Why Science and Philosophy Need Each Other | Lauren Ross & Megan Peters - Mind-Body Solution

Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:08):
Megan, Lauren, thanks so much for joining me today.
We have an exciting colloquium planned.
The topic is why science and philosophy need each other.
I can't think of the perfect twoguests to be on the show because
Lauren, you could be seen as a ascientifically informed
philosopher. Megan, you could be seen as a
philosophically informed scientist.
However, one could argue that both of you are actually both.

(00:30):
So I think that brings us to this first question, which is
fundamental to setting the stage.
Why do both of you think scienceand philosophy are often seen as
separate, even antagonistic, when historically they emerged
together? Lauren, do you want to perhaps
start? And then Megan, you can take it
from there. Sounds great.
It's a very interesting question, and it partly has to

(00:54):
do with how we think of both science and philosophy in modern
times. For science, usually we have
more awareness and understandingof what we're referring to, but
with philosophy, that's less thecase.

(01:16):
Philosophy is this label that can be applied to many different
things, many different types of thinking.
I mean, philosophy is also a field, of course, that involves
different types of work. In everyday life conversations,
we use the term philosophy to sometimes refer to places where

(01:37):
you have questions without answers, a place where anything
goes, where maybe we're interested in your own
subjective views, your beliefs and your thoughts.
And sometimes philosophy in thisspace is viewed as the opposite
of pragmatic. It's your philosophical musings

(01:59):
or, or you'll sometimes hear even scientists say that they're
asking a philosophical question where they mean there's this
sort of unbounded open question.Or you'll hear the expression
near philosophy. So I I think that starts to show
and paint a picture for how science and philosophy can be

(02:20):
seen as opposites. And I would say mistakenly,
because that picture of philosophy is not how we think
of science. This picture of philosophy
questions without answers and the sort of unbounded anything
goes, well, science doesn't operate like that.
And part of what can help here is to specify different types of
philosophy, and in this case, philosophy of science,

(02:43):
philosophy of mind, philosophy of cognitive science, where we
don't have an anything goes typeproject.
We have an interest in the principles of science.
We have an interest in precision, clarity, rigor.
And so this would be one reason I would suggest for why they're

(03:03):
sometimes seen as separate or incontrast.
It'll partly depend on who you talk to.
Many scientists do view them as helpful colleagues that kind of
need to work together or that when they do work together can
lead to various types of successes.
So that would be a first answer that I would give.

(03:26):
Megan same question. Yeah, actually there's a lot of
of what Lauren said that I agreewith, but I'm actually kind of
surprised by this question in general, because from my
perspective, I don't find them antagonistic at all.
And I guess it's my privilege that the philosophers and
scientists that I tend to hang out with might agree with me

(03:48):
that they are not antagonistic. But again, that's, that's my own
privilege in the space that I, Ichoose to occupy and that I'm
privileged to occupy. I I feel that, yes, there is
this general tenor, this generalfeeling that philosophy and
science might be separate, antagonistic, because there's

(04:08):
the empirical scientists who aredoing the real work.
And it's the philosophers who are over here in their armchairs
kind of, you know, deciding thatthere is a difference when maybe
there isn't really a difference in in the real world.
And so there could be that tension where it's like, do we
really need to be having that particular type of conversation?
Does that really matter to the experiments that I'm going to be

(04:29):
doing? But I think that again, the
folks that I tend to interact with within the science and the
philosophy space recognize that these are truly synergistic, not
even just friendly, that they both can learn from each other
in the ways that that Lauren pointed out.

(04:52):
But I will disagree, Lauren, with one thing that you said,
which is that the realm of questions without answers might
be more the philosophical realm.And I feel like there's just so
much of science that is that too.
There's so much where the purpose of what we're doing as
scientists at the at the cuttingedge, at the forefront of our
knowledge is, well, maybe there isn't an answer yet, but the

(05:15):
role of philosophy and the role of philosophically informed
science could be to discover, isthere an answer to be had here?
So we don't know the answer yet and we need to decide, is this
something that we could actuallygo after scientifically?
So a lot of what I do is a question without an answer yet
too. So, yeah, I think that, sure,

(05:39):
the general tenor might be that these are antagonistic, but I
think that there are quite a lotof people who also disagree with
that general assessment. Yeah, I think that a lot of
people actually forget that whenyou do a PhD in in anything
scientific, you're just fundamentally getting a
doctorate in philosophy because I mean, this, this is an

(06:00):
enriched history of just philosophy that's expanded over
time. Natural philosophy has become
what science is today. Lauren, for you, as a
philosopher of science and a trained physician, I know as a
medical doctor, it's that most people who when I worked in the
medical field, when I do work inthe medical field, is that it's
very uninformed philosophically.It's a quite a common theme.

(06:23):
They're not not interested in it.
It's just that they just don't have the time perhaps to explore
this as much as maybe someone like like myself does.
So what do you think scientists most often misunderstand about
what philosophy actually contributes?
I think that scientists, if theyare misunderstanding what

(06:45):
philosophy is here, primarily philosophy of science and what
it can contribute. It's this view that philosophy
is sort of so open and so anything goes that it isn't
useful and that it's this base where you can't always get

(07:10):
traction on questions or how to understand the world.
Part of what's coming up in thisquestion and part of, as you
mentioned, Ted, why this is so interesting, the kind of current
views in some areas that these are disparate types of study is
that they very much were part ofthe same program in early work.

(07:32):
So natural philosophy, we refer to early scientists as natural
philosophers. Aristotle is both someone we
think of as an early biologist and also a philosopher.
And in in Darwin's lifetime, theterm scientist was created.
And so he was a natural philosopher perhaps early on in

(07:52):
his career and then was only referred to as a scientist
later. So they very much do have a
shared root and a shared history.
Currently, the common misconceptions I see is that
philosophy is so open that you can't use it to get guidance.

(08:14):
And that's, as Megan is suggesting, very much
antithetical to the kind of workthat you see in a lot of
philosophy of Cogsai, a lot of philosophy of mind, a lot of
philosophy of neuroscience. This is a space where you have
scientists, sorry, you have philosophers who are doing work

(08:34):
where they're interested in, I'dsay 3 main things.
There's an interest in getting precision about foundational
topics and methods in science. They want to know the principles
and the justification that are guiding those concepts, those
methods. And then they want to know, and

(08:56):
they want to be able to specify how something works.
If you have a scientist giving an explanation, how do you know
it's a good one? If you have scientists debating
how we should understand causation or what the mechanism
is for something, how do you know when it works and when it

(09:16):
doesn't? So here we're often looking at
science from a functional perspective where scientists
have goals and you can assess the success of science with
respect to when scientists are reaching those goals.
And so in this space, we think of science as a as a, a practice
that gives us our best understanding of the world.

(09:38):
And it often involves this theorizing that we sometimes
call philosophy, that scientistsare very much doing and
philosophers of science are engaged in as well, where
you're, you're looking at these fundamental scientific concepts
and practices that scientists engage in.
If science does give us our bestunderstanding of the world, we

(10:01):
should be able to say how it does.
And that's where here it's helpful to get precision about
what is an explanation in science, what is causation?
What is getting information about the causal structure of
the world? What are the principles that
scientists use that we can identify to help guide work in

(10:22):
this space? And then how do you know when it
works? How do you know when scientists
have met the standards of their field?
And that partly involves specifying what they are.
And so I think it's sometimes surprising to physicians,
healthcare practitioners who aremore in a professional space and

(10:42):
they aren't necessarily theorizing the way that other
types of scientists are, to hearthat this is a kind of
philosophy of that this is a kind of work that happens in
philosophy and philosophy of science.
So, yeah, there's a kind of difference, I think, across

(11:06):
types of researchers. Some of them are more on the
front lines of professional work, and maybe others are more
engaged with research. And you have some of these
researchers who are working withphilosophers and kind of
interested in these theoretical questions that show up in
philosophy of science. Megan As a neuroscientist deeply

(11:30):
grounded in philosophy, what do philosophers sometimes overlook
about how scientific practice really works today?
You're asking me to to say what's wrong with all my
colleagues. So I think for me this is a, a
challenge that maybe philosophers face more than some
scientists, but scientists certainly face this challenge as

(11:52):
well. And that is, I hinted at this
earlier, this idea of is this a difference that makes a
difference? So philosophers of science,
philosophers of mind, philosophers of modeling of
cognitive science will often tryto drive at the the conceptual
distinctions that provide clarity with respect to the

(12:13):
questions that we're asking and an assessment of the validity of
the methods that we're using to answer those questions.
But sometimes I think philosophers and scientists to a
certain extent as well, we get so into the details of finding
the joints in nature, you know, finding the separation between

(12:34):
two concepts that ultimately, ifwe were to take a step back and
say, all right, well, maybe there is this separation in
concepts that you've identified,this difference between concept
A and concept B that you've started to really home in on.
How could we ever know if that'sa real difference, if that, And

(12:55):
not just that this is a difference that we can
conceptualize, that we can come up with, that we can describe,
but that this is a real difference in the world.
This is a real joint in nature. And I think that sometimes the
pushback that scientists will give towards philosophers is

(13:16):
this like, Oh, well, you're making distinctions that don't
really have any bearing on anything that's physical, that's
real, that's empirical. And so you're, you're really
just kind of in this space as asLauren said, that where
everything, anything goes, like you've, you've discovered a
difference and you've decided that that's an important
difference. But I think that the hard part

(13:38):
is, is not just dismissing thesedifferences or these
distinctions and saying, well, Icould never test for them.
So it's not, it's not a meaningful distinction.
The hard part is deciding whether there is a meaningful
distinction there. And so deciding whether this is
a problem where philosopher of science or cogs I or modeling

(14:00):
has come up with this distinction that may or may not
be empirically testable. And the challenge is to say, do
we care to empirically test this?
And if we do care to empiricallytest it, can we even come up
with something that would allow us to see whether this joint in
nature is actually present? And so I think that that's a

(14:21):
hard hump between science and and some philosophy, where some
more pure philosophers of science will see the intrinsic
value of making the distinction and clarifying it to begin with.
And some empirical scientists will say, well, that's great,
you can write it down, That's lovely.

(14:41):
You can draw a picture. But like, do I actually care?
Is this a thing that I can go and find with some sort of
empirical study? So that, I think would be the
closest thing that I can think of to a a kind of something that
philosophers might overlook or that they the relative value
placed on that enterprise is different between philosophy of

(15:05):
science and empirical science. Megan, before we started, while
we were waiting, Lauren, we've accidentally sent.
I might have missed mistakenly not sent the ring link to the
right place, but we were chatting about one of our
favorite heroes, who's Dan Daniel Dennett.
And, and I often looked at Dan growing growing up as a as a
neuroscientist and a philosopher.

(15:25):
He was someone so ingrained intoboth of these fields.
And and he often touched on thisdeeper reflection culturally.
So this whole this, does this reflect something deeper,
perhaps so like objectivity versus reflection or shut up and
calculate versus anything goes. What do both of you think about
this? And how might we bridge this

(15:47):
divide to transform how we studythe mind and consciousness?
It partly relates to what has come up already because, as
Megan suggested, it's sometimes is confusing to think of there
being a difference between this kind of philosophical work in
science, because you see scientists who are engaged in

(16:11):
philosophical questions and theorizing.
So from my perspective, I look at what they're doing and
they're doing philosophy, and then looking at these
philosophers who are interested in providing analysis and
accounts that, as Megan was suggesting, kind of latch onto
the world matter. You can do something with them.
You can show why this would be agood account to have or not.

(16:36):
You see how both of them are really interrelated types of
projects. And so I think it partly boils
down to sometimes we have cartoon pictures of both.
We have a kind of cartoon picture of a scientist who just
takes out a measuring device andgoes out and studies the world.

(16:57):
And what you miss if you look atthat picture, is all the
theorizing that took place before you set up that
experiment. There's so many assumptions
involved. There's so many methods you can
choose from. There's so many questions that
scientists need to and do ask themselves an answer before they
just go out and get the objective facts about the world.

(17:19):
And it's going to depend on the questions they ask, which is
partly what Megan brought up. And in some cases, you've got to
ask the right kind of question too, or appreciate the different
questions require different methods and then they give you
different answers. This is also just fascinating
from the standpoint of how complex the world is.
Scientists have to deal with that and they want order.

(17:42):
And it's fascinating how they'reable to do that given the
complexity of the world. And so, you know, they are able
to do that. We kind of look at the places
where they've done it, and then we're looking at these other
situations where there's a a complex new question, there's
some new territory they're trying to understand, right?

(18:04):
Is the brain the most complicated machine on the
planet? You know, the brain is so
complicated, the world is so complicated.
And so you see how they're making decisions about what to
do with that because they can't cite all detail that's out there
and not all of it matters. So they have to figure out what

(18:27):
details of the world matter and how to carve out questions that
allow them to give principled answers to those kind of topics
of interest. And so this is both science and
philosophy, the way I think Megan and I often see it.
But if you have a toy picture ofscience and a toy picture of

(18:49):
philosophy, they look very distinct.
And there are types of philosophy, as Megan is
suggesting, where it's more armchair type work.
And in this case what we want iswe want philosophy that's useful
for these kind of scientific questions.
And we see many examples of that.
That might be that would be partof the kind of answer I would

(19:12):
give. I, I totally agree with the, the
cartoonification of science versus philosophy and in in
particular this version of science, which is that you pull
out your, you know, measurement O meter or whatever, and you
point it at the thing and you get some sort of objective
answer. And so this mischaracterization

(19:35):
of philosophy as this subjectiveanything goes and science is
objective and like, definitely we're just measuring the world.
No, like there is no such thing as objective science.
Sorry, but there just isn't that.
We carry, as Lauren you said, the the assumptions that we make

(19:55):
about the structure of reality, about the types of measurements
that are going to be useful, about the types of models that
we can build that will be usefulto answering a particular type
of question or retrieving a particular type of explanatory
goal. There's so many cases where if
you actually do kind of a historical overview of a

(20:16):
particular niche field. For example, like this
particular type of model of decisions and reaction times in,
in neuroscience, you know, how, how do people make decisions in
a noisy environment and how longdoes it take them to come to a
decision under the conditions ofnoise in the world?
You know, you're driving down a foggy Rd. it's foggy.
How long do you take to decide what you're seeing, right?

(20:40):
And what do you decide that you're seeing?
There's models of that kind of decision process and those
models have been successful for literally decades since they
were developed. And there's been a lot of really
beautiful work to say this is now like the dominant
explanation of how we make thesetypes of decisions.
But they have assumptions and those assumptions, Dr., the

(21:02):
experiments that are done to to generate the objective empirical
data that then goes on to validate or 'cause those models
to be modified a little bit. And if you take a step back and
you look at those assumptions, they have constrained the space
of inquiry in a way that obscured potential alternative

(21:25):
explanations. So this is a particular hobby
horse of mine because we've got a couple papers on this
recently. But I think the general
principle applies across all of science, not just cognitive
science and psychology and you know, complexity sciences within
neuroscience, but in general, the, the way you think the world

(21:46):
works and the models that you'vebuilt and their relative success
in capturing the components of the system that you're trying to
explain that gives you myopia. And if you don't get out of
that, if you don't take off the blinders, you're going to miss a
whole lot. And simply the recognition that

(22:07):
you have blinders on in the first place allows you to
acknowledge that science is not an objective enterprise, that
there is always a scientist in the picture, and that we are
human beings and we have biases and we have preconceived notions
and we have assumptions and we we shape the way that we go
about trying to understand the world in ways that we not, we

(22:30):
may not be fully aware of at all.
Those biases and implicit assumptions are, are implicit.
They are deeply buried and and they're going to shape the new
models that we built. So I fully, fully agree with
Lauren here. And this is another case where I
think there's a, you know, seeming divide between

(22:50):
objectivity and subjectivity, philosophy versus science, that
kind of thing. And it's we're kidding ourselves
if we think that science is truly objective, because it just
really is not. Well, OK, so the stage is set,
and I think now would be a greatway to explore both of your work
together while trying to illuminate each other's work.
So in that, with that being said, let's try this.

(23:13):
Megan, perhaps could you tell uswhy Lauren's work helps
illuminate? So let's say, why does her
philosophical work help illuminate science?
And then I'm going to ask you, Lauren, do the same question but
in reverse. Sure.
So I as probably was, was said, you know, in, in my
introduction, you can go like Google both of us.

(23:35):
So I am a a philosopher and scientist of subjective
experience. I study the brain and the mind.
I try to reverse engineer the software that's running on the
wetware of our brains and how that creates the subjective
experiences that you have of theworld and the models that you
build and query and kind of run forward to predict what's going

(23:57):
to happen in your environment and how you're going to interact
with it. So the kind of work that Lauren
does is really helpful to me because it brings this
conceptual clarity. You know, the consciousness
science as a broadly writ field is a little bit all over the
place. You've got everybody from folks
who are studying this from, you know, kind of the, the quantum

(24:20):
or mathematical side. And then you've got the
cognitive neuroscientists who like to go look at brain
activity. And then you've got the
theoreticians. So it's, it's a little bit all
over the place, like a lot of fields.
Sure, you've got a lot of interdisciplinarity, but the
nature of what we are studying as folks who are interested in
subjective experience is even less objectively identifiable

(24:43):
than basically anything else in the world because it is the
thing that lives inside your head by definition.
And so having clarity on those concepts or seeking clarity on
those concepts, what do I mean when I say consciousness, when I
say subjective experience, when I say qualitative experience?
This gives us Lauren's work. And, and I saw this very clearly

(25:08):
actually at the Southern California Consciousness
Conference that we both went to,I don't know, last spring where
Lauren kept pushing the rest of us scientists in the room to
say, what is it in what actuallyare you trying to explain?
What is the target of your explanation?
Because every time you all say the word consciousness, I'm
paraphrasing here, Lauren was a lot more, you know, diplomatic.

(25:30):
But basically, you know, every time that we said the word
consciousness, everybody in the room meant something slightly
different. And it wasn't.
This isn't just a taxonomic or linguistic problem.
This is a conceptual clarity problem.
And so I think that for the kindof work that I do and even more

(25:51):
expansive, the kind of work thatany cognitive scientist or
computational neuroscientist does, where we're really trying
to reverse engineer the softwareof the mind.
In a lot of ways, the target of the explanation itself is
unclear from the beginning. And it's really hard to come up
with a nicely constrained littlebox to live in and say that is

(26:12):
the thing that I want to explain.
And so this is where someone like Lauren and Lauren is
particularly good at doing this in a way that corrals the cats
and herds the cats into coming up with something useful.
It's it's really valuable because without that clarity,
we're just going to have the same conversations over and over

(26:34):
and over again and they will always devolve into what is it
that we're even trying to understand.
Lauren, same question, but aboutMegan's work.
Perfect. It's, it's so important as a
philosopher of science to talk to actual scientists to make

(26:55):
sure that the way you're characterizing what they do
makes sense, is accurate, and itsort of keeps you in check a
bit. One of the challenges of my
field is that sometimes philosophers will have toy

(27:15):
simplified characterizations of what scientists are interested
in, what they want to explain, and then what they're doing in
the 1st place. And so one of the areas I work
on is scientific explanation. How do scientists give
explanations? How do you know they've got a
real one? What are the standards that need

(27:35):
to be met? Well, one thing you need as a
philosopher of science, if you're going to do that well, is
you need to capture the actual explanatory targets that
scientists are interested in. And so one of the many values of
talking to Megan is looking at the types of explanatory targets

(27:55):
that she's interested in her work and then in her field,
they're far more complicated than a lot of the more simple
models we have for how explanations work.
And so if we're going to providehopeful, accurate accounts,
scientific explanation, we need to make sure that we're not just

(28:19):
talking about explaining how if you throw a rock and a bottle,
it shatters, which is a, you know, there's these kind of
classic examples that show up a lot in philosophy that are often
quite simple. They have an explanatory target
that's binary. It sort of happens or it
doesn't. And you can even think of these

(28:40):
examples that are more sciency. So you might want to explain eye
color in a fruit fly. There's different colors that
will show up and you want to know well what explains why it's
got red eyes or white or black. Or you might want to explain the
height of a plant. You have genetically identical
plants and they've got differentheights.
What explains that? Those are getting us real

(29:03):
scientific examples, but those are so much more simplified and
not complex when you compare it to something like explaining
subjective experience. When you look at explaining
consciousness, even when you look at explaining disease
outcomes that are harder to identify and measure.

(29:27):
And so keeping us honest, right?And so that's one of the main
advantages of of working with Megan is it keeps your
philosophy honest, both in termsof are we actually capturing the

(29:47):
phenomena in the world that scientists are interested, that
they're studying and then how they do it.
So another nice thing that Meganmentioned is that scientists,
you know, and humans, when we'rereasoning in everyday life and
in scientific context, we have limited information about the
world. We don't have that picture where

(30:10):
you've got information about allof the details.
And so one of the features we need to include in our accounts
is that limitation. When humans reason, there's
limitations in terms of computational abilities,
computational power, the time scale in which they're making

(30:31):
decisions. Scientists are humans and so
what's what's important is our accounts of explanation need to
include those limitations, but also they managed to be
successful despite those constraints.
And so part of what is so helpful about interdisciplinary

(30:51):
connections of being a philosopher of science, working
with an actual scientist is thatwhen we're coming up with
accounts of how scientific practice and explanations work,
you can actually check it with the practice of scientists that
are right next door to you. You can talk to them about it.

(31:13):
You can make sure that you have clarity on what their goals are,
right? That's, that's something that's
very important for, in order forus to provide criteria for
explanation or ways of understanding causality that are
useful, that we need to know what goals scientists have.
And then are these concepts useful for their goals.

(31:35):
And so there's a a whole host ofreasons why working with Megan
and talking with Megan kind of helps keep my philosophy honest
in a way that I wouldn't be ableto do on my own, right?
Because he's doing that kind of scientific work in a way that

(31:57):
I'm not. So it's a big advantage of this
interdisciplinary approach. Yeah, I completely agree.
I think that both of you work works together.
It's a very symbiotic relationship.
It's it's something that should be seen as one.
And I think that by the end of this conversation, hopefully you
both do identify as both philosopher and neuroscientist.
But Megan, let's let's go to your work for a moment.

(32:18):
In computational and cognitive neuroscience, models attempt to
capture how the brain handles uncertainty.
What can these models truly reveal about something you just
touched on earlier, subjective experience.
So if this is truly subjective, are these models going to give
us any sort of objective information?

(32:39):
Yeah, great, great question. And this is maybe not the hard
problem, but this is one of the hard questions, right.
So the, the idea here is can anyempirical science give us any
sort of foothold or toehold or fingernail hold on, on something
that we might refer to as the hard problem?
And then the nature of subjective experience.

(33:00):
And I, I think, you know, I'm gonna use a couple overused
examples here maybe to explain where I'm coming from.
But a lot of folks that in the philosophically informed science
of consciousness might say that consciousness science right now

(33:20):
is in the state that life sciences was, you know, several
100 years ago where there was this magical force that we
called life. And it was this vital force.
And we didn't know what it was, but it was like a thing that was
out there and, and it was magic.And that as we learned more
about biology, the problem just kind of dissolved that we found
ways of describing and explaining what was going on

(33:44):
that made it very clear, well, this is a thing that's alive and
this is a thing that's not alive.
And this is a thing that's maybehalfway in between like viruses
and we're not really sure whether they're alive or not
alive by, by different definitions.
But it kind of doesn't matter where the the bifurcation where
we put that binary point anymore.
And I feel like I agree with thefolks who will, who will state

(34:05):
that consciousness science may have a similar future ahead of
it, where right now we have thismonolithic thing that we call
consciousness or subjective experience.
And it seems like there is this massive explanatory gap, but the
reality very well could be that as we approach that explanatory

(34:27):
gap, it it shrinks and it, it appears to be this big chasm
from over here. But as we take tiny baby steps
towards it, it turns out that that was an illusion or a
barrage or, or something. So I think that the work that
we're doing on how the brain deals with uncertainty, how it
arrives at the best that it it does, that kind of inference to

(34:49):
the best explanation. You know, your brain is itself a
natural philosopher and that it's trying to understand the
environment and build a model ofthe environment all the time.
It's doing what scientists are trying to do with, you know,
with help from philosophers of science.
And so I think that in a way, understanding how the brain is,

(35:11):
is building these models of the world.
The result of those models is ultimately somehow magically our
subjective experience. Unless you want to deny that
subjective experience exists. And that again, might be like,
OK, I'm going to leave that overthere for the the folks who want
to to argue that maybe subjective experience doesn't
exist. But for me, it's a useful

(35:33):
assumption to say, Yep, subjective experience exists,
conscious awareness exists. So I'm going to try to build
ways of capturing variants in itand linking that variance to
simplified components of models that I build.

(35:56):
If I twist this knob in my model, it predicts that some
sort of output on the subjectiveexperience side is going to
change in a particular way. I'd go do an experiment.
Yeah, it did OK when people say,oh, I have a stronger subjective
experience. OK, so maybe I'm on to something
there. I'll link it up with the brain
and say, OK, if I twist this knob, then I see like this area

(36:17):
of the brain lights up more or the pattern changes or
something, then I can say, OK, Ithink that this is the nature of
the information being represented in the patterns of
neural activity. And it maps onto this component
of the model and it maps onto this report of your subjective
experience. So that's how I'm trying to go
about it. I'm not going to say that any

(36:38):
work that I'm doing is, is building, is solving any sort of
hard problem or jumping any sortof explanatory gap.
But I think that if we sit over here and we say, hey, look at
that explanatory gap, it's the size of the Grand Canyon.
I'm not even going to bother approaching it to see how big it
is. I don't think that that's a

(37:00):
useful enterprise. So I want to take, I want to
create approaches to take those baby steps.
And that's some of the work thatwe're doing on metacognition
specifically is not just understanding how the brain kind
of builds models of the world orhow the mind builds models of
the world, but how it also puts itself into those models, how it
builds models of itself. And the subjective experiences

(37:23):
that we have are ultimately the reflection of a combination of
the model that we've built of our environments and kind of our
own understanding or introspective insight into that
model that we built that we can query and and evaluate that
model and look at it. So that's how I use uncertainty

(37:43):
or noise or variation is to lookfor how it dictates how it
interacts with the subjective experiences that we can report
in these kinds of experimental approaches.
Lauren, my question for you it, you'll notice that these
questions sort of inform each other and then bounce back and

(38:05):
forth. So your research as a, as a sort
of reply to Megan, your research, Lauren, distinguishes
between types of explanation. So, mechanistic causal
unification based. When a neuroscientist claims
that they've explained something, for example
consciousness, which form of explanation are they actually

(38:25):
offering? I think the short answer to this
question is that it's still a bit of an open, It's still a bit
of an open question, what we expect these explanations to

(38:45):
meet in terms of the criteria. I also think that few people in
this space suggest that they have a full explanation or
almost any explanation of consciousness.
So let me back up a little bit. Here in my work in philosophy of
science, we study scientific explanation.

(39:07):
What does it take for a scientist to have an explanation
and to give an explanation? Something that's very important
about this space is that is saying a little bit about what
explanation is. So we often think of explanation
as one of the most important things that scientists do.

(39:29):
It's a very difficult thing for them to do.
We think of explanations as giving deep understanding of the
world. So in this sense, explanation is
different from other types of projects that scientists engage
in that are very important projects, like giving
descriptions of the world. So I can describe the color of a
leaf on a tree, but I haven't explained why it has that color.

(39:53):
So that's a description. Scientists engage in
classification. They sort things into helpful
categories. That's also not an explanation
of something in the world. And in other cases they give
predictions. And giving a prediction is of
course very useful, but it's notyet giving an explanation.
We think of explanations as answering why questions.

(40:15):
So why is it the case that that leaf is green?
Why does this person have a disease as opposed to not?
Why does this? Plant have a certain height as
opposed to having another height.
And so and so a first thing to point out is that explanations

(40:37):
offer deep understanding of the world.
We want to know what criteria they need to meet to know that
we have good ones right. How do you know when you have a
right, a good or an appropriate explanation of any kind of
phenomenon of interest, right, Adisease outcome, social
inequalities, right? This doesn't just extend to
neuroscience. This is all scientific domains.

(40:59):
So the two parts that you see here for an explanation or two
parts that show up is you first need to ask an explanatory why
question, or you can couch your explanatory target in terms of a
why question, right? What explains consciousness is
going to be a sort of start or you can put in any kind of

(41:20):
target of interest. So you ask a why question.
The explanation is the answer. So why is it the case that this
patient has measles as opposed to not?
Well, part of the explanation isthere's some virus that they
encountered and then there's a bunch of other interactions in
the immune system that explain why they have that disease

(41:42):
outcome. SO2 parts of an explanation,
explanatory why question and then your answer to that
question. So in order to give an
explanation for something, you need to say what you want to
explain. And that's where that why
question shows up. And there's actually a lot of

(42:04):
features involved in providing awell defined explanatory target.
And so right now in consciousness research, there's
debate and investigation and discussion about what's the
explanatory target and then what's the answer?
What's the stuff that explains that target?

(42:24):
And as Megan dimensioned, there are many different explanatory
targets that are showing up in consciousness research.
And part of the challenge is being very clear about which one
a scientist is interested in. So saying you know what explains
consciousness, that's not a welldefined scientific question yet.

(42:45):
It's not yet a well defined explanatory why question for two
reasons. First, you need to define
consciousness and we don't have a consensus definition.
So then you need to be precise about which one you have in
mind. And then the second is you need
a contrast. You always have to say as
opposed to what? So if I'm interested in

(43:06):
explaining why someone has a loss of sensation in their hand,
I can't just say what explains why they have a loss of
sensation in their hand. I have to say as opposed to
what? As opposed to full sensation in
their hand, or as opposed to a loss of sensation in their leg,
right. If I don't specify the contrast,

(43:26):
you don't know what answer to give me.
And so part of what? Philosophers of science do here
is we're looking at what are thethings that need to be met to
have a well defined explanatory target.
And you see them in other scientific fields.
So we're looking at cases where we have scientists who've
successfully given explanations and we're looking at the

(43:48):
criteria. And then we're looking at
consciousness research and theseother spaces where you have
scientists working on answering really difficult questions that
we don't yet have answers to. But you first have to ask the
right kind of question before you can get an answer.
And so there's two main challenges.
What's the right question? And then in terms of what's the

(44:11):
right answer, here's where you start to see what do you need to
give an answer. Do you want causal information?
Do you want a causal explanation?
Do you want a functional explanation?
We sometimes think that computational explanations are.
There's something there that we need that's going to help answer

(44:33):
that question. Mechanism, of course, shows up
in philosophy of science. We have different categories of
explanations. Causal is a main one.
I would put mechanistic explanation that's just a causal
explanation. Mechanisms are just saying
you've identified. Well in most cases mechanism is

(44:53):
a causal explanation. In other cases there might be a
non causal mathematical explanation.
So I guess 3 categories I would pin down are causal explanation,
non causal, mathematical. There's a lot of debate about
what those look like. Functional explanations you
could think like evolutionary explanation.

(45:14):
That's not quite what we're interested in here.
And computational, there's a question what you know, are
computational explanations causal?
Do they, are they a subcategory of causal?
But for the most part, we're often interested in causal
explanations. So you're, you're looking for

(45:36):
the main factors that 'cause that target of interest.
And there's also debate here about in consciousness research.
Do you have the right factors there?
If you're interested in correlates, neural correlates,
there's often a bit of slippage in how that's used.

(45:58):
But if something is a mere correlation with your target,
then you don't yet have causality.
So this is where a philosopher is working with scientists to,
to to help determine what are your different explanatory
targets, because that's going tohelp you get the right answer to

(46:19):
that question. And what I would say is there
isn't one question here. There almost never is for
complex systems. There isn't A1 complete full
theory of everything explanation.
It's piece meal. And so you're asking different

(46:40):
why questions about a complex system.
And that's the sort of trick that scientists have to manage
this complexity. But part of what that shows are
those two pieces of an explanation.
Your explanatory why question the X we, the fancy word here is
explanondem. This is what you want to
explain. And then the explanons is what

(47:02):
answers that question. What gives you the explanation?
Usually some kind of causal information causes explain their
effects. And so there's a whole challenge
of once you have a well defined explanatory target going out in
the world and identifying the main causes that are relevant to
that target. I think let's try and bridge

(47:23):
these two together. So Megan, taking all of that
into account, these levels of explanation, your work in
metacognition or your research in neuroimaging modelling,
etcetera, how would you then address what Lawrence talking
about using your work as a as a guide for us?
Yeah, great. Great question.
That's kind of the whole enterprise, right?

(47:44):
I think there's a couple things that Lauren said that really
resonate with me. And this is the nature of being
very clear about the the questions that you're asking.
So in actually this is this is what we try to instill in our
students at Neuromash is the asking and answering the right
kind of question is, is the primary thing that you should be

(48:07):
looking at. The technique can come
afterwards. You have to pick the technique
later in order to answer the question, but you got to get the
question right first. And, and I just had a piece come
out recently about how to come up with good scientific
questions and what that really looks like.
And there's been a lot of work in computational neuroscience

(48:28):
and cognitive neuroscience in how to think about the
interaction between the questions that you're asking and
the goals that you have as a modeler or as a scientist in
general. And the plurality that Lauren
noted is absolutely right that there's, you know, depending on
who you ask, there's what, how and why questions.

(48:49):
That's classic Diane and Abbott 2005.
There's, you know, Mars levels of analysis, which are
computational and algorithmic and implementation.
You can ask questions about eachof those levels of, you know, at
each of those levels of inquiry,you can have questions that
target different levels of granularity.
So you have micro versus macro versus organismal versus like

(49:12):
societal. And so this plurality of
questions and plurality of approaches I think is really
critical because as Lauren said,there is no one question to rule
them all. There is no one answer or one
explanation to rule them all. There's no one ring to rule them
all. It's just not going to happen.
So I think that from our perspective, this is this is

(49:36):
actually something that I try toinstill in all of not just my,
you know, doctoral trainees, butthe undergraduates that I teach
and the folks that we reach out to at Neuromatch as well.
Is that this the recognition that the heterogeneity is, is a
feature, not a bug that I think is is really, really critical.
There was something else that that Lauren said to earlier

(49:58):
though about this, which is in coming up with your your type of
question, you have to have a little bit of an understanding
of the the way in which you might go about building that
explanons that explanation of the target and the level of
visibility that you might have into the system.

(50:18):
The level of access that you might have into the system.
Because you can come up with this amazing question that is
actually unanswerable with the tools that we have available to
us. And you can also come up with a
question that might be answerable in.

(50:39):
So it's, it's answerable in principle, but not in practice.
That's one kind. But then there are others that
might like not be answerable in principle, at least not yet.
Because we don't have, it's not that we don't have the right
tool, the right neuroimaging technique or the right model or
something. It's that we don't know how to
ask that question by the right way yet.

(51:01):
And and you said something, Lauren, that that really struck
me, this kind of limited visibility into the world idea
that we always have this, these barriers that shape the types of
explanations we seek, the types of questions that we can shape,
types of answers that we can go out and look for and the kinds

(51:22):
of data that we can acquire. But I do think that there are
there are other kinds of limitations that are not these
kind of practical like, you know, the parts of the world are
unobservable. I think that there are other
limitations that that we should acknowledge in building these
questions as well. So, you know, imagine a case

(51:44):
where you have I've built some sort of magical machine in the
future, some magical brain imaging device that has perfect
visibility into everything that the that every neuron is do on
every every synapse. I have the morphology, the shape
of every neuron. I have the structure of the
dendritic tree. I have all the chemical
interactions. I have literally everything
about the brain. I still could shape all sorts of

(52:07):
different kinds of questions. I can't just take that model of
the brain and shove it into someartificial intelligence and be
like, poof, great. I understand.
I have an explanation. It's still like even if we had
perfect visibility, the questions are still going to be
the primary driver and the lack of visibility into certain kinds
of systems is still going to be the limitation.

(52:30):
And that lack of visibility is now not coming from the tools
that we have available. It's like the lack of conceptual
clarity, the lack of, of being very precise about the target of
explanation. So yeah, I think it's, it's all
got to come down to the questions that you that you ask,
the shape of those and how thosequestions interact with the

(52:51):
goals that you have as a, as a scientist.
So do you want to build an explanation that has clinical
impact? Do you want to build an
explanation that is beautiful and intuitive and simple and
like easy to explain to others? Do you, you know, So what is?
What is the kind of explanation that you want to build too, not

(53:12):
just the kind of question that you want to ask?
Lauren, anything you want to addto that?
Absolutely. We sometimes discuss this in
philosophy in terms of having a God's eye view of the world or
the Laplacian demon sort of knowledge about all of the stuff

(53:32):
that's out there. And it can be very tempting as a
philosopher, sometimes a scientist too, to think there's
all of this stuff out there. If I just knew more about all of
the stuff, I would get the perfect complete explanation.
And the challenge for that picture is we currently don't

(53:58):
have that information yet. We're successful at navigating
the world. So part of what we're looking at
here as philosophers is how scientists reason and how
they're successful. But also in everyday life, we
give explanations, we engage in causal reasoning, and we do that
pretty well. Are we perfect?

(54:18):
No, but we do it pretty well. And we just don't have that kind
of full, complete information about the world.
So the question is, how do we dothat?
It looks like we don't need thatsort of information.
And if you wanna provide an an account of how a human or a

(54:38):
scientist ever studies the world, you can never include
that kind of picture because it's just that's, that's a
fantasy story, right? Where all scientists are humans
and they're engaged with the world.
And if you want to talk about having all those details, you're

(55:01):
talking about a future science that doesn't exist.
And I'm not sure my future science is going to match up to.
So what we want to talk about iscurrent science and and past and
what has worked. And so one of the fun parts of
doing this kind of work for me, I think Megan has this too, is

(55:21):
you're looking at what has worked in these different
scientific contexts, and you have a sort of domain, general
view of real scientific practiceand how scientists manage those
limitations to get information about the world.
And so, yeah, it can be very tempting.

(55:42):
There's interesting temptations and interesting pictures we have
in everyday life, philosophers and scientists about getting
full detail is very attractive to us.
Also reduction, which I think will come up.
If we just could get more information about stuff at lower
levels, we could get better explanations or the view that

(56:04):
that's where we should look to get the right kind of
explanatory account. So yes, very much, very much
compatible with this kind of realistic picture of scientific
practice and scientific work, asopposed to this idealized view

(56:31):
where we ever had access to all of the details.
There's a, let me just follow upon that for two seconds.
There's a, a favorite paper thatI like to send students in my
neuroanalytics class to, to kindof highlight this.
If only we had perfect access toeverything, then we would
definitely understand. And it's this paper that Conrad

(56:53):
recording and, and some colleagues wrote, I don't know,
10 to something years ago. It's it's called could a
neuroscient, Could a neuroscientist understand a
microprocessor? And they have this toy example
where they say, OK, I've got this microprocessor and it runs
like Donkey Kong and Sonic the Hedgehog and Mario or something
like that. And they go about dissecting

(57:15):
this microprocessor using all ofthe fancy available tools, all
of the models that we would use in neurosciences.
So they measure all the resistors and they measure all
the synapses between all the, you know, nodes in the
microprocessor. And it's a simulation
microprocessor. So it literally they have
perfect access, right? There's like no noise in the
system. And they do like inactivation

(57:39):
experiments and they measure like the network connectivity
and the like the state transitions and they do all the
tricks. And they still don't end up with
an explanation for why poking the thing in this way makes it
unable to run Mario or why poking the thing in this way
versus that way has no effect onwhether Mario can hop over the
thing or not. So it's really like a, a kind of

(58:03):
cheeky demonstration that it really matters what you think
you're measuring. Like do you have perfect access
to the system? Do you have perfect access to
all of the things that are actually the, the parts of the
system that you need to have access to?
And there they have access to all the physical system, but
they're not like reading the software.

(58:23):
And they'd have to come up with,you know, the software in order
to, to kind of build more of an explanation for how the software
and the hardware interact. So I don't know, it's, it's a
fun one. If you haven't seen that paper,
those of you who are out there listening, I I suggest you go
and have a look because it is. It's fun and it's cheeky and
it's also quite profound. I love that concept, Megan, if

(58:46):
you've got the link, please share it with me so I can put it
in the. Yeah, I will.
In your work, from all the work you've done, what are some of
the ground breaking things you guys have figured out during
this time that allow us to ask deeper philosophical questions?
I think that some of so some of the things that that I am
interested in here is as we said, the nature of

(59:08):
metacognition and subjective experience and how those two
interacts and metacognition being the process and subjective
experience potentially being like the output of that process.
I like this approach because it combines, it combines
neuroscience and behavior and psychophysics and psychometrics

(59:31):
and also computational models ina way that tries to build like
this piece meal. Small, tiny explanation of why
it is that if I change this particular aspect of the world,
it changes your subjective experience in this particular
way, and it changes your subjective experience in a way

(59:51):
that's different from changing your ability to just interact
with the world in a meaningful, goal directed, kind of
evolutionarily optimized way. And So what I mean by that is
that for us, when we process theworld, you can think of a lot of

(01:00:12):
what's happening in that processing is going on under the
hood, so to speak. There's a lot of heavy lifting
that the brain does that is not available to us.
Consciously, subjectively, anything like that.
And I'm not even just talking reflexes, I'm talking all the
processing that gives rise to the fact that you see the world

(01:00:32):
in 3D. Can you like kind of consciously
intervene on that and say like, no, it's, I know that it's
actually a 2D image on my retina.
No, like you just see the world in 3D.
It just happens magically somehow.
And so there's a lot of this complex processing that goes on
under the hood. I'm a vision scientist.

(01:00:53):
So vision science is my my typical workhorse here.
But you can play this game for alot of other things too.
For any way that you interact with the world, you see a
complex, noisy, stochastic dynamic environment and you are
standing like you're standing ona sidewalk and you're deciding
whether to cross the road or not.

(01:01:15):
And you hear things and you see things and you have to decide is
it safe? And, and that decision is going
to impact your ability to survive, right.
If you if you get it wrong, you get hit by a car.
But so much of that could be said to be done potentially
without conscious awareness. And so a lot of the work that

(01:01:36):
we're doing is taking all of these tools in our tool kit and
pointing them at trying to dissociate the conscious
experience or subjective experience part from all the
other stuff that like a Tesla could do or a Roomba could maybe
do a very smart Roomba. And there's nothing that it's
like to be a Roomba, presumably.So I think the favorite thing

(01:01:59):
that I'm doing right now that might hopefully have some impact
is the conceptual and methodological distinction
between understanding the behaviors, computations and
neural correlates that give riseto adaptive goal directed
interaction with the environment.
Not dying, not stepping in frontof the car and separating that

(01:02:22):
from the computations and neuralcircuits and neural
representations that are uniquely driving or uniquely
correlated with the subjective experience part.
And I think the reason, so I'm not the only one who's working
on this, by the way, there's like quite a lot of us who see
that distinction as being reallyimportant, but there's also

(01:02:44):
quite a lot of people who don't.And I would say that there are
some folks who are in the artificial intelligence space,
for example, who conflate intelligent behavior with
probability of being, of someonebeing in there, of being
conscious, or even worse, conflate intelligent looking
behavior with, you know, subjective experience or

(01:03:05):
consciousness with threat. You know, we say, oh, no,
Skynet's going to wake up and it's going to kill all of us.
And I think that some of the work that we're doing,
admittedly with very simple stimuli and very simple models
will help drive at that really important distinction that just
because you've got a system that's seeming intelligent, it's

(01:03:26):
seeming like it can solve problems, that doesn't mean that
anyone's in there, that it's that there's anything that it's
like to be that system. It also doesn't mean that it's a
threat. Things can be threatening
without having subjective experience and without being
intelligent. And so all of those things are
independent. So I think that was maybe the
thing that I would say is potentially the impact of the

(01:03:47):
work that we're doing. Lauren, if you had to envision a
philosophically informed neuroscience infused with your
work regarding mechanism, explanation, etcetera, what
would that look like in practice?
An experiment, design, theory, formation, or even peer review.
Part of what it would look like is a kind of neuroscience that

(01:04:13):
we partly already see. But part of what it would
highlight is clarity about the types of causes and causal
systems that neuroscientists study and that that researchers
are studying in this space. One way to see this is we often
find the term mechanism being used to refer loosely to any

(01:04:37):
kind of causal system. Part of what my work has done is
specified that there are different types of causes out in
the world that scientists study.They have very different types
of features, and those matter for how we study the systems.
They matter for the behaviors they produce.

(01:04:59):
And you start to see some of these distinctions show up
already when scientists talk about causes that are
probabilistic versus deterministic, causes that are
more or less strong, more or less stable, or when scientists
use terms like referring to a causal system as a pathway or a
circuit or a cascade. There's a reference here and

(01:05:20):
then analogy to different types of causal systems.
So part of what my work is compatible with and can
encourage is far more clarity about the types of causes that
are out there that we study. We partly need words to refer to
the different types we have, different features that they

(01:05:44):
have, and this is partly going to inform the standards that we
have, basically being clear about the standards that we have
for the kind of causal information that neuroscientists
need to provide right now. The standard is usually phrased
as a mechanism. Scientists needs to provide

(01:06:06):
mechanistic information about a system.
You see this in grant calls, yousee it in journal publication
guidelines. In order to get funded, in order
to get published, A researcher needs to identify a mechanism or
provide mechanistic insights. But then the editors very
quickly follow up by saying thatthey can't tell you what a
mechanism is. And then it's often the case

(01:06:30):
that two or more researchers reviewing the same paper
completely disagree about whether the same paper provides
mechanistic insights or not. So right now you have a causal
standard for the field that is this word mechanism, and we have
different people defining that term in different ways, and
there's no consensus on what exactly it refers to.

(01:06:53):
Is it lower level cellular details?
Is it higher level network information?
You have researchers pointing toboth as real mechanisms.
Partly we need to put mechanism aside, and when we're interested
in causal explanation, we just need to talk about these as
causal systems, causal relationships, the way that

(01:07:13):
causes are organized and arranged.
How do you know you have the right kind of causal information
that's explanatory, relevant to your target?
So part of it is clarity on the standards for the field and
getting more clarity on what we mean by mechanism if that's the

(01:07:34):
current specification of the field standard.
Yeah, I'm, I'm looking forward to having both of you separately
on the channel so we can explorespecific aspects of both your
work. But I think at this point to get
back to the main topic here, if you both had to look at science
and philosophy and moving into the future, bringing them both
together, what would this new, what would a new picture of mind

(01:07:55):
would emerge from this? Would it be something different,
do you think, do you think it would change anything?
And what advantage does this have for new fields specifically
that will arise? I'll take that.
I I love that you said advantages for new fields
because I think that one of the challenges that we have in
neuroscience, again, I'm a neuroscientist, so that's where

(01:08:16):
I'm coming from, is that this isstill a new field.
It's really young, especially like the neuroscience of, you
know, consciousness or somethinglike that.
Like psychology, yeah. Has been, oh, it's been around
for, you know, 150 years in its present state and, and, you
know, quantitative computationalpsychology.
Yeah. But like, that's not very long.

(01:08:37):
That's really not very long. Modern science is not very old
in general, but when it comes tomodern science, philosophy or
psychology and neuroscience specifically, like our first
neural signatures are only about100 and some odd years old, you
know, when EEG was first invented.

(01:08:59):
And so this is a really young field.
And so I think that new fields and emerging fields like this is
where the value is. This is where we need help at
getting conceptual clarity because a lot, in a lot of cases
for new and emerging fields, the, the major tool that we have

(01:09:21):
to say, well, where do we even begin is something like
intuition. I came up with an idea and like,
let's just run with it and see what happens.
And as we, I think all have probably discovered through one
point or another in our lives, what you think is happening and
what is actually happening is never like your first guess is

(01:09:43):
never the right one. And so recognizing the value of
philosophy of science in young and emerging fields and fields
that have yet to emerge, I thinkis really powerful.
And as Lauren said, some, you know, this idea that especially
at the beginning in a young science, seeing the

(01:10:04):
commonalities in the structure across this new emerging field
and maybe a more established discipline that has kind of
already figured out some stuff. So we've got a lot of really
precise terminology in how we understand mechanism, whatever
that prestige wars is, by the way, like, yeah, we all, we all
know that we want to go for a mechanistic or causal

(01:10:25):
explanation and get what even isthat?
But there are, there are even among the young modern science
fields, there are some that are very, very young, you know,
they're children. And then there are some that are
a little bit more middle-aged. And so on the surface they're
all gonna have these extremely different features, these
extremely different kind of surface level properties of or

(01:10:47):
observables. But causal and mechanistic
explanations are a unifying principle.
And so recognizing that the shape of the problems that we're
trying to solve might actually be quite similar in this new and
emerging field to a more established field.
But when you're a scientist and you are reading the science

(01:11:08):
journals and you're kind of likein your little your little box,
you don't have time to pop your head out and go read some
astrophysics journal. It's just not going to happen.
Or some material Science Journalor something.
And so having though this targetof building explanatory models,

(01:11:29):
of getting conceptual clarity, of understanding the types of
causal and mechanistic explanations that we can go for
that can provide a bridge. And you say, OK, well, we're
talking about completely different systems, completely
different targets of explanation, but the kinds of
explanations that we're trying to build might actually be quite
similar. And I have experiences with this

(01:11:51):
where I, I wrote this, this paper with my one of my graduate
students and another professor and his graduate student.
And he's a microbiologist. He studies the microbiome of
pregnant women and how the microbiome of pregnant women
impacts birth outcomes and maternal outcomes.

(01:12:12):
I don't do that at all. I have no idea even what half of
the vocabulary is that he says. And yet through talking with
him, we discovered that the shape of the problem that we are
trying to solve was actually very similar.
And so we wrote a paper about that and how like these kind of
modern, you know, machine learning tools might be able to
help us with that. And that's what we need in these

(01:12:33):
young emerging fields is to see,well, someone else solved a
problem that had a similar shape.
And, and if we can get that right, it will propel the new
fields that we have right now forward and emerging fields that
come in the future. I think that that will be a
major step forward in building better science, building more

(01:12:57):
coherent science that is self perpetuating. 100% agree with
Megan here and. We partly already see nice

(01:13:19):
features and aspects of current work in neuroscience that show
this interdisciplinary aspect. We've got lots of neuroscientist
philosophers who are aware of both fields.
You see this in Megan's work. You see it in the work of Anil
Seth, Danny Bassett, other cognitive scientists like Nadia

(01:13:43):
Cherniak, Karen Walker. There really are lots of
scientists and academics who areengaged in this
interdisciplinary approach. We 100% need it too.
For many of these challenging questions, we have an all hands
on deck type situation. We.

(01:14:04):
Need many different people from many different perspectives to
help out with these questions. The challenges are it can be
pretty uncomfortable to do this kind of work because you're
never the main expert. When I'm talking to scientists,
I mean, they are always so much more of a deep expert in their

(01:14:24):
area of work than I than I couldever be.
And I'm, it's partly, it's the way it has to be.
I'm talking to social scientists, I'm talking to
cognitive scientists, I'm talking to neuroscientists.
It's a bit of a a stretch sometimes, but for me, it's very
important for me to put myself in their perspective.

(01:14:45):
What are they interested in? Can I bring philosophy of
science? That's useful and that's helpful
in order to figure out if it's aperson on the team or a hand on
your deck. They do need to be useful and
they do need to be helpful. And it's not easy for
philosophers to fill those shoessometimes with respect to
scientific work because it can be uncomfortable.

(01:15:07):
You have to learn a lot of science, and you're still never
going to know, You're never going to have the same kind of
picture. But these discussions can show
you types of philosophy that will be really helpful for
scientists to have. You don't.
We also don't want to reinvent the wheel.
And we've seen this in cases where you have researchers that

(01:15:28):
aren't interacting with each other, right?
Someone spends a lot of their career developing an approach
that someone built basically 3 decades earlier.
So you don't want to reinvent the wheel.
You do want some pushback. I need it from Megan.
I try to give it to her too. I think the, the standard thing

(01:15:49):
you'll probably hear scientists say about philosophers is
they're sort of the one asking that question of, well, what do
you mean by mechanism? You know, And then you give an
answer and then we think of counter examples and it's like,
well, if that's what you mean, then there's a problem that
shows up. Or if that's what you mean by
explanation, you're including all these cases that you don't

(01:16:09):
want to include. So we are trained to think
abstractly and we are trained tokind of want that precision.
And so that is something that wecan attribute.
And there are scientists who lean into this interdisciplinary
approach by bringing philosophers on board, but of

(01:16:30):
course scientists from all sortsof other domains.
There's interesting examples where philosophy can suggest
ideas that scientists go on to study in their empirical work
that they might not have thoughtof originally or as quickly,
because it's a little easier to see them in some frameworks.
We see this in Cogsai with studies of different types of

(01:16:53):
causal relationships, things like stability and strength, for
example. But yeah, I think part of what
we, part of what I envision for a kind of future here, and the
advantages come from this interdisciplinary work.

(01:17:15):
I also think part of what would be helpful is for scientists to
have a little bit more time and space to to do theorizing right.
So, yeah, so they're the, you really start to appreciate the
challenges of the scientific work when you look at the sense

(01:17:38):
in which they're trying to tackle new problem spaces.
You, you know, often the funding, the funding incentives
are for this tried and true method.
And if you kind of already know it works, you can do a lot of
that. If you're expected to publish a
lot, that doesn't always incentivize taking the time to
think about all these different routes you could take and being

(01:18:02):
able to discuss which one, whichones you should follow.
So I think having a little bit more space for scientists to
have the time, like Megan mentioned in philosophy, we have
a little less of the pressures that they have.
But part of it is in having the kind of time and and incentives

(01:18:26):
to take advantage of interdisciplinary connections in
work. And that's not always easy for
scientists to do, given the constraints that they have.
You know, this, this question oftime and, and you know,
publisher perish mentality, there's a a lot of people who
are probably out there listeningright now.

(01:18:46):
We say like, why do we even carethat you're publishing papers?
Who, who reads those papers? And to a certain extent, you're
absolutely right. Like we, you know, the metric of
our success and the thing that allows us as academics to
proceed through the ranks and get promoted and, you know, do
our jobs well and so on is to get grants and to publish

(01:19:08):
papers. And it feels very insular.
It feels very much like you're kind of in a little echo
chamber. And I think that that's that's a
correct way of looking at this, that this is an old school way
of thinking about how we should go about the enterprise of doing

(01:19:28):
science. And it shouldn't be contrasted
with the way that industry professionals are doing science,
which is to produce products andto do to engage in activities
that have the potential for clinical or societal benefit,
basic science and foundational science.
And it has to be there in order for those those kind of more

(01:19:49):
applied approaches to have legs to have a foundation to stand
on. But I do think that the model of
do a thing and then write a paper and then get a grant to
continue doing the thing and then writing another paper is,
is doomed. Ultimately, to put it bluntly,
that it is striking to me that in 2025 we are still doing

(01:20:10):
science the way we did in the 1800s, that we've got scientists
who are doing. Science and then writing a
little paper that other scientists will read and that
maybe makes a big splash and hassome sort of impact on some, you
know, applied science later. The basic science has to be done
right. The reason that we have

(01:20:31):
technologies like GPS, for example, is because someone at
some point was like, huh, I wonder if we can do that.
And so they figured out how to do the technological basis, the
foundation that became GPS. And it wasn't because they went
about trying to invent GPS from the beginning and an applied
technology. It's because they did the basic

(01:20:51):
science work first. But this practice of just
writing a little paper and then like, you know, packaging it and
tying it up with a nice little bow and sending it to a journal
and paying thousands of dollars to publish it and then having it
be locked behind a paintball. This is a rant, but it's also, I
think, a recognition that in order to realize the future that

(01:21:12):
Lauren and I have been really talking about, we need to change
this. Because it's not just a time
constraint, it's a societal and like expectation constraint on
the way that we as basic scientists and academics are
engaging this enterprise. It's hamstringing us.

(01:21:33):
It's preventing us from engagingin this future that Lauren and
I, I think have laid out and that we're both very excited
about and I think that others are excited about too.
That we need to find a way to bemore interconnected, to
capitalize on the fact that we do have a global scientific
community that doesn't need to wait for a paper to get

(01:21:57):
published in order to learn about some new scientific
finding. There's got to be a better way.
And it isn't social media. We need something in between and
so, and it's not, you know, conferences only.
I think that there's got to be abetter way to do this.
And I don't know exactly what itlooks like, but there's a call
to action for the folks listening in here that if you

(01:22:21):
think that this future sounds cool and exciting and powerful,
think about how to make it a reality.
And this is something that I think about a lot.
And then some of the activities I'm engaged in are trying to do,
but, but I think we need more people.
So there, that's my my plea to get involved in making this
future a reality. Let's get back to the this idea

(01:22:42):
of consciousness, computation, and causation.
Megan, you've described the brain as a probabilistic machine
navigating uncertainty. Would you describe consciousness
as a byproduct of computation oran adaptive feature of it?
Yeah, I, I don't know if I want to weigh in on that and, and
pick a hill to die on. This is a big question.

(01:23:04):
Is consciousness in every phenomenon, is it just kind of
there as a by product or does itserve some kind of meaningful
function in our ability to, you know, from an evolutionarily
evolutionary perspective, stay, stay alive, engage, procreate,
that kind of thing. So I think an important

(01:23:25):
component of this question is todifferentiate among a, a
potential function of consciousness versus a potential
function for consciousness versus functions associated with
consciousness. So there's you're asking, is
consciousness an epiphenomenon? That would be there is no
function at all. It's just kind of it happens

(01:23:46):
because you know, that just is with the way the universe is set
up. I, I personally think that it's
probably the case that it's not totally an epiphenomenon that it
is emerges as a component in a giant functional system that
probably was evolutionarily optimized in some way.
So I think that there is a function of consciousness it, it

(01:24:06):
has a purpose, there is something that it does that is
adaptive and facilitatory for the Organism that possesses it.
It allows you to bring information into a global
workspace so that you can manipulate it in a kind of a
domain general way, or it allowsyou to differentiate between

(01:24:28):
something that is real out therein the world and something that
you just kind of hallucinated ormade-up in your head or just
noise. So this is sometimes called
reality monitoring. And so the, the presence of
phenomenal experience is the result of some reality
monitoring tagging system that says these are the components of
the world that are probably real.
And these are the components of your internal representation

(01:24:50):
that are probably just noise or you just made it up.
And then then there's a function, you know, for
consciousness that is the internal machinations that gave
rise to the the conscious experience that's very different
than the reason that we have it.And then there would be all the
other things that go along with consciousness in us anyway, like

(01:25:13):
language and executive functioning and reasoning and
problem solving and, you know, stuff like that, that seem to be
present when you are conscious and seem to be absent when
you're not, or seem to be present when you are conscious
of a particular piece of information and absent when
you're not. So there was a big debate for a
while about can you do math unconsciously?
Can you do arithmetic or addition unconsciously, that

(01:25:35):
kind of thing. So the truth is we don't know if
consciousness has a function. I think that something like the
ability to decide when to updateyour model that you've built of
the world based on new incoming information, that seems like a
useful thing for a reality monitoring or similar mechanism

(01:25:56):
to do. I don't know that phenomenal
experience per SE is the component that has the
functional, that is the functional or like kind of
causally efficacious knob in thesystem.
But all indications seem to point to in in my mind that

(01:26:18):
without phenomenal consciousnessyou cannot do some things that
it does have some sort of facilitatory function for us.
So I think that there is a function.
It probably has to do with learning adaptive behavior,
updating of world models. Pretty hand WAVY answer, but I

(01:26:39):
don't think it's an EPI phenomenon.
I think that there's probably a reason that it's there.
Lauren, when it comes to the philosophy side of this and the
question of what is consciousness, are we even
asking the right question? I think that there are many
questions that are being asked right now in this space.

(01:27:02):
It's, it's a mistake, I would say, to think that there is one
question. And it's helpful to consider the
sense in which we're trying to figure out, even if there are
many questions that we're asking, that any given question
involves, there's a lot of boxesthat need to be checked to make

(01:27:24):
sure that it's well defined. And so there are as, as Megan
has suggested and as we see froma cursory understanding of
research in this space, there are really different types of
topics of interest that consciousness researchers are
focused on. One helpful thing we can do is

(01:27:47):
to separate out those questions.It would be unhelpful to think
that there's one. I'm also skeptical about the
need for some unifying theory that they all need to strictly
fall under, although that might take and require a longer set of

(01:28:11):
discussions. I think there is some kind of
unification that's helpful, but it's somewhat loose.
What we do want are very principled, clear questions.
And so we don't have this anything goes, you know, ask
whatever question you want. There's all these different
facets. No, the questions that we ask in

(01:28:32):
this space need to be so precisethat one of the main challenges
is asking the right question, right?
That's something that's been showing up repeatedly in this
discussion. It reminds me of, there's this
great quote, I think it's from the band U2, which is we thought

(01:28:52):
we knew the answers, it was the questions we had wrong.
And so a big challenge in scientific space is asking the
right questions. And we often think of that as
the starting point for giving anexplanation.
I can't give you an explanation for something until you first
tell me exactly what it is you want explained.

(01:29:14):
And we sometimes start on that path and we get stuck at the
1st, that first step, specifyingthe target.
And that's where a lot of discussion is in this space.
It would be silly to think you could give the explanation if
the target isn't sufficiently precise yet.
There are different targets of interest, that's just fine.

(01:29:38):
I can't think of many scientificspaces where that's not the
standard for any kind of system.There's so many different
questions you could ask. There's some that we might want
to put outside the space of an interest of a consciousness
researcher. So that's up for debate too.
What's the, what are the bounds on the space of explanatory why
questions here for consciousnessresearch?

(01:30:00):
We're interested in consciousness.
What are we, what are we interested in explaining?
So I think it's helpful to thinkthat an important part of
scientific work is asking the right questions.
And I don't think that in this space there's a lot of fixed

(01:30:28):
consensus on exactly what those are.
But that's the way science works.
And it's helpful to think that that's the first step that you
need to accomplish before you can get the proper answer.
So if you want to skip that stepand start looking for the
answer, you're going to be wading through a mess of stuff

(01:30:51):
and you just won't have the right guidelines because you
don't yet know what you're looking for.
And sometimes in science, we start with a rough question and
we go and we look for the causesand based on what we find, we go
back and we refine the question.You see this happen in medicine,
psychiatric medicine, right? We start with the disease
category. We think we've got the right and
then we go and we look for what the causes are.

(01:31:13):
We might re describe the target on the basis of what we find.
That's a kind of brick. It's a very smart strategy that
scientists use to get order in the world.
So I don't think we're there yet, but it's and I don't think
there's one question in that space, but a lot of the research
is focused there as I think it should be.

(01:31:36):
Megan, when it comes to consciousness, it's almost
impossible to nowadays have a conversation about it without
mentioning AI. So I feel like we have to touch
on this. So can AI systems or large
language models ever genuinely experience uncertainty, or will
there always be simulations without subjectivity?

(01:31:57):
You really want a definitive answer to this, don't you?
I so I there's two big, there's two big camps in the
consciousness science field about this and you've
articulated them very nicely. 1 is that artificial systems have
the potential. I think most people would agree
that they don't now have some sort of consciousness, but that

(01:32:19):
in the future they have the potential to manifest subjective
experience or phenomenal consciousness or whatever
terminology you want to use for someone being in there, the
lights being on, etcetera. And then there's the other camp,
which is kind of the more biological naturalism camp,
which says like, no, there's really something very special
about biology and silicon based systems or something.

(01:32:43):
Something that is not biologicalis never going to be able to
instantiate this type of this type of thing.
And you have really smart peopleon both sides arguing both
camps. So, you know, Anil Seth has just
written a piece in Behavioral Brain Sciences that's one of
those kind of target article. And then there's a bunch of
commentaries that come out associated with it that will say

(01:33:06):
things like, so Anil's piece says it argues for the point
that, you know, there is something special about, as he
puts it, being a beast machine, that biology does have
components that allow it to maybe manifest the types of
computations that are necessary in order to instantiate
consciousness. But he actually issues the idea

(01:33:27):
of computational functionalism in general and says it's not a
function that there really is something special about, you
know, synapses and biology and the squishy piece of wetware.
And the philosopher Ned Block has, you know, written a
commentary that kind of agrees with him that says there's
something that, you know, might be, although I don't want to
mischaracterize Ned. But then there's other

(01:33:47):
philosophers who and scientists who have argued against this.
And I tend to be more in the more functionalism camp.
So Matthias, Michelle also argues that, yeah, we can say
that there is something special about biology, but the special
thing about biology might be that it has the particular, it's
the, it's the only kind of substrate that can instantiate
that function. But the function is the key

(01:34:09):
component. The function or the computation
is the key component that gives rise to consciousness.
And so in the future, it is possible that maybe we figure
out what it is that might have been special about biology and
we actually build an artificial system that has all those
special components and now it can instantiate consciousness as

(01:34:30):
well. So that's a very tight view.
There's also a more general viewthat says, oh, well, maybe
neuromorphic systems might be able to instantiate
consciousness. Neuromorphic really just is a
fancy word for brain inspired and it can mean it either
instantiates the algorithms thatwe are discovering in the brain

(01:34:54):
or more likely neuromorphic refers to something hardware
based that there's this particular kind of spiking
neural network that is in that is manifested or instantiated on
a particular kind of hardware. Where we did some material
science to come up with the resistors and stuff that would
actually like look a little bit more like brain as opposed to

(01:35:16):
traditional. You know, when you think vacuum
tubes style 1960s computers, memory is over here and
computation is over here. And so then you move information
between memory and computation and then you put it back in
memory. And so there's, when we talk
about artificial intelligence, anything that is not biology is

(01:35:37):
in this big pile, but we have tothink about differentiating it a
little bit more. And then the very abstract
version of this is it doesn't matter what the substrate is.
It could be a traditional, it could be a neuromorphic system.
It could be a von Neumann architecture, which is like
this. You know, memory is over here
and compute is over here. It could be your laptop.

(01:35:59):
It could be some technology thatwe haven't come up with yet, all
of those. It could be a large language
model that runs on a server farm.
It could be kind of anything. And it's the computations that
matter. It it doesn't matter what the
hardware is at all. It's just the computations and
the type of like representationsthat the system can have.
And so from that perspective, maybe large language models are

(01:36:22):
like this close to waking up. I tend to be more on the
computational functionalist side.
So that was a long winded way ofsaying I think it's the
computations that matter. I don't think that there is
anything particularly magical orspecial about biology, except
perhaps that it can instantiate certain kinds of computations

(01:36:44):
that we don't yet know how to door that might end up being
impossible to do in certain kinds of non biological systems.
So from that perspective, I would say yeah, probably in the
future artificial systems could wake up.
Is it around the corner? Probably not.

(01:37:05):
Don't think that GPT 5 is on thecusp of having subjective
experiences and maybe this is not the place to go into this
necessarily. But let's say that you disagree
with me and you say no, it does.How would you test for that?

(01:37:26):
How would you know? That's like a whole other whole
other conversation that maybe wecan get into it at another time.
But this idea that we have ways of evaluating whether someone is
in there or not for us, for, youknow, neurotypical awake

(01:37:48):
behaving human versus a neurotypical asleep human who is
not behaving or a human who is in a coma or that kind of thing.
Like maybe those tests work pretty well for a clinical and
bedside. But as soon as you get outside
of the population on which they've been validated, like
what do you do? You can't apply them to the

(01:38:09):
artificial systems. You fall back on tests of
intelligence, which is, as we'vediscussed, not the same thing.
So I think it's very possible that artificial systems will be
able to have subjective experiences in the future.
It is not a hill that I'm going to die on.
And the the way of answering. Has that happened yet or at some

(01:38:32):
point in the future, When does it happen is really, really
hard. It's really hard to answer that.
Lauren, do you do you have anything to add to that?
I do, I think part of what can be helpful in terms of looking
at progress explanations and work in this space is that this

(01:38:56):
is an explanatory target that isso much different from many
others that we're interested in in science.
And that's part of the challenge.
And that can partly explain why we we don't yet have an answer,
but also have specifying the standards is difficult.

(01:39:17):
So this is a type of thing we want to explain that is
different from other types of things we want to explain in
science. And then we have explained.
And so we partly need to figure out what those differences are.
And then the second is this interesting feature where for
the types of targets that we do see across different life

(01:39:38):
sciences that we are interested in explaining, we give
explanations for. There is often a set of
challenges that show up with respect to how much detail you
need to cite to give an explanation.
And one thing I find is that there's sometimes a kind of
confusion between stuff in the system or stuff that's necessary

(01:40:03):
and stuff that's explanatory. And this partly relates to
reduction and just figuring out what details.
That a scientist needs to cite and should cite in their
explanations. And this is where we find
various interesting, confusing things that show up when we're
interested in giving explanation.

(01:40:23):
So how, how low do we need to goin giving an explanation?
And how far back in the causal history of something do we need
to go? Is another question that shows
up. Explanations are selective, they
are choosy and they pick some ofthose details, not all of them.

(01:40:45):
One confusion, confusion that can show up is you can admit a
kind of physicalist position fora biological system, a neural
system, and agree that there's physical stuff at lower scales,
but that doesn't mean it's explanatory.
And when someone is referring tofactors at a higher scale is

(01:41:07):
explanatory, they're not denyingthat physicalist picture.
And sometimes those get confused.
And so we need to separate explanatory relevance from
physicalism because they're verydifferent.
I mean, if we needed to cite allof that physical stuff, we would
almost never be able to give an explanation, but we also don't
need to. And so the the way I think about

(01:41:28):
many of our causal explanations here is that a causal
explanation isn't a game of how low can you go, but a game of
what gives you control. And depending on your
explanatory target of interest, the factors that give you
control might be at a higher scale.
And so this is partly where we need to kind of make these

(01:41:51):
helpful distinctions to solve these kinds of things that can
get tricky, where a scientist might think that if you include
more and more lower level detail, you're always giving a
better explanation. Or that network neuroscientists
deny physicalism when that's notwhat they're doing.

(01:42:11):
If they're making an explanatoryclaim.
Or there's a puzzle that that philosophers sometimes run into
where they think The Big Bang, since it's in the causal history
of of everything, that it's something you should cite in
your explanation. So do you need to cite The Big
Bang in explaining why we're allhere today or why a patient has

(01:42:32):
a disease? That sounds so silly to us.
A philosopher's job is partly tosay why that's silly and why
that's not explanatory. But we get stuck on those cases.
So we get stuck on reductionism and we get stuck on the entire
causal history and sometimes distinctions like physicalism
and explanatory relevance and necessity too.

(01:42:55):
Something can be necessary for an outcome doesn't mean it
explains it, right? The Big Bang is necessary for my
having asthma, but it doesn't explain why I have it.
If I went into the position and they started and I asked, you
know, why do I have asthma and The Big Bang is that what's that
doesn't sound right. So part of what we see in these

(01:43:19):
spaces are really important questions about how to how
scientists are making are makingprogress, the types of
explanatory targets they have and important distinctions we
need to make to get over these puzzles that that show up that
can kind of lead us astray and that don't capture the rationale

(01:43:43):
that does underlie our explanation.
I mean, you partly see it with control, right?
The Big Bang isn't something, ifyou were to hypothetically
manipulate it, it doesn't control whether a patient has
measles or not. So it doesn't explain that
outcome. So, so scientists.
Sorry, sure. Are you sure?

(01:44:04):
It doesn't explain why they havemeasles?
I'm just being cheeky. I mean in terms of what's
currently on offer in. Fair enough.
But but part of what philosophy of science, when it's at its
best, it can help with a bit of this science communication
element, which is what is the justification for why physicians

(01:44:27):
say that there's a virus that causes measles and not
fundamental physics or not The Big Bang or, Yeah, why
neuroscientists are working so hard to explain something like
consciousness and why this is actually more difficult than
explaining just any kind of trait in biology.

(01:44:51):
So yeah, just a bit of a follow up there and support of various
things, Megan said. Megan, I asked Lauren about what
a philosophically informed neuroscience would look like, So
I'm curious to know from your side, what would a
neuroscientifically informed philosophy of mind look like in

(01:45:12):
practice for you? I think I will go back to
something that Lawrence said actually very much at the
beginning, which is this recognition of the complexity of
the system that we're trying to understand, that we're trying to

(01:45:35):
explain that in some cases toy examples and and simplified
models are really the only thingthat we have available to us.
And they can be very powerful. And sometimes a really highly
oversimplified explanation or model or description of what's
going on is surprisingly powerful.

(01:45:59):
It's it's really kind of remarkable how something as
simple as, well, I'll use an example from my own field,
Signal detection theory can actually do a remarkable job at
explaining how, or at least describing, maybe I shouldn't
use the word explanation, but describing how an observer like

(01:46:19):
you or me is going to deal with noise in our environment or in
our own minds. And signal detection theory, it
turns out, was not even developed for psychology.
It was developed to understand and characterize the noise in
electrical circuits in like the 1950s.
Yeah. How do you find the signal in
the noise? That's basically what it's
trying to do, really. Almost stupidly simple

(01:46:41):
explanation. Stupidly simple system does a
pretty good job at targeting howand maybe why certain kinds of
behaviors emerge in certain kinds of situations from, you
know, a human or or animal observer model.
And yet ultimately the thing that we are trying to capture to

(01:47:04):
explain is things that we know exists.
It's one of the most complex things on this planet.
Brains are really hard. They're really highly non linear
dynamical systems. There's a lot of components that
we have no visibility into. There's a lot of stuff that we
are still kind of floundering around in the dark to try to

(01:47:27):
build even just a just so post hoc story of why the system did
what it did. What are the kinds of
informational structures that are present?
What even could the software look like?
Is it software are we like, whatare we even doing here, man?
And so the recognition of just the sheer mind boggling,

(01:47:53):
unfathomable complexity of what it is that we're trying to
reverse engineer. I think that and, and the gulf
between that and billiard balls on the table, which is a causal
explanation of why this ball went into the pocket or didn't
go into the pocket or something that I think would be we would,
we would all do very well to to recognize the size of that gulf

(01:48:17):
and to try to try to shrink it alittle bit.
So for young researchers who feel pressure to pick a side,
scientist or philosopher, what would both of you tell them
about integrating both parts meaningfully?
Anyone can start. Don't pick a side.
Look at me and Lauren. We didn't pick a side.

(01:48:39):
And maybe this discussion is also highlighted the, the
extraordinary value of not picking a side of not burying
your nose in the sand and just kind of doing the one thing.
And that, yeah, it's uncomfortable, as Lauren said,
to maybe not always be the expert in the room.
I'm certainly no, not the expertin the room in a lot of ways

(01:49:02):
that there's a lot of things that I, I want to have my
fingers in a lot of pies. I want to understand a little
bit about a lot of things. And I do have deep expertise in
a couple areas, but there are a lot of spaces that I have been
in where the folks around me know way more about a particular
than I do. And that direction that that can

(01:49:24):
be the norm and that that's OK. And that a lot of other people
in the room might seem like theyare topic matter experts in
something that you understand. And they are, but you're also a
topic matter expert in somethingthat they're not and you see
things that they're not able to see.
And there's a one example of this from my own life is that,

(01:49:45):
you know, I sometimes go to these, these conferences or
workshops that are really focused on computational and
theoretical neuroscience and, and even neurotechnology.
I'm not a neurotechnologist. I know things about that, but I
definitely am not that person. And there are things that I can
bring to the table as someone who's a little bit more of a

(01:50:06):
generalist, like that's really like, like bringing in.
I, I remember recently I broughtin actually some of Lauren's
work. I said, what you're doing is
trying to build an explanation of, you know, how the brain does
something in order to drive likea neuroprosthetic, for example.
And it would it, you really don't want to just drive the

(01:50:27):
neuroprosthetic, which we can doalready using neural recordings.
But in order to optimize that, it would be really great if you
could understand why that kind of model is working better than
this other kind of model, or whyone type of model is more or
less susceptible to neural drift.
Like once you put the implant inand you train the model, you
come back next week, it doesn't work anymore.

(01:50:48):
Why? Why did that model fail and this
other model might not fail? Like those kinds of explanatory,
those kinds of explanations could be really useful from a
practical perspective. And a lot of the folks in
neurotechnology do not think about explanation.
They don't prediction ability tocapture variance in a system
that is the target and that is the thing that matters to them.

(01:51:13):
And so differentiating between prediction and explanation and
differentiating between, you know, models and targets of
different levels of complexity is something that I can bring to
the table. And I can't help them optimize
their neural implant, but they can inform me about what they're
doing and I can inform them about what what I'm doing.

(01:51:33):
And so I guess learning to buildcalluses and tolerate that
uncertainty and that discomfort of not being the expert in the
room. No one is the expert in
everything though. And so like to a certain extent,
even if you build deep expertisein one area, you're going to
have to navigate spaces where you're not the expert anyway.
So you might as well get used toit now.

(01:51:56):
Lauren, anything you want to addthere?
Yeah, there's a few things I would add.
I think that finding work that you like, finding people and
researchers that are doing work that you're interested in, and

(01:52:17):
as Megan says already may be interested in science and
philosophy is helpful. These are academic fields.
I mean, academia is still prettysiloed, so I don't always get
easy access to scientists and I'm not always credited for

(01:52:40):
working with them or taking the time to talk to them or even
writing in writing publications that get published in scientific
journals. So there are interesting
standards of my field philosophy, philosophy of
science, that are quite different from various
scientific fields where, I mean,they also might not get credit
for talking to a philosopher or writing with one.

(01:53:03):
In really different scientific fields value philosophy very
differently. One of the advantages of talking
to neuroscientists is they already value philosophy a bit
and they're already more aware of it than other scientific
fields. So when I talk to a biologist, I
might have to do a little bit more legwork to tell them what I

(01:53:25):
do and to persuade them that I'msomeone useful to be in the room
in the first place. That's not the case with many
neuroscientists, cognitive scientists, same kind of thing.
You know, cogs size, a field that views itself as
interdisciplinary, and one of its areas is philosophy,
computer science, psychology. So it really does depend on the

(01:53:47):
scientific field and that you'reinterested in.
And it helps to talk to people who work in that space because
they know a bit about the norms in the field and the
expectations. You know, Megan has different
expectations on her than I do inthe field of philosophy.

(01:54:07):
We're both probably doing more than the standard person in the
sense that, you know, people aren't expecting me to get
grants. They're not expecting me to work
with neuroscientists. But I care about that work, and
it's important. One thing I sometimes say is
that being a philosopher of science is a bit odd because you
sometimes feel like you're telling scientists why

(01:54:28):
philosophy matters and philosophers why science
matters. And so there's also philosophers
who I'm talking to, and they do not think that the way to
understand the fundamental causal structure of the world is
to look at anything scientists are doing.

(01:54:49):
Why would you do that? Why would you?
So not only do I, you know, not get credit for interdisciplinary
work in that sense, but they don't see why they should care
about science if they're interested in causation or
explanation in some cases or understanding the world.

(01:55:12):
So in in all of our fields, there are different groups of
people who are approaching problems in different ways.
It's helpful to find the work that speaks to you, the
researchers who are doing thingsyou're interested in.
Also to look more pragmatically at.

(01:55:33):
I mean, it's one thing to study philosophy and to study
neuroscience as more of a hobby,but in terms of going into it as
a APHD student or a professor, you know, there's certain types
of aspects of the, of those cultures that it's helpful to
learn a bit about. And there's also differences,

(01:55:53):
right, in terms of different types of people.
But it is fascinating the differences across fields.
But I have the advantage of, I mean, Megan started studying
philosophy before I did, you know, I started studying
philosophy. I took my first class was
basically at the end of undergrad, and then it shows up

(01:56:14):
a lot later. So I don't have to do as much
legwork when I'm talking to Megan.
And. But when I am working with
scientists, a main goal is to bring the philosophy that's
useful for what they're interested in.
If they want to get pulled into some of the philosophical

(01:56:34):
debates, we can do that too. There's jargon that we're using
that. Yeah, You know, I don't that I
don't want to kind of burden people with, but part of these
interdisciplinary connections islearning how to speak to people
who use very different vocabularies.
When you have someone that knowsa bit of the philosophy already,

(01:56:56):
they already know the vocabulary.
And then, you know, I've trainedin medicine, so I know a bit of
theirs too. But there is still this needing
to be comfortable in an uncomfortable situation where
you're not the main expert and you're leaning on other people

(01:57:17):
and looking for their input too.But once you start to see the
value of that discomfort and that approach, and you're you're
among academics who have the, the, you know, ideal disposition
of being open to being wrong, topursuing big ideas and taking
risks, but also reorienting the Sky's the limit.

(01:57:40):
And then you do have a kind of team in a kind of group that can
start to ask the right kinds of questions so that we can
ultimately get helpful answers. But you know, it's very
interesting to think of the differences across fields.
And as a philosopher of science,it's non trivial to convey to

(01:58:04):
different types of scientists what it is that I do, how it
might be useful. And the same goes for the public
or any kind of audience. But part of what those
interdisciplinary connections help you learn is just, is doing
just that, you know, speaking todifferent audiences and and
working to do that effectively or well.

(01:58:28):
I, I think you know, something as you're talking, Lauren, I, I
feel like something's really just crystallized in my mind
that in this type of discussion,we really say, well, we're, you
know, if you're a philosophically informed
scientist or vice versa, you're the bridge between the, you
know, potentially domain matter experts.

(01:58:49):
And so maybe you're not the expert in the room on, you know,
one of whatever it is that's, that's being spoken about, but
you know what? You are the expert in the room
on making bridges, on finding those connections like that is
your expertise. You're not an expert in, you
know, the, the measles or whatever, right, But you are an

(01:59:11):
expert at, at finding the shape of the problem and, and building
those bridges and this science communication, this, this
ability to translate between specialized vocabularies.
That itself is an expertise areaand it's valuable not only in
academic or scientific or even industry spaces, that kind of

(01:59:34):
thing. It's also valuable in, as you
said, communication to a broaderaudience to translating to make
sure that the people that you'relistening to and the people that
you're speaking to can actually understand each other.
You're a translator, you're a, abridge between disciplines.
You're a holistic trees level orsorry, forest level, not trees

(01:59:57):
level kind of perspective. That is the domain expertise
that someone who wants to occupythis space will bring.
And, and it brings with it the, the requirement of developing
another skill too, which is not just talking, but also
listening. And I think that as a lot of
domain matter experts like we, we tend to want to talk a lot.

(02:00:20):
We tend to want to, you know, come up with our own description
or explanation for what's happening and and push on that.
But it's harder to learn how to listen, especially when you
don't really speak the language.And so having essentially a
translator in the room is such avaluable asset.
And being that expert, can be can be the difference between a

(02:00:43):
breakthrough or just kind of continuing on in parallel with
our blinders on and reinventing the wheel.
Yeah, I think that's. Interesting.
Oh, sorry. There's something you have said
before to Megan about a researcher showing as opposed to
just saying they're doing something.

(02:01:06):
What was that? Do you remember that expression?
Yeah. Like show don't tell, you know,
like don't tell us that you found the explanation for
something. Show us what that explanation is
and how you're writing about it and the story that you're
telling and the narrative that you're constructing.
You know, you want to, you want to take the listener or the
reader and guide them by the hand so that they have that aha

(02:01:26):
moment along with you. This is, you know, this is what
you want to do in storytelling and narrative buildings, what
you want to do in film and media, right?
You want to show the audience, don't tell them.
No one wants to read a a story about that's a list of
accomplishments. They want to take the journey
with you. So this is the same kind of
thing. Yeah.

(02:01:46):
So I wonder if in terms of talking about this
interdisciplinary approach and perspective and an academic who
does this and does it well, picking up on what you said, one
element of it is being a listener in part.
And then another element of it, I would add is this sort of,
you're open. I mean, when you think

(02:02:08):
something's right, you really stand by it, but you're also
open to being wrong. And one of the challenges of
some work and philosophy is someone becomes known for a view
and then they don't want to change it because they're sort
of known for it. So they're not really open to
modifying it or being wrong. And some of the most impressive
academics I know are truly open to that.

(02:02:31):
And it allows them to reach certain types of peaks that they
wouldn't have had access to. So listening, being open, being
open to considering new ideas, maybe even being wrong.
But then also there's this interesting piece where you do
have to pitch things and you do have to tell a story and getting
a grant, I mean, Megan's more ofthe expert here for sure.

(02:02:55):
But, but also, I mean, when we write papers, you are pitching
an idea. When we're writing arguments,
I'm trying to persuade someone I'm in in similar with a grant
or in any kind of communication,science communication, there is
a lot to the story that you tell, but the best academics,

(02:03:16):
they can back it up. And it isn't just a tell me,
it's also a show me. So they can do both.
And maybe they don't even put that story together until they
know that they could show you. And so sometimes you see, I
mean, scientists are engaged in a social, you know, this is a
this is a social space. If I tell you I've got a

(02:03:37):
mechanism and I tell you I've got an explanation, and I'm
coming from a fancy university and I've done a couple things, I
can, you know, that might go a long way.
And, and we do need to be able to communicate well and some
people can check that box. But if you really want to do the
best work, it's not just being acommunicator, you've got to back

(02:03:58):
it up. And so then when someone asks
you, what do you mean that you say you have an explanation, how
is how is this explanatorily relevant?
What's your guiding principle? You need to have an answer.
Or when they say, what do you mean by causation here?
How is this a call? What do you mean by mechanism?
Right? They need to have an answer.

(02:04:19):
And so we have these buzzwords, they're status terms.
And part of playing the game well is knowing how to use words
that that gain some traction. But if you want to play the game
the best, you just have to back that up.
And really science should be something that we can back up in

(02:04:39):
that way. So that's that's a tall order
for a scientist or a researcher,but it shows you how they're
willing to adapt and that they can really tell you the value of
their work and the justificationfor it.
But you start to see the kind oftheorizing that a philosopher
might do and that scientists aredoing with the scientific

(02:05:03):
practice, and then this interesting aspect, which is
their need to pitch this work right, to communicate it to
other people in papers, grants and so on.
Yeah. So if you wanna, you wanna do
that communication, well that storytelling, well what better
way than to wear 2 hats? Philosopher and scientist.

(02:05:23):
Yeah, I think you both are excellent in in both fields and
and that skill you were talking about Megan, that the fact that
Lauren has that skill, I think you both technically do.
You both are these translators in both fields and, and, and I
think you can see this becoming a thing where most up and coming
scientists, researchers are trying to make sure that they
understand both, both sides nowadays.

(02:05:45):
So when you do look at young researchers, they're, they're
ingrained into multidisciplinaryfields like never before.
You'll see someone doing mathematics, AI, consciousness
research all in one go. And, and it's kind of
surprising, but but super exciting as well because it
means that the future is kind ofbright in that regard.
What do you think that we shouldclose off with?
Anything that you feel you haven't said?

(02:06:06):
Is there anything about this conversation, why science and
philosophy need each other, thatyou feel you'd like to just hone
in on A drive home before we close for me?
I think we've covered quite a lot of ground here, but one
theme that maybe has been a common thread throughout this
is, is the need for recognizing that whatever you're doing,

(02:06:30):
whether you're a scientist or a philosopher or some, you know,
blend of both, you're not doing it in a vacuum.
There's all these other folks around you and that, you know,
doing good science and good philosophy is it's a social and
it's a, a networked enterprise and that no one researcher, no
one expert is an island. And this isn't just you gotta

(02:06:53):
read stuff. Everybody knows you have to read
the literature and it's gobs and, you know, piles of, of
literature all the time. And if you're especially in like
artificial intelligence or machine learning, like good luck
keeping up with archive, good luck.
But it's not just that. It's not just reading and, and
thinking and making connections yourself and working with your,
your local research group and soon.

(02:07:15):
It's, I think really trying to get out and make your network as
big and as interdisciplinary as possible.
You don't necessarily have to bethe true bridge.
If that's not your bag, that's fine.
But recognizing the value of allthese different kinds of
approaches and kinds of ways of doing science as a community,

(02:07:39):
rather than as a collection of individuals, that there's,
there's an emergent property that we should be going for
here. And the, the way to do that is
to recognize value and, and really celebrate the different
kinds of expertise that we can all bring to the table.
So the community aspect, I thinkis something that's been a

(02:08:01):
thread throughout all of this that maybe I'll just bring to
the forefront at the, the end isthat you too, all of you
listeners, you can also be part of this community.
And I'm sure that you already are.
Lauren you. Great.
Yeah, just building on that and adding to that the type.

(02:08:25):
So why do science and philosophyneed each other?
Part of the answer is that the projects that are involved in
both are intimately related. Many scientists I know are doing
theorizing and theoretical work that is similar to the kinds of

(02:08:48):
philosophy of science that I'm engaged in and that other
philosophers are engaged in. So there's a sense in which it's
hard to separate them if you're looking at scientific research
and if you're looking at scientifically informed
philosophy. I think if we're looking at
current research that scientistsare doing, where they're

(02:09:11):
interested in big questions and they're at the forefront and
they're trying to uncover and understand new things that we
don't yet understand. If you're looking at those open
questions in the sort of cuttingedge of science, or if you're
looking at justifying scientificpractice as it's taken place for

(02:09:35):
decades and centuries, philosophy of science is very
useful for both of those projects.
Philosophy of science here is a kind of work that's focused on
foundations of science, precision in the concepts and
the methods that scientists use,the principles that guide their
research and how it how it is that it works, the success that

(02:09:59):
they get, how they reach the goals that they have.
And so this isn't something thata philosopher can do in a
vacuum, right? We're studying and hopefully
working with scientists to get that precision principles and
those kinds of goals to show, you know, how they actually

(02:10:19):
work. And that's a, a kind of
philosophy that scientists do, that philosophers of science do.
And it's helpful for both being able to justify the scientific
method, how science gives us ourbest understanding of the world,
but also when scientists are tackling these big questions, it
helps to see and to look at witha clear lens scientific practice

(02:10:45):
and all of these domains in the principles that you find across
those contexts and across those domains.
So, yeah, very much fields that in some sense can be continuous.
And I think it's very much the case that you find many
scientists engaged in theorizingthat we can think of and in many

(02:11:11):
cases should think of as philosophical.
But of course it's going to depend on what we mean by
philosophy. And that's maybe also one point
of our discussions is that philosophy, philosophical work,
philosophical thinking can mean very different things in
different contexts. But here it's focused on

(02:11:33):
critical thinking, argumentation, and in particular
kind of foundations of of science and how scientific
thinking, reasoning and explanations work and how
they're so successful. Well, I just want to say thank
you both for this wonderful conversation.
You both are definitely experts in the field.

(02:11:55):
I can't wait to dissect your work individually as well,
showcase it highlights as much as possible.
It's a true privilege and honor to have you both.
And yeah, thank you so much. This is a wonderful discussion
and I really enjoyed it. Have thank you so much for
having us. This has been really, really fun
and engaging. Looking forward to the next
time. And it's always fun to hang out
with Lauren and talk about science and philosophy.

(02:12:16):
It's one of my favorite things. Oh yeah, always, always fun to
talk more, learn more from Meganand Ted.
Yeah, Thank you so much. Great to be here and looking
forward to more.

All Episodes

Why Science and Philosophy Need Each Other | Lauren Ross & Megan Peters

Episode Transcript

Popular Podcasts

Stuff You Should Know

Medal of Honor: Stories of Courage

Dateline NBC

.css-15opob5{left:0;position:absolute;top:0.8rem;} All Episodes

.css-14f5ked{margin:0;word-break:break-word;display:-webkit-box;-webkit-box-orient:vertical;box-orient:vertical;-webkit-line-clamp:2;overflow:hidden;}Why Science and Philosophy Need Each Other | Lauren Ross & Megan Peters