When I was a girl of about 9, I was stunned when I realized that the time could be changed by an act of human decision making. I had always taken what my parents, teachers, and even my older siblings told me at face value as Truth. So when daylight savings time came into my consciousness, the idea that time could be changed, something I thought of as given, something True and immutable, was beyond my comprehension. The idea that we could collectively agree to change the time to one hour later made the concept of time itself seem completely arbitrary. (The next question I asked was “Why would we do that?” But that is a different subject.) I remember when I realized that money was also not True, but a creation of our collective agreement – that it had no inherent value. This did not have as big an impact as the first, which had shaken my world view. I continued to accept many things at face value, but less completely, and by age 15 I was a dedicated skeptic, doubting everything, and embracing data. One of my favorite quotes was, and still is, W. E. Deming: “In God we trust, all others bring data.”
But gradually over time I find myself increasingly questioning data itself. Data is completely neutral, but as soon as it enters our consciousness, it loses its neutrality. We have already filtered it, interpreted it, made meaning of it. In the act of collecting it, we have already altered it. So often I have seen people use data to support their case, rather than to try and understand what is happening. They present it to explain, rather than explore. But at the heart of it, all data is just like time – it only has meaning to the extent that we give it meaning. If we all agree on what it means collectively, that meaning is very powerful – we have co-created our reality, with a shared mental model. That collective agreement is rare, though – most of the time people disagree about the meaning of data, and it is in that very disagreement and the accompanying dialogue that we collectively learn. Which is why I am quite skeptical about the drive many organizations have to a “single version of the truth” – I am not sure it is possible, and even if it were, not sure it is necessarily a good thing. It implies a group-think mental model that does not have room for dissenting voices, something that seems a risk for organizational learning and ultimately for organizational success.
Similarly, the term “data driven,” previously an aspiration, now makes me slightly uncomfortable because it implies that data can decide things. But decision making is an inherently human quality – it requires making sense of something, applying judgment. At best, decisions can be “data informed” – for even if there was a process that could be completely modeled by machine learning such that no human need be involved in the process choices, a human would have still been involved in the creation of the model itself. I would be delighted if more decisions were truly informed by data, rather than opinion and belief, but there is room for opinion and belief in the decision process.
I welcome your thoughts. Please share in comments.