Glad to see Sabourin's work and other responses to this question.   
Sabourin's 2002 analysis incorporates several advances that show the  
mid-80s influence of British Columbia and the 90s influence of  
Australia.  It is essentially reflects my own philosophy of  
classification scheme design, though we go a bit farther in examining  
potential for numbering conventions that are not articulated in the  
paper so as to provide options for fitting a scheme to operational  
reality--and underpinning both risk & compliance management in the  
context of corporate knowledge resource management.

Lee's question is big and touches on fundamentals that are fun  
(really!) to explore and debate toward coming to a position in a  
project.  I think it is a necessary process within projects to enable  
stakeholders to move along in a common understanding.  On the list,  
this is the stuff of going deeper in ways that ground practice. So,  
diving in with advance warning--it's a long one.

When working with a team of top notch thinkers in the mid-80s on how  
to conceive, design and implement an enterprise-wide RIM program, we  
had many, really great debates over things like the implications of  
setting retention at this level or that, the advantage of a separate  
or integrated retention scheduling practice, the means of  
representing the system for enterprise-wide control versus a simpler  
representation for front-line users.  Later, working with several  
major programs, we had to consider numbering schemes that could  
retain integrity despite unpredictable  mergers, dissolutions, etc.   
This is serious RM.  In my view, it also grounds taking RIM into the  
executive suite and offers opportunity to transform the vision of RIM  
from administrative nuisance and compliance headache to management of  
a critical resource that grounds corporate thinking, decision making,  
etc.

Ginny's example suggests that a breakdown based on organisational  
structure can work.  Many organisations have such a structure and in  
my experience that works best where the organisation is itself  
structured along functional lines--even better if the divisional  
names reinforce their function.  It would be interesting to know if   
the Public Utilities' divisional structure reflects an essential  
functional split, or more about how the necessary relationship  
between record value and purpose is addressed in scheduling.

Larry's post on the importance of language and meaning, focuses on a  
critical issue.  Terms like "big bucket" and even "schedule" are  
interpreted very differently by various practitioners in various  
contexts. "In active" does not actually mean inactive in most  
jurisdictions that use the term, semi-active may better describe  
activity and has been adopted by some jurisdictions and there,  
inactive means "no activity" and suitable for a decision final  
disposition.   Certain "subject-based" classification schemes are  
not.  As Lee suggests, some are hybrids (as in the system for the  
British Columbia government I was privileged to be a part of  
designing many years back).  That scheme broke ground in integrating  
the retention schedule, and the "buckets" differed for classification  
versus semi-active retention/disposal/archival retention.  For  
example, 7 retention bands for reasons not related to information  
value, but aggregating defined periods for efficiency in operation of  
an outsourced records centre service; classification structure  
limited to two (visible) levels with positive effect in getting buy  
(looks so simple!) but less positive effects in requiring a great  
number of primary headings to achieve the necessary segregation of  
series (lesson: simple isn't necessarily so;  a good RIM program  
should be able to handle complexity without sacrificing  
practicality).  Even the term "hybrid" is used in various ways.   
Typically, it means a media-mix, not a mix of structural elements in  
scheme design. Nonetheless, mixing structural elements can work and  
is not new.

As Pilar notes, there is more than one way and consistency is key.  A  
critical factor is to develop a very clear view of what the program  
is to achieve, and how, and whether the program is limited to  
compliance and retention or to knowledge building (which requires  
additional metadata). The degree to which systems evolve and are  
effective has a lot to do with surrounding processes and cultures.   
The use of technologies should enable--not drive the solution--and  
the selection of tools that are compliant with various standards  
needs vetting against management's understanding of what that means.   
(Application of system functionality results from configuration and  
customization--entirely a result of the RIM program's decisions  
points--or that functionality will be wasted or will be activated  
without understanding of the implications, limiting value and posible  
causing problems).

How the system is presented to various user groups is another thing  
altogether.  Complexity in IM is inevitable. Complexity does NOT have  
to be complicated. It does have to be engaged--or something gets  
missed.  User groups all have different needs and the communication  
strategy should articulate who needs to know what and why so that the  
system and its various aspects can be presented when needed to  
support users' needs.  In my (strongly held) view, systems that do  
not work for people, do not work.  Systems designed for one user  
group at the expense of another will result in the latter group  
creating independent solutions - entirely predictable causes of  
ultimate failure.  If one cannot afford to engage all relevant  
stakeholders effectively, it is better (IMO) to carve out a scope and  
be very clear that the result is limited.

The questions relate to the challenge of arranging metadata for  
defined purposes.  Sometimes merely suggesting that metadata be  
captured is enough to scare users--never mind that the actual meta  
data required is not yet identified so the assumed impact on workload  
is entirely speculative.  In some regions, existing processes are  
more burdensome that capturing new metadata (an advantage of not  
being at the leading edge).  This can vary within one organization.   
Much metadata exists without user effort as a byproduct of work--but  
the metadata and its value for defined purpose has to be mapped to  
the desired functionality (whether manual or electronic). The  
complexity necessary to capture relationship across several distinct  
for interdependent workflows can be presented in a simple checklist-- 
some users like that. A spiderweb sort of map of the  
interrelationships is not needed to support the work of the CEO, or  
the line worker. (But it can have real value in showing a CEO that  
there is more to RIM than s/he thought.)

A mapping exercise--the process--has great importance for the  
individual who has to develop--and defend--decisions on what  
information exists, when and in what media at what cost.  It also can  
inform processes that reveal worker experience being developed  
through work-streams with potential for capacity building and  
matching competencies to emerging needs.  Such a map is also  
valuable, I would argue, for archival decision making because today,  
limiting appraisal to post-event, retroactive analysis of accessions  
does a disservice to the organization (ok, minor rant not to be  
developed here ;-)  It is the process that provides broad benefit.   
As in gaining the CRM, it is the journey that is transformational.   
RIM is all about transformation.

In this way, RIM becomes becomes relevant in ways that surprise.   
Such considerations were instrumental being able to take a 12 person  
box storage buried under ground level (in space that was flooded,  
workers gassed by pipes feeding the BBQ for execs upstairs and which  
later became criminal  holding cells when I moved my team out to  
space built-to-suit)  and transform that into a 30+ program  
recognized as direct support to Executive--as well as serving almost  
500 offices with the most comprehensive RIM program in a geographical  
area the size of France and Germany combined (ok, make that  
California, Oregon and Washington State ;-)  This program proactively  
analysed cost implications (savings of 4.2 million in one budget  
alone and garnering a 300% improvement of funding for work across the  
entire government necessitated by new FOI legislation).  An RIM  
program can be (should be?) a vital contributor to strategy and  
outcome achievement--not limited to repository management. We're  
always happy to support the latter goal, but are just plain thrilled  
when the full potential of RIM can be enabled. It is so rare.

So, knowing what you want the program to do is the big piece.   
Clarity in program leadership and in articulating where the program  
is going is critical.  All else flows from that.  It is worth  
spending whatever time is necessary to decide what role RIM will play  
in the organisation.

Sometimes, pragmatic decisions compromise the idea.  It is helpful to  
remember that nothing is perfect...look at the "patches" that are  
essentially workarounds for system flaws. The key is to have enough  
knowledge about the context and variables to make good decisions--how  
and why any compromises should be made, and a strategy to overcome  
the effect of pragmatic decisions.

Re duplication, again language and meaning are critical. For example,  
a multi-part form is a legitimate record and why each part may a  
legitimately different retention period based on the function/work  
flow it serves (in both hard copy and electronic media).  Is that  
duplication?  At the design level, is it duplication to have a slot  
for contracts in all relevant business functions/flows? Is it  
duplication to have a central collection of records to support a  
financial function and dispersed accounting records in remote  
locations?  There are options around all such questions.   
Identification of the "office of record" versus offices subject to  
process audit is only one approach to transforming what some see as a  
duplication into a series that reflects a well defined, easily  
comprehensible purpose.  Purpose is key. If you have duplicate  
purposes and work flows replicating what is done elsewhere--then you  
have an opportunity to deliver a corporate benefit in revealing that  
and reducing duplication.  If you have similar or identical content  
used in multiple workflows for different purposes, then you do not  
have duplication, you have a well defined picture of how your  
information resources are used (and are less likely to screw things  
up should someone come along ans say, "that record's over in office  
X, lets eliminate the duplication").  Out of context conclusions are  
dangerous and the stuff of many errors made with good intent.

I'm going on too long, and I hope the response is helpful.

Lee, you are not over-thinking.  it is this kind of thinking and  
exploring that needs to be built into your program, IMHO, because  
many of the conclusions made  require an understanding beyond the  
obvious among various stakeholder groups.  Important to do your own  
stuff, but not to forget that the process that results in your own  
clarity may be a journey that others need to go on.  Learned that one  
the hard way!  Some very capable people get left behind (maybe even  
become opponents) if the process of change isn't managed to enable  
the result that you intend.

Look forward to others' thoughts and views.

John James O'Brien, BA, CRM, MALT
[log in to unmask]

Partner & Managing Director
IRM Strategies
Hong Kong: +852 3101 7359
Bangkok: +66 2 207 2530
www.irmstrategies.com

Associate Partner, S4K Research
Stockholm www.s4k.com
______________________________

Note: This email correspondence (including all attachments and  
content conveyed hereby) is intended for the addressee, in person or  
post, only.  If received in error,  use, distribution or action based  
on this email is prohibited. No    rights of ownership are waived or  
lost through misdirection or interception. If you are not the  
intended addressee, kindly notify the sender immediately and delete/ 
expunge all record of this email from all relevant systems. Legal  
action may arise from any misuse of this correspondence.


List archives at http://lists.ufl.edu/archives/recmgmt-l.html
Contact [log in to unmask] for assistance
To unsubscribe from this list, click the below link. If not already present, place UNSUBSCRIBE RECMGMT-L or UNSUB RECMGMT-L in the body of the message.
mailto:[log in to unmask]