Guilty Robots, Happy Dogs: The Question of Alien Minds
By David McFarland
OUP £16.99, 256 pages
FT bookshop price: £13.59

What is it like to be a dog? How does it feel to chase your own tail, or to think sniffing someone’s behind an appropriate form of greeting?

In this land of dog lovers, we may believe we have a pretty good idea of what is going on in Rover’s head. But according to David McFarland, we don’t. The eminent zoologist claims in Guilty Robots, Happy Dogs that not only will the urge to bury bones in the garden forever remain foreign to us, but we may be mistaken in believing that dogs are conscious at all. We will never be able to get into Rover’s mind, for the simple reason that he doesn’t have one.

To anyone who has a dog, the thought that those funny things their pet does with grandma’s slippers are merely the reflexes of a mindless automaton will seem to refute the wide-eyed, panting obvious.

But McFarland, who is also an expert in artificial intelligence, claims that all animal behaviour can be explained by simple formulae that do not presuppose anything so mysterious as a mind.

What is more, we are now able to build robots whose behaviour is just as complex as Rover’s – but these machines have nothing in their heads but wires and transistors. So, he concludes, if we are not inclined to ascribe finer feelings to our hardware, then we shouldn’t to our pets either.

The gap between flesh-and-bone and nuts-and-bolts is clearly narrowing: the first predatory, carnivorous robot, for example, is already among us. The SlugBot hunts only slugs – and isn’t even very good at that, but these are early days. As they evolve, keeping our gardens slug-free will be only one of many chores for which we will rely on robots. But to be truly useful, McFarland predicts, they will need to become more independent – we don’t want to have to worry about recharging their batteries, or explaining to them the difference between a slug and the garden hose. In other words, robots will have to become more like animals.

This confronts us with a crucial question: should we be generous and say that both robots and animals are thinking, feeling beings; or should we be stingy, and say that neither are. McFarland is inclined to be stingy. He is therefore following a century-old tradition of zoology that, terrified of appearing unscientific, attempts to explain animal behaviour without crediting humbler creatures with “human” feelings. Disciples of this tradition are highly adept at describing animals in purely mechanical terms – and at explaining away the vast body of evidence that suggests they feel pain and pleasure just like we do.

It is unsurprising that after decades of treating animals like mindless machines, we should start to see machines as more like animals. This approach may well help us to build better robots, as we emulate the tricks of a biology that evolved over millions of years. But it also legitimises our treatment of fellow mammals as mere meat and milk factories. This book’s prognoses about robotics may be cutting-edge, but its view of Rover is pure dogma.