Simone Harvey studies in front of the Valley Life Sciences Building at the University of California at Berkeley, 12 May 2014. Photo by Noah Berger/Reuters


Getting smarter

Brain-training games won’t boost your IQ, but a host of strategies can improve your cognitive abilities one piece at a time

by Jeffrey M Zacks + BIO

Simone Harvey studies in front of the Valley Life Sciences Building at the University of California at Berkeley, 12 May 2014. Photo by Noah Berger/Reuters

Is it just me, or is everybody out there looking for a quick fix? There is something highly compelling about the idea that there is a secret switch we can flip to become suddenly smarter, to reveal cognitive abilities hidden inside each of us. It is a notion that certainly has commercial appeal. Over just seven years, the games-maker Lumosity rocketed from zero to 50 million users, promising rapid improvements in general intelligence by playing brain-training video games for just a few weeks. (Lumosity recently settled with the United States Federal Trade Commission for making unsupported claims that its product was scientifically validated.) ‘Memory health’ nutritional supplements have sales of more than $1.5 billion, and ‘smart drugs’ – pills to enhance cognitive performance – have become prevalent on college campuses. Purveyors of products based on subliminal messages promise to teach us foreign languages and cure our addictions while we sleep. And makers of headgear that attaches electrodes to our scalps promise to rev up our brains to improve gaming performance and other cognitive activities.

These are global trends but, living in the US, it seems to me that there is something particularly American to them. We have a long tradition of positivism, progressivism and self-improvement. Some of this is great. Thomas Edison was convinced that human existence could be dramatically improved with rapid technological innovation, and it is hard to dispute the transformative value of electric light and recorded sound. Henry Wallace looked at the variable and unpredictable yields of Iowa corn farmers and was convinced that scientific breeding and hybridisation could do better than allowing judges to select the most attractive ears at corn shows. His hybrid corn transformed grain farming and launched him on a political career that landed him as Franklin D Roosevelt’s vice president. (However, Wallace’s self-improvement enthusiasm also extended to investigations of psychical communication with the dead, which probably contributed to his getting booted off Roosevelt’s third re-election campaign in favour of Harry S Truman.)

Some of this progressive self-improvement tradition is essential, and some is charming in a kooky way. But part of the drive to engineer a quick fix for what ails us is alarming and dangerous. Americans have endured generations of rapid weight-loss schemes that don’t work and are often dangerous to health – especially diet pills from amphetamines to fen-phen.

As the pace of life continues to accelerate, there is increasing pressure for a quick fix to boost our cognitive capacities – for study, for work and for recreation. So, is there anything out there that works?

Sadly, most of the rapid cognitive enhancers currently being peddled are not very effective. Let’s start with brain training. None of the commercial brain-training software being marketed is backed up by strong scientific evidence. Lumosity started off claiming that its product was supported by scientific research, but a number of us in the cognitive neuroscience field were struck that the research section of the company’s website didn’t seem to have any peer-reviewed studies demonstrating this claim. Making such a claim without evidence is what got the company sued.

Lumosity, Posit Science and other brain-training companies differ in many of their particulars but share some common features. First, they ‘gamify’ training, so that it is more appealing. Second, they are adaptive, which means that, as you get better, the game gets harder. Both of these are solid principles of cognitive training. Third, many (but not all) of these systems focus on training an ability called working memory, which is the skill to keep multiple bits of information ready to hand, and juggle some of them while using others.

Working memory is an attractive target for cognitive training because measures of its capacity – how many bits of information you can juggle – correlate strongly with measures of problem-solving and reasoning, and because working-memory tasks are easy to program into games. Some systems also train attentional control, which is the ability to pick out task-relevant information from among distracting information. Like working memory, training attentional control makes theoretical sense and can be incorporated into interactive games. So, the thinking behind these interventions is totally reasonable.

The problem is that experiments to test the types of techniques used by commercial products to boost working memory and attention have not yielded significant results. The most directly relevant was a large study carried out by researchers at the British Medical Research Council’s Cognition and Brain Sciences Unit in Cambridge. They recruited 11,430 participants through a TV show called Bang Goes the Theory. One-third of them were randomly assigned to practise a broad set of tasks focusing on memory and attention, similar to those trained by the commercial products. One task required participants to quickly click on symbols appearing on the screen, but only if they matched a set of target symbols. Another task had them watching pieces of luggage going through an airport X‑ray machine. From time to time, the conveyor belt would stop and the participants had to report how many bags were on the machine.

people got better at the tasks they practised but this did not transfer at all to other measures of cognition

Another third of the participants were trained on a set of reasoning and planning tasks: one of these required them to find which shape did not belong in a set, another required them to slide tiles around on a grid to place them in numerical order. The remaining third were assigned to a control group and received no training. Everyone completed a separate set of IQ-like benchmarking tests before and after the training period to measure a range of cognitive abilities. For example, in one test people had to verify whether a rule (eg ‘the circle is not smaller than the square’) applied to a picture, and in another they had to repeat back increasingly long strings of digits. Both of the training groups improved dramatically on the tasks they practised – but showed no more gain on the benchmarking tests than the control group.

Another large experiment in the US called the ACTIVE study (for Advanced Cognitive Training for Independent and Vital Elderly) aimed to figure out what might stave off the cognitive decline associated with ageing. In this study, the researchers recruited 2,832 people and randomly assigned them to practice tasks of memory, reasoning or cognitive speed, or to a control group that received no special training. Again, people got better at the tasks they practised but this did not transfer at all to other measures of cognition.

For instance, people in the memory group were taught strategies for memorising world lists. Sure enough, they got better at memorising words, but not at reasoning out the patterns in series of numbers or travel schedules. Conversely, people in the reasoning group were taught strategies for finding patterns, and they got better at that task but did not improve at memorising words. (On a follow-up test with a five-year delay, there was a hint of an effect that those participants who were in the reasoning training group were helped somewhat with everyday tasks, so perhaps all hope is not lost. But this is definitely not the magnitude of effect that one would expect to be meaningful in boosting intelligence, and it was not apparent in the short-term follow-up.)

Cognitive-enhancing drugs actually have a bit more evidentiary support than brain-training. It has been known for a while that caffeine and nicotine allow people to respond faster, stay more vigilant and score higher on some kinds of standardised tests. Prescription stimulants such as Adderall (amphetamine) and Ritalin (methylphenidate), which have been shown to be effective in treating attention-deficit disorder, also improve cognitive function in healthy people. These effects are not huge, and they are not always consistent, but they are real.

What is less well-known is the trade-off. All of these drugs have pretty short half-lives. What happens when you come down? An ideal cognitive supplement would boost performance and then return you to your baseline. However, a likely possibility is that stimulant drugs produce a rebound effect, reducing cognitive function once they wear off. This might still be an acceptable trade-off: if you know when you have to be ‘on’, it might be perfectly fine to be a little duller for a while afterwards. This trade-off is particularly acute for habit-forming stimulants such as nicotine. If you don’t have much control over when you use the drug, then you have less ability to time its effects to be helpful.

More concerning to me is the use of stimulants in developing brains. One of the important things we have learned about brain development over the past 40 years is that the brain undergoes more protracted maturation than any other body system. The prefrontal cortex is the slowest part of the brain to mature, undergoing substantial change well into the 20s. And the prefrontal cortex is absolutely critical for higher brain function – for organising behaviour in extended tasks, for exercising self-control, and for planning actions in the future. The effects of cognitive-enhancing drugs on brain development is still imperfectly understood, but for all these drugs there is a real risk that abuse can harm long-term brain development.

Subliminal training programs claim to change your brain by inserting information without your awareness. For example, many of these products claim to have a verbal message recorded at a very low level along with a much louder recording of nature sounds or relaxing music. The idea is that words presented too quietly to be consciously detected are able to sneak past some cognitive defences and reprogram the brain. Better yet, purveyors of subliminal training systems claim that because you don’t need to be aware of the training messages, they work even while you’re sleeping.

Suppose you want to kick cigarettes or lose weight. Wouldn’t it be great if you could just drive around with a tape that played soothing music or sounds with subliminal messages buried in them that could change your behaviour? Or suppose you wanted to learn Portuguese for that trip to Brazil? Why spend hundreds of hours drilling vocabulary and grammar if you can play the language while you sleep and learn it? The trouble is that this idea makes no sense in terms of what we know about the mind and brain, and there is very good evidence that it simply doesn’t work. People who listen to weight-loss tapes lose no more weight than those who listen to tapes with no messages or no tapes at all, and people who try to learn words in their sleep don’t learn.

Electrical stimulation – at the right dosage with the right technique in the right location – can produce modest enhancement of some kinds of learning and motor function

What about electrical stimulation? It’s certainly the most sci-fi cognitive enhancer out there. Electrical stimulation of the brain has actually been around for a long time. In the 19th century, physiologists began studying the nervous system in non-human animals by inserting electrodes directly into the brain. This technique was picked up by neurosurgeons, and forms an important component of surgical planning for brain surgery to this day. But direct stimulation of the brain is highly invasive and therefore rare. More practical for human use are techniques that use external stimulators to induce electrical changes in the brain.

The one that has received the most attention for cognitive enhancement is called transcranial direct current stimulation (tDCS). It works by applying a mild electrical current to electrodes placed on the scalp. But tDCS does not directly cause particular neurons to fire; it shifts the electrical potential in a part of the brain, which makes it more or less likely that neurons there will fire when stimulated. Scientific attention to tDCS has been building over the past 20 years – and has been picked up by hobbyists and entrepreneurs. DIY tDCS explorers have been building their own electrical stimulators, and at least one company has commercialised a tDCS product aimed at gamers.

Do not try this at home. Really.

As it is used in the lab and in the clinic, tDCS is very safe – a recent review of data from 3,836 subjects found that adverse effects are mainly sensations of itching, tingling and the like, with a smaller number of people experiencing headaches. My lab at Washington University in St Louis has started working with the technique and I have been a subject; I find it easy to work with and totally innocuous to experience, and I encourage people to participate in tDCS research if they have the chance. (My guess is that the headaches are mostly due to the headband needed to hold the electrodes in place; headaches occur as often in controls as in people undergoing stimulation.)

However, to make tDCS safe requires equipment with proper safety controls, and some basic knowledge about neuroanatomy and physiology. Unfortunately, enthusiasts in the DIY tDCS have suffered burns and headaches from stimulation gone wrong. Worse yet, there is a real risk that applying electrical stimulation improperly could cause permanent injury. There are intriguing findings that tDCS – applied at the right dosage with the right technique in the right location – can produce modest enhancement of some kinds of learning and motor function.

For example, in one study Lauren Richmond, now a postdoctoral fellow in my laboratory, and colleagues trained people to hold increasingly long sequences of letters or spatial locations in mind. During training, half of the participants received tDCS over their left prefrontal cortex, a part of the brain known to be important for this kind of memory. The other participants were wired up during training, but the stimulator was set to shut itself off so that they received a trivial dose of current. Everyone improved on the tasks over a fortnight with 10 training sessions, but those who got the stimulation improved faster. However, at this point the effects are too small and variable to come anywhere close to counterbalancing the risks of trying this on yourself outside of a reputable clinic or lab. Also, we don’t yet know if a tDCS-induced gain in one cognitive ability comes at a cost to some other ability.

So, does all this mean that there is no hope for cognitive enhancement? My view is that there is actually lots of hope, as long as we are not unrealistic about what we can achieve. Right now, I think the best bet is cognitive training – but not the kind we’re being sold. Firms such as Lumosity, Posit Science and Jungle Memory are trying to do something that learning scientists call ‘far transfer’ – training on one task, and transferring that gain to a very different task.

For example, suppose I want to get better at keeping track of baseball statistics or learning the names of new acquaintances. The far transfer approach goes like this: I sign up for a brain-training program, boost my working memory capacity and attentional control, and this improves my ability to track stats or remember new people. The big advantage of this approach is that, if successful, it gives broad benefits. If I really could increase my general cognitive ability, it would help me out in many facets of my life. The trouble is that far transfer is really hard to come by. That is what is shown by the data I described before.

What is the alternative? Scale down your expectations and aim for near transfer. If you want to get good at baseball scorekeeping, practise that. If you meet lots of people and need to learn their names, practise that. For both scorekeeping and name-learning – and many other cognitive tasks – there are well-known techniques that have been shown to be effective.

For example, there are two techniques that are very helpful for learning new associations between faces and names. First, practise the association before you lose it. When you meet someone, repeat his name to yourself and use it in conversation if possible. After an encounter, review the people you met and rehearse their names. Second, come up with a meaningful association with the name. It can be silly – silly actual works better. For example, if you are out for pizza and meet a new guy named Pete, let his virtual nickname be ‘pizza Pete’. This totally works. I teach a class of two to three dozen students each fall, and I try to learn everyone’s name before the first day. This doesn’t come naturally to me, so I print out the roster with their pictures and practise a bunch of times naming each picture, rehearsing my virtual nicknames each time.

For training to be effective, you have to stick with it, and it is much easier to do so if the training is fun, and if it embeds regular rewards along the way

For memorising numbers such as baseball statistics, the rise of memory competitions has pushed dramatic improvements in training techniques. The British memory champ Ben Pridmore is a great example. He was the first person to successfully memorise a deck of cards in under 30 seconds, and has won all kinds of championships. One competitive event requires that you spend 15 minutes listening to a string of digits and then write them all down. In 2008, Pridmore managed 153. (The current record, held by Lance Tschirhart of the United States, is 456.)

Pridmore wrote How to Be Clever (2011), which lays out exactly the techniques and practice regimens he adopted to achieve this performance. The achievements of such memory athletes show how powerful near transfer can be. But getting good at one of these tasks does not make you smart in general. Pridmore tells the reader this right on page three: ‘Intelligence I can’t do anything about.’

There is one piece of advice, however, that the brain-training products have exactly right: gamification. For training to be effective, you have to stick with it, and it is much easier to do so if the training is fun, and if it embeds regular rewards along the way. Psychology and neuroscience are learning a lot about how to design training systems that people want to use, and that produce efficient results.

Does this mean that nothing can make us smarter in general? Not at all. Far transfer in training, or safe and effective smart drugs or brain stimulation, could be right around the corner. None of these are things that shouldn’t work in principle. (I wouldn’t bet on subliminal messages working out, though.) It could be that one of these technologies – or a new one – will break through and really make us smarter. In fact, in the past century our species experienced a massive increase in smarts that happened very fast in evolutionary terms – too fast to have been due to genetic changes. In 1984, the psychologist James Flynn was studying intelligence testing; he reported that, from 1932 to 1978, IQ scores in the US had increased 13.8 points. Around the world, he found increases ranging from five to 25 points.

These are massive increases. Take the US number as a reasonable overall estimate. An increase of 13.8 percentage points means that a score which would have placed you squarely in the middle of the distribution in 1932 would now land you below the 20th percentile, with more than 80 per cent of people scoring better than you. Nobody knows exactly what drove these increases, but a good guess is that it was largely due to basic improvements in nutrition and education, which allowed brains to better develop to their full capacity.

If you are reading this, you probably benefited from the nutritional and educational improvements of the 20th century, so you’ve already got your brain boost. But there is still something you can do to improve and maintain general cognitive ability: get some exercise. Specifically, cardiovascular exercise such as walking, running, biking or swimming. Improving cardiovascular fitness has been shown causally to improve cognitive function, and this has been traced to the growth of new neurons in the brain and to improvements in the function of existing neurons.

we’re already smarter than our grandparents. And if we live an active lifestyle, we can get smarter and stay smart throughout our lives

For example, in one study older adults were asked to make quick visual judgments about alphanumeric characters while avoiding being distracted by nearby characters. Then, half of them were assigned to improve their cardiovascular fitness by taking increasingly intensive walks, maxing out at 40-45 minutes of brisk walking. The other half were assigned to do stretching and toning exercises. After training, the initial task was repeated. The group who walked improved at avoiding being distracted by the extraneous, and showed changes in brain activity while doing the task. The stretching and toning group did not change. A particular focus of this research has been on ageing. The results suggest that exercise can help older adults maintain and improve cognitive function. Given that exercise has lots of other benefits, what’s not to like?

Now, neither good rearing nor cardiovascular fitness is a quick fix that will make us cognitive superheroes. But I think the story is mostly good news. First, if there is something you really need to get good at, chances are that there is a training technique that will be effective. Second, we’re already smarter than our grandparents (on average). And finally, if we live an active lifestyle, we can get smarter and stay smart throughout our lives.

Sometimes, the function of science is to overturn common sense, achieving a deeper understanding of nature. Other times, common sense turns out to be right. When it comes to getting smarter, common sense turns out to be on target so far. Six-week mental workouts, smart drugs, learning while you sleep and zapping yourself into genius all sound a little too good to be true – and for now they are. But we can all think better in specific domains if we engage in focused practice, and be smarter, happier and healthier if we take care of ourselves.