A post. A philosopher. A pulpit.
I came across a post by Winfried Adalbert Etzel this week. One of those references you go to when in doubt. Fellow DAMA International® member. Someone who is out there doing the work, not selling the dream. All the way from Norway, bringing ideas that land like a well placed sermon on a Sunday morning.
And it made me think. Which is exactly what good posts are supposed to do.
But before I get into it, let me make a distinction I have been carrying for a while: the difference between a preacher versus evangelist, an evangelist focuses on conversion, a preacher focuses on teaching and equipping. I do not trust evangelists, I always feel they are trying to sell me something I do not need. Winfried falls squarely in the second category, and in our field, that matters more than most people admit. Data preachers are rare, we need more of them.
And so, data brothers and data sisters, let me step up to the pulpit for a moment. Words are important...Definitions matter...Context is relevant.
I can see myself up there, sharing the good word of the Book. And by Book, I mean the DMBOK 2.
Now. Back to business.
Winfried brought Wittgenstein into the room. He is right.
Seventy five years after his death, Wittgenstein is still making data people uncomfortable. Good. We deserve it.
If the name only rings a faint philosophy 101 bell: Wittgenstein was the stubborn Austrian philosopher who argued that meaning does not live inside words. It lives in how we use them. That one idea is still blowing up data meetings today, and he has been dead since 1951.
His insight is simple and devastating. Ask Sales what a "revenue" is (more on that HERE). Then ask Finance. Then ask Marketing. Three different answers. All potentially correct. That is not a data quality problem. That is a human organization doing exactly what human organizations do.
Winfried Adalbert Etzel connects this to AI and he is right to do so. AI systems do not renegotiate meaning over a cup of coffee. They act on definitions as if definitions were meaning. Feed an AI system a term that means three different things across three departments and watch what happens. Actually, do not watch. It is not pretty.
That is a governance gap we cannot catalog our way out of.
Hat tipped. Torch passed. Now let me add my floor to the building.
The glossary is a tool. Just a tool. Same as AI is a hammer.
What surprises me is how often I see colleagues expecting governance to succeed or fail based on whether they have the right tool in place. A consolidated business glossary is just that: a tool. A tool to align people. Nothing more. Nothing less, same with AI.
I always tell people: AI is a hammer. Everyone thinks they know how to use a hammer. How difficult can it be? Tell that to my wall. A hammer is useful if you know how to use it. If you do not, it is dangerous. Not because the hammer is evil, because you skipped the part where you understand what you are actually trying to build, or do to someone's head.
I have seen teams celebrate the go live of a glossary that nobody ever opens. Full ceremony. Stakeholder sign off. A nice slide in the all hands deck. And then six months later the same three departments are still using three different definitions of revenue and nobody can explain why the reports do not match. That is not a tool problem. That is a people problem wearing a tool problem as a disguise.
A glossary in the hands of a team that has never talked to each other is not governance. It is a document waiting to be ignored.
The tool is not the point. The people holding it are...THEY ARE THE KEY
The tower is comfortable. That is the problem.
Data governance is a discipline that can make you feel very important from a very safe distance. The frameworks are real. The DMBOK 2 is scripture. The methodologies are solid. And somewhere along the way it becomes tempting to build yourself a very nice tower, surround it with diagrams, and come down on Fridays to let everyone know that what they are doing is wrong.
I know this character. I have met this character. On my worst days I have been this character.
The tower is seductive because it feels like clarity. But what it actually is, is distance. And you cannot govern people you have never talked to. You cannot align meaning across an organization from the eighteenth floor with a PDF.
Governance without proximity is just opinion with a logo.
There is a version of this job where you produce beautiful documentation that nobody reads, attend meetings where everyone nods and nothing changes, and then wonder why adoption is low. The answer is not a better framework. The answer is a conversation you have not had yet.
Come down from the tower. The coffee is better down here.
Here is what actually works. A technical data steward who sits close enough to the business to hear when a term has shifted in practice before it shifts in the data. I have seen this happen. Marketing starts calling leads "customers" two stages earlier than the contract says they should. Nobody announces it. It just happens. And if your steward is not close enough to catch that in conversation, it will show up six months later as a discrepancy in a dashboard that nobody can explain.
A data owner who does not see governance as compliance theater but as the mechanism by which their domain actually functions. A glossary that is never finished because the business is never finished.
When you open real lines of communication, when you invite people into the process instead of presenting them with conclusions, something shifts. They stop being subjects of governance and start being authors of it. They imprint their knowledge, their context, their practice into every definition, every rule, every decision.
And then something works. Not because the framework is better. Because the people own it.
Wittgenstein said the meaning of a word is its use in the language. I will add: the meaning of a governance initiative is its use in the organization. If no one uses it, it means nothing. If everyone helped build it, it means everything.
Data is not just what we measure. It is how we understand each other.
The best governance work I have ever seen looked less like a framework and more like a long conversation. Messy at times. Slow at times. Occasionally derailed by someone bringing up something from three years ago that nobody had documented but everyone remembered differently.
And it worked. Because the people in the room felt that the process was theirs.
That is the only governance that lasts. Not the governance that is handed down from the tower. The governance that is built at the table, with real people, with bad coffee, with someone writing on a whiteboard and someone else saying "wait, that is not how we use that word in my team."
That moment. Right there. That is governance happening in real time. That is Wittgenstein sitting in your data meeting whether you invited him or not.
Data brothers and data sisters: come down from the tower. Bring the DMBOK 2 if you must. But leave the PDF at the door and start with a question instead of an answer.
Because at the end of the day, data is not just infrastructure. Data is how organizations see themselves, understand each other, and make decisions together.
Data is everything.
Data is love.
Winfried Adalbert Etzel , thank you for bringing Wittgenstein back into the conversation. Read his full piece in the comments. Worth every minute.
People in this piece: Winfried Adalbert Etzel