Masthead CMC Magazine July 1, 1995 / Page 3

COVER STORY

The Ethics of Web Site Engineering

by Chris MacDonald (chrismac@ethicsweb.ca)

The analogy between electronic communications and physical spaces is not a new one. Since well before Gibson's Neuromancer [1] introduced the term cyberspace, the denizens of the Internet have thought of the Net as a place to meet, exchange ideas and information, and make friends. The names and numbers used to designate individual computers and even documents on the Internet are already known as addresses. Various parts of the Internet are joined by bridges and gateways, and individual users interact with their computers by means of windows and desktops. The Grand Analogy in Johnson's Computer Ethics [2] suggests that we think of cyberspace as a newly discovered island, waiting to be explored and civilized. Most recently, perhaps, we have seen endless hype over the "information superhighway" which will one day soon (it is said) connect the world via high-capacity data transfer media. This analogy between electronic communications and physical spaces is clearly the dominant one. But it behooves anyone thinking seriously about the ethics of electronic communication to consider carefully this analogy.

Obviously, there are some real similarities between real places and cyberspaces. People can interact in both. Individuals and groups carve out niches in both. One's access to both depends to some degree on both financial status and personal connections.

But just as obviously, there are also real differences. Real space is strictly limited in quantity. The size of cyberspace is limited only by our capacity to finance larger and faster machines. Virtual space, unlike physical space, can grow -- and is growing exponentially -- as a result of the work of interested human beings. In "real" spaces, much of our interaction is affected by the fact that there is a great deal of permanence to our surroundings. In cyberspace, a place you visit today may be radically changed or disappear tomorrow. Cyberspace also differs in the way points are connected: any given point is (or can be) just one step away from every other point.

How good a guide, then, is this analogy between electronic communications and spaces? How misleading is it? Most importantly for the field of ethics, should we allow the inevitable establishment of norms in cyberspace to be guided by the differences or the similarities between real and virtual spaces?


It is clear that the spatial metaphor is useful psychologically. But is it useful ethically?

This question can take at least two forms. First, why should we expect cyberspace to be subject to different norms than other domains? Norms that exist in other domains -- the norms of face-to-face communications, for example -- have evolved over generations and have arguably served us quite well. Why go to the trouble of trying to establish new norms, just because we've established a new way of communicating? What's wrong with existing norms?

Alternatively, we might ask, since cyberspace is different in a number of ways, why shouldn't it be subject to different norms than those that prevail in other domains? For one thing, cyberspace has different possibilities and limitations than real space. The goals of cyberspace are also different. Michael Philips [3] puts forward an instrumentalist view of ethics which suggests that specific moral standards are to be evaluated according to how well they serve us. But this evaluation of standards, according to Philips, must be sensitive to the needs of particular domains -- that is, sensitive to what people value within those domains. By domain, Philips has in mind such areas of life as the business domain, the family domain, and the religious domain. But in the same spirit, we might ask why norms should be constant across the domains of cyberspace and physical space, given that these two domains serve different functions.

The region of cyberspace most recently being built and explored is the system of distributed information storage know as the World Wide Web. In the pre-Web era, cyberspace was, to most people, a forbidding frontier. With the advent of the Web and its subsequent popularization by such intuitive and inviting software as Mosaic and Netscape, cyberspace has finally had its borders opened to even casual visitors. Clearly, the current Netscape-driven rush to colonize cyberspace is bound to result in escalating social tensions within the online world. New interactions will need to be mediated by norms and standards of behavior, be they codified or unspoken. Whether those norms and standards should be ones long known and trusted or ones new and custom-made will be the first of two guiding questions for the rest of this essay.

The Ethics of Web-Site Engineering

Were we to take this analogy between electronic and physical space seriously, then we would have to consider the building of a Web site to be akin to architectural design and engineering. To what extent do the guidelines for architecture and engineering help us in building virtual spaces?

Engineering a Web site involves making a wide variety of decisions, from the type of server software to use, to the scope of the information to be made available, to the way in which the information will be presented. Many of these decisions will be ethically significant. Among the decisions which have the most obvious ethical importance are, for example, the decision as to whether or not to include materials that involve copyright infringement or are sexist or racist, and the decision whether to use pirated server software. Decisions with more obscure ethical importance might include (among many others) the choice of server software (will you cater to high-end or low-end users?), decisions as to what sorts of information to point to (are you willing to accept responsibility for pointing to materials provided by others who infringe upon copyright?), and decisions regarding the methods for keeping the information you present up-to-date (if people come to rely on your information, is your organization committed to its upkeep?).

In some cases, these decisions will be made by institutional policy makers. Policy makers seem as likely as technical staff to consider these to be more-or-less technical matters. Most institutional managers will be unaware of the norms and possibilities associated with a medium even as relatively well-established as the Internet. A fortiori most managers will be unaware of the norms and possibilities of the Web, which is arguably still in its infancy. In many cases, however, important decisions -- many of which will not be purely technical -- will be left to the technical personnel responsible for actually setting up the institution's server and building the institution's home page. No matter who ends up making these decisions, it is preferable that value-laden decisions be recognized as such, and that the value-assumptions that go into them be thought out clearly in advance.

In this essay, I will examine just two key Web-site features with ethical importance: clarity of administrative responsibility, and provision of Web-space for staff and, where applicable, students.

'Personal' Space on the Web

One crucial question that needs to be answered -- at both the micro and macro levels -- is that of who will have access to the Web. Like most valuable resources, the Web is subject to limitations on availability. It is worth remembering that not everyone can afford a direct or SLIP connection and a machine powerful enough to run Web software. Unlike most previous forms of electronic communications, on the Web the individual's ability to view documents is only half the story. On the Web, full access means the ability to publish.

I take it that one of the major factors in the rapid growth of the popularity of the Web is the relative ease with which individuals and groups can publish their own information. On the Web, publishing basically requires having one's own home page. Since the majority of Web sites are owned and administered by universities and corporations, having one's own home page is, for most people, a matter of being granted one. But there are obvious disincentives for organizations to give out such access. Administrative convenience usually mandates the provision of fewer, rather than more services and benefits. Often there will also be financial costs in providing the kind of connectivity that Web access requires. And the possibility of being held liable for content is another disincentive to which I will return.

Another disincentive is the computational load created by any good site. An individual given Web space might post information that becomes problematic due not to its controversial nature, but rather due to its popularity. Information provided over the Web might become so popular externally that it becomes unpopular internally. Web is a system which makes down-loading so simple (just click the mouse and wait a few seconds [4] ) that it risks placing an enormous burden on server machines.

In order to give a reasonably complete accounting of the costs and benefits of provision of home pages to individuals, it is necessary to look at both financial costs and benefits and moral costs and benefits. My main concern here is the moral costs and benefits, since I think the financial side is relatively, though not perfectly, clear. The moral costs seem to lie primarily in the risk that some individual to whom an organization gives home page privileges will do something which goes against the norms set by the organization. A corporation which gives Web space to employees runs the risk, for example, that some employee will use that space to display pornographic images for all the world to see. An academic department giving home pages to students runs the risk that some student will use that space to help organize a neo-Nazi rally. Keep in mind, here, that this information, though authored by the individual, has an URL (Uniform Resource Locator, the Web's system for giving any document a unique address) that includes the organization's domain name. Documents which I personally make available, for instance, all have URLs which include ethics.ubc.ca, the domain-name of the institution which provides me with Web access. This is in some sense the equivalent of using company letterhead for personal communications.

The moral benefits accruing to an organization which provides members with Web home pages are less obvious than the costs. It should be clear, at least, that there is a moral benefit to the individuals given such access--they gain moral goods, namely freedom of expression and a means of joining a robust online social and professional community. Those individuals given Web access also gain a valuable educational experience, and a broadening of their horizons. Yet it seems to be in the very nature of moral goods that providing them is in itself a moral good. There seems to be a clear moral benefit, therefore, not just to individuals, but to the organizations which provide the opportunities for expression and learning which go along with access to the Web.

Organizations commonly provide a wide range of benefits to their members. Corporations often provide not just workspaces, but employee lounges and kitchens. Universities provide libraries and study space. Many organizations provide funding so that individuals within the organization may travel to a variety of conferences and workshops not strictly required by the individual's position. Some will provide leaves of absence so that individuals can pursue personal projects or educational goals. Providing these benefits is seen both as a means of maintaining morale and as a matter of good organizational citizenship. It might be argued that it is in the same spirit for organizations to provide individuals with access to the Web.

How useful is our spatial analogy in this respect? How might it affect our intuitions? Might the provision of Web-space be compared helpfully with the provision of physical space to members of institutions? Will the similarities or dissimilarities between the two kinds of spaces be most relevant to this question?

On this point, I want to be careful to avoid providing dogmatically a yes or no answer. Whether the spatial analogy will prove useful is not something that can be wholly determined in advance on a theoretical basis. Whether it is a useful analogy or not depends precisely upon whether those faced with tough decisions find it to be useful -- that is, whether it helps them to make good decisions. I will, however, offer a few suggestions as to whether relevant considerations in such decisions will spring from or defy the spatial metaphor.

It seems that most of the relevant considerations involved in deciding whether or not a given institution should provide Web-space to its members involve differences between real and virtual spaces. The financial costs of Web-space are very much lower than the financial costs of physical space (enough Web-space for thousands of students can probably be provided by a computer costing under $10,000, whereas a enough study space for a similar number of students would run well into the millions). The moral costs (or rather moral risks -- potential moral costs) of providing Web-space spring largely from the fact that any virtual place (a Web page, for example) can be accessed directly from millions of other places, making it difficult for institutions to oversee (or regulate) the use to which virtual spaces are put.

The moral benefits, too, of providing Web access seem quite different from those of providing physical space. The moral benefits of providing study space, lounges, and offices relate to the fact that physical spaces meet physical needs (rest, etc.), and sometimes social or psychological needs (interaction with co-workers or fellow students). The moral benefits which I have suggested are associated with providing Web access seem to be of a different character: they arise mainly from meeting intellectual and educational needs and providing individuals with an opportunity to expand their worlds.

Again, I do not wish to state authoritatively whether or not the spatial analogy is useful in this respect. My own suspicion is that this metaphor will not prove helpful in making good decisions. My suggestions as to its potential usefulness notwithstanding, the actual usefulness of the metaphor can only be determined through application. I do suggest, however, that keeping the spatial analogy -- and relevant disanalogies -- in mind during the decision-making process may be a useful way of systematizing discourse. This in itself may be of significant value.

Administrative Responsibility

One of the best features of the Web is that it allows individuals to "travel through the online world along paths of whim and intuition" [5].

The downside of this, however, is that it is often difficult to keep track of whose virtual space you are viewing/visiting. As the amount of information available on the Web increases--which it currently does exponentially--administrative responsibility for that information becomes more relevant.

The primary means of signifying responsibility for Web pages is, of course, the author's name and e-mail address, or a link to those, at the bottom of the document. The "Suggestions for Web Pages" put out by NAIC (the Network Applications & Information Center at NASA) advises that "A point of contact should be indicated for every Web page...." This sage advice is often ignored, however. Besides, in such a visually-oriented medium, it seems a both unnecessary and insufficient to rely entirely on textual markers such as signatures to indicate borders of responsibility.

I suggest that a significant portion of the ethical risk involved in unclear responsibility for Web materials lies precisely in the seamless manner in which Web-based information is often presented to the viewer. Part of the beauty of the Web is that the user can follow links from one site to another, guided by a stream of consciousness, blithely unaware of the institutional borders he crosses. I suggest that here there is a qualified lesson to be learned from the analogy to physical space. Architects of physical space talk of the importance of 'transition space' between public and private spaces.[6]

These spaces act both as a signifier of the transition from one sort of space to another, and as a place where the decision whether to enter or not can be made. Why not make use of such transition spaces on the Web?

In attempting to apply this principle of 'transition space' to the Web, we see both the value and the limits of the spatial analogy. Transition spaces, though easy to implement, go against the very nature of the Web, at least if implemented aggressively. While exploring the Web, no one wants to see pages that are nothing but notifications of an institutional border. On the Web, anything that contains no information or entertainment seems otiose. On the other hand, some transition spaces are absolutely crucial to the construction of a good Web site. By this I mean the sorts of transition spaces represented by introductory pages featuring indices pointing to further information. When viewing a Web site's introductory page, no one wants to have pages and pages of text and megabytes of graphics thrust upon them. The very nature of Web surfing makes desirable the provision of 'areas' where visitors to your site can look around without committing themselves to long downloads. This is precisely what architects mean by 'transition space.' The spatial analogy here is useful, in that it reminds us of this important architectural concept. This example also illustrates, however, that we must be careful to tailor principles when applying them to new domains.

Another solution which we might import to help solve the problem of signifying ownership and responsibility lies in the very diversity which makes the Web so interesting. Rather than transition spaces, I advocate the marking of transitions by means of giving the pages associated with a site a specific look. This is akin to the attempt on the part of architects to give each building a distinctive look to differentiate it from its neighbors. Physical buildings seldom present a problem in terms of determining ownership or responsibility. The need to exit one building into the street before entering the next means that it is pretty much always clear when you have left one institution and entered another. The Web arguably has no streets--just innumerable adjoining rooms. Web surfers can move from one institution to another without even noticing. Giving the Web pages associated with any one institution a specific look or feel is akin to marking the transition from school to corporation by a shift from brick to marble.

A brief survey of a few institutional Web sites reveals that the best ones all have some unifying motif, whether it be a corporate logo or merely a consistent color-scheme. All the most visually interesting ones make it very clear when you enter and exit their 'space.' In a medium in which links to other people's information is the norm, these sites have made it clear which materials are their own, and which are not. It is in the interest of accountability to have such clearly delineated domain boundaries.

Norms in Virgin Territory: to Lead or to Follow?

It would be an understatement to say that the Web is growing quickly. One recent conservative estimate suggests that the Web is currently growing at an annual growth rate of 3000%. Much attention is given to macro-level issues involved in the growth of the Internet (e.g. the Clinton administration's frequent policy statements concerning the "Information Super highway," the battle between the Telcos and the cable industry, etc.,). My suspicion is that, while waiting for the Big Bang, 10,000 instances of micro-engineering may make a lot of decisions by collective fiat. Many decisions have already been made in this manner.

Building a Web site is a chance to participate in, and perhaps influence, the organic construction of the Internet. The Web is subject to the standard mechanisms of evolution: genetic crossover, mutation, and natural selection. But a few thousand years of animal husbandry has taught us that evolution need not proceed undirected. The architects of individual Web sites have the chance to push the evolution of the Web along a variety of continua, to whatever small extent, according to their own aesthetic and moral views. Popular Web sites will prove to be the template[7] for hundreds, if not thousands of other sites.

A crucial question for thoughtful Webmasters at this point in the Web's evolution is whether to follow precedents set by existing sites or create new standards instead. Most likely, of course, anyone setting up a new site will will end up doing some combination of following and trail-blazing. Clearly, the existence of a common language, HTML, takes care of one crucial bit of coordination. It would be foolish for any individual to invent and unilaterally implement her own HTML codes, since no one else's software would recognize them. But there is a nearly infinite number of characteristics which any given site can either have or leave out. Anyone who has spent any time at all surfing the Web knows that the Web is at present characterized by a fair bit of diversity.

In my own research, I have examined sites from a variety of institutions to look for just the two key features with ethical importance discussed above: clarity of administrative responsibility, and provision of personally administered home pages for staff and, where applicable, students. A surprising degree of diversity showed up with regard to these characteristics. The result is that if the novice Webmaster surfs the Web looking for examples to follow, it is likely that no real pattern will be apparent. So even the decision 'just to follow suit' does not settle much.

Perhaps the most that can be said in this regard is that those charged with the design and construction of new Web sites should follow three main guidelines. First, remember that the Web, and the Internet more generally, is a community. Explore it extensively in order to get a feel for what the existing norms and standards are. Second, follow those examples which best suit both the style of the organization and the nature of the information to be presented, and ignore those precedents which seem, upon sufficient consideration, irrelevant. It matters whether you are building a Web site for a professional association or a rock band. Third, be aware that many (most? all?) of the decisions to be made in designing a Web site will have ethical implications at some level. The very fact that a decision has ethical implications implies that it merits discussion, rather than unilateral decision-making. [CMC TOC]

Notes:

  1. William Gibson, Neuromancer. New York : Ace Books, 1984. (Return to text)
  2. Deborah G. Johnson, Computer Ethics, 2nd ed. Englewood Cliffs, NJ: Prentice Hall, 1994. See esp. p. 13-14. (Return to text)
  3. Michael Philips, "Business Ethics: Thinking Systematically." Earl R. Winkler and Jerrold R. Coombs, eds., Applied Ethics: A Reader. Cambridge, MA: Blackwell, 1993. (Return to text)
  4. Granted, Web documents containing megabytes of images or dozens of pages of text can take much longer, especially over a slow modem, but the Web's advantages over Gopher and FTP when it comes to accessing information are still quite clear. As a result, some popular sites currently receive document requests on a scale exceeding one per second. Much greater demands are easy to imagine. (Return to text)
  5. Gary Wolf, Wired 2.10, p. 116 (Return to text)
  6. Eric Fruhauf, UBC School of Architecture, personal communication. December 13, 1994. (Return to text)
  7. One site can literally be the template for another, since most web-browsers allow the user to download the source code for any given web-page. (Return to text)

Chris MacDonald is a Graduate Research Associate, at the Centre for Applied Ethics, University of British Columbia.

This work was supported in part by a University Graduate Fellowship from the University of British Columbia, and a Research Assistantship at the Centre for Applied Ethics. Thanks are due to Peter Danielson for the course which inspired this work, and for his many excellent comments on an earlier version. Thanks also to Bashir Jiwani for his comments.

Copyright © 1995 Chris MacDonald. All Rights Reserved.


This Issue / Index / CMC Studies Center / Contact Us