Paul Rand: Have you ever wondered why some people seem to have a really great memory while others struggle to remember their friend’s birthday or their neighbor’s name or even where they put their keys?

Edward Awh: So there’s been this fascination ever since people started studying the mind, this fascination in understanding why is it that we can’t think about everything at once? Why is there a limit to the scope of our conscious awareness?

Paul Rand: That’s Edward Awh, Professor of Psychology at the University of Chicago, where he co-directs a neuroscience lab along with his colleague, Ed Vogel. Their research examines how the brain controls focus, memory, and attention.

Edward Awh: It seems like our conscious awareness is actually interestingly quite limited despite our subjective impression that we’re taking in a huge amount of information at any given moment in time.

Paul Rand: Even in today’s world where it feels like we’re overloaded with information from social media to the news, to the text messages on our phones, Awh says there’s only so much information that our brains can actually store.

Edward Awh: And there’s some fascinating work that suggests that if you try to measure the maximum amount of information that people can hold in mind at one time, it turns out to be very similar across people who have excellent performance and people who have not so good performance. And what really distinguishes the high performers from the others is the consistency with which they’re able to load information into memory.

Paul Rand: But being a high performer or having a so called good memory doesn’t actually reflect how intelligent you are or even how studious you are. It might just have to do with one single trait.

Edward Awh: And it’s interesting that this ability is a very stable trait of an individual. If we measure your working memory capacity today, we’re going to find a very similar measurement a year from now or six months from now.

Paul Rand: We all want to be more productive, whether in our personal lives or at work. So is there a life hack that we can use to actually control our attention?

Edward Awh: Use the remote control of your mind to decide on which aspects of a complex environment that you’re going to focus your attention on and thereby control which things are going to occupy your really limited space in working memory.

Paul Rand: From the University of Chicago Podcast Network, welcome to Big Brains, where we explore the groundbreaking research and the discoveries that are changing our world. I’m your host, Paul Rand. Join me as we meet the minds behind the breakthroughs on today’s episode, The Science Behind Our Memory and Attention.

Well, I wonder if we can just start off right at the beginning for you and tell me what your research is focused on and what problems are you trying to understand about the human mind?

Edward Awh: Yes. So we have a cognitive neuroscience laboratory here, and so we’re interested in understanding with brain and behavior how the mind works. And one of our key focuses in our lab is the idea of attention and working memory. We see these as very intertwined ideas with attention referring to an observer’s ability to determine which aspects of a really busy environment they’re actually focusing their mind on. And working memory refers to their ability to maintain that information, even if it’s not present any longer in the environment.

And what we’re trying to understand is the sharp capacity limits that have been seen in working memory. It seems like our conscious awareness is actually interestingly quite limited, despite our subjective impression that we’re taking in a huge amount of information at any given moment in time. And we try to combine those types of information to understand why memory and attention are limited and what the consequences of those limits might be.

Paul Rand: And when you talk about working memory, can you explain how that fits into the other types of memory, please?

Edward Awh: When I refer to working memory, what I mean are the memories or the thoughts that are currently in mind, the ones that you’re currently thinking about and aware of. And we could contrast that with long-term memory, which refers to all of the memories and experiences that you have stored over your lifetime in your brain. And we know that we can retrieve those long-term memories in a fairly amazing and efficient way, but the ones that are actually on stage in your mind, those are the ones that we’re referring to as working memories.

Paul Rand: And when did we start first understanding that the brain actually had limits?

Edward Awh: So there’s been this fascination ever since people started studying the mind, this fascination and understanding why is it that we can’t think about everything at once? Why is there a limit to the scope of our conscious awareness? Now, in modern literature, I would actually point to a really seminal study that was done by Steve Luck and Ed Vogel, where they asked this question about the scope of working memory, but they did so in this really clever way where they used extremely simple stimuli, just colored squares, nothing more. It was the simplest objects that one could conceive of. And they came up with an objective method for just measuring how many simple colors can an individual hold in mind at one time.

So now, there’s no debate about whether each object is an object or whether they’re too complex to take in. They were as simple as possible. And yet, despite the simplicity of the objects, the apparent capacity was only about three or four of these simple colors.

Paul Rand: Got it.

Edward Awh: And this really comes as a surprise to most people whose intuitions are that they can have a much richer and vaster landscape that they’re thinking about at any given moment. And I think that this study, it’s been cited many, many thousands of times because it really captured the field’s imagination to try to understand why the limit is so, so short.

Paul Rand: And once it was understood that there is a limit, did we get around to asking, “Well, why? Why can we only hold such a small amount of information?”

Edward Awh: Yes. I think that this is kind of the million-dollar question, Paul.

Paul Rand: Yeah.

Edward Awh: Why is there this limit? And I think that there’s a long history here and the history of the neuroscience of looking at short-term memory function. About 50 years ago or so in the ‘70s, people started to find neural signals that would actually track the content of the information that an animal was holding in working memory, for example. Like if a monkey was holding a color in short-term memory, they could actually find cells in the inferotemporal cortex that would represent that color during that delay period.

And there was kind of this real excitement about this type of observation, because it seemed like neuroscientists were able to be looking at the memories as they were being held online in this neural activity. And so this really inspired a lot of theories of working memory limits that focus on limits in those brain regions that are representing the information of interest.

So color selective neurons in the visual cortex or face selective cells in inferotemporal cortex, perhaps these are the places where these capacity limits are actually arising. But I will point out that Ed Vogel and I actually have a kind of different perspective on these capacity limits. And instead, we’ve kind of been focused on the idea of a limit in a process that is actually separate, a distinct facet of memory from the maintenance of specific pieces of content per se. And so if you’ll give me a moment, I can try to explain what I’m talking about here.

Paul Rand: Please, go for it.

Edward Awh: So if you think about your everyday experience in life, almost every experience that you have is in a dynamic environment, even just reaching out to pick up a coffee cup. Well, what you’re having to do is track your fingers and your hand and the position of that cup simultaneously as you bring them together and then you lift it to your mouth and take a sip. And what we’re realizing through kind of visual cognition and the philosophy of visual cognition is that this simple act of picking up an object or targeting an object, acting upon an object, even touching an object, requires you to be able to track that object through time and space. In other words, we would call this in more jargony terms, refer to this as context binding. How do you actually take an object that you’re interested in and perceive it within the context of a dynamic event?

And what we think we’ve been able to find is a neural signal that actually tracks the number of items that a person is currently binding to the event that they are perceiving. And so for example, in that simple colored square memory task that I was just mentioning to you, if we vary the number of colors that a person is holding in working memory, we can see a sustained neural signal that will actually tell us how many individual colors are they able to hold in mind at one time. But a really fascinating thing about that neural signal is that it is general and does not seem to care too much about what the content is of the items that are being maintained in memory.

So for example, we in a recent paper found that if you look at the neural activity that tracks the number of colors that a person is holding in memory, we see that a very similar, if not identical, neural signal is generated when people are holding different numbers of sounds in working memory as well.

Paul Rand: Okay. Interesting.

Edward Awh: So even though they’re separate sensory modalities, there are going to be kind of a unified system for tracking the number of items that are currently being attended or held in mind.

Paul Rand: So help me understand this. Some more recently than others have played the memory game. There’s always somebody at the table that blows everybody away and can remember multiple numbers of things. Are they special? Are they exercise their memory in different ways? What constitutes the differences in human performance like this?

Edward Awh: Yes. Well, I think there’s two interesting answers to this question. First, we could talk about the normal variation in human memory ability. So if we measure working memory capacity across a large group of people, we will find a pretty wide distribution of abilities across people. Some people are very good at these memory tasks and other people are not so good. And it’s interesting that this ability is a very stable trait of an individual. If we measure your working memory capacity today, we’re going to find a very similar measurement a year from now or six months from now. It seems to be kind of a trait-like aspect of cognitive ability.

Now, you might ask why are some people better and some people worse. And an intuitive explanation is to say that, well, people with better memory simply have more space in their working memory. They have room for a larger amount of information, but I think this turns out to be an incorrect intuition and there’s some fascinating work that suggests that if you try to measure the maximum amount of information that people can hold in mind at one time, it turns out to be very similar across people who have excellent performance and people who have not so good performance. And what really distinguishes the high performers from the others is the consistency with which they are able to load information into memory.

Paul Rand: What do you mean by that?

Edward Awh: For example, perhaps what you’d find is that in a group of a hundred subjects, the vast majority of people will have a top limit of about three items that they could hold in memory at a time. But what you’ll see is that some subjects accomplish that nearly every time that they try, and other subjects only rarely get to that maximum limit. So there’s something about consistency and attentional focus that is providing an advantage to some people, and it does so in this very consistent and reliable way across many different contexts.

Paul Rand: Now, your research also has shown that when cognitive loads increase, the brain doesn’t just slow down, but it actually drops information. And I think this gets to where it starts, to me, starts getting particular interest. How do you know that?

Edward Awh: Well, there’s two ways we can demonstrate this fascinating fact. One of them is we simply use behavioral measurements to objectively measure how many things a person can reliably report from a display that has varying numbers of targets. So perhaps we could vary it from one item all the way up to eight items. Eight items being well beyond the typical capacity of a human observer. And what we’ll see is that as we count up the number of items that are actually held in mind, one, two, three, we see a rise from one to two to three. But then actually as we go beyond three items, we’ll see that their ability to report that information actually declines. So that with six items, they can report fewer things back to you than they could with three items. Even though they had more things that they had the opportunity to report, it’s as though they drop that information when their memory system is overloaded.

Paul Rand: Got it.

Edward Awh: One interesting fact is that this tendency to drop information with overload is much more pronounced in individuals who have lower memory ability. So this is one of the things that actually distinguishes the high from the low performers.

Paul Rand: Okay. Now you have used, as you mentioned, EEGs to identify some neuromarkers that talk about and predict when attention’s about to fail. What are you seeing in the brain at that moment, and is that addressable?

Edward Awh: Well, so yes, I mentioned that we can look at EEG activity, which will provide us a pretty precise and time resolved measure of how many things a person is thinking about. And if we go back to your question about overload, if we continue to look at that neural activity as we vary the number of items that a person is trying to take in and report, we’ll see that neural signatures of storage actually also decline as you’re going to higher and higher loads. And this actually is a pretty good predictor of which subjects will also show a decline in their behavioral performance.

Now, another kind of neural signal that you can look at is ongoing activity, rhythmic brain activity, looking at activity in different frequency bands. And Ed Vogel has found in the past that certain frequencies of activity in the beta range of the frequencies will actually be predictive of when individuals are likely to experience a drop of information during an upcoming trial of the experiment.

So by scrutinizing these types of brain signals, you actually can see a little bit of a prediction of the future of how likely it is that somebody’s going to be dropping information. And this raises the exciting idea that perhaps we could use that type of neural activity to try to minimize those kinds of errors.

Paul Rand: Yes.

Edward Awh: If it were possible to measure it in real time and thus prepare somebody for the likelihood that they’re about to make that kind of error, that would be a really fascinating way to try to boost people’s performance, even if we’re not increasing their native capacity.

Paul Rand: Have our brains evolved over time with the amount of new inflows that are coming into it? Or have they stayed relatively consistent, but we’ve had to figure out how to manage the capacities that are innate to us?

Edward Awh: I think that I would lean toward the latter perspective that if we look at measurements of human memory performance, for example, over the last century, there’s a fascinating study that I talk about in class by a man named Jevons, who measured his memory capacity over 120 years ago at about four items, and we see very similar limits today in modern subjects using modern methods.

And we can kind of look at that trajectory of performance across many decades, where there’s actually been a fairly massive change in the density and amount of media information and the complexity of our environments with industrialization, and there’s no real compelling evidence that our mental faculties have changed in some fundamental way due to these really dramatic changes in the information load of our environments.

Paul Rand: If what we’re being told is that the brain is not necessarily evolving to process additional information, that gives us a pretty huge insight about how this new world that we’re going into, that we’ve been in, but even more importantly, what we’re going into, how we ought to be thinking about it, doesn’t it?

Edward Awh: Yes. And certainly, it shows that there’s a premium on our ability to take these limited cognitive resources that we have and to distribute them wisely, to apply them to exactly the right aspects, the most relevant aspects of our busy environment.

Paul Rand: Okay. Based on your research, can you talk about who, what kind of people, what kind of minds function better in complex information rich environments? And why is it they’re able to do so?

Edward Awh: So again, here I would return to this fundamental ability to use the remote control of your mind, to decide on which aspects of a complex environment that you’re going to focus your attention on, and thereby control which things are going to occupy your really limited space in working memory. So the individuals that tend to succeed in complex environments are the ones who can take their goal for that particular event or task and use it to voluntarily control which aspect of the environment they’re currently focusing their mind on. This determines which aspect of the environment will then enter into their working memory, and that is the online workspace of our ability to reason and solve problems.

Paul Rand: And do people have that innately, or have they learned how to do that better?

Edward Awh: I think that it is a fundamentally innate process. That is, it seems as though people have a certain level of ability to focus their attention. It doesn’t seem as though it’s that amenable to training or change.

Paul Rand: Okay.

Edward Awh: And that it really seems to be a stable trait of the individual with the exception that there are ways to damage your natural capacity. For example, if you are malnourished or if you have extended periods of stress or anxiety, we know that these types of things can undermine your ability to achieve the potential that you have. So if we have a child who has lived in a malnourished or stressful environment and we’re able to bring them into a more nurturing environment where they are well-fed, well, it turns out that their cognitive capacities will improve. And likewise, well, there is one thing that we can do that I am convinced actually does-

Paul Rand: I’m all ears.

Edward Awh: ... have measurable effects on memory and attention function, and that is exercise. It turns out, the brain is a metabolically very greedy organ in the body. It weighs about 4% of our weight, and it uses about 20% of our metabolic energy. There are many, many neurons and brain cells that allow us to think, and it turns out that they use a large amount of glucose and energy and oxygen when they are firing. And so this organ is highly dependent upon a consistent and reliable blood flow and vascular system.

Paul Rand: Wow. Okay.

Edward Awh: So if you are having good aerobic conditioning, then you will be giving your brain everything that it wants. And by contrast, if you have very poor aerobic function, it has measurable effects on cognitive ability. And I think that this has been some of the most convincing work that I’ve seen in the literature that has examined-

Paul Rand: That is very fascinating.

Edward Awh: ... what would we do to safeguard and even improve cognition. When you take older adults, for example, and they’re assigned randomly to a condition where they do regular exercise versus one where perhaps they just do some stretching exercises and have the same amount of contact with the experimenters. What they can show is that there are measurable changes in brain structure and function and in cognitive performance that seem to be supported by being in better physical condition.

Paul Rand: Okay. Very interesting. So there is this quest to improve attention/working memory. What do you think people misunderstand about that? And is the answer just to exercise more and not worry about types of other things?

Edward Awh: Well, I think that we have to acknowledge, there’s a multi-billion dollar industry out there that is trying to convince people that they can train their brain with video games or other types of cognitively challenging games. And there are many companies that people will be familiar with who are marketing these types of apps on phones, for example. And the typical argument that they make is an analogy argument. They’ll say, “Well, if I want to strengthen my legs and arms, I go to the gym and I work out my legs and arms, and that exercise leads them to be stronger and more effective. But where can I find a gym for my brain?” And they show you an app that has games on it and they basically argue, “This is like a gym for your brain.” But I think that this analogy just turned, while being a compelling analogy, just turns out not to be correct.

When you think a lot using a brain training game, it doesn’t actually change your native cognitive capacity according to a lot of data in the literature. Now, I’ll point out that the literature actually has debate and controversy in it. There are papers that have been published that have reported improvements in these types of cognitive ability measures after certain types of brain training exercises were being performed.

Paul Rand: Were they sponsored by chance?

Edward Awh: Well, they do tend to be related to, or they tend to be run by individuals that may have some interest in promoting a brain training game perhaps, but I guess the thing that caught my eye is that when I’ve seen large scale attempts to replicate these findings by independent laboratories, they just haven’t fared very well overall.

I wouldn’t close the door and say that it’s impossible that this would ever work because I think that would be a little too dogmatic here, but what I will say is that I’m still cautiously skeptical of this claim and I think that knock-down evidence for this idea that you can train the mind with a video game, it requires more work for us to really determine what the answer is.

Paul Rand: In the age of AI, and it seems to becoming stronger and more used by humans by the day, do we need stronger minds, or is it that we need to stop asking our minds to do the kinds of work that they were not designed to do and to learn to use these external tools better?

Edward Awh: I have a lot of thoughts about AI and in particular in the context of training our minds, which of course is one of the core goals of a university is to help students to train their minds to be as effective as possible by adding knowledge and wisdom and by adding skills that they didn’t have before. And I guess my worry at times about the use of AI in the classroom is that it may make it easier for students to settle for the work that AI can do for them rather than to sharpen their own ability to generate those products.

So for example, we all know that it is difficult to be an excellent writer, to be able to express ideas clearly and in a compelling fashion, in a written form, takes many years of practice and hard work. Now, what will happen if you could instead press a button and get a B+ on that essay? Will you engage in the hard struggle that it requires to become an excellent writer instead of a mediocre writer? And I worry that it may actually change the motivation of students who are overworked and busy, who will be more likely to rely on those crutches.

Paul Rand: It doesn’t even seem like that’s a question, does it? Well, It’s inevitable.

Edward Awh: It’s certainly something that we’ve seen instances of, and so that is a concern that I have is that there may be fewer students that take that important journey to really work hard to overcome those challenges. Now, having said that, I think that if you’re already an excellent, well-trained thinker on a particular topic, that AI can really facilitate your ability to express those ideas much more efficiently than you would have been able to express them without AI.

But in this case, we’re talking about somebody who has already kind of achieved the training and the analytical ability that leads to new or creative insights, and then using AI to bolster their ability to communicate the insights that they have, as opposed to using AI to substitute for the work of developing those insights in the first place. And so I think that there are many legitimate and genuinely helpful ways to use these tools.

And the only concern that I was raising there is, will it actually change people’s motivation to kind of make themselves the sharpest thinkers they can, who could use those tools in the most effective way? And so, one fun project that we’ve been working on is to develop a mobile app, not to train your brain, but rather just to measure your current capacity. How are you doing right now in your working memory and attention focus? And we’ve developed a task that can be run in just like one or two minutes. And what we’re hoping is that if people were to use that regularly, they would then have a robust baseline to know what their typical cognitive performance is and when.

And a really interesting thing about our standard of medical care right now is that our ability to assess cognitive insults, to assess when somebody’s cognition has been harmfully affected by a neurological incident or by an accident or something like that, it’s really very blunt tools that we have for measuring cognitive function in the hospital. We have things like the MMSE, which would detect if somebody no longer knows who the president is or what day of the week it is. So they’re detecting really serious and deep changes in cognitive processing, but they would be terrible at detecting subtle changes or less obvious changes in mental function.

So just take a hypothetical example of somebody who’s a brilliant person with an IQ that is two standard deviations above normal and they have a bicycle accident and they hit their head, they go to the hospital and now their IQ is average. They’ve lost two standard deviations, a huge amount of mental function. Nothing would be detected by the standard medical process that we have. If we had baselines for each individual, then we would actually be able to detect far more subtle changes in their cognitive processing.

This could also be a really useful approach as we have a larger and larger number of people becoming older in our society. We’ve got a growing number of seniors in our society. And so there’s a lot of interest in understanding cognitive decline and what may lead to things like Alzheimer’s or other types of mental degenerative diseases. And if we had measurements of people’s baseline cognitive ability, we would be in a far better position to detect early when there might be some change in their functions. And so, one thing I’ve been saying is it would be great if cognitive capacity or working memory capacity were treated more like blood pressure.

Paul Rand: Oh, interesting.

Edward Awh: Or we go to the hospital. They measure your blood pressure. Whether you have a problem with it or not, they check it because it’s just one of those basic parameters that they want to keep tags on. And I think that it would be a great idea for us to start doing the same thing with cognitive abilities.

Paul Rand: Okay. In terms of my final question, I think you’ve touched on this, but the question of you don’t see any time coming with Neuralink or any types of other tools where we’re going to be able to upgrade the RAM in our brains and just simply increase their capacity to do more of the things we’re talking about.

Edward Awh: Right. That’s correct. I think that at this point, while that is an exciting goal and mission for us to try to come up with such technologies. And as a neuroscientist, I would never shut the door on trying to find those types of approaches. That’s part of why we’re interested in building a precise neural model of how the mind works, is that it would hopefully give us the opportunity to have far more effective interventions or cures or even enhancements that we may someday develop.

But I just think we should have a sober view of what has already been shown to work in a compelling fashion and that there are some companies out there that may be presenting far too optimistic a picture of what current technology looks like. And so I would say instead, let’s optimize what we have. Let’s figure out how it works and how we can best support what we have, exercise being one of those things, and then let’s move forward and kind of keep a sober eye on how we can develop real technologies that may work in the future.