Humanist Discussion Group, Vol. 13, No. 443.
Centre for Computing in the Humanities, King's College London
<http://www.princeton.edu/~mccarty/humanist/>
<http://www.kcl.ac.uk/humanities/cch/humanist/>
[1] From: Jan-Gunnar Tingsell <tingsell@hum.gu.se> (29)
Subject: Re: 13.0439 how to get what we need: build it
[2] From: "Osher Doctorow" <osher@ix.netcom.com> (21)
Subject: Last and Next 5 years
[3] From: Willard McCarty <willard.mccarty@kcl.ac.uk> (52)
Subject: if only we could speak it easily?
--[1]------------------------------------------------------------------
Date: Fri, 25 Feb 2000 20:38:01 +0000
From: Jan-Gunnar Tingsell <tingsell@hum.gu.se>
Subject: Re: 13.0439 how to get what we need: build it
At 06:42 00-02-25 +0000, C. M. Sperberg-McQueen wrote:
>The bad news is that getting input from users on new features for software
>is notoriously hard work. Eliciting, from users, a description of a new
>kind of software is even less likely to succeed easily. So I don't have
>much hope of survey-driven software development -- even if we get surveys
>designed and performed with more sociological acumen than those of which
>the lash-wielder complains. I am more sanguine about the prospects of
>success if people who know they want a particular kind of software build
>it, and show it to other people to see if they like it. The results will
>have no statistical validity or significance at all. But they may include
>some useful software.
>....
I do agree. It is an impossible question to ask most people. My experience
is that they don't know what is possible to wish.
Most of the "Internet concept" is (of course you know this already) that
people write their own, often rather small programs they have need for and
then share the programs with others. Many people then say: "Oh, that is
just what I have been looking for!", they run the programs and sometimes
write or suggest improvements. Collaboratively they (or shall I say we)
use and appreciate the great amount of code we have access to in this way.
Why not copy this concept in the Humanities?
I would like a co-ordination of all good ideas I think exist out there,
a place where it is easy to inform about programs and an archive where
it is easy to identity free "Humanistic" software code.
-- Jan-Gunnar Tingsell <tingsell@hum.gu.se> Centre for Humanities Computing tel: +46 (0)31 773 4553 Gteborg University fax: +46 (0)31 773 4455 URL: http://www.hum.gu.se/hfds/--[2]------------------------------------------------------------------ Date: Fri, 25 Feb 2000 20:38:48 +0000 From: "Osher Doctorow" <osher@ix.netcom.com> Subject: Last and Next 5 years
Concerning the very interesting new topic on what the field needed and needs in 5 years, I would guess just on the basis of looking at the email of the last few weeks that a shift from the local to global orientation might be useful. The use of automated reasoning to answer local questions is the Bayesian approach, while the Logic-Based Probability approach is interested in both global and local questions. To begin using LBP very effectively, it would be wise to formulate some global questions, for example about origins of various language families and related conjectures. You can see a sample of some global questions on my University of Vienna Institute for Logic abstracts site. For those who do not have the time to look, I would suggest an algorithm that goes something like this: 1. Is it uniformly distributed? 2. Does it involve a boundary of an object in space? 3. Does it involve the boundary of an object in time, including initial and terminal conditions. 4. Does it involve rare event(s)? 5. Does it involve a central or critical point of an object? 6. Does it involve an internal distinguished structure of an object? 7. Does it involve a subset of an object? I would go so far as to suggest supplementing the Socratic method with these questions, with or without a computer, although eventually it should not be too hard to build up enough data and questions and answers for a mechanization. (I can just see the Bayesian alternative: "keep on sampling," or "given a sample which is based on a sample given a second sample which is...,".).
--[3]------------------------------------------------------------------ Date: Fri, 25 Feb 2000 20:40:06 +0000 From: Willard McCarty <willard.mccarty@kcl.ac.uk> Subject: if only we could speak it easily?
In Humanist 13.439 Michael Sperberg McQueen responds to the anonymous critic of surveys:
>The bad news is that getting input from users on new features for software >is notoriously hard work. Eliciting, from users, a description of a new >kind of software is even less likely to succeed easily. So I don't have >much hope of survey-driven software development -- even if we get surveys >designed and performed with more sociological acumen than those of which >the lash-wielder complains. I am more sanguine about the prospects of >success if people who know they want a particular kind of software build >it, and show it to other people to see if they like it. The results will >have no statistical validity or significance at all. But they may include >some useful software.
He goes on to recommend the building of prototypes.
Since for us a properly done sociological survey is but a means to an end, we can easily follow his cogent recommendation in the short term by making prototyping easier, in the long term by working on a high-level programming environment that the ordinary humanist could use him- or herself.
Some years ago I was persuaded by Geoffrey Rockwell that one good way to help humanists along our path institutionally is to set up prototyping labs or services from which people with ideas could get prototypes built. I suppose that this happens informally in a number of places; have these labs been formally set up anywhere? One could easily imagine a lab staffed by apprentices from CS programmes who would get very useful training that way. A good common ground on which the humanities and CS could work out a working relationship?
THe high-level programming environment I've been tossing about mentally for the last several months. What I've been imagining is based on something called Explorer, a piece of scientific visualisation software John Bradley showed me years ago. It provides a visual interface with Lego-like pieces one sticks together; data goes in one end of the construct, results come out at the other end. I am also stimulated to think like this because of something Antoinette Renouf (Liverpool) has used to illustrate her work on neologisms in English: a diagram reminding me of an industrial process, huge amounts of newspaper text going in one end, neologisms coming out the other. (For her work see <http://www.rdues.liv.ac.uk/>.) There's much more to humanities research than that, of course, but we know that mechanical processes are a part of what we do. What if we had a visually-orientated "programming" environment of Lego-like primitives we could plug together? One thing we'd get is a lot of people making things, playing, discovering, trading instantiated ideas.
As I've suggested before, all this rests on identifying a set of more or less standard primitives, a modern UNIX toolbox. The research question is, what are those primitives? One approach to finding out would be a good sociological study of how we do research, yes? Not asking what we want -- we don't know, we're not good at articulating all those deep-down desires -- but asking, what do we in fact do?
Yours, WM - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 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 : Fri Feb 25 2000 - 20:52:13 CUT