Systems are episodes in the life of information June 24, 2014

Simpsons 500th Episode Marathon - Professor FrinkOne of the issues we are talking to agencies a lot about at the moment is information longevity, primarily the fact that an awful lot of business information will need to outlive the system it is part of.

This concept however, still does not have significant traction or understanding.

In most business environments, systems or services are still regularly regarded as primary, and information is not understood as the constant that will need to be sustained through system or service change and evolution.

Thinking through this issue brought to mind a quote from one of my archival heroes, Chris Hurley.

In his paper, Problems with Provenance, first published in Archives and Manuscripts in 1995, and available at Chris’s excellent Description Guy website, Chris debates how we can accurately and accountably describe the genesis of information. He concludes that we need to have rich and complex means of describing the context and provenance of information in order to truly understand and use it through time.

Back then, Chris wrote ‘Agencies are episodes in the life of a function’. While this is still such a valid statement, today I have altered it slightly to also claim that systems are episodes in the life of information.

What do I mean by this? I think I mean two things.


The need for information by design

Firstly, in high risk/high value business areas, I mean that information management by design needs to be actively built into system and service procurement.

In these environments it is likely that information will be needed for long periods of time. Systems and services tend to have lifespans between 12 month and 5 years. If necessary business information is needed for periods beyond this 12 month to 5 year window, strategies must be consciously developed to enable all necessary information to be transitioned out of one system or service environment and into another.

These strategies will be more effective if they are developed ‘by design’ or prior to system or service implementation. In addition, these strategies need to strongly consider exactly what information is needed by the business to support its operations and accountabilities now and into the future.

For example, in a lot of the business environments I work with, there is not a good understanding of the need to think through and identify the information needed to ensure information remains trustworthy.

For example, a lot of process-specific cloud applications have excellent metadata for documenting authorisations, approval dates, revisions, alterations, workflows etc, but I have not seen any procurement assessments or decision making processes that assess the exportability and sustainability of this information.

This information is really critical for the ongoing usability and viability of the information in the system, but there is not yet a strong understanding of how much this information, as well as the data it relates to, is needed to support trustworthy useable business information through time. We are not deeply engaging with issues around how much information is changing and how proactive our actions need to be in order to sustain all relevant aspects of information that will be needed for multiple purposes in the future.


The need for data in context

The second thing I mean, I think, when I say that systems are episodes in the life of information, is that, although different systems may, in the scheme of things, be quite transitory in the history of long-lived information, each of them is significant. Each of them is context for the business – how it was performed, the rules and controls and metadata that were applied, the forms of information that were generated.

Currently, however, we don’t have well developed rules or understandings for how this important context should be documented.

We currently don’t do enough when decommissioning one system and commissioning another, to understand how integrally business and information about business were connected to the technology, what this relationship and dependence meant and how relevant information needs to be brought forward and defined in relation to its new technological context.

Today information really is data in context. We can’t understand the data unless we have the context and today systems, as episodic or transitory as they may be, are key components of the context of information. We all need to consider means by which our high risk/high value data can be understood and kept in context, so that it maintains its meaning, value and accountability through time.


Systems are episodes in the life of information

So systems are transient, information is a constant. And a multitude of components form the complex context for today’s digital business information.

How do we better build this understanding into our business environments, to truly support the business needs and accountabilities of the organisations we work within? I would love to hear your views.

And thanks to Chris Hurley, whose fantastic work continues to inspire.


photo by: Doug Kline
Leave a Reply

You must be logged in to post a comment.