Submitting the Theory of Quantum Gravity Part 2: Needs More Data

Do you know what the most frustrating phrase in science is? It’s “It needs more data”. As in “I acknowledge that you state that you are observing a pattern emerging from your data, but I think there’s a good likelihood that you’re just attributing meaning to random noise”.  The reason why this phrase is frustrating is because the phrase is both kind of right and kind of wrong.

Very often in science, the first signs of a fundamental discovery are vague. You observe an anomaly and try to make sense of it. You formulate an initial hypothesis, which is oftentimes at least partly wrong, even if the hypothesis contains a grain of truth. You try to show this grain of truth to others, but they can only see the noise, because the others haven’t spent years, or decades, tackling the problem.

If you’re skilled enough, you device experiments that test your hypothesis. And when you observe that the test produces the results you’re expecting, you get excited: “Now I’ve verified my hypothesis”, you think. But when you show the results to others, they see the uncertainty and not strengthening of the outlines of the initial hypothesis.

And rinse and repeat: you device experiment after another, always building on the previous results. And each time the outlines of hypothesis become stronger and an inkling of a theory begins to emerge. Here we need to pause for a moment to discuss the differences between a hypothesis and a theory. A hypothesis is a (highly) educated guess, providing a qualitative explanation for an observation. However, a theory is something more rigorous. Often (but not necessarily always) quantitative. With physical phenomena the theory includes equations describing it.

Then you take your equations and your data and show them to others, hoping that they’re convinced. But still, all they see are the blank spots where the equations and experiments don’t meet. Or in my specific case: the equations are so fundamental in nature that they describe minimally disturbed motion. And the only thing that experiments can show (almost by definition) is (maximally) disturbed motion: a situation where a liquid has turned into a solid. Literally, not figuratively. So, can equations that describe undisturbed motion be proven by something that described disturbed motion? Can such equations ever be experimentally proven?

I’d like to answer by saying: “I don’t think I can prove the equations, but I don’t think I need to”. That is, I’ve introduced the idea to the world and sufficient initial evidence to allow the theory to be taken seriously. It shouldn’t be up to me to provide conclusive data to back my theory up.

If you’re not a scientist, you might consider my statement reasonable. However, if you are one, you just might feel that my statement is a copout. Shouldn’t I be the person who provides the data to prove the theory? After all, truly new theories don’t come along that often.

And here we come to the title of today’s post: there is never enough data. I’ve spent the last five years gathering data and there’s still not enough. I know the manuscript could be better, but this time a reasonable reviewer should see the merit in the paper even through its flaws. I’m just hoping that I’ll get such reviewers. Or that I’ll even pass through the editor’s desk and not get rejected without peer review.

I’ll just have to hope for the best…

And here is the mandatory picture of the day. It’s a 3D model of an adhesive bond in a glulam  sample made with my lignin adhesive, generated by stitching multiple x-ray images in (industrial) computer tomography.

P.S. Very soon I’ll be presenting the version of the manuscript that’s going to be submitted to Nature. There’s not too much change in the substance, but the structure had to be changed quite a bit to fulfill the demands set by the paper.