Internet Grammars

[P J KAHREL]

The `Internet grammars' sounds like an excellent idea. But the
comments that some people (Bernard Comrie, Elena Maslova,
Frans Plank, Leon Stassen, John Verhaar) have made about it
make it sound both attractive and difficult to achieve. I
think, however, that Internet Grammars are feasible, both in
regard to quality (content) and to technical or practical
implementation (form). And looking a bit further, Lena's
(implicit) suggestion of ALT acting as a publisher is, I
think, feasible and perhaps even desirable. I will deal with
these three issues separately and make a couple of concrete
proposals. At the end of this letter I will give directions--
to those who are interested--to see how things could be
done: the inevitable demo. All comments are welcome.


1. Quality.

The ideas, proposals and reservations expressed so far can be
reconciled if we allow for two separate series. One series
would deal with material out of copyright (as suggested by
Verhaar) and with things that will, for all sorts of reasons
that do not matter here, never be `officially' published
anyway (Comrie's point 4 and Plank's original idea). A bit
like working papers, where people can publish things that they
think (rightly or wrongly) will not be published in refereed
journals--with all due respect to working papers! This series
can be a varied lot: field notes, grammars, simple grammatical
sketches, dictionaries--basically anything. Whether this
series should or should not be refereed I don't know. If not,
then, depending on the reputation of the writer(s), there
should be a huge caveat emptor stamped on every issue (as
Comrie said). This series can be organized relatively easily,
mainly because there are no copyright issues. It would be a
matter of just doing it. As to the technical aspects, see
below.

The second series would be the kind of thing Leon Stassen was
talking about: this would be the high-quality grammars that
Mouton and Longman, for example, are publishing. ALT could do
things on the same quality-level. As Lena said, `If not ALT
then who?'. The thing Bernard brought up was that well-
established scholars can afford to publish electronically, but
that not-so-well-established scholars are still dependent upon
books published by Prestiguous and Well-Known Publishers like
OUP, CUP, Mouton, and Benjamins. But, again, Lena said ALT
comprises so much expertise that from its members it must be
possible to form a most respectable editorial board, one that
will impress every academic body. It may be the case that this
form of publishing needs a bit more time than the one outlined
above. But as Bernard said, (1) publishers may soon be
reluctant to publish grammars because of the cost; and
(2) many parts of the world are serviced by the Internet but
not by libraries. So electronic publishing may become a
necessity sooner than we care to think. ALT as a publisher,
then?



2. ALT as a Publisher?

There is no reason at all why ALT could not be a publisher.
The difficulty with publishing used to be the copy-editing,
typesetting, printing, storage, and distribution of the book.
ALT as a publisher would not be faced by any of these problems
in a major way. Printing is not an issue because, as far as
ALT is concerned, the book will live an exclusively electronic
life. Storage, then (together with the typesetting the most
expensive bit of a book), is not an issue either: disks, both
hard disks and CDs, are very cheap. Distribution is not an
issue either: a book is advertised via the internet (Lingtyp
list, Linguist list, Functional list and who knows what other
lists). Perhaps the copy-editing and the typesetting might be
a problem, but within ALT there is enough expertise to deal
with that aspect as well.

This means then, that ALT, as a publisher, could produce
(electronic) books at a fraction of the cost that
`traditional' publishers face.


3. The technical/practical part

How to publish things? Lena mentioned putting things on the
internet as Word (or whatever) documents. That would not be a
good idea: not everybody uses Word; in the spirit of the Web
we should be platform independent (there is no cross-platform
word processor); if you supply word-processor files you also
have to supply the fonts associated with that file. This font
business is especially relevant for things that ALT is dealing
with, as things published by ALT will always contain IPA
characters. And IPA characters have not been standardized by
any application, and Unicode is still beyond the horizon...
No, word processor files are not a good idea.

The way to overcome all technical problems is to supply files
in PDF-format. PDF (an invention of Adobe's) stands for
Portable Document Format and has a number of things going for
it:
(a)  it is platform-independent, so Mac, PC, Unix, and other
     users can handle it;
(b)  a PDF file has all fonts embedded in it, so it is not
     dependent on the user's environment;
(c)  a PDF file can be indexed with hypertext links, giving
     new meaning to things like Table of Contents and Index;
(d)  a PDF file can be printed on any printer.
With a PDF-file, then, you have the benefit of two versions:
an electronic version that is easy to use because of the links
embedded in it, and a paper version that you can browse at
leisure.


4. The demo.

Seeing is believing. Therefore, and since PDF is a relatively
new technology, I have put on the Web a book as a PDF file so
you can see what it is all about. To view the book, you will
need a program called `Acrobat Reader'. This program is
available for all operating systems/environments and can be
downladed free of charge from any of Adobe's web sites. Point
your Webby thing at http://www.adobe.co.uk/ (Europe) or
http://www.adobe.com/ (rest of the world).

If you have Acrobat Reader (or Acrobat Exchange) installed,
start your web browser and go to ALT's home page. I have put
an invisible link on the home page to the example book: go to
the very end of the home page and hover over the space between
`Thanks' and `are' in the sentence `Thanks are due to . . .'
You will see, on the status line of your browser, a message
that ends in `book.pdf'. Click there and your Reader will be
started, and the book will be loaded.

The book (my Phd thesis, by the way, and you are welcome to
it) has been partially hyperlinked. To display the `hot'
contents, press Ctrl+7. The contents will appear on the left
of the screen. The small triangles pointing to the right
indicate that there are sections--click a triangle display the
sections and maybe subsections. To go to a certain chapter or
section, just click that chapter or section. Two things have
been (partially) hyperlinked, the Tables, Figures and Maps and
first four entries in the Language Index. To use any of these
links, just hover your mouse cursor over anything displayed in
red. If the cursor changes from the spread-hand cursor into
the pointing finger, you are on a link. Click it to follow the
link to the destination.

Apart from the links provided in the file, you can search it
(Ctrl+F or \Tools\Find). You can also copy parts of the file
to the clipboard and insert it into other applications. And,
as mentioned, you can print it on any printer. You can
download the file to your local disk for faster access: press
the Shift key before clicking on the `invisible' link (this is
in Netscape Navigator--I don't know how that works in other
browsers).