Like the card-carrying member of the snowflake generation that I am, I have recently spent a lot of time thinking about my ‘purpose’. I am, nominally, a scientist, which makes this a harder question then it might otherwise be. If I was a bike mechanic, a midwife, or a farmer, at day’s end I could gaze at an array of working bikes, healthy babies, or ploughed furrows, and sleep easy at night. But, unfortunately, I am involved in the production of scientific knowledge, which no-one has yet managed to eat or ride to work. What use is it then? Why am I doing it?!
If you ask the government, whose tax revenues fund much science, then they will pragmatically and firmly tell me one set of answers. I am supposed to figure things out so that:
And these practical concerns are definitely active in how I, and many other scientists I’ve met, think about their work (at least 1. and 2.). But it’s far from the whole truth.
The real motivation to continue attacking a research problem is often not its direct application, but a driving curiosity to understand a phenomena; applications are a happy corollary to successful understanding. When everything clicks, making sense of observations, that is when the job is done. A large part of why I became a scientist is chasing the enjoyment of this ‘clicking’.
But ‘clicking’ is quite a fuzzy concept. If you have two plausible explanations how do you choose between them? Well, experiments! We are believers in empiricism after all - as da Vinci said (maybe) ‘Experience is the mother of all certainty.’ If your theory predicts flying pigs and the imminent arrival of a supernatural messiah you are unlikely to hold up to the harsh light of evidence.
Well, perfect, we think, patting ourselves contentedly on the back. Let’s just try and predict the world, and whoever does it best is officially, objectively, the best. Their understanding of the world must be the most accurate picture we have, and before long we’ll build amazing iphones and cure cancer; bring on the technocratic utopia.
But here, I am afraid, I must disagree. I posit to you that predicting the world is only one component of why we do science, falling into the same category as utilitarian desires to improve technology and medicine – a useful corollary. If 10,000-layer bottomless neural networks have taught us anything, it’s that we can predict things perfectly in wholly intellectually unsatisfactory (though still useful!) ways. In fact, sometimes theories that don’t match up with all experiments are more valuable!
Rather, science is the business of creating useful schemas for thinking. The real universe, as far as we can tell, is vaster than I am able to conceive, full of unimaginable numbers of elementary particles just doing their thing. They interact, and as a result a sun explodes, or I trip over my shoelaces. We cannot think our way around this. We conceptualise quarks, gluons, and electrons as little things whizzing around and obeying the rules we wrote down – and maybe that is exactly what they do – but they don’t care what maths we’ve written down, and they definitely don’t care about my little intuitive pictures of little coloured spheres with attached magnetic dipoles. Our theories describing these particles are just stories, very quantitative and useful stories, good bedtime stories only for very peculiar types of children, but just stories that we tell each other to explain things. Stories to make a ‘click’ of understanding in your head.
I want to push this a little bit further before moving on. Scientific theories do not have to correspond to ‘reality’ – whatever that is – at all. In fact, they probably shouldn’t. Maybe reality is 10100 eleven-dimensional vibrating strings, or an omniscient god that makes sure all the particles move at just the right speed to trick us into thinking there’s a rule at work, or perhaps the universe is just a simulation. In all of these ‘realities’ ideas like gravity, fluid dynamics, neuronal action potentials, and my memory of my grandfather are completely invented fabricated nonsense ideas to help us explain otherwise impenetrable things. There’s a temptation to be sad when you realise we’re just peddling convincing fictions, but I think this should actually make you excited. How much more manageable does future scientific progress now seem when we don’t have to understand reality, we just have to write useful stories? Now we just need to work out what we mean by useful!
When Newton said “If I have seen further it is by standing on the shoulders of Giants” he wasn’t talking about predicting better using the previous generation’s most highly optimised neural network hyperparameters. He was talking about the power of previous generations of researchers to produce a human-intelligible picture of complicated phenomena, allowing him to use their advances to understand new things. We all know this. Things are much easier to understand when we’re guided through the process of building our intuition and understanding, versus coming to these conclusions ourselves. That is why it took humanity many thousands of years and one Einstein to come up with General Relativity, our most successful description of gravity yet; but I, along with 140 other doofuses, was able to derive the key equation in a one semester course.
So, what I actually want to do as a scientist – at least until the funding dries up – is contribute to our generation’s layer of ‘the Giant’ from which future generations can gaze. Useful additions to the Giant tend to correlate with those that help us predict Nature’s behaviour, or create better therapies, but this isn’t their primary goal. Instead my desired contributions are those that, in helpful and powerful ways, guide the mind towards the understanding of a system we have spent many man-hours exploring so that no other poor sods have to do the same thing. A concise, preferably quantitative, picture, that students of the next century can cover in 5 rushed minutes at the end of a shoddily delivered holographic lecture so that the next time they find their mind wandering and wondering towards this part of mother nature they will be able to use the intuitions we’ve developed without having to go through the pain of producing it fresh.
And this is why I think scientists are artists. A machine can make pots that hold your food. A potter makes art: they’re still pots that hold food, but they communicate human meaning (okay, I haven’t yet worked out my philosophy of aesthetics, maybe that’ll be next issue – lucky you). What transforms someone from a maker to an artist? I’m not quite sure, but it’s something to do with escaping the bounds of pure utility to communicate something to other people.
Similarly, we are tasked with telling stories. These stories must transmit understanding of the world around us, and hence are heavily constrained by our observations – like a pot that has to hold your food. But really, we’re spinning tales that can be communicated, providing the fuel and inspiration for the next generation of scientific discoveries. If this isn’t one of the most human, engaging and enriching activities around, a story that started millennia ago to which we are able to contribute our footnote, if this isn’t art, then I will resign from humanity and meet you at the exit.