"It kept saying it would navigate us a road," said a Japanese tourist in Australia who, while attempting to reach North Stradbroke Island, drove into the Pacific Ocean. A man in West Yorkshire, England, who took his BMW off-road and nearly over a cliff, told the authorities that his GPS "kept insisting the path was a road". In perhaps the most infamous incident, a woman in Belgium asked the GPS to take her to a destination less than two hours away. Two days later, she turned up in Croatia.

These episodes naturally inspire incredulity, if not outright mockery. After a couple of Swedes mistakenly followed their GPS to the city of Carpi (when they meant to visit Capri), an Italian tourism official dryly noted to the BBC that "Capri is an island. They did not even wonder why they didn't cross any bridge or take any boat". And an Upper West Side blogger's account of the man who interpreted "turn here" to mean onto a stairway in Riverside Park was headlined: GPS, Brain Fail Driver.

But some have tragic endings - like the couple who ignored "Road Closed" signs and plunged off a bridge in Indiana last year. Disastrous incidents involving drivers following disused roads and disappearing into remote areas of Death Valley in California became so common that park rangers gave them a name: "Death by GPS." Last October, a tourist was shot to death in Brazil after GPS led her and her husband down the wrong street and into a notorious drug area.

If we're being honest, it's not that hard to imagine doing something similar ourselves. Most of us use GPS as a crutch while driving through unfamiliar terrain, tuning out and letting that soothing voice do the dirty work of navigating. Since the explosive rise of in-car navigation systems around 10 years ago, several studies have demonstrated empirically what we already know instinctively.

Cornell researchers who analysed the behaviour of drivers using GPS found these drivers to be "detached" from the "environments that surround them". Their conclusion: "GPS eliminated much of the need to pay attention."

As a driving tool, GPS is not so much a new technology as it is an apotheosis. For almost as long as automobiles have existed, people have tried to develop auto-navigation technologies. In the early 20th century, products like the Jones Live-Map Metre and the Chadwick Road Guide used complex mechanical systems connected to a vehicle's wheels or odometer to provide specialised directions.

In the 1960s and 1970s, Japan and the United States experimented with networks of beacons attached to centralised computers that let drivers transmit their route and receive route information.

We seem driven (so to speak) to transform cars, conveyances that show us the world, into machines that also see the world for us.

A consequence is a possible diminution of our "cognitive map", a term introduced in 1948 by the University of California, Berkeley psychologist Edward Tolman. In a groundbreaking paper, Dr Tolman analysed several laboratory experiments involving rats and mazes. He argued that rats had the ability to develop not only cognitive "strip maps" - simple conceptions of the spatial relationship between two points - but also more comprehensive cognitive maps that encompassed the entire maze.

Could society's embrace of GPS be eroding our cognitive maps? For Dr Julia Frankenstein, a psychologist at the University of Freiburg's Centre for Cognitive Science, the danger of GPS is that "we are not forced to remember or process the information - as it is permanently 'at hand',' we need not think or decide for ourselves". She has written that we "see the way from A to Z, but we don't see the landmarks along the way". In this sense, "developing a cognitive map from this reduced information is a bit like trying to get an entire musical piece from a few notes". GPS abets a strip-map level of orientation with the world.

There is evidence that one's cognitive map can deteriorate. A widely reported study published in 2006 demonstrated that the brains of London taxi drivers have larger than average amounts of grey matter in the area responsible for complex spatial relations. Brain scans of retired taxi drivers suggested that the volume of grey matter in those areas also decreases when that part of the brain is no longer being used as frequently.

"I think it's possible that if you went to someone doing a lot of active navigation, but just relying on GPS," Dr Hugo Spiers, one of the authors of the taxi study, hypothesised to me, "you'd actually get a reduction in that area."

For Dr Tolman, the cognitive map was a fluid metaphor with myriad applications. He identified with his rats. Like them, a scientist runs the maze, turning strip maps into comprehensive maps - increasingly accurate models of the "great God-given maze which is our human world", as he put it. The countless examples of "displaced aggression" he saw in that maze - "the poor Southern whites, who take it out on the Negros", "we psychologists who criticise all other departments", "Americans who criticise the Russians and Russians who criticize us" - were all, to some degree, examples of strip-map comprehension, a blinkered view that failed to comprehend the big picture.

"What in the name of Heaven and Psychology can we do about it?" he wrote. "My only answer is to preach again the virtues of reason - of, that is, broad cognitive maps."

GPS is just one more way for us to strip-map the world, receding into our automotive cocoons as we run the maze. Maybe we should be grateful when, now and then, they give us a broader view of it - even if by accident. Mr Santillan's response to his misbegotten journey was the right one. When he reached Siglufjordur, he got out of the car, marvelled at the scenery and decided to stay awhile. Reykjavik could wait.

NEW YORK TIMES

• Greg Milner is author of the forthcoming book Pinpoint: How GPS Is Changing Technology, Culture And Our Minds.