Hi,
Earlier this year there was a thread on the need for corporate memory to
support organisational learning. If I remember correctly, the bottom line
was that we agreed that corporate memory was a good thing to have.
I'd like to investigate how individuals use this corporate memory for
novel, intellectually challenging tasks. My theory is that it is very
difficult to reuse someone elses analysis.
For example, one of the largest problems in the software development
process is documenting the analysis and design knowledge used by
developers so that other people can maintain the software. Again and again
we hear that it is easier to recode the software than try and maintain
someone else's code. It can be argued that this results from a lack of
"standardised" ways of developing code - but in novel situations things
can only be standardised in hindsight (and I wonder whether we lose a lot
of the contextual information that was important in developing the
analysis in the first place in this standardisation process).
My understanding of corporate memory is that it is a database where
documents are stored for later reuse. In the software development process,
I put my code and all the relevant supporting documentation into the
corporate memory, and at some later point in time, the maintainer then
accesses all these document and "knows" what my code did, and how to
maintain it.
I don't think this approach will ever work for any problem that is not
routine (i.e. that is not solved over and over again by the organisation
e.g. the types of problems handled by transaction processing systems).
I think there is a lot of intellectual value in "doing" the analysis
process yourself, not just reading someone else's analysis (hence the ease
with recoding software rather than maintaining someone else's). Secondly,
I don't believe it is possible to document all the contextual knowledge
that you used in the analysis process (how do you know what is and isn't
important, how do you know what your implicit assumptions were, and then
there's the AI frame problem - how do you limit what you need to
document).
This starts raising some doubts about the corporate memory process - it is
not just about document storage, it is about enabling people to "redo" the
analysis so they have an intellectual stake in the analysis - and this
leads to the problems of different people have different mental models and
learning strategies.
I think at this point I need to ask some questions:
1. is the fact that individuals require an intellectual stake in the
analysis process actually a real problem (or am I just making life hard
for myself?)?
2. if so, has any research been conducted in this area as to how people
can "buy-in" to the analysis process over time?
3. corporate memory (as I understand it) has been developed on the basis
of document storage. Has anyone investigated how individuals can access
and reuse this corporate knowledge (especially in non-routine
problem-solving situations)?
I welcome your comments on these ideas :->
-- John O'Neill DSTO C3 Research Centre, Australia email: jao@itd.dsto.gov.au