Humanist Discussion Group, Vol. 33, No. 560. Department of Digital Humanities, King's College London Hosted by King's Digital Lab www.dhhumanist.org Submit to: firstname.lastname@example.org Date: 2020-01-18 12:20:59+00:00 From: Jan Christophe Meister
Subject: Re: [Humanist] 33.558: untranslatable text and illustrative analogies or metaphors? Dear Willard, no (1) is easy: Russel's paradox (aka the barber paradox) marked up in (first order) predicate logic. Whether Zermelo et. al. provide a solution to this is another question (to be answered by Michael Sperberg-McQueen et al ... I profess ignorance.) Probably a more tangible variant from a Humanist perspective are the various manifestations self-reflexivity encountered - or rather, encoded - in symbolic artefacts, such as metalepsis in narratives which Genette (1993) describes as the "deliberate transgression of the threshold of embedding [...]: when an author (or his reader) introduces himself into the fictive action of the narrative or when a character in that fiction intrudes into the extradiegetic existence of the author or reader." For a more detailed definition of metalepsis see Pier, John: "Metalepsis (revised version; uploaded 13 July 2016)", Paragraph 1. In: Hühn, Peter et al. (eds.): /the living handbook of narratology/. Hamburg: Hamburg University. URL = (http://www.lhn.uni-hamburg.de/node/51.html) Using this example let us assume you want to markup the ontological levels (intradiegetic = story world of the characters; extradiegetic = 'world' of the narrator) in a narrative and then use this markup for, say, the purpose of analysing narrative embeddings computationally. Unless you provide the system with a "cut" condition it will loop - for a demonstration see my 2003 article "The /Metalepticon/: a Computational Approach to Metalepsis" (available at https://www.jcmeister.de/downloads/texts/jcm-metalepticon.html) where I did this using a PROLOG encoding. I think that from a philosophical point of view it all boils down to a simple fact: the problem with computers is that they don't "have" problems in the emphatic sense. Nothing in the design of a computational machine will inspire it to look out for the type of existential problem that motivates our human intellect and culture. We can, at best, try to capture those problems using declarative markup in order to operationalize them for the machine, but then the problem looses its existential, referential edge. Unless of course you want to equate a piece of looping code (does the machine experience anguish and anxiety as it begins to run out of RAM?) with our human existential dilemma. As for me running a syntax check and encountering questions of being just don't fall into the same category - and so we'd probably be well advised to meet the current (and cyclically recurring) AI-hype with a bit of Kantian foundationalism along the lines of his remarks on metaphysics in his 1783 "Prolegomena": there are questions which are irresolvable for the 'ratio'. And luckily we happen to be beings that don't just run on that principle. Best, Chris Am 18.01.2020 um 09:49 schrieb Humanist: > Humanist Discussion Group, Vol. 33, No. 558. > Department of Digital Humanities, King's College London > Hosted by King's Digital Lab > www.dhhumanist.org > Submit to: email@example.com > > > > > Date: 2020-01-17 14:28:02+00:00 > From: Willard McCarty > Subject: an example of untranslatable text > > I'm in need of two rhetorical devices, as follows: > > (1) a few very short examples of text that would put up > obviously insurmountable resistance to a satisfactory rendering in any > computational markup language. A problem that would pose such an > impossibility would be an ambiguity that must not be resolved and whose > range of renditions is indefinite. For my purposes the text should be in > ordinary, everyday English. > > (2) analogies or metaphors that were or are used to characterise > artificial intelligence. Historical examples are 'fast moron' and > 'idiot savant'. The first of these (according to Pamela McCorduck) > was invented at IBM to allay fears of its customers, but its persistence > suggests the ring of truth. It's more or less vanished, but the second > remains popular. > > The problem is not to bad-mouth AI, rather to illustrate to a novice > that any possible intelligence of which the digital machine could be > capable has specific characteristics that can be traced back to its > design. These characteristics, I am wanting to argue, are limited, > as the characteristics of any definite thing are limited, but not so > as to make an AI necessarily less 'intelligent', only differently so. > > Many thanks for examples -- or for more intelligent framing of the > problem :-). > > Yours, > WM > -- > Willard McCarty (www.mccarty.org.uk/), > Professor emeritus, Department of Digital Humanities, King's College > London; Editor, Interdisciplinary Science Reviews > (www.tandfonline.com/loi/yisr20) and Humanist (www.dhhumanist.org) ------------------------ Dr. Jan Christoph Meister Universitätsprofessor für Digital Humanities - Schwerpunkt Deutsche Literatur und Textanalyse - Universität Hamburg, Institut für Germanistik Überseering 35 / Raum 08064 22 297 Hamburg +49 40 42838 2972 +49 172 40865 41 http://jcmeister.de _______________________________________________ Unsubscribe at: http://dhhumanist.org/Restricted List posts to: firstname.lastname@example.org List info and archives at at: http://dhhumanist.org Listmember interface at: http://dhhumanist.org/Restricted/ Subscribe at: http://dhhumanist.org/membership_form.php
Editor: Willard McCarty (King's College London, U.K.; Western Sydney University, Australia)
Software designer: Malgosia Askanas (Mind-Crafts)
This site is maintained under a service level agreement by King's Digital Lab.