source: publicdomainpictures.net
Scientific Numerology is the idea that, since there are some cases in which natural events can be reduced to equations with a few variables, that this must be true of every event. Like any other idea, this idea that complex systems, like ecosystems, societies, and people can all be reduced to simple equations, has consequences. Namely, believers would argue that the solution to any problem is to data mine the world until the problem can be plotted into a graph, and the graph then serves as a "map" to show what we should do.
To sum up the history of this idea in as short a manner as possible, Scientific Numerology is a part of a larger "cult of expertise" which teaches that the rule of enlightened experts will lead to a perfect world. This idea has been around a long time, but in more recent years improvements in computing power led to a resurgence of the idea that enough scientists cranking out data would, one day, find a perfect set of laws, regulations, and whatever else that would give us a perfect society. And there is at least some evidence for this: we can build better buildings, better vehicles, and better machines now compared to any other time in human history, so why not give the eggheads double the funding so we can get our flying cars? However, there is a flaw in the idea of this reduction to numbers and equations which presents itself very quickly, and the consequences of this flaw is concerning.
This is the issue: We assume that the facts generated by a scientist are absolute and self-interpreting.
When a scientist does research, he often has an end in mind. This can cause bias to be introduced in two ways. First, he can choose his methods so that only the data he is interested in is collected. History research is particularly vulnerable to this: if the researcher doesn't include some information and he isn't called on it, then his misinterpretation, intentional or accidental, may be considered "truth." Second, the job of any scientist is to interpret his data to extract a narrative out of it. In my engineering research field, we often call this step "Finding the story to be told," or "Making sense out of the data," since data on its own means nothing without interpretation.
To prevent these biases there is peer review, but what if everyone on the review board is biased in the same way?
For example, what if all economists with some authority assume that the best thing a government can do is maximize a GDP function, when this assumption is false? They may spend decades trying to fine-tune the GDP function, but in the end, the method they use might cause more harm than good.
Another example is in health: if a health indicator such as, say cholesterol, is too high, is the issue one of just trying to make the number go down, or is there a larger disease behind the symptom?
Further still, I remember a food science professor brag in a presentation about how he came up with a way to get record food stamp participation in his state. What about poverty? Where was his solution to help the growing numbers of poor people become independent of food stamps by coming up with creative ways to make healthy food cheaper?
These are the kinds of questions that need to be asked of researchers, but the reason they aren't asked within research communities is because there is a known fixation on magic numbers within certain fields and making them go up, or, making the numbers fit a narrative. I will discuss more about the creation of narratives next time.