Columns

I doubt that anyone has captured the internal dynamics of a flock of birds quite as well as Richard Wilbur, in “An Event”: 

As if a cast of grain leapt back to the hand,  

A landscapeful of small black birds, intent  

On the far south, convene at some command  

At once in the middle of the air, at once are gone  

With headlong and unanimous consent  

From the pale trees and fields they settled on.  

What is an individual thing? They roll  

Like a drunken fingerprint across the sky!  

Or so I give their image to my soul  

Until, as if refusing to be caught  

In any singular vision of my eye  

Or in the nets and cages of my thought,  

They tower up, shatter, and madden space  

With their divergences, are each alone  

Swallowed from sight, and leave me in this place  

Shaping these images to make them stay:  

Meanwhile, in some formation of their own,  

They fly me still, and steal my thoughts away.  

Delighted with myself and with the birds,  

I set them down and give them leave to be.  

It is by words and the defeat of words,  

Down sudden vistas of the vain attempt,  

That for a flying moment one may see  

By what cross-purposes the world is dreamt. 

“Drunken fingerprint” is good. I’ve seen several kinds of birds go through these mass aerial evolutions: New World blackbirds, dunlin and other shorebirds (especially when pursued by a raptor), and, on film, quelea finches in Africa and dickcissels on their wintering grounds in South America. Absent evidence to the contrary, though, I suspect Wilbur’s “small black birds” were European starlings. 

The flight maneuvers of birds have fascinated lay and scientific observers alike for centuries. How do they do it? How do they coordinate their moves so as to avoid colliding with each other? Is there a lead bird providing cues too subtle for human observers to recognize? Is telepathy involved? The otherwise sane British ornithologist Edmund Selous, writing in 1931, leaned toward that explanation: “It is transfused thought, thought transference—collective thinking practically. What else can it be?” 

The why of flocking is more intuitive than the how. Being part of a group reduces a bird’s chance of falling victim to a predator. A location in the group’s center is even better. But everyone wants to be in the center, so the shape of the flock continually morphs as individuals shift. 

About fifty years ago, a Russian biologist named Dimitri Radakov found that every fish in a school coordinates its movements with those of its nearest neighbors. Others suspected that bird flocks were using a similar algorithm. Wayne Potts, studying dunlins on Puget Sound in the 1970s, proposed that the shorebirds kept an eye on more distant individuals, since a wave of movement could sweep through a flock three times faster than the nearest-neighbor rule could account for. He called this the “chorus line hypothesis.” 

Later researchers, mostly in Rome (which has an overabundance of starlings), have used high-speed stereoscopic photography and software developed for molecular-level materials analysis to map the three-dimensional structure of flocks in motion. According to physicist Andrea Cavagna, each individual starling appears to track six or seven neighbors. More data than that might overload the birds’ brains. I recall a report of similar attention parameters among chorus-singing frogs. 

Most recently, Charlotte Hemelrijk at the University of Groningen in the Netherlands has used the Rome group’s computer models to simulate how a starling flock circles above a roost. The ground rules: birds are attracted to other birds, they move in the same direction, and they try to avoid collisions. Hemelrijk and her colleagues found that the dynamics were different than in a school of fish. Fish on the outside of a turning school accelerate while those on the inside slow down, so each fish keeps its original spot and the school maintains an oblong shape. Starlings, however, turn individually, which accounts for the fantastic toroid formations of the flock as a whole. Flock shape varies most when more birds are involved, fewer of them interact, and individuals roll into the turn. The models didn’t account for variables like wind, air turbulence, or predator evasion, but the research, just published online , is a significant advance in understanding flock behavior. Caution: contains equations. 

Scientists with an interest in the emergent properties of systems—the way an ant colony can perform complex functions based on a set of simple decision rules--are paying attention to the starling studies. Some are looking for analogies with human voting or consumer behavior. The pestiferous European starling (introduced to North America by a badly misguided Shakespeare buff) may yet turn out to have its own contribution to make, for better or worse.