BJKS Podcast

118. Lauren Ross: Causation, mechanism, and explanation in neuroscience

Share episode

Lauren Ross is a professor of logic and philosophy at the University of California, Irvine. We talk about her work on causation, mechanism, and explanation in neuroscience, Lauren's background in medicine, how to write clearly, and much more.

BJKS Podcast is a podcast about neuroscience, psychology, and anything vaguely related, hosted by Benjamin James Kuper-Smith.

Support the show: https://geni.us/bjks-patreon

Timestamps

0:00:00: Why Lauren studied medicine

0:04:23: Differences between medicine and philosophy

0:21:19: Why Lauren switched to philosophy of science

0:25:30: How to learn to write clearly

0:30:21: Are doctors practitioners of causality?

0:34:25: What's so difficult about causality?

0:38:46: Causal structures: mechanism, pathway, cascade, circuit.

1:02:11: The practical use of thinking about causal structures and varieties

1:11:35: What's the difference between a circuit and a pathway? And what are you trying to do?

1:20:31: Secondary features of causation/causal varieties: strength, stability, speed, specificity

1:29:29: A book or paper more people should read

1:30:45: Something Lauren wishes she'd learnt sooner

1:33:29: Advice for PhD students/postdocs

Podcast links


Lauren's links


Ben's links


References

Alon (2006). An introduction to systems biology: design principles of biological circuits. [There's a lecture series by Alon that seems to be based on the book: https://www.youtube.com/watch?v=N6VZeWuME_A&list=PLLbr-B8cNbo6v4kc68JowzUeAYdh6gdQH]

Barack, Miller, Moore, Packer, Pessoa, Ross, & Rust (2022). A call for more clarity around causality in neuroscience. Trends in neurosciences.

Forsyth (2013). The elements of eloquence: How to turn the perfect English phrase.

Hempel (1965). Aspects of scientific explanation.

Ross (2021). Causal concepts in biology: How pathways differ from mechanisms and why it matters. The British Journal for the Philosophy of Science.

Ross & Bassett (2024). Causation in neuroscience: keeping mechanism meaningful. Nature Reviews Neuroscience.


[This is an automated transcript that contains many errors]

Benjamin James Kuper-Smith: [00:00:00] I mean, we were talking mainly about, uh, most of the episode is gonna be about your work on causality and mechanisms in neuroscience in particular. I guess you, uh, from what I understand too, your work is more broad than that, but mine isn't so, or limited by by that. Um, uh, and yeah, I often like to ask kind of how people got into what they're doing and that kinda stuff, and I think in your case it's particularly interesting because it's, at least on the surface, it seems a bit unusual.

Um, but, so from what I understand, uh, or maybe I'll start like this. When you were in high school, was it just you wanted to, when a doctor, or was it more a bit open and then decided later on to go to medical school? Or what was the kind of, what do you think you were gonna be doing when you like entered university?

Lauren Ross: Yeah, good question. When I was in high school, I was actually not that clear on what I wanted to do for my career. [00:01:00] I wasn't really thinking about it all that much. I was interested in school, but it was, it was more when I got to college that I started to think of different career options. And when, when I was at the end of high school and the beginning of college, I was very interested.

I mean, I've always been interested in science. Um, my father's a microbiologist. My mom is an elementary school teacher, and so I kind of always had parents who were giving me an environment where kind of science and learning about the world was coming up. And so I knew I loved that. At, at that point I was more focused on, you know, nutrition, science, actually partly because of athletics and just an interest in health and, you know, human physiology. So when I started college. I fortunately had the option to, [00:02:00] you know, think more about what career path I wanted to take. And what I ended up doing was this very interesting process of elimination. So basically when I started college, I thought, well, I'm a nutrition science major and I'm interested in that.

So the standard career is a dietician. And then I shadowed dieticians. I learned more about the job and I was like, well, I don't really think I wanna do that. And I just went through a sequence of careers. I think after that I shadowed physical therapists, I shadowed pharmacists. And then I ended up, I was in, you know, clinical settings and hospitals, and then I shadowed physicians. And in particular I, remember watching, um, open heart surgery. And once I shadowed physicians, but in that case, you know, in surgical [00:03:00] settings, in ER settings, then I was somewhat hooked and just thought, this is amazing. You know, you're learning so much about the human body and you're seeing people making a difference in the lives of people coming in just on a daily basis and having this massive positive impact.

So fortunately I was kind of looking at the day-to-day job of these different, um, in this case, healthcare practitioners. And so yeah, that's, it was at the very end of my first year of college where I decided to become pre-med after shadowing physicians. And then I was just sort of off to the races. You know, being pre-med is a, a good amount of work. And so that's when I really became interested in medicine. [00:04:00] And then, um, yeah, I ended up going to medical school after that, of course. And then you, I'm a philosopher of science now working in academia. So the, the career trajectory changed more later, but that's how I got started and that's how I became interested in, in medicine 

in college. 

Benjamin James Kuper-Smith: What's interesting to me is that they're all. At least the ones you mentioned now applied. Right? It's all, uh, in the healthcare system applied, trying to help in a sense, well, specific people get better in that context, which I guess is, uh, again, on the surface different from being a philosopher of science.

Lauren Ross: Yeah, I think that's a really important point, especially with respect to, and this will probably come up later, especially with respect to the kind of philosophy that I do and the kind of philosophy of science that I do, which I mean inevitably being an academic and philosophy is less, [00:05:00] um, clinical and hands-on than being like a physical therapist or a surgeon of course. But the kind of philosophy that I do expects that a philosopher is familiar with real scientific practice. And so that's something that we sort of pride ourselves. Um, in knowing something about and making sure that our work is meeting that kind of standard, we, we don't want to be doing philosophy in a way that's removed from actual scientific work.

So the advantage of having this, you know, interesting early start to my career is that I was just really immersed in, um, both the kind of background scientific knowledge that you need for this kind of clinical work, and then just hands-on how, um, clinicians are operating, how researchers are operating in their lab. And so it, it did end up influencing my academic work [00:06:00] in philosophy, but, um, certainly a, a unique kind of start to being a philosopher of science. Although that being said, philosophy of science is a space where you. You often find individuals who start in a science and then they come to the philosophy later. They start studying physics, they start studying biology, and then they start studying chemistry, and then they gravitate towards these more philosophical, theoretical questions. And I fit that picture of kind of migrating toward philosophy of science in a similar way to that kind of picture.

Benjamin James Kuper-Smith: So to, you know, fill in the gap between you deciding to study medicine and then doing philosophy, what was, did you think you were then gonna be a surgeon or was that just an example that kind of hooked you initially? Or what was kind of your, what were you, what were you particularly interested in? In the, in the medical field itself?

Lauren Ross: I was interested in [00:07:00] surgery early. On, especially early on when I was in medical school. But then I started to gravitate more towards internal medicine, which is basically a different kind of problem solving. And for me, what really happened is I could tell that when I was, you know, working in clinics, working in hospitals, I was often drawn towards these more systems level, bigger picture questions about, you know, how this, um, how this system was working, what were the different factors involved in someone being predisposed to getting a disease, um, someone being predisposed, being healthy.

There were sort of all sorts of explanations for why these people who were coming into the. [00:08:00]To the hospital, to the er, to the clinic. Were, were suffering in many cases. And the explanations weren't all genetic and they weren't, um, you know, they weren't all,

um, social, they, so there were these kind of really complicated, um, questions about how to understand causes of, um, human health and human disease.

And so I found myself kind of gravitating more toward that, more academic orientation, I would say.

Benjamin James Kuper-Smith: Yeah, I guess it, you know, me having obviously done none of it, uh, it does sound a bit more intellectually stimulating than an orthopedic surgeon or something where someone fell on their arm and broke it. Now they have to bend it back in shape and wait for it to heal. I mean, it's, I know it's more complicated than that, but, uh, I see what you mean in terms of like contrasting different types of medicine.

Lauren Ross: Yeah, there's very different [00:09:00] types of medicine and there's different types of things that people are interested in doing and, and good at. So, um, being in academia requires just a very different skillset than a lot of this work taking place, um, by these clinical practitioners. So yeah, it's a very different, there's just very different types of, of work and some people kind of gravitate more towards certain types than others. For me, I was fortunate to be in a medical school that allowed me to get a master's degree, and that was sort of part of the training, and so I I was interested in getting my master's degree in philosophy. And so that was sort of the, the draw. That was one element that drew me into studying philosophy from medicine is I was sort of already planning to get a master's degree, [00:10:00] um, got it in philosophy and then was just sort of drawn to academia where in this case I could use my medical and scientific background to engage with these philosophical and theoretical questions and to do so in a way that was, you know, taking place in a more academic setting. So, yeah.

Benjamin James Kuper-Smith: the different skill sets. Uh. Basically the simple question is, uh, do you think you would've made a good doctor? And the kind of the, the motivation in particular I'm asking that is because I, I remember I had this once, this, this point, you know, after my masters when I was like, oh, you know, I think lots of scientists have this.

When at some point, especially when you do basic science right, that it's completely removed from any like, real world application. At some point you have this thought of like, maybe I should do something clinically, or, you know, maybe I should study medicine or something like that. And I happened to once spend a, yeah, uh, roughly a week in a, in a neurosurgery department.

And I, and I really like, again watching them basically. Uh, and I really, [00:11:00] I, I really liked that whole environment and all that kinda stuff. And I had this, this brief thought like, maybe I should do that. But then I also realized like. I am too experimentally minded. I think. Like if I, you know, it's like, well, we know that surgery works this way, let's try it a different way like that.

I don't think that's the best, uh, uh, set of interests of someone to be a surgeon, to constantly want to, you know, to discard the ways that work for ones that you don't know whether they work or not. Um, I'm sure that has some place, uh, but probably not on a, on a daily basis. Um, so, you know, I mean, let's be honest, I, I never seriously considered making the switch, but I think that, that kind of thought made me realize that maybe that's not the kind of thing I should be doing.

Um, so I was just curious from what, from what you've described, do you think that would've actually sued you, you very well or, or not?

Lauren Ross: Yeah, that's a great question. The advantage of medicine is that there's so many different [00:12:00] types that one can work in, and so it's, it's. Pretty easy to find this really massive heterogeneity of types of work. If you're a radiologist, if you are, um, a trauma surgeon, if you're a pediatrician, if you're an ob, GYN.

There's just so many different ways to be a physician that I think it can be. Um, there's an advantage for that in the sense that you can find something that's a good fit for very different types of people. In my case, I do think I would've been able to find an area of medicine that was a great fit. Um, I also feel extremely fortunate to have found my current work, which also feels like a great fit.

The thing I'll say is that with medicine, being a physician is, is really difficult in a way that I [00:13:00]think, um. People don't always appreciate, it's a job that is very taxing, and it's really hard on healthcare practitioners in a way that isn't always common knowledge. One element is just, I mean, there's this kind of interesting aspect where they're taking care of the health of other people, but they don't always have an environment where they can take care of their own health. Maybe, uh, they don't always have time to get enough sleep or they're working so many hours. Um, and then you add on top of that, this massive responsibility when something goes wrong in my work, I miss write a word and a paper, you know,

and or. People get slightly confused. Or if I provide my own account of mechanism and it's different from someone else's, they get frustrated because they think that, [00:14:00] you know, that's the level.

Or I'm teaching a class and I'm not clear the making mistakes in, in my current job as an academic. No way to compare the kind of, of responsibility and the kinds of, um, decisions that positions are making and the things that they can be blamed for, whether, you know that's right or wrong. It's a very kind of heavy set of responsibilities and that takes a toll on, on people.

And I think we don't always appreciate that. So I guess what I'd say is I have a lot of respect for physicians, for healthcare practitioners. They make the world go round in a way that. Other types of careers don't. Um, they have real important, powerful, positive impact on people in, in everyday lives. I can point to important positive impact that philosophers of [00:15:00] science make, but it's different.

Um, and so I, I just, um, yeah, I think there's, there's not always an appreciation for the, the kind of massive responsibility that comes with that career on a daily basis. And, um, it's just very impressive to see people working in that kind of way and so successful in that type of work. And I agree with you.

It's very different from other types of careers in all sorts of ways.

Benjamin James Kuper-Smith: Do you, I mean, obviously understand the, you know, having that much responsibilities and maybe not something you always want, but do you in some sense, miss. Parts of that. I mean, what I meant a little bit earlier also about, you know, scientists, you know, you're having a bad day, things are not going well, and then, you know, your stupid experiment doesn't [00:16:00]work and you go like, oh, well, who even cares?

Like, you know, uh, I guess it's the flip side of not having responsibility, right. Where, uh, uh, I'm just wondering for someone who, uh, you know, has, has seen in that sense, both sides of this, is that something where you, uh, I mean, is it just something that maybe, you know, one's work doesn't have to have, but other parts of one's life can have?

Or is it, or do you Uh, I mean, I guess it's not that it's so important, but the, the, the distance to the direct application, it's just far larger. I'm just wondering how you, is that something you think about at all or not?

Lauren Ross: It is, I mean, it's something that I think about more and more in current times where. We see lots of disparities when it comes to healthcare and when it comes to health, of course, it's hard to know what to expect healthcare practitioners to do, but I do miss direct impact with people and patients [00:17:00] and you know, the healthcare practitioner friends that I have and my experience in the profession, there's nothing that compares to the kind of support you can give to people who really need it. The program that I was in here at uc, Irvine, was particularly focused on serving various. Groups, especially ones that don't often get access to healthcare, the Latino and Chicano population. I, when, where I grew up, I was working in free clinics in Guadalupe near Rio Grande. And so I would often see patients who were working in, there were farm workers working in fields and they, you know, were coming in with, um, coming in with all sorts of health issues.

And, um, you know, [00:18:00] there were, there were those patients, there were others. I think it was both frustrating because there were limits on what you could do, but it was rewarding in the sense that you were there for them, you were, you were able to do certain things. Um, and of course that levels up when you're at that point.

I was pre-med when you're actually, um. A healthcare practitioner in a hospital, it's different. Um, but even then there's just limits on what physicians can do. And this is sort of what drew me to a more systems level picture. You have an ER physician and they're prescribing the right medications and they're listening to the patient and they're taking all their time. But if that patient doesn't have, you know, uh, access, if they don't have the finances to get the medication in the first place, like what are you gonna do? And if there's other ways in which they're not thriving in their everyday lives due to their socioeconomic status or you know, their struggling to get their kids access to [00:19:00] education, it's pretty, it's pretty easy to feel like a tiny cog in a big system. And so I think there's both no way to compare. I mean, the, the type of positive impact one can have in that type of career is massive. I think it can also feel very frustrating. Um, it, I do miss that kind of positive impact in the work that I do Now. I can write about it, I can talk about it, I can work to clarify some of these issues, you know, what's explaining various social inequalities or, you know, um, when we're thinking of providing clarity on disease outcomes in, in neuroscience or pathologies, I can work to provide clarity in those contexts.

And I think, um, you sometimes have access to certain types of audiences and resources when you're an academic, but, um, it's very different [00:20:00] from, from the kind of impact that physicians have. So, yeah, maybe in some ways it's sort of like there's people on the front lines and then you kind of need people in political settings.

You need people in academic settings. I guess the way I think about it. Now in terms of making a positive impact on the world is that we have sort of like a all hands on deck type situation. So we need lots of different types of people. We need lots of different perspectives and, and it's sort of this team effort of a varied group of individuals working to make things working, to make things better. But yes, it's very different, the sense in which, you know, an academic, especially a philosopher, the kind of direct impact they [00:21:00] have on, um, on people, right? Or just, um, things in current society when you compare it to someone like a clinician who's seeing, um, many, many patients every day. So yeah, lots of interesting, interesting differences.

Benjamin James Kuper-Smith: And so you made that switch at some point. At some point you had to decide to do PhD. What was kind of lead me up to, up to that. So, uh, yeah, I, I guess, yeah. I'm not even gonna restrict where you wanna start that. I dunno where. This inkling started or how the process was for you, but at some point you had to sign up or apply for a PhD program.

Lauren Ross: Right. Well, so I was getting my medical degree here at uc, Irvine, which happens to have a logic and philosophy of science department. And so I started to realize I had this more systems level, academic perspective more in my third year of medical school. And so I took some classes [00:22:00] in the LPS Logic and Philosophy of Science department here on campus. I mentioned that there was that master's program where I, so I applied to study history and philosophy of science at Pittsburgh. I ended up going there and really just. Got hooked on the combination of medicine, biology, neuroscience, and philosophy. I was working with two advisors there. One of them had his medical degree and his PhD.

Ken Schaffner philosopher of science. The other is Jim Woodward, philosopher of Science PhD. He works on causation and explanation, and so I was just in this with both of them and this whole community of philosophers of science. I was in this amazing community of philosophers who were, you know, training students and doing their own work on theoretical philosophical [00:23:00] questions.

And so there I could use my background in science, my experience with medicine, biology, neuroscience, I could use that to start to answer these. These questions that are showing up and that do show up in scientific context and philosophical context. So once I, once I started studying philosophy of science, I realized it was a great fit. Of course, that didn't mean I was able to, um, to do that as my job, getting a career in academia at that point. Then involved applying for positions. So I, so I start, I, I actually finished the medical degree. I started getting a master's degree, got very interested, finished the PhD, and then I went on the job market for academic positions.

And at that point I got the position that I currently have. I am a professor [00:24:00] in the logic and philosophy of science department at uc, Irvine. So I ended up coming back to the same university where I got my medical degree, which is, which is fascinating and. And yeah, that's the, that's the trajectory. It was sort of this weaving together of, and this and this, you know, interesting path of moving from more of a science focus to philosophy of science.

And that wasn't easy in the PhD program at Pittsburgh in history and philosophy of science. You know, you are a philosopher who has to write argumentative papers and I didn't know how to do that really. I didn't really know how to do that yet. I had taken some classes at uc, Irvine, in philosophy of Science, but I only a, only a few. And so I really had to get to, to get trained in terms of the types of clear [00:25:00] argumentation that you find in the best kinds of philosophy of science work. And so that's what I ended up doing. But I had this advantage of having. A kind of scientific background, but the disadvantage of not knowing yet how to clearly write, but I was at the, uh, just a great place to learn how to do that kind of writing.

So it all sort of came together there, and then it's been, um, it's been a kind of wild ride ever since then.

Benjamin James Kuper-Smith: How, how does Juan, or how did you learn to write clearly?

Lauren Ross: Oh my gosh.

Benjamin James Kuper-Smith: I'm sure it's a easy two step process.

Lauren Ross: Yeah. Couple words on the page. You know, the, the most helpful thing for me was reading, writing. That was clear. That was one thing I did. And then another was just practice. It's one thing to kind of digest, and that was what was so [00:26:00] nice. So the advantage of going to Pittsburgh and having almost zero experience writing. Argumentative papers is that I hadn't learned the wrong way how to do it. I hadn't learned how to, I mean, in medicine you're scratching notes, you know, you're not being trained how to write essays. I hadn't, I hadn't learned, you know, poor skills for writing essays. Um, I just hadn't really learned many skills at all.

So I was sort of in the right place to pick up, um, examples of clear writing and to try to create my own arguments that meet, that met some of those high standards. So reading, writing, that's very clear and that's also making a point. And now the way, yeah. Um, now the way I think about it, we're often in the business of sort of talking with our students about clear writing and maybe what standards [00:27:00] we think we try to shoot for with our writing.

Usually in philosophy. At least the work I do. Um, there's sort of three things I think about. One is having an argument that's positive, where you're, you're not just saying why someone is wrong, but you're saying how we should think about things. So an argument that's positive, where we're often focused on providing arguments that are novel. So I need to say something new, of course. So an argument that's positive, an argument that's novel, an argument that's important, that's a sort of harder, you, you have to have some intuition for what topics are important in this space. And then the kind of fourth thing I think about is, is it correct? So we expect our work to capture actual scientific work and practice.

So if I am 

providing an account of causation, I want it to be positive. I wanna [00:28:00] be saying something. How, how should we think about causation? I wanna say something novel, right? I can't, I need to say something no one has said before. It has to be important, right? You might say something sort of trivial that we already know. And that fourth thing is, is it correct? Does this correctly capture how scientists think of causation or how they causally reason or how we do in our everyday lives? So those are four sort of standards I think about with my own writing, but it is this very interesting sort of intuitive process also where you sometimes get your own voice and you can just see is this sentence, does it have content?

And is it also clear? And maybe also does it, does it sound, does it sound nice?

There's some kind of interesting aesthetic to the writing too. 

Um, 

Benjamin James Kuper-Smith: you have a few examples of clear of, of writing that I mean. Other than your own, but, uh, of maybe the things that, I dunno. You mentioned you, you know, when you were doing your PhD, you read some [00:29:00] examples and that kinda stuff. Is there anything in particular that you wanna point towards that people can

Lauren Ross: Mm-hmm.

Benjamin James Kuper-Smith: a particularly good example of that or.

Lauren Ross: Yeah. One great example is the work of Carl Hempel. He's known for being a very clear, a very clear writer. Um, also one who is providing new novel insights. He has a book called Aspects of Scientific Explanation in many other books and papers. He is a great example of a philosopher who writes very clearly. Um, so that's one, that's one nice example there. Yeah, there, there are many, and I think it's. One advantage for an academic in this position is to sort of see what writing they gravitate towards. There's usually a premium on being clear first and then maybe getting style later, or at least that's how I think about it.

My main [00:30:00] goal is, is always to be clear and I, and that's something that Hempel just does very well. So he would be, uh, a classic

example, I'd say. 

Benjamin James Kuper-Smith: Okay. Yeah, I'll put all this stuff in the description so people don't have to find it. I mean, the reference, you have to find the book yourself. Um, uh, I have one final question about the medical background philosophy switch or parts of your career. Um, which is, so one thing I was, uh, uh, thinking about is that in particularly your focus, one of your folk of research, um, causation.

One thing that occurred to me, uh, and I dunno whether this is just, uh, a coincidence or whatever, but it occurred to me that. Um, the way I see it, at least from as an outsider, doctors are basically, uh, practitioners of causality. [00:31:00] So you, you have a situation and all you do all day apart from, you know, admin or whatever, is you see, how can I influence that system to lead to a desired outcome.

And now you're basically doing the theoretical side or on a broader scale of that. I was wondering, is that a complete coincidence? Is that, uh, has that occurred to you? What's the, what do you think about that? What.

Lauren Ross: That's really interesting, the practitioners of causality. Yeah. So I think what's really nice about that, um, insight is that one of the things we want out of causation is control. You want, you have some interest in changing things and if there is one thing that a physician is interested in doing, it's changing things for the better. And so someone comes to you with an ailment, [00:32:00] an issue or a problem, and the. Uh, main goal of a physician or healthcare practitioner is to help, um, and in many cases, to give advice or treatments or medication that's going to change things and make things better. That's causality, if anything is. There's, um, so yeah, I think there's a couple other things to mention here too, which is that medicine gives us helpful examples of causality.

I mean, in, in medicine there's an interest in figuring out what the causes of diseases are, of course, because we want to treat them, prevent them, cure them. We wanna explain them, predict them, and you need causation for all of that. So it's central, but causation is central to many sciences. It's just medicine maybe sometimes gives us, um, helpful examples and really, um, practical clear ones.

I mean, what's interesting is the work of Hempel, right? [00:33:00] Hempel, who is this? Classic in philosophy of science. Many of the examples he gives when he talks about explanations, they come from medicine. He's bringing up measles, he's bringing up syphilis. And so it is fascinating that medicine is showing up in some of the classic texts in philosophy, on explanation. And yes, physicians and healthcare practitioners are very much interested in causation. They can sometimes do their jobs without knowing the full causal story, of course, and sometimes they don't know the cause, but they can still prescribe in a way that is helpful. So there's still certain things to distinguish there. You can use signs and symptoms to predict whether a patient gets a disease, even if those signs and symptoms aren't the cause of the disease. So we get into interesting, interesting cases where you can. Kind of separate out causation from other types of information. Um, but yeah, I think [00:34:00] the training and the background I've had has involved many different types of examples of causal systems, types of causation and interest in causation.

And then you see what scientists are able to do with it too. And medicine is a great example of that. So yeah, I like that insight.

Benjamin James Kuper-Smith: Okay. Uh, okay. So let's maybe start talking a bit more about your actual work. Um, so what's so difficult about causality? I mean, uh, in a way it's, it's, you know, it's one of those things everyone uses all the time. You would, you would figure that since there's been writing, you know, the old group Aristot will probably figure that out.

Um, what, what's exactly, what's the problem that, uh, uh, or the sets of problems that, uh, people like you are working on and that we still haven't figured out.

Lauren Ross: Perfect. The best way to capture some of the biggest problems [00:35:00] are that, first, everyone agrees that it's important and foundational to science and how we understand the world in, in our everyday lives, but also in scientific context. Everyone agrees that it's foundational for science and understanding the world, but they don't agree on what it is and they can't always clearly say what it is.

So it's this really fascinating case in science and in everyday life. And there's been debates about causation in philosophy and in science for over, you know, for over 2000 years. Um, much longer, but that's if we're starting from Aristotle. So, yeah, the, the puzzle is, it's so important. We all agree it's important.

What we don't agree on is. What is it? How do you identify it? What distinguishes real causal relationships from non causal ones or mere correlations? [00:36:00] That's the, that's the puzzle. Um, and you know, there's a little bit more to say. I guess the second question is, okay, like, well, does that matter? And the reason why it matters is that causation is very powerful. If you can say what the cause of an outcome is, you can do a lot of things with that. A you say you're able to give an explanation. So if you wanna know what causes a disease or social inequalities, or you know, car crashes in society, if you can identify the cause, the main cause or the causes of those outcomes, here's what you get.

You get a, the ability to give an explanation. B, you get the ability to make predictions. Right. C and and maybe most importantly, control causes or factors that give control. And we wanna control all of these outcomes. We wanna prevent diseases, [00:37:00] we wanna cure them, treat them. We wanna, we want to diminish social inequalities.

We want to prevent car crashes. If you want control over those things, you need to target their causes. The fourth thing you get is the ability to attribute responsibility and blame. So when you identify the main cause of an outcome, part of what you're saying is this is what is responsible for the outcome. And if you say that someone's disease is due to their genes, versus, um, not taking a medication versus a social inequality where they don't have access to a kind of resource, those are very different things to attach the blame to. So causation is powerful, and if you get it wrong, it's also. It's also dangerous if you identify the wrong cause of a disease. Um, there's risks for that. If you identify the right one, oh my gosh. If you identify the right cause of a disease right here, you get explanation, [00:38:00] predictive ability, you can now focus on targeting that to control it. And then you can attribute responsibility and blame. So causation is very powerful. We wanna get it right and we all agree it's important.

The disagreement is okay, what is it? Um, and that's what we're interested in here in philosophy of science. And scientists are interested in this of course. And that's the, that's the space here in philosophy of science, where we care about how scientists answer that question. We care about their success in studying causation and causal reasoning.

And then we bring to that normative or philosophical frameworks to help address challenging questions that show up in this space.

Benjamin James Kuper-Smith: So one thing I found, so, you know, as I said before we started recording, I didn't have much to any background in this kinda stuff. So, um, you know, most of what I'm gonna refer to is your work, because that's what I did specifically to [00:39:00] prepare for this interview. Uh, one thing I found quite interesting, uh, is that, you know, as someone who hasn't thought about this that much, I think to me causation often seemed like a bit like a singular thing almost.

Like there's, there's the concept of causation and um, but it seems to me that a lot of your work has gone into, uh, I think one. Phrase you particularly used is cause of varieties, um, or cause of structures. Um, and how, uh, yeah, different causal structures can lead to different, uh, yeah. Types of situations.

I dunno exactly what, what the correct abstract noun is to use here. But, um, yeah. Can you, can you maybe talk a bit more about that in terms of that the, the, how should I put it? The, the, the taking a part of causation in that sense into different types and varieties.

Lauren Ross: Yeah, of course. The, this is a nice question partly because there's this [00:40:00] sort of first puzzle of what is causation and then there's this second puzzle of, when we think of that question, we talk about causation. Like it's one thing, but it shows up in many different ways in science and in our everyday lives. And we need to be able to capture the different ways that it shows up. Um, the way, so how does it show up? Well, if you look at neuroscience, biology, medicine, you see that scientists. Distinguish types of causes from each other. Some causes are probabilistic versus deterministic. Some causes are fast, others are slow, some are specific.

Some are non-specific causal systems with many causal factors. They're sometimes called mechanisms. In other cases, they're pathways. In other cases, they're cascades or circuits. Here, what you see is a kind of rich, causal language that's being used to refer to very different types of [00:41:00] causes in the world.

The way that I think about this is that it's helpful to distinguish what makes all of these systems causal versus non causal. I think of that as there's primary features of causation. That's the hallmark defining feature of causality. And then what you have after that is you have secondary features.

These are just extra bells and whistles and accessories to causation. So it's one thing to note that X is a cause of y. There you need like a primary feature of causation. Something that distinguishes cause from non cause. But then once you have that X causes Y, there's other things you wanna know about it.

How fast is it? How stable is it? How strong is it? And so here we get into this space of causal varieties that those examples I'm giving have to do with causation at a single step, right? How fast, slow, stable. If you look at more complex causal systems, you see other distinctions [00:42:00] between, for example, mechanism, pathway, cascade, and circuit. What you see here are in many scientific domains, a researcher analogizing a causal system to something in everyday life in order to highlight the main features. So the unique features of the system mechanism, it's often analogized to. Different types of machines. It's usually used to refer to causal systems that have lower level causal parts that all mechanically interact to produce a higher level outcome. This is our like standard simple machine. Think of a watch mechanism or a car engine. Now we can come back to mechanism. This is the most commonly used causal term. I'd say it's used in many different ways, but that's one main usage. Uh, the other causal concept that I mentioned next was the notion of a pathway. Biochemical pathways. You have neuro pathways, vascular pathways. [00:43:00] Here scientists are often analogizing the causal system to a roadway or a highway or a city street. You have some. Fixed set of steps on your pathway and something is flowing along them. Like the cars along a roadway, you have neural signaling along your neural pathway, you have blood flowing along a vascular pathway. So here the focus isn't on lower level mechanically interacting parts. You have a set of interconnected roots and scientists will sometimes call these pathway maps. So you have a set of of roads that are representing your blood vessels in one case or your metabolic pathways in another. And this allows you, this is a different way of representing a causal system.

It's a different kind of causal system in the world. And here you're focused on, well what are those roadways and where can this object move along? This more fixed set of roots? So that's pathway other, in other cases, you see use of cascade. Here a system is [00:44:00] often analogized to a waterfall, like a cascade is another word for a waterfall where you have, um, amplification.

So this is a key feature of many uses of cascade. We have examples like the blood coagulation cascade, disaster cascades in society. There's cascading reactions in physics and in chemistry you basically have a small cause and it explodes and it produces a huge amount of some effect, and you get this amplification along multiple steps until you get this just huge effect from a very small cause.

Scientists also analogize this to the snowball effect or the ripple effect for that same reason, small, tiny snowball at the top of a mountain causes an avalanche. These are systems that are very powerful. They can be very dangerous. Um, and they have this unique feature of amplification, which the idea here is that scientists are using the term cascade [00:45:00] to highlight the unique feature of the causal system. Another example of cascades is how COVID is spread through the population. One person with COVID can spread it to two. They can each spread it to two more. So you get this cascade model one cause many different effects. And then the circuit concept is one that's found very often of course in neuroscience circuits, um, that are productive of more complex outcomes and behaviors, but also reflex circuits here, what you find is often analogies to circuits in electronics. And computer science context where you have some kind of fixed connectivity in a system and then you have the flow of some computation along that connectivity. There's a sort of similar part of what the, part of what's being emphasized in a a causal system that's a circuit is you have fixed wiring and then you have computations that are moving along [00:46:00] that wiring and usually there's a kind of input and an output. So if you think of simple reflex circuits and neuroscience, there's some input behavior and then there's some kind of output behavior. And this is a very different kind of system than a mechanism where you go all the way down. It's very different from a cascade of course, too. So part of what you see a philosopher of science doing in this space is, is, working to clarify what are the different types of causes and causal systems that scientists are studying and. What are the implications of the different features that they have? Right? These are systems that have very different types of behaviors. They give very different types of control over their outcomes. And sometimes these are all called mechanisms and other times they're not. And a, a kind of challenge here for neuroscience in particular is mechanism is a status term. It's often used [00:47:00] to determine what work gets published and what work gets funded. I mean, so much so that you find this listed explicitly in grant calls and in journal publication guidelines. The, the challenge is mechanism is a causal word, but it's not always clear what kind of causal system it refers to.

So the main thing philosophers of science are working to do is to help provide clarity on what scientists do mean, and sometimes what they should mean by causation and also by causal words like mechanism. And that's part of what my work and other work in this space is aiming to do, is to give that kind of clarity.

And maybe part of what you can see here is we're doing it in a way that's balancing an attention to actual scientific work. We call that like a descriptive focus. I care about what scientists are actually doing. But there's also a normative element where you're looking at the goals [00:48:00] of science and you're using philosophical frameworks to not just capture how scientists do think about causation, but also to say how they should. And so there's kind of perspective to this, these topics like causal varieties. That involves this very careful balance because you don't just want a philosopher to show up and say, I've thought really hard about causation in my armchair, and here's what I, here's what I figured out. You know, they, they need to know what scientists care about.

They need to know what their goals are. They also need to know the scientists are very successful in reaching their, their goals. Not always, but in many cases they are. And so the right kind of philosopher of science is attempting to capture that success. How are scientists able to untangle these causal systems so successfully? We wanna capture how they, how they do that, and the way they, they do that successfully [00:49:00] relates to how we should think of causation. Um, and part of what you can maybe also see here is that it matters for science communication. We need to be able to communicate to the public. The success of science. How are scientists?

What are they interested in doing? How are they so successful at doing it? Why and how does science give us our best understanding of the world? A philosopher of science here is often working toward clarity on topics that relate to foundational questions in science. You know, what is causation, but also ones that are viewed as more basic.

If we're gonna communicate to the public, you know, how neuroscientists are operating, why in this research space they're able to successfully uncover causes of disease. You know, that involves providing justification for that. And philosophers are often trained to think about those types of questions [00:50:00] and hopefully also to work with scientists in giving answers to them. Um, so there's lots to say about causal varieties.

Benjamin James Kuper-Smith: Yeah.

And yeah, I definitely wanna ask a bunch of questions about some of the things you mentioned. Um, but before that one kind of, um, meta question almost about this is that, um, so you mentioned the, what is it? Four, five, the mechanism pathways, circuit cascades, I forget one. Uh, anyway, those, and those, and maybe, maybe those one more you mentioned, but those, um, and you've, you know, I've, I've seen some like, uh, other talks you gave where you also mentioned those or focused on one of them or something like that.

I'm curious, is that because like why those, is it basically, 'cause those are just the main ones people use. Is it because these are the most useful ones? Is it because that's how far you've gotten so far and you know you're gonna add in another one soon, or, um, yeah. I'm just curious because it seems to me that, uh, uh, you know.[00:51:00]

It's probably a bunch more. Um, but yeah, so I'm just curious why, why those couple in particular,

Lauren Ross: Good. The more so the direct answer that relates to work in philosophy is that there's debates about mechanism, and for probably two decades, I'd say so for a, for quite a while in philosophy of neuroscience and philosophy of biology mechanism was viewed as the only game in town. So the suggestion was in this research and is for many new mechanists, they call themselves. The way that neuroscientists give explanations causal explanations is they cite mechanisms. This is the only way that explanations work. This is the only kind of [00:52:00] causal structure. That scientists study in this space. So the reason why I wrote about Pathways and Cascades and circuits was to show that there are different types of causal systems in these scientific domains, and they're not always helpfully understood as mechanisms. So the, the reason for a focus on those was to show there's diversity and variety in types of causal systems. Um, these are more complex terms that scientists use when they find a very unique kind of causal system. They don't always need to use a term like this. Sometimes they're more interested in. Other features of the system, like the organization of causal steps in the system, maybe the, they're interested in whether there's a linear organization or a feedback loop. So there's, these are just bigger, more complex terms for [00:53:00] systems that you see showing up repeatedly that are similar. But the, the main point was to show, and really the main question here is, what do you mean by mechanism? And that's a question for both scientists and philosophers. If what you mean by mechanism is a system with lower level causal parts. And if what you suggest is that the way that neuroscientists and the only way they give causal explanations is by repealing, is by citing systems with lower level reductive causal parts.

Well, what a part of what my work has argued for is. Is, it doesn't look like. That's the only way neuroscientists give explanations. They don't just cite reductive mechanisms when they give causal explanations. They also cite circuits. They also cite pathways. They also cite cascades. And these are different types of causal systems.

So it was sort of to break us out of this view, that mechanism is the only game in town. [00:54:00] Now, it depends on what you mean by mechanism, of course, and that's one of the most important questions in this space. There's at least three things that I commonly find. Um, there's three different ways mechanisms are, are understood in a lot of work.

The first way is the reductive way that I've mentioned. This is our kind of standard machine type mechanism. Lower level, reductive, causal parts, mechanically interact, you, you know, lower level, um. Descriptions of neurons and how they fire meet this picture. If you have enzymes that are interacting lots of lower level detail in biology and neuroscience. So that's one notion of mechanism reductive. A second notion is just that you have some intermediates in between a cause and an effect. So if you're performing an experiment on a drug in an animal model and you wanna see, does that drug [00:55:00] cause some outcome, you can identify there's a causal connection or relationship between drug and outcome Mechanism here means do you have the intermediate steps in between?

That's a second notion of mechanisms different from the reductive. One. Third notion is very broad and it it just means causal system. So a third notion of mechanism is just. Causality. Like when a, when a scientist here or a philosopher says they have a mechanism, they just mean they have a causal system. And in some ways that's our, that's a very slippery and sometimes a troubling notion because it sounds like you're offering so much more when you say you have a mechanism. And it's a notion that includes so many different types of causal systems and it makes it harder to distinguish those types. But a main challenge is just that mechanism is one word used in many different ways, and it's a word that shows up as the [00:56:00] standard, causal standard of neuroscience work. But neuroscientists use it in very different ways. And so this leads them to talk past each other sometimes. And if this is a kind of foundational causal structure that should guide the field, we need to be able to say what they are. Part of what my work can help us do is to give ways of, of saying what a mechanism refers to, what kind of causal system it is. But yes, I very much am on board with what you're saying, which is that there's just many different types of causal systems and causal relationships in the world. Part of what I think we need in neuroscience and in philosophy is a systematic way to capture those differences in those types. We don't quite have that yet. The mechanism pathway, cascade circuit, this is giving us some ways to get traction here. And also if you look at [00:57:00] distinctions at single causal steps, like how strong is a causal relationship, how stable. How specific is it? What's the speed? Those are also differences in causation. Causal varieties you might think, or as Woodward sometimes calls them distinctions within causation. What we really need is a clear framework for talking about distinctions within causation or causal varieties in science, and hopefully we'll see kind of more progress on that topic in the future. But it's a project that involves attention to scientific work and scientists and then also philosophers who are working to, um, kind of provide frameworks that are helpful and clear and that capture the success of science.

Benjamin James Kuper-Smith: Yeah, it's really what I find really interesting is the, uh, um, uh, because it relates very specifically to discussions we've had in the lab are working, uh, about whether something is [00:58:00] a mechanism or not. And because, so you. The hierarchy, I think is, uh, is, is the first thing that I always kind of, uh, go. Yep.

Um, when, when I say that, because it's basically, I remember in, in, in our lab, there's often the discussion, um, uh, or very much a focus on is it a descriptive model or often there was mechanistic model or account, mechanistic account or whatever. But basically the, the, the ques the, the question is, does your model describe or does it have some sort of explanation in it?

Um, does it explain why something's happening or are you just showing that something is happening? Um, and in which way, in particular, and what was interesting to me is just that we've, you know, occasionally you are, you're here. Like is it really a. Mechanism or not, and what does that mean? And we're never really quite sure, I think what, what exactly to, to make of that question.

But, um, you know, if I, if I, if I understand you correctly, [00:59:00] then I think, 'cause in our case, you know, this is human neuroscience, sometimes purely behavioral research, right? Um, this is definitely, uh, in many cases, not something where we say, okay, the ION channel does this, and if you go one step lower, then this happens.

Um, and I think if I understand it correctly, then in, in our case, it would probably be more useful and accurate to not call it a mechanism, but to say there is some sort of causal structure here that we're trying to capture. And, you know, what particular type it is is maybe an open question and maybe it doesn't have to be answered specifically, but, uh, uh, yeah, in a, in a way I feel like the word mechanism has become a bit of a, just a syn synonym with causation and causal structure in that case.

Lauren Ross: I absolutely agree with, and I think you've identified a really helpful two step piece of advice in this space. Um, I, I guess the first thing to say is that yes, mechanism is [01:00:00] very often used to just mean that you have causality or a causal system. The problem is, it means other things too, and sometimes one slips between all of those meanings even in a single paper, and that's a problem, right? That's, um, that's a, that's a problem when it comes to clarity and being able to, you know, demonstrate something empirically or be clear about it in, in any context. But exactly what you said is a helpful suggestion for moving forward, which is that instead of using the word mechanism, what I think is most helpful to do is to put that word aside and to ask, do you have causality?

Do you have a causal system? Is this causal? Do you have, you know, one cause do you have many causes working together? So it's really, do you have causation? And then the second question is, okay, what kind? Because a mechanism is just, I mean, it either [01:01:00] means you have causation or you have a special kind of causation and you know, whichever thing you use it for.

We need to answer both of those questions. Do you have a causal system? You know, what's your evidence for causality? And then what kind do you have? Because causal systems are so different, you know, you have neurons that are firing and they produce very different types of, of control over downstream outcomes.

They can, you know, increase the probability of something happening. They can decrease the probability of something happening. They can be connected up in different ways. So exactly what you said is. A helpful kind of two step piece of advice for getting clarity in these spaces where mechanism can make it really hard to clearly, um, talk about what one is interested in with a scientific project, and then [01:02:00] what empirical research is, is suggesting or what it supports.

Benjamin James Kuper-Smith: Uh , so one, one thing I find also, so I think.

Sometimes I, I, you know, not, not, uh, I'm not saying this is the case here, but sometimes some of these debates can, I feel like sometimes, you know, veer into the semantic was like, is it this or that? It doesn't really matter. Like, you know, uh, okay, it is maybe clarifies communication a bit, but, you know, who cares?

But, um, uh, one thing that I find quite interesting from just a purely practical perspective, right, is that, you know, it just opens up lots of questions and ways to think about what you're doing to think whether it's this system or that system or that structure, whatever. And the, funnily enough, the analogy I have here is one that goes a little bit back to what we talked about earlier in terms of clear writing, is that one thing that I like doing but never do, and I should do more 'cause I keep forgetting to do, is, um, I find.[01:03:00]

Rhetoric extremely useful for clarifying your writing to, to go. There's a, there's a great book called E Elements of Eloquence, which is like a very funny run through a bunch of, uh, rhetoric devices. And, uh, I think using that to clarify your writing is really useful because you, you know, you might look at a specifically specific rhetoric device, let's say a parallelism or something like that, where you go, you know, these two sentence structures to show that they're similar, should be written in a similar, you know, noun verb object structure or whatever.

And whether you do that or not in your actual writing is kind of irrelevant. The question is more, do you think this is the kind of structure that supports the argument you're trying to make? And I find this, uh, again, I always forget that this is a really useful way to improve your writing, just to go through them and figure out which kind of works for the point you're trying to make.

And it seems to me that also when you're trying to figure out the system you're studying or the question rather you're trying to answer. You know, just, you know, taking the shape of a [01:04:00] particular structure and going, does it fit this one? Yes or no? And if not, in which ways does it not fit? It, I think is a really useful way to ask questions about, you know, for example, with a wiring diagram or whatever where you say, um, a, a, um, circuit, sorry.

Um, we say it has a fixed structure. They go like, well, does this one, and if not, why not? Or how does it change? Or whatever. So from a, from a purely practical perspective, I, I think this is quite useful to have these like, uh, well I was gonna call 'em metaphors, but they're ster, uh, serotypes where they're metaphors, right?

These kind of metaphors to go like, is it like this or like that, and why not?

Lauren Ross: Absolutely. There's a sense in which, as you can see with this type of research, you're paying close attention to the words and the language that scientists use, and you're using that as a starting hint for what kind of causal system they're interested in. And they're using [01:05:00] very rich, colorful language, and that's fascinating. Um, and they're, that's fascinating in its own right. What I think we need here, just as you're saying, is, I mean, we're not interested in, in what's called you make just semantics. It's, we're not just interested in. Flowery colorful language that a scientist is using. We're interested in language that's highlighting something they found in the world that is some kind of fixed type of, in this case, cause or causal system in the world. And one of the best ways to figure out what unique features they're studying in causal systems is to look at what they're calling them. Part of this philosophical work, it appreciates that the way we understand the world is, you know, from, uh, scientists or humans, right? So you have humans that are studying the world and the way they talk about the world is with language [01:06:00] of course. And then there's hints in how they're using that language to talk about what they find and to talk about their conclusions in a, in a study. And so there, there, is this interesting detective work of looking at why do they keep calling this a circuit? Why do they keep calling this a cascade? And there's fascinating history of science examples where when it's McFarland and a number of other physiologists, when they're first uncovering the blood coagulation cascade, they're trying to figure out what to call it because it has this weird structure, it amplifies, and they're, they're comparing it to electronic systems that involve amplification, which is fascinating. They basically found this really weird causal system where you have, you know, one coagulation factor or one triggering event that causes many different, um, enzymes to produce outcomes. They find this one to many causal relationship in a sequence of steps, and they're, [01:07:00] they basically wanna make that complicated system very accessible to their audience.

And so they start analogizing it to electronic systems that have amplification. And then there's both waterfall and cascade models of blood coagulation, and they're using both of those to capture. That it has this one to many sequential kind of amplification profile. And so there is this element of taking very seriously the language that scientists are using to pin down the type of causal system that they're studying. But then you also have to follow it up with some kind of clarification about how this isn't just about words. You know, this isn't, you know, a rose by any other name would smell just as sweet because here we're using a word not just to refer to something in the world, you're using the word to refer to certain features of something in [01:08:00] the world. When you call it a cascade, you're saying that it involves amplification. When you're calling it a circuit, you're saying it has some fixed structure that is unchanging on a certain timescale, and then you're interested in computation, something that flows, some kind of signaling or computation that flows along, that those are claims about the world.

And so here you see you're using, you're looking at the language that scientists use to talk about fixed stuff in the world. And the debate here is about what is that fixed stuff in the world like, is it always mechanistic in a certain sense or does it sometimes have these other features? And, um, it's, I mean, it's, it's philosophy is absolutely a space, as you mentioned, where you do sometimes find debates that are more, see more about semantics, you know, how is a word, how is a word being used or meaning [01:09:00] more of a focus on that here, what we're interested in. Is a project that attends to causal language, but is always focused on capturing types of causal stuff in the world, types of causal structures in the world, and how do we capture different types of structures? So the, the words give you a hint, but your justification needs to always involve more than the words.

And so the part of the justification is, well, when scientists intervene on these triggers in the upstream cascade, it produces this explosive downstream outcome. And so that, you know, it isn't just words, we're talking about different types of causes and, and causal systems in the world. But of course one of the, you know, very interesting things about philosophy is there's many different ways of doing it.

And you certainly find, uh, academics sometimes getting caught up [01:10:00] mainly with words. I guess one thing to emphasize is that. Being a philosopher of science involves this challenge of needing to know a good amount of scientific work also. And that's not easy for philosophers of science to do. It's not easy for them to have that background, but they need to, if they're going to engage in certain types of projects and to do so in a way that's fruitful, and that I think is useful to our understanding in both science and philosophy.

So, um, so an interesting field, right, where you see attention to scientific practice, but rigorous philosophical frameworks. Um, and here, you know, semantics is showing up in a certain kind of way. Attention to rich causal language is showing up in a certain kind of way. And I guess another thing to add is, another thing you see is. A continuity between how we use causal [01:11:00] language and how we reason causally in our everyday lives and in scientific work. Philosophers here often see that as continuous when we, you know, um, we think of causal systems in our everyday lives too. We're also successful in causally reasoning. Day to day.

We're not perfect. Um, but you know, we see that kind of success in science too. And so there's a similar way of thinking about causation across many different domains, many different scientific domains, but also everyday life to scientific context as well.

Benjamin James Kuper-Smith: Yeah. Uh, very specific question, just because I keep wondering about it and it is probably not gonna fit in, so I'm just going to force it in now. What exactly is the difference between a circuit and a pathway? It's never been quite clear to me because it seems like they have a lot of similarities.

Lauren Ross: They do. Circuits typically involve a focus on computation and pathways don't. So if you look at [01:12:00] circuits in neuroscience contexts that produce more complex behaviors or reflex circuits, simple reflex circuits, you see a focus on the flow of signals or the flow of computation. And if you look at another example are circuit motifs and systems biology. There's there also a focus on computation. These are systems also in computational biology. And with the circuit there's more of it of an input and an output focus. In pathways, you don't need an input or an output and you don't need computation. A pathway is sort of like a set of roadways and you have some often physical object moving along them. So you have metabolic pathways where you have the flow of a carbon skeleton. You have vascular pathways or blood vessels where you have the flow of blood, and in this case, they're often represented with kind of [01:13:00] roadmap. Now you still have the fixed connectivity. That's a similarity between pathways and circuits, but the circuit has a sort of goal and like a computational job. When you get this input, the system needs to produce a certain kind of behavior. When you get a, you know, with these reflex circuits, when you step on a nail, there's this. Crossed extensor reflex circuit where you lift that foot up very quickly and all your muscles shift and calibrate so that you can stand on your other foot.

You don't think about it, it happens just automatically, right? This reflex circuit that's serving a goal for the system, and we think of circuits in neuroscience, is often having that goal-oriented feature. If you look at metabolic circuits, we don't think of them as having a sort of single input and output in the same way.

We don't always think of them as goal-oriented in the same way. You're just, you're following something that travels through, um, a, a [01:14:00] set of different tracks and you can explain where it can go and where it can't go. But in this case, there are some similar features across both types of both types of systems.

What will matter is what the scientist wants to emphasize and highlight, given what they're interested in and maybe what they're studying and what they wanna explain.

Benjamin James Kuper-Smith: I mean, it seems to me to some extent that almost, uh, a circuit requires a pathway to be there almost, if I understand, like in the neuro sense. I mean, I'm sure there are, maybe, maybe this is a neuroscience specific thing, but it seems to me that, you know, if you don't have a connection between two neurons, then it can't really be connected causally in any sense.

Right. Uh, so if, if the pathway is just a connection between these two, uh, then, uh, so in, in a way it seems to me it's, it's almost like a, a circuit is a more specific subclass, maybe almost of pathways or.

Lauren Ross: Part [01:15:00] of what it will depend on is what a scientist wants to explain. The way that I think about a lot of these causal terms and really explanation in general is that it's piecemeal. So if you look at some part of the brain or some part of the body. In order to provide an explanation or to invoke a kind of causal structure or system as the one that kind of matters.

And the way to understand the system, you usually have to say, we are interested in first, what do you wanna explain? So if you want to explain this particular behavior, this crossed extent or reflex behavior, um, that's one kind of target. If you want to explain just a single neuron firing, that's a different kind of target. And if you want to explain, so part of what part, the way that I see this is explanation is very piecemeal [01:16:00] for a given, you know, physical thing in the world. If you specify different explanatory targets, different types of causal systems will matter or won't matter for that target. So it's also the case that if you drill down in a circuit, you can also find mechanisms if you think of mechanism in a reductive way, right? So that might make it seem like mechanisms are always there or more fundamental. This is sometimes what new Mechanists will say. And the, the, point here, and the way that I approach this is in order to figure out what causes or causal systems matter, you always have to say, matter to what you have to say.

Like what's the effective interest, what's the explanatory target? And once you pin that down, that allows you to identify is it a circuit or a cascade or just what are the causes in general. And what that also allows you to do it is, is it allows you to, to say what doesn't matter as much in, in many cases [01:17:00] in neuroscience and other fields, we don't think you need to go all the way down to lower level ion channels to give explanations.

And so the way we can make sense of this is once you specify a target of interest, I mean, and that's one of the. Advantages of circuits is they are this more mesoscale causal system. So why is it the case that a neuroscientist can give an explanation that cites this mesoscale higher level of causal system without going down and down and down? Well, the, the way they do this is they say what they're interested in explaining some kind of behavior. And the idea is that explanation isn't this game of how low can you go, but it's a game of what gives you control. And so sometimes the factors that give you control for your target of interest are at a higher level, but you always have to say what the target of interest is. You know, you can't give an explanation for something unless you first say like, [01:18:00] what's the thing you wanna explain? And so this is an approach that is piecemeal in the sense that different explanatory targets pick out different types of causes and causal systems. Sometimes they'll pick out a mechanism, maybe other times a pathway, other times a cascade, or other times just a, a set of interacting causes.

Maybe they do or don't have one of these labels, but it'll tell you why it's those causes that matter and not causes all the way down at fundamental physics or causes all the way back in the history of the world. Because we don't think that all explanations in neuroscience should cite the big bang all the way back in the causal history, or we don't think they should go all the way down to fundamental physics.

And the question is, well, why not? And I think the answer is because those don't give you control over that fine-grained thing of interest. If you want to explain why a patient has this particular behavior as opposed to not. The big bang doesn't give you control over that specific contrast, but there's a [01:19:00]circuit in the brain that does. So part of what we want here also is to figure out what causes matter in our explanations and which ones don't. And that kind of control piecemeal perspective, I think starts to give us some ways of answering those questions.

Benjamin James Kuper-Smith: Um, yeah, it's, it's so funny to me that like, um, this aspect of like, but yeah, what are you trying to undo?

Like, what are you trying to understand is such a basic common sense question to ask. And yet I think when, uh, uh, I'm really glad you highlighted that again, because I think sometimes when I read about this, 'cause I just stuff it, it kind of like these, you know, very common sense kind of questions that in a way.

Not dictate, but like guide you to what the relevant, you know, level of resolution is or whatever you wanna look at. I think it's very easy to kind of forget about that a bit. [01:20:00] Uh, and, and that also like, and that it's not a, a bit of an excuse to not, you know, to, to, to, to, as I put it, that it's, that it's a valid specific point to why you have to look at something in a particular way.

Um, and not just a way of saying like, well, you know, always sounded like an excuse for when you can't do something, uh, when you can't find a mechanism or something like that. It's like, well, yeah, but that's not what we're doing here. Uh, so I'm glad you highlighted that. So as a, as a kind of final question, uh, about causes, causality, seen, causes, structures, um, you mentioned very briefly in passing earlier that you know, some of the factors that matter are speed or timescale and these kind of things.

And um, yeah, I was just hoping you could slightly clarify, found that, because I think, yeah, this is one of those things that was slightly confusing to me until I then in one of your talks, heard a clarification and then it made complete sense. [01:21:00] So, um, yeah. How does speed in this sense relate to causality?

Lauren Ross: Perfect. The way that I think about speed is that it's one of those secondary features, which is a feature that captures how causes can differ, but it doesn't capture what makes something a cause. A primary feature of causation tells you whether you have causality or not, or sometimes what you, you find in the literature, you know, causation exists when you have certain kind of feature. This is debates about causation, but speed is a secondary feature. It's a, a feature that captures distinctions across causes, basically. Some causes are fast and some causes are slow. That's, that's the simple way of putting it. In which we see this shows up all over the place in scientific context. And of course saying whether a cause produces [01:22:00] its effect in a way that's fast or slow.

It always depends on specifying a timescale of interest. So if your timescale is, you know, nanoseconds, that's a lot different from a geological timescale. So if you have scientists looking at enzyme reactions, they take place on a much shorter timescale. And scientists studying geological processes, that's a, a much different timescale. The first reason why speed matters, you can see in that case, is the methods that scientists use to study systems. There's different methods they need and evidence they need for causality if your causal relationship takes place on shorter or longer timescales. Another reason for why it matters is control. So if you think of two drugs that a patient can take, and they both cause pain relief, but one causes it in two minutes and the other causes it in two hours, well you've got. Two different drugs that both cause pain relief, but they cause it in different ways. One is causing it [01:23:00] very quickly and one is causing it much less quickly.

One's causing it slower. So I mean, another reason for why, why it matters is control. And if you look at, there's a, as a side note, there's some really fascinating examples in, um, systems biology, the work of Uri Alon and others, where if you look at small circuit motifs where you have connections of feedback loops, for example, the speed of the steps of these systems matters for the special behaviors that they produce. So, you know, systems in our body, they don't just need to produce outcomes like hormones or. Any kind of outcome, they need to produce them on a very particular timescale. And so our body, when it's kind of regulating outcomes, it doesn't just care about putting together the causes that produce them. It's they, they need to be produced in a very particular timescale.

So control is a, a [01:24:00] second reason. And a third interesting reason is that if you look at cognitive science research, which is something we care about in philosophical accounts of causation, we're looking to cognitive scientists because they study how humans engage in causal reasoning. They call it causal cognition.

Sometimes children are quite good at identifying causes in the world, so are adults, and they study how. Children and adults do that. Something that's really fascinating is we're sometimes biased a against causes that have certain features, in particular evidence shows that we're biased against slow causes. Another way to put this is that we think causes that are fast, are more real, they're more paradigmatic and they're sort of genuine, legitimate causes.

There's ways of understanding why 

Benjamin James Kuper-Smith: uh, coming, uh, where the effect comes shortly after the cause or, 

Lauren Ross: That's right. yeah. [01:25:00] Yeah. And in this case, it's on a timescale typically of like our everyday lives. If you, you know, flip a light switch up and it takes, you know, the light, the light turns on very quickly, we'll view that as more of a, a real causal relationship than if you flip the light switch up and the light turns on a minute later, for example. So there's here, the suggestion is we have research that shows we have a biased for fast causes and against slow ones. But in some ways this can be a problem because when we have scientists communicating to the public, there's sometimes in the business of communicating slow causal processes like climate change or the spread of COVID or pa.

For patients. Radiation effects take place long after radiation exposure, and this is puzzling for us. It's puzzling because we're not geared to accept that slow causes are [01:26:00] legitimate real causes. So part of why this matters is in science communication. We need to keep this in mind and add extra detail and information. And know that our audience can be biased against these causes so that we can provide further information that suggests that they are real. These are very real causal processes. They can be just as strong in the sense of the cause. When it happens, it will produce this effect with a high probability. It just won't happen fast.

It's gonna take a longer time to do that. So the timescale on which causal a cause produces its effect, whether it's fast or slow, it matters for these reasons that have to do with control. The methods scientists use to study causes and also these, um, topics that have to do with communicating about causation and the biases and preferences we have for causes that have certain kinds of features, like, um, speed being a secondary feature that captures [01:27:00] differences across types of causes and causal relationships.

Benjamin James Kuper-Smith: Is then the probabilistic versus deterministic also one of those secondary features, or it just occurred to me because that's, I guess, another thing that makes climate change so difficult is that it's not, it doesn't get s steadily warmer, but you'll have years that are cold on some of the warmer and all this kind of stuff.

Um,

Lauren Ross: Oh, absolutely. So yeah, this is the project of clarifying causal varieties, which you had mentioned earlier. Speed is one of them. So if you think of a single causal step, you have some candidate cause X and some effect Y, and you have evidence that there's causality here, but then there's different secondary features.

One of them is speed. Another the one you mentioned here, sometimes called probabilistic or deterministic. This has to do with strength in part, which is how probability boosting a cause is for its effect. If it's more strong, then it increases the [01:28:00] probability of the effect more. Um, another is stability, which has to do with if the causal relationship holds, if you change background conditions and then yet another that you find is specificity. Um, one notion of specificity is fine grain control. So if I have a dimmer switch that allows me to change the causal handle to different values, I get fine grain control over the amount of light that's being omitted. That's specific control and it's in contrast to binary, just an on off switch. So those are at least four distinctions that you find. Um, capturing distinctions within causation here. Secondary features of causation, they don't, they don't have to do with whether you have causation or not. They tell you what kind you have. So part of what we need here is this part of what you can see. We can start to provide is a systematic [01:29:00] framework for specifying what are the types of causation that we care about in science and in everyday life.

What are the different features of causal relationships? And so these secondary features help capture many of these varieties. Speed, strength, stability, and specificity are just, um, some of, some examples here.

Benjamin James Kuper-Smith: Um, unless you have anything else, I'll go to the record questions now. Okay.

So, uh, at the end of each episode, I ask my guest the same three questions. Uh, the first is, uh, what's the book called? Paper or really anything? Uh, it can be a blog post, but, uh, what's something you think more people should read?

Can be odd, new, famous, completely unknown, uh, just something you wanna, you know, give a bit more attention to.

Lauren Ross: Good. Two of them came up in part in

our earlier discussions. The first is. Carl Heel's book, aspects of scientific explanation, [01:30:00]fantastic piece of philosophical work. And the other is a scientific book, it's Uri Allan's book called An Introduction to Systems Biology. This is just a fascinating, um, introduction to scientific work that is very principled, very clear, and that allows you to see more of these causal varieties, but also in a very kind of engineering functional perspective where you start to see how the body has structured and set up causal systems that are functional or serving various goals.

So those are two, two books that are favorites of

Benjamin James Kuper-Smith: Okay. Uh, second question is, what's something you wish you'd learned sooner? It can be from your work or from your private life, whatever. Just something that, uh, you know, would've helped to figure that out a bit sooner. And maybe if you can or want to, uh, how you discovered it or what you did about it, or, yeah.[01:31:00]

Lauren Ross: Yeah, I'll give one thing that has to do both with career and everyday life, and it's that you can't please everyone. That's, that's something I wish I had learned earlier. It partly comes up in philosophical work or academic work because you are, you're in this space where lots of people are disagreeing and sometimes you find a researcher who has provided just an incredible argument.

It's very clear. It makes a lot of progress, but it's not going to convince everyone, and it's been helpful for me to see examples of that. You have just a brilliant researcher, but their work never convinces everyone. And this is maybe more in philosophy and a helpful lesson from that is that it's, that's not a goal that one should have.

You wanna convince most [01:32:00] people and a lot of people, but if you don't convince a hundred percent of them, don't be hard on yourself because that's either not possible or that's not the right kind of goal in mind. And it, it's with everyday life too. I think it's a helpful piece of advice. We're, you know, we're, there's different types of people in the world and you know, not everyone is going to be our best friend or not everyone will be the kind of person that we have the best fit with, or that likes how we live or our ideas.

And hopefully we're all respectful of each other and we can all live in a space where we kind of co-exist. But you know, it's also. It's also kind of fools errand to try to please everyone, in both an academic setting and an everyday life. More important to make sure you're doing excellent work in, you know, academic life and that you're, you know, [01:33:00] in everyday life, um, supportive of and respectful of other people, but also able to be your own self as well. Um, so

Benjamin James Kuper-Smith: Yeah, it's, uh, the idea of being best friends with everyone sounds exhausting.

Lauren Ross: Yeah,

Benjamin James Kuper-Smith: I'm not sure I even want that.

Um, uh, but still be respectful as you said. Um,

Lauren Ross: Mm-hmm.

Benjamin James Kuper-Smith: final question is, what is it? Oh, yes. Well this is the question that ages with me. Uh, so it used to be advice for PhD students slash postdocs, but now I'm second year postdocs, so it's only about postdocs now.

Uh, any, I mean, maybe here we should, uh, I dunno what it's like in philosophy, but I have noticed that especially around what a postdoc means, can be very different for different, uh, departments and that kinda stuff. Uh. I mean, in my situation, it's the kind of thing where most people take a couple of years of postdoc.

Uh, but I guess you can interpret however you want to. [01:34:00] Um, but uh, maybe clarify that first so it's, uh, yeah, because there can be very different expectations there.

Lauren Ross: Well, advice for the general advice that I think pertains to people in your situation, but also many others, is finding a kind of balance between things you're passionate about and then. A kind of pragmatic orientation to work and career. Sometimes there are certain things that you work on in your studies in a lab that are particularly well supported in the current time that you're alive or in the current space, or the people are just interested in. And it's, there's a sort of practical or pragmatic element to taking advantage of that. You know, we have to pay the [01:35:00] bills, we need to have a job for kind of practical reasons. And so, um, not holding yourself back from those opportunities is the kind of practical piece. And then the, the one, the kind of passion piece is to see if you can get both of those together and you can orient yourself to working on something that you. Do really enjoy, doesn't everything doesn't need to be the top of your, you know, passion list, but if you can inch your way towards things that you could spend a lot of your time on, that's only going to serve you well in the future if you get the opportunity to work more on those kinds of projects because you're inevitably going to need to work on them quite a bit.

And if you're more interested in them, it's easier to do the work. So for me it's this balance of kind of pragmatics and, and passion trying to, to get both of those to a certain degree in one's [01:36:00] in one's work. And it's a, a different kind of math depending on the situation, but I think it's an approach that can be helpful and has certain kinds of advantages.

Benjamin James Kuper-Smith: kind of good, not too extreme idealism in either way. Not, not too much of,

yeah. Yeah. Yeah. I think that's, that's a good idea. Uh, anyways, yeah, those were some of my questions. I think they have many more, but it would be far too long. Uh, so thank you very much for your time.

Lauren Ross: Thank you. Yeah,

really appreciate it.