Выбрать главу

Job done.

I think of starting down, but having overcome my qualms to get this high, I cannot resist the opportunity to poke my head out of the top. At the back of the electrical gear is a small ladder which leads to an access hatch in the roof of the housing.

I go up the short steps of the ladder, undo the catches, and heave open the access hatch.

My knees wobble a bit. I push my head through the hatch, like a tank commander. I look around. There’s a rubberised walkway on top, and a set of low handrails, so in theory I could go all the way out and stand on top of the housing. But I’ve never done that, and I doubt that I would ever have the nerve.

Still with only my head jutting out, I look back at the hut, a huddle of pale rectangles. The perimeter circle is hard to trace from this elevation, but eventually I glimpse the spaced-out sentinels of the tripods, with the scratchy traces of my own wheel tracks between them. Then I pivot around and try to pick out the causeway. But it’s harder to follow than I expect, seeming to abandon itself in a confusion of marsh and bog. I squint to the horizon, looking for a trace of its continuation.

Strange how some things are clearer to the eye at ground level, than they are from the air.

Birds must know this in their bones.

THE NEXT DAY, the computers running again, I squeeze our data until it bleeds science. With the vector tracking, we can trace the response to the sparrowhawk across all possible interaction lengths. Remarkable to see how effectively the ‘news’ of the sparrowhawk’s arrival is disseminated through that vast assemblage of birds.

Because there is no centralised order, the murmuration is best considered as a scale-free network. The internet is like that, and so is the human brain. Scale-free networks are robust against directed attacks. There is no single hub which is critical to the function of the whole, but rather a tangle of distributed pathways, no one of which is indispensible. On the other hand, the scale-free paradigm does not preclude the existence of those vector domains I mentioned earlier. Just as the internet has its top-level domains, so the brain has its hierarchies, its functional modules.

Would it be a leap too far to start thinking of the murmuration as hosting some level of modular organisation?

I jot down some speculative notes. No harm in sleeping on them. In the meantime, though, I write up the sparrowhawk results in as dry and unexciting manner as I can manage, downplaying any of the intellectual thrill I feel. Passive voice all the way. The sparrowhawk was prepared for remote control. Standard reduction methods were used in the data analysis. The murmuration was observed from twenty spatially separated viewing positions – see Fig. 3.

The way to do science is never to sound excited by it, never to sound involved, never to sound as if this is something done by people, with lives and loves and all the usual hopes and fears.

I send the latest version of the paper back to the journal, and cross my fingers.

I OPEN THE glove compartment and take out the latest version of the paper. I skim it quickly, then go back through some of the more problematic passages with the yellow highlighter, before adding more detailed notes in the margin. My initial optimism quickly turns to dismay. Why in hell have they opened up this whole other can of worms? I squint at an entire new section of the paper, hardly believing my eyes.

Sparrowhawks? Robot sparrowhawks? And pages of graphs and histograms and paragraphs of analysis, all springing from work which was not even foreshadowed in the original paper?

What are they thinking?

I’m furious at this. Furious with the journal editor, for not spotting this late addition before it was forwarded to me. Furious with the authors, for adding to our mutual workload. Furious for their presumption, that I will presumably be sufficiently distracted by this to overlook the existing flaws – like a magpie distracted by something shiny. (Except that’s a myth; corvids are not attracted to shiny things at all.)

Furious above all else that they are prepared to squander this good and original science, to slip it into this paper like a lazy afterthought, as a kind of intellectual bribe.

No, this must not stand.

I put the 4WD back into gear. I must settle my thoughts before firing back an intemperate response. But really, I’m enraged. I bet they think they know who I am. I imagine encountering them out here, running them down, feeling the bounce of my wheels over their bodies. I’d stop the 4WD, get out, walk slowly back. Savour the squelch of my boots on the marshy ground.

Their whimpering, their broken-boned pleas.

‘You think this is how we do science, do you?’ I’d ask them, entirely rhetorically. ‘You think it is a kind of game, a kind of bluff? Well, the joke’s on you. I am recommending your paper be rejected.’

And then I would walk away, ignoring their noises, get back into the 4WD, drive off. At night their cries would still come in across the marsh, but I would not let myself be troubled. After all, they brought this on themselves.

But that sparrowhawk, I’ll admit, was beautiful.

I GO OUT to the walk-in traps and collect the overnight haul. There are almost always some birds in the snares, and almost always some starlings. It is how we ring them, bring them in for study, assess their overall genetic fitness. Generally they are none the worse for having been caught up in the nets overnight.

I collect ten adult specimens, let the others go, and take the ten back to the hut.

A firm in Germany has made the digital polarising masks for us. They are elegant little contraptions, similar in design to the hoods fitted around captive birds of prey. These are smaller, though, optimised to be worn by starlings, and to offer no significant resistance to normal flight. Each hood is actually a marvel of miniature electronic engineering. Bulging out from either side are two glassy hemispheres.

In its neutral mode, the bird has an unrestricted view of its surroundings. Each hemisphere, though, is divided into digitally-controlled facets. These can be selectively darkened via wireless computer signals, effectively blocking out an area of the starling’s vision.

The consequence of this – the point of the masks – is that we can control the birds’ collision-avoidance response remotely. By making a given bird think it is about to be struck by another bird, we can cause it to fly in any direction we choose.

Again, it is asking too much of human reflexes to control a bird at such a level. But the computers can do it elegantly and repeatably. Each of our ten hooded starlings then becomes a remote-controlled agent, under our direct operation. Like the robot sparrowhawk, we can steer our agents into the murmuration. The distinction is that the hooded starlings do not trigger a threat response from their seven neighbours; they are absorbed into the whole, accepted and assimilated.

But they can influence the other birds. And by careful control of our hooded agents, we can initiate global changes in the entire murmuration. We can instigate domains, break them up, make them coalesce. Anything that happens under the influence of natural factors, we can now bring about at our will.

By we, of course, I mean I.

Old habits die hard. Science is always done in the ‘we’, even when the work is borne on a single pair of shoulders. But frankly, I am starting to doubt the commitment of my fellow researchers. There is always a division of labour in any collaborative enterprise, and sometimes that division can seem unfair. If the brunt of the work is my responsibility, though, I fail to see why I should not receive the lion’s share of the credit.