Home About Subscribe Search Member Area

Humanist Discussion Group

< Back to Volume 32

Humanist Archives: Feb. 26, 2019, 5:47 a.m. Humanist 32.491 - editions, in print or in bytes

                  Humanist Discussion Group, Vol. 32, No. 491.
            Department of Digital Humanities, King's College London
                   Hosted by King's Digital Lab
                Submit to: humanist@dhhumanist.org

    [1]    From: C. M. Sperberg-McQueen 
           Subject: a correction (50)

    [2]    From: C. M. Sperberg-McQueen 
           Subject: nothing replaces the object - or does it? (104)

    [3]    From: Dr. Herbert Wender 
           Subject: Re: [Humanist] 32.489: editions, in print or in bytes (62)

    [4]    From: philomousos@gmail.com
           Subject: Re: [Humanist] 32.489: editions, in print or in bytes (140)

    [5]    From: savoye@eapoe.org
           Subject: Re: [Humanist] 32.487: editions, in print or in bytes (24)

        Date: 2019-02-26 05:38:02+00:00
        From: C. M. Sperberg-McQueen 
        Subject: a correction

I must apologize to anyone misled by my characterizations, in my
answer to Dana Paramskas, of the positions various people have taken
in the discussion that has been running on Humanist first under the
rubric "the McGann/Renear debate" and in the current thread.

I wrote:
     At least one or two participants in the discussion have argued
     that any attempt to represent the text of multiple textual
     witnesses in a single electronic document will necessarily cause
     painful difficulties in the electronic document, and further that
     the hierarchical structure of SGML and XML documents makes the
     difficulties even worse than they would otherwise be.

It appears that in my effort to provide as plausible a statement of
each argument as I was capable of, I inadvertently attributed to the
makers of this argument a position they do not in fact hold.

Desmond Schmidt has objected to the paraphrase on the grounds that
those whose views I was attempting to describe (which group includes
him) have not in fact ever admitted that the satisfactory
representation of multiple witnesses is difficult, but have on the
contrary only argued that it's difficult in SGML and XML.  In the
notation he has invented, he says, this task is not painful but
"simple and easy to edit".

I apologize for mischaracterizing DS's argument.

DS's posting touches on a number of other issues, but adds no
information useful for those trying to understand his claims in
earlier postings.  I will spare myself and any remaining readers of
this threat the point by point response and the correction of DS's
striking misreading of the IBM manual on the GML starter set.
DS and I have not, I regret to say, reached a level of clarity and
mutual understanding which would allow us to agree on a
description of the areas in which we do not agree,  We have
merely reached a point at which I doubt that further discussion
of DS’s views on the origins of SGML is likely to be useful,
enlightening, or entertaining.

C. M. Sperberg-McQueen
Black Mesa Technologies LLC

        Date: 2019-02-26 05:36:27+00:00
        From: C. M. Sperberg-McQueen 
        Subject: nothing replaces the object - or does it?

In Humanist 32.489, Willard McCarty writes:

     My particular concern was with human and artificial agencies in
     the act of note-making and what I called knowing-by-doing -- not
     just Ryle's 'knowing how' but the knowing in enacting how. The
     psychology of art would seem a place to go. I would appreciate not
     just specific pointers into Rudolf Arnheim's work but also any
     other suggestions that you might have.

I wonder if I could induce you to expound a bit on how the topics you
mention connect with the topic of scholarly editions.  Is it the task
of a scholarly edition to capture and represent the performative
knowledge of which you speak?  Or is it the task of an edition to
shape itself in ways that support the development and profession of
such performative knowledge in the reader?  In either case, it would
be good to have some clearer idea of what you are talking about.

     I would be so bold as to say that nothing replaces the
     face-to-parchment (skin-to-skin) work with manuscripts, even
     though getting to see these rare items is expensive and

I agree, but WMcC may not find my reasoning wholly satisfactory.

My paleography instructors made a point of stressing that no edition
of any manuscript can legitimately be made without autopsy of the
manuscript.  They came as close as statute would allow to making us
swear in blood that we would never publish a transcript made only from
a microfilm.  I have never had a reason to think they were wrong.

So, yes, nothing can replace autopsy.

But it would be a very ill wind that blew nobody any good.  Nothing
replaces face-to-parchment work, but for manuscripts we find hard to
read, it is equally true that nothing replaces work with image
manipulation software.  With a scanned image, we can perform
transformations which librarians and archivists would frown on us
attempting on the original manuscript.  If a student came to me
with a transcription of a difficult manuscript full of passages
marked uncertain or illegible, and explained that they had not
used existing scans of the manuscript to attempt to render those
passages legible, I do not think I would be able to advise that they
publish an edition based on their transcription in its current

And in the same vein, nothing replaces work with a word frequency
list.  The list could be extended a long time, perhaps indefinitely.

That is, I agree with what you say, but I do not think that physical
contact with a manuscript is distinguished thereby from work with
a digital surrogate.

Nothing replaces a useful tool.

     Similarly, I'd say that nothing replaces the physical manipulation
     of notes. Perhaps a Minority Report device could equal index cards
     on a table or floor, but who could afford that?

On the same principle as above, I think this risks being a truism --
no thing replaces any other -- but I think you wish to suggest that
physical manipulation of notes is better than the alternatives.  I
think I part company from you here.  In my experience scholars have an
absolutely bewildering variety of styles in the way they construct and
work with their notes, and it is by no means the case that physical
manipulation of notes plays an important role for everyone.  If you
believe that no one can ever work as effectively with a SQL database
as with a shoebox full of 4x6 slips (or, for the unenlightened, of 3x5
cards — or [God forbid] an erratic mixture of cards and slips,
3x5 and 4x6), then I think you are falling victim to a failure of
imagination about the inner lives of others who work with notes.

If a distinction between any X and Y is perceptible, it is almost
certain that some people will prefer X and others Y.  If you prefer playing
solitare with your notes in order to organize them, someone will
prefer tossing them into the air above a staircase and challenging
themselves to write a compelling discourse starting with the
topics mentioned on the lowest stair and ending with those on
the highest.  By a wide margin the conference talk on which I got the
most compliments in my life so far was one in which my notes were lost
entirely,  my hard disk died, and I wrote the slides the night before the
talk with the help of a soft-hearted graduate student,  using software I
had never seen before under an operating system set up for a
metropolitan French keyboard on a machine with the keys labeled for a
Canadian French layout (so every other slide or so I had to consult the
student about which key to press to obtain a particular
character).  The experience suggests to me that in my case perhaps
the right thing to do would be to replace physical manipulation of
notes with complete physical inaccessibility of notes.

     (You are not hearing the sound of a cane thumping the floor! :-)

Well, no, but I could turn up my hearing aid, if that would help.

C. M. Sperberg-McQueen
Black Mesa Technologies LLC

        Date: 2019-02-26 00:45:09+00:00
        From: Dr. Herbert Wender 
        Subject: Re: [Humanist] 32.489: editions, in print or in bytes

Just two cents from Wikipedia:

"Broadly, any metalanguage is language or symbols used when language itself is
being discussed or examined. In logic and linguistics, a metalanguage is a
language used to make statements about statements in another language (the
object language)."

I would mean that in most caseswe can agree in stating the hierarchy

S-GML, i.e. Standard for (a certain typpe of) generic markup language*s*

XML in a TEI conformant specification

NL, i.e. natural language 'chatacter data'

Insofar the 'content objects' underlying a given markup are NL expressions, the
markup as usage of the specified metalangugage  allows to exxpress certain
assertions on (parts of) the underlying textual object.(s)
[BTW: One shouldn't wonder about ambibuous usage of the acronym SGNL with
regards to the imprecise name (ending in the singular 'language') and the
parallel usage as filetype (f.e. OTA's file "1824.sgml" bearing the TEXT-part to

Coming back from some offline days I was surprised by the new name and a certain
turn in the thread begun under the names McGann and Renear. It will probably
take some times to order my thoughts on the new questions. But with respect to
the older ones I wish to express quite now my disappointment regarding the new
massterpiece of clarification what may be meant by the term 'text', and this in
a manner that makes rhetorically disappear some problems already stated and yet
not ansswered, f.e.
- what kind of 'content object' is Goethe's mdash in the mentioned ROSENKNOSPEN
- would a 'floating' speaker name break a TEI rule or not?
- what hierarchy governs the relationship Speaker-Speech (understood as spoken
lines of text)?
- why the CONCUR option foreseen by SGML never (?) was implemented?

And after all these details I would like to come back to the general question:
Can we really expect that using a formalized language suited to model language
competence and (in the case of SGML conformancy) invented to regularize highly
formalized communications in governmental, technical and scientific (narrower
sense) contexts will ever be fitted to sufficiently describe language
performances as in the writing of literary authors?

My point of view is decidedly not the one which has shown Desmond Schmidt in the
new turn of the debate on analog / digital divide. The label "editions, in print
or in bytes" fails to reach the point in which I'm interested in - and it
neglectss definitely nowadays pubishing workflow: Where is a newly printed
edition which wasn't prior in bytes? Why, the hell, no one is speaking here
about DTP? (i.e. Desk Top Publishing) Why no mention to PDFs? (Page Description
Could anyone explain why f.e. the Stroemfeld facsimile edition of Kafka's "Der
Process" or the edition of Hölderlin's Works by the same publisher aren't
valuable as scholarly editions? Otherwise asked: What can legitimize the great
funding efforts in SGML/XML/TEI conformant encoding projects as "Heinrich Heine
Portal" or "Faustedition"? Where ist the "stupid softwate" to process such
"smart data" (Sperberg-McQueen) ?

With kindly regards (and the usual apologies)

        Date: 2019-02-25 18:22:00+00:00
        From: philomousos@gmail.com
        Subject: Re: [Humanist] 32.489: editions, in print or in bytes

Whereas most of us think in sentences, Michael appears (to me anyway) to
think in full, coherent paragraphs. I don't think there is any attempt on
his part to bludgeon you into silence.

> I presume by these two long postings your intention is to swat the
> annoying fly that I have become. I'm sorry to disappoint you.
> > At least one
> > or two participants in the discussion have argued that any attempt to
> > represent the text of multiple textual witnesses in a single
> > electronic document will necessarily cause painful difficulties in the
> > electronic document, and further that the hierarchical structure of
> > SGML and XML documents makes the difficulties even worse than they
> > would otherwise be.
> This is precisely what my model of text addresses and delivers on: Any
> number of versions in one electronic document, each of which is simple
> and easy to edit. If you agree here with this summary the SGML/XML
> model fails us in this essential task.

Since it is trivially possible to have more than one document to represent
a 'version' (I assume by this you mean a witness, or an editing stage, or a
previous edition, depending on the circumstances), and moreover since doing
so means you can put them in a version control system (different sort of
version), I still fail to see how this is a win.

> > The assumption that "elements and attributes" constitute "metadata" is
> > also not one I think can be taken for granted.  The idea that "markup"
> > is always and only "metadata" is not hard to find, and is often useful
> > when teaching beginners the rudiments of markup, but it's hard to take
> > seriously as a philosophical statement and -- like the concept of
> > "metadata" itself -- does not (in my limited experience) withstand
> > sustained scrutiny.
> I used the term metadata because I just wanted a general term to hang
> elements and attributes off. Both are "data about data" - the
> definition in the dictionary - which is good enough for me. In
> SGML/XML they both describe the text nodes, the content, in layman's
> terms the stuff not in angle-brackets. They still do that even when we
> philosophise about whether metadata are also data.

This is not completely correct. Markup can also represent (e.g.) no text at
all (), or unrepresentable text (), or structural divisions
(,,pb>). There's not in fact a bright line between text nodes as
data and the element and attribute nodes as metadata.

> As for your long discussion about what can or cannot be done with
> interlinked elements I'm afraid my interest in XML waned after 2004
> when I saw that it couldn't do what I wanted. IDs and IDREFs are in
> any case attributes and any *connections* between them are outside the
> grammar of the language.

This depends what you mean by "grammar", and what you mean by "the
language". TEI these days doesn't rely much (at all? I'd have to think
about it a bit) on IDREFs. We tend to use URIs to point at stuff. xml:id
attributes are IDs, certainly.

> Nowhere in your extremely long postings do you provide any explanation
> as to how elements, attributes and hierarchies arose. They did not
> arrive magically one day on a cloud. They arose from print, and just
> because the GML->SGML->XML textual model has been widely used for
> digital text doesn't mean it was designed originally for that purpose.

You keep hammering on this point, but I don't think it matters.
Technologies are repurposable. Java was originally designed for embedded
systems, but it found its home in server-side application development. The
Java Virtual Machine was designed for Java (obviously), but plays host to
many languages (including Clojure, JavaScript, PHP, Python, Ruby, and
Scala). Really, so what if XML's distant origins were focused on print?
That wouldn't mean it can't be used elsewhere. Type and fonts were designed
for print. Ought we to stop using them as well? It would put an end to this

> When many digital humanists complain about problems of overlap or
> interoperability in their texts you have no answer other than to turn
> those texts into a kind of digital spaghetti of interlinked elements,
> whose significance and function depend on the encoder and his or her
> mood on a particular day. We need better.

[ - - - ]

> -----------------------
> Katherine,
> when you said:
> > I'm less interested in the who's in and who's discussion and more
> > interested in how to expand the digital scholarly edition beyond the
> > limitations of the codex without having to spend $1million+ to get it
> done.
> I couldn't agree more. But the problem with existing standards is that
> $1million cost for a digital scholarly edition is no exaggeration.
> Instead I want to make them easy for anyone to create and maintain
> without specialised technical training.

Perhaps here we get to a point that's fundamental and interesting for DH in
general: what do we want to be when we grow up? Desmond envisions a system
that is easy to use and that requires no expertise. My own feeling is that
expertise is valuable and should be nurtured. I think things that are hard
to do are worth doing, and that they get considerably easier as we develop
expertise (so your marginal costs go down, even if the initial investment
is high). And I think that DH (perhaps simply because it is embedded in
academia, and because of its project- and grant-orientation) tends somewhat
to devalue skill development (with exceptions to be sure) and (again with
local exceptions) does a poor job with career development off the faculty
track. What do we want the future to be? Will it look more like a factory
or more like a cooperative?

All of this requires planning and investment at an institutional and
trans-institutional level, something I believe the NHPRC program Katherine
referenced hopes to foster.

> I'm not sure though that the MLA are in a position to assess the
> technical or practical issues surrounding the creation of digital
> scholarly editions (DSEs). The theoretical stuff sounds fine (I only
> had time to read "Scholarly Edition in the Digital Age"). I'm all for
> the ideals expressed there, but the interoperability of XML is just
> assumed. It isn't true for DSEs.
> I'm not really sure who else would be (well, for the disciplinary areas
they represent, anyway). I thought that document was quite clear that
interoperability doesn't come for free.


        Date: 2019-02-25 10:17:59+00:00
        From: savoye@eapoe.org
        Subject: Re: [Humanist] 32.487: editions, in print or in bytes

I wonder if we are having a slight problem in resolving a conflict of
purpose. Printed texts do not necessarily require one text (although
publishers might), but most readers will demand exactly that. One of the
complaints I get most regularly for my own website is that I provide
multiple texts. (I reproduce each historical text separately, as well as
a number of noteworthy reprints and most of the texts as presented in
significant scholarly print editions.) The question is usually posed in
some form of "which is the right text," which reflects a basic
misunderstanding about the whole concept of revision history, a
misunderstanding that we, as scholars, will probably never be able to
win in the court of public opinion.

I present comparative texts (mostly for my own purposes, but also useful
to some researchers) separately since any apparatus for variants tends
to be clunky and to risk overwhelming the text itself. One problem I
have noted in using some digital presentations that follow the idea of a
fluid text is that the interface renders the text almost unreadable, and
reading should probably still be of primary concern to an editor.

Jeffrey A. Savoye
The Edgar Allan Poe Society of Baltimore

Unsubscribe at: http://dhhumanist.org/Restricted
List posts to: humanist@dhhumanist.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.