In This Section

4C Partners

Deutsche National Bibliothek
Keep Solutions
National Library Estonia
The Royal Library
Statens Arkiver
UK Data Archive
University of Glasgow

'Uncertainty: the final frontier' by José Borbinha

In simple terms, we can define an “engineered system” as a “bunch of things” one puts together (a structure) for a purpose (concern) that it is expected to help addressing (a behavior). We also say a system is “open” when its internal behavior produces outputs that can be perceived outside of the system (in engineering we are used to classify those outputs, for pragmatic purposes, as “matter”, “energy”, or “information”). In fact, being more comprehensive, a system is open when it also gets inputs, thus an open system is one that “exchanges” matter, energy, or information with its exterior.

All this means systems exist in contexts (the exterior that interacts with the system), so “systems thinking” should care not only about what systems have and do inside, but also what they get and give to the context that relies on them (and decides about them…). In this sense, I expect it to be consensual that we can consider most of the issues concerning digital curation are about some kind of system.

Continuing this game with concepts, and motivated more by pragmatics than by academic rigor, I’d add one more dimension of inputs and outputs to this “systems thinking,” that is considering the main concern for curation: time. I’m assuming most readers already are motivated by concerns of what we are used to calling “digital curation” (that I’ll call from now “our domain”), or at least “digital preservation”, so I believe there is no need to detail why “time” is important (recall when we say that “digital preservation is interoperability in time”…). However, maybe I still can have a new way to look at “time”...

In our domain we are used to understanding “matter” as made of multiple classes: buildings, silica computer’s hardware, carbon computer’s hardware (people ;-), etc. From this perspective I also enjoy thinking of software as “dematerialized matter” (we do things with it that give us a “physical” sense, such as “sell” or “buy”, “install”, “customize”).

Concerning energy, besides the classic “electrical energy”, to keep our silicon stuff working, I also enjoy thinking about “stamina” (or “will”?)  as the energy that keeps the carbon hardware working. Finally, I also like to think the infamous “money” as energy, as it serves for nothing specific, but nowadays is a fundamental requirement for any hardware to work…

Concerning information, I also believe we all agree to understand it as made up of two parts: data, which we measure by “size” (Byte, Kilobyte, Megabyte, Gigabyte, Terabyte, Petabyte …); and metadata, which we measure by “quality” (maybe we could agree to measure it as bad, fair, good, excellent… uau!!!).

Finally, on time, which we can measure as “past”, “present” and “future”. But, it might not be so simple… Just ask any Historian, and you’ll learn how “past is not so simple…”! Or just watch any “decision maker”, and realized how the present can be “distressful”! And, about the future, just ask anyone…! But maybe time can be a “workable” dimension, anyway, if we can “measure” it, for example, by “events”… Let us explore that!

From we learn:

Event, noun \i-ˈvent\

  • something (especially something important or notable) that happens
  • a planned occasion or activity (such as a social gathering)
  • any one of the contests in a sports program

Hummm… no good! Not really this… After looking more, I found what I was looking for, at

2.17 Event

  • An event could be one occurrence, several occurrences, or even a nonoccurrence (when something doesn’t happen that was supposed to happen). It can also be a change in circumstances. Events are sometimes referred to as incidents or accidents.
  • Events always have causes and usually have consequences.
  • Events without consequences are sometimes referred to as near-misses, near-hits, or close-calls.

This is interesting! It looks like events can be certain or… uncertain:

“Uncertainty (or lack of certainty) is a state or condition that involves a deficiency of information and leads to inadequate or incomplete knowledge or understanding. In the context of risk management, uncertainty exists whenever your knowledge or understanding of an event, consequence, or likelihood is inadequate or incomplete.”

Visiting the ISO31000 standard now (in fact I cheated, as it was from back when I started…), we learn that “risk” can be defined as “an objective exposed to uncertainty”, and so that “uncertainty” is simply “the lack of sufficient relevant knowledge for a particular objective”. Exploring the same reference, we can also realize of more these interesting concepts:

  • Vulnerability – an uncertainty that could have a negative effect for a particular objective
  • Threat – the exploitation of a vulnerability
  • Loss – a negative effect on an objective from a threat
  • Opportunity - an uncertainty that could have a positive effect for a particular objective
  • Gain – a positive effect on an objective from the exploitation of an opportunity
  • Risk Management - The systematic process of identifying and analyzing risks and responding by defining controls
  • Risk Register - record of information about identified risks

Accordingly, if we scope these concepts in or domain, maybe we are authorized to say that:

  • Control – is a measure that we can put in practice to minimize loss (the main concern of digital preservation) or also to maximize gain (the maximum concern of digital curation)
  • Costs – is what we have to give up for the controls!!!

In this sense, a control is anything we consider to apply either to minimize negative impacts (mitigate threats) or to take advantage of opportunities to produce value! It means maybe we also can consider these concepts:

  • Relative value – This concept can be defined as a comparative ranking of a set of assets, for which we do not have an absolute value (this can be often the case of assets in the cultural heritage sector). This is important in a risk management approach if our focus is not one asset alone, but a set/collection of assets where we realize each can be exposed to different uncertainties, thus different controls have to be considered and consequently decisions on priorities have to be made based on the relative value of the assets.
  • Added value – In this context, this concept can be defined as the value that results from a control if applied to explore an opportunity. If we can measure that value, than we can claim we can quantify an added value for the related asset as a result of that control (this is curation, which, please note, can be considered with no need to have defined an initial value for the asset –which is often the case in the cultural heritage sector…).

So to conclude, here is my hypothesis: maybe we could develop a risk management method that (adapted from ISO31000) uses proved relevant risk assessment techniques (taken from ISO31010…). This can be proved to be useful in supporting decision making in scenarios where costs of controls have to be considered, making “time” a new dimension to “measure” the challenge of the digital curation (and especially in its concerns of preservation).

We will never be able to accurately preview the future, but we already know something about the vulnerabilities of the technology and the contexts of curation, about the threats to our assets and also about risks. We also know about some controls to apply, etc… so… it might be a good idea!

Uncertainty: the final frontier. These are the voyages of data curation. Its continuing mission: to explore strange new worlds, to seek out new life and new civilizations, to boldly go where no one has gone before.

Jose Borbinha, IST (Lisbon Technical University, Portugal) and INESC-ID.

INESC-ID is leading the task to create a report which will look at a range of inter-related issues from a perspective of Risk Management. Using case studies, the role of risk and risk assessment will be considered in relation to curation as one of the principal drivers for governance. In that sense, not only cost but also benefit, impact and value (and its relation to cost efficiency) will also be considered terminologically and by sector to try and characterise the influence of these factors as determinants.