| Yes, although Michael Heim and I come from very different backgrounds - his is Heideggerian, mine Peircean - we appear to have reached rather similar conclusions. In the nineties I was more of a techno-optimist, Michael more of a techno-pragmatist. Later developments have sobered me considerably; in my teaching I have found that e.g. search engines have brought my students losses as well as gains, in that their capacity for sustained reading appeared to have diminished.
Best, Peter Sent from my iPhone Peter, List, Thank you for adding this further background information, Peter. Your comparison between machine translation and translation-memory tools was especially illuminating. The latter seems to me almost a paradigm case of what we've been discussing semiotically: not replacing the interpreter but, rather, extending the network of signs, habits, and prior interpretants upon which further interpretation -- and translation -- depends. It also reinforces what I have taken to be the broader significance of your paper on virtuality. For, if thought is already virtual in Peirce's sense, then technologies such as these are valuable precisely insofar as they augment the continuing development of the semiosis involved in any inquiry (taken in the broadest sense). I also took a look at a review of Michael Heim's The Metaphysics of Virtual Reality. Reviewing the book for Computers and Society, Zach Hildebrandt characterizes Heim as arguing that the computer -- and now one might add, AI -- should function as a component rather than an opponent of human thought, and that the highest goal of virtual reality is not technology for its own sake, but the enlargement of imagination and human experience. If that is a fair characterization, then there may be more common ground between your Peircean account of virtuality and Heim's phenomenological approach than I had initially thought. Best, Gary R Thank you
for that further elaboration, Gary. In yesterday’s post, while I cited my
Macintosh’s GUI as an instance of Engelbart’s intelligence augmentation, I
might have mentioned that this was no accident, since Engelbart invented GUIs.
Engelbart and Ivan Sutherland, I should say, since Engelbart would not have
gotten very far without Sutherland’s bit-mapped graphics.
To
continue. In the nineteen nineties, while teaching philosophy part-time at U
Mass Lowell, I also managed operations in a foreign-language translation
company. In that capacity I was also offsite manager of Compuserve’s online
translation service for a couple of years, until AOL bought Compuserve and
promptly cancelled the translation service. During those years I was profoundly
impressed with the contrast between computerized machine translation (MT) and
computer-aided translation using translation memory tools such as Trados, which
allows human translators easily to retrieve and reuse previously translated
chunks of text – sentences as well as modules – thus dramatically increasing
the efficiency of human translators. Again, I was faced with a striking example
of computerized intelligence augmentation as articulated by Engelbart, and as
explained by Peirce’s semiotic model of the mind.
MT at the
time presented a very different spectacle. Here the machine was doing the
actual thinking, i.e. translating, while in its then primitive state MT
required the assistance of human translators to carry out such mundane tasks as
cleaning up the machine’s output. Not having kept up with the technology I
expect that today MT has benefited substantially from AI, but that was not the
case at the time. Of course, MT arose in the nineteen fifties from the
hypothesis that human languages are codes and that translating a text from one
language to another was in effect decoding, something machines had proven quite
successful at during World War Two. By the nineties reality had set in, and MT
developers understood for instance the need to build semantics into the
programs, but the results were still rather primitive.
That
provides a bit of subtext for my virtuality paper, as well as other papers
published in the nineties. I should add my ideas about the relevance of Peirce’s
semiotics for computerized intelligence augmentation were originally rather
rough and were sharpened and refined through discussion on Peirce-L starting in
October 1994. The late Joe Ransdell, in particular, was exceptionally helpful
with his incisive comments and questions.
Enough for
now.
Best,
Peter
Peter, List,
I'm gratified that you found my summary and interpretation of your paper "clear and perception." I was in truth eager to recommend it to the List. Thinking again about your work (and Doug Engelbart's), I have come to believe that you were both ahead of your times.
Thank you for pointing to Michael Heim's book in your post. I found this review of Heim's Electric Language especially interesting when read alongside your article. See: "The Corrupt Computer: Electric Language: A Philosophical Study of Word Processing by Michael Heim," reviewed by Edward Mendelson, The New Republic. https://mendelson.org/Michael-Heim-Review.pdf
Mendelson largely treats the computer as a technology that competes with and often diminishes human cognition. His view seems more skeptical than Heim's in arguing that computers often compete with rather than support reflective thought. Heim himself appears to hold a somewhat more ambivalent position: Electronic technologies transform thought in significant, even profound ways, opening new possibilities for thought -- but also threatening traditional forms of cognition. On the other hand, I read your article, 'Peirce, Virtuality, and Semiotic' as suggesting that what I've referred to as 'external symbolic media' are constitutive of thought itself. Whereas Mendelson worries that computers may replace reflective and creative thought, you emphasize that thought has always been mediated by external signs and symbolic tools.
Further, while you, Heim, and Mendelson all appear to see computation as significant for writing, especially philosophical writing, yet you begin from very different traditions: Heim from Heidegger and continental phenomenology, Mendelson from the literary-critical tradition, and you from pragmatism and semeiotic; and with some significant consequences. Your work tends to shift the focus from the phenomenology of the user or the practices of the writer to the community of inquiry. For you technologies are neither mere tools (as Mendelson tends to treat them) nor essentially existential 'contexts' (as in Heim), but actual, albeit devative, participants in the continuing growth of signs, thought-habits, and communal inquiry. I think that distinctively Peircean move has considerable potential for reframing contemporary debates, not only about contemporary technologies, but for thought itself.
I look forward to your post on foreign language translation,
Best,
Gary R
Gary, List,
Thank you, Gary, both recommending my paper and for the clear and perceptive summary and interpretation of it. When I wrote this paper in the nineteen nineties I regarded IA and AI as rival research programs: one regarding computers - like pencils, pen and ink, etc. - as tools for human thinking, the other seeking to locate thinking in the computer itself. As noted in an earlier post, I am not really up to date on current AI research, but I have come to see the two research programs as to some extent complementary, as one use of AI is to improve human-computer user interfaces.
What motivated me to write about these matters in the nineties - other than my obvious interest in Peirce’s thought - were two very practical experiences, to wit, word-processing and foreign-language translation. I found my days spent working with an MS DOS computer in the office while writing about philosophy with a Macintosh at home, and I found myself thinking differently - more easily - with the Macintosh than with the pc, as my thinking was not distracted by the requirements of the program. Around this time I also read Michael Heim’s ‘Electric Language’, published in 1990, which discussed the effects on thought of using command-based word-processing. It was actually from Heim’s book that I first heard of Engelbart, and my virtuality paper was presented at a panel headed by Heim at the 1997 Pacific Division Meeting of the APA.
I have to run now, but will post a few words on foreign-language translation tomorrow.
Best, Peter Sent from my iPhone List, I'd like to strongly recommend Peter Skagestad's paper, "Peirce, Virtuality, and Semiotic," which seems remarkably prescient despite having been written decades before today's AI systems. But perhaps more important is what it suggests about virtuality in human semiosis. https://www.bu.edu/wcp/Papers/Cogn/CognSkag.htmSkagestad begins with Peirce's own definition from Baldwin's Dictionary of Philosophy and Psychology (1902): "A virtual X ... is something, not an X, which has the efficiency (virtus) of an X." Virtuality, in this sense, concerns functional capability, not appearance. He then argues that this idea is deeply embedded in Peirce's semeiotic itself: Since every Sign stands in an irreducibly triadic relation to its Object and Interpretant, and every Interpretant usually becomes a further Sign, semiosis is an open-ended, developmental process. So the meaning of a sign lies not in its present occurrence but in its possible future interpretants and interpretations. From this follows what Skagestad takes to be one of Peirce's most radical insights: Thoughts are themselves signs, and because signs have their being through their possible future interpretants, thought itself is fundamentally virtual. As Peirce puts it, "the meaning of a thought is altogether something virtual." Mind is not something fully present at any instant but an ongoing developing process extending through time. This perspective also helps clarify Peirce's famous remark that if someone stole his inkstand "the very thoughts would not come to me." Skagestad makes clear that Peirce isn't saying merely that writing records thoughts. Instead, pens, paper, diagrams, logical notation, and other symbolic media participate in the production of thought itself. As argued in other related threads, human intelligence has always depended upon such external symbolic technologies, from alphabets and books to computers and, today, networks which extend -- and I believe that List members agree, not replace -- cognition. Skagestad concludes that Peirce's semeiotic offers perhaps the most promising philosophical framework for understanding both human thought and virtual technologies. Because thought consists of signs, and signs are essentially virtual, and that cognition depends upon external media, then modern computational technologies are best understood as augmentations of human inquiry rather than as substitutes for human minds (read: Douglas Engelbart's IA). I believe that Skagestad's paper deserves renewed attention. Although written before the contemporary AI era, it anticipates many of the questions we are now asking and does so by returning directly to Peirce's own conception of signs, virtuality and, in my view, the growth of inquiry. Best, Gary R
|
