Hello everybody,

Several people have asked me to summarize the responses that I’ve received 
regarding handling case files with a functional classification system. I must 
explain first the advice that we’ve received from our consultant regarding 
paper case files and the functional classification system (well, it’s still a BCS 
right now, and not even a complete one – we’re finishing up the work). She 
proposes that hanging files are used to keep active case files. Each hanging 
file has a plastic tab with the case identifier. Inside it, different folders contain 
labels for records. Each hanging files has several folders labeled with 
their “correct” functional classification”. When the hanging files are passed on 
to the file room, the folders are pulled out, indexed, and then filed on the 
correct shelf for their function/activity. If retrieval is needed multiple pull-outs 
from shelves are needed, and the users will receive all the folders that make 
up the “case”. I guess if the files are in semi-active storage at the Alberta 
Records Centre, which is a warehouse, they’ll have to pull several boxes down 
to re-constitute a “case”.

I’ve checked with records delivery staff here and they cannot see any gain in 
a system that would reduce their productivity. In an EDRM system I can see 
how metadata can quite easily create virtual case files, but I decided to ask 
the list-serv how they would handle paper case files in a functional system. 
The advice that I’ve received can be summarized into the following points:

•	Good functional classification design is always pragmatic and not 
strict and academic. They must support the business otherwise they will 
become a long-term liability.
•	For a classification scheme to work there has to be a compromise 
between theory and usability.

•	If possible, tweaking the model to keep it functional at the top end 
and less functional at the lower levels could make it more usable and therefore 
accommodate case files. 

•	One solution suggested was to use the term “Case Management” at 
the sub-function level even if this was not a descriptive term because that did 
not matter either for searching or sentencing, and the third level would 
contain the cases files themselves, identified by unique descriptor.

•	Another suggestion was that the functional classification scheme 
has three levels (Function/Activity/Transaction) but a file plan really needs 5 
levels to work: Function, Activity, Transaction, File and Item. The top two 
levels are fixed and the rest are flexible enough to meet the needs of the user 
requirements.


•	I did get some very good guidance on how to manage casefiles in an 
EDRMS using metadata, however, I failed to mention that the problem was 
that in several areas the classification system will be deployed with no EDRMS.

•	One respondent did not believe that functional classification provided 
benefits over subject based classification. Regarding case files, an interesting 
comment:“In May of this year, I met with Luciana Duranti of UBC School of 
Information and Archival studies on the InterPARES project, which is all about 
preservation and access to long-term digital records. She yet again affirmed 
to the group that any one record does not exist alone. Each record must 
stack together with the other records that relate to a case file, and all the 
records together provide the "record" for an event / case / project 
representing an action of the organization. Furthermore, how do you manage 
case files? Answer: you connect the case file to classification in order to 
derive accountability and retention. Luciana's re-affirmation makes me more 
strongly stick to the subject classification system with the insistence that 
every electronic and paper record receive both a case file identifier and a 
classification number to be of value to the organization.” Patricia Daum, BCIT.

•	One respondent said that at his organization a dual system is in 
place.  For those still using paper then we stick to case files.  When the 
business unit comes into the EDRM system we use virtual case files by using 
metadata to link the records that constitute the “case”. The same respondent 
said the kind of approach that our consultant has proposed would not get 
even past the approval stage, because projects need to demonstrate that 
efficiencies can be gained before they can go ahead. (By the way I could not 
agree more)

All in all, I had excellent suggestions from everyone. I have not just cut and 
pasted their responses because I’m not sure yet of list-serv etiquette, so I 
just summarized the points.

What came through over and over in the responses was that flexibility and 
pragmatism must trump over academic theory when designing classification 
systems. Interestingly, I’ve found that common sense is a harder sell than 
theory!

A big thank you again to everybody that helped me with my query.


Paula Mello McClure, MLIS, IT Dip.
R.I.M. Analyst
Service Alberta
10155-102 Street
Edmonton, AB, Canada T5J 4L4