Theory is not raw data. To borrow a bit from Kant: “perception without conception is blind, conception without perception is empty.”. Without applying any reasoning or any concepts to the data, it won’t be very useful. Theory is about analysis of data. It’s about linking previous knowledge and research together with newly found data to analyse. Theory is what makes the data relevant. Theory also serves as a cornerstone for us to understand things in life.
I thought it was quite interesting that papers are sometimes rejected on the basis of too little theory. This really shows which significance the theory actually has. The results aren’t in themselves very interesting but it’s when we start to analyse it and try to understand the reasons of why the results came out as they did that it starts to become interesting.
One thing that I’ve thought about is whether theory is qualitative or not. Surely it is, but is it only the qualitative part of for example a research paper? If it was then theory would be subject of the writer’s interpretations, which it is to some degree. Theory is dependent on already established knowledge and it will have observations or data to support it. These observations and data should be captured as objectively as possible. The amount of data backing up a theory is then what determines a large part of the strength of the theory. Combining data from different places to form an argument is also important for the strength of the theory. It’s similar to how cross referencing in journalism will increase the credibility of a news report.
What I think is important to point out is that theory is a layered construct. Theory is built with the help of earlier theories. The wheel isn’t reinvented for every research article for example. What kind of implications might this have then? A naive belief would be that since theories in part comes from earlier theories this will lead to an always expanding theory (if we use the term to encompass all existing theories). This would be wonderful but theories might actually be wrong. Maybe an observation wasn’t actually what you thought it was or a data set was tampered with. You would still be able to theorize on this and might actually come up with a reasonable explanation. Someone could then build his/her research on this and it could potentially start a chain of flawed theory. Obviusly the sooner you discover the flaws in a theory, the easier it will be to retract it and it will face less resistance than for a well established theory. One example would be the theory of evolution which still to this day faces criticism from creationists (though much less than before), even though it has a lot of evidence to back it up.
It’s interesting to see what a big role the natural sciences have come to play in the forming of theories. The scientific community is actually serving as a gatekeeper and it’s deciding what papers and with that theories, that get published. This should result in that only strong enough theory is published.
The notion of theory is not black and white. At first I didn’t really see the point of dividing theory into different types. After discussing this topic more I do think that this classification can be quite useful. Mainly to see that all research doesn’t have the same purpose. Some can focus on only explaining while other research also takes on the task of prediction. It emphasizes that there’s some flexibility in the word theory.