere is designed to study things “as a macro unit of analysis” for things like the 2000 U.S. elections or the 2002 U.S. Presidential elections [8]. Centering on themes, rather than particular sites, Web spheres offers a mode of analysis that seems similar to genre criticism, with a large group of sites identified and subsequently analyzed. However, the approach does little to address the rhetorical approaches of individual Web sites or events where there is no clear sphere that can be identified.

The articulation of Web spheres offers one approach to scholars, but far more qualitative criticism has acknowledged the issues of using traditional tools of analysis without theorizing how the material conditions of the Web are different from those of other media. Without this step, textual critics are left without a tool with which to analyze individual Web sites. However, looking at how hubs and terminal sites can be identified on the Web can help scholars identify a research object for textual analysis, while focusing on the material conditions surrounding the construction of Web sites.

 

++++++++++

Hubs and terminal sites defined

Hubs seek to create movement among texts that were initially designed or are currently hosted in different locations. An initial way to think about a hub is to consider a links and resources or page, which provides several links to similar kinds of sites. In doing this, the “original” site displaces itself in certain ways, promoting movement away from the current center and outward through hyperlinks to another site, or to the outer surface of a wheel. These sites may host multiple voices as different, originally discrete texts are bound together, requiring a significant reconceptualization of what “text” and “author” means for these texts.

Furthermore, a lack of clear boundaries puts an additional burden on scholars to define where a text ends. By offering a live hypertext link within a site, the Web presents unique opportunities, potentially significantly different than the intertextual connections typically found offline; bringing additional elements into the frame of what may have been a discrete research object (cf. Haraway, 1997; Landow, 1992; Bolter, 1991). By creating movement among texts, sites also can enact a sort of ideological triangulation, with additional links adding to a chorus supporting a view to clarify the position of the original text. By enabling connections among and giving voice to multiple texts, messages are changed, requiring critics to adapt when dealing with heavily connected Web sites.

Perhaps the purest example of a hub on the World Wide Web is the search engine. Search engines are designed to move users from one central site to others in order to obtain desired information. Certain engines, like Yahoo! (http://www.yahoo.com/) or Ask Jeeves (http://www.ask.com/) are designed in such a way that they take measures to keep users tied to the home site. Yahoo!’s additional services are intended to do far more than simply give access to Web sites of interest, and Ask Jeeves’ search results are framed with a link back to their home page. On the other hand, Google (http://www.google.com) is designed in a way that heavily emphasizes movement away from its home site, making it a destination, but in most cases a destination only insofar as it is necessary to find information about other sites.

Google debuted later than many search engines, but rose quickly in popularity owing to its user–friendly interface and ability to quickly connect to the kinds of Web sites its users desired. Eschewing the category–based searching Yahoo! was using at the time, Google employed an algorithm based on the hyperlinks to and from sites, ranking links in order of importance on certain issues [9]. When users type in key phrases, the algorithm compares the request to the links in their database to compile a lengthy list of sites, sorted by relevance. In doing this, Google’s method of searching not only is based on hyperlinks, but also on the idea of collections of sites online [10]. Google searches seek to give you the “best” site for the information you are looking for; in other words, the potential center of a hub on that topic.

As an example of what a hub can look like on the Web, Google is a prime choice because it presents little content of its own. Individually, Google is little more than a compiler, even as additional services are added [11]. Visits to Google are designed to find out things about other sites, not to spend a long period of time on Google’s site itself, even though additions like Google Books and Google Maps may increase the time that visitors spend in “Google’s” space. Content on Google is designed to create movement among sites, originally from Google to another site and then perhaps farther, but to come back to Google it is necessary either to use the back button, access the browser’s history, or retype Google’s Web address.

In addition to the central search engine, Google functions as a hub through its use of advertising. The most direct way it does this is through the use of sponsored words on its search engine. When people search for certain terms, corresponding ads appear at the top of the search page that can move users to other sites. However, Google expands this by using its AdSense program to place advertisements on smaller sites. AdSense matches a list of advertisers who have contracted with Google to sites that are interested in hosting advertisements, enabling sites without their own advertising departments to attract ads without having to solicit them [12]. To participate in the program, sites must turn a portion of their Web space over to Google, which then matches the content on the site with complementary advertisers.

In placing these ads, Google takes its role as a hub on the Web beyond the reaches of its own domain and to a variety of sites by adding external links, another way in which Google is predicated on creating movement from site to site. This was also expanded into the ads placed within Google’s Gmail service, as words in e–mail messages are scanned to select and present a network of advertisers that were deemed by a computer algorithm as best fitting the content in a message.

In the narrowest sense, a terminal site has no connections and is limited to a single page. A Web page or Web site without hyperlinks is certainly a terminal site, but a more meaningful definition is achieved by contrasting the idea of an extreme terminal site to that of a hub. The most extreme terminal site is one that does not have any external links. Under this conception, a page could have many links within it, but if all of those links are contained within a closed system designed by the operators of a site, then it is an end, a terminal point. This would include hypertexts like early hypertextual novels, where the allowed hypertext paths were many, but all were predetermined within a closed system, leading to a substantially different reading experience than if the site linked to places that did not originate from the same author(s) or along the same storyline(s) (cf. Landow, 1992; Bolter, 1991; Aarseth, 1997).

Terminal sites tend to be far more “flat” than hubs. They may promote movement, but they do it without the multiple levels of discourse that tend to exist in hubs. This is largely because the moves they make are internal; they create movement around a series of pages that have direct, original relationships among them, whereas a hub can lead to movement that spans far from the site of origin. As terminals tend to have clearer boundaries, they are more likely to have identifiable authors, and the texts may resemble something found off of the Web.

Terminal texts are fairly good matches for traditional tools of analysis because, even if they employ some of the technologies of the Web, they are more likely to do so in a manner that does not challenge fundamental assumptions about how to conduct analysis. Terminal Web sites may include substantive technical developments and investments, however, even if they are closed systems. One of the key ways contemporary terminal Web sites make use of the technologies of the Web is to create movement around a given site, using a central page to provide access to mountains of information contained within the site at large.

There are two broad categories of terminal sites on the Web. The first is sites with limited ability to link elsewhere. These are most often found in the case of pages that were designed in the early stages of the Web when there were few pages with which to link or in the case of pages designed to convey information in a traditional format, rather than create movement among Web sites (photos uploaded onto a Web page, essays originally written for offline consumption, etc.).

The second, far larger group, attempts to forestall movement to other sites either to sell their products online, in the case of companies like amazon.com (http://www.amazon.com), or to serve as another location at which to communicate with their potential customers, as in the case of Microsoft (http://www.microsoft.com). Using what is known in the Internet design business as “stickiness,” these sites use hyperlinks to redirect users around the site, whether it is through suggestions of other books you might like on amazon or product listings and promotional materials available on the Microsoft page that lead to detailed descriptions and similar products. The primary difference between these kinds of sites is one of scale, with the latter group deliberately using technological tools to foreclose external options. However, both groups share a limited ability to move from the current page to one not sponsored by the central site, requiring utilizing the browser’s back button, history of visited sites, or inputting the URL of a new site in order to leave the terminus.

Considering Microsoft as a terminal site helps to elucidate many of the differences between terminal and hub sites. Centered on a homepage, Microsoft’s site exemplifies many of the characteristics of a hub, with its massive number of links and its complex architecture. The vast scope of the site and the need to encompass the wide variety of products ensure a significant use of Web–based technologies, which is only compounded by their need to present themselves as a leading computer company. However, the site also offers a strikingly consistent corporate voice because, even though the site was constructed by a team of many workers, it is written with the corporation as its sole author. In this sense the author can be identified and analyzed with traditional analytical techniques.

Text offers similar potential, because it has its limits and is bound by the domain of Microsoft; even if the number of pages composing the site is vast and regularly updated, there is an end to the site. Microsoft’s site is terminal, because the linked pages are controlled, limited, and designed by the entity offering the links. Obtaining any relevant information requires moving off the home page and through the network of links, but the site sets bounds by which movement terminates. Analyzing parts of the site would require critical analysis of how movement can be promoted within, but the inability to move outside of the realm of Microsoft–authored rhetoric characterizes this as a terminal site.

Beginning to identify the markers of hubs and terminal sites offer critics a measured way of investigating sites, appropriately representing the features of sites while stabilizing texts enough to conduct analysis. Looking more closely at three sites, NYTimes.com, Penny Arcade, and Townhall.com demonstrates how the radically different design philosophies present on the Web require adaptation for critics to constitute a stable text for analysis and why recognizing where a site fits on the hub/terminal continuum makes for a fitting start for textual analysis.

 

++++++++++

Online news of record: NYTimes.com

Representative of one of the longest running offline U.S. media sources, NYTimes.com (http://www.nytimes.com) is an extension of the New York Times Company’s media empire. Structurally, NYTimes.com has shifted its presentation from a site designed to resemble an offline paper, as it was in its original incarnation on the Web, to a markedly different mode of presentation. As currently structured, the homepage of NYTimes.com provides a “center” for the site and employs hyperlinks to create movement to other articles and connections.

A digitized version of the New York Times banner is displayed across the top of the front page, framed by two small ads. Above the banner are recently added tabbed browsing components that offer visitors a connection to specific, Web–only portions of the site. Links offer the opportunity to register or sign in, as well as the option to subscribe to the premium TimesSelect service. If the visitor is automatically logged in via a cookie in their browser, they are greeted by member name and presented with links to the member center and the option to log out.

The bulk of the front page is split into columns, with a frame on the left that links to “sections” of the online paper. As one moves to these links, the sections often are expanded to detailed subsections about the topic that is being viewed, ensuring that the most relevant links are available to users. On other pages on the site, similar sectional links are offered, but they are placed as the top of the page and expanded sectional options for the section can be provided. The center of the front page is used in a fashion similar to the above–the–fold section of offline newspapers. A handful of key articles are briefly described and hyperlinks connect readers to the entire article.

Beyond this handful of articles that could be found online or off, NYTimes.com places things throughout the page that reify its status as an online paper. The central articles are framed by advertising and stock market information to the right and by video and breaking news service headlines underneath. This information is fully integrated into pages online, something that is not paralleled offline. Although the ads are on the margins of the core information, they frame the site and are a stark difference from what could be found in the offline paper, where ads are found within the paper, but not on such prominent pages and rarely integrated so fully into the landscape surrounding the articles.

In the midst of ads and video links, there is a box that offers tabbed links to content from four different sections of the online paper: real estate, autos, job market, and all classifieds. Each selection provides a number of items from that area of the paper, from articles to actual real estate listings, but the use of the tabbed layout enables the online paper to provide more links and information in comparatively less space. This is a newer addition to the site and it adds an element of personalization for each visitor, as they are able to select the section that is of the most interest to them. Opposing this, on the far left side of the page, are navigational links to sections of the online paper. Much like the separate sections of an offline paper, these offer users the ability to move directly to an area of their interest. Unlike the offline paper, an advertisement is included at the top of the list and the first three listings are for sections that are traditionally found deeper in offline papers, including the “job market,” “real estate,” and “autos.” The vast majority of the links are for sections of the traditional paper, although there are also online only sections included for areas like blogs and podcasts.

The motif of more information in less space is continued in the middle of the page, as the home page offers a section called “Inside NYTimes.com.” Unique to the Web site, this feature is composed of a series of links to articles from a number of different sections of the paper. Including both regular articles and TimesSelect content, this advertises a premium subscription, while offering content to all users. Stories are generally promoted with a headline or a picture, but the most innovative use of this area of the home page is the pair of arrows that moves the stories left and right, allowing the site to place far more articles in this section than would fit on any single user’s screen. Furthermore, the section facilitates interactivity, as users are able to scroll through links to find the articles that best fit their interests. This design enables NYTimes.com to offer more links throughout their site and ties additional content to the home page.

Below the bar of articles is the largest mass of information on the home page. The far left column is devoted to advertising, but the next three columns contain headlines from each of the sections of the paper. This concentrated information enables users to scan headlines by section and get a quick overview of what it is the paper, without having to go to each section of the site. The far right side of the page is about the size of two columns and begins with a tabbed group offering the ten most e–mailed articles, the ten most blogged articles, and the ten most searched words or phrases. Each entry offers a link to the relevant article or search and the pane includes a link to the complete list of articles and searches.

Below the links to the most popular information is a box promoting TimesSelect, a premium subscription service. Boxed off from the rest of the page, the TimesSelect feature generally contains a graphic that is tied to a premium article, as well as about three other articles only available through TimesSelect that are promoted through their headline, byline, or a combination of both. The column is rounded out by more links to sections of the online paper, akin to the links on the left of the page, and links to more service oriented sections of the Web site, including information about the site and about the hard–copy newspaper. The final components of the column are links to their RSS feed and NYTimes company ads, largely promoting home delivery of the hard–copy paper.

In addition to the central page, there are two other templates of the site to consider. The first is that of the individual articles. As users select the articles they wish to read, they are taken to a page with a similar feel to that of the front page, but with several important differences. The top of the page is quite similar, with tabbed options and log in information. The New York Times banner is replaced with an appropriate sectional header for the given article and the navigational bar is moved from the left-hand side of the page to the top. A search bar is also included, which can search a section of the paper or the site as a whole and, in the case of articles from the business section, find stock quotes by ticker symbol. Ads are also included at some points in the header, often in the form of a sponsor for searching NYTimes.com.

The right hand portion of the page is consumed by a variety of different links, elements from the home page, and advertisements. The column starts with a link to articles about similar topics, an option to sign up for an e–mailed list of daily headlines, and a large advertisement. The remaining two elements of the column are both borrowed from the front page, a tabbed list of the most popular content and the TimesSelect promotional box. More text–only advertisements are sometimes located beneath the article, under an agreement with Google’s AdSense program, and links to related articles and searches can be found under those ads. The page ends with the “Inside NYTimes.com” bar from the front page and the site’s footer, which includes navigational links to sections of the paper and information about the site.

The final template to consider is the sectional homepages. In many ways they are a combination of the central page and the individual article pages, although they generally use more graphics. The sectional homepages do not tend to be as large as the home page; traveling to them opens up additional subtopics on the navigational bar at the top of the page to enable users to pursue even more specific sectional subdivisions. These sectional pages offer more surfaces for advertising and enable users to browse in a manner akin to a traditional newspaper, but with advantages of being able to look at articles written for a variety of different editions of the hard copy paper.

The electronic subdivisions also allow articles to be placed in multiple sections, as electronic storage is more forgiving than codex in that respect. Effectively, these sections function as sub–hubs, rerouting traffic through areas of interest, while proffering additional surfaces for advertising. Certain elements of the pages are recast, as the most popular box is targeted by section on these pages, but the general layout of the page keeps with the overall theme of the site, providing enough information via links to enable site visitors to move to articles within the sections.

 

++++++++++

NYTimes.com as an emergent hub

Given a general understanding of how the site is laid out, it is possible to conduct an analysis of how the links on NYTimes.com dramatically affect the composition of site. The first area to analyze is how the site uses its links, how it can be seen as developing as a hub. Although NYTimes.com is full of hyperlinks, few of them lead to places on the Web that are not controlled by the New York Times Corporation generally and NYTimes.com specifically.

The primary source of external links stem from various advertisements. Stories on NYTimes.com have recently started to include hypertext links when relevant. Originally the site would include a URL, but not a link, then they shifted to sporadic links, largely in articles with a focus on technology, but most current articles referencing a Web site will now include a link. However, these links are still constrained to those that either serve the larger interests of the NY Times Company or that aid in the reporting of a story related to the Web.

The selection of links is limited, as NYTimes.com is highly unlikely to link to other organizations in order to increase the amount of information available in a story. Just as it would be shocking to see a page of the Washington Post included in the hard copy version of the Times, it would be alarming to move from NYTimes.com to another news–related Web site. Stories may be reprinted or posted on multiple sites, but in the case of the AP and Reuters news services, NYTimes.com chooses to give the services sections on NYTimes.com in which to present their articles (http://www.nytimes.com/pages/reuters/index.html), rather than moving users to the actual Reuters site itself (http://www.reuters.com). By doing this NYTimes.com retains control, enforcing the bounds around its Web site and reinforcing its status as a terminal site.

The ads require some degree of adjustment in order to evaluate whether or not they constitute a hub. Surely these links lead off of NYTimes.com and to somewhere else; in that sense, the advertisements are hub–like. However, the movement is palpably different than the circulation obtain by links predicated on movement. Advertisements create movement that is paid for, contracted, so that users’ attention is sold in exchange for the link. On NYTimes.com, clicking on the ads opens a new window in which the advertiser’s site is displayed. By doing this, NYTimes.com still retains the underlying hold on users, as their window is still present even after closing the advertiser’s window, reemphasizing how NYTimes.com seeks to keep users on its site, rather than moving them freely in a search for information. The only exceptions to this general procedure are for some of the products offered by the New York Times Company, as the user stays within the corporate family.

The links within the articles are designed to increase the ability of NYTimes.com to communicate a particular story. In this case, links are functioning in a manner that parallels a design based on circulation. However, these links still open another window, holding users within the article, rather than creating an unfettered flow of information. This type of link requires an additional command written into the code of the site, instructing the browser to open the link in a new window. In doing this, NYTimes.com makes a deliberative choice, placing emphasis on sharing additional information, while holding users within the realm of the home site.

NYTimes.com is a hybrid, centered between a hub and a terminal site and driven by the commercial aims of the overall corporate structure in which it is designed. Increasingly taking advantage of the hyperlinks enabled by its presence on the Web, the site is becoming more like a hub. However, the initial use of links was exclusively for profit and that drive is still apparent. Furthermore, the insistence on opening a new window prevents users from fully moving from one site to another, placing NYTimes.com in the middle of the hub/terminal continuum.

 

++++++++++

A central site of conservatism: Townhall.com

Originally a project of the Heritage Foundation, Townhall.com is a clearinghouse of conservative information on the Web (http://www.townhall.com). Although Townhall.com produces some opinion columns, the majority of the information available on their site is culled from other conservative sites or as part of their larger corporate partnership with the Salem Communications Corporation. Townhall.com bills itself as a place to find conservative news and information, which makes it different than a site like NYTimes.com, which attempts to pursue objectivity in reporting. Townhall.com is straightforward in their politics, indicating that their site is not an attempt to correct the news media or to give an overall take on the day’s news. Instead, the site is labeled as “conservative news and information” and aims to amplify conservative voices for upcoming election cycles. Building on an ideologically consistent user base, the commercial elements of the site are designed to capitalize on the conservativism espoused by Townhall.com.

Given the design philosophy predicated on gathering the original information of others, two main components are employed to expand beyond the bounds of Townhall.com’s limited production. The first element is straightforward and easily noticeable on the site’s front page: the dozen–plus articles linked to each day from the Townhall.com front page. As a further attempt to compile information on the Web, Townhall.com also offers to e–mail headlines to interested parties. Both of these efforts are designed to make Townhall.com a site where conservatives can go to get information from a number of places without having to check each one on their own.

Another manner in which Townhall.com stretches beyond the bounds of its own site is through its promotion of membership. Conservative sites can join Townhall.com to obtain a number of benefits, largely based on access to the millions of hits obtained by Townhall.com each month. In a letter to prospective members, an associate editor of Townhall.com writes:

By joining the Townhall.com community, your ideas will be in the spotlight. Your Website [sic] will experience increased traffic, lending heightened effectiveness to your efforts and giving you the opportunity to influence a wide audience. (Biddison, http://www.townhall.com/Partners.aspx).

By gaining member sites, Townhall.com shifts focus from its own production to other locations on the Web. This situates Townhall.com as an important node within a hub of conservative thought.

Townhall.com is highly interconnected to other sources of media. A primary way in which this is seen is the transition from the Heritage Foundation to Salem Communications Corporation. This transition switched control of the site from a think tank to a company that is primarily based in conservative talk–radio broadcasting. Acquisition by a talk–radio company benefited both Townhall.com and Salem Communications, as Salem obtained another site at which they could disseminate information and build the brand identities of their top talk–radio talents, including Bill Bennett, Michael Medved, and Hugh Hewitt, but it also gave Townhall.com the primary thing the site needs, content. This changed the source of some of the content on Townhall.com, but the primary sources of information are still culled from the product of others.

The central page is laden with information, primarily based on textual links to articles with advertisements and navigational links framing the page. The top of the page has a banner announcing the site and providing basic navigational links to information about the site and different sectional home pages. There is space for an advertisement directly under the banner and above the majority of content on the page. Part of the banner is dedicated to a sign up for visitors to receive articles and talk–radio content via e–mail.

The page is divided into two primary columns. The left–hand column is larger and contains most of the content on the page. The top story is given and usually presented with a picture and a variety of headlines are listed to the right. As one moves down the page there are links to blogs, today’s opinion pieces and editorial cartoons, or “funnies,” and a poll. Below the opinion pieces of nationally known conservatives there is a space for the three most viewed, most forwarded, and most linked columns to be listed under the header “Your Opinion.” Pull down boxes are provided that give links to both the Townhall partners’s Web site and to various Townhall columnists. The majority of the columnists have radio shows that are broadcast via Salem Communications, while the partner organizations pay to be associated with Townhall.com.

The largest feature on the page is found below these two pull down boxes and is one of the primary things that Townhall.com can offer to partner organizations. Seventeen contemporary political topics are listed, from “A Culture of Life” to “Immigration,” and links are provided to a section page on the topic, as well as to five articles on the topic written by member organizations. This section of the site puts other organizations headlines in front of far more viewers than likely traffic any, or all, of the partner sites and is a clear way in which Townhall.com creates movement around the Web.

The right–hand column consists almost exclusively of promotional content and other advertising. The top of the column is entitled “Talk Radio Today” and promotes the topics discussed by five of Salem Communication’s top talents. The rest of the column pairs the information and content on the left side of the page with a variety of different advertisements for things ranging from the GOP to conservative books and information about how to retire overseas. The only real content on this side of the page is a promotion for Townhall.com’s Action Center. The Action Center is based on online tools used in George W. Bush’s 2004 campaign for President and provides a conduit by which visitors can more easily get in touch with elected representatives or write letters to the editor of various newspapers.

Little of the front page is designed to point users toward information developed by Townhall.com, instead most pushes them to a content produced by other pieces of Salem Communications Corporation, to advertiser’s sites, or to sites maintained by member organizations. The only real option for staying on Townhall.com is to move to the information that is literally about the site and its evolution or to move to the Action Center, which is predicated on communicating with people or entities outside of Townhall.com.

Many of the internal pages on the site are structured to encourage movement among pages on and off of Townhall.com. Several of the sections are designed to connect conservatives “in real life” through access to Townhall.com. In all of these cases users are led to situations in which they are able to link up with thoughts, ideas, and people that, at least in terms of basic orientation, share the same thought process. In all of these cases, Townhall.com takes a role that is largely based on serving as an entry point for users, but quickly pushes them to other sites by gathering links and offering basic information on real world events, rather than offering up original material to attract and maintain attention. Townhall.com flips many traditional notions about Web sites, developing a site based on compiling the information of others, rather than creating their own.

 

++++++++++

Spreading a message through a hub

Townhall.com is clearly a hub, but it fits the category of hub in two discrete ways: by pushing people to other sites on the Web and, in a novel way, by pushing people together outside the bounds of the Web. The purpose of Townhall.com seems to be to connect people and texts together outside of the realm of Townhall.com, to attract them to Townhall.com first and then enable them to go to their radio, another site, location, or place in the real world to find things of interest to conservatives.

In doing this, Townhall.com is performing the most vital aspects of a hub. This is supported by the philosophical introduction to the site as provided on their “About Us” page:

Townhall.com is designed to amplify conservative voices in America’s political debates just as the 2006 and 2008 election cycles begin to heat up.

By uniting the nations’ top conservative radio hosts with their millions of listeners, Townhall.com breaks down the barriers between news and opinion, journalism and political participation — and enables conservatives to participate in the political process with unprecedented ease. (http://www.townhall.com/aboutus.aspx).

Townhall.com sets out to build a community, which it attempts to fulfill by connecting texts and by connecting people. In doing this, Townhall.com can fulfill a role within the conservative political movement that allows sites to come together under a common banner, while creating connections in a manner that might not have happened without Townhall.com.

The clearest way that Townhall.com connects texts together is in the numerous references made to other sites through the listed news articles. Most of the articles on the front page take users to another site, quite possibly one they would not have encountered on their own. These links are propagated through programs like the e–mail alerts advertised on the site and through the heavily promoted connections to talk–radio hosts that take the message off of the Web and into a more established medium of communication. All are efforts to move users away from Townhall.com toward other locations. This is a move indicative of a hub, the value of Townhall.com is dependent on pushing users to other sites.

Efforts to connect people via Townhall.com are slightly more scattered, but are fascinating as they are predicated on taking the messages contained on a Web site and making them “real,” as their success or failure is measured by occasions in the offline world. These efforts, like the “Conservative Calendar” and the “Opportunities for Conservatives” have real, practical impacts.

The desire to expand user’s abilities to connect to the site and each other has been expanded through a blog creation tool that enables site visitors to create their own conservative blog and create a space of their own on the Web. Blogging not only enables people their own page, but a chance to meet with other, similarly motivated people who have sought out the same location for their blog. An additional example of this type of hub can be found in a frequent advertiser, a conservative dating service that attempts to connect conservative singles with one another. Townhall.com cannot be reduced to its connections, but it is certainly based largely on hypertextuality. One could not conduct analysis of the site without analyzing how its connections define it. In this case, without the idea of a hub, it would be quite difficult to determine the bounds of this text in order to analyze it.

Townhall.com uses the Web in an interesting way. Pushing the meaning of a hub to include connectivity created outside of cyberspace, the site seeks to ensure a circulation of their ideology throughout the Web and in real world interactions. Building beyond the connections enabled by NYTimes.com, Townhall.com enables thorough movement to other sites, functionally becoming a center of conservative information online. Their prominent status and the number of connections available through the site mean that users have a reason to keep coming back. This increased loyalty permits complete connections. By trusting their users and focusing on circulation, Townhall.com has a more open feel than NYTimes.com, something that is expanded on in the third site, Penny Arcade.

 

++++++++++

Loyalty through humor: Penny Arcade

Penny Arcade (http://www.penny-arcade.com/), one of the premier comics on the Web, is an enterprise led by four men, one who writes for the site, a graphic artist who draws the comics and occasionally chimes in with his opinions, and two men who handle the business end of things. The brainchild of Mike Krahulik, the artist known as Gabe on the site, and writer Jerry Holkins, who goes by the name of Tycho, Penny Arcade was initially founded as a Web site in 1999, but has gone through multiple business models before attracting Robert Khoo, who helped turn Penny Arcade into a venture that could become their full–time jobs.

Drawing about 175,000 viewers a day, Penny Arcade managed to turn their personal site, a place they used to express their thoughts and feelings about their lives and the gaming industry, into careers, placing the Web site firmly in the realm of the commercial, but simultaneously keeping it personal. As Gabe observed after the birth of his son, Penny Arcade is more than a Web comic; for him it is a diary, in which he can talk about what it was like to be a new father and where Tycho can talk about his outstanding achievements playing a new video game in the waiting room (http://www.penny-arcade.com/2004/09/13). This idiosyncrasy is characteristic of the site, which uses the two men’s opinions and humor to keep users coming back.

Capturing the intent of Penny Arcade is somewhat difficult because there is something extra that makes the site what it is, a dynamic prone to escape simple description. There is an aspect to the Penny Arcade site that makes it more like an experience, increasing the bonds of users and making the site’s producers cult heroes. Gabe and Tycho have a rabid fan base evidenced by the page views, as well as their ability to draw a substantial crowd at the conventions they attend as representatives of the site. Capitalizing on their popularity, Penny Arcade produced their first annual PAX convention in August of 2004, drawing almost 3,000 people to a consumer convention on video games. In the years since, the convention has grown to the point where a larger venue has been needed each year, with 37,000 gamers attending the conference in Seattle in 2007 (http://www.pennyarcadeexpo.com/). PAX, like some of the efforts of Townhall.com, takes the hub off of the Web and uses the Web site as a starting point for like–minded people to meet in the real world.

Penny Arcade features a basic, straightforward layout that focuses attention on the site and what it contains. The top of the page features the Penny Arcade logo centered on the screen and framed by images of Tycho on the left and Gabe on the right. Joining them on the header of the page are links to: the latest comic, the archive, the forums, PA Presents, PAX, and their online store. This is the default header for almost all of the pages on the site, with the exception of the forums. The “comic” link takes users to one of the prime attractions of the site, Gabe’s comic strips, which cover a range of issues, but never stray for too long from gaming, and generally feature the Penny Arcade takes on the industry or chronicle their lives as gamers.

The top of this page features a banner ad and the bottom of the page features the same footer as found on the home page, links to a variety of their sponsors and projects like PAX and Child’s Play, their non–profit organization. The archive tab takes users to a searchable archive of past comics that runs back for years. This enables users to catch up on their Penny Arcade history, but it also is used to cite Penny Arcade elsewhere, as many message board posts on other sites reference Penny Arcade strips, something that would be difficult to do without a searchable archive. An oft cited strip is drawn as an educational filmstrip, with the intent of teaching people how to use punctuation correctly (http://www.penny-arcade.com/comic/2002/10/11). This comic typically shows up in response to the type of poster described in the strip, a teenager who simply does not use punctuation, and when the strip is linked to on a bulletin board, it regularly generates a number of posts from people praising Penny Arcade. This strip is indicative of much of the content on the site as it is harsh, funny, and sarcastic. This is one way in which the readers of Penny Arcade expand its reach, as they effectively colonize other sites on the Web with Penny Arcade’s content.

The header described above frames the site, but the rest of the material rounds out the shape and the form of the central page of Penny Arcade. Presented on top of a blue background surrounded by grey, the pages feature ads, if sold, underneath the page header and on the right–hand side of the page. These ads are fairly large and, on occasion, include movement. Commonly the ads will be drawn by Gabe, which adds a personal touch to the site and makes the advertisements match the overall feel of Penny Arcade. Content on all of the pages is centered and consumes most of the page.

On the central page, Tycho writes an essay that introduces each of the publications, regularly done on Mondays, Wednesdays, and Fridays. Within his posts Tycho will provide the context for the comic strip that has been posted and then engage in stream–of–consciousness ramblings. Generally these essays are targeted at the video game industry, but when appropriate, they take on issues more indicative of life in general. On busier days, additional posts are often added by either Tycho or Gabe to add context to the original posting or offer information about other items that may be of interest to readers. When links are included within the posts, they are designed to carry the user, in the current page, to the new site, forcing them to go away from Penny Arcade to make sense of what is being said and then return to Penny Arcade to complete the essay. The final part of this central posting is a link to the archive, which enables users to move backward to previous posts.

 

++++++++++

A hub of trust

There are three primary ways in which Penny Arcade enacts its status as a hub. The first is through the ads scattered throughout the site. The ads function as hubs, either by opening a new window for the link or by providing the material in the same window. Penny Arcade does not have that many ads, but the combination of banner ads and sponsorship deals with game publishing companies creates a situation in which sponsored links to other sites are integrated into Penny Arcade. As a Web site about video games, there is a natural link to ad content. The ads chosen fit the landscape of the site and form a larger hub that constitutes a particular space for Penny Arcade within the context of the electronic gaming industry. This is different than the ads on NYTimes.com or Townhall, as these ads are fundamentally tied to the content of the site in a manner unlike that of NYTimes.com linking to Tiffany & Co. or Townhall linking to a dating site. The ads on Penny Arcade are contracted, but they are, in effect, more content about gaming or the gaming industry.

Although these advertisements are one way in which Penny Arcade functions as a hub, a more important factor is the way that links are integrated into Tycho and Gabe’s essays on the front page. These links, written directly into posts, offer depth and detail about the subjects being discussed giving the links power to frame all of the content composed by Penny Arcade. However, Penny Arcade uses these links to make their case, to complete their argument. These posts do not seek discussion on a topic; they intend to establish a point, which is often illustrated through an external link. In some cases, the links are informative, offering users information that is not replicated on Penny Arcade, including references that facilitate critiques of sites, ideas, companies, games, and trends in the gaming industry.

The process of reading in this case requires that users follow the links in order to complete the narrative arguments made in the post [13]. After following the link and getting the necessary information, readers must return to Penny Arcade in order to finish the post. Sometimes this process can be repeated three to five times in a single essay to understand the point being made more completely. By using links in this way, Penny Arcade is a hub, but it is a special sort of hub that pushes users externally, while implicitly drawing them back. The external moves are controlled and contextualized, and the move to the linked sites is temporary because a move back to Penny Arcade is needed to finish the story, even if users meander around the Web before returning.

A third key to circulation on Penny Arcade are the various external projects taken on in order to promote specific video games. First completed for the UbiSoft game “Splinter Cell: Pandora Tomorrow,” Gabe and Tycho have composed comics for a variety of games. These books are typically posted a page at a time on the game publisher’s site and can be linked to through Penny Arcade. Generally, each time a new page is available, it is announced on Penny Arcade, but to view the page it is necessary to go to the publisher’s site, creating an ongoing circuit of advertising. A link to the comic then stays in the Penny Arcade Presents page, enabling publishers to maximize their advertising dollars. These comics also range beyond the Web, as they are often one of the things that Penny Arcade hands out at the various conventions they attend. The comics are interesting in their own right and quite worth reading if one is a fan of Penny Arcade, but there is no mistaking that they are a giant, functional advertisement, dependent on establishment of a hub and connecting an advertising vehicle with an intended audience and Penny Arcade.

Penny Arcade is designed to communicate through what they do and through the texts they cite. By including hyperlinks in the postings on Penny Arcade users are required to move throughout the Web in order to understand what is covered in Penny Arcade. To a degree this also requires users of Penny Arcade to follow gaming in general to make sense of what is happening on the site. Most of what is needed is explicit, but there are points at which background information, which is not supplied, is necessary. Overall, this means that Penny Arcade is only part of the story. Simply viewing the posts on the site or sticking to what exists on Penny Arcade misses the dynamic connections made with the rest of the Web. This ignores the ability of a server to get “wanged” by a link from Penny Arcade owing to the immense number of hits a link from a news post on the site can cause. Most wanged sites are smaller and not used to tons of traffic, but the victims include Penny Arcade’s own forums [14]. In these cases the number of users following the links off of Penny Arcade crash a server that is not set up to handle the traffic a link on Penny Arcade often causes. These connections and linkages are a significant part of what constitutes Penny Arcade and makes it distinct. Because links are opened in the same window as Penny Arcade, the designers of the site trust that interested parties will move away from the site and then come back in order to finish the story. Penny Arcade depends on other sites, without links to them; many of the postings would fall flat.

Penny Arcade is certainly a quirky site. The virtual incarnation of two men on a steady diet of video games and comic books, the site is alternately funny, informative, and touching. Using a controlled circulation to move users around the Web and then bring them back to Penny Arcade, there is a reliance on a form of citation and sharing that unique to the Web; Penny Arcade is designed as a hub built on a trust that users will keep coming back to continue the development of messages on the site. Other Web sites are fully integrated into Penny Arcade; they are necessary for the posts on Penny Arcade to make sense. However, it is the hub–like design combined with the presence of the men producing Penny Arcade has turned it into a tremendously popular site and changed it from an enjoyable hobby on the side to an endeavor that pays the bills in its own right.

 

++++++++++

Determining textuality

To conduct any sort of textual criticism, one must have a text. Analyzing a Web site is fundamentally different from analyzing more traditional texts. Instead of grabbing a book, looking at a speech transcript, critically analyzing a film, or even looking at a hypertext contained on a disc, analysis of the World Wide Web requires looking at more than a single Web page. Having a means by which to identify a fair boundary around a site enables a more systematic form of Web criticism that focuses on the importance of the link in Web–based hypertexts.

Each of the three sites discussed above offers a different perspective on how the hub/terminal continuum can function to stabilize a site. NYTimes.com has economic reasons to control users on the site. The terminal aspects of the site’s design are representative of the commercial motives of the site, while the news–reporting elements sometimes lean to a hub–like design. Townhall.com is bolstered by a hub to connect itself to similar ideological counterparts and attempt to spread their message offline. Penny Arcade uses links to offer readers an opportunity to leave the site, trusting that they will come back later. Each site is defined by more than the sum of its pages and charting a site offers a perspective on how links can be used to define the site as a text.

The hub/terminal continuum is designed to rectify issues with textual criticism of Web sites by providing three clear benefits to critics. First, looking at how a site connects throughout the Web stabilizes the site by looking at its larger design, rather than the particular manifestations of that design on a given day. Of the sites examined in this paper, all three are substantially modified at least three times a week. This dynamic means any analysis of a site will be dated, depending on the version selected for study.

Unlike studying a speech, television show, or other text written offline, Web texts change by design. Looking at patterns of linking is one way to minimize the problems presented by this changes over time. In doing so, critics are forced to look at the overall design of the site and how links are deployed, rather than the particularities of a site on a given day. Sites can certainly change, but overall design and linking philosophy is less likely to change quickly than the surface appearance of a site.

A second area this continuum resolves for critics is the ability to draw a boundary around a site. Analyzing sites like Townhall.com or Penny Arcade can be virtually impossible as other texts are woven into the site. Both of these sites are defined in large part by how they tie themselves to texts originally designed independently. In doing this, a dilemma is presented for critics; where does the text for the site really end? Examining the links and their functionality helps to determine how the site functions as a terminus, hub, or something in between.

The final advantage of this approach is that it defines sites based on the unique features of the Web. Designed as a massive hypertext system, the Web is structured to allow associations to be drawn between text created independently.

The Web bring communities of texts together in altogether new ways. Analysis of Web–based text is complicated The hub/terminal continuum does this, evaluating sites based on how they choose to connect to other locations on the Web and how they use, or choose not to use, additional material to enhance or change their site. End of article

 

About the author

Christopher A. Paul is an Assistant Professor at the University of Alabama in Huntsville.

Comments may be e–mailed to paulc [at] uah [dot] edu.

 

Notes

1. Landow, 1992, p. 42.

2. Katherine Hayles (2002) makes a fundamentally different argument about textuality, predicated on how the development of hypertextual thinking alters how people engage and conceive of “texts.” She connects texts to their material mode of construction, arguing that electronic texts can lead to the recognition that print is “a medium and not a transparent interface” (p. 43). Looking at the material conditions structuring Web sites is part of what stressing the importance of the link functions to do. Although this approach can lead to a more stable text than Hayles is likely to be comfortable with, it does fundamentally stress how the material conditions surrounding the establishment and reading of hypertexts structure the ways in which Web sites are encountered.

3. Michele Jackson (1997) analyzes how the “structuring of relations among nodes is a powerful strategic tool unique to communication in Web–based media,” but her article largely makes a case for network analysis as an analytical tool, rather than a means by which to engage in textual criticism.

4. Mitra and Cohen, 1999, p. 180.

5. Mitra and Cohen, 1999, p. 181.

6. Schneider and Foot, 2004, pp. 116–117.

7. Schneider and Foot, 2004, p. 118.

8. Ibid.

9. For an analysis of how the debut of Google impacted the role of links in defining what is seen on the Web, see Jill Walker’s “Links and power: The political economy of the Web.”

10. For a brief explanation links play a role in Google’s PageRank, see http://www.google.com/technology/.

11. Google’s value is certainly determined by the content of the database of links it possesses, as that is how the site generates its search results. However, reducing the site to its index misses the fundamental connection between Google and other sites. Google cannot be reduced to its database, as the database is largely a means by which Google can apply its algorithm and attract users. The point of a visit to Google as a search engine is to query its database and find another, previously unknown, site, solidifying its status as a hub.

12. More details about AdSense can be found at https://www.google.com/adsense/.

13. Certainly a link need not be anything more than an invitation to visit a linked site, but, in the case of Penny Arcade, the links are so thoroughly worked into the narrative argument of the site that making sense of what is being said regularly requires the following of a link. In this case, PA is actively creating a hub where their text depends on the connected link to be understood.

14. Examples of this are discussed at http://www.penny-arcade.com/2003/02/19 and http://www.penny-arcade.com/2003/02/24. Thank you to PA forum denizens neville and Yar for the links.

 

References

Espen J. Aarseth, 1997. Cybertext: Perspectives on ergodic literature. Baltimore: Johns Hopkins University Press.

Nancy Baym, 1997. “Interpreting soap operas and creating community: Inside a electronic fan culture,” In: Sara Kiesler (editor). Culture of the Internet. Mahwah, N.J.: Lawrence Erlbaum Associates, pp. 103–120.

Jennifer Roberts Biddison, “Become a partner,” at http://www.townhall.com/Partners.aspx, accessed 3 September 2007.

Jay David Bolter, 1991. Writing space: The computer, hypertext, and the history of writing. Hillsdale, N.J.: Lawrence Erlbaum Associates.

Locke M. Carter, 2000. “Arguments in hypertext: A rhetorical approach,” In: Ken Anderson (editor). Hypertext 2000: Conference on Hypertext and Hypermedia — Proceedings of the Eleventh ACM on Hypertext and Hypermedia (30 May–4 June, San Antonio). New York: Association for Computing Machinery, pp. 85–91.

Rune Dalgaard, 2001. “Hypertext and the scholarly archive: Intertexts, paratexts and metatexts at work,” In: Hugh Davis, Yellowlees Douglas, and David G. Durand (editors). Hypertext 2001: Proceedings of the Twelfth ACM Conference on Hypertext and Hypermedia (14–18 August, University of Aarhus, Århus, Denmark). New York: Association for Computing Machinery, pp. 175–184.

Steven J. DeRose, 1989. “Expanding the notion of links,” Hypertext ’89: Proceedings (5–8 November, Pittsburgh). New York: Association for Computing Machinery, pp. 249–257.

Donna J. Haraway, 1997. Modest_Witness@Second_Millenium.FemaleMan©_Meets OncoMouse: Feminism and technoscience. New York: Routledge.

N. Katherine Hayles, 2002. Writing machines. Cambridge, Mass.: MIT Press.

Christine Hine, 2007. “Connective ethnography for the exploration of e–science,” Journal of Computer–Mediated Communication, volume 12, number 2 (January), at http://jcmc.indiana.edu/vol12/issue2/hine.html, accessed 3 September 2007.

Michele H. Jackson, 1997. “Assessing the structure of communication on the World Wide Web,” Journal of Computer–Mediated Communication, volume 3, number 1 (June), at http://jcmc.indiana.edu/vol3/issue1/jackson.html, accessed 3 September 2007.

George P. Landow, 1992. Hypertext: The convergence of contemporary critical theory and technology. Baltimore: Johns Hopkins University Press.

Adrian Miles, 2001. “Hypertext structure as the event of connection,” In: Hugh Davis, Yellowlees Douglas, and David G. Durand (editors). Hypertext 2001: Proceedings of the Twelfth ACM Conference on Hypertext and Hypermedia (14–18 August, University of Aarhus, Århus, Denmark). New York: Association for Computing Machinery, pp. 61–68.

Ananda Mitra and Elisia Cohen, 1999. “Analyzing the Web: Directions and challenges,” In: Steve Jones (editor). Doing Internet research: Critical issues and methods for examining the Net. Thousand Oaks, Calif.: Sage, pp. 179–202.

Wendy Morgan, 1999. “Electronic tools for dismantling the master’s house: Postructuralist feminist research and hypertext poetics,” In: Jörg Westbomke, Uffe K. Wiil, John J. Leggett, Klaus Tochtermann, and Jörg M. Haake (editors). Hypertext ’99: Proceedings of the 10th ACM Conference on Hypertext and Hypermedia (21–25 February, Darmstadt, Germany). New York: Association for Computing Machinery, pp. 207–216.

Stuart Moulthrop, 1991. “Beyond the electronic book: A critique of hypertext rhetoric,” In: Hypertext ’91: Proceedings of the Third ACM Conference on Hypertext (15–18 September, San Antonio). New York: Association for Computing Machinery, pp. 291–298.

Alex Soojung–Kim Pang, 1998. “Hypertext, the next generation: A review and research agenda,” First Monday, volume 3, number 11 (November), at http://www.uic.edu/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/628/549, accessed 3 September 2007.

John Paolillo, 1999. “The virtual speech community: Social network and language variation on IRC,” Journal of Computer–Mediated Communication, volume 4, number 4 (June), at http://jcmc.indiana.edu/vol4/issue4/paolillo.html, accessed 3 September 2007.

Christopher A. Paul, 2005. “Re–imagining Web analysis as circulation,” First Monday, volume 10, number 11 (November), at http://www.uic.edu/htbin/cgiwrap/bin/ojs/index.php/fm/article/view/1291/1211, accessed 3 September 2007.

Francisco J. Ricardo, 1998. “Stalking the paratext: Speculations on hypertext links as a second order text,” In: Kaj Gronbaek, Elli Mylonas, and Frank M. Shipman (editors). Hypertext ’98: Proceedings of the Ninth ACM Conference on Hypertext (20–24 June, Pittsburgh). New York: Association for Computing Machinery, pp. 142–151.

Steven M. Schneider and Kirsten A. Foot, 2004. “The Web as an object of study,” New Media & Society, volume 6, number 1, pp. 114–122. http://dx.doi.org/10.1177/1461444804039912

Rob Shields, 2000. “Hypertext links: The ethic of the index and its space–time effects,” In: Andrew Herman and Thomas Swiss (editors). The World Wide Web and contemporary cultural theory. New York: Routledge, pp. 145–160.

Jill Walker, 2002. “Links and power: The political economy of linking on the Web,” In: Kenneth M. Anderson, Stuart Moulthrop, and James Blustein (editors). Hypertext 2002: Proceedings of the Thirteenth ACM Conference on Hypertext and Hypermedia (11–15 June, College Park, Maryland). New York: Association for Computing Machinery, pp. 72–73.

Barbara Warnick, 2002. Critical literacy in a digital era: Technology, rhetoric, and the public interest. Mahwah, N.J.: Lawrence Erlbaum Associates.

Barbara Warnick, Michael Xenos, Danielle Endres, and John Gastil, 2005. “Effects of campaign–to–user and text–based interactivity in political candidate campaign Web sites,” Journal of Computer–Mediated Communication, volume 10, number 3 (April), at http://jcmc.indiana.edu/vol10/issue3/warnick.html, accessed 3 September 2007.

 


Editorial history

Paper received 4 September 2007; accepted 28 October 2007.


Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.

Hub and terminal: Developing a method for textual analysis on the World Wide Web by Christopher Paul
First Monday, Volume 12 Number 11 - 5 November 2007
http://www.firstmonday.org/ojs/index.php/fm/article/viewArticle/2025/1891





A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2014.