Dear Metafolk,
I have had a correspondence with Frans Van Assche that I think has
historical interest for us, and he has kindly permitted me to forward this
note to the list. It particularly tickled me to see that from the
beginning, the notion of meta-meta was recognized as a threat to clarity.
Many thanks to Frans and his colleagues for reminding us that our history
goes back further than a few weeks.
stu
---------------------------------------------------------------------------
Dear Stu,
My old friends (we are all older than fifty now:-) have come up with some
information about the term meta-data:
1. Dr. T.W. Olle thinks that the term metadata was first used by Peter
King who wrote a seminal paper on data dictionary work sometime in the late
sixties. However the oldest reference we could find (credit to prof. dr.
Hannu Kangassalo of the University of Tampere) is the following:
Sundgren,B. An Infological Approach to Data Bases. URVAL No.7, Statistiska
Centralbyrån, Stockholm, 1973. Ph.D. dissertation. page 104 - 105:
"3.2.5 Metainformation
As was pointed out in section 2.1, the infological apprach to data bases
stresses the distinction between:
(a) the real-world phenomena that we are interested in, the object
system
(b) information about the the object system
(c) data reperesnting information about the object system
.....
.... Accordingly, the data base should contain quality information and
other information about the information contents of the data base. We shall
refer to such information as "information on information". Information on
information is made up by messages, the object components of which are
themselves messages.
Information on information is one important category of meta-information.
The latter concept also covers
- information about the basic constituents of the particular
infological model underlaying the data base; eample: formal and informal
definitions of attributes and object types
- information about the data representation of the information
contents of the data base; example file descriptions.
Remark. Data representation of meta-indormation will be called meta-data.
Sub-systems of data bases containing mata-information and data
representation thereof may be called meta-data-bases.
Remark. If we exclude from the object system the phenomena informed about
meta-information and represented by meta data, we are left with a
sub-system of the ofject system which we shall now and then refer to as
"the object system proper"."
2. The oldest implementation of a proper metadatabase we could trace
(credit Prof. Dr. Meersman and myself)
is in the EDMS database management System developed by the Data
Management Research Lab
of Control Data Corporation. Ref: User Manual van EDMS v.5,
Control Data Corporation, Brussels, 1978.
Note: earlier versions of this manual may exist.
I remember the term metadata and metadatabase was used in our lab from the
mid 70ties.
3. Dr. T.W. Olle concurs that the term was in use in the seventies. He
adds "However, we (the UK delegates) discouraged its us because it was
confusing once you start talking about meta-meta data and meta-meta-meta
data and it is not to be found in the standards on Information Resource
Dictionary Systems."
4. My personnel recollection is that already in the seventies it was a
standard term in the database research community and especially in the
Northern part of continental Europe: Scandinavia, Benelux, Germany, which
was at that time leading in information modelling methods.
Best regards,
Frans Van Assche [[log in to unmask]]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|