13.0488 science, formal methods &c.

From: Humanist Discussion Group (willard@lists.village.virginia.edu)
Date: Tue Mar 14 2000 - 08:21:28 CUT

  • Next message: Humanist Discussion Group: "13.0487 conventions - the list"

                   Humanist Discussion Group, Vol. 13, No. 488.
           Centre for Computing in the Humanities, King's College London

             Date: Tue, 14 Mar 2000 08:16:50 +0000
             From: Willard McCarty <willard.mccarty@kcl.ac.uk>
             Subject: science, formal methods &c

    Following is an exchange of letters between Richard Giordano and myself on
    the topic introduced by my announcement of the Colloquium "Humanities
    computing: formal methods, experimental practice" here at King's College
    London 13 May. Comments are of course most welcome. --WM

    >Date: Mon, 13 Mar 2000 13:35:21 -0500 (EST)
    >From: Richard Giordano <richardg@HOOVER.MIT.EDU>
    >To: willard.mccarty@kcl.ac.uk
    >This is a post to you personally. I don't mind if you post it on Humanist,
    >but I don't think the wider group would be interested in reading this.
    >It's your call. By the way, I'm the person who posted the questions, "Is
    >there a difference between a formal method and a structured method? Are
    >the use of structured methods more appropriate in this regard? Why is
    >scientific investigation privileged? Are methods of design and
    >investigation in technology as appropriate as science?" I was finishing a
    >paper (which I gave over the weekend at Stanford on, uh, knowledge in
    >technology communities) and wrote those questions to you very quickly when
    >the Humanist email arrived.
    >As you probably know, I got my PhD in American history, but have worked in
    >the science and technology communities since about 1988, having taught CS
    >at the University of Manchester. At MIT, I am working on a few projects
    >concerned with the nature of knowledge in technology communities,
    >especially, roughly speaking, knowledge generation, knowledge transfer,
    >the emergence of communities of practice, and what not. All this is to
    >say that I have had a long and complex exposure to different work and
    >thought styles over the years. Hence the question on science. Why
    >reference science as a system of knowing? Why not think more broadly about
    >the sociology of knowledge in general and see how hat plays out? Collins's
    >work is at root, at least to me, concerned with the sociology of coming to
    >know. And there are many ways to know, even in science. Harry Collins,
    >whose work I both know and admire, would agree with this. I'm just a bit
    >confused why you're concerned with how scientists come to know. What
    >makes them special?
    >About 'formal' v 'structured' methods. When engineers work on a problem,
    >they go about within a framework that guides their actions. This helps
    >them to know which steps to take when, and helps them to manage
    >complexity. This is called a structured method. It's something of a
    >design method. I am familiar with formal methods from computer science.
    >A formal method is a verifiable method. Typically, the method refers to
    >mathematics for verification. This helps you to know when your work is
    >correct, altough not neccesarily complete. By using the term 'formal
    >method' in the context of science, I'm wonderin if your using it in this
    >context, if you mean a structured way of working, or if you want to avoid
    >using the phrase scientific method. If you mean a verifiable method, then
    >the question is do we mean that research results can be replicated? Is it
    >a structured methodlogy? If it's the latter, what discipline does not use
    >a structured methodology? Because I get so easily confused, thought that
    >going back to first principles and looking at the sociology of knowledge
    >in general would help.

    >Date: Mon, 14 Mar 2000 08:11:21
    >From: willard.mccarty@kcl.ac.uk
    >To: Richard Giordano <richardg@HOOVER.MIT.EDU>
    >At the end of his book Changing Order, Collins says that he hopes that
    >with the other work he has done and that of his colleagues it "will be
    >seen to have had three consequences: to have changed the way we study and
    >understand the history, sociology, and philosophy of science; to have
    >changed the relationship of science with other cultural endeavours; and to
    >have become a foundation for the science of knowledge" (p. 191).
    >So, yes, his objective (and I think ours) is aimed at understanding how we
    >come to know, what we mean by "knowledge", well beyond the confines of the
    >sciences. My look toward the sciences is part of a project to understand
    >and articulate the practice of humanities computing. It seems to me a
    >rather obvious thing to do, to look toward the practices of others who
    >also use equipment as part of their epistemological activities, who also
    >make this equipment the centre of their collaborations, who share the
    >activity of modelling with us and who, for various reasons, have attracted
    >the attentions of historians, philosophers and sociologists, through whom
    >we can begin to understand the commonalities. Why turn down or aside from
    >some rather obvious parallels? Paying attention to the sciences is not
    >necessarily to privilege them, esp when one pays attention to them via the
    >likes of Galison, Hacking and Collins, with a good dollop of Searle on the
    >The analogy of the experimental sciences, it seems to me, raises some
    >rather interesting questions, the answers to which are not obvious, or at
    >least not to me. Once one gives up on the "spectator view of knowledge",
    >as Hacking calls it, then how does one formulate the difference between
    >the sciences and the humanities? Lots of quicksand here, but I see no
    >other path ahead that is quite so compelling. And the topic is so clearly
    >in the air, which means we have many interesting, smart people to talk to
    >about what we're doing. Not only is there the flood of books and articles
    >in those disciplines but also discussions in the wider community. The
    >latest issue of the London Review of Books, for example, has a long piece
    >by Richard Rorty, "The End of the Epistemic Wars", on this topic, more
    >precisely, on the convergence of the sciences and humanities because of
    >the current work in philosophy, history, sociology of science. I don't
    >want to oversimplify the complexities of the arguments involved, nor to
    >pretend that I can follow them all, but I am utterly convinced that our
    >minds need to be in the midst of the debate.
    >Our colleagues make reference all the time to what we do as "scientific",
    >and we continually puzzle over our relationship with computer science.
    >Would it not be better if we understood what "scientific" means, when it
    >means anything at all beyond the purely, dangerously honourific term that
    >Searle would rather avoid? Would it not be better if we actually
    >benefitted from the experience and wisdom of others in and around the
    >sciences? At the end of my project it may become obvious that the analogy
    >with experimental science is not so good after all, but knowing what we're
    >not about is a positive step. It is most certainly the case that as you
    >say the sciences represent only one part of the epistemological scene. All
    >I'm saying is here's a good place to begin.
    >About "formal methods". In the context of the Colloquium this term refers
    >to an argument that one of our speakers, Tito Orlandi, will be making. I
    >don't want to try to say what he will say on 13 May. If your Italian is in
    >good working order, you can start with his 1997 publication, INFORMATICA,
    ><http://rmcisadu.let.uniroma1.it/~orlandi/formaliz.html>. In a forthcoming
    >paper I go at this question by assuming the most ambitious claim one could
    >make, i.e. the claim of strong AI; the papers in the Stanford Humanities
    >Review 4.2, previously announced on Humanist, are helpful. Replication, as
    >you know from Collins and Hacking, is a very complex idea; they also deal
    >with the multiplicity of scientific methods. So, to answer your question,
    >I didn't mean much in particular by "formal methods", though I think that
    >the formalisation forced upon us by use of computing is a very fruitful
    >topic indeed.
    >PS I am publishing your note and my reply on Humanist; BOTH authors have
    >given their permissions :-).

    - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
    Dr. Willard McCarty, Senior Lecturer, King's College London
    voice: +44 (0)171 848 2784 fax: +44 (0)171 848 5081
    <Willard.McCarty@kcl.ac.uk> <http://ilex.cc.kcl.ac.uk/wlm/>
    maui gratia

    This archive was generated by hypermail 2b29 : Tue Mar 14 2000 - 08:38:58 CUT