Monday, December 30, 2013

Sal Khan's Academy and the Purpose of Myelination

ADDENDUM BELOW: Please see the addendum at the end of the post - it seems that KA has redone their video, correcting the mistake I pointed out.

In recently learning about Khan Academy and the debate over their place in the world of education and whatnot, I stumbled across this quote from Salman Khan's 2011 interview with Charlie Rose:

But the main thing I do... whether it's history or chemistry or anything is get the scaffold and then really immerse myself for however long it takes and then make sure that I can make intuitive connections between everything that happens.  [...] Understand why something - if I'm doing it on the neuron [...] a biology book will tell you the signal goes across because there's a myelin sheath.  I'm like yeah but how does putting a little tissue around the neuron make the signal go faster, right?  And no biology book will tell you that answer.  And I'd kinda ponder it a bit and then I was like it's kinda like a fiber optic system and one's kind of amplifying the signal but if you amplify too  - and then I'd call up some buddies who are either biologists or communications engineers or whatever and I was like "Does this make sense?" and they're like "No, I think you're right."  And so it's a big deal for me if I'm doing biology - why does the myelin sheath make the signal go faster?

As a fledgling neuroscientist, this struck me as a somewhat weird way to think about the purpose of myelination, so I went to check out the video in which he attempts to teach this concept.  Here he is discussing what myelination does (skip to 8:35ish if you don't want to watch the whole thing):

Now - the axons have these huge spaces where they're covered up.  They're covered up by these Schwann cells or by these myelin sheaths.  So even if there are gates underneath them they're useless.  They can't interface with the outside sodiums.  They can't interface with the outside sodiums and potassiums.  So over this - over this distance of the axon you cannot have a straight-up action potential. 
But what happened is your action potential it made everything a lot more positive than it would have been had there not been that opening of the gate and now you can have an electrotonic effect again.  So you kind of boosted your signal so now this is going to - So now this might get to +40 maybe over here we only get to -50 but -50 once again is enough to trigger - to trigger a sodium gate which then makes everything really positive again.  It - it boosts the - it boosts the signal.

He seems to think that the purpose of the myelin sheath is to smother voltage-gated sodium channels on the axonal membrane.  By preventing access of these channels to the extracellular fluid [this is his speculation, I don't think that's really 100% true] the myelin ensures that the signal can only propagate electrotonically in that region of the axon.

But what happens if we ask a student (or ask yourself if you don't know how myelin works and just learned from his video) the following question:

Suppose that voltage-gated sodium and potassium channels involved in action potential generation are clustered predominantly at the axon hillock and nodes of Ranvier, and are largely absent from the neuronal membrane wrapped in myelin [this happens to be true].  Now suppose we remove the myelin sheath, but keep those channels in the same locations.  Would anything be different about how the action potential moves down the axon?

The answer, based on what Khan teaches, is no.  But the real answer is yes.  Myelin does not simply exist as a measuring stick for how far apart action potential regeneration should be to guarantee reliable saltatory conduction.  Most importantly, myelin is an insulator!  It speeds conduction velocity along the membrane by increasing resistance across the membrane.

This illustrates one problem with Khan Academy: when you have just one guy learning about a lot of subjects and then trying to teach them, he's going to get some things wrong.  And because his videos just gloss over the surface facts without explaining the concepts behind everything (to be expected because you can't teach everything in a short video), his students aren't going to know to ask the right questions to realize where he might have made a mistake.



Well one day after this blog went up, the video I initially referenced went private (you may have noticed) and a new video on saltatory conduction was added in its place:

This video corrects the earlier mistake!  I don't know if it had anything to do with my little blog (this particular entry only had like 69 views...) or if others had pointed out the mistake earlier and they had the video ready to go and were simply prompted to get on it by this post, or if the timing was entirely coincidental.  Still, credit to them for taking the time to redo the video and correcting the mistake!

Wednesday, August 21, 2013

Can We Chill Out About How Complex the Brain Is?

Time for busting another lazy hook:

"We won't be able to understand the brain. It is the most complex thing in the universe," says Professor Sir Robin Murray, one of the UK's leading psychiatrists. (BBC)


This is SCIENCE FRIDAY. I'm Ira Flatow. Your brain has nearly 100 billion neurons, and one of my next guests compares that complexity to the Amazon rainforest. In fact, he says there about as many trees in the Amazon as there are neurons in your brain. Think about what the Amazon looks like for a second.

And the roots and the branches and the leaves and the vines, all of that can be compared to the tangled network formed between your brain cells because many of your neurons are in fact wired to tens of thousands of other neurons. That incredible complex network is packed into a soft, three pound organ inside your head, making it, as my next guest [Christof Koch] says, the most complicated object in the known universe.


The center's goal is to understand the biological parts and processes behind such phenomena as consciousness, moral behavior and logical thought. "There are more synapses in the brain than stars in the galaxy," Cohen notes. "We are studying the most complex device in the known universe." (Princeton)

1.  The brain is not the most complex thing/device/object in the un/known universe.

If the brain is so complex, then isn't the human head more complex?  It has more stuff in it than just the brain, therefore it must be more complex.  Hey, why not the human body?  Wait a minute, why not everything on planet Earth?

If you really care, the most complex thing in the universe is, in fact, the entire universe.

2.  There's a lot of stuff in the brain and we don't know how it all works yet.

This is what people mean when they talk about how complex the brain is.  More trees than the Amazon!  More stars than the Milky Way!  Stick that in your pipe and smoke it!  Numerically unfathomable = complicated.

3.  The cool part about brains isn't how much stuff is in them - it's what they do with that stuff.

I know people like big numbers (for some reason), but I think us scientists and science communicators can more effectively hook our audiences' attention by talking about what the brain does instead of how many synapses it has.

So can we let go of the needless hyperbole? 

Monday, August 12, 2013

Criticizing a New Criticism of the BRAIN Initiative

At The Stone, the NYTimes web blog for philosophy, Benjamin Fong has written a rather curious critique of the BRAIN Initiative.  His arguments are so ridiculous, they're actually interesting.  Let's have a look:

Corporations are evil:

"On one level, its proponents are simply naïve about the corporate wolves with whom they run. George Church, a genetics professor at Harvard and one of the faces of the initiative, describes his sponsors, including Chevron, Procter & Gamble and Merck, as institutions that are “very pragmatic and practical about helping our world get better.” This willful ignorance regarding corporate influence is even more disturbing in the case of the Brain Initiative, which promises a very fine control over the seat of consciousness. With the help of this research, today’s neuro-marketing – marketing researched not with focus groups but M.R.I.s – may soon look quite primitive."

1.  George Church is not the entire BRAIN Initiative, and those sponsors have nothing to do with the BRAIN Initiative as far as I know.  Church was one of the guys who helped get the ball rolling on the project but he isn't in charge of it.  So the Chevron-P&G-Merck thing is all strawman.

2.  Yeah okay I guess corporations are probably going to corporation-y things with any new research.  But doesn't this argument just go into the tired "science isn't inherently evil, its what people do with it" stuff?  I dunno.  It feels like this Fong guy thinks that a bunch of corporations invented and are supporting the BRAIN Initiative and whatnot.  That just isn't the case - the project came from scientists and was pushed by scientists from the beginning.

Biological solutions to disease are given too much precedence over psychosocial solutions:

"The real trouble with the Brain Initiative is not philosophical but practical. In short, the instrumental approach to the treatment of physiological and psychological diseases tends to be at odds with the traditional ways in which human beings have addressed their problems: that is, by talking and working with one another to the end of greater personal self-realization and social harmony." 
"We know, for instance, that low socioeconomic status at birth is associated with a greater risk of developing schizophrenia, but the lion’s share of research into schizophrenia today is carried out by neurobiologists and geneticists, who are intent on uncovering the organic “cause” of the disease rather than looking into psychosocial factors. Though this research may very well bear fruit, its dominance over other forms of research, in the face of the known connection between poverty and schizophrenia, attests to a curious assumption that has settled into a comfortable obviousness: that socioeconomic status, unlike human biology, is something we cannot change “scientifically.” That it is somehow more realistic, “scientifically,” to find a way to change the human being itself than it is to work together to change the kind of environment that lends itself to the emergence of a disorder like schizophrenia." 


1.  Alright, there's no argument from me that we should reduce low socioeconomic status in our society, or develop strategies to provide effective mental health care for those populations.  But schizophrenias are still biological disorders.  Why not study both angles?  Is his argument that the money is being spent disproportionately?  Okay, fine.  But that's not the biologists' fault.

2.  "[T]alking and working with one another to the end of greater personal self-realization and social harmony" does not cure Alzheimer's, Parkinson's, or Huntington's.

Neuroscience detracts from other approaches:

My point is simply that the attempt to gain control over life itself has severely detracted from the work of figuring out how we talk to and work with one another in order to better ourselves and our world. To be clear, I do not believe that this communicative project is easier or more efficient than the instrumental approach – how we go about changing socioeconomic conditions is a problem we have not even begun to solve – but only that it is an important part of what it means to be a human being. And no matter how precisely we can manipulate the brain with drugs, electrodes, and other such contrivances, the emerging insights of neuroscience will never provide sufficient help with this work. 

1.  How does neuroscience "severely detract" from social work?  Maybe he means that the government is spending too much on science and not on social programs/research?  Doesn't it feel like he's just saying "Spend more on humanities and not on sciences"?

2.  Neuroscience will also not cure malaria, generate a unified theory of quantum physics and general relativity, or bring world peace.  And changing socioeconomic conditions will not end Alzheimer's disease.


In closing, it's worth reflecting that none of his argument really needs the BRAIN Initiative.  Really, he just doesn't like neuroscience because... I don't know.  Some people are just reflexively anti-reductionist I guess.


Friday, July 5, 2013

There is No "Primitive" Part of the Brain

Was watching some old interview about "emotional intelligence" (itself a fraught term, but that's for another day) last night, when the interviewee (Daniel Goleman) calmly stated that emotional intelligence comes from your "primitive brain" that "arose earlier in evolution".  ARGHGHSAGHGHGH

(Click for time of offending remark)

This idea of there being a "primitive" brain crops up all over the place, and from reputable sources:

"This means that areas traditionally considered to be more primitive were just as important during our evolution." (Professor Robert Barton, Durham University, said that).  N.B. please see this comment from Dr. Barton:  As one of the people quoted, I am going to cry foul! I do completely agree with your thesis, and that was the point of my comment about "areas TRADITIONALLY ASSUMED to be primitive". I have been on a mission to repudiate the "ice cream scoop" idea of brain evolution, and have shown that cortex and subcortical structures co-evolved. So keep up the good work, but please don't attack those who agree! Robert Barton  But note that despite his good intentions, his quote eventually led to this article - the number one hit on google for "primitive brain" - to be written.  The phrase has power!

"They suggest that new learning isn't simply the smarter bits of our brain such as the cortex 'figuring things out.' Instead, we should think of learning as interaction between our primitive brain structures and our more advanced cortex. In other words, primitive brain structures might be the engine driving even our most advanced high-level, intelligent learning abilities" (Picower Professor of Neuroscience Earl Miller, MIT, said that).

"It's like adding scoops to an ice cream cone.  So if you imagine the lizard brain as a single-scoop ice cream cone, the way you make a mouse brain out of a lizard brain isn't to throw the cone and the first scoop away and start over and make a banana split — rather, it's to put a second scoop on top of the first scoop." (Professor David Linden, Johns Hopkins, said that).

Now let me explain why this is all complete BS.

First, semantics.  What is "primitive"?  How do you measure a tissue's "primitivity"?  In the common usage of the word, primitive means simple, especially in the context of a task, idea, structure, way of life, etc that was employed a long time ago.  Cavemen were more primitive than us, for example.  Unfortunately, this means that "primitive" is a word that refers to things both "ancient" AND "simple".  Which, as we'll see, is a big problem when you start applying it to mean only one of those things as occurs with the "primitive brain" meme.

Second, what are people actually talking about when they say "primitive brain"?  This is confused as well, but in general the thought is structured like this:

Most primitive - brain stem, pons, cerebellum.  (The hindbrain, basically).

Also primitive - the limbic system where "limbic" means "border, edge".  This includes the hippocampus, amygdala, parts of the thalamus (who knows why), some of cortex, and some other bits.  It's supposed to do "emotion" and this is what Daniel Goleman is referring to when he talks about the "primitive brain".  Really though it's just a lumping together of all the structures right near the inner border of cortex, because why not lump it all together?

Gloriously advanced - the mighty cortex (you know, the wrinkly part on the outside).

Third, why do people say that these particular brain structures are "primitive"?  The idea is that evolutionarily, one came after the other.  As in, the first vertebrate just had a brain stem.  Then it evolved more stuff after that.  Then as things kept evolving, more and more stuff got added on.  This is the "ice cream cone" model that David Linden espouses.  It's also incredibly stupid (or at least misleading).  Let's break it down.

Evolution did not happen like this:

Evolution did happen like this:

 I hope everyone pretty much understands the concept of a phylogeny and the fact that every vertebrate came from one common ancestor.  Yes, the common ancestor was a kind of fish.  No, today's fish aren't the same as the common ancestor.  They evolved from it just like everyone else, although "rates" of evolutionary phenomena like mutation and drift can vary and that's beyond the scope of this post.

The point is that the "primitive" brain meme is born in the idea that the brain components fish, lizards, mice, and humans share in common must be evolutionarily ancient and were likely shared in common by the common ancestor.  So, homologous structures across the phylogeny indicate "ancientness".  And "ancientness" = "primitive".  (Except it doesn't, but more on that in a second).

And since we all share structures that resemble the brain stem, voilà!  That's the most primitive part of the brain.  Here's where things go astray.

First, we don't just share the brain stem with all animals.  Here's the real "ice cream cone" of the brain:


And when I say "the" brain, I should say "all vertebrate brains".  Every fish, bird (including reptiles), amphibian, and mammal has a brain that starts out looking like the pictures above.  Each colored bump (or "scoop of ice cream") represents a region of the brain very early on in development, when the whole nervous system is basically a simple tube.  Each bump goes on to expand to varying sizes into varying kind of structures and yadda yadda depending on the species.  The point, though, is that all vertebrates have a forebrain, a midbrain, and a hindbrain.  And the hindbrain, by the way, is the "primitive" brain stem.  But clearly, humans, fish, lizards, and mice all evolved from a common ancestor that had all brain regions, not just the hindbrain.

This is why David Linden's ice cream analogy is so dumb.  He's implying that first you start with one scoop, the hindbrain, then add on another (the midbrain), and finally one more (the forebrain).  Or at least he led NPR to believe that's the case:

When mammals like mice came along, the lizard brain didn't go away. It simply became the brain stem, which is perched on top of the spine, Linden says.  Then evolution slapped more brain on top of the brain stem.
But that's not what happened at all.  All the scoops were there to begin with.  Then as adaptation took its course, different scoops got bigger or smaller or just different as you began comparing across the entire phylogeny.  Yes, humans have an enormous cortex and lizards don't.  And yes, lizards simply evolved a different-looking kind of forebrain.  That's all.

Second, homology does NOT imply "ancientness".  Even if the hindbrain looks pretty similar across the vertebrate phylogeny as it exists today, that doesn't make it "ancient".  The hindbrain has been evolving just like the midbrain and the forebrain.  Maybe it's not been changing as much, but it's still been changing.

This leads me to why the "primitive" notion is so misleading, and should be avoided:

(1) Calling a part of the brain "primitive" suggests what David Linden articulated: that brain evolution happened like stacking scoops of ice cream.  It implies that our brain stem is no different than that of a lizard, or of a mouse, or of a fish.  Yet despite their vast similarities, they are clearly not the same.  You can't plug a human forebrain into a lizard hindbrain and expect the thing to work.  The hindbrain of humans HAD to adapt to having an enormous forebrain.  There's something seductive in the idea that inside all of us is a primal being, a "reptilian brain".  There isn't.  It's a human brain, top to bottom.

(2) Calling brain parts "primitive" because they are shared across phylogeny is often used to justify how amazing our human cortex is.  Look at what makes us us!  We're so great!  Well, I guess.  But we are just one little excursion among many that evolution has produced.  The lizard brain is adapated for what lizards need to do.  The fish brain is adapted for what fish need to do.  They don't have "primitive" brains.  They have highly adapted brains, just like any other vertebrate.

(3) Simply using the word "primitive" makes the casual reader think of cavemen.  It just does.  And that's even more ridiculous, because ancient Homo sapiens were still Homo sapiens.  Read what this poor misinformed blogger has written:

So, let me explain the Primitive brain in simple terms. We have an Intellectual (rational) part of the brain and a Primitive (emotional) part of the brain. In the picture above, the Primitive brain is around the Hippocampus and Hypothalamus areas. In some texts, it has also been called the Limbic System.
The subconscious Primitive part has been there ever since we were cavemen and cavewomen, and houses our fight/flight/freeze response (in the amygdala in between the Hippocampus and the Hypothalamus). Its function is to ensure our survival.

AHHHHHHHHHHHHHHH.  You see?  You see??????

 (4) There is not just a "primitive, emotional brain" and a "complex, intellectual brain".  That is so.... wrong.  Factually wrong.  Yet people like Daniel Goleman sell books about emotional intelligence claiming that people need to develop their "emotional brain" and then bloggers like Carrie (above) start internalizing and spreading the misinformation.  Folks.  Let me be clear.  You have ONE brain.  One.  It has many parts, which is to say that humans looking at brains have found ways to subdivide them into various areas and regions and structures and whatnot.  Regardless of all that, the whole damn thing works together.  It's one big circuit that does emotion, rationality, sensation, movement, the whole shebang.  There isn't a simplistic "emotion" part and an "intellectual" part.  The cortex is involved in emotions and feelings.  The basal ganglia are involved in cognition.  In fact, the whole idea of splitting emotion and reason into two separate camps is fraught, as emotion turns out to be critical in reasoning tasks like decision-making.

(5) The "primitiveness" of lizard brains is vastly overstated.  Things like this get written about the "primitive brain":
A lizard brain is about survival — it controls heart rate and breathing, and processes information from the eyes and ears and mouth.
This implies, to the casual reader, that lizards are just sitting around breathing.  Maybe there's some "survival instinct" in there: look out for that big hawk!  Yeah, okay.  But guess what?  Lizards gotta do other stuff too.  They have to reproduce, find food, navigate their environment, learn, remember, make choices, etc.  They aren't just breathing and instinct machines.  And because they aren't, that means their brains aren't just doing that either.  And why is it always lizards and reptiles?  You'd think fish would get picked on more.

(6) "Primitive" in the context of a brain part primarily means "ancient".  But the word "primitive", as we already saw, connotes simplicity.  This leaves laypeople with many misconceptions.  First, that the brain stem, or the "emotional brain", or whatever is simple.  Or even that they're simpler.  Nope.  Not really.  Pretty much every part of the brain is complex.  Second, it reinforces, in the case of the "emotional brain", that emotions are beneath intellect.  They came first, they are older, they are simpler, they are the stupid part of your brain.  Again, just no.  You need emotions to survive just as you need your intellect to survive.  Fish need their emotions (an emotion, after all, is just a representation of bodily state imbued with a certain postive/negative valence) just like they need their reasoning abilities as well.

(7) People using "primitive" as scientists do it because it can sound cool and surprising.  Look at how Earl Miller framed it, from above:

"They suggest that new learning isn't simply the smarter bits of our brain such as the cortex 'figuring things out.' Instead, we should think of learning as interaction between our primitive brain structures and our more advanced cortex. In other words, primitive brain structures might be the engine driving even our most advanced high-level, intelligent learning abilities"

Look at that result!  A primitive thing did something advanced!  The forgotten thing is important!  Or maybe - this is going to sound crazy - the whole system evolved together in order to support such essential tasks like learning.  There never was a primitive part or an advanced part, despite two areas or regions being labeled as such.  Every part of the human brain has been evolving for the same amount of time as every other part, and has been evolving to work as best as possible with each of those other parts.

(8) Finally, let's return to Daniel Goleman, who argues that "emotional intelligence" arises from the "primitive emotional brain".  Then he waxes on and on about the value of emotional intelligence, particularly as it relates to social abilities.  Ability to understand your own emotions.  Ability to perceive those of others.  Ability to interact well with others on the basis of understanding their emotions.  Et cetera.  That's all fine, but by saying this comes from an ancient, primitive, emotional brain might make people think that ancient vertebrates really had to know themselves, be able to read others, and interact socially.  But there's a whole lot of solitary, nonsocial vertebrate species out there folks, and they have brainstems and limbic systems too.

Okay folks.  Hopefully never again will you refer to a part of the brain as "primitive."  Some structures probably more closely resemble their homologous counterparts in the last common ancestor of vertebrates, but all the basic parts were there from the beginning.  And remember, evolution didn't happen just to make humans.  We aren't more advanced in an evolutionary sense than fish, lizards, or mice.  Each species is just adapted to the roles it finds itself in, and continues to adapt.  Our sense of being "advanced" comes purely from our own self-regard and anthropocentric tendencies. The human brain is not the best brain, nor is it the most advanced brain, because there's no scale on which to measure how good a brain is.  

To reiterate: we don't have primitive brain parts.  Your urges, desires, rages, and other feelings you think you shouldn't have can't be excused away to a "primitive" brain.  You are your brain - the whole thing.