A scientific New Year's resolution

The modern paradigm of ‘Big Data’ is all well and good – but does harvesting it leave room for inspiration? Russ Swan would like us all just to slow down a bit and have a think…

THE ONLY constant in the world these days is the constancy of change, as someone cleverer than me once said. To which I might add that the constant, in mathematical terms, it is a geometric rather than linear one: change is not only ever-present, but accelerating.

We see this on almost a daily basis in the laboratory. Every new project seems to demand a faster turnaround, every new instrument promises to perform more analyses in less time, every new protocol aims to speed the process from samples in to data out.

With almost risible lack of irony, the newer, faster, better lab equipment is promoted with claims that its speedier performance will free the scientist for more important tasks. It seems to me that the reality is the exact opposite – that higher throughput and faster turnaround actually leads to less time for the hapless human. More automation means more vigilance, with scarcely time to make a cup of tea before the assembly line hooter demands yet more human intervention.

What we’ve lost, as a result of the inexorable industrialisation of science, is the very thing automation was supposed to give us: time.

This isn’t ‘spare’ time or ‘free’ time, the very idea of which is enough to make management apoplectic. It is auxiliary time, time for mental processing and (dare I say it?) imagination. For science should not be a drudge, a robotic action performed by mere mechanical executors. Science is, I think, a creative process that demands at least some opportunity to kick back, consider the situation, solve the crossword, and wait for that holy grail of scientific endeavour – inspiration.

These may be merely the rantings of a prematurely middle-aged geek, but there already exists a small and growing movement to promote ‘slow science’. Taking their cue from the slow food (ie, anti-McDonalds) movement, slow science is opposed to the widespread use of performance targets and promotes the old-fashioned notion of research driven by curiosity.

I think there is an interesting philosophical point here, that we should all take heed of. Modern big science projects inevitably demand the collection of huge quantities of data, and it is in the processing of this that the worlds of fast science and slow science may, gently, collide.

Take the Galaxy Zoo project, in which highly automated telescopes scanned the heavens and produced hundreds of thousands of images. That part is the fast science, the industrialised collection of data. Slow science comes in the interpretation, with tens of thousands of volunteers employing the most sophisticated image sensor in the known universe to consider these pictures and classify them.

The image sensor in question is, of course, the human eyeball, which when coupled to the human brain provides better pattern recognition and detail observation than the best CCDs and computers ever built. The same truth probably applies in your lab.

I was struck by this last autumn when I was given a chance to play with the new flagship Olympus IX83 inverted microscope. Coupled to an advanced image processing system and displayed on a large high-definition monitor, the instrument gave the sort of high-quality pictures you’d expect. I was moderately impressed. But placing my eyes against the oculars, the difference was a revelation. The improvement in what could be seen and interpreted was partly due to the stereoscopic view presented optically, and partly the result of collecting the actual photons that had fallen on the sample instead of the machine-processed synthetic ones.

So rather than classifying laboratory activities as ‘fast’ or ‘slow’ science, we might instead think of them as digital or analogue. The Human Genome Project and the Large Hadron Collider are examples of digital science – vast quantities of data are generated and processed mechanically.

These remind me of the way a Victorian naturalist might act, by collecting every butterfly in the garden and mounting it according to its size, shape, and habitat. Once completed, the observer might hope to glean some insight into the nature of butterflies, although of course the only thing that is guaranteed is that there are no more butterflies in the garden.

The analogue approach might be to observe the beasts and make copious notes, and then to consider what patterns emerge. This is really the way many great scientific breakthroughs have been made – like vaccination, penicillin, and even relativity. They all came from observation and thought, rather than data collection and number crunching.

So here’s my suggestion for a scientific New Year’s resolution. Every now and then, turn away from the monitor and peer into an instrument. Think about what you are seeing. Do less work, and thereby do more real science.