Computer Ethics: The Evolution of the Uniqueness Revolution

Transcription

1 Computer Ethics: The Evolution of the Uniqueness Revolution Computer Ethics: The Evolution of the Uniqueness Revolution by Donald Gotterbarn and Simon Rogerson Presented at the ACM Computer Ethics Conference CEPE'97, Erasmus University, The Netherlands, 1997 Abstract: We examine, what has been considered a defining characteristic of computer ethics, namely the claim that it is unique and argue that there is a significant conceptual muddle about the statement, "Computer ethics is unique." This muddle has led to multiple inconsistent views of computer ethics and in fact this muddle has contributed to some claims which are counter productive for the discipline. In this paper we clarify the claim and show why these counter productive claims are mistaken. Contact: CCSR Web Master Last update Fri Feb of 1 16/2/2002 6:29 ìì

2 Computer Ethics: The Evolution of the Uniqueness Revolution by Donald Gotterbarn and Simon Rogerson Abstract: We examine, what has been considered a defining characteristic of computer ethics, namely the claim that it is unique and argue that there is a significant conceptual muddle about the statement, "Computer ethics is unique." This muddle has led to multiple views of computer ethics which appear to be inconsistent and in fact this muddle has contributed to some claims which are counter productive for the discipline. In this paper we clarify the the different senses of this claim and show why these counter productive senses are mistaken. 1 Introduction In the evolution of computer ethics there has been an intermittent discussion of the statement that "Computer ethics is unique". The discussion seems to re-emerge with each new change in technology and change in predominant area of interest about the way computers are changing our social institutions. One of the earliest positive definitions of computer ethics was given by James Moor [1985]. In this paper Moor argues that computer ethics is unique by first defining a computer ethics situation as one in which a computer is "essentially involved". This involvement contributes to the uniqueness of the situation because of the computer's unique properties of "logical malleability" and "speed". He argues that the pervasive alteration of social and cultural situations by this machine has created policy "vacuums" about how to use computer technology which require us to adopt a multi-disciplinary approach to determine how to fill these vacuums. From this he concludes that the domain of computer ethics consists of those situations which have been altered and that computer ethics is worthy of study because of its practical importance. The rise of the Internet has led to several new papers on the question. Deborah Johnson[1997] recently suggested that several aspects of the Internet have had an impact on the issues within computer ethics She appeals to the pervasive "anonymity" of individuals and "reproducibility" of information in ways not possible without the Internet. Krystyna Gorniak [1996] also appeals to the Internet and argues that there are no good analogies available for cyber-situations. On the surface it may appear that these authors are making similar claims based on the deceptively simple claim "Computer ethics is unique" (Hereafter CEIU). We shall see that there are many meanings to CEIU. There are several axis of understanding of CEIU. One way to see the difference is to look at the position that CEIU is used to further. In the papers cited above the authors assume that one approach to ethical reasoning is by analogy. We determine which normative principles to apply and how to judge a new situation based on its similarity to other familiar situations. Moor points to a policy vacuum where we have not yet discovered the analogies to the new situations, while Gorniak talks about the absence of analogies for cyber-situations. From this she concludes traditional ethics, for which we have analogies, which may be difficult to discover, does not apply to cyber-situations. Moor's claim for uniqueness leads to a search for usable analogies; Johnson uses traditional norms to derive some general rules for behavior; while Gorniak's claim about cyber-situations means that traditional ethics is not applicable to cyber-situations. This is an example of one difference in CEIU, namely some interpret CEIU to mean "Computers have made an evolutionary change in the ethical landscape and we have to work hard using traditional ethics to remap that landscape", while others interpret CEIU to mean that "Computers have made a revolutionary change in the ethical landscape such that the principles space and time can no longer be used to guide us in traversing the landscape." The evolutionary version of CEIU is about particular ethical decisions, while the revolutionary version is about the methodology of ethical reasoning. This is but one of several variations in CEIU. In this paper, we maintain that: this is not a simple statement, but that it has several quite distinct meanings, 1 of 8 16/2/2002 6:30 ìì

3 this assertion has some significant and dangerous consequences, until these meanings of uniqueness are clearly defined, discussions based on CEIU results in inconsistencies and a failure to understand the consequences of this claim, and the revolutionary interpretation of CEIU has not yet been proven. 2 CEIU: traditional responses 2.1 The assertion of CEIU is benign Discussions about the uniqueness of computer ethics is sometimes viewed by philosophers and novice computer ethics scholars as just another benign philosophical discussion. The answer to this question is viewed as simply a part of the definition of computer ethics. Saying computer ethics is unique is almost like giving an ostensive definition. People who have dealt with this question have rarely claimed to be establishing results in the discipline of computer ethics. CEIU's primary use is as a justification for the discipline of computer ethics or the study of these special unique problems. The uniqueness discussion is sometimes viewed as so irrelevant to the progress of the discipline, that the discussants show little interest in how others have approached the question or in testing the correctness or precise meaning of other's claims that computer ethics is unique. This claim is sometimes made to justify the creation of another discipline or academic department, for example, Walter Maner [1995] states, " Computer ethics is unique so its study is justified." As indicated above, the consequences of such a claim, when understood in a revolutionary way, are far from insignificant. 2.2 A framework for understanding A framework for understanding the various meanings of this assertion will be useful. One way to view the brief history of computer ethics is to order the primary areas of interest of computer ethics. Broadly speaking computer ethicists have addressed subjects of: 1) abuses committed with computers including fraud and theft, 2) the effects computers have on changing society, and 3) ethical issues related to the development of software systems. In the first stage of concern with computer ethics the order of the interests was: abuses, changes to society, and development issues. In the current stage of computer ethics the order of these areas is changed to society characterized by the Internet, abuses done using computers, and software development issues. Donn Parker was one of the early writers about ethical issues in computing. In his early works Parker used the fact that each answer to short scenarios was unique and that this disagreement proved that the reason computer ethics is not taught in computer science departments is that there is no such subject as computer ethics. It is so unique, it is a non-discipline [Parker 1977]. He used a voting methodology to decide what is ethical in computing. He gathered the opinions of people from several professions, ranging from accountants, attorneys, psychologists and philosophy professors to computer professionals. They were asked to vote on the ethics of individuals described in very short scenarios. He says, this work was not guided by a concept of computer ethics nor was there an attempt to discover ethical principles. He called this approach "micro-ethics." Not only was there an absence of a concept of computer ethics but the primary direction was an emphasis on proscribed activities. The only direction was that the scenarios were " written in such a way as to raise questions of unethicality rather than ethicality [Parker 1977]". Donn Parker used the diversity of opinions expressed about these scenarios to argue that there was no such thing as computer ethics. Parker's extreme conclusion is not justified for two reasons. First, it does not follow from his own evidence. Second, the evidence examined has little to do with Computer Ethics. The presumption that there can be no 2 of 8 16/2/2002 6:30 ìì

4 agreement in ethics can be so strong that it corrupts one's view of the evidence. For example, at a conference, Parker described the results of his 1977 workshop. He began by saying that there was agreement on many scenarios and then went on to say, "We got a lot of very close votes. In other words, we were not able to obtain a consensus on what is unethical and not unethical in the computer field. [Parker 1982]" This conclusion requires that he ignores all the places where there was agreement. It does not follow from the existence of some gray areas in a domain that there are no clear areas in a domain. The existence of hard problems in math - like the discrete decimal value of 1/3 - is not proof that there are no solutions in math. He handles the evidence in the same way in his revision of the 1977 book [Parker 1990]. The only cases he brings forward to the new book are those which generated the highest degree of diversity of opinion. He ignored those on which there had been a significant degree of unanimity. In the first stage of the concern with computer ethics, James Moor had argued that computer ethics is a viable subject but that the pervasive impact of computing had caused some policy vacuums about the practical use of computers. For Moor, uniqueness refers to the new situations More recently Krystyna Gorniak, arguing that Moor has not gone far enough [Gorniak 1996] maintains that although there may be an area called computer ethics, it is so unique that traditional ethical concepts do not apply to it and we have to wait for the Newtonian revolution in ethics to develop or discover the concepts that do apply. The computer has unique effects on the world "...the more computers will change the world as we know it, the more irrelevant ethical rules will be and the more evident the need of a new ethic. This new ethic will be the computer ethic." The uniqueness of the way the computer transforms the world will "...require us to go back to the old question: what is ethical". The old ethics has been made irrelevant by the unique nature of the computer revolution. According to Gorniak, we are in an ethical "vacuum" in which the prior ethics (Utilitarianism, Kantianism, etc.) are increasingly irrelevant.. Some computer practitioners may appreciate this, since no one can, with good philosophical foundation, say their actions are ethically wrong. Others consider this an ambiguous and ill-founded situation which restricts all of their actions because they want to only do the right thing. Notice how "uniqueness" has shifted from referring to a socio-cultural phenomena and it now refers to the principles of normative ethics, which are so unique they are yet to be discovered. The answer to the uniqueness question is not insignificant. What is puzzling is that, as we have just seen, the same answer "Yes, computer ethics is unique." can lead to such radically different conclusions about the viability and nature of the subject "computer ethics". There are even those (for example Oz) who have used all the uniqueness variants to promote an inclusive form of computer ethics. How can there be such diverse conclusions drawn from such a simple statement? William James, the great American philosopher, said that many philosophical problems can be solved if we clearly define the question. 3 The Variety of Answers The variety of interpretations of the uniqueness claim can be approached from several directions. We present the views of several authorities in compute ethics. We consider their starting positions, the resultant perspectives that they adopt and use these to identify the overall view of computer ethics each of them promotes. In 1985, Deborah Johnson considered computer ethics to be a study into the way in which computers pose new versions of standard moral problems and moral dilemmas that resulted in having to apply moral norms in new situations. Her stance allows little room for the concept of computer ethics uniqueness. This stance is modified in the second edition of her book [1994] in that more emphasis is put on computer ethics being concerned with new species of problems. This evolution of stance continues in her work with Keith Miller [1996]. Johnson and Miller explain that computing has fundamentally changed who we are as individuals, communities and nations and that computing has challenged and modified social culture, laws, economics and politics. They 3 of 8 16/2/2002 6:30 ìì

5 explain that computer ethics is about working on something new whilst drawing on something old. Significantly they state that most computing issues can be understood using traditional moral concepts but there are special features that make a moral difference. The implication is that if most are covered, have analogies with standard ethical issues, some must not be covered and these therefore are unique. They also recognize that the issues are significant in their own right and are thus arguing for the existence of a type of uniqueness. In addition to the way computers transform some social issues, they explain that issues arise in unique ways that computer professionals must address in the course of their work. It is clear that the genus-species argument has thus evolved to recognize there is a kind of uniqueness in computer ethics and that there is some recognition it is multifaceted. Moor[1985] advocates that the field is concerned with policy vacuums and conceptual muddles regarding the social and ethical use of computing. Moor appears to be suggesting that there is the possibility of new situations but does not advocate abandoning existing theories to address these new situations which is the possible course of action that Gorniak [1996] proposes. Moor also implies that addressing these issues will require an amalgam of disciplines. Oz [1994] claims that it is the speed of change which has left a big ethical vacuum and this has meant it is difficult to address issues of misuse that inevitably follow the advent of this universal tool [Moor 1985]. Oz describes three categories of misuse requiring ethical consideration:offences that existed before the advent of computers but were facilitated by computers, offences against computers, computer equipment and software, and the invasion of privacy. The first of these is the genus-species concept, the second includes offences that did not exist before the advent of computers such as the destruction of computer files and the illegal copying of software, and the final one is a specific fundamental human right. Oz is unique in that all of the varieties of the uniqueness claim are maintained simultaneously. Maner [1990] started to define an area of study that could be labelled computer ethics. In a later paper, Maner [1996] advocates that computer ethics examines the ethical problems aggravated, transformed or created by computing and that the "Lack of an effective analogy forces us to discover new moral values, formulate new moral principles, develop new policies, and find new ways to think...". Maner makes a clear statement of uniqueness in the definition of computer ethics. Bynum [1993] advocates that computer ethics is concerned with how to integrate computing and human values in such a way that computing advances and protects human values rather than harms them. It is probably the broadest perspective and embraces applied ethics, sociology of computing, technology assessment, computer law and other related fields. Such an approach appears to place computer ethics as an "umbrella" discipline unique in its holistic form but not when decomposed. Huff and Finholt[1994] adopt a different stance. They focus on the computer professional, suggesting there are three areas of commitment in the search for excellence: commitment to designing and implementing a quality product, awareness of ethical issues in design and implementation, and concern about the social context that influences the use of computing. This appears to be an approach of practical usefulness recognizing that the field requires some special attention given its special characteristics. This approach is similar to the work of Gotterbarn [1992] in this area who is concerned with the way in which computer professionals undertake their work so that the outcomes of their endeavors are socially and ethically acceptable. Anderson [1994] suggests that computer ethics applies not just to computer professionals but to all who use computers or come into contact with their inputs and outputs which means every one living in industrialized societies. It is this, the last part of his definition, that restricts the focus in a potentially dangerous manner as it does not take into account the needs or impacts on the world at large simply focusing attention on the industrial society. Spinello [1995] explains that any technology tends to create a new human environment and never more so 4 of 8 16/2/2002 6:30 ìì

6 than computing. He is concerned that this new environment does not violate personal rights or the values of fairness and justice. He insists on the inclusion of all stakeholder considerations when deciding upon technological issues and appears to advocate a Rawlsian approach. Miller provides some interesting input to the computer ethics debate [Collins and Miller 1992, Collins, Miller et al 1994]. He suggests that software is different from other manufactured products and thus raises different and difficult ethical issues. He further suggests that such issues can be addressed in a rational and systematic manner and advocates a method for ethical decision making that is computing-specific. Without doubt there is an implicit uniqueness claim in these suggestions. It can be seen from this brief review that there are many views of what computer ethics comprises. This richness of opinion is probably due to a difference in perception of computing characteristics, the adopted focus - broadly split between the outcome/symptom of computing or the undertaking of computing, and the disciplinary background. Starting points are certainly different and include the properties of computer technology, the concept of computing, the application of computing and the human value impact of computing. The perspectives can be analyzed using several sub categories: social - considering the computing-impact on society professional - considering the manner of professional activity universal activist - inclusive of communities around the globe parochial - restricted consideration of current and future issues ethics only - drawing only from ethics theory multi-disciplinary - blending theory from several disciplines These are not mutually exclusive. Using starting points and perspective categories the different stances can be classified as shown in the following table. Starting Points Johnson 85&94 Moor Gorniak Oz Maner Bynum Huff & Finholt Gotterbarn Anderson Spinello Miller Parker Properties X X X X Concepts X Applications X X X X X X X Environment X Value X Perspectives Social X X X X X X Professional X X X X X Universal X X X X Parochial X Ethics X X X X X X Multi-disciplinary X X X X X X 5 of 8 16/2/2002 6:30 ìì

7 So from this we can determine that, for example, Moor adopts a universal conceptual multi-disciplinary stance, Gorniak adopts an environmental universal ethicist stance, Maner is a properties ethicist and so on. What does this mean regarding the uniqueness claim? Those that have a very strong emphasis on the technological makeup are likely to see computing as unique [for example, Maner and to a lesser extent Oz]. Those adopting a universal stance are more likely to subscribe to elements of uniqueness [for example, Bynum, Gorniak and Moor]. Those who have strong application elements in their stance are more likely to question the wholly unique position [for example, Johnson, Huff & Finholt and Gotterbarn]. In the context of this paper, it is interesting to analyze to what purpose these particular stances have been put. It is clear that the all the authors reviewed in this paper have argued a particular stance to explain why and how certain situations and decisions have occurred and to give some insight to the future. It is also clear that these stances have then been used to substantiate the importance of a particular direction of computer ethics or justify some holistic label for computer ethics. The table below shows, for each author, the main thrust with respect to uniqueness and the underlying message as to the nature of computer ethics that the uniqueness stance has been used to justify and promote. For example, Johnson's genus-species stance is the justification for using existing theory for computer ethics in other words the "use the old order" holistic label. In contrast, Miller's assertion of software as a different product is the justification for a special approach called computer ethics. Bynum's stance of computing being all pervasive is used to argue that computer ethics is about the human values and technology equation whilst pervasiveness is used by Anderson to promote the idea that computer ethics is important for all people in a technologically-dependent world. Author Overall thrust regarding uniqueness Computer ethics message Johnson genus-species use the older order Maner special technology bounded and worthy of study Gotterbarn professional specificity codes and other guides are essential Miller different product need for special approach Moor vacuum need for serious multi-disciplinary consideration Oz special technology special discipline Bynum all pervasive human value to computer equation Anderson all pervasive computer ethics for all Gorniak all pervasive global new order Spinello special technology inclusion of all interests and concerns Huff and Finholt all pervasive inclusive approach Parker non consensus no new discipline 4 A Range of Views The evidence for the whole panorama of uniqueness claims is drawn either from some special characteristic about the computer or from the type and pervasiveness of the impact on social institutions. When the machine is used as evidence, its novel characteristics are referenced, for example, logical malleability [Moor 1985], or information storage [Maner 1996]. The machine is unique if some parts of it which are unique can be identified. The appeal to social phenomena is not as absolute as the appeal to the machine. It is in the type and range of social phenomena used in the uniqueness claim that we also find evidence for the amorphous nature of CEIU. 6 of 8 16/2/2002 6:30 ìì

8 The views on the scope of the impact of computers on social institutions ranges from the universal to the very narrow. Both Moor and Gorniak start from the claim that the impact of the computer is pervasive, on all aspects of society, yet one concludes that we need to apply ethics to the policy vacuums created by this machine while the other questions the applicability of traditional ethics to global ethics. Terry Bynum follows Moor on the pervasiveness question, but Bynum wants computer ethics to advance and protect human values. This is an evolutionary view. Others argue that the range of impact is limited and their view of computer ethics is in part determined by what they include in the range of unique impact. Gotterbarn [1992] argued for a very narrow range of unique situations, namely those situations which are relevant to the practising professional and he claimed that all other ethical situations have analogies to traditional ethical issues. The unique professional issues could be addressed using principles of John Rawls. In her first book, Deborah Johnson [1985] simply addressed a set of issues which had "a different moral flavor" but could all be managed by traditional moral principles. In a later piece with Miller[1996], there are clear indications that some of the transformed situations cannot be addressed using traditional normative concepts. Collins and Miller [1992] also used Rawlsian techniques, but they applied them to a much broader range of situations than did Gotterbarn. Effy Oz, limits the scope of ethical issues to computer abuses and uses traditional normative theories to address these issues. In the brief period of time that computer ethics has been discussed, the truth of Moor's insight about the increasing pervasiveness of computing and the need to address new issues is quite evident. As the technology called computing has worked its ways into new facets of life, the scope of computer ethics has changed. One of the things this investigation has revealed is that several of the authors have broadened the range of situations that they subsume under computer ethics. These elements are unique in the sense that they are new elements in the human situation. However, they are still part of the human situation which can be understood in terms of traditional ethics. It is these authors who have been at the forefront of evolutionary ethics. The arguments in support of revolutionary computer ethics have not been very convincing. For example, Gorniak argues that Kantianism is not relevant to global ethics, "because Kantianism developed in the West, and computer ethics is global, Kantianism is not applicable". She extends this claim to all moral theories developed in the West.. She argues that they are not global because they are not universally accepted, respected, and followed. Gorniak rejects the applicability of a moral principle based on its failure to be a "universally accepted ethical system".. Maner's arguments are equally unconvincing. He moves from claiming the uniqueness of a select group of moral situations to claiming that we must discover new moral principles; but he has not demonstrated that the ways in which these situations are unique entails the irrelevancy of traditional moral norms to these situations. 5 Conclusion The claim that computer ethics is unique is correct in only a limited sense. It is correct that some of the particular decisions in computer ethics are unique but the revolutionary view that the methods and fundamental principles of ethics are unique has not been demonstrated. We have seen that several situations are so transformed by the computer that they are unique. There have been a variety of answers as to the scope of the unique situations: all social interactions that are transformed by a computer, cases of computer abuse, or issues related to software development. If the domain of computer ethics is defined by those elements which are unique, we first need to understand the sense of uniqueness we are talking about when dealing with computer ethics. Those who study views of various computer ethicists are often misled by the simple claim that "Computer ethics is unique." To avoid the continuation of these difficulties it is necessary for those who assert this claim to clearly define their concepts using the categories we have defined above. To paraphrase James, "If we clearly state the elements of our standards of the uniqueness we can spend less time arguing for the uniqueness of computer ethics and more energy doing computer ethics.". 7 of 8 16/2/2002 6:30 ìì

9 References Anderson R E (1994), The ACM Code of Ethics: History, Process and Implications, in Huff C and Finholt T (eds), Social Issues in Computing: Putting Computing in its Place, McGraw-Hill, 1994 pp Bynum T W, Maner W and Fodor J, eds. (1993), Teaching Computer Ethics, Research Center on Computers & Society. Collins W R and Miller K W (1992), Paramedic ethics for computer professionals, Journal of System Software, Vol 17 No 1, pp Collins, W R, Miller K W, Spielman B J, Wherry P (1994), How good is good enough, Communications of the ACM, Vol 37 No 1, 1994 pp Gorniak K (1996), The Computer Revolution and the Problem of Global Ethics, Science and Engineering Ethics, Volume 2 No 2, pp Gotterbarn D (1992), The use and abuse of computer ethics, Journal of System Software, January Vol 17 No 1 pp Gotterbarn D (1991), Computer Ethics: Responsibility Regained, National Forum: The Phi Kappa Phi Journal. Johnson D G (1985), Computer Ethics, Prentice-Hall, Johnson D G (1994), Computer Ethics, 2nd edition, Prentice-Hall. Johnson D G (1997), Ethics Online, Communications of the ACM, January Johnson D G and Miller K W (1996), Ethical Issues for Computer Scientists and Engineers, in Tucker A B (ed), The Computer Science and Engineering Handbook, CRC Press pp Johnson D G and Nissenbaum H F (1995), (eds), Computer Ethics and Social Value, Prentice-Hall, Englewood Cliffs, New Jersey. Huff C. and Finholt T (1994), (eds), Social Issues in Computing: Putting Computing in its Place, McGraw-Hill. Maner W (1990), Starter Kit on Teaching Computer Ethics, Helvetia Press. Maner W (1996), Unique ethical problems in information technology, Science and Engineering Ethics, Volume 2 No 2, pp Moor J (1985), What is computer ethics?, Metaphilosophy Vol 16 No 4, pp Oz E (1994), Ethics for the Information Age, Wm C Brown Communications. Parker D (1977), Ethical Conflicts in Computer Science and Technology, AFIPS, no date but based on a meeting in Parker D (1982), Ethical Dilemmas in Computer Technology, in Hoffman W M and Moore J M (eds), Ethics and the Management of Computer Technology, Oelgeschlager, Gunn & Hain. Parker D (1990), Ethical Conflicts in Information and Computer Science, Technology, and Business, QED Information Sciences, Inc. Spinello R (1995), Ethical Aspects of Information Technology, Prentice-Hall. publication details GOTTERBARN, D., ROGERSON, S., Computer Ethics: The Evolution of the Uniqueness Revolution, the ACM Computer Ethics Conference CEPE'97, Erasmus University, The Netherlands, of 8 16/2/2002 6:30 ìì