GENIRE-L Archives

Archiver > GENIRE > 2003-03 > 1046515319

From: Don Moody <>
Subject: Re: Genealogy - what and where is it?
Date: Sat, 1 Mar 2003 10:41:59 +0000
References: <005401c2df88$fc53eca0$d4a1a5c2@computer>

In message <005401c2df88$fc53eca0$>, Jane Lyons
<> writes
>In the past, when people criticised any person who posted lists of names or
>newspaper articles to any genealogy list, I used to 'react' so to
> anger with the critic, and defence of the poster - it
>didn't have to be me.

You miss the entire point. You can publish lists of any length you want
under any uninformative title you like, and it won't be useful genealogy
or any sort of aid thereto. What you are doing is merely repeating
information which already exists. Somewhere. You are therefore not
adding in any way to the sum total of human knowledge. But you are
consuming resources for this zero net benefit.

The only way of making a non-zero net benefit is to make the information
available in a form which is much easier to search than the same
information was originally. That requires sorting out who asks what
questions of the database, and remodelling the database so that those
questions can be answered. In genealogy, and whether you like it or not,
the prime questions will always be about names. Therefore any compendium
in any form which is not fast-searchable by name, whether on title or
body text, is effectively useless in the wider genealogical community.

'Extracts of Bogston Recorder, issues 1 to infinity' is and always will
be useless to people doing genealogy. In the first place it requires
them to re-read the whole extract to find out if any of their names of
interest are recorded, and they don't have the time. In the second place
if no criteria are stated for the extracting then failure to find a name
in the extract proves nothing, and sets the searcher the task of reading
every issue of the Bogston Recorder cover to cover, for which there is
even less time and which makes the extracting effort pointless.

No search engine means no use. Search engines which direct to a 'page'
with thousands of names, but which do NOT highlight the name searched
for wherever it occurs on the page, are not useful. To use jargon, you
have to understand for any field of information searching what are its
'indexable concepts' and how the searchers operate in 'fast scan' mode.

The game is all about dumping at least cost and as fast as possible the
vast mountain of 'irrelevant' dross. And doing it under the constraint
that some of what is classified as dross by A may very well be
classified as useful by B. In other words you cannot make
value-judgements when pre-structuring the database.

In genealogy we have the specific problem that the usual search term, a
name, does not necessarily have invariant spelling through space and
time. Search engines still need human support to find some of the less
routine transformations of names, and it is a good thing if they can be
made capable of 'learning' from the humans.

So either publish in an efficiently searchable way, or don't waste
effort on publishing at all. THAT is the point, and the decision.

Dr D P Moody, Ashwood, Exeter Cross, Liverton, Newton Abbot, Devon,
England TQ12 6EY
Tel: +44(0) 1626 821725 Fax: +44(0) 1626 824912

This thread: