Everyone. Welcome to Non Trivial. I'm your host, Sean mcclure. In this episode, we look at how neuroscience tries to explain human behavior by identifying specific regions in the brain. I'll discuss how such targeted approaches are problematic in science and how neuroscience is just another example of a complex science acting as though it's far more simple and deterministic than it really is.
We'll look at the myth of root causes, why nature doesn't localize behavior and what needs to change to study complex phenomena correctly. Let's get started. Neuroscience is the science that tries to understand the nervous system that highly complex system of neurons and signals permeating the human body. This includes the brain, spinal cord and an intricate network of nerves. The nervous system is the major control system.
Our body uses to coordinate our physical and behavioral attributes by transforming sensory information into usable signals. Nervous systems enable bodies to respond to changes in their environment. Now, neuroscience looks to benefit from advances in everything from molecular biology to computational science motivating the effort to explain human behaviors in terms of how the brain is structured.
Neuroscience inspects how our brains have evolved physically hoping to demonstrate how evolved structure leads to both normal and pathological behaviors. Neuroscience is touted as our window into the biological basis of learning and memory. It looks to bring a kind of rigor to our understanding of brains, something more tangible than the ethereal models of psychology. There is an apparent concreteness to the explanations of neuroscience because they rest on physicality.
Neuroscience has a number of tools at its disposal. Researchers can investigate patterns of neuronal connectivity using molecular biology and genetics to track changes, attempting to correlate those changes to various biological functions. Now, beyond the usual quest to understand nature, neuroscience may also contribute to the treatment and diagnosis of brain disorders. A better understanding of the brain could shed light on things that affect our lives.
Neuroscience is one of those research areas that appears worthwhile for both scientific and ethical reasons. Now, people can't help but be interested in what makes us unique. We are bombarded with messages of who comes from where and how cultures do things differently. Now on its face, there isn't much wrong with this. I mean, after all differences make us uncommon and interesting.
There also seems to be some utility here if we can understand what makes us different, that might lead to a better society. Since knowledge about why people are different could lead to useful interventions, but understanding what actually makes us different is difficult. So first, so called differences are ultimately just mental labels, right. Nature is under no obligation to adhere to the demarcations we create in our minds.
We envision borders between things to make it easier to comprehend and navigate our world. Second, whatever actual differences exist, become increasingly difficult to ascertain under complexity. While it's easy to notice when things look different, this isn't the same as knowing how structure leads to function. Former is interesting but not necessarily consequential. The latter requires uncovering causes connecting the appearance of something to the behaviors we observe.
This is the notion of locality where we attempt to identify the specific regions responsible for something we observe. More generally. Neuroscience makes its way into these different discussions because many of the discrepancies in people's behavior must stem ultimately from our brains.
If we could find structural differences in the brain, then these might explain some of the behavioral differences we see across individuals in extreme cases like patients with epilepsy, such identification might dramatically improve someone's quality of life. Since to know a cause is to conceive of an action that might change outcomes.
One way neuroscience chases local causes is by looking for structural differences using functional magnetic resonance imaging or F MRI F MRI scans reveal areas of the brain that light up when people experience emotions or conduct some mental process, fear, anticipation, happiness and reasoning are all mapped in the brain by correlating these behaviors to specific places in the brain.
Now, when causal explanations are given, they present signs with targets, right, for potential future intervention. Now, if these targets are correct, then treatments can target the problem area and potentially improve outcomes. But if the targets are wrong, then at best they waste people's time and money and at worst, they heavily discriminate against or even damage parts of the population. For bogus reasons.
Finding root causes defines much of science since scientists are largely concerned with uncovering the why behind what we observe? Explanations are the grand purpose of today's science. With an explanation. In hand, we have a theory about how something supposedly works. But while peeling back the layers of various phenomena can show us its parts. Such reverse engineering cannot tell us how nontrivial outputs are arrived at. Structural differences may play a role.
But there's no way to know what that role is because the notion of role is misplaced under complexity. A role relates to an individual thing function. It is some specific activity intended towards a specific goal, but that's not how complexity works. Specific things don't produce specific outputs because the outputs we observe and measure in complex situations are arrived at in aggregate.
It is intellectually dishonest to conduct science under the premise that root causes can be uncovered under complexity chasing the root cause of a complex situation will always look like a wild goose chase because there's nothing to chase. It is not a matter of peeling back more layers. It's a matter of reasoning under an incorrect paradigm pieces interact in fantastically intricate ways to produce what we observe in complex phenomena.
Outputs are achieved in a multiplicative fashion, not some simplistic sum of inputs. If we look at the history of Neurosciences attempt to attribute observed behaviors to specific regions of the brain, we see the unsurprising jumping around of observation. One study points to memory in the hippocampus only to be revised when some memories are found to reside outside this region, extraction of some piece of the brain stops seizures but also renders other cognitive abilities lifeless.
Every attempt to isolate the cause is met with some other region of the brain that was critically involved in achieving the overall function. The correct scientific stance on anything complex is that what we observe is arrived at via the entire system. While such truth goes against much of the convenient reductionism at the heart of today's science, it's still truth complex systems call upon all the pieces of a system in order to produce what we observe.
It is a concerted effort, structural differences in the brain as with any physical system, undoubtedly produce different information. But those differences are used to achieve what's needed in aggregate not to play a specific role. The human need to create mental demarcations can corrupt any science that looks to understand complexity.
If such misplaced concreteness is taken too literally a common way to model complex systems is by using networks, networks are just a bunch of nodes connected by lines, networks can be analyzed for their behavior and properties, providing insights into complex systems. The human brain can be modeled as a complex network. Since the overall structure of any brain is a network of neurons connected by synapses with you know, neurotransmitters moving across these bridges, right?
A common behavior seen in complex networks are power law relationships. This is when a quantity varies as a power of another. So think of doubling the length of a side of a square, the area ends up being multiplied by a factor of four. So you change a little bit in one part and then the other part changes by a lot. Now many real world phenomena appear to obey power laws including unsurprisingly, the activity patterns of neurons, power laws are characterized by a number of properties.
One property is scale invariants whereby some feature of the object does not change. Despite altering the scale by which we observe it, many geological phenomena possess scale invariants. For example, the size distributions of rock fragments of volcanic eruptions and earthquakes all show this property. So imagine you're looking at some image and it looks like a bunch of rocks scattered over the surface, but you have nothing to give you the scale of what you're looking at.
You know, there's not like a pencil in the picture or a hand or a human being or a building, you just see a bunch of rocks on a surface Well, there's actually no real way to know what scale you're looking at. Unless you have something for comparison, it could be literally a nanos cop or microscopic image of particles scattered across some smooth silicon surface, uh or it might be boulders that are, um, you know, scattered across the landscape.
And unless you have something for reference, you can't really tell what it is you're looking at. And that's because the patterns, a lot of the patterns that we see in nature actually scale and variant, the things that you see at the nano and microscopic scale are very much the same thing you see at the very very large scales. You know, the same with, you know, Galaxies where you get the spiral formations of Galaxies at these massive scales.
If you go out even more to, you know, the literally the the kind of uh fibrous structure of the entire universe and the way the stars are connected to these large star systems, they have some of that uh patterns that you see at the very smallest scale as well. So the point is is that scale invariance is uh is is is these properties that exist essentially at all scales.
Now, the thing about scale and variance is that it's a type of universality universality occurs when the properties of a system are independent of the details of the system leading to many different physical systems exhibiting the same behavior. So for example, uh the behavior of water. And CO2 are almost identical near the boil near the boiling points.
You know, despite them being entirely different substances, so universality upends the notion of locality because universal properties are agnostic to the underlying physicality of the system. For something to be locally important would mean that one part of a system that exists at a specific scale is producing what we observe. But if the properties that matter are agnostic to scale and physicality, then attributing overall behavior to a region isn't reality, it's fiction.
OK. So let's use some examples. So take an example of uh influencers in a social network, right? We talk about social networks all the time, whether we're talking about Twitter or something else. And uh and you'll have influencers in that network. Now, these are individuals who for whatever reason, become popular hubs of network activity, right? Influencers have the most followers and quite literally influence the overall behavior of the network.
So when an influencer says something many people will be exposed to that message uh and and apply more weight to it compared to some random Joe on the network, right? But hubs don't exist without the rest of the population. A hub is a location where we see the most intense activity, but that activity is arrived at via a multi way street.
In other words, an influencer is only defined by virtue of being embedded inside a network of countless individuals without that context, there is no hub, an influencer has nothing to influence or without a crowd, but it's more profound than that. The information that makes an influencer what they are fully depends on the flow of information from the rest of the network. The influencer is not really a concentrated region more causally connected to the behavior of the network.
That's just a convenient way for our minds to think of them. An influencer informational output materializes out of the multiplicative interaction between many individuals.
Sometimes they show these uh Power law relationships as a, as a distribution and it's got like a big spike squashed on one end and the rest is just this really, really long tail, you know, and, and we tend to, you know, it's kind of like, you know, the 80 20 rule, right, where 80% of the value comes from 20% of whatever the people or the land or whatever it is, right? Um But you need that long tail, that long tail is what makes the spike possible, right?
It's, it's you don't just have the one thing. So, so complex systems have both their concentrated parts like those, those hubs in the network or the spikes in the graph. And they also have the rest of the population. The idea that one is more important than the other is really just an artifact of the way people demarcate sensory information into regions of importance, right? An influencer uh is not and cannot be a target. It is too interconnected and codependent to be thought of as such.
Wiping out, let's say a terrorist cell or a tyrannical leader will undoubtedly impede terror attacks, but only momentarily. Right. What makes complex networks so resilient is their distributed nature. There are two things that happen to complex systems in response to an intervention. They either self heal and adapt or they collapse.
Now, if we collapse a, a terror network, we're not merely wiping out, you know, terrorist cells, we're wiping out food distribution, education and and potentially an entire economy. I mean, whatever it is, you're working with these things are deeply, deeply interconnected, right? If we extract a region of the brain to stop seizures, we'll be stopping a lot more than just someone's seizures.
There's no such thing as targeting and or extracting some isolated portion of a network and then deterministically controlling the outcome. At least not under complexity. It is only the human need to create contrived labels that we assign a location to what we observe no different than the false narrative that geniuses are responsible for human progress. The appearance of major parts of activity does not equate to the source of behavior.
We observe regions are just names, we assign to locations that appear to concentrate activity. The behavior we notice depends, you know, not on some unique local structure, but instead on the way information works across the entire network. Nature does not localize its behavior because the very purpose of a network is to leverage long range correlations to establish what's needed. Now, look, there are a lot of good things to study in neuroscience, right?
The brain is obviously a worthwhile test subject and you know, perhaps the most important one, if someone suffers from some extreme condition such as you know, debilitating mental illness, right, or ep- epilepsy or something like this, then perhaps an intervention is worthwhile. But research that runs counter to how complex systems function helps nobody. The chasing of f MRI hotspots will forever be in vain because that's not how complexity works.
It's not some error or a lack of sophistication and measurement that produces the mercurial firing patterns in the brain. It's the natural and fully expected behavior inside a fantastically intricate network. Trying to find the location of behavior is a remnant of outdated science. Neuroscience needs to detach itself from the reductionism that plagues so much of today's complex sciences. Neuroscience is not physics.
The brain cannot be understood by reverse engineering it down to components with specific roles. A new generation of neuroscientists need to reset the bogus aspects of their field. They need to refound their vocation absent the false premise that complex behavior stems from regions. They need to embrace the fundamental opacity that precludes the notion that specific structures produce specific behaviors, focusing instead on universal properties that all complex systems share.
Neuroscience is just another example of what degrades so much of today's scientific enterprise. Today's paradigm can't help but try to reverse engineer the phenomenon we observe. You know, since the so called enlightenment, humanity has oriented its scientific efforts around extraction and isolation, leading to targets and subsequent intervention many times. Now look that worked well for simple systems studied, you know, in physics and the machines created during the industrial revolution.
It does not work for the overwhelming number of phenomena that define our real complex world. That's it for this episode. Thanks so much for listening. Until next time, take care.