Not brain regions, brain networks

People love to talk about brain regions. So you probably know the amygdala is the fear centre, and the prefrontal cortex is the smart bit, and maybe you know that the hippocampus does memory, or the nucleus accumbens does reward. These are all common things Instagrammers and writers for Forbes like to waffle on about.

But none of this stuff is true. The amygdala isn’t the fear centre, the prefrontal cortex isn’t the ‘smart’ bit, the hippocampus doesn’t do memory all by itself, and the nucleus accumbens is far from the most interesting part of the reward system.

Really, what people are doing is the old brain science confidence game:

You will often hear people say something like ‘doing x thing releases y neurotransmitter which has z psychological benefit’. I have never once heard a construction like that where the reference to the neurotransmitter was anything more than cosmetic. To say ‘going for a run will trigger a release of dopamine, which will make you feel good’, has precisely the same informational content as ‘going for a run will make you feel good’. You haven’t learned anything because dopamine was included in there. Indeed, that’s not even how dopamine works

Now, in that article I’m talking about neurotransmitters, but people do this with all sorts of brain bits. Here, people might say something like ‘mindfulness activates the prefrontal cortex, which enhances your decision-making’. But everything activates the prefrontal cortex. It’s like a third of your brain! It’s silly. It’s the equivalent of saying something like ‘drinking water activates your liver, which enhances your hydration’. The liver is 30-ish% of your organs, and it’s certainly involved in hydration, but what exactly is including it helping you understand?

Water hydrates you. Mindfulness can be helpful.¹ Livers and prefrontal cortices are involved. That’s all you get. It’s just a cute way of pretending you know more stuff than you do.

Which is a shame, because there’s plenty of interesting brain stuff that isn’t just word salad, as I talk about quite a lot.

But to get to those, we need to stop talking about brain regions and start talking about brain networks.

¶Brain regions that do specific stuff aren’t interesting by themselves

You can tell that some brain regions really do seem like they’re responsible for certain stuff.

So for example, I’ve written about the ‘language centres’ of the brain. These really seem to be involved in processing and producing speech. This is because they’re located near a part of the brain that’s all about producing motor (movement) plans.

The language regions live right at the bottom of the primary motor cortex. If I take a magnetic stimulation device and buzz the primary motor cortex, I can make your fingers twitch and your toes curl. And right next to the primary motor cortex is the primary somatosensory (feeling of touch) cortex. So, if I prick your finger, I can see a tiny segment of this part of the brain get very excited. These regions of the brain are basically maps—maps of touch information that comes in, and maps of movement that goes out. Look:

A representation of the 'map' of our body in our brain. On the left is what we feel. On the right is how we move. Two strips of the brain that have an almost 1-to-1 correspondence to our body parts. From Wilder Penfield, c. 1940.

All of the primary sensory regions of the brain are like this. They’re the parts of the brain where your senses plug in, so they map the kinds of stimuli that you might sense with your sense organs. As I describe elsewhere:

At these sites of connection between brain and body, we have beautiful ‘maps’ of features of the world. At the sides of the brain, where the ear is connected, we have small clusters of cells that fire for specific frequencies of sound. At the back, where the eyes plug in, we have clusters of cells that respond to the very basic features of the visual world—colours, specific orientations, changes in contrast and so on. And where our bodily organs plug in, we have a map of of the body—a map of what we feel and a map of how we move:

So, in addition to the primary motor cortex (a.k.a. M1, or the precentral gyrus) and the primary somatosensory cortex (a.k.a. S1, or the postcentral gyrus), I’m describing the primary auditory cortex (a.k.a A1, or the superior temporal gyrus), the primary visual cortex (a.k.a. V1, or the central pole of the occipital lobe, and bits of striate cortex). We also have the primary olfactory cortex (a.k.a. piriform cortex and parts of the medial temporal lobe), and gustatory cortex (a.k.a. the frontal operculum and bits of the insula), which do smell and taste respectively.

But what makes these regions interesting isn’t they they’re so specific, nor their fancy names, it’s that you can also look around these primary regions and identify bits of brain that are more specialised in more complicated ways. So at the side of the head, you have the ‘fusiform face area’, which seems to be very interested in recognising faces, and the ‘parahippocampal place area’ which seems to be really interested in recognising places. But this shouldn’t really surprise us:

what the cortex appears to be doing is storing information related to the interaction of those more primary sensory regions nearby—about less specific and more general features of the world that relate to multiple kinds of information.

For example, as you move away from the region of the brain that codes for body parts we move and toward the vision part and you start to see clusters of cells that enthusiastically respond to things like motion perception—something that is both about moving and seeing movement.

The language centres are one of these. I told you they live right under the motor regions of the brain, but they also live right on top of the hearing part—speech is about hearing and moving our lips to speak.

A representation of the information streams in our brain. As we move from primary sensory areas to places in between, we get cells that activate in response to combinations of information. Adapted from Kandel et al. 2013 Principles of Neural Science.

And these ‘category selective’ brain regions, like the fusiform face area and the parahippocampal place area live sort of between the primary visual cortex and a bit that seems to be involved in storing memories. You could probably imagine where other ‘category selective’ brain regions are in the head, just be knowing what they’re about. Like, the ‘visual word form area’ is near the language bit, but closer to the vision bit.²

And this is the point, these regions aren’t interesting by themselves. It’s only when you look at them in context with each other that they become interesting. In this case, we’re being encouraged to view the brain like a map of our memories:

In essence, certain regions of the brain appear to represent aspects of the world in detail, and as one moves away from these regions, it represents more general features of the world.

Probably, then, a great deal of memory relies on the pathways we create and develop in the processing space of the cortex. Those beautiful neural maps already describe the world and your interactions within it. Why would you use a seperate structure to remember all that information? Wouldn’t be very efficient. So, as we move through the world, we are actively mapping it, and much of our memory is likely the product of this map we are inscribing into our neurons.

¶Brain networks are the things

This is a much more interesting way of using the brain to help us understand thinking and behaving. So, you could say “the amygdala is the alarm bell,” and talk about how when we’re threatened, we start to feel intense emotions that can cloud our judgement and control our actions, and your amydala is ringing for the wrong shit so you have this messy response to lots of stuff that you shouldn’t. But if that’s true of you, you already knew that, probably, and blaming the amydala isn’t really helping you understand this better. Maybe it reassures you that you’re not crazy, which is helpful, but it doesn’t help you fix the problem, especially since that’s not how the amygdala works.

Instead, we could say, “hey look, your brain is sort of like a map of your history of being in the world. Things that happen in a similar way will all be put in a similar place, and will be connected up to similar processes. So stuff that’s like stuff that scared you is probably going to make you feel scared.” It’s not a hijack, it’s a category error. And you don’t need to just give in to this fucked up alarm bell in your brain. You just need to train your brain to recognise the difference between actually scary stuff and not-scary stuff, so it ‘maps’ that information somewhere else.

That’s actual brain science. It’s not as cute as the ‘alarm bell’, but it seems like it’d be more useful. And there are a good handful of these that exist, that we could talk about. They’re called ‘brain networks’, and you’ll see these start to dominate the business- and pop-psychology market pretty soon, so let’s get ahead of the game, and take a quick little spin through the main ones.

¶The ethological map network

Key Regions: Primary sensory cortices and cortical regions that sit in between them, often referred to as ‘association cortex’ even though it’s not really one bit of brain but lots of regions. Also, bits of your prefrontal cortex, where maps of the most complex combinations of information live.

Now, this ‘map’ I’ve been describing isn’t strictly one of the networks brain scientists like to write papers about. ‘The brain is a map of the statistical structure of the world and you in it’ is an old idea, and there isn’t much research nous left in that juice worth squeezing. But it’s my favourite one, so I’ll give it a name. We’ll call it the ‘ethological map network’, after Graziano’s ‘ethological action maps’. But it seems like a much more useful network to talk about than the subnetworks that are often published about. I don’t think they need their own paragraph, but you have:

• The visual networks, which are normally described as two ‘streams’: one that’s about what you’re looking at, and one that’s about figuring out where something you’re trying to see it.
• The motor network that span from the primary motor cortex where the big, rough and jerky movements are planned, through the basal ganglia and the cerebellum, which smooth things out and get the details right.
• The language network, which have the language regions obviously, but also rope in a bunch of the visual streams and some memory bits too.

There is the so-called ‘semantic hub-and-spoke network’, which describes a bunch of regions around the sides of the head, apparently describing how concepts are mapped out in the brain. But I don’t really see this as fundamentally different from the rest of the map. Your concept of a barking dog needs visual, audio, and memory information, so it lives at the sides with all the other stuff that needs all that information. We’re not doing research on it, so trying to understand more won’t be very rewarding.

¶Default Mode Network (DMN)

Key Regions: Medial prefrontal cortex, posterior cingulate cortex, precuneus, angular gyrus, and lateral temporal cortex.

This one is already popular in the more forward thinking blogs, mainly because it often pops up when people are meditating. It’s like the ‘rummaging through your own head’ network because all these regions start firing away when you daydream, reflect on the past, reflect on imagined futures, and meditation-styles that have you focusing your attention internally. No one has been able to convince me that it’s actually interesting, other than the fact that it exists, and I don’t have anything particularly fun to say about it. However, there seems to be some evidence that the DMN is often quite active just before ‘a-ha’ moments, which is another good case for the idea that these moments of insight are more likely to happen when your mind is wandering off-task, rather than focused on a task. But I already wrote a whole article on that.

¶The Salience Network

Key Regions: Anterior insula, anterior cingulate cortex (ACC), and parts of the fronto-opercular cortex.

You’ll hear people call the anterior cingulate cortex the ‘conflict monitor’, and I fully predict it to become a new ‘amygdala’, because you often see it get excited when something violates your expectations, or when you have to overcome something easy to do something hard, like name the colour of this word: blue. But the ACC is better understood in the context of the rest of this salience network, which seems like they work together to figure out what you’re supposed to be paying attention to. You see it get excited often when you’re transitioning from daydreaming to paying attention again, and is probably involved in that ‘what was I supposed to be doing again’ feeling.

¶The Central Executive Network (CEN)

Key Regions: Dorsolateral prefrontal cortex and posterior parietal cortex.

This one, also called the frontoparietal network, and by my old lab group the ‘Multiple Demand Cortex’, is active all the time, but is more active when a task is harder than when a task is easier. This is probably a good time to point out that all these networks overlap, and lots of the Salience Network figures in this network too. But unlike the Salience Network, which is more about surprise and attentional focus, this one seems to help you solve problems when problems need real thinking about. And that’s interesting, that they overlap but are meaningfully separate. Just because something is taking a lot of your attention doesn’t mean that you’re thinking very hard. You actually know this. You’ve probably heard of a ‘flow state’—total concentration on what you’re doing. But you’re not necessarily thinking there. In fact, the executive network seems to indicate that you’re not really thinking that often at all.

¶Affective networks

Key Regions: Amygdala, insula, cingulate and prefrontal cortex regions.

Paying attention to stuff and thinking about stuff isn’t the same as feeling some type of way about stuff. Your affective networks are the ones that seem to work this out. What’s interesting about this network is that it includes the insula, which mainly seems to be about interpreting signals that come in from your viscera and your bodily feelings. I think the fact that these regions all like to happen together when you’re working out your emotions lends a lot of weight to the dual-factor theories of emotion: we first feel emotions as a visceral, autonomous response. A simple feeling of ‘goodness’ or ‘badness’. Then we hand that information over to other parts of the mind, which tries to work out what to do with it. Has it happened before? What else is going on? How good or bad does this feel (the amygdala’s actual job)? This seems to be what the affective networks are doing.

¶Reward network

Key Regions: Nucleus accumbens, ventral tegmental area, prefrontal cortex, basal ganglia and the hippocampus.

I wrote a whole article about this recently, so I’ll just crib:

Reward comes in and gets to the VTA, and the first thing it does is release dopamine into the striatum … [and] … the striatum gets very keen when the VTA showers it in dopamine … from there, the system loops in a few more regions. It talks to the hippocampus, which remembers the context around the reward. It talks to the insula, which remembers the feelings happening in your body around the reward. It talks to the amygdala, which remembers the emotional significance of the reward. And it walks its way to regions at the front of the brain (the PFC), where it seems to interrupt some of the work the PFC does managing impulsivity, risk-taking, and evaluating how valuable stuff is.

All that to say that the reward system doesn’t actually reward you at all. It just remembers what contexts were rewarding and starts getting very excited when you recognise something that reminds you of that context. Something lots of companies like to hijack, but something you can use yourself just as easily.

¶Outro

Hopefully you’re convinced. Brain regions are boring. Brain networks are more interesting. You can actually use these to figure out why people do what they’re doing. And maybe I’ll write some more articles than the ones I link to above to flesh that out a bit. But until then, at least you can ask any idiot that starts gabbing off about the prefrontal cortex what bit they’re talking about, so you can work out if they know the difference between the bits that’re involved in ‘decision-making’ and the bits that are involved in daydreaming. If they don’t, don’t buy their book. Thank me later.