You know that feeling when you walk into a room and immediately forget why you came in? Maybe you were there to fetch your keys. On your way to the room, you were thinking about grabbing your keys. But once you arrive, your keys have completely disappeared from your mind.
This is sometimes known as the doorway effect , since it often strikes when you walk into a new room. Why does it happen? The answer has a lot to do with a faculty called working memory. Information gets stored in working memory when we need it for the tasks that we are engaged in right now (like remembering to grab your keys).
What makes working memory so intriguing is its close link to consciousness. The doorway effect suggests that when information is removed from working memory, it immediately seems to leave consciousness. It also suggests that it is easy for information in working memory to be forgotten.
The link between working memory and consciousness is getting increasing attention in psychology, philosophy and neuroscience. Could working memory somehow give rise to consciousness? In my new book , I explore the complex relationship between the two.
Working memory: both rich and poor
To understand the doorway effect, we'll need to know a bit about working memory. One thing that makes working memory so special is that it's so rich, both in terms of the information it has access to, and its processing power. According to recent models of working memory , it can draw information from sensory channels (vision, touch, smell etc), as well as from other memory systems such as long-term memory and also the brain's system for processing language. In other words, working memory is where a lot of the information in your brain comes together.
Once working memory has that information, there's a lot it can do with it. Inside working memory are a host of different smaller systems for specific tasks, including visual and spatial reasoning (like solving a Rubik's cube) and storing chunks of information (like a phone number). There's even a "central executive" system (my favourite). The executive is like a merciless boss, assigning tasks to the different systems within working memory and keeping everything under control.
In other ways, what makes working memory so special is that it's simultaneously very poor. Despite the riches of information available to it, working memory can only actually store a tiny amount of information at any one time.
In one classic experiment reported in 1997 people were asked to view a screen with several coloured shapes on it, which they were told to remember. The shapes then disappeared for about a second, and a new set of coloured shapes appeared. One of the new shapes might have changed colour. Participants were asked to spot whether there had been any changes between the two sets of shapes. This is called "change-detection".
People were almost perfect at this when there were only 1-3 shapes involved in each set, but got steadily worse as the number of shapes was increased from 4-12. The experimenters argued that this is because it gets harder to store information as the number of shapes increases. This is because the capacity of working memory isn't big enough to store lots of shapes. The experimenters concluded that the capacity of working memory is only about four "slots". Once those slots are taken, working memory is full up: there's simply no more room for any new information.
The idea that working memory has "slots" is closely related to something called "chunking". Here are two strings of letters (nine in each). Try to memorise them both:
BBC FBI WWF
ZQK EWP WLJ
I bet you find the first string of letters easier to memorise. This is because they're familiar, and so naturally your brain sorts them into three chunks. They can then be stored as three different chunks in working memory . For this reason, the first set only takes up three slots in working memory. The second string of letters is unfamiliar, and so requires us to store all nine letters as individual chunks. This is difficult because working memory quickly runs out of slots.
But like many features of working memory, its capacity is a hotly debated issue. A growing number of scientists have rejected the idea that it has "slots", arguing instead that its capacity is more of a flexible resource that can be differently distributed across different pieces of information. According to this view, working memory's capacity - far from being four rigid slots - might be more like a tank of water to be used in watering your garden: you can give a little bit of water to lots of different areas, or lots of water to just one or two areas.
In the same way, working memory might be able to store a little bit of information about lots of objects, or very detailed information about just one or two of them.
One study from 2004 supports this flexible resource view over the slots view. This experiment also used change-detection with shapes.
Crucially, the experimenters tried this with different kinds of shapes. Sometimes they used only very simple shapes, sometimes very complicated ones. They found that people seem able to store information about more of the simpler shapes in working memory. They were much worse at storing information about the complicated shapes. In fact, the experimenters suggest that the capacity of working memory for a very complicated object (like a cube with many different coloured sides) might only be between 1 and 2.
This seems to show that the capacity of working memory is "soaked up" much more when it tries to remember very complicated objects. This suggests that working memory doesn't have a fixed number of slots, but that its capacity depends on how complicated the information you're trying to store is.
To me, there's something romantic about how rich working memory is in terms of how much information is available to it, and how poor it is in terms of its small capacity. It's like it can always see the vast riches available to it, but can only ever sample a tiny portion at a time.
The low capacity of working memory can help us understand why the information it stores is so easily forgotten, like in the doorway effect. There's not much room in working memory, so when new information comes in, old information needs to go.
As I mentioned at the beginning, research suggests that the very action of walking through a doorway might trigger forgetting. One experiment showed that people find it harder to remember things when they walk through a doorway, compared to people who walk the same distance but don't cross a threshold. When we enter a new room, its like the brain flushes away the old facts from working memory, to get ready for the fresh information that we might need in our new setting.
From an evolutionary perspective, the doorway effect makes sense: forgetting old information is important in helping us to stay open and alert to novel information in the new environment.
Thinking about the capacity of working memory can help shed light on why it's so easy to forget things, even when we were just this second thinking about them. But there's an even more tantalising possibility here. When we forget things like our keys, they seem to drop away from our consciousness entirely. This raises the suggestion that working memory and consciousness might go hand in hand.
Working memory and consciousness
Consciousness is perhaps the biggest mystery facing both science and philosophy today. By "consciousness", I mean the subjective experiences that we have of the world. Consciousness includes the visual image of a beautiful sunset, or the taste of chocolate, as well as emotions like love and anger.
Many philosophers and scientists have thought that working memory can help us understand consciousness. .
The close link between consciousness and working memory is clear from some of the major theories of consciousness in psychology and neuroscience today. Personally, I'm a huge fan of the global neuronal workspace theory , which suggests that consciousness arises as a result of information being "broadcast" in a "global workspace" in the brain. This workspace is like a central information store , which can process information and distribute it globally to many different systems in the brain.
Does that sound like working memory to you? If it does, that's no coincidence: the global workspace and working memory are similar notions . Their similarity can even be seen in the brain. There are still lots of open questions about where working memory is located in the brain but one important area is the prefrontal cortex. This is at the front of your brain, just above your eyes and behind your forehead. The same area also seems to be important for the sort of global broadcasting that global workspace theorists think is responsible for consciousness .
Scientists friendly to the global neuronal workspace theory have suggested that when we pay attention to information that is stored in working memory, that information gets boosted in volume and is broadcast across the brain - and that is what makes it conscious. According to this view, consciousness arises when working memory and attention work together.
The idea that attention and working memory might both be important for consciousness seems to fit with our own experience. When you're trying to remember a phone number in your head, your attention stays on the phone number and you're conscious of it. If someone distracts you by asking you a question, your attention is pulled away from the phone number and it immediately gets deleted from your consciousness. According to this picture, no attention = no consciousness .
The importance of both working memory and attention also fits with experimental data. One of my favourite experiments studied people walking across a courtyard on a spring afternoon. It was found that 75% of people who were on their mobile phones completely failed to spot a purple and yellow clown unicycling around the courtyard. This is even though the clown could have crossed their path, potentially causing a dangerous collision. They were on their phones, their attention was elsewhere, so the clown didn't get into their consciousness. Again, this suggests that no attention = no consciousness.
I have a lifelong fear of clowns. The idea that there might be a clown nearby that I just haven't spotted fills me with dread. (I mean come on, they're clearly terrifying ).
But like everything to do with consciousness, the link between consciousness and working memory is controversial. Some think that there's just too much consciousness to fit into working memory. Others say that some bits of working memory aren't conscious at all. Let's have a look at these arguments.
Is working memory too small for consciousness?
We've seen that working memory has a small capacity. This raises an obvious question: if working memory is responsible for consciousness, doesn't that mean that consciousness must have a small capacity as well?
This can be a difficult idea to swallow. Imagine you're looking out at a countryside scene. You see rolling hills, the vibrant sunshine and a herd of cows. You hear the birds, smell the fresh cut grass and feel the wind on your skin. Surely you are conscious of this whole scene all at once. But we know that working memory has a capacity that is far too tiny to fit all of this information in at one time. If consciousness arises from working memory, then how can I be conscious of all this stuff at once?
Indeed, some philosophers and scientists have argued in just this way, saying that consciousness overflows the capacity of working memory. If this is true, it would be a problem for those who think that consciousness arises from working memory.
In response to this problem, supporters of the link between working memory and consciousness have argued that consciousness isn't as big as you might think. With the countryside scene, you might think that you're conscious of all those sights, sounds and smells together. However, according to this view, really you're only ever conscious of a few scraps at any one time. The reason it feels like you're conscious of more is because, whenever you pay attention to something, that attention boosts the information into consciousness.
This is known as the refrigerator light illusion . Imagine someone who thought the light in their fridge was always turned on, because whenever they open the door to check, the light is on. Obviously, the problem is that the very act of opening the door causes the light to come on.
In the same way, the very act of checking to see if you're conscious of birdsong causes you to direct attention to it, which brings the birdsong to consciousness. According to this view, we are only ever conscious of a few little bits at a time, but the ease with which attention can make things conscious fosters the illusion that we're conscious of a lot more.
As if it weren't bad enough that doorways make us forget, or that phones make us ignore unicycling clowns, now we have to deal with our fridges hoodwinking us about our own consciousness.
Working memory without consciousness?
Another area where consciousness and working memory might come apart concerns unconscious information. We know that lots of processing in our brains occurs unconsciously. When it comes to information in the brain, we only get to be conscious of the very tip of a large iceberg . Some psychologists have suggested that some of the information in working memory is completely unconscious. If this is true, this would mark an important difference between consciousness and working memory (since by definition, unconsciousness can't be part of consciousness).
One key experiment from 2011 involves showing participants an image of a rippled patch, tilted at a specific angle. In the psychology world, this is called a Gabor patch. This patch was only shown to the participants for the tiniest flash of time. It's there on the screen for only 16.67 milliseconds, about 17 thousandths of a second. This is about as long as a bee takes to flap its wings three times.
Flashing the patch on the screen so briefly prevents the patch from being consciously seen. Because of the brevity with which the patch appears, information about the patch enters the participants' eyes, and participants see the patch but are not consciously aware of it. They see it, but unconsciously.
Still, we know that information about things we're not conscious of must be processed at quite a high level in the brain. After the first patch disappeared, they were shown another different patch. This one was shown for longer, so it could be seen consciously. Participants were asked to indicate (by clicking a button) whether this second patch was tilted to the left or the right of the first patch that they had not consciously seen. Amazingly, they were able to do this at a level above chance. Even though the first patch was unconscious, people could still use information about it to make comparison judgements.
This is crucial for working memory and consciousness. The experimenters claimed that information about the briefly flashed patch must be stored in working memory, even though participants were completely unaware of it. From this, the experimenters concluded that some information in working memory is not conscious. If so, the link between consciousness and working memory is weaker than we might hope.
My idea: not all-or-nothing thing
When it comes to the experiment just discussed, I want to close with some of the ideas that I've been exploring in my book . I think that working memory might not be an all-or-nothing thing, that information might not have to be either "in" or "out" of working memory.
Sometimes, we can slip into the trap of thinking that everything in the human mind must be either one thing or another. This mindset is very natural, but in my book I argue that it must be rejected.
I suggest that there could be some information that is not "in" or "out" of working memory, but somewhere in between. I argue that working memory comes on a spectrum . Some information is definitely in working memory, and some is definitely not. But in between (I argue) is a large grey area where there is some information in our minds that falls in between being fully stored in working memory and not being stored.
This is certainly an unusual suggestion. But I argue that thinking about working memory as a spectrum sheds new light on the experiments I talked about above. Remember the patch that was flashed up in three beats of a bee's wing - so fast you couldn't consciously see it and preventing it from being fully encoded in working memory? I argue that this information sits in the grey area between being fully in and fully out of working memory.
This raises an interesting possibility with respect to consciousness. If we think that working memory is closely linked to consciousness, an obvious question is whether there can be a grey area between something being conscious and not being conscious. Indeed, some philosophers have suggested that there might be such a grey area.
At first sight, the experiments I've talked about might look a bit strange and obscure. When I tell people about these experiments at parties, they sometimes ask me how I can spend so much time thinking about doorways, unicycling clowns, or how we remember random shapes. I get visibly excited talking to people about all this, and I can almost see them making a mental note not to invite me back.
But these experiments are exciting. To see this, we need to step away from the technical details of these experiments and take a broader view. These experiments and others continue to unearth the mechanics of working memory. There is still much more to find out, such as where and how it is brought about in the brain , and the ways in which injury to the brain can impact on working memory . As these mysteries are slowly explained, perhaps we will be in a better place to tackle the arguably biggest conundrum in science: consciousness itself.
Henry Taylor has received funding from The Royal Institute of Philosophy, The Leverhulme Trust and the Arts and Humanities Research Council.
