> I am on a task force at the National Museum of American History trying to
> define what the minimum amount of documentation is needed to make our objects
> useful to the telling of our story. What do you in the field consider the
> minimum? What is the "practical ideal."
 
In the design phase you can sell yourself short by working with
'minimum' or 'core' information.  I recommend you take an information
modelling approach and account for every piece of information you can
associate with your enterprise and its 'objects'.  An entity-relationship
model is relatively easy for users to understand and is a good place to
start and can lead easily in to building a working database on a variety
of widely available platforms - but there are other options to describe
what you are interested in.  Once you have defined the information
requirements of your enterprize, sorted out where things fit, removed
duplication and resolved inconsistencies, you can decide on acceptable
levels of information to make a system work.
 
Design for the maximum and build for the optimum within the constraints
of your technology, time, people and money.  If you specify a minimum,
that may be all they will let you build and within a vey short time
(days, hours) there will be users who ask for (no, *demand*) something
your system can not deliver and that can not be easily implemented in
the future.
 
> Does information storage technology
> have an impact on the data collected? How far beyond administrative and
> descriptive data should we go in documenting a historic object? What is the
> best means of making this collected documentation accessible to the museum's
> staff, scholars and the public?
 
Yes and no.  If you are limited to paper-based technology, the design
exercise is still useful but the options for retrieval and manipulation
of data are extremely limited - all the information in the model could
be collected and stored on forms and cards in an extremely rigourous
way.  Any contemporary computer system will be able to handle the
information on, about and concerning your 'objects' and how they relate
to the world at large.
 
Your should go as far as your resources will allow, and even if you do
not collect all the data to start with, you should have clearly
identified places for it to go, either in the working system or in the
design to be implemented at a later stage, so that you will be in a
position to take advantage of the hugely increased resources for museum
activities that are coming 'real soon now' ;-)
 
As far as making the documentation accessible to staff and the public,
I highly recommend planning at the outset to use the internet and
in particular gateways to gopher and the World Wide Web.  By using this
system to deliver information to your staff, you have provided it to the
world at no extra cost.  This is an approach being taken by several
natural history collections (our gopher and WWW servers, those at the
museum of paleontology at Berkeley, the Harvard biodiversity gopher, the
Smithsonian botany department gopher, to name but a few)
 
> What forms might this documentation take?
> Photographs? digital images? Letters? Journal articles? Catalog worksheets?
> Computer files?
 
All of the above.  Whatever you have.  Thes items can all be stored in
databases or computer files systems and delivered to the WWW in an easy
to use and visually exciting interface.  Because the interface is also
used for a variety of other application, user training becomes a minor
component of your task.  Once the database gateways have been
implemented, and the user knows how to drive a stardard WWW or gopher
client, they can get straight into using the system - there is no having
to learn how to drive an Oracle interface, a Sybase interface, an
Ingress interface, a dBase interface, or whatever - and it is all the
same on all the common computing platforms in use today.
 
> Any answers?
>
> -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
>       NMAH,  Dept. of the History of Science & Technology
>           Phone: (202) 357-2314    NMAH 5119, MRC 638
 
Design properly and in detail from the outset, and if you don't have any
professional expertise in house, pay for it, or borrow it from a sister
institution.  A lot of the design work has already been done, a lot of
the mistakes already been made, and it makes little sense to do that
all again (after all, there are heaps of new mistakes to make).  Have a
look at the work done in natural history museums and herbaria - their
whole enterprize is based on historical artifacts (specimens) and I do
think you will find too much there that is incompatible or too far
removed from what you do.
 
A lot of work on information models have been done by a working group of
the Association of Systematics Collections and the Natural History
collections staff of the Smithsonian.  An ASC information model for natural
history collections as a postscript file is available from a number of
FTP sites.  The Smithsonian staff have built (or in the process of doing
so) an elaborate system for the management and transactions of their
collections.  Information model junkie, Stan Blum, who used to work at
the Smithsonian and who I think monitors museum-l, was behind both of
these activities and could provide more details.
 
jim
___________________________________________________________________________
Jim Croft           [Herbarium CBG]               internet: [log in to unmask]
Australian National Botanic Gardens                  voice:  +61-6-2509 490
GPO Box 1777, Canberra, ACT 2601, AUSTRALIA            fax:  +61-6-2509 599
                          URL=http://155.187.10.12:80/people/croft.jim.html
______Biodiversity Directorate, Australian Nature Conservation Agency______