Context

All data is interpreted, but I don’t think that’s anything special about data. I have simultaneously a very high and very low standard for objectivity, depending on what exactly we define objectivity as. It comes down to math, which I can’t help but think of as objective. Lots of people share this belief, but not all of them seem to really understand what math is. Math doesn’t give naked facts. There is no objective truth to any single sentence that a logician can write down, even cliché examples like 1+1=2. What math can talk about are implications. Math can tell you which assertions lead to which facts, and those implications are objective, but the facts themselves are meaningless without context. I’ve heard people tell me that my definition for objectivity is their definition for subjectivity, in which case my standard seems pretty low. I’ve also heard people claim that all things are inherently true or false, in which case my only taking things that can literally be proven without room for disagreement as objective seems like a pretty high standard.

While I’d say that objectivity to be the solely in the domain of mathematics, this idea that things require context to be meaningful is not at all unique to math. Every discussion needs a context, a language, a setting, and an agreed upon reality from which to build. Every idea comes inside a framework. Without the framework the idea is meaningless. Music is enjoyable because it builds, breaks, or completes patterns, which is a somewhat meta framework in and of itself. Art is created whenever it is consumed, since the viewer puts it in a new context, thereby creating a new idea. To perceive is to create.

Even physics seems to agree. It is an accepted model these days to say that position and momentum both share the same information, and the more information exists about one the less exists about the other. It is only by choosing what to measure that any meaningful ideas can emerge.

So by these axioms subjective things require context, and so do objective things. It would be extremely surprising to learn that some list of numbers somewhere is actually bare reality with no need for addition interpretation.

So yes, all data is interpreted, and through one lens a dataset can say something entirely contradictory to what it says through another lens. This is ok. Everything is like that. Even the things that really are objectively true are like that. We can solve this ‘problem’ in the exact same way the mathematicians do: by owning our own context. In math we never stray too far from our axioms. We can always see them in the rear view mirror, and they are invaluable tools. We never claim to find facts, only implications. We are explicit about which things we are assuming. Reasonable people can agree to which context they are using. If there’s some disagreement about which context should be used, then that question needs to be settled long before any data can be brought to bear, as it is one of philosophy, a field in which data is notoriously unhelpful. Unfortunately, that conversation would also require a framework, so it may take a while to sort it out.

This is by no means controversial. Scientists put units on their data to immediately provide context. They also accompany their data with papers, to provide yet more context. Conversations come with an implicit decision for which languages in which they will be conducted. Settings come with a cultural understanding of which things are acceptable and which things are rude. Nothing is objective without context, but context it not too hard to come by.