In her response to Kerridwen Harvey's request for experiences of using
SHIC, Pat Reynolds <[log in to unmask]> lists some bad points which
I feel deserve some comment.  Whilst I speak as a SHIC user, I must
declare an interest and admit that I am also one of the originators of
the system.  However, I hope my response is reasonably unbiased.

>Bad points:
>* I can't remember numbers (compare 4.323 to history USE personal
>smoking - the old Buckinghamshire Classification system)  And I know a
>box of matches isn't 4.323 - I can't remember what 4.323 is - something
>to do with the chemical industry???.

You don't have to use the numbers.  They are simply a shorthand
notation.  Each digit represents a subject heading in the hierarchy.  In
words, the true category name is:
        PERSONAL LIFE - FOOD, DRINK AND TOBACCO - TOBACCO
This conveys exactly the same information as 3.63 (the corresponding
SHIC notation) and could be used in place of the numerical code.
However, most people find that the hierarchical string of heading names
which are required to specify a particular category is too clumsy (and
often no more memorable than the shorthand notation).

Compared with this, the Buckinghamshire Classification's:
        PERSONAL - SMOKING
is refreshingly simple, but the hierarchy is not as extensive as that in
SHIC so the comparison is not of like with like.  If you don't need a
classification with an extended hierarchy then it is probably better to
use a simpler system that avoids the need for notation.

In an attempt to get round the need to specify the full *pathname* in
order to uniquely define a category, Martin Norgate (Documentation
Officer for the Hampshire Museums Service here in the UK) has devised a
simplified list of SHIC category names.  However, this does not indicate
the path.  So it is impossible to reconstruct from the name the
relationships implied in a true SHIC name or number.  Martin's unique
SHIC headings are part of his excellent _Object Format Rules_.  I have
not checked, but they are probably available in the version of OFR
available on the Hampshire Museums Web Site at:
        http://www.hants.gov.uk/museums/ofr/index.html

So Pat is right, the numbers are a pain but for most of us they are the
best way of summarising the lengthy category names.  I look forward to
the day when we are all using computer systems which can exploit the
SHIC paths transparently and let us get on with using the concepts
without worrying about arcane codes.

>* It's not just me. We use ModesPlus, and Intouch, which allows the
>public to select by classification - history AGRICULTURE harvesting
>wheat means something, 1.354.45 doesn't, to your average museum visitor.

I certainly wouldn't recommend using SHIC numbers overtly in any public
access system (although it is worth remembering that public libraries
have been operating for many years with public access sytems based on
decimal notations).  The trick is to use SHIC behind the scenes to
classify the data in a highly structured way but give the public an easy
way into it which exploits the structure without making specialised
demands on the user.

Some interesting work is being carried out in this area by Doug Tudhope
and his team at the University of Glamorgan.  They have been
experimenting with using SHIC in a hypermedia museum exhibit to support
intelligent machine-assisted navigation by topic using an on-line
photographic display.  They are using Pontypridd Museum as a test bed.
Initially the interface between the public and SHIC was rather crude but
they are working on this.  I am not sure if details are published on the
Web but I will find out.

>* ModesPlus files history AGRICULTIRE harvesting wheat right after
>history AGRICULTURE harvesting barley, whereas it files 1.354.45 right
>after 1.232, and files 1.34 before that.  So it's a bit difficult to
>'grab' a 'logical' part of the database (ModesPlus problem, not a SHIC
>problem).

SHIC notation must be treated as a character string to achieve sensible
sorting.  If your database software can do this you will not have a
problem.  MODESPlus is probably unique in this respect.  At present its
only field type is Character but it will detect anything which looks
like an accession number and sort it using a cunning algorithm which is
brilliant if the string really is an accession number but disastrous if
it is not!  A case of being too clever by half!!  This may change with
the forthcoming Windows version of MODES.

>* A mis-type of history AGROCULTIRE harvesting is a hell of a lot easier
>to spot than a mis-type of 3.444 (should have been 4.333!)

I can't argue with that!  Ideally, your database software should also
display the code in text form making errors easier to spot (MODESPlus
can do this, and by using the old Buckinghamshire classification in
parallel Pat is achieving much the same result).

>* it's process-driven.  Take that box of matches - it would have a
>different number (or more than one number) depending on whether it was
>used in a display stand by a shop, or in a trade exhibition, by Mr Jones
>to light his pipe, or by Mrs Jones to light the aga.

As Leonard Will has already remarked, this is the whole point of a
subject classification and hardly a bad feature.  As Leonard makes
clear, SHIC should be used alongside a separate system which groups
material by physical form and function.  Earlier classification systems
(from which SHIC has developed) tried to combine these approaches.
Whilst it might be more convenient to use just one system for both
functions, we found that better results could be achieved by
concentrating on activities rather than object types.

>* the subjects covered reflect the interests of the museums involved
>(all subjects are covered, but whereas, for example, there is detailed
>sub-numbering for Railways, there isn't for the Lock manufacturing or
>Lace industry, and while the campaigns of British Army is well covered,
>there's no specific numbering for campaigns in which they weren't
>involved)

This is a bit misleading.  We tried to cover _all_ aspects of life in
the main part of the classification.  We then devised a model
subdivision which could be broadly applied to _any_ particular activity.
In order to demonstrate how this might be applied to specific activities
we gave some examples.  The idea was that users would then be able to
apply the general model to any activity for which they had sufficient
material to warrant more detailed classification.  It would clearly be
impossible to provide detailed subdivisions for every conceivable
activity from the outset.  We are always interested in receiving
additional activity subdivisions from users with specialist collections
and expertise.  We hope in due course to publish these in some form
(possibly on the Web).

The bias toward the British Army reflects the fact that SHIC was devised
in the UK.  As international interest grows we may have to rethink this.
Currently there are other examples of bias within SHIC, in religion for
example.  As British society becomes ever more multi-cultural SHIC in
its present form does not even reflect the contemporary scene in the UK.
We will try to put this right in future.

>* a personal niggle - the costume sub-divisions go:
>costume : sex of wearer : broad > narrow
>As I tend to do textile exhibitions which focus on the broad or narrow,
>rather than the sex, I find this inconvenient!

The costume sub-divisions are based on the _ICOM Costume Vocabulary_.
Since they are largely based on form and function rather than activity,
they really have no place within SHIC and with hindsight it would
probably have been better to omit them.  Users could then choose to use
alongside SHIC whichever object based system(s) best suited their needs.

My apologies for responding at such length.  As Leonard has said, more
information and a sample of the classification are available at:
        http://www.holm.demon.co.uk/shic.htm
This includes some interesting comments from Pat Reynolds regarding SHIC
and 20th Century Collecting.

Regards

Stuart
------
Stuart Holm, Heritage Documentation Projects     Tel: +44 1603 870772
2 New Road, Reepham, Norwich NR10 4LP, UK     E-mail: [log in to unmask]
-------------   World Wide Web - http://www.holm.demon.co.uk   -------------