快猫短视频

ON HER morning drive to work, Debbie drinks her coffee, eats her breakfast, checks her email and chats with her mum, often all at once. She hates wasting time. This isn鈥檛 just a female talent: Alun can conduct a job interview on top of answering emails and surfing the web. These people are examples of an elite species of human: multitaskers. They can juggle more activities in 5 minutes than our ancestors did in a day. Or so they like to think.

The debate about multitasking has lately come to a head worldwide over the cellphone-while-driving issue. If Alun hires a dud, the consequences are unlikely to be fatal. But if Debbie fails to see something in the road or is too slow to react to it, they could be. Examples abound in the real world: accidents caused by drivers talking on their phones killed 13 people and seriously injured 400 in the UK in 2005. That same year there were an estimated 2600 deaths and 330,000 injuries due to cellphone use while driving in the US, according to the Human Factors and Ergonomics Society, based in Santa Monica, California.

Talking on the phone while driving isn鈥檛 the only situation where we鈥檙e worse at multitasking than we might like to think we are. New studies have identified a bottleneck in our brains that some say means we are fundamentally incapable of true multitasking. If experimental findings reflect real-world performance, people who think they are multitasking are probably just underperforming in all 鈥� or at best, all but one 鈥� of their parallel pursuits. Practice might improve your performance, but you will never be as good as when focusing on one task at a time.

The problem, according to Ren茅 Marois, a psychologist at Vanderbilt University in Nashville, Tennessee, is that there鈥檚 a sticking point in the brain. To demonstrate this, Marois devised an experiment to locate it. Volunteers watch a screen and when a particular image appears, a red circle, say, they have to press a key with their index finger. Different coloured circles require presses from different fingers. Typical response time is about half a second, and the volunteers quickly reach their peak performance. Then they learn to listen to different recordings and respond by making a specific sound. For instance, when they hear a bird chirp, they have to say 鈥渂a鈥�; an electronic sound should elicit a 鈥渒o鈥�, and so on. Again, no problem. A normal person can do that in about half a second, with almost no effort.

The trouble comes when Marois shows the volunteers an image, then almost immediately plays them a sound. Now they鈥檙e flummoxed. 鈥淚f you show an image and play a sound at the same time, one task is postponed,鈥� he says. In fact, if the second task is introduced within the half-second or so it takes to process and react to the first, it will simply be delayed until the first one is done. The largest dual-task delays occur when the two tasks are presented simultaneously; delays progressively shorten as the interval between presenting the tasks lengthens (See Diagram).

There are at least three points where we seem to get stuck, says Marois. The first is in simply identifying what we鈥檙e looking at. This can take a few tenths of a second, during which time we are not able to see and recognise a second item. This limitation is known as the 鈥渁ttentional blink鈥�: experiments have shown that if you鈥檙e watching out for a particular event and a second one shows up unexpectedly any time within this crucial window of concentration, it may register in your visual cortex but you will be unable to act upon it. Interestingly, if you don鈥檛 expect the first event, you have no trouble responding to the second. What exactly causes the attentional blink is still a matter for debate.

A second limitation is in our short-term visual memory. It鈥檚 estimated that we can keep track of about four items at a time, fewer if they are complex. This capacity shortage is thought to explain, in part, our astonishing inability to detect even huge changes in scenes that are otherwise identical, so-called 鈥渃hange blindness鈥�. Show people pairs of near-identical photos 鈥� say, aircraft engines in one picture have disappeared in the other 鈥� and they will fail to spot the differences (if you don鈥檛 believe it, check out the clips at ). Here again, though, there is disagreement about what the essential limiting factor really is. Does it come down to a dearth of storage capacity, or is it about how much attention a viewer is paying?

A third limitation is that choosing a response to a stimulus 鈥� braking when you see a child in the road, for instance, or replying when your mother tells you over the phone that she鈥檚 thinking of leaving your dad 鈥� also takes brainpower. Selecting a response to one of these things will delay by some tenths of a second your ability to respond to the other. This is called the 鈥渞esponse selection bottleneck鈥� theory, first proposed in 1952.

Last December, Marois and his colleagues published a paper arguing that this bottleneck is in fact created in two different areas of the brain: one in the posterior lateral prefrontal cortex and another in the superior medial frontal cortex (Neuron, vol 52, p 1109). They found this by scanning people鈥檚 brains with functional MRI while the subjects struggled to choose among eight possible responses to each of two closely timed tasks. They discovered that these brain areas are not tied to any particular sense but are generally involved in selecting responses, and they seemed to queue these responses when presented with multiple tasks concurrently.

Bottleneck? What bottleneck?

But David Meyer, a psychologist at the University of Michigan, Ann Arbor, doesn鈥檛 buy the bottleneck idea. He thinks dual-task interference is just evidence of a strategy used by the brain to prioritise multiple activities. Meyer is known as something of an optimist by his peers. He has written papers with titles like 鈥淰irtually perfect time-sharing in dual-task performance: Uncorking the central cognitive bottleneck鈥� (Psychological Science, vol 12, p 101). His experiments have shown that with enough practice 鈥� at least 2000 tries 鈥� some people can execute two tasks simultaneously as competently as if they were doing them one after the other. He suggests that there is a central cognitive processor that coordinates all this and, what鈥檚 more, he thinks it uses discretion: sometimes it chooses to delay one task while completing another.

Even with practice, not all people manage to achieve this harmonious time-share, however. Meyer argues that individual differences come down to variations in the character of the processor 鈥� some brains are just more 鈥渃autious鈥�, some more 鈥渄aring鈥�. And despite urban legend, there are no noticeable differences between men and women. So, according to him, it鈥檚 not a central bottleneck that causes dual-task interference, but rather 鈥渁daptive executive control鈥�, which 鈥渟chedules task processes appropriately to obey instructions about their relative priorities and serial order鈥�.

Marois agrees that practice can sometimes erase interference effects. He has found that with just 1 hour of practice each day for two weeks, volunteers show a huge improvement at managing both his tasks at once. Where he disagrees with Meyer is in what the brain is doing to achieve this. Marois speculates that practice might give us the chance to find less congested circuits to execute a task 鈥� rather like finding trusty back streets to avoid heavy traffic on main roads 鈥� effectively making our response to the task subconscious. After all, there are plenty of examples of subconscious multitasking that most of us routinely manage: walking and talking, eating and reading, watching TV and folding the laundry.

But while some dual tasks benefit from practice, others simply do not. 鈥淐ertain kinds of tasks are really hard to do two at once,鈥� says Pierre Jolicoeur at the University of Montreal, Canada, who also studies multitasking. Dual tasks involving a visual stimulus and skeletal-motor response (which he dubs 鈥渋n the eye and out the hand鈥�) and an auditory stimulus with a verbal response (鈥渋n the ear and out the mouth鈥�) do seem to be amenable to practice, he says. Jolicoeur has found that with enough training such tasks can be performed as well together as apart. He speculates that the brain connections that they use may be somehow special, because we learn to speak by hearing and learn to move by looking. But pair visual input with a verbal response, or sound to motor, and there鈥檚 no dramatic improvement. 鈥淚t looks like no amount of practice will allow you to combine these,鈥� he says.

For research purposes, these experiments have to be kept simple. Real-world multitasking poses much greater challenges. Even the upbeat Meyer is sceptical about how a lot of us live our lives. Instant-messaging and trying to do your homework? 鈥淚t can鈥檛 be done,鈥� he says. Conducting a job interview while answering emails? 鈥淭here鈥檚 no way you wind up being as good.鈥� Needless to say, there appear to be no researchers in the area of multitasking who believe that you can safely drive a car and carry on a phone conversation. In fact, last year David Strayer at the University of Utah in Salt Lake City reported that people using cellphones drive no better than drunks (Human Factors, vol 48, p 381). In another study, Strayer found that using a hands-free kit did not improve a driver鈥檚 response time. He concluded that what distracts a driver so badly is the very act of talking to someone who isn鈥檛 present in the car and therefore is unaware of the hazards facing the driver.

鈥淣o researchers believe it鈥檚 safe to drive a car and carry on a phone conversation鈥�

It probably comes as no surprise that, generally speaking, we get worse at multitasking as we age. According to Art Kramer at the University of Illinois at Urbana-Champaign, who studies how ageing affects our cognitive abilities, we peak in our 20s. Though the decline is slow through our 30s and on into our 50s, it is there; and after 55, it becomes more precipitous. In one study, he and his colleagues had both young and old participants do a simulated driving task while carrying on a conversation. He found that while young drivers tended to miss background changes, older drivers failed to notice things that were highly relevant. Likewise, older subjects had more trouble paying attention to the more important parts of a scene than young drivers.

It鈥檚 not all bad news for over-55s, though. Kramer also found that older people can benefit from practice. Not only did they learn to perform better, brain scans showed that underlying that improvement was a change in the way their brains become active.

While it鈥檚 clear that practice can often make a difference, especially as we age, the basic facts remain sobering. 鈥淲e have this impression of an almighty complex brain,鈥� says Marois, 鈥渁nd yet we have very humbling and crippling limits.鈥� For most of our history, we probably never needed to do more than one thing at a time, he says, and so we haven鈥檛 evolved to be able to. Perhaps we will in future, though. We might yet look back one day on people like Debbie and Alun as ancestors of a new breed of true multitaskers.