Stanford University.
--push off into new terrain, here-- yet another bucket.
But before doing that, various bits of feedback
from office hours.
Hearing from TAs.
Getting a sense of where the grand, gaping craters
of confusion are so far.
And apparently an awful lot of them
were provided in the last two days of lectures
with behavior genetics.
Various issues that came up.
Clearly, one of the most complicated, inaccessible,
subtle, pain in the neck concepts
in the whole class, which is this heritability business.
So memorize the following two sentences,
because it all comes down to the difference between something
that is inherited and how heritable a trait is.
The fact that humans overwhelmingly
have five fingers reflects the fact that the number of fingers
is an inherited trait.
The fact that, when there are some circumstances of humans
having other than five fingers, it
is overwhelmingly due to environmental
something-or-others is an indication of the fact
that, nonetheless, variability around the number of fingers--
heritability is essentially 0%.
So get those two sorted out, and you
have those two concepts all under your belt
and very useful.
Why is it useful?
First off, why is this one useful?
Because it is important to know the distinction.
Everybody, for one thing, out there
tends to view these as telling the same thing.
And insofar as they think it says the same thing,
they think that both terms are referring to this.
Something that is genetically regulated, genetically
determined-- whatever.
Two totally different terms.
So one reason to obsess over this
is that when the newspapers give us
our sound bite of "scientists report that,"
it's usually this one that they're reporting,
with one of those marvelously misleading numbers.
Which then sets people up for thinking they have just
been told how much do genes "determine"
some average feature of this trait.
So, important to tell them apart.
The other reason of getting this sorted out
is not to only unlearn the nonsense aspects
and misinterpreting this term but because understanding
that term gives you a lot of insights
into when and how you are getting
gene-environment interactions.
Caveat with that.
Saying "when and how you're getting"--
you're always getting gene-environment interactions.
Remember the quote from the other day.
It's like saying, which contributes
more to the volume of a square, or the height, or the length,
or the roundiness, or whatever?
Yes, it's always gene-environment interactions.
Saying, in what ways are there some
of the more interesting ones, dramatic ones, in what realms--
Understand this, and you avoid this confusion.
Understand this, and it gives you
insights into gene-environment interactions.
OK.
Other issues that are coming up.
By now, we've looked at three different broad approaches
to the biology of social behavior.
The evolutionary stuff-- broadly stated-- the molecular,
the behavior genetics.
And what should be clear by now is
if you were living inside one of those buckets,
you hate each of the other two and trash them entirely.
So we get to the first of our great, conflicting points,
which is, does that mean one of these is wrong?
No, one of them is not wrong.
None of them are wrong.
Some are more right than others, and some
are a lot better than others.
But nonetheless, these are all-- again,
from the very first lecture-- different levels
of description.
Another way to begin to think about this, in terms
of what we've now been focusing on--
and this sort of coming up from some questions afterward.
So by now we've had the term "epigenetics."
That's come in in the last couple of lectures.
And there are at least three different levels,
three different buckets, with which you can define the term.
What is epigenetics?
Epigenetics is the way the culture, environment--
all of that-- affects biology.
That's a certain broad way of stating that.
NodA stated that way, you are making
biology synonymous with genetics, which ain't so.
But nonetheless, that's a certain broad level
of stating it.
In another discipline, a very sort
of first pass at molecular stuff,
what is epigenetics about?
It's the way in which environments turns genes
on and off.
And at another level of explanation, more reductive,
what is epigenetics about?
It is regulation of chromatin remodeling and methylation
of genes and all of that.
Do not panic, if that's not the level
you want to know it about.
That's the business from the other day,
about changing access of transcription factors
to DNA-- jargony way of doing it.
Don't worry.
The main point being that this is a completely different level
of defining this.
This is the whole point in here.
We are beginning to see different disciplinary
approaches.
We are beginning to see where one discipline has decided
they have answered a question.
This is how culture affects biology?
Give me a break.
Show me which genes we're talking about!
We're talking about genes, here.
Give me a break.
Show me-- is this chromatin remodeling-- what's
the mechanism for it?
How reversible is it?
One discipline's answer is the next one's starting point--
blah, blah.
One discipline's wonderfully satisfying scientific answer
is the basis for the next discipline
to be totally contemptuous of them.
You call that science?
This is sort of the whole point, here,
beginning to see how we could chip out a way between these.
More bits of clarification.
Amid that, one of the things that really came through
is-- in the sort of ratio of praising
to trashing-- I was clearly spending a lot of the last two
days trashing the behavior-genetics approaches.
So a little bit of clarification, there.
First off, we can broadly divide what
came the last two days into classical behavior genetics.
And that's the comparing monozygotic and dizygotic
twins.
That's the adoption studies, that's
the twins adopted at birth-- that's all of those approaches,
there.
Those are all the ones where you were just inferring really
indirectly that there's something genetically going
on there.
The other half, and much more sort of the modern behavior
genetics, is marrying these traditional approaches
with molecular biology.
And that was the business at the end-- you know the gene,
and you kind of have an idea of what it might do.
How does that map onto behavioral variability
in humans?
You know the behavior, and you've
got some sense of its variability.
How does that map onto variability in genes?
This is the much more powerful, contemporary end of it.
So what is this end good for?
All it's good for is pronouncing that something's genetic,
or it's 73% genetic and then you trash it
because that's jibberish.
What it's good for mainly is--
OK, so you've got an adoption study--
classical behavior-genetic study--
where they're adopted right at birth, within seconds,
and raised in different households.
Oh, you haven't ruled out environment!
Don't forget prenatal environment.
Oh, you haven't ruled out environment!
Remember the nonrandom assignment
of adoption-- all of that.
Does this mean this approach is useless?
No.
What it's good for is demonstrating nonetheless well
we've just ruled out all sorts of realms of environment
that people would guess is consequential.
It's less consequential than you think.
We haven't ruled out environment entirely.
And just because it's all gene-environment interaction,
we certainly can't come up with a stupid number like that.
But what this is good for is at least showing,
here's domains where people-- a lot of people--
would have assumed there is big environmental effects.
Much less than you would think.
So that's the much more conservative, sort
of sobrietous thing that people can do with that field.
The fact that far too few of them actually
do that is reasons to trash-- no, well, nonetheless, there's
lots of good things in behavior genetics.
But that's the limited domain where it's useful.
OK, so just beginning to get a sense, here,
of the various things that are confusing.
Obviously, no one on earth in any exam in here
is going to be asked to choose which field is better
and which field gives more-- you know, kum by ya
and all of that.
But just recognizing the different approaches
and the wonderful rainbow diversity of ways
to think about mating in fruit flies or whatever.
And just beginning to see by now,
this is what the whole class is about.
Speaking of that, I recall from the first class on that.
There's somebody in here who I think was an English Lit
grad student.
If you have a chance, email me about how it's going in here.
Let's see.
What else?
Next week.
[LAUGHTER]
No, I'm curious.
I am very pleased whoever you are--
assuming you haven't fled after the first class-- is in here.
Let's see.
Other stuff.
OK.
Schedule.
Next Monday, this coming Monday, we
will have a lecture on yet another of our disciplines,
ethology.
And we will see that's a totally different way
of doing it-- blah, blah.
Wednesday, Friday, and the following
Monday are the catch-up lectures in class, taught by the TAs.
Two of the lectures, introduction
to the nervous system.
The third one, introduction to endocrinology.
It will be a broad overview.
It is explicitly designed for people who have
no prior background in this.
But what I think is probably a good idea is,
even if you believe you have prior background in it,
maybe assume it might be a good thing to get a refresher on it.
And this will be very useful.
Following that, we will have two more lectures--
more advanced topics and basic features
of neuroendocrinology-- and then staggering into the midterm.
And then-- oh, then-- the second half comes.
Just a sense of what's coming.
OK.
So that's where we're at.
So now we transition to the next subject,
here, which is-- no, actually, before I forget--
how many-- did people see this-- the article in the New York
Times this morning about the study-- the Chutes and Ladders
study?
Did anybody see it?
It was posted this-- nobody saw it?
This was-- you know the game Chutes and Ladders?
OK, we all played Chutes and Ladders.
This was this massive study funded by the World Health
Organization where what they did--
this is going to be the definitive study
on the subject.
They showed that people from Nepal
are better at playing Chutes and Ladders than are
people from Belgium.
[LAUGHTER]
Cool study!
I don't know how you guys missed it.
OK.
You need to know about this.
This is important.
What do you want to know about this study?
This massive study shows that people from Nepal
are better at Chutes and Ladders than people from Belgium.
Ask me questions, since you were terribly, woefully
underinformed about this.
What more information do you need
to be impressed and tell everybody about it
in the dorm tonight?
What else you want to know about this study?
[INTERPOSING VOICES]
"What does 'better' mean?"
OK, social relativist.
And does it mean that you learn more about yourself
playing Chutes and Ladders?
[LAUGHTER]
Does it mean that you make the world a happy-- you win!
You win, win, win!
OK, so winning is the end point.
Good question.
What else do you want to know about the study?
Yeah.
Why?
Why?
Why?
You ask that?
Read Aristotle!
Since humans first pulled out of the mud,
this has been the thing we have wanted to know.
[LAUGHTER]
And now we do.
[LAUGHTER]
Son, the difference between knowledge and wisdom--
OK, so, why?
OK, what else do you want to know?
What are your questions?
Yeah.
What the methods of the study were.
The methods.
Very good.
OK, various methods.
Do you want to break it down into more details?
Given some of the critical tools you
have by now, in terms of methods,
before deciding how impressed you are or not with this study.
What sort of questions, now, in the methodological realm.
Yeah.
What population they're sampling?
Ah, very good question.
Because we see a first methodological issue.
Are you getting a decent sample size,
so that you're confident that you can actually say something
about the population at large.
And that's a great question.
It was done on everybody in Nepal and everybody in Belgium.
[LAUGHTER]
Good study!
OK, what else do you want to know?
What other questions?
Yeah.
In terms of their method, did they pit people from Nepal
against people from Belgium in one game, or--
Well, that's an excellent question.
Every single person in the study from both Belgium and Nepal
played one game against every single other person
in the study.
[LAUGHTER]
That's why you haven't been hearing
much about either country in the news lately.
They've been very busy with this study.
OK, so, good methodology, large sample size,
random assignments of games.
What else you want to know?
Yeah.
Why-- the question and subject to ask worth reading about,
and why did they [INAUDIBLE]?
Why'd they-- OK, that one again.
It's self-evident.
Know thyself, or know the Nepalese and the Belgians
or something.
This is a-- I'm glad I know this now.
I'm glad they went and did that.
OK, what else do you want to know?
Obviously, a matter of subjective
taste as to what counts as an important scientific subject.
this strikes me as critical.
What else do you want to know, in terms
of what sort of conclusions?
They kind of hint-- they're not positive--
but they're kind of hinting at a genetic component to it.
Yeah-- question in the back, there.
Yeah.
What skills are involved in Chutes and Ladders?
What skills are involved?
It involves, um-- what is it?
They have a whole list, there.
Um-- let's see.
Various tasks involving spatial memory.
Various tasks involving reversal performance and telekinesis.
[LAUGHTER]
So it taps into all of those domains.
What else do you want to know about the study?
Yeah
Is it inheritable?
Is it heritable?
Great-- oh, ho, ho, is it heritable!
So which are you asking about?
OK, well, they did it.
They are suggesting there's some degree of heritability.
And they did it right.
They went after all the issues we've learned about by now.
Specifically, what I'm saying is that every single person
in Nepal and in Belgium was cross-fostered to somebody
in Ecuador.
[LAUGHTER]
In fact, they did it fetally, right after conception.
[LAUGHTER]
So they can-- the prenatal stuff.
What else do you want to know, in order to decide,
am I impressed about the fact that people from Nepal
are better at Chutes and Ladders than people from Belgium?
Yeah.
Were they all good at the exact same amount
of Chutes and Ladders as children?
Oh!
Yes.
In fact, none of them had been exposed to it before.
They were all raised in hydroponic gardens,
without Chutes and Ladders--
[LAUGHTER]
--and they were only released in time for doing this.
OK what else?
Yes.
What's the amount of variance within each individual group?
Oh, good question!
Here we are, getting to that-- all of that.
Everybody got the exact same score from Nepal,
and everybody from Belgium got the exact same score.
So that's kind of impressive.
Are you impressed yet?
Are you impressed enough yet to go run out of here,
screaming with this news?
What else do you need to know?
Yeah.
Who went first?
Who went first?
They randomized it.
They randomized it by-- the entire populations played, um--
oh, what is that--
Roshambo.
Roshambo, yes.
They played roshambo.
And, best of all, they had to play roshambo in Esperanto.
Yes-- more.
What was the environment in which they test [INAUDIBLE]?
OK, they were released from the hydroponic garden,
and then all of them were placed in a completely sterile bubble
environment, where all they had in there
were copies of People magazine in a language they didn't
understand.
So it was well--
[LAUGHTER]
Environment was very well controlled-for.
What else do you need to know?
How much better were they?
Like, 0.001--
Ah!
Ah!
There we have it.
They had a huge sample size.
They cross-fostered as fetuses.
They controlled for environment.
They got everybody in there.
They did the right techniques.
They randomized at every possible choice.
That is so impressive.
That is so impress--
How big is the difference?
Is it a big difference?
And this is a critical thing to start putting there
in your armamentarium of critical questions,
to start having skeptical ones.
Great.
You've got a whole bunch of tools by now.
Whoa!
They said "genetic."
Did you control for this?
Did you control for that?
Whoa, they see a difference, there.
Well, wait a second.
Did they all have the same exper-- whoa,
did they have a big sample size?
All of it covered.
Great, wonderful, perfect science.
They get every single bit of it covered, there.
But then the critical thing you better ask
is, how big is the difference?
Is it impressive or not?
Because we all think we've got the basic tools-- or hopefully
we all have the basic tools-- for going at these issues.
OK, was a study done in a way that
was clear-cut and unobjective?
Were people blinded to whether or not
this person was Nepalese or Belgian
or-- was there appropriate blinding?
Was it done in a way that you can falsify your finding--
in a sense, the definition of experimental science?
Was it designed in-- was it independently
replicated by anybody else?
We all have those tools under our belt by now.
But far too often, we are not trained, at that point,
to say, wait a second.
Before I get all excited, how big of a difference was it?
Let me give you a real example of this.
And this was a paper, about three years ago,
in the journal Science-- which you should get a sense by now
that the journals Science and Nature are
the two official biggie ones on this planet,
in terms of credibility.
And this was a paper having to do with IQ.
And this was a paper having to do with IQ and birth order.
And what they showed in this paper,
with spectacular statistical confidence and as much
of, like, all the controls in place
as you can ask for from the hydroponically gardened fetuses
in Nepal-- they did absolutely perfectly.
What they showed in the definitive study,
with 250,000 18-year-olds in Norway,
was that there is a reliable IQ difference
depending on whether you were firstborn or latter-born.
OK, how many of you are firstborn?
Whoa!
How many of you are only childs?
OK, how many of you are number 2?
How many of you are more than number 2?
OK, so, how many of you think the highest
IQ comes from the firstborn?
[LAUGHTER]
OK, and we'll assume the converse with the other group.
Did anybody just vote against their own birth order?
[LAUGHTER]
Whoa-- OK.
Well, that's-- impressive.
We salute you.
OK, so they found a difference.
They found a difference, which is,
firstborns-- in this very statistically reliable way--
have higher IQ than latter-borns.
They restricted the analysis just to second-borns
to keep things clearer.
And they showed firstborns have higher IQ than second-borns.
250,000 people-- as close as you can
get to all the people from Nepal and Belgian.
Huge sample size.
So they report this.
This was the most thorough study ever done.
So what sort of questions do you want to ask?
Give me hypotheses for what's going on.
Obviously, sort of one is that, like,
the parents pour protein into the ears of firstborn
and the second one just gets Fritos or something.
But give me other hypotheses for what could be happening.
Possible explanation for this, in terms
of biological, sociocultural, endo, immuno, psycho-- yeah.
I think we already know that firstborns tend to be more,
like-- well, I mean-- it's like the parents-- they're more
worried about making mistakes, and they kind of raise them
harder, so to speak, and then they sort of relax
on the second-born child. [INAUDIBLE]
firstborns tend to go to college more and stuff like that--
do more traditional [INAUDIBLE].
OK, so, great hypothesis.
It's more parental investment in the firstborn,
and those are the ones with the parents freak out
with everything.
And by the time there's a third one,
they're all foraging on their own
when they're six months old.
[LAUGHTER]
OK, so, parental investment.
They went after that one correctly.
Here's what they showed, to rule that out.
Which was, if you are an only child,
you have a lower IQ-- on the average,
blah blah-- then firstborns who have younger siblings.
So it's not parental investment.
It's something about being the firstborn of multiples.
Yeah.
[INAUDIBLE] firstborns are expected
to kind of like [INAUDIBLE].
OK, so pressure on the firstborns to be first born.
How would that raise IQ?
Um-- all right, just kind of like-- kind
of like responsibility [INAUDIBLE]?
OK, so that's one of the models out there.
A variant on that is, firstborns get
pushed into a tutoring position, early on.
And the well-known fact that occasionally now
within teaching something actually
causes you to know what you're talking about.
So the firstborns, because of this tutoring role-- oh,
that would control for the single child
versus the firstborn difference, there.
So maybe it's a tutoring phenomenon.
What else could be happening?
What other possible notions?
Yeah.
It depends what age you're looking at the kids at.
Because if you're looking at, like, actual children
and comparing IQs, older children
will probably have higher IQs because they've had longer time
in life to learn.
OK.
Great idea.
Up to age 12, latter-borns tend to have
higher IQ than firstborns.
By age 18, which is when this study was carried out,
it flips the other way around.
So, any hypotheses for why this difference occurs?
Why, by 12 years of age, latter-borns
tend to have higher IQs than firstborns?
Why does that then flip afterward?
Yeah.
[INAUDIBLE]
OK, so is it IQ testing-- some biases with that.
What would rule against that, though,
is that the age controlling for only
childs-- only children-- versus firstborns.
So that would definitely be a possibility.
That was ruled out.
Yeah.
I think that I [INAUDIBLE] in another class
that you said it had to do with the ratio of adults
in the environment when the kids were growing up.
So when there's only one child, you
have more adults with higher language skills, and--
Good.
OK, so another version of a parental-investment model,
there, which is, the fewer the children, the more
parental energy.
So, again, that's one of the standard things in the field.
They ruled that out with comparing the only children
versus firstborn child.
So that's been a dominant model in the field,
so they had good data against that.
Yeah.
Could it be the age of the mother
when the baby is a fetus?
Oh, OK!
What's your idea about that?
Well, younger mothers-- maybe the fetal environment
is better, and obviously older children
have younger mothers than their siblings whose-- the mother
[INAUDIBLE]?
OK, so we've got an intrauterine effect.
So we've got egg quality and age of mom and all of that.
They controlled for that-- the age of the mother.
Can anybody think of something-- another intrauterine mechanism,
though-- for what's the difference between being
the first fetus who hits the womb
of your mother versus being the second or third or fourth?
What's one of the biological things that might happen?
Yeah.
Stress levels.
[INTERPOSING VOICES]
OK, stress levels, which is a way of stating something
about intrauterine environment.
The more times you've done this, perhaps more
stressed you are in there.
What else could biologically-- that's definitely
one of the things.
What else biologically can happen, over time?
Remember progesterone, the other day,
from that lecture of making new if-then clauses
and glucocorticoids.
And this is a bit of a stretch.
This takes some sort of fair amount of OB/GYN knowledge.
What's a danger as you have more and more pregnancies,
in terms of your immune system?
[INAUDIBLE]
Yeah-- wait-- I heard that mumbled!
Sure!
It gets surpressed?
Yeah.
OK, immune suppression.
So mom could be getting a lot less healthy,
which is another version of tapping into that notion,
there, in terms of egg quality with age.
Separate of age, the number of times you've gone through this.
What could be another possibility, immunologically,
with repeated pregnancies?
Yeah?
A mother [INAUDIBLE] more, like, form antigens [INAUDIBLE]?
OK, despite that immune suppression, on the average,
with later pregnancies you have a greater likelihood
of having formed antibodies against aspects of the fetus.
They controlled for that.
They showed, if you were second-born
and there were kids after that-- if you were second-born
and the firstborn died, you reverted to the firstborn IQ.
Showing that it was not anything about,
oh, you were the second fetus in there, with more antibodies.
They controlled for that.
That's actually one of the ideas they brought up in there.
What else?
How about that business about up to age 12,
the second-born does better?
After that, by age 18, the firstborn does better.
Any theories with that?
Yeah.
Did it have to do, maybe, with how fast they grow,
or something?
I know we talked about how for fathers it's,
like, a child grows faster [INAUDIBLE]
more later [INAUDIBLE] grow faster then
that takes a toll on your [INAUDIBLE] health?
OK, so an early advantage, and you pay for it later,
type deal.
So that's a possibility.
What else could come in?
Think about-- back to the idea, there,
about if you're a firstborn.
Some of the firstborn responsibility stuff.
How that plays out in family dynamics early on.
Why are the second-borns doing better
in the first dozen years?
Why is it reversed later on?
More ideas about that.
Yeah.
The second-born benefits from the tutelage
of the firstborn early in life.
But later on, having the older having had that
experience of being more responsible [INAUDIBLE].
Exactly.
That's one of the main theories.
That's probably the predominant one proposed
to explain that age switch.
You get the second kid show up, and suddenly you
have a neotenized, dumbed-down environment,
where suddenly the 8-year-old kid
is watching Tinky Winky again for the first time in six
years.
That it's an environment that's then dominated more
by having a younger one around.
And what the older one is mostly doing is the tutoring.
And it takes a number of years for the advantages of that
to finally come through.
That's the main model that's given for that.
Any other ideas?
How about parental-resource stuff?
The fact that the larger the family, on the average,
in westernized countries, the lower the socioeconomic status?
Run with that one.
Where does that fit in?
What else?
What else could be happening with that?
Yeah.
If the socioeconomic status of the family is lower,
then wouldn't that suggest that the children might
bear more burden, in some respects, with more children
around, so that they had to grow up more quickly
and be exposed to the real world?
It's the later ones who are out hunting squirrels and not
getting the violin lessons.
That's another version of the parental-resource one, one
being because there's a smaller ratio of parents
to the kids, the other being because the more kids, the more
expensive for the same parental income.
And you have got less to spread to each child.
So they controlled for that, looking at within family
rather than just between family.
They covered all of this.
This is going to be the definitive study
for the rest of all of time, showing what's
going on with IQ by birth order in 18-year-olds in Norway
in 2007.
What was the magnitude of the difference for this study?
2.3 IQ points.
And thus, coming back to what could
have been the very first question sitting there
when these guys were ready to announce this
to the world and sort of start selling
their "be like a firstborn IQ" self-help books and all
of that.
Lost in there-- and this was picked up all over the press
and, like, no doubt, endless snotty comments by David
Letterman or--
And nobody-- 2.3 IQ points!
You sneeze while you're taking an IQ test
and have to wipe your nose for eight seconds afterward,
and that's going to cost you 2.3 IQ points, because you
get distracted for a second.
An amazing example of this whole business of-- yes,
impeccable science that these people did-- phenomenal!
I don't know who possibly gave them the money
to do a study like this.
And at the end of the day, totally cool,
irrefutable, statistically totally reliable--
which is very different from saying "important."
But what you get at the end of the day
was this mammoth study producing 2.3 difference.
And this is a great demonstration.
The difference between how solid the science
is, how confident you are of the finding-- which
takes in all the variables of sample size and objectivity--
how confident you are of the finding,
and how big of a finding it is.
And those could be two entire differences.
So, as a [INAUDIBLE] you to all the stuff that's
going to come in the second half of the course, one
of the next tools you have to have in mind,
in addition to all the ones that were apparent
here in the questions you were asking-- another one
is to keep saying, well, is this a big effect?
"Is this a reliable effect?" is different
from "Is this a big effect?" "Is this a consequential one?"
So, another tool to have in hand.
So, with that in hand, go and tell everybody tonight
about Chutes and Ladders.
OK.
So what we jump to now is something
that's been running through a whole bunch of the lectures
already.
OK, we've got all those evolutionary models
of individual selection and kin selection and in-group
and out-group.
And we've got something about the molecular biology
of why it is that you share 50% of your genes
with this relative and 25 and 12 and a half, and all of that.
And somewhere in there, lurking through all of it,
is a question which is what we'll
focus on now, which is, how do organisms, how do animals,
how do individuals recognize relatives?
Because you can't do any of that evolutionary-biology,
kin-selection-theory stuff, where it's all predicated
on degree of relatedness, unless you know how related
am I with this individual.
So what we're going to be looking at, here, is, why--
or how-- sure, let's go for "how" instead of "why"--
how do animals, how do social animals, recognize kin?
What's clear is, it does not take a very fancy organism.
And there was a great example of this,
which I think I stuck in the extended notes, last minute--
a paper published just a couple weeks ago looking at deer mice.
Deer mice, and much like their vole cousins--
if they are cousins-- some deer-mice strains are
monogamous, and some are polygamous.
This appears to be a frequent theme
in these little rodent things.
And with the deer mice, what you find
is, with the polygamous ones one female mates
with a bunch of males.
And as a result, one female will have
sperm from a number of different males afterward in the vagina.
And what you get is, evolutionarily,
from all the rules we learned by now, perfectly logically,
you get intrasexual competition between the sperm
from the different males.
We already heard a version of that
with the flies, back the other week,
there, where the sperm of one releases
toxins that kill the other sperm but in the process could
damage the female's future-- all of that.
This is a theme that runs through a lot of the literature
on sexual competition.
Sperm competition.
And there's even hints that something like that
goes on in humans.
OK.
So what form does it take?
In these deer mice, as follows.
I don't begin to understand the mechanics of this,
nor do I want to.
But apparently, with deer-mice sperm,
if they all clump together-- you know,
many hands on the oars, or something.
If they all clump together, you get this macro sperm thing
which swims upstream faster.
And the paper had all sorts of diagrams
of this which I did not want to look at in much detail.
But there's this--
So you've got-- with the polygamous strains,
you've got this problem.
If your sperm want to do things absolutely correctly,
they only want to form one of these big-old,
you know, pleasure-boat aggregate crew things,
with sperm from themselves-- with sperm from only
themselves.
And, following all of our theorizing,
to a lesser extent with close relatives and not at all with
sperm from some other guy.
And that's precisely what they showed in this paper.
You take sperm from monogamous strains,
and you put sperm from different males together,
and they all happily form this big cooperative clump of sperm,
there.
But you take them from species that
have evolved under the selective pressure of polygamy,
and the sperm there know who they're related to
and will form these clumps only with the ones from themselves.
You can immediately design all sorts of lock-and-key stories
for how it's pulled off.
You could immediately come up with some approximation
of what the molecular mechanism would be.
But for a first pass, what's striking here is, oh,
how do organisms recognize relatives?
There are, out there, single cells
that can do this under exactly the evolutionary
circumstances-- models that we've got already.
So, as we begin to look, now, at how whole organisms do it,
even cells can do it.
And we're going to see lots of different possible mechanisms.
In lots of species, what you have is some equivalent of what
those single sperm are doing, which is,
there is innate recognition of relatives.
How do you show this?
We already know the classic ways of doing this, which is
the crossed-fostering approach.
You take a litter, newborn litter, of your rodents,
and you cross-foster them to other females.
And if, later on, they can behaviorally
differentiate between their siblings and nonsiblings,
there's something innate about it.
Oh!
Wait a second, wait a second.
What about prenatal environment?
Wasn't there something about-- so now you
do the prenatal cross-fostering--
the fetal-transplant approach-- and you get
the exact same thing.
There is innate recognition of relatedness
in all sorts of rodent species.
Another way of doing it even cleaner.
You have two different litters from the same mother and father
rodent.
And then they meet together.
So there was no shared prenatal environment.
And you can show recognition there.
You put the rat, later on, into the cage where
there is the urine of its sibling,
and there is the urine of a perfect stranger,
and they will prefer going to that one.
You could show that it's even more subtle than that.
They will prefer to go to the urine of a full sibling
versus a half sibling, a half sibling versus a first cousin--
all the way on down.
They could take it out to about third or fourth cousins.
Incredible discrimination, there, that can go into it.
And it has to be that way, or else all of the theorizing from
the other day-- you can't figure out how to give up your life
for two cousins or eight brothers or eight brothers or--
[FEEDBACK]
--two cousins, or whatever the math
is, unless you know who's who.
And in some species where it's done entirely instinctually,
that would be the way you demonstrate it.
So what's the mechanism, , there, in those cases?
The most-studied ones are olfactory.
Olfactory signatures.
"Pheromones"-- we've already heard that term in here.
We're going to hear tons more about it.
But pheromonal communication.
What does that begin to require?
If you have pheromones, odorants, coming out of, say,
the urine from different individuals,
telling you your degree of relatedness to them,
it requires two things.
It requires qualitative differences in the urine,
reflecting the genetic makeup of the individual who provided
that urine to the grad student.
And it requires some mechanism, some olfactory brain-processing
mechanism, to be able to pick up whatever those differences are.
And both have been shown.
OK.
The way you begin to get the differentiation
at the end of how the urine smells differently.
What you've got-- and referred to back
with the transposable stuff is-- you remember,
in vertebrates you've got some of your highest rates
of transposable events in your immune system,
your genes devoted to immunity, where you juggle them around.
And that's how, with any luck, you come up
with an antibody that will recognize this completely
novel pathogen. All of that.
There's an additional stretch of genes
in that neighborhood where what happens with that is,
it also undergoes huge amounts of splicing and transposition
and juggling and all of that.
And what do you do then is, you create a completely unique
protein.
You do it in enough of a combinatorial way
that it would take statistically 400 quadrillion
googleflex-whatevers to come up with another organism
with the exact same protein signature.
When you make a protein from that,
you have made up one that no other organism on earth has,
with a very high statistical reliability.
This is a stretch of genes called
the "major histocompatibility complex"-- MHCs.
And what you see with those is histocompatibility,
that whole business with organ transplants-- how
compatible of a donor is it, how closely related,
how much of this-- jargon, for those of you
who know it-- how much of a shared antigenic determinant,
how genetically similar is this fingerprint, this identifying
ID of a protein?
That determines things like histocompatibility--
how well organ transfers work.
That's the origins of the term.
So every single one of us, every single organism out there,
has made an arguably unique juggle of these genes
and comes up with this signature protein
that it sticks on the surface of every single one of its cells.
What's that good for?
That's good for your immune system
learning, if we run into a cell with one of those things,
it's us.
Don't attack it.
Don't form antibodies against it.
And if we run into anything else in here that
doesn't have one of those, it's an invasive pathogen,
and go attack it.
This is the basis of the self/nonself recognition
ability of the immune system.
And what autoimmune diseases are is when your immune system
screws up and begins to mistake one of your signature proteins,
your major histocompatibility gene
derived signature protein, as, in fact, being invasive.
And one of the other things you hear about-- the other day,
we heard about this tropical parasite, trypanosome.
What it does, as you heard, was it
keeps juggling its surface proteins.
So just as your immune system is all set to attack it,
because it's got antibodies to recognize it,
it has changed its signature protein.
There's another tropical parasite, schistosomes,
where what they do is they steal your major histocompatibility
proteins from the surface of some of your cells
and glue it on themselves.
And they are wolves in a sheep major histocompatibility
proteins or some such things.
So that's one major domain where this unique protein-- derived
from a unique gene-- unique protein signature
lets your immune system work properly.
So it turns out there's another whole domain with it, which
is, these proteins can become soluble.
Which means they're no longer anchored to a cell,
they're just floating around.
And ultimately they're floating around in your saliva,
in your urine, in your armpit exudates or whatever it is.
And what they begin to do-- complicated mechanisms, which
I think I will bypass.
What they do is give a unique signature to the pheromones
coming off of you.
And, as we will see in the lectures
to come, animals of all sorts of species can tell,
is this individual of my species,
are they the same gender, are they an adult,
are they sexually mature, are they healthy,
are they pregnant-- whatever.
But thanks to this major-histocompatibility
business they can also tell, is this a relative?
Now the juggling of events, the splicing and the juggling
of the genes, has some degree of statistical relatedness
the more closely related you are.
In other words, you smell your own urine--
if that's your hobby-- and the major histocompatibility genes
in there will obviously exactly match your own.
You smell those of a full sibling--
even more questions to be asked-- and you do that,
and there will be a greater degree, statistically,
of similarity of the structure of that protein
than with a second cousin, than with a perfect stranger,
than with a Nepalese if you're from Belgium.
Whatever it is, what you find in those cases is,
that's how you not only get innate olfactory
recognition of, is this a relative or not,
but how related of a relative?
So that's half of it.
That's how you generate the unique signature
at the olfaction at the pheromone end.
The other trick is, how do you generate
an olfactory system that can make that discrimination?
And all we've got to go with, at that point, is-- you can guess,
if you think about it a bit-- is,
we've got to have some version of olfactory receptors
that do the old lock-and-key business.
Just as you make a protein which has a certain shape,
indicating that this is your unique signature.
What you want to do is have receptors
that will have the uniquely complimentary shape
for the lock and key so that you can do a-- oh!
This fits perfectly.
Let's transduce this to a signal to the brain, saying,
I'm smelling my armpit.
And if, instead, you've got a protein that
fits in there like a lock and key but not quite as well,
you send the message of, oh, I'm smelling my sibling's armpit.
And if it fits in there not quite as well, and all the way
down, you could begin to see exactly how you designed this.
If you've got 1,000 of these receptors of this shape
and every single one of them it has them fitting in well enough
to stay in there for three seconds, so that all 1,000
of them send the signal.
It takes three seconds of binding there
to cause the signal to happen.
All 1,000 of them send the signal that means "it's me."
It doesn't fit quite as well, so, statistically, only 800
of them stay in there for three seconds, so 800 of the cells
are reporting, oh, that's a full sibling.
I don't know if that's the mechanism.
But this would be a way of constructing it.
That's exactly how it could be, along the lines of lock
and key, genes produce proteins of certain shapes,
certain functions-- all of that.
And that's how you begin to do it.
What have we just gotten?
We've gotten a protein-- a molecular basis
of our theorizing, the other day, of an if/then clause.
If and only if this is a close relative,
then send a message to the neurons that do "cooperation."
We already know that's gibberish,
to say that there's neurons that do cooperation.
But you could begin to see how this is going to work.
This is the "if" part of all the conditional if/then
clauses built around degree of relatedness.
So how does your olfactory bulb do this?
Very interestingly, people are beginning
to get a sense of two hormones that are relevant to this.
One is called "oxytocin," and the other
is called "vasopressin."
And what we will see in lectures to come is, particularly
in females, oxytocin has long been
known to play some, like, plumbing, nuts-and-bolts job
in giving birth, and vasopressin in uterine contractions.
And take your average, like, off-the-rack endocrinologist.
And what these hormones are about
is, like, your uterus contracting and giving birth.
But oh, that's so little of what they do.
What they're much more interestingly involved with
is what happens next.
Which is now beginning to learn how
to recognize the smell of the individual
you just gave birth to.
Because it turns out, what oxytocin and vasopressin do
in the olfactory bulbs-- the olfactory system,
the olfactory equivalent of your eyes and ears--
is they tune up the cells that recognize
major histocompatibility signals.
They make them attuned to, is this a relative or not?
And there's an if/then clause.
If the levels of this hormone have the certain high levels
indicating that I just gave birth,
and I smell something whose signature odorants fit really
well to this receptor, then this is
someone who I'm going to nurse like crazy, unless it turns out
to be my mother or grandmother.
OK, let's put it another-- if they're very little and cutesy
and make cute little mewling sounds,
then I will take care of them and nurse them
and all that sort of thing.
And it's turning out that that's what oxytocin and vasopressin
are doing in there.
You generate mice with genetically-- knocking out
oxytocin or vasopressin or their receptors--
if these are totally new terms, this will come by the week
after next, as an introduction.
You knock out those genes, and you get what
is called a "social anosmia."
Anosmia is the inability to smell something.
A social anosmia is, your nose is working just fine--
if you're a rat, you could discriminate different food
types-- completely arbitrary odors-- you just
can't distinguish between individuals.
And there was a paper, a couple of weeks ago,
showing for the first time-- what the model has always been
is that circulating oxytocin and vasopressin
get into your olfactory system and have those effects.
What this paper showed for the first time is,
you're making those hormones right in your nostrils,
to begin with.
This is what tunes it up.
Very interestingly, there is also
a literature emerging suggesting mutations
in genes relevant to oxytocin and vasopressin
in families with a high incidence of autism.
Autism, a disease where-- one way of characterizing it is,
there are enormous deficits in normal socialization
interaction, social bonding, social affiliation.
And this suggests that has something to do with it.
OK, so that's a first mechanism.
That's how it might work innately.
Very cool study in recent years, also showing one facet of this.
One of the things you get taught in Intro Neuro--
if you've taken anytime in the last 5,000 years--
is, when you've got an adult brain,
it doesn't make any more neurons.
You've got all the neurons you're ever going to get,
by the time you're three years old,
and all you can do thereafter is squander them away on, like,
stupid weekend binges or whatever.
And it turns out that, nonetheless, this is not true.
And arguably this is the biggest revolution in neuroscience
in the last decade or so.
Adult "neurogenesis."
And it turns out, this adult neurogenesis
happens in only two areas in the brain.
The first one is really interesting.
Because it's this part of the brain called the "hippocampus."
Hippocampus-- learning, memory.
It's totally cool.
And a gazillion studies now showing stuff
like, you learn a new fact, you stimulate neurogenesis
in your hippocampus.
You get put in an enriched environment,
you exercise, you do all sorts of stuff--
you make new neurons there.
You get stressed, you make less new neurons there.
This whole new field.
And 99% of the studies have been about these populations
of neural stem cells in the hippocampus.
Totally ignored has been the second, little pocket
of these neural stem cells that can make new neurons.
Nobody's interested in it.
What's this about?
It's totally boring.
Where's the second pocket?
Just behind the olfactory bulb.
And what this study showed-- and this is one in your reader,
just at the abstract.
What it showed was, if you have a rat right
around the time she gets pregnant,
she starts to make a new neurons out the wazoo--
not from the exciting hippocampal pocket,
but from this boring, little olfactory.
And what goes on with the onset of pregnancy--
female rodents do this massive renovation
job of all their olfactory neurons going on, there.
What it is, they showed in this study--
it is driven by the prolactin levels that
rise during pregnancy.
And what have you got there?
What they showed was, right around the time
she gives birth, she's got this spanking-new, completely
renovated olfactory system, just in time
to do one of the most important olfactory things of her life,
which is quickly figure out which ones are her babies.
To quickly do the social bonding to them.
What hasn't been tested yet, but what
I guarantee has to be there, is that vasopressin and oxytocin
has to have some sort of role going on in there.
Really interesting.
Interesting implication of this.
So think about this.
If you were pregnant-- and assuming this works in mammals
other than rodents-- and if you were pregnant--
so the whole time you were pregnant, what's going on?
You're doing this huge job of ripping out
the walls and the plumbing in your olfactory bulb and, like,
putting in new stuff.
And it's a total mess.
Like, you spend your pregnancy with your olfactory system
totally cockeyed.
No wonder stuff smells weird, and no wonder foods
taste weird and all of that.
There is a whole adaptationist literature
out there on, why should it be that you suddenly
want pickles and these foods you can't stand the taste of?
And it's to avoid inadvertently eating toxins.
A really unconvincing, spandrel-filled
literature out there.
It may be an inadvertent spandrel byproduct of,
you're ripping apart the whole olfactory system,
so you're all set to recognize the smell of your kids.
You just got screwy olfaction and taste
all throughout pregnancy.
The main point of this, though, is,
this is endocrine regulation driving,
not to make you able to recognize
a relative-- because you've already got the genes in place
for that-- just making sure that your olfactory bulb is
at the very best at that time for doing it.
OK.
What else does one want to know about that?
OK, one additional thing you could do with that information.
Which is, OK, so, why do you want
to know who your relatives are?
It's who you mate with, it's who you cooperate
with, it's who you try to kill, it's
who you take care of, it's who-- all that sort of thing.
All of these domains.
Also it's who you pay attention to socially,
in terms of gossip and such.
One interesting study that was shown,
which was with baboons-- and this
was the same folks at the University of Pennsylvania.
What they did was they recorded the voices--
You notice this business about playing
the voices of some animal in the bushes
and looking at the response in everybody else
is one of these standard tools.
What they did, in this case, was,
the voice of two animals from that group, from that troop.
And what we heard was the voice of the lower-ranking animal
giving a dominating vocalization,
and the voice of the lower-ranking individual
giving this terrified subordinating noise.
So they obviously were not getting a terrified
subordinating noise out of number 1,
but they had to sit around and get recordings
of number 2 and everybody else, at some point or other,
so now they could put them down there.
And they would play this.
So everybody else is sitting there and saying, what?
Number 4 is terrified of number 27?
What's going on?
And what they showed was, everybody paid a huge attention
to this, if they were hearing-- number 27 was trashing number
4 if they weren't relatives.
But if they were in the same family,
and there was this dominance reversal nobody paid attention.
Crazy relatives.
Who knows what's going on in that family?
They're just squabbling.
They distinguish the social implications
of a dominance reversal, depending
on relatedness or not.
One additional thing with it, before we
go to our next way of recognizing relatives,
after the break-- hold on.
One additional thing is, of course,
telling you who to mate with.
And it's obvious who you're supposed to mate with,
in species after species-- someone
who is not related to you.
Because if you do, you may inadvertently
produce babies with two tails and seven fingers
and all of that.
Nobody picked up on the fact that actually we
have 10 fingers instead of five, but we'll let that slide.
But what you get there is, oh, avoid inbreeding.
Don't breed with relatives.
But we've already heard about a counterargument, which
is, breed with relatives because of the inclusive fitness,
the kin-selection advantages of doing that.
And, from the first minute people
started getting these theoretical models,
it was clear that, in fact, there
were contradictory pulls between doing
major outbreeding in your mating and doing
inbreeding with your mating.
And people did all these theoretical models
of econometrics of where you optimize the difference.
And they came out with the conclusion
that, in all sorts of species, the optimal balance
of avoiding inbreeding disasters with advantages
of kin selection would be to mate with something
like a third cousin.
And you go and look at all sorts of species out there,
and that's precisely what they do.
That's where it balances out.
And, again, you can't do that unless you know relatedness.
Then, a couple of years ago, along came this researcher--
Martha McClintock, University of Chicago.
She's the person who discovered the Wellesley effect back when.
And for all of you guys who are gearing up for senior honors
theses, this was her senior honors thesis-- discovering
the Wellesley effect.
So this was, like, one impressive study.
Yeah?
Did Darwin marry a second cousin? [INAUDIBLE]
Who did-- he married a first cousin?
First cousin [INAUDIBLE].
First cousin.
There you go.
When we all know he would much rather have married a Galapagos
tortoise, but his parents forced him--
[LAUGHTER]
Well, interesting exam-- OK, so you thus catapult us, here,
into the human realm.
What McClintock did was a study, a couple of years
ago, where she took swabs-- she has this whole paradigm of very
high-tech-- getting cotton balls and rubbing it
on people's armpits and putting it in a jar,
there, and then getting these volunteers who
can [SNIFF] smell it and give some assessment of how good it
smells or not.
Which, in and of itself, is pretty wild.
And you run that with humans.
And of a curve of relatedness, whose odor gets rated
as having the most appealing?
Third cousins.
Yuck!
Think about that one later.
We are just another species, in that regard.
All of this suggesting, even in humans,
you are balancing this disadvantages
of inbreeding with the kin-selection advantages.
And, again, there's no way you can do that without knowing
who's related to what extent.
OK.
Five-minute break.
And what we'll then transition to are,
species that don't do it innately but instead
have to imprint, early on, after encountering them.
OK, let's get going again.
Let's see.
First off, thanks to the wonders of Wikipedia
never being that far away from us, we now know-- here
were Charles Darwin's parents, who were third cousins.
And then Charles Darwin married his first cousin.
So there you have it-- something or other.
But this apparently was rather common at the time.
And is not quite the optimal, according to Martha McClintock.
OK.
So, pushing on.
So we've now seen why you would want
to recognize your relatives and recognize
the degree of relatedness.
All of our models of who to compete with, who to mate with,
who to nurse, who to take care of,
who to be voyeuristic about.
And we've seen the first domain where
you can do that, which is to recognize somebody innately.
And one very confusing aspect of it,
which I managed to make confusing,
is-- so, what's up with olfaction
with that and oxytocin and vasopressin and prolactin?
OK, as follows.
It is innate that you will have receptors, olfactory receptors,
in your olfactory neurons, in your olfactory bulb--
in your nose.
It is innate that you will have olfactory receptors that
will be able to detect degree of relatedness--
how close an olfactory signature is to your own.
That is innate.
That will be there.
What appears to be the case is, oxytocin and vasopressin
make you more likely to make those receptors-- increase
the number of those receptors.
So this is not oxytocin and vasopressin making you suddenly
be able to recognize relatives.
It's just making you better at doing
that-- more sensitized to it.
So that you have 100 receptors reporting instead
of 10 of them-- more accuracy.
What prolactin appears to be doing is--
and this is under study, but the best bet is,
this is another way of getting more of these olfactory
receptors online, right around birth,
to innately recognize your relative.
And this time, instead of making neurons
make more copies of the receptors,
it's making more neurons.
There's going to be more subtlety than that.
But broadly, those are two different ways
of making you better at doing something
that is innate in you.
Rather than making you suddenly able to learn how to do this.
So we now transition to the second way
in which relatives are recognized,
where it's not innate.
It requires imprinting.
It requires some sort of learning which leaves
a long-lasting message of it.
Imprinting.
So now we've got imprinting in yet another use of the word.
And major use of it coming next Monday.
Imprinting-- how an animal learns who its mother is.
How a mother learns who its babies are.
How it imprints on the smell, on the sound,
on the whatever of its offspring or parent--
how that learning goes on.
What is clear is, that's a case where
the learning that the learning occurs at that point is innate.
What is learned is experiential.
Important sort of distinction, there.
OK.
So, what goes on with this?
So in lots of species, you learn the sound
of your infant's voice.
In lots of species, you learn the odor.
In lots of species, you learn what they look like.
Different species, different modalities that dominate.
What goes into, for example, learning what
your offspring smells like?
Remember, it's not innate in these cases.
For example, a goat, a sheep-- whatever-- they do not innately
have the means of recognizing, oh, this
is somebody whose major histocompatibility proteins are
fairly similar to mine.
In fact, they share 50% homology.
Oh, this must be my child.
That's not done that way.
These are not innate cases.
So what sort of rules might you have for how
to recognize your offspring?
Here's a simple one.
OK, there's a whole bunch of babies out here,
and which ones are mine?
And I'm a goat trying to figure out
which one I'm going to nurse.
What would be a good rule that is not innate,
instead building on learning something at that point?
I know-- I'm going to start taking care of whichever kid
there smells like my vaginal fluids.
That's a pretty reliable way of figuring out
this is somebody you just gave birth to.
Or, this is someone who I lick as they're first coming out.
And, for a while afterward, I figure out, who am I nice to?
Someone who smells like my saliva.
Someone who I scent-mark right after this
is someone who smells like my whatever
glands I'm using, there.
This is someone who smells like my amniotic fluid.
This is someone who, after I get that learned,
I then nurse them for the first time-- this is someone whose
mouth smells like my milk.
You could see, now, in this case,
it is learning being built on top
of whatever your own recognition system is of smell.
You can have elaborations on this.
How do you recognize a sibling in species where it is
learned in this imprinting way?
Oh, I'm going to imprint as a relative on somebody who
smells just like mom.
Somebody who smells like mom's vaginal fluid or her saliva
or any of that same stuff going on, there.
Or, I'm going to be nice to someone who smells like someone
I mated with back when.
That's another strategy in various species.
You could begin to see how all of these
are ways of just getting logical information.
Someone who has the voice like someone
who I heard peeping when they were still inside the egg.
Oh, it's them!
I'm going to nurse-- no, I'm not going to nurse them,
I'm going to give them worms or whatever it is birds do.
But all these ways of using sensory information
to say, oh, that's the one!
That's the one.
That's how I know it's them.
So, all sorts of ways in which this could be done.
How can you prove that this is not hardwired-- this is not
absolutely dominating?
A technique we've been hearing about already, which is
that cross-fostering business.
Which is, you take newborn whatevers,
and you switch mothers on them, and the mothers
will take care of them.
What does that tell you?
It means other attributes of these pups coming over
to them-- these rat pups-- override the fact-- wait,
these kids don't smell like my vaginal fluid!
Wow, they're cute, though, and there's no other moms around,
and they sure look cuddly, and--
So maybe it takes you 10 seconds to decide, this is one of mine,
instead of three seconds.
You have different things being played out, there.
Contrasting sort of signals coming through.
So that's another domain of doing that.
So now we move to us and how we do recognizing relatives.
And initially what the answer seems to be
is that we have a different version of it.
We do not recognize relatives innately,
nor do we recognize relatives by imprinting.
We don't do that deal of, like, we smell our parents right
after they're born, or we sniff the vaginal fluid
and that's how we know who mom is forever after, or stuff
like that.
So we don't-- instead, what do we do?
We do it cognitively.
We figure it out.
We think about it.
We think about it, with active, conscious, cognitive rules
of how you know a relative is.
And, of course, what we're going to see shortly
is, that's not really how it works, a lot of the time.
But, for a first pass, we do it cognitively.
So now you do it--
Instead of, this is someone who smells just
like my vaginal fluid, the way the goats are doing,
or, this is someone who I innately recognize as my child,
the way the rats are doing it, you're saying,
well, that's the baby I just gave birth to,
because they haven't taken him out of the room yet.
So that must be-- a cognitive strategy.
How do I know who the father is?
Not because I can smell in the baby's 50%
sharing their major histocompatibility gene.
This is the only person I had sex
with during the cycle that I conceived.
A cognitive strategy.
This is what humans do, again, [INAUDIBLE]
because we can think.
We can go through stuff like that.
We can also do other versions of it,
which is, well, this is someone who I saw mom give birth to.
Or at least when they took me out of the room,
and then they, like, gave me some stuffed animals to keep me
from, like, getting too upset.
And then they said, here's your new baby whatever.
Oh, OK.
Which is a variant on, this is someone
who's been around ever since I saw mom give birth
to that individual.
This is someone who looks like me.
This is someone who looks like a family member.
All of these cognitive strategies, going into that.
We think about it.
We think about it.
And, it turns out, we have brain mechanisms that
are good for doing that, too.
But we're not the only species that
thinks about it in that way.
For example, baboons-- wonderful study, a few years ago.
Baboons can do this with some sort of statistical thinking.
OK, baboons are polygamous.
Females will mate with a bunch of different males
during her cycle, and conceive, and thus it's not at all clear
who the father is.
But everybody there plays a guessing game.
And everybody does some statistics.
The usual rule is, baboons, being
a highly tournament species-- males do
no parental care of offspring.
Turns out that's not quite the case.
Some males do.
When they're pretty sure who their kid is.
And here's how it's done.
You are a baboon, and you are just hitting puberty.
You're female.
And what goes on is, for, like, your first half dozen cycles,
you're cycling but you're probably not ovulating yet.
You're not quite fertile.
This happens in humans, as well.
And none of the big, high-ranking guys
are terribly interested in you, probably because you're not
pumping out a whole lot of interesting pheromones yet.
So who do you wind up with?
You wind up with some poor, like, adolescent schnook
who has no chance to mate with anybody else.
And nobody's contesting his ability
to have a consortship with you.
So you do all of your mating with him.
The vast majority of the time, they're not fertile,
because you're not really ov--
Every now and then, though, you get
this, like, junior-high-school baboon
guy who gets his girlfriend pregnant,
and he's the only one.
And what tends to happen then is, when she gives birth,
he gives a fair amount of parental care to the offspring.
So what does he have as a rule, there?
If I'm the only one who mated with her--
if they're using a cognitive strategy-- then
I'll take care of the kid to some extent.
And it is a sight to behold, how incredibly inept
an adolescent male baboon is when
he's trying to be paternal.
But what you wind up seeing there is,
well, maybe he's imprinted.
Maybe he's doing major histocompatibility gene innate
stuff.
Or maybe he's actually thinking, hey, I was with her 24/7,
so it's gotta be me.
I'll take care of the kid.
Now you see the more complicated circumstance,
where it's a more desirable female-- more mature one--
where there will be a bunch of males contesting.
And what you will tend to see is, two days before
and after she ovulates, she'll be mating with number 10.
A day before and after, she's mating with number 3.
The day of ovulation, she's mating with number 1.
That tends to be the pattern.
Stay tuned.
It doesn't fit that perfectly, because female baboons also
have some opinions about who they want to mate with.
But, nonetheless, there winds up being
this pattern, just like that.
So there winds up being this pattern.
And thus-- you're the male, afterward,
when she gives birth, and you're trying to decide,
is this my kid?
Or, should I give some male parental investment?
And what is shown is, baboon males do statistics.
They do probability.
What they do is, if this was a male who
was mating during her prime ovulatory day,
he is more likely to take care of the kid
than a male who was mating during this window
or this window.
They are playing probability.
They're not very good at it-- no surprise.
But nonetheless this pattern emerges.
That's not innate recognition.
That's not innate recognition of some smell.
That's just thinking through it.
Was I with her?
Yeah, but she sure smelled a lot better the day after,
and that other guy was with her, so I
guess I'm in this category.
OK, well, I'll smile at the kid now and then tell him I
like their piano playing.
But I'm not going to, like, invest anything--
kind of stuff, there.
Here, you see a conscious, cognitive strategy.
So, OK, so us and other smart beasties like primates.
Here's another version of it, occurring in fish-- in sunfish.
And this was research done by that guy
David Sloan Wilson, that multilevel-selection-evolution
evolution guy, who actually has done research in, like,
so many different fields.
Incredibly creative guy.
Here's a study that he did.
Sunfish males, surprisingly, are quite
paternal in their taking care of their eventual offspring.
They're almost as good as Nemo's dad.
And what you've got there is, in this version, he--
I'm not going to draw fish mating.
Forget that!
OK.
So the guy mates with the female,
and she eventually gives birth to kids.
And he helps take care of them.
Now, instead, he mates with a female--
and you, the savagely heartless researcher,
puts him in the next tank over, where
he can see what's happening.
And you fiendishly, at that point,
drop down a clear plastic box right next to the female,
with another male in there!
In reality, he's not mating with the female,
but he's right there.
And this guy, who's stuck on the other side of the barrier
and going out of his mind with jealousy and petulance
and all of that immature stuff.
And what happens is, after she eventually
gives birth, she-- this apparently being the female,
in this diagram.
After she gives birth, he doesn't take care of the kids
as much.
There is no difference in any of the sensory cues.
Because he, in fact, is the only one who mated.
This one is kept inside.
There is some sort of cognitive stuff even going on in a fish.
Remarkable.
So it's not just us.
So we've got this broad realm-- innate strategies.
One's on very all-or-none sensory imprinting.
And then there's the folks who think through it.
We're in the mainstream of that, but we're not the only ones.
So how do our brains do that?
There is a part of the human cortex called the "fusiform
cortex."
And what it's good at is recognizing faces,
which is a remarkable thing.
It specializes in recognizing faces, facial expression,
degree of relatedness.
You show someone a good portrait of someone else,
and you will get that part of the brain
to activate as if it was their face.
You show a good cartoon of somebody.
It will do the same thing, maybe with a little bit
less confidence.
This is the specialized part of the cortex
that does facial recognition.
Remarkable.
You look at people with autism, and this part of the cortex
doesn't do a whole lot.
You show a nonautistic individual
a picture of a well-known loved one.
Fusiform cortex activates a whole lot.
You show them a picture of someone
they don't know a whole lot.
It activates somewhat to a lesser degree.
You show them a picture of an armchair.
Doesn't do anything at all.
You take someone who is autistic, and you show them,
and you get the same low-degree activation for all three.
You know, mother equals stranger equals armchair.
This in some ways is the core of what autism is about.
You see that going on, as played out in this part of the cortex.
So that's really interesting.
And then it turns out we're not the only species that
has this specialized fusiform cortex.
Primates have this, as well-- nonhuman primates.
Sheep have it, it turns out.
Pigeons who could recognize pictures of each other.
And why they would want to do that, god knows.
But pigeons have a very proto version of the fusiform cortex.
This seems to be part of what goes
into this conscious, cognitive strategizing of, this
is a face, this is a face I've seen
a whole lot more over the years.
So this activates a lot more.
This seems to suggest some sort of cortical specialization
for recognizing individuals.
More features of how humans do it.
Something that humans are very good at,
if they are human mothers, is recognizing
the smell of their baby right after birth.
And this has been shown to have a major histocompatibility
component to it.
In other words, that's not a purely cognitive strategy.
There's some innate, instinctual olfactory
signaling going on there.
So, a first bit of evidence is that we are not
just purely rational beasts at figuring out
who we're related to.
Babies, very shortly after birth,
are already spectacularly good at distinguishing the smell
of Mom versus somebody else.
How do you tell that with a baby?
You take a newborn baby, and you give them, like, on this side,
some armpit smell of Mom, and there's some armpit smell of,
you know-- I don't know-- Margaret Thatcher.
And what you see, then, is that the baby, the newborn,
is more likely-- will spend more time turning
its head towards Mom's smell.
That's how you know it.
Newborn babies cannot distinguish between the smell
of Dad and any other male.
It's not instinctual, in that case.
Newborn humans are doing some version of proximity to mom,
vaginal fluid smell-- whatever.
This seems to be bringing up a question of,
is there a difference in bonding in those early stages
between vaginal-birth offspring and cesareans?
I don't know, but that suggests that should be happening.
And maybe I should even find that out.
OK.
Other features of it.
Newborn kids, as we already know from a couple lectures ago,
can recognize the voice of Mom.
How?
From all that resonance, there, inside the amniotic fluid,
as Mom reads them War and Peace.
And, as we also heard, they can't
recognize the voice of Dad.
Not instinctual.
In this case, getting information
on the sensory route-- the amniotic environment
as a good resonator for Mom's voice.
So we're now already getting a mixture, here,
of some instinctual olfactory stuff,
major-histocompatibility-complex stuff, some acquired,
hardwired sensory-driven stuff.
Ooh, does this sound like the person whose voice I heard
for those last nine months?
And then some cognitive stuff-- trying to figure out
who this relative is.
Ooh, who did they come with to the party with?
All of that-- the cognitive stuff.
So, already a suggestion that we're not just
such pure cognitive machines.
So where is that most interesting?
And this is in a totally fascinating series
of studies that were done over the years,
showing just how little conscious cognition might
play out in us humans in some really critical realms.
Which is, who you decide you are interested in mating with.
OK, so how does that work in humans?
All sorts of different ways.
But here's one way in which it does something interesting.
And this was a classic study done by an anthropologist named
Joseph Schaeffer.
And this is what he did.
He studied people who grew up in Israel on kibbutzes.
Kibbutzes-- these are these traditional, socialist sort
of communes, where one of the-- none of them
are like this anymore.
But one of the early principles was, one parent
is as good as another parent.
All the kids are raised in these big, communal bathtubs
together.
And it's one big, sort of socialist [INAUDIBLE]
of everybody bathing naked together.
And what Schaeffer found was this very interesting thing.
Which is, you are brought up in your age group.
All the kids born this year are raised
in the same communal group.
And they take their baths together,
and they play together, and they've
got the same-- one parent takes all of them
for one afternoon, every Tuesday afternoon,
and the next one comes for the next shift.
And this big communal business, there.
And what Schaeffer discovered was,
if you were raised in the same age group as somebody,
anywhere up to six years of age, you never,
ever wind up marrying them.
And this was not with a sample-- an appropriate sampling.
This was a study he did of every individual who ever grew up
in the kibbutz system in Israel.
We appear to have a rule as follows.
If this is somebody who you spent a whole lot of time
with before six years of age, this
is someone who you sure don't want
to grow up and marry-- yuck!
And you love them-- they're incred--
but they feel like my sister.
They feel like my brother.
Whoa, that would be totally grotesque!
There has never been a case of people brought
up during the first-- grow up in your first six years of life,
taking a whole lot of baths with someone,
and you are not going to discover an amorous passion
for them when you're 20 years old.
They're going to feel like a sibling for you,
for the rest of your life.
What is that telling us?
We have a very noncognitive strategy in there.
Yes, who are relatives-- who are appropriate people
to mate with?
Well, if we know that this is the daughter of Mom's sister,
then this person is not appropriate,
and this whole cognitive-- spend a whole lot of time
naked with somebody, taking baths and playing pattycake
and counting their toes, the first six years of life,
they feel like a sibling.
Something very similar was then shown by Arthur Wolf,
here, in the Anthropology department--
a different cultural version.
Traditional Taiwanese marriages, where
there's some equivalent of that, and where either you wind up
with your future spouse at some insanely early age
and basically get brought up with them from infancy,
or it happens later.
And if you're brought up from infancy with them,
you have a disastrous marriage later on,
because they feel like a sibling for the rest of your life.
So what this shows us here is, yes, we
are these wonderful, rational, cognitive machines.
We've got all sorts of innate strategies
and sensory-imprinting stuff going on, instead, in us.
We are not a whole lot fancier than hamsters.
What does this set us up for?
A topic that's going to be real important when
we come to the lectures on aggression and cooperation
and competition and all of that.
Which is, if we spend a lot of our time figuring out
who we are related to-- cognitively--
or, even more importantly, if we are
malleable in the way these kids growing up in Taiwan,
or in these kibbutzes-- if we are
malleable in who we feel related to,
it is possible to manipulate us in lots of ways
to feel more related to individuals
than we actually are, or to feel less related to individuals.
And the terms for these are "pseudo kinship"
and "pseudo speciation."
And what we will eventually see is,
when you make sense of human violence and human cooperation
and human aggression and all of that,
we are so easily manipulated as to who counts as an "us"
and who counts as a "them."
And one of the brilliant things militaries all the world over
do, whether you're talking about clans with warrior classes
all the way up to sort of state militaries--
what you see in all these cases is a brilliant understanding
of how to make nonrelatives feel like they're a band of brothers
and how to make "them" seem so different they hardly
even count as humans.
So this ability for us to be manipulated
in these subliminal ways, as to who counts as a relative,
will come back to haunt us big-time
in making sense of a lot of human social behavior.
In other words, we are not purely rational.
OK.
So what we will do on Monday is switch another bucket, now,
to this field of ethology, which, once again, is trying
to make sense of, what behaviors are hardwired,
what does environment do?
You will see a completely different approach--
For more, please visit us at stanford.edu.
