THE FUTURE OF THE TEI: SOME IDEAS
 
SMH, 2 June 1991
 
I believe we should address the future (i.e. what happens after June
1992) in the following way:
 
(1) Consider what needs to be done
 
and only then
 
(2) Determine a mechanism for doing it
 
 
1. What needs to be done
 
1.1 What will we have by June 1992
 
We will have more or less what we said we would produce, i.e. some
guidelines for encoding machine-readable texts. I expect that we will have
enough coverage in the guidelines for people to use them, but they will be a
bit lacking in detail in some application areas and some users will
think of topics which we haven't thought of.
 
The first document we will produce is a technical specification. I believe
we must have that by June 1992. We need to plan in detail at the 16-17 June
meeting what other publications will be useful. I suggest
 
(1) a tutorial guide, with some reference material.
(2) a cookbook of examples.
(3) a formal specification for software developers
 
It seems to me that (1) and (2) will be the most widely used, followed by
the technical specification and then (3).
 
 
1.2  What we committed ourselves to doing in the future
 
In the memorandum of understanding the ACH, ACL and ALLC agreed to
 
(clause 10)  'establish and maintain some mechanism (e.g. a standing
committee or bureau) for the continuing development of the guidelines and
interchange format in cooperation with the other sponsoring organisations
after completion of the first version. These mechanisms shall be
responsible for accepting comments and suggestions from the membership
of the sponsoring organizations, monitoring the success of the
recommendations, and considering revisions and extensions in the light
of experience.'
 
Although this was written before we started, it does reflect what I think
our main objectives after June 1992 should be, i.e.
 
To establish a mechanism for
 
(1) continuing maintenance and development of the guidelines in the light of
user reactions
 
(2) to monitor the success of the recommendations within the guidelines
 
 
1.3 What we now see is needed as well
 
'Success' in (2) above can be intepreted to mean
 
(1) that the guidelines work in terms of being able to handle the
complexities of the texts
 
and
 
(2) that they become widely used
 
We are already dealing with (1) with the affiliated projects, but inevitably
more technical queries will arise as more people use the guidelines.
(2) follows on (1) and I think is politically very important for us as we
have had about $1,000,000 of mostly public funds. It will be the way in
which the members of the sponsoring organizations and the user community
judge the success or failure of the project. It will be a consideration for
funding agencies, if we decide to ask for more money.
 
To achieve (2) I think we need
 
(a) software to generate TEI SGML which is
 
(i)   easy to use
(ii)  relatively cheap
(iii) runs on PCs, Macs and UNIX (as a minimum)
(iv)  does not require a substantial change in the method of working of the
      user who is preparing a text or document.
 
It does not seem to me that the existing SGML software meets criterion (iv)
and, other than the deals which the TEI has obtained for development,
criterion (ii).
 
Steve Siebert's comments at the Tempe conference were enlightening in terms
of the investment needed to put TEI in Nota Bene. The calculations may
possibly be different for other developers of wordprocessing software.
To the best of my knowledge Steve Siebert is the only one who has shown
interest so far. Potential users are being asked to change their encoding
scheme. They may feel happier doing it if they can use their existing word
processing program.
 
(b) software to analyse text marked up in TEI SGML.
 
Some software for SGML-encoded text is now appearing. Cost seems to be the
major problem here. My indications are that it is the cost of developing the
SGML parser which is the major factor.
 
(c) training and education of users of the TEI guidelines.
 
We have already started doing this.
 
(d) acceptance by more funding agencies
 
Another way of ensuring user acceptance is to reach the situation where
major funding agencies will only consider the funding of a text database if
coded in TEI SGML. It is not reasonable to expect this to happen until the
software tools are in place.
 
(e) conversion of existing archives
 
This was not one of our original objectives, but to make text available in
TEI format obviously helps to promote the TEI. I do not think it is
reasonable to expect existing archives to fund this out of their own
budgets, at least at this time. They may do it some time in the future if
they see that this is the only way to proceed. Many of them also have
software which is built on their own encoding scheme.
 
 
2. What do we need to do this?
 
What we need in outline seems obvious:
 
(1) an organizational structure
(2) money
 
The detail is less obvious.
 
2.1 Organizational structure
 
I think the following considerations apply for the organizational structure
(they are not in any particular order):
 
(1) In 1988-9 we set up what now seems to be a very complicated project
structure for the development of the guidelines. Involving a lot of people
on committees, work groups etc was probably necessary to get user
acceptance, but it generated a lot of managerial work in terms of keeping
track of them, dealing with payments from different sources etc.
 
It seems a good time now to review the organizational structure.
 
(2) We need something which retains the international perspective,
both North American and European, also probably with the Japanese.
 
(3) We need some kind of central organization, otherwise the problems in (1)
above will get worse.
 
(4) The original sponsoring organizations ought to be represented - since
they committed themselves to continuing the project.
 
I would like us to determine whether the following three points should also
be considered:
 
(1) The need to have some kind of Advisory Board or watchdog to keep an eye
on the TEI. This would have representatives from the user community.
It may or may not be the same as the existing Advisory Board. It would be
there to ensure that we do continue to serve the user community. It would of
course be a charge to the budget unless it can meet at a major conference or
be combined with some other Advisory Group.
 
(2) Do we have any intellectual property rights in the TEI? If so we need to
ensure that the TEI exploits them as best it can, rather than somebody else.
 
(3) Various people have suggested that we should approach ISO to get the TEI
accepted as a standard. Will this benefit us and how? When is the best time
for this? How do we do it? This may well incur extra costs as I assume
somebody well have to prepare a document in the right format for ISO?
 
 
2.2 Money
 
After summer 1992 we have no more income as yet. It seems that we will need
two kinds of money:
 
(1) general running costs, for maintaining the TEI in terms of secretariat,
some travel, answering queries etc
 
(2) for specific activities, e.g. software development, workshops etc.
 
There are possible sources of money for (2). NSF?, NEH? Computer Board? (I
think the Paul Ellison project will possibly lead to some software
development in the UK, but not for another 2 years). Workshops need support,
even though the participants may be expected to pay something to come.
 
Without the travel, (1) would seem to need a minimum of one-half time person
plus secretarial help, say a total of about $25,000 per year plus any
overheads. More will be needed initially to complete the other publications
outlined above. Meetings could be arranged to coincide with major
conferences but it should be noted that some of us have no other travel
funds for attending such meetings.
 
Sources for (1) which I can think of are:
 
(a) finding some institution which will support this anyway (unlikely
unless something is already in the budget for it). Putting it somewhere
together with another related activity which already has some support is
cost effective and seems sensible. Also if payment is made to the
institution in return it for providing a 'TEI service', it does oblige the
institution to ensure that the work is done. This is better than relying
only on somebody's goodwill.
 
(b) royalties from any publication, but these will be small in comparison to
what we need, for example to sell 1000 copies of a book which costs $30
would give us only $3000 at the normal 10% rate. And we would not get much,
unless we can negotiate some kind of advance, for probably a year after the
book comes out.
 
(c) possibly royalties from software, once it is developed. This is more
profitable than book publication, but it would need investment to start,
or if it was funded by a grant, would need to be done with the cooperation
of the grant giving body. I do not know what the NEH or NSF's attitude to
this is.
 
3. Legal organization
 
The options are:
 
(1) Continue as before, in effect a 'joint venture' of the three
organizations (2 US-based and 1 UK-based but all international).
 
Advantages:    (a) simple and cheap to set up
               (b) not identified with any one country
               (c) flexibility in fund raising
               (d) funds belong to the sponsoring organizations jointly
                   and can take adavantage of their tax status
 
Disadvantages: (a) complicated to administer, but a project manager might
                   help
               (b) ownership of IPR would have to be clarified
 
(2) Set up a new legal entity
 
Advantages:    (a) better/more permanent image
               (b) could hold IPR
               (c) limited liability - is this important for us?
 
Disadvantages: (a) cost to set up
               (b) would have to set up in one country
                     EEC attitude to US entity?
                     NEH attitude to non-US entity?
 
The options for a new legal entity are
 
(a) in UK
 
(i) Charitable Trust. This is the normal form if limited liability is not
needed. The ALLC has this status. It needs to be approved by the Charity
Commissioners. Most income is then tax exempt. There is no limited
liability. The cost to set up would be between 500 and 600 pounds.
 
(ii) Company limited by guarantee. This is the normal form if limited
liability is needed. The cost to set up is about the same as (i), but the
running costs are much more expensive. Accounts have to be prepared in a
statutory form, audited and filed.
 
(b) in USA
 
not-for-profit?
 
[can somebody supply the relevant information?]
 
(c) in another country
 
This doesn't seem sensible for us and would probably be more expensive.
 
(3) set up a pair of entities, one in the USA and one in the UK, with links
between them.
 
Advantage:      more flexibility in funding (even if one was only a 'feeder'
                of funds to the other
 
Disadvantage:   double set of
 costs