Carolyn Miller

Carolyn Miller (USA)

Carolyn Miller[i] is professor of English at the Department of English at North Carolina State University, where she teaches courses in their Master of Science degree in Technical Communication [ii]. Her research interests include the rhetoric of science and technology, writing in the professions, classical and contemporary rhetorical theory, and on how values, interests, and prior knowledge affect the ways that individuals and groups interpret and respond to communication. Her current research applies these interests to risk communication; she is also studying the role of novelty and tradition in scientific rhetoric. She has published articles in a wide range of academic journals and several scholarly books, and is a co-editor of the award-winning book, New Essays in Technical and Scientific Communication. She was also instrumental in designing and implementing the Master of Science in Technical Communication at NC State and served as its director from 1988 to 1995.

Carolyn’s contribution is entitled Writing in a Culture of Simulation: Ethos Online. Here, we meet Julia, a kind of mini computer robot known as a ‘chatterbot’, that visitors to networked communities in MUDs, MOO’s and Internet chatrooms are meeting more and more. Julia has been designed to take the Turing Test, which, if successful, means in practice to manage to convince, at least for a while, a human interlocutor meeting and interacting with her online, that ‘she’ has a human intelligence. As an artefactual ‘inhabitant’ of online text-worlds or distributed virtual environments capable of interacting with humans, the case of Julia is of particular interest, claims Carolyn, in that ‘she’ may perhaps be able to help us understand better some of the specific effects which the ‘culture of simulation’ (Turkle 1997) now being opened up via networked writing and interaction environments, is having on writing practices, or as Carolyn often refers to it herself, rhetorical action in general. Declining the opportunity to discuss trends in writing instruction related to the transition from a ‘culture of calculation’ – the world of DOS and UNIX – over to the culture of simulation offered by Windows and Macintosh interfaces, and also the issue of ‘the death of argument’, which tends to focus on loss of authorial control and fragmentation of linear discourse by hypertext, she elects to pursue the issue of what kind of interaction we are actually involved in when we engage in ‘computer mediated communication’. Her contention is that the Turing test, as instanced in human interactions with a chatterbot like Julia, is not a test of intelligence at all, but rather a test of rhetorical ethos, that particular qualitative aspect of discourse which allows us to infer the character of our interlocutor.

Work some time back now with a computer program that simulated a psychiatrist (ELIZA) demonstrated the so-called Eliza effect, i.e. that we often tend, as Turkle puts it, “to project our own complexity onto the undeserving object”, attributing more intelligence to computer programs than they actually possess. In a culture of simulation people quite readily come to treat computers as social actors. This can be coupled with the fact that in interactions with other previously unknown people, we readily tend to form specific impressions of the personalities of our interlocutors, in spite of having little real evidence to allow us to do so. Given these preconditions, it emerges that exclusively computer mediated interactions between real people can sometimes facilitate the development of highly intense (hyperpersonal) emotional relationships between people who meet online. In this kind of situation, argues Carolyn, there is a strong need to establish in reliable ways in encounters online who we can trust, who we can learn from, whether they are like us or strange and challenging, whether we can dominate them or them us, whether they will enthral or disgust us. Being able to do so in the longer term is almost more important than knowing whether the other we are communicating with is a real human being or an artificial agent. Her main contention then is that testing for rhetorical ethos – mobilising to the maximum our ethopoetic impulse, both in representing ourselves online and in evaluating our interactions with others we meet there, will be more and more important as time goes on as an aspect of our increasingly networked everyday lives.