The end of theory?
This is a transcript of a 3-minute thesis I gave on this paper about Big Data in biology in 2018.
Modern Science is moving faster than ever. We are now in an information age, and the sheer quantity of measurements is making Big Data bigger all the time. This explosion in data offers a tantalising new way to conduct science; away from a hypothesis driven, to a data driven approach.
A famous historian once declared the fall of the Berlin Wall to be the end of history. Similarly, scientists in some circles are claiming the Big Data regime heralds the end of theory. With Big Data, so the story goes, if we can find the right correlations between certain “measurables”, we need not bother finding the underlying laws which give rise to these correlations. The authors of my paper argue that this is misguided, and serves only to slow progress in biology. Instead, they suggest a different approach.
There is a difference between fact and concept: * Facts are the “what” – the data we are now gleaning at an unprecedented rate * Concepts are the “how” – they tell us about links between facts, as well as suggesting new areas and experiments to look for them Conceptual frameworks provide the context in which we view the facts. Adrift in a sea of Big Data, we need conceptual frameworks to guide us, to direct us onwards.
In science, often it is the change of conceptual framework which leads to breakthroughs. Consider Schrödinger’s proposal in What is Life: that an “aperiodic crystal” contains genetic information in the form of covalent bonds. It was this conceptual framework that inspired Watson, Crick & Franklin to propose the double helix structure of DNA, founding the field of molecular biology along the way.
Science is not merely an accumulation of facts about the world, but a process, whereby we make conjectures (or hypotheses) about how we think the world might be, before testing those guesses with experiment.
And the most precise theoretical framework for formulating predictions that we know of is mathematics. By mathematising our hypotheses, we gain at least two benefits: * We provide a quantitative prediction against which we can test our theory, by experiment * Maths also sharpens our thinking by forcing us to formally state all thoughts and assumptions Mathematical theory in biology has revealed the similarity in seemingly disparate concepts (for example, the broad applicability of the Boltzmann distribution), and can illuminate trends in the data.
It is responsible for many of the biological breakthroughs in recent years, so with the advent of Big Data, we should not cast theory aside, but rather help us navigate through all of this new information.
Plus, without mathematical theory in biology, we wouldn’t have this class (!).
Naysayers may object by saying: “Biology has made a lot of progress in the past without excessive mathematisation. Why go through all the effort?”
To that, I’ll answer with a quote from Von Neumann: “If people do not believe maths is simple, it is only because they do not realise how complicated life is.”
Life is the most complicated thing there is, but mathematical theories can help us break down and understand it, revealing insights that might otherwise remain hidden.