Björn Brembs

In a discussion about what decisions are, [John
Krakauer](https://en.wikipedia.org/wiki/John_Krakauer) emphatically
pronounced that "decisions happen for reasons", in answering 'no' to my
question if it wasn't a decision with which foot to start walking from a
stand-still.

A [recent article](https://www.nature.com/articles/s41586-024-07890-3)
from the laboratory of Carolina Rezaval in Birmingham studied a
decision-making process in male *Drosophila* fruit flies where the
reasons for each decision seemed apparent. When a male fly was presented
with a looming stimulus (which mimics a threat from a nearing predator)
while in the early stages of courtship, it stops courting and either
freezes or tries to escape. In later stages of courtship, however, the
same threat stimulus is no longer able to interrupt courtship and elicit
escape. While it is less clear if the male fruit fly is able to
articulate them, the reasons for him to behave in that way seem obvious:
in the early stages of courtship, the male escapes as it is still not
clear if the courtship attempts will be successful. In a later stage,
once sufficient information has accumulated about the receptivity of the
female and/or copulation is close, it's worth trading some safety for
the prospects of getting one's genes into the next generation. In
biology, such reasons are commonly called "ultimate causations" of the
behavior, while the neurobiology by which they are implemented are the
"proximate causations". The recent article concerned the proximate
causations of these decisions in *Drosophila*.

In the early stages of courtship, the authors found, a class of visual
neurons in the fly's optic lobe called LC16 detect and respond to the
threat stimulus. Via several other neurons, these LC16 neurons connect
to P1 neurons in the fly brain, which collect multimodal information
(e.g., female sensory cues) and can mediate courtship behavior. If LC16
neurons are activated by a threat stimulus, downstream neurons release
serotonin, which inhibits the P1 neurons, halting courtship. This
inhibition between different behavior circuits is a common theme
observed in many other preparations, preventing two behaviors occurring
at the same time. In this case, LC16 neurons mediate escape behavior and
inhibit other behaviors, such as courtship, which could interfere with
the escape.

The fact that LC16 no longer inhibits courtship at later stages implies
that there is more to this story than the simple inhibition known from
so many other experiments with different animals. In the course of
courtship, as the duo gets closer to copulation, a population of
dopamine neurons in the male slowly ramp up their activity. These
dopamine neurons make direct synaptic connections with LC16 neurons and
inhibit them. In other words, the lack of escape is not due to a failure
of LC16 to inhibit courtship, but due to a failure of LC16 to respond to
threat stimuli.

Taken together, much like what was known from other preparations, there
is a mutual inhibition between circuits mediating different behaviors:
escape behavior circuits inhibit courtship circuits and courtship
circuits inhibit escape circuits. Until now, such mutual inhibition was
mainly studied in central-pattern-generator driven behaviors such as
swimming/feeding in molluscs (cited by the authors). In these cases, the
inhibition was identified very close to the motor side of the circuits.
In the case of male *Drosophila*, the relation of the neuronal processes
appears much closer to sensory input: Both LC16 and P1, while not being
sensory neurons, respond to sensory stimuli: LC16 to visual threat and
P1 to female sensory cues. Their inhibition can be interpreted as a lack
of sensation. The authors themselves do not seem to have made up their
minds about this. While they start their article by emphasizing the
animals' "decisions that require balancing opportunities and risks",
title and abstract of their work invoke more attention-like processes by
using words such as "love-blind". Which is it, decision or attention?

Is the decision to court or to escape really a decision, where the
"risks and opportunities" are weighed and then one of the two options is
chosen, or is this just an example of sensory competition, where
eventually one "blinds" the other, such that the losing stimulus is
simply not perceived? If only one of the two options is perceived, it
can no longer be a decision, can it? As neither LC16 nor P1 are directly
involved in sensing the stimuli, it is fair to assume that both the
looming stimulus and the female sensory cues are processed appropriately
even if any of the two classes of neurons are inhibited. However, if the
perceived looming stimulus is not longer assessed as a threat, or the
female sensory cues no longer perceived as sexually attractive, aren't
the animals then just simply responding to the single remaining
stimulus, without needing any reasons? As an aside: if so, what are the
'losing' stimuli perceived as?

The flies cannot tell us what they perceive and how, but this is not the
first time an animal's behavior that outwardly looks, sounds and smells
like a decision, loses these properties upon neuronal inspection.
Leeches commonly respond with local bending of their body wall to
mechanosensory stimulation such as light touches. Increase the intensity
of the touch a little, and the animal will start crawling away from it,
and with further increases, the animal will start swimming to escape.
However, when the animal is feeding, not even the strongest
mechanosensory stimuli can get the animal to do swim or crawl away. The
reasons for this decision are clear: if the leech has itself attached to
the animal it is sucking blood from then this animal is likely moving
around in the pond or stream where the leech lives. Given the common
environments for this leech species with plenty of vegetation in the
water, the moving animal will likely touch an number of obstacles and
chances are that these obstacles will also touch the leech. If it were
to start crawling away or even swim away in that phase, it would not be
able to obtain a sufficient blood meal for survival and procreation. So
the leech will literally hold on for dear life until it has had a
sufficient meal.

Decades of research have provided a pretty good understanding of the
circuitry for the processing of mechanosensory stimuli in the leech. Way
back in [2009, Gaudry and
Kristan](https://www.nature.com/articles/nn.2400) reported on the
neuronal mechanisms underlying the decision of feeding leeches to ignore
mechanosensory stimuli. Analogously to the male flies where a
sufficiently progressed courtship inhibits the transmission of threat
stimuli, feeding in the leech also leads to the release of a biogenic
amine, in this case serotonin. Serotonin then inhibits transmitter
release from the mechanosensory neurons. Conceptually very similar to
the male flies, the sensory stimuli are still perceived by the relevant
sensory neurons, but the transmission of the signal is blocked via
central processes. In male flies, dopaminergic neurons active during
courtship inhibit threat transmission during courtship, in the leech,
serotonergic neurons active during feeding inhibit mechanosensory
transmission during feeding. In the case of the leech, the authors take
an unambiguous position: Title, abstract and text all call the
inhibition of responses to mechanosensory stimulation during feeding a
'choice' or 'decision'.

In both flies and leeches, their behaviors were tied closely to sensory
stimuli: looming stimuli, female cues, blood, touch. It is precisely
because of these stimulus situations, that we can articulate the reasons
for the decisions the animals make. The animals' decision appear very
"reasonable" for the human observer. However, the two examples above are
just two of a growing list of examples where this very intuitive sense
of the behavior being a "decision" starts to disappear with our
understanding of how the neurons are doing it. If one of the two stimuli
the animal has to decide about gets shut out so that only one remains,
where is the decision?

A classic example of a decision in humans is "red or white wine?" If
human nervous systems would mediate such a decision analogously to flies
and leeches, the decision would be that between, say, an empty glass and
a glass filled with wine -- very different from how we envisage and
experience such decisions. But then again, the decisions studied in the
examples above are life-and death decisions for the animals. Everybody
knows that decisions about loved-ones are clouded by our love -- the
world "love-blindness" exists for a reason. Everybody has experienced
products in a grocery store becoming much more attractive when we are
hungry. It is conceivable that our minds similarly bend our sensory
experiences when in a life-threatening situation. If the animal examples
were anything to go by, decisions with reasons would more like reactions
to stimuli and less like actual decisions?

One can think of two arguments why studying the neurobiology of
decisions without reasons may actually be a more fruitful endeavor for
understanding decision-making, one conceptual and one practical.

Conceptually, as long as we haven't understood the whole process, we can
never be sure how much choice any human or animal ever has in a given
situation. We simply cannot know how coercive, ahem, compelling the
reasons for the decision actually are. Practically, it is also not the
most efficient way to just study one decision after another until one
happens to stumble upon one where the 'reasons' aren't really just
competing stimuli eliciting responses where, ultimately, one wins in a
very deterministic way. And even if one finds such a lucky situation, as
is seems to be the
[case](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0256560)
with
[photopreference](https://royalsocietypublishing.org/doi/full/10.1098/rsob.160229?rfr_dat=cr_pub++0pubmed&url_ver=Z39.88-2003&rfr_id=ori%3Arid%3Acrossref.org)
in flies, even then the practical problem remains, that without complete
knowledge of sensory processing, it is exceedingly difficult to know if
one's manipulations are actually affecting the decision-making process
itself or just some aspect of sensory processing.

A decision involves that there actually are at least two real options.
In the animal examples above, it looks, to the observer of the behavior,
like there were two such options, but once the neuronal mechanisms
mediating the 'choice' have become clear, one of the options was gone.
This is not the case when studying decisions that have no reasons at
all. Under such circumstances, the decision is always 50-50. The outcome
is, if the experiment is properly designed, unpredictable and happens
exclusively in the animal and cannot be dictated by external stimuli.
There are many such examples: isolated buccal ganglia of the sea slug
*Aplysia* produce different motor programs in a petri dish, without any
sensory organs. Analogously, [isolated leech nervous systems start and
stop producing swimming motor
programs](https://www.youtube.com/watch?v=-DmH3NxnUvM) in the dish.
[Tethered fruit flies produce different flight maneuvers even if flying
stationarily](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0000443)
inside the center of a ping-pong ball. In these cases, there are no
obvious, outwardly discernible 'reasons' for the animals/nervous systems
to pick one behavior over another. The reasons for the decision come
solely from within the animal itself. Conceptually, these situations
appear much closer to what we call "decisions" and practically, they
avoid the sensory confounds altogether.

These two kinds of decisions have traditionally been referred to as
"picking" and "choosing" in a [1977 seminal
paper](https://www.jstor.org/stable/40971175) by Ullmann-Margalit and
Morgenbesser.

So what really is a decision and how do we study them in neuroscience?
Picking or Choosing?

In human psychology, there is a
[distinction](https://archive.charlespence.net/papers/2016_Animal_Decisions.pdf)
between "judgments" and "decisions" (i.e., the judgment and
decision-making, JDM, framework):

> 'Judgments' refer to how individuals acquire and process information
> to arrive at an understanding of the situation or state of the world.
> 'Decisions' refer to the processes by which individuals use judgments
> to arrive at a course of action. In other words, individuals can vary
> in two main ways: how they 'see' the world, and what they decide to do
> about it.

It appears that the leech and fly work cited above merely concerns the
"judgment" aspect of JDM, leaving the decision-making untouched. In this
JDM framework, does it not seem wise for neuroscience experiments to
circumvent the practical problems of judgments by studying decisions
that happen in the absence or at least with equivalence of sensory
stimuli?


In a discussion about what decisions are, John Krakauer emphatically pronounced that “decisions happen for reasons”, in answering ‘no’ to my question if it wasn’t a decision with which foot to start walking from a stand-still. A recent article from the laboratory of Carolina Rezaval in Birmingham studied a decision-making process in male
<em>
 Drosophila
</em>
fruit flies where the reasons for each decision seemed apparent.

What is a decision?

I am skeptical about agent-based models. Here are my thoughts on their problems.

Blair Fix

In the opening
[post](https://economicsfromthetopdown.com/2019/04/11/economics-from-the-top-down/)
of this blog, I described my 'top-down' approach to studying society.
This means studying *groups* of people without trying to reduce
everything to the actions of individuals. It's not that I think
individual actions are *unimportant*. Of course they are important. The
problem is that we are hopelessly far from understanding the
complexities of the human psyche. In this post, I'm going to dive into
the problem.

Let's talk about the difficulties of modelling human behaviour. To model
human behaviour, we need to create a little simulated 'agent' that makes
decisions just like a human would. We call these 'agent-based' models.
We create a whole bunch of little simulated humans. Each has an
algorithm for making decisions. We then let these agents loose and see
what happens. The results can be very interesting.

But 'interesting' results are not what science is about. Science is
about the search for the *truth*. So ultimately we want to know if these
models mimic *actual* human behaviour. The problem is that this is not
easy to test. Modellers typically let loose their simulation and then
look at the *aggregate results*. If the aggregate results are consistent
with real-world data, the modellers celebrate. The model 'explains'
human behaviour!

But not so fast. A model can be completely wrong and still produce
results that are consistent with observations. Consider the distribution
of human height. It follows a bell curve, centered around some average
height. But there are many different mechanisms that can produce a bell
curve. So suppose we create a model that produces the correct bell curve
for human height. It does not mean that our model is *correct*.

Suppose we assumed that adults' height varies randomly through time.
This is easy to model mathematically, and we could make it produce the
correct distribution of human height. *But the model is completely
wrong*. Adult height does not change (except when we reach old age).

So the hard part of modelling is not testing the model *results*. That's
easy. The hard part is testing the *assumptions*. You see, testing the
assumptions is something that has to be done independently of the model.
It's part of experimental and observational science. It's actually
really hard, and few economic modellers try do it. Almost 50 years ago,
Wassily Leontief went on an [epic
rant](https://blogs.warwick.ac.uk/files/dennisleech/leontieftheoassnonfacts.pdf)
about this very problem. His complaint: economists modelled without
testing their assumptions.

Sadly, most economists did not change their ways. Was it because
Leontief was just some crank? No! Leontief was the President of the
American Economics Association at the time. Economists did not change
their ways for a far more perverse reason. If they had actually tested
their assumptions, they would likely have found that most of their
assumptions were *wrong*. Or worse yet, that they were *impossible* to
test.

But enough about the sorry state of mainstream economics. Let's look at
how hard it is to model human behaviour.

### The Ghost in the Machine

To model human behavior, we need to simulate the 'ghost in the machine'.
This is the (metaphorical) demon in our brains that drives our
behaviour. I'm going to break the ghost down into three parts, based on
what is accessible to our conscious minds. I call these the "three
ghosts in the decision-making machine":

#### The Three 'Ghosts' in the Decision-Making Machine:

1.  Awareness that we are making a decision
2.  Awareness of the trade offs that we are weighting
3.  Awareness of the decision-making algorithm

For any given decision, we can be aware of 0, 1, 2, or all 3 of the
ghosts. We can use this awareness to define four types of decisions:

#### The four types of decisions:

1.  Instinctual decisions
2.  Intuitive decisions
3.  Quasi-rational decisions
4.  Rational decisions

Let's go through each one.

#### Instinctual Decisions

*Instinctual* decisions occur when we don't know any of the three
ghosts. We don't know the decision-making algorithm, we don't know the
trade offs being weighted, and most importantly, *we don't even know we
are making a decision*! This may seem absurd, but these instinctual
decisions probably outnumber all the others.

Take the act of *walking*. This takes decisions. How long should your
gait be? How high should you lift your feet? How hard should you push
off? How far apart should your feet be?

Are these really *decisions*? Yes! We figured this out when we tried to
teach [robots to
walk](https://www.wired.com/story/the-clever-clumsiness-of-a-robot-teaching-itself-to-walk/).
It's remarkably hard. The robot must be given an algorithm for making
these decisions. The robot must explore the possible ways to walk. Then
it must judge which way is best, and make a choice. We make these
decisions unconsciously. Some people --- dancers --- receive special
training so that they have conscious control over all the possible
decisions. But this takes years of practice. The rest of us are happily
oblivious to these choices. We can thank evolution for this instinctual
ability.

#### Intuitive Decisions

*Intuitive* decisions occur when we are aware that we are making a
decision, but we don't know the other two ghosts. We remain blind
(either totally or partially) to the trade offs, and we are blind to the
decision-making algorithm.

Choosing a life partner is an intuitive decision. We are fully aware of
this decision (most of us, anyway). But we are unaware of the second
ghost --- the trade offs we are weighting. We're aware of *some* of the
trade offs when choosing a mate. But certainly not *all* of them. And
here's an unsettling thought. Are the criteria that we *think* we are
using the ones that we are *actually* using? This is hard to know and
hard to find out. But I suspect we may often delude ourselves.

Lastly, when making intuitive decisions we are oblivious to the
decision-making algorithm at work. We just get an intuitive feeling for
what to do. When choosing our life partner, we call the algorithm
'love'. It remains happily beyond our understanding. Again, we can thank
evolution for this ability.

#### Quasi-rational Decisions

Let's move up the ladder to *quasi-rational* decisions. Here we know
that we are making a decision, and we know the trade offs being
weighted. But we do not know the decision-making algorithm.

This happens in situations like courts. Here we clearly define the
decision to be made and the trade offs to be weighted. A judge is
presented with a specific set of facts, and is supposed to consider
these and only these facts. The judge's job is to weight these facts to
determine a verdict. You may protest that we *do* know the
decision-making algorithm. It's dictated by the law, judicial precedent,
and so on. But alas, this is an illusion.

If the decision-making algorithm was determined by the legal system,
then all judges would render the same verdict if given the same facts.
But we know this is not what happens. Judges have different opinions and
different values. In the language used here, they have different
quasi-rational algorithms for making decisions. These algorithms are not
known to us, and it is doubtful they are known to the judges. Thanks to
evolution, we are equipped with a mental machinery that can weight
different facts. But we do not understand how this machinery works.

[Daniel
Dennett](https://www.theatlantic.com/technology/archive/2012/06/-a-perfect-and-beautiful-machine-what-darwins-theory-of-evolution-reveals-about-artificial-intelligence/258829/)
calls this "competence without comprehension". It's one of the
miraculous things about life. Organisms are little robots that are good
at reproducing. We don't need to understand why we do what we do. We
just need to do it.

#### Rational Decisions

Let's climb to the final rung of the ladder: *rational* decisions. Here
we know all of the ghosts in the machine. We know we are making a
decision, we know the trade offs, and we know the decision-making
algorithm. In humans, this type of decision is confined to very
artificial scenarios, such as mathematical logic.

The beauty of formal logic is that it *provides* the decision-making
algorithm. It defines how we weight different trade offs. For instance,
logic dictates that 5 is greater than 4. So if we need to decide which
number is greater, there is only one correct answer.

Economists use the term 'rational' with wild abandon. But I use the term
quite restrictively. I reserve 'rational choice' only for those
decisions where the algorithm is known to us. And because the
decision-making algorithm is known to us, we can prove that there is
only a single correct answer.

### Defining the ghost by what the machine can do

Having defined four types of decisions, let's return to agent-based
models. The curious thing is that virtually all such models live in the
rational decision-making space. In fact, we are really good at making
these models. Look at the success of chess-playing programs such as
[Deep Blue](https://en.wikipedia.org/wiki/Deep_Blue_(chess_computer)).
These programs are little computer agents that make rational decisions
about how to play chess. They weight every move by the probability that
it will lead to winning the game. And they are spectacularly successful.
They crush us humans.

But before we celebrate, we need to reflect on the goal of agent-based
models. The goal is not to *beat* humans at chess (or any other game or
task). No, the goal is to play the game *like a human would play*. This
means the simulation must make all the same mistakes that humans are
prone to make. The program must make *quasi-rational* decisions.

This task is more difficult than building a rational chess-playing
computer that can trounce humans. Only in the most contrived situations
do humans make purely rational decisions. As the logic gets harder, we
make mistakes. The algorithm of formal logic becomes too hard and we
start to rely on algorithms given to us by evolution. These algorithms
make mistakes. A correct model of human behaviour must simulate these
mistakes.

So why do most agent-based models live in the rational decision-making
space? It's simple. Rational decisions are the ones where we *know the
algorithm*. In games such as chess, the algorithm is defined by the
rules of the game. Because we know the decision-making algorithm (we
defined it when creating the game), we can teach it to a machine. We may
not be able to do the computations ourselves, but this is what computers
are good at.

The problem is that the real-world is not a game. We don't get to define
the decision-making algorithm. Most of the decisions that matter in our
lives are either intuitive or quasi-rational. This means the
decision-making algorithm remains unknown to us. Of course, this
algorithm should be scientifically studied. But we don't need
agent-based models to do this. We need experiments.

If we want to build agent-based models, it's far more fruitful to go to
the *lowest* ghost in the machine: *instinct*. Why? Because here the
decision-making algorithm is easiest to mimic. Take the behaviour of
people in [crowds](https://en.wikipedia.org/wiki/Crowd_simulation). In
crowds, humans behave instinctually. We move with the crowd, but also
protect our personal space. This makes our behavior fairly easy to
model. Other animals, such as
[penguins](https://phys.org/news/2018-04-king-penguin-colonies-fluids.html),
show similar behaviour.

There is another philosophical reason to model instinct. It guarantees
that our models will not overstep their bounds. To model instinct, we
need to find very specific situations where our instincts take over.
Crowd movement is a good example. Having created an agent-based model of
crowds, no one would dream of using it to explain state formation. Its
limited applicability is clear. This guards against [economic
imperialism](https://en.wikipedia.org/wiki/Economics_imperialism).

### Introspection: Does the ghost play tricks on us?

Microeconomics is a form of 'armchair psychology'. The early
microeconomists used self reflection to postulate algorithms for how
humans make decisions. For these economists, their introspection
indicated the self-evident truth of their theories. Here is [Lionel
Robbins](https://onlinelibrary.wiley.com/doi/full/10.1111/j.1468-0335.2009.00793.x)
making this claim with bombast:

> \[T\]he chief ... postulates \[in economics\] are all assumptions
> involving ... simple and indisputable facts of experience ...
>
> We do not need controlled experiments to establish their validity:
> they are so much the stuff of our everyday experience that they have
> only to be stated to be recognised as obvious.

We are lucky that engineers don't think like Robbins. Imagine an
engineer saying: "We don't need controlled experiments to establish the
strength of a bridge. Its strength is the obvious stuff of everyday
experience". This doesn't pass the laugh test. Yet this thinking forms
the basis of microeconomics.

My point is that we need to question the validity of *self reflection*.
Is it a valid tool for understanding how we make decisions? Suppose I
make a decision, and then reflect on how I made the decision. Here is
the problem. How do I know if my reflection is correct? I can't just
trust that it is. Self-delusion is one of the hallmarks of the human
experience. (I know this because I am the greatest scientist on Earth).
Having self-reflected, we must test the hypothesized algorithm to see if
it predicts *future* decisions.

This takes difficult experimental work. We need to set up
decision-making scenarios and test how humans behave. Then we make an
algorithm that mimics this behaviour. In other words, agent-based
modelling must be an outgrowth of *experimental psychology*. Cue
shudders from economists.

### When the ghosts talk to other ghosts

To make things even harder, it is not enough to model the decisions of
*individual* humans. The model must also mimic how humans behave in
*groups*.

The problem is that human groups are more than the sum of their parts.
When humans interact, we update our beliefs. So not only must we model
how individuals make decisions, we must also model how the
decision-making algorithm gets updated as individuals interact.

In groups, we adopt shared ideologies that influence our decision-making
algorithms. We're prone to groupthink. The resulting group behavior can
go in all sorts of directions. It can lead to military atrocities, to
great humanitarian works, to triumphs of science, to mass suicides, to
vain monuments, to creative visions, and so on.

When you stop to think about how complex this is, it boggles the mind.
Are we ready to model this complexity? I think not. We should be modest
and admit our ignorance on this front. It's better to admit we know
little, than to stubbornly hold on to theories that are wrong.

### The proper place for agent-based models

Having outlined all of the difficulties with agent-based models, I admit
that they have a place in science. They are useful as *thought
experiments*. We suppose that humans have a specific decision-making
algorithm. We then use an agent-based model to explore the consequences.
This is a perfectly valid tool for investigation.

But we don't get to pretend that this thought experiment says anything
about *actual* human behaviour. To do that, we need to test our
assumptions. We need to show that our decision-making algorithm is how
humans actually behave. As I've tried to show here, this is extremely
hard to do.

If agent-based modellers want to make strides in understanding human
behaviour, they will have to become a sub-discipline of experimental
psychology. Oh the horrors!

------------------------------------------------------------------------

#### Support this blog

Economics from the Top Down is where I share my ideas for how to create
a better economics. If you liked this post, please consider becoming a
patron. You'll help me continue my research, and continue to share it
with readers like you.

[![patron_button](https://i0.wp.com/economicsfromthetopdown.com/wp-content/uploads/2023/05/supporter_button-1.png?w=220&ssl=1){.size-thumbnail
.wp-image-1180 .aligncenter decoding="async" attachment-id="1180"
permalink="https://economicsfromthetopdown.com/patron_button/"
orig-file="https://i0.wp.com/economicsfromthetopdown.com/wp-content/uploads/2020/02/patron_button.png?fit=434%2C102&ssl=1"
orig-size="434,102" comments-opened="1"
image-meta="{\"aperture\":\"0\",\"credit\":\"\",\"camera\":\"\",\"caption\":\"\",\"created_timestamp\":\"0\",\"copyright\":\"\",\"focal_length\":\"0\",\"iso\":\"0\",\"shutter_speed\":\"0\",\"title\":\"\",\"orientation\":\"0\"}"
image-title="patron_button" image-description="" image-caption=""
medium-file="https://i0.wp.com/economicsfromthetopdown.com/wp-content/uploads/2020/02/patron_button.png?fit=300%2C71&ssl=1"
large-file="https://i0.wp.com/economicsfromthetopdown.com/wp-content/uploads/2020/02/patron_button.png?fit=434%2C102&ssl=1"
recalc-dims="1"}](https://economicsfromthetopdown.com/membership/){target="_blank"
rel="noopener"}

------------------------------------------------------------------------

#### Stay updated

Sign up to get email updates from this blog.

:::: {.jetpack_subscription_widget}
##  {#section .widgettitle}

::: {.wp-block-jetpack-subscriptions__container}
Email Address

Keep me up to date
:::
::::

------------------------------------------------------------------------

[![](https://i0.wp.com/economicsfromthetopdown.com/wp-content/uploads/2021/03/by.png?w=150&ssl=1){.aligncenter
decoding="async"
recalc-dims="1"}](http://creativecommons.org/licenses/by/4.0/){rel="license"}  
This work is licensed under a [Creative Commons Attribution 4.0
License](http://creativecommons.org/licenses/by/4.0/){rel="license"}.
You can use/share it anyway you want, provided you attribute it to me
(Blair Fix) and link to [Economics from the Top
Down](https://economicsfromthetopdown.com/).

------------------------------------------------------------------------


In the opening post of this blog, I described my ‘top-down’ approach to studying society. This means studying
<i>
 groups
</i>
of people without trying to reduce everything to the actions of individuals. It’s not that I think individual actions are
<i>
 unimportant
</i>
. Of course they are important. The problem is that we are hopelessly far from understanding the complexities of the human psyche. In this post, I’m going to dive into the problem.

Agent-Based Models and the Ghost in the Machine

Few insect behaviors are more iconic than the proverbial moths circling
the lamps at night.

::: {#attachment_1408 .wp-caption .aligncenter style="width: 510px"}
![Artist: Dave
McKean](https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/3246231_orig1.jpg){.size-full
.wp-image-1408 loading="lazy" decoding="async"
aria-describedby="caption-attachment-1408" width="500" height="500"
srcset="https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/3246231_orig1.jpg 500w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/3246231_orig1-150x150.jpg 150w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/3246231_orig1-300x300.jpg 300w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/3246231_orig1-50x50.jpg 50w"
sizes="(max-width: 500px) 100vw, 500px"}

Artist: [Dave McKean](https://www.davemckean.com/)
:::

These observations are prime examples of the supposedly stereotypic
insect responses to external stimuli. In contrast, in [our new paper
that just appeared
today](https://rsob.royalsocietypublishing.org/content/6/12/160229), we
describe experiments suggesting that insects appear to make a
value-based decision before approaching the light. However, compared to
us, an insect's decisions can take very different aspects into account.
For instance, when a fruit fly (*Drosophila*) decides whether to
approach light, it takes its flying ability into account. If any
parameter of flight is sufficiently compromised, it is better to hide in
the shadows, whereas the full ability to fly emboldens the animal to
seek out the light. In a way, these results are reminiscent of the
confidence with which a certain cartoon bee approaches a window:

Perhaps the most beautiful result of this work is the first
investigation into the neurobiological mechanisms underlying the
different valuation of light (or dark) stimuli in flying and non-flying
flies. We found that the tendency of flies to approach or avoid light is
not an all-or-nothing decision, but that different fly strains and
different manipulations of flying ability show different degrees of
approach/avoidance as well as indifference. Experiments with transgenic
flies showed that we can push the flies' preference back and forth along
this 'photopreference' gradient by activating or inhibiting neurons that
secrete either dopamine or octopamine, respectively. Octopamine and
Dopamine are so-called neuromodulators, known to be responsible for
valuation processes in other experiments across animals and in the case
of dopamine also humans. Commonly, they do this by modulating the
activity of neurons involved in processing sensory stimuli, such that
the value of these stimuli to the animal changes after the modulator is
applied.

In our case, activating dopamine neurons made flightless flies which
would otherwise avoid light, approach it. Activating octopamine neurons,
on the other hand, made normal flies, which approach light, hide in the
shadows, despite the manipulation leaving their flying capabilities
intact. The results obtained after inhibiting these neurons were mirror
symmetric: blocking dopamine neurons from firing made the flies seek
darkness without affecting their flying ability. Wingless flies with
their octopamine neurons blocked approached the light as if they could
fly. These results inspired the following illustration of how these
neuromodulators may cooperate to orchestrate the evolutionarily
advantageous decision for insects when faced with a light/dark choice:

::: {#attachment_1410 .wp-caption .aligncenter style="width: 510px"}
[![Illustration of the hypthetical balance between
octopaminergic/tyraminargic (OA/TA) and dopaminergic (DA) neurons
stabilzing the valuation of light/dark stimuli in fruit fly
photopreference
experiemnts.](https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/Fig-8-300x121.jpg){.wp-image-1410
loading="lazy" decoding="async"
aria-describedby="caption-attachment-1410" width="500" height="202"
srcset="https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/Fig-8-300x121.jpg 300w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/Fig-8-768x311.jpg 768w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/Fig-8-50x20.jpg 50w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/Fig-8.jpg 1005w"
sizes="(max-width: 500px) 100vw, 500px"}](https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/Fig-8.jpg)

Illustration of the hypothetical balance between
octopaminergic/tyraminargic (OA/TA) and dopaminergic (DA) neurons
establishing the valuation of light/dark stimuli in fruit fly
photopreference experiments.
:::

Future research will show whether these same neurons indeed change their
activity when the flying ability is manipulated, as one would expect
from these results.

It is possible that the evolutionary origin and ultimate ethological
relevance may be found in the behavior of flies which have just emerged
from their pupal case. The wings of these young adults are still folded
up as the insect first needs to pump blood into the veins of the wings
to expand them. During this time, the flies perch underneath horizontal
surfaces and avoid light. Even when the wings are fully expanded, but
not capable of supporting flight just yet, this behavior persists. Only
once the wings are ready for flight, do the flies perch on top of
horizontal surfaces and approach light.

::: {#attachment_1411 .wp-caption .aligncenter style="width: 530px"}
[![newly eclosed
fly](https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/dmel-antp-e-rnai-2013-04-18-16-09-471-1024x768.jpg){.wp-image-1411
.size-large loading="lazy" decoding="async"
aria-describedby="caption-attachment-1411" width="520" height="390"
srcset="https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/dmel-antp-e-rnai-2013-04-18-16-09-471-1024x768.jpg 1024w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/dmel-antp-e-rnai-2013-04-18-16-09-471-300x225.jpg 300w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/dmel-antp-e-rnai-2013-04-18-16-09-471-768x576.jpg 768w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/dmel-antp-e-rnai-2013-04-18-16-09-471-50x38.jpg 50w"
sizes="(max-width: 520px) 100vw, 520px"}](https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/dmel-antp-e-rnai-2013-04-18-16-09-471.jpg)

Source: hawaiireedlab.com
:::

Adult flies in the wild that live on rotting fruit probably face the
challenge of the sugary liquid of the fruit occasionally gumming up
their wings, only for the next rain to wash them clean again. We
mimicked this situation in one of our experiments by gluing the wings
together with sucrose solution and subsequently removing the sugar glue
from the flies' wings with water. As expected, the flies approached the
light before the treatment, avoided it when the wings were unusable and
approached it again after the 'shower'.

The first mention in the literature of adult flies with compromised
wings being less attracted by light was by [Robert McEwen in
1918](http://blogarchive.brembs.net/comment-n704.html). In the
intervening 49 years, we could not find any mention of this phenomenon
in the scholarly literature. Only in 1967, one of the founding fathers
of *Drosophila* neurogenetics, Seymour Benzer, [published
work](https://www.ncbi.nlm.nih.gov/pmc/articles/PMC335755/) mentioning
adult flies with deformed wings being less phototactic, but without any
insight into the underlying neurobiology. It took yet another 49 years
after Benzer's work without any mention in the literature, before our
paper described the first neurobiological components of this case of
insect behavioral flexibility, 98 years after the original discoverer
McEwen. Our postdoc, E. Axel Gorostiza, the first author of the paper,
will start his own laboratory on this topic, so it seems unlikely that
it will take yet another 49 years for the fourth publication on this
topic to appear.

Of course, all our raw data are available from
[figshare](https://figshare.com/articles/Phototactic_flexibility_Raw_data_/1502427).
This was also the first paper from our laboratory where all authors were
listed with their [ORCID](https://orcid.org/) IDs and all materials and
protocols were referenced with their
[RRID](https://scicrunch.org/resources)s and
[protocols.io](https://www.protocols.io/) DOIs, respectively. All
previous versions of the article are available as [biorxiv
preprints](https://biorxiv.org/content/early/2016/11/24/023846) as well.

------------------------------------------------------------------------

Original research article:

[A Decision Underlies Phototaxis in an
Insect](https://rsob.royalsocietypublishing.org/content/6/12/160229).
[Royal Society Open Biology]{style="text-decoration: underline;"}

*E Axel Gorostiza, Julien Colomb, Björn Brembs*

Freie Universität Berlin, Universität Regensburg, Germany.

Abstract:

Like a moth into the flame -- Phototaxis is an iconic example for innate
preferences. Such preferences likely reflect evolutionary adaptations to
predictable situations and have traditionally been conceptualized as
hard-wired stimulus-response links. Perhaps therefore, the century-old
discovery of flexibility in *Drosophila* phototaxis has received little
attention. Here we report that across several different behavioral
tests, light/dark preference tested in walking is dependent on various
aspects of flight. If we temporarily compromise flying ability, walking
photopreference reverses concomitantly. Neuronal activity in circuits
expressing dopamine and octopamine, respectively, plays a differential
role in photopreference, suggesting a potential involvement of these
biogenic amines in this case of behavioral flexibility. We conclude that
flies monitor their ability to fly, and that flying ability exerts a
fundamental effect on action selection in *Drosophila*. This work
suggests that even behaviors which appear simple and hard-wired comprise
a value-driven decision-making stage, negotiating the external situation
with the animal's internal state, before an action is selected.


Few insect behaviors are more iconic than the proverbial moths circling the lamps at night. Artist: Dave McKean  These observations are prime examples of the supposedly stereotypic insect responses to external stimuli. In contrast, in our new paper that just appeared today, we describe experiments suggesting that insects appear to make a value-based decision before approaching the light.

Why did the moth fly into the flame?

This is our first poster at this year's SfN meeting in San Diego. It's
about decision-making in fruit flies. We find a probabilistic form of
decision-making that suggests that without understanding the mechanisms
behind this fundamental uncertainty, we will never fully understand
decision-making.

[![gorostizasfn2016](https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/gorostizasfn2016-1024x613.png){.aligncenter
.size-large .wp-image-1401 fetchpriority="high" decoding="async"
width="520" height="311"
srcset="https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/gorostizasfn2016-1024x613.png 1024w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/gorostizasfn2016-300x180.png 300w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/gorostizasfn2016-768x460.png 768w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/gorostizasfn2016-50x30.png 50w, https://bjoern.brembs.net/wp/wp-content/uploads/2016/11/gorostizasfn2016.png 1352w"
sizes="(max-width: 520px) 100vw, 520px"}](https://bjoern.brembs.net/download/do-flies-in-groups-make-individual-choices/?wpdmdl=1398)Clicking
on the image will let you download the PDF Version of the poster. This
is our abstract:

In every behavioral population paradigm where groups of animals are
being exposed to forced-choice situations, there is the question whether
or not the individual animals can be assumed to make their own choices.
We approach this hypothesis by testing *Drosophila* fruit flies for
their photopreference in a light/dark T-maze. Approximately 75% of a
randomly chosen group of wild type flies decide to approach the bright
arm of the T-Maze, while the remaining 25% walk into the dark tube.
Taking these subgroups of flies and re-testing them revealed a similar
75-25 distribution in each subgroup.  
In order to increase the number of choices each subgroup makes without
losing too many flies in the process, we used the classic phototaxis
experiment developed by Seymour Benzer in the 1960s. In this experiment,
flies are exposed to a light source while they are confined in
transparent tubes aligned with the direction of light. Each round of the
experiment consists of 5 consecutive choices were the animal can either
stay or walk towards the light (positive phototaxis). At the end of a
round the original group is split into 6 subgroups according to their
sequence of choices.  
We discovered that while the test/re-test distributions were similar,
there was a tendency of the extremely phototactic animals (positive and
negative) to skew their distributions towards their respective end.  
These results are consistent with observations in single-animals where
individual choice probability was discovered to be itself distributed
over a population of flies (Kain et al., 2012). To test for potentially
confounding effects of general activity and walking speed, we tested
individual flies after their phototaxis experiments in Buridan's
Paradigm, where flies walk between two opposing black stripes. We
detected small walking speed and general activity differences,
suggesting a quantitative interaction between general and light-specific
processes contributing to the performance scores in Benzer's phototaxis
experiment.  
Kain JS, Stokes C, de Bivort BL (2012) Phototactic personality in fruit
flies and its suppression by serotonin and white. Proc Natl Acad Sci U S
A 109:19834--19839.


This is our first poster at this year’s SfN meeting in San Diego. It’s about decision-making in fruit flies. We find a probabilistic form of decision-making that suggests that without understanding the mechanisms behind this fundamental uncertainty, we will never fully understand decision-making. Clicking on the image will let you download the PDF Version of the poster.

Do flies in groups make individual choices?

*[During my [flyfishing vacation last
year](http://blogarchive.brembs.net/comment-n862.html), pretty much
nothing was happening on this blog. Now that I've migrated the blog to
WordPress, I can actually schedule posts to appear when in fact I'm not
even at the computer. I'm using this functionality to re-blog a few
posts from the [archives](http://blogarchive.brembs.net) during the
month of august while I'm away. This post is from [February 9,
2011](http://blogarchive.brembs.net/comment-n704.html):]{style="font-size: 12px;"}  
*

 

In this week's journal club, we talked about an old paper from 1918!
"[The reactions to light and to gravity in Drosophila and its
mutants](https://onlinelibrary.wiley.com/doi/10.1002/jez.1400250103/abstract){target="_blank"
rel="external"}" by Robert McEwen, in the *Journal of Experimental
Zoology*.

As the title says, the author studied how the fruit fly
*[Drosophila](http://brembs.net/learning/drosophila/){target="_blank"
rel="external"}* responds to light and gravity. He tested this in
walking flies and compared flies both with intact wings and clipped
wings, wing mutations, clipped antennae, glued wings or clipped middle
legs. He discovered that flies without wings or with mutated wing shape,
show less movement towards light (i.e., less phototaxis). This finding
was later confirmed by one of the founders of modern neurogenetics,
Seymour Benzer
([1967](https://www.pnas.org/content/58/3/1112.full.pdf+html?sid=a682d90d-38bb-468e-a0fe-516988a69491){target="_blank"})
and we also find this in our experiments. We have now set out to find
out which neuronal mechanisms are involved in this drastic change in
behavior.

In order to get the flies to show phototaxis, McEwen developed a machine
to gently tap the tube in which the flies were placed, to get them to
walk. He described the necessity for flies to be active in order to show
a consistent orientation towards a light source: without walking
behavior being either initiated spontaneously or by the tapping machine,
flies would not walk towards the light themselves. If the flies were at
rest, light was not an orienting stimulus for them. This key insight was
formulated by McEwen at the very end of the paper:

> <div>
>
> Lastly, it may be well to emphasize the peculiar relation which exists
> in
> [Drosophila](http://brembs.net/learning/drosophila/){target="_blank"
> rel="external"} between general activity and phototropism. This
> phenomenon has been clearly recognized by Carpenter and in general I
> agree with this author's conclusions. The fact seems to be that this
> insect is not phototropic unless it is in a certain physiological
> state brought on by, or at least accompanied by, activity. When the
> fly reaches a certain degree of activity, induced by various means, it
> suddenly becomes phototropic. When it quiets down, however, it may
> still crawl about but ceases to be phototropic. Thus, when an insect
> has been exposed to constant illumination for some time, it no longer
> orients to light but wanders aimlessly up and down the tube.
> Eventually such an animal may even come to rest with its head away
> from the source of light.
>
> </div>

The technique described mimics what other colleagues have later
developed in other fly paradigms based on vision and walking, such as
the
"[fly-stampede](https://www.landesbioscience.com/journals/fly/article/9139){target="_blank"}"
paradigm. But the insight reaches much further than that. More recent
research has shown that the state of the animal has minute control over
how the environment is processed. For instance, leeches respond with
various behaviors to local mechanosensory stimulation (i.e., touch).
However, when they feed, the biogenic amine serotonin is released and
prevents the mechanosensory neurons from transmitting the stimuli -- the
animal becomes unresponsive when it feeds ([Gaudry & Kristan,
2009](https://www.nature.com/neuro/journal/v12/n11/full/nn.2400.html){target="_blank"}).
Another study showed that motion-sensitive neurons in the optic lobes of
the fly brain increase their gain when the fly is flying, as opposed to
when it is not flying ([Maimon et al.,
2010](https://www.nature.com/neuro/journal/v13/n3/abs/nn.2492.html){target="_blank"}).
Analogous results were obtained when walking vs. sitting flies were
compared ([Chiappe et al.
2010](https://www.sciencedirect.com/science?_ob=ArticleURL&_udi=B6VRT-50KKY4V-4&_user=4113557&_coverDate=08%2F24%2F2010&_rdoc=1&_fmt=high&_orig=search&_origin=search&_sort=d&_docanchor=&view=c&_rerunOrigin=scholar.google&_acct=C000047720&_version=1&_urlVersion=0&_userid=4113557&md5=587c1ad8c1256e6f1543427990606546&searchtype=a){target="_blank"}).
In another, also very sophisticated study, Haag et al.
([2010](https://www.pnas.org/content/107/46/20104.full){target="_blank"})
showed how an identified motor neuron responds more strongly to visual
input when the animal is flying than when it is at rest. Finally, Tang
and Juusola
([2010](https://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0014455))
report evidence that the direction in which a fly attempts to turn
changes the way in which the optic lobes process the visual information
on the side towards the fly attempts to turn, compared to the
contralateral side.

All these groups have, largely independently of each other, discovered
the biological mechanisms for something that already McEwen (and
Carpenter, cited there) had understood: animals don't just respond to
stimuli in always the same, stereotypical way: all animals have many
different ways to treat and evaluate the incoming sensory stream,
depending on what they are doing at the moment. The decisive factor for
understanding animal behavior is not the environment, or the sensory
organs, it is the animal itself. Apparently, this profound insight was
known long ago and we're just rediscovering it now, in various places,
all over the world.

Something *was* new in all the recent studies, though: they all provide
first mechanistic insight into how brains balance internal and external
processing. All these studies show that there seems to be a smooth
gradient between decision-making and attention-like processing, even in
invertebrates: Gaudry and Kristan call it decision-making, when their
leeches 'decide' to ignore stimuli while they feed, even though the
incoming sensory stimuli are blocked already at the very first synapse.
Chiappe et al., on the other hand, relate their phenomenon to attention
and Haag et al. also mention attention in their paper, with their
effects being observed many synapses downstream of the sensory neurons
-- the word 'decision' does not occur in either of the two papers. It
appears as if future neurophysiological research is bound to show that
the distinction between attention-like mechanisms and decision-making,
which seems so intuitive and clear-cut, may dissolve when we start to
unravel how brains actually do it. Now when will we come accross the
ancient text that already pre-empts that insight?
![nuetral.png](http://blogarchive.brembs.net/e107_images/emotes/yellow/nuetral.png){decoding="async"}

References:

------------------------------------------------------------------------

McEwen, R. (1918). The reactions to light and to gravity in
[Drosophila](http://brembs.net/learning/drosophila/){target="_blank"
rel="external"} and its mutants Journal of Experimental Zoology, 25 (1),
49-106 DOI:
[10.1002/jez.1400250103](https://dx.doi.org/10.1002/jez.1400250103){rev="review"}  
Zhu, Y., & Frye, M. (2009). Neurogenetics and the "fly-stampede":
dissecting neural circuits involved in visual behaviors Fly, 3 (3),
209-211 DOI:
[10.4161/fly.3.3.9139](https://dx.doi.org/10.4161/fly.3.3.9139){rev="review"}  
Gaudry, Q., & Kristan, W. (2009). Behavioral choice by presynaptic
inhibition of tactile sensory terminals Nature Neuroscience, 12 (11),
1450-1457 DOI:
[10.1038/nn.2400](https://dx.doi.org/10.1038/nn.2400){rev="review"}  
Maimon, G., Straw, A., & Dickinson, M. (2010). Active flight increases
the gain of visual motion processing in
[Drosophila](http://brembs.net/learning/drosophila/){target="_blank"
rel="external"} Nature Neuroscience, 13 (3), 393-399 DOI:
[10.1038/nn.2492](https://dx.doi.org/10.1038/nn.2492){rev="review"}  
Chiappe, M., Seelig, J., Reiser, M., & Jayaraman, V. (2010). Walking
Modulates Speed Sensitivity in
[Drosophila](http://brembs.net/learning/drosophila/){target="_blank"
rel="external"} Motion Vision Current Biology, 20 (16), 1470-1475 DOI:
[10.1016/j.cub.2010.06.072](https://dx.doi.org/10.1016/j.cub.2010.06.072){rev="review"}  
Haag, J., Wertz, A., & Borst, A. (2010). Central gating of fly optomotor
response Proceedings of the National Academy of Sciences, 107 (46),
20104-20109 DOI:
[10.1073/pnas.1009381107](https://dx.doi.org/10.1073/pnas.1009381107){rev="review"}  
Tang, S., & Juusola, M. (2010). Intrinsic Activity in the Fly Brain
Gates Visual Information during Behavioral Choices [PLoS
ONE](https://plosone.org){target="_blank" rel="external"}, 5 (12) DOI:
[10.1371/journal.pone.0014455](https://dx.doi.org/10.1371/journal.pone.0014455){rev="review"}


<em>
 During my flyfishing vacation last year, pretty much nothing was happening on this blog. Now that I’ve migrated the blog to WordPress, I can actually schedule posts to appear when in fact I’m not even at the computer. I’m using this functionality to re-blog a few posts from the archives during the month of august while I’m away. This post is from February 9, 2011:
</em>

Messages de Rogue Scholar

What is a decision?

Agent-Based Models and the Ghost in the Machine

Why did the moth fly into the flame?

Do flies in groups make individual choices?

Flashback: Nothing new in science?