Allowing data to breathe: the impermanence of governance
There is a balance to be found in data governance.
We need periods of data governance to build out understanding. In a way, it says, “explore the depths; develop nuance; sense and try to articulate connections yet untold.” We can use the construct to better understand existing information.
But there always comes a point in our information perigrinations where data cannot, should not, and will not fit the existing construct. That’s when governance can fail us.
Governance, holding too dearly to its authority, will often not bend to the novel forms. The structure will basically gaslight the information, telling the data that it’s wrong, or junk, according to the governance structure. Some rare instances the governance structure might suggest the data is ‘miscategorized’ — not quite dismissing it’s very existence, but saying it has no place in this construct, no matter it’s synergy and connection with other data that has found a place within it.
Data — an information nugget brought to its foundations in hopes to find a meaningful place in the spot it seems to belong — usually has a good idea of where it fits. The problem is often not the data, but that governance didn’t foretell its coming into articulation. The construct was moribund in light of new understanding and meaningfulness.
Allowing data to inhabit the information construct where it fits — potentially in all the places it fits — just makes sense if the goal is to continually increase our understanding of the universe, whatever that universe entails. It makes sense if our goal is to set people up for success.
People do not thrive in a world based on misinformation; missing out on information because it doesn’t fit a too-narrow construct is ultimately such an easy thing to fix, it’s mind boggling to let it linger.
Governance needs periods of expansion, where it checks in with all the data to let change flow.
Thinking about it in terms of content metadata
I think anyone who has ever touched the construction of tagged content on a website or in a content library understands the chaos that ensues when tags are left open for a long time. They become replete with misspellings, almost-same meaningfulness, transposed words, and oddball single-use tags that never form outward connections. It takes a huge amount of time and effort to weed through it, make sense of it, talk to people involved who might hold a particular use very tightly. Depending on the culture, we can butt up against people who obstruct as a knee-jerk reaction. They find it too unimportant to sit down and talk; but at the same time too necessary to a few touchpoints with which they want to spend as little time as possible to allow a change to happen. It’s a form of information dominance through passive aggression.
Run into this a few times, and a governance-first approach seems like the easier, more elegant, less time-consuming route. The decision gets put on the existing taxonomy: out of control, outside of interpersonal behavior, beyond the ken of the average content builder, manager, and moderator. It becomes information dominance by way of getting to the playground first.
Neither form is truly useful to a broader range of people. Both forms resort to dominance to get their way. But if the metadata is a nothing-issue to the business at large — just a conduit to look good from the outside — speed and efficiency will get to the profit-based answer. Governance wins.
But here’s the problem: a governance-first mindset will assume that all information that doesn’t fit has someone playing the passive-aggressive game: not conforming for the sake of their own comfort and in spite of everything and everyone else running smoothly. This behavior, as a summation of more conversations that I could begin to reasonably count, is the exception.
It’s far more likely that the data is simply breaching into a previously unarticulated realm, or that governance was emplaced too early on barely-understood information. Or that the governance was too tightly regulated, and left out key aspects that on consideration — through conversation — left valuable information out of the arena. Or that the mental modeling is overlapping instead of the same, and it’s a bigger issue than simply “taxonomy”. Most of the instanced can be fixed through: gaining trust, talking it through, broadening the conversation, and ultimately integrating information.
Yes, reopening the taxonomy — whether for categorization, metadata, ontology, or connections — will be messy. It will be a form of chaos, more or less all-encompassing and time consuming based on just how off the constructs are compared to current meaningful understanding. It definitely does not adhere well to a budget that assumed it would take a few hours of expert time and decision making.
The world will not stay put just because it’s willed by a few to stay put (even if the “few” is a few percent, and the quantity runs into the millions). If an organizational entity wants to stay abreast with changing reality and ever-enlarging scopes of information and understanding, their information structures have to periodically be opened to change. They need to be built with more care than expediency.
It still strikes me as an art to know when and how to expand an information construct — whether the at-hand structure, strata, or metadata; and whether they can be expanded based on conversations, workshops, surveys; or have the field be freeform for a while to capture a solid set of tags to then work through with people. It’s messy, and chaotic, and uncertain, and takes a certain courage to walk into that storm with assurance in self and others that we’ll make sense of the construct anew.
Building change into governance
In my experience it’s the rare corporation that has allowed for the construction of a process through which change can be codified and managed, with documentation and enculturation to adhere to change as the status quo, even as employees shift in and out of the relevant roles. Even in design system documentation, the focus is usually constrained to making sense of what exists, not pathways to bring up change or understanding all the touchpoints that need to be checked in with as that change is explored. Without including an always-on process for change, though, we lose understanding. We ignore what needs to be explored and we dismiss where we should wonder. So while a top-down-minded management is focused on budgetary constraints, they are building a system that will eventually either break or slowly disintegrate.
There are structures that brilliant minds formed that seem to be mostly able to keep up with the changing scope of our understanding. The two that most easily come to mind are the Library of Congress classification system and species taxonomy; neither were quick or easy to formulate. Both still work generations after inception; both are predicated on a systematic expansion, where the assumption is that no content will be left out; both can still take a chunk of time to incorporate expanding information. At no point is something that fits the general form — media or living being, respectively — told it doesn’t count because of another arbitrary characteristic. If it exists, it’s brought in, and a process exists to manage the inclusion — not to manage through exclusion. It’s intrinsically bottom-up.
This is not how most governance systems work today. Most governance systems default their top-down reading as how it’s managed. Even if it’s not what was intended as it was built, even if the process to expand was kept out of the official documentation but intended to shift and grow with the content through manual manipulation of the structure, the governance as it’s used by the people managing it (or desisting management of it) stays the same. Often for no other reason than, “that’s the way it’s done.” It never ocurred to the governance managers that it would/could/should shift, and they wouldn’t know where to start, so…why fix what’s not broken? And from the top-down, it appears unbroken; it’s only from the bottom-up — from the content that doesn’t fit — does the brokenness become so apparent that it unavoidable. The bottom-up content can’t comprehend why the top-down view insists everything is fine.
Fractal application
This truly is about information, and yet even as I write it I see fractal patterns through so many of our ‘governance’ structures. I’ve said this before, and I’ll probably say it ad infinitum: information is everything. Culture is a form of information structure: it’s accepted signaling of information shortcuts within a cohort so they get to more nuanced and ambiguous information faster — the stuff that is changing, unusual, rare, unique, or novel. If there’s no checking in between the holistic and details, the same dissonance between inclusion and exclusion happens as in pure information structures.
And here’s what some sense: to allow the governance of information to loosen and expand and contract as appropriate, implies that — fractally — we should allow the governance of people and our shared understanding. Through culture, through laws, through regulation, through our sciences both soft and hard, through our interrelations and our accepted -isms and cognitive biases, we should through the fractal application of information understanding allow those variations of information structures to do the same. Once you allow something as simple and intrinsic as information to take on the changability of life, where does it stop? How much of what is known will need to be updated? How often? Why can’t what I didn’t want to understand in the first place just stay the same?
If change is the topmost thing to fear, they are not wrong. As an information architect, as someone who is fascinated with right-sizing, balancing, and setting up as many people for success as possible, change is not the most fearsome thing. Change, actually, may be the only reasonable status quo. The most fearsome thing in my mind is misapplying information, creating a fractal cadence of mistakes so deeply rooted and intertwingled it becomes a wonder of how to even find the points from which to pivot.
If our goal is to set up people — all of us today, into a future as long and bright for our species and our world as we can manage — for success, allowing information to breathe just makes sense. It not only keeps our information in all of its various forms more closely aligned with changing reality, it reminds us all on a regular basis that checking in with the assumed taxonomies and structures is normal. It’s not the exception, it’s not unusual, and it’s not a thing to be shamed or feel shameful about being out of step. It’s just part of our expanding universe.
There is an intrinsic kindness to monitor information structures intended for glancing or introductory information, to make changes in chunks as the density of changed information hits a tipping point. But that’s a writing for another day.
This supplements my explanation the concept of expansion and contraction of data in systems thinking.
