C (3:10)

Thank you, Carrie.

B (3:12)

I'm really excited to be talking about compositional abduction and scientific interpretation. A granular approach. Before we get to the book, tell us a bit about yourself, how you became a philosopher, why and how this book came about.

C (3:31)

Okay. So how I became a philosopher, I guess I was in what they called the advanced program in high school. I guess it's a. They call magnet programs now. And I guess my best teachers were literature teachers, but I found literature unsatisfying. And when I went off to college, by then I'd gotten the idea of philosophy. And in my first semester, I took Intro to Philosophy and Introduction to Logic, and I liked it. And I guess it's all been downhill since then. What am I interested in? Well, I've mostly been a philosopher of psychology. I started out working on connectionism and then the systematicity of thought, and I wrote a book on that. And then I got interested in extended cognition, or maybe irritated by extended cognition, and I wrote a book on that. And then I started working on multiple realization in psychology, and I started writing a book on that. But then I got sidetracked, I guess. So I had been, you know, I had read Craver's book on explaining the brain. I guess I read it in manuscript and I taught it a few years, and it. And it seemed to me that this mutual manipulability stuff, that, to me, it just came out of nowhere. It didn't fit, and it didn't seem to me like a very plausible view of how scientists confirmed their hypotheses. But to me, at the time, you know, if I was reading it in, you know, like 2005 in manuscript, and. But I just wasn't taken with the mutual manipulability stuff. And it seemed to me that, well, you know, everyone is reading this book and a lot of people are working on it, so someone's going to realize that there are probably problems with this mutual manipulability. And that probably a better account is something like mechanistic constitution as mechanistic constitute explanation. Ibe some combination like that. So I just, you know, I said, okay, well, someone's going to come up with that. And so I didn't really worry about it. And then there was a critique of Craver's book by Baumgartner and Cassini and Gebharder, and that Looked like, well, that looks pretty compelling. This approach is not going to work. And then in 2019, I got wind that Craver wasn't going to give up on mutual manipulability. I said, okay, well if Pravor is not going to give up, he's not going to look at something else. I'm going to look at something else. And so that's when I started thinking, you know, about 2019. Well, let's look at the idea of, of inference to the best explanation with mechanistic constitutive explanation. So that was the start about six years ago.

B (6:25)

And.

C (6:28)

The vision of the book was simple, right? How did Hodgkin and Huxley confirm their, their theory of the action potential? And so I just began reading the original papers from 1952 and working it out from there. And so the book is a science first approach. It's not like, it seems to me, at least at the time, a methodologically groundbreaking approach. But it does now seem to me that it's just not the way new mechanists have done it. They've just really not gone back to Hodgkin and Huxley. I mean, they talk about the Hodgkin Huxley model all the time, but it's not from a historical perspective. It is to me from a textbook perspective. So.

B (7:12)

Okay, so let me stop you there for a second. So we've kind of gotten into the weeds before. We kind of scoped out the trees in a way. So let's stop back because I'm glad you mentioned Hodgkin Huxley model of the action potential. I mean, all this stuff is going to. Because that's your prime. But not only example, but for a lot of people who have any background in philosophy of science and for those who don't really, actually a familiar sort of idea is the idea of singular causation. A billiard ball hits another billiard ball and that particular spatio temporally located event, the first ball's hitting the second ball is an event, a singular event that will be given some sort of a causal explanation. And then there's debates over, well, you know, singular causal explanation versus some sort of general or law based, you know. But the idea of singular causal explanation of a singular causal event is pretty familiar in the, in the philosophy, science, literature. You are talking about, you know, sort of compositional, you know, events or properties or something like that and explanation of those, those things. So from a basic ontological perspective, what are the singular events or properties or whatever that you're calling your compositional things. What are those things that we will then go on to the account of explaining them.

C (8:56)

Okay, so. So again, Hodgkin Huxley is, Is the parade case. So there's a Central figure. Figure one from their 1952 paper. And it is a figure showing a current that they recorded sometime in the summer of 1949. It's just one trace of. Of one action potential. So there's what occurs on the oscilloscope, which they took to measure something that was going on in the action in. In the axon. And then what they want to explain is what is sort of the underlying physical basis of this action current. This current.

B (9:40)

This is the current that goes through the neuron from one neuron to the other.

C (9:44)

And, you know, so, so think of the axon as a tube. I'm talking about the current going into the axon. That's the way they conceived it. And what was the current? It was sodium ions moving in.

B (9:58)

Okay. Through the.

C (9:59)

That's what they, that's what they say. Why do we have this current? Oh, it's sodium ions moving in. That's maybe the short story of. So that's what you're trying to explain. Why do we get this current? Well, it could have been anything. It could. It could have been potassium ions running out. It could have been chloride ions running out. It could have been any number of things. But they pinned it down to it's sodium moving in. So that's what you're trying to explain.

B (10:20)

And they have. Those are charged. They're charged. Right?

C (10:23)

The sodium ions are charged. They have unit positive charge.

B (10:26)

Okay, so it's a spatiotemporal sort of this instance of a positive charge flowing through a membrane. Okay, great. So that's our sort of paradigm case, at least, or poster case of a singular composition event, I guess. Okay, yes. So what's compositional about that?

C (10:56)

Well, it's. You've got this hole which is the axon, and a property or activity of that hole is the current. But then you explain in terms of these, these microscopic parts. I've talked just about the flow of atoms, sorry, the flow of sodium ions that they couldn't see. They are sort of the paradigmatic philosophers, unobservable entities or unobserved entities at the very least. But in addition, they. There were changes in the permeability of the membrane. So they knew at the time that the membrane was. Was highly resistant in its natural state. But then when you trigger it with a depolarization, all of a sudden the permeability changes. And sodium starts to rush in. So they postulated all that. So that's what's composition. You're so the membrane. There's something unseen and unmeasured and tiny in the membrane, but they had really no idea what it was. Now they know it's ion channels, but at the time they had no idea what was going on, but they knew there had to be something letting the sodium ions in. So, so compositional explanations are ones in which you explain, say for example, the activity of a hole, this current in this axon, in terms of the little parts like ions flowing in, and little microscopic changes in the axonal membrane. So that's what makes it compositional.

B (12:20)

Okay, good. Yeah, good. And just to, you know, I mean, it is, the book is clearly a granular approach. I mean, you go into, you know, you know, important detail that actually makes a big difference in terms of how we think about the interpretation of results and things like that. But could you just mention some other sorts of cases that would be of this same type of compositional event?

C (12:51)

Oh, yes. So in the book it was, I focused on this one case, or there were two cases. One is the Hermann Bridd illusion. But let me take another one that I've worked on since. And so in the 19th century, Jacobus van't Hoff had a theory of carbon atoms. He thought carbon atoms had tetrahedral structures. There aren't little planes. At one time they thought that the carbon atom had atoms attached to it in a flat plane or structure, like with atoms attached to it north, south, east and west. Instead, Van taught they were set out as a tetrahedral that, that things were attached to it, pointing out to the corners of a tetrahedron. Now he's coming up with the theory of atomic structure. Something that you cannot see, right. And, and what he did was he, he, he relied on experiments where you purified a KEMP compound, passed polarized light through it. And so then it turns out that when you pass the light through a sample, there are two things that can happen. Sometimes the light will turn to the left and sometimes the light will turn to the right. And so he has this elaborate structure where he has to connect. You know, he's got this sample that he can't really see and he's, he's going to use this, these rotating optical changes to postulate changes in, postulate atomic structure. It was a revolutionary idea in the 19th century. So that's compositional, or here's maybe the most famous one in the world. When Watson and James Watson saw Rosalind Franklin's Photograph 51. It got it to click for him. Oh, I know why this photograph has these features, and it is because DNA has a double helical structure. That's probably the most famous compositional abduction in all of science. No one's, I think, theorized it that way, but that's, I think what was going on. So I think it's all over the place.

B (15:01)

Okay, good. And these are. You call also. You know, again, I'm kind of trying to set things up so that later you can just talk and we can take this for granted. Intra level? Yes, you know, as opposed to interlevel. So could, could you. These are all intra level. Could you, could you clarify that just a little bit?

C (15:21)

Sure. So, so here's what's going on. So in, in the axon case, you've, you do something to the axon, you voltage clamp it and then it induces a current in that axon. So you do one thing to the axon, then you measure something of the axon, and then you postulate some third unseen unmeasured thing that's. So the experiment is intra level. You voltage clamp the axon, you measure the axon, and then you postulate something at a lower level to explain this intra level experiment. Now.

B (16:00)

Okay, go ahead.

C (16:02)

But what Craver, for example, and does thinks is going on is you voltage clamp. And then he seems to, he seems not to have looked at the case, but all the cases he examines are you, you do something like you put a subject in an FMRI and then you measure blood oxygenation levels at a lower level. So that's interlevel. And he thinks you do a bunch of in, you do a bunch of interlevel experiments to get at these compositional relations. So that's so, so that's the big difference between us. He doesn't. He doesn't, I think. And the new mechanists don't even entertain the kind of story I'm talking about where you, you voltage clamp an axon, measure an axon, then postulate something you don't see and don't measure.

B (16:46)

Okay, so the intra level bit is the, you know, you're doing something to the axon and you're, and you're measuring some outcome of that. That's all the intralevel.

C (16:59)

Yes, that is intralevel.

B (17:01)

But the explaining part is you said lower level.

C (17:07)

Yes, you're appealing to something at a lower level. Some parts that bring about the upper level stuff. That's why it's compositional that connects the lower level to the higher level is what connects. That's where compositional explanation comes in.

B (17:21)

Okay, and that's not interlevel.

C (17:24)

No, that's in that. No, that. The explanation, what you're explaining is when you do the one thing, get the other result, you explain it by something that's going on a lower level. We'll take another. Maybe take another example.

B (17:36)

But it's a. It's compositional. Yeah, it's that the. It's that the. It's that the. The phenomenon. I use that word advisedly, but the. The. The. The event that's happening, you know, with. With the ion membrane or whatever, is composed of these things happening in the sodium ions.

D (17:58)

New year, new me. Cute. But how about new year, new money? With Experian, you can actually take control of your finances, check your FICO score, find ways to save and get matched with credit card offers giving you time to power through those New Year's goals. You know, you're gonna crush. Start the year off. Right. Download the Experian app based on FICO scoring model offers an approval not guaranteed. Eligibility requirements and terms apply subject to credit check, which may impact your credit scores. Offers not available in all states. See experian.com for details.

C (18:28)

Right.

B (18:29)

Okay. All right, so it's the compositional bit that I want to make clear there.

C (18:33)

Yes.

B (18:34)

Okay. Because everything else kind of follows from. I mean, that's what makes this a kind of a special, very specific, you know, form of abduction and form of explanation. Right. So. So. So, yeah, okay, so we're talking about, you know, you know, pausing these theoretical entities in order to explain some singular event at a particular level. So singular. So what is a, you know, singular compositional explanation? And then what is singular compositional abduction in particular?

C (19:13)

Okay, so a compositional explanation is. You can explain, well, why did this. Why did this current appear? Oh, and it's because it was generated by sodium ions fluxing into the axon. That's the explanation. And then abduction works by taking. I say, taking the explanons to be confirmed by the explanation.

B (19:36)

Okay, can you expand on that? Because that's, you know.

C (19:42)

Yeah, well, sure, sure. So here. Here's like the textbook story, right? So, so you. You walk out in your front yard and you see these little marks in the snow, and you say, oh, a deer walked through there. Okay, so the deer explains the tracks. Right?

B (19:55)

Right. Or the passage of the deer.

C (19:57)

Yeah, the passage of the deer explains the tracks. And so the deer explains the tracks. And that's actually also A reason to believe there was a deer and simple. So, so the, so what you. Abduction works, I say by confirming the explanans. That's the idea. The explaining thing is given a reason to, to, to be something you take seriously. The explanans has been to some degree confirmed. That's how I think abduction works and that's how I think scientists thinks it works.

B (20:31)

Okay, well, there's a number of different, you know, aspects to your view. You know, you compare it with some more well known views of, of abduction and confirmation and things like that. So let's, let me get, let's get a bit into, into those sorts of, of weed. What, what is the singular compositional abduction in particular? I mean, what is. You have a different, I should say you have a different view of abduction than is, you know, than is maybe dominant or something.

C (21:07)

Yeah, so, so it's, it's totally individual to me, as far as I know, totally idiosyncratic.

B (21:14)

What would you say at the, at now, now is the, maybe the dominant view of abduction or maybe more than two. I mean, you go through the Persians and then. Gil Harmon, um, what is a standard, you know, well known view of abduction? And what is your view?

C (21:37)

Okay, well, I, I suppose the Persian view is a standard view and I think it exists largely apart from, I think the more commonly entertained view that abduction is ibe. So in the inference to the best explanation literature, which descends from a paper by gil harmon in 1965, it's just common to say abduction is IBE. And they just say that as if it's completely obvious and then they move on. And so for them, when I say abduction is not ibe, they, they just, it just makes reviewers angry. They just can't. What is wrong with you? I mean, it's just not a position they are, they seem prepared to entertain. It's a. You're just, you are just out of your mind. You just do not know what you are talking about. And same thing with the Persians. I say abduction is used for confirmation. They say, no, it's not. And they respond with, you know, angry and dismissive rebukes. But let me explain why the difference is.

B (22:41)

Yeah, please.

C (22:42)

So, so Peirce thought abduction is just a way of introducing hypotheses, but it doesn't confirm that. So go back to the tracks in the snow, the marks in the snow. You say, well, you know, maybe a deer walk through here. And Peirce thought, well, that's no reason to believe that the deer walked through there. It's just a reason to think about maybe a deer walk through there. And if you want to really confirm it, you used a hypothetical deductive method. That's, I think, an interpretation of Peirce. Peirce is hard to interpret many times, but I think a very common view is that Peirce thought abduction is used just to put hypotheses on the table. A deer walked through here, and then you have to go and check it. And I think, no, that's not how it works. I think it's always used to confirm. Well, it's not always used, but it is used to confirm. So Hodgkin, Huxley, you know, they did a whole bunch of experiments, a whole bunch of variations, and they're all abductive. And there's no point in doing multiple experiments like this and reasoning abductively multiple times because after the first experiment, the hypothes, this has already been introduced. Why do you keep doing new experiments and continuing to engage in abductive reasoning? It's because you're, you're trying to amass more confirmation. So, so I think. And Persians don't think that abduction is used for confirmation. So that's the big difference between me and them.

B (24:07)

Okay, can I, can I. Before you. So the abduction itself, I mean, this is kind of something that I was thinking, you know, abduction as a general category, as a general type of reasoning implicates explanation. I mean, that's, you know, it's, that's why it's, you know, whether it is or is not identical to inference to a best explanation. You agree that abduction involves, you know, explanation in it. Right. As part of the, as a part of the reasoning process involves some sort of explanatory bit.

C (24:47)

Correct.

B (24:49)

So could you just say, like, if you're using abduction to confirm a hypothesis, you've got an explanation embedded in the abduction.

C (25:02)

Correct.

B (25:03)

And then you've got a separate hypothesis that's being confirmed.

C (25:10)

L' explanance is being confirmed. So there's a whole bunch of stuff in the explanans, and I think all of it's confirmed. I mean, maybe that's not right, but it looks like it is.

B (25:20)

Okay, so then where does the explanation that's in the abduction come from?

C (25:28)

Well, I think it's largely amassed from the background information that scientists have. They know, scientists come to these experiments with a lot of background knowledge. So when Hodgkin and Huxley were doing the experiment, they, they knew basic physics. That current is the movement of charge. So they then say, well, okay, so what things in the cell could be movements of charge. And they said, well, it's gotta be ions moving. Right. And so then they just had to find out, okay, what ions. So they had a lot of background. They knew, basically, they knew the basic physics of charge and its movement of current. I'm sorry, they knew the basic. Sorry. They knew the basics of current. It's just the movement of charge. So then they have to find, okay, what things that carry charge are moving. And they figured it out. It was sodium now, but they had to add an extra thing. They had to do a study. They had to rely on a study, a background knowledge to say, oh, there's more sodium outside of the axon than there is inside. So if you change the permeability of the membrane, that sodium is going to rush in. That's what they. And so they had to bring that to bear. So the explanation comes from all the background information that the scientists have. They just know stuff. So think about, again, the explanation of the marks in the snow in your front yard. You say it's a deer if you're, say, from New Jersey because you know that, well, deer live in New Jersey. Right. But you don't think it's an elk because you don't think elk live in New Jersey. Right. So you're hypothesizing that it's a deer. Depends on what you know about New Jersey.

B (27:05)

Okay, so the explanatory bit is all these facts about New Jersey.

C (27:10)

It's the mustering of these facts that the scientists know. They put it together to form the explanation. That's where it comes from.

B (27:17)

I see. Okay. So anyway, so, okay, so the abduction is a way of mustering, as you put it, these background facts to bring them to bear on an explanation. On the explanation of the, you know, the tracks.

C (27:38)

Yes.

B (27:39)

Right. And confirming.

C (27:40)

Yes. Contrary to what Hearst thought.

B (27:43)

So it's, it's, it's, it's. It is confirming in itself. Right? Not, not.

C (27:49)

It's not proof. Say it's not maybe compelling evidence, but it's some evidence. And because it's some evidence, that's why scientists do a lot of experiments, because they want to get enough evidence. That's sort of the outline of how I think it works.

B (28:03)

Okay, so. So the explaining and the confirming is the same thing in the, in this case.

C (28:08)

Yeah. You get the, you get the confirmation out of the explanation. Yeah. And chapter four has a story about how that works.

B (28:15)

Okay. Okay, so then. Okay, so. So that's the, that's part of the abduction, part of the, of the book of Singular oppositional Abduction. Another thing you mentioned just before, which I thought was really, really interesting and probably controversial, but also to me, pretty eye opening, was this idea that abduction is not inference to the best explanation. In particular, following from your science, first, let's look at what scientists actually do, at least as far as. Let's look at what they write up in their published results or published papers. They don't go about entertaining and eliminating hypotheses in one experiment.

C (29:10)

No.

B (29:11)

Yeah. So can you tell us a bit about this? The whole abduction is not ibe, you know, what. What's the usual view, and where do you think it's wrong and what's your alternative here?

C (29:26)

Well, I mean, if you just want to. You just want to stipulate abduction as ibe, you can, of course.

B (29:32)

Well, no, yeah, right.

C (29:33)

But here's what. Here's. Here's. You know, I've been thinking about this for six years, and I. And here's what I think is. Is maybe the simplest story. Ibe is a package. It's a package that tells you. First, it studies the relationship between H and E, and it's usually treated as.

B (29:50)

H potentially explains E. I mean, hypothesis.

C (29:54)

Hypothesis, H explains some experimental result, E. So it's a relationship between a hypothesis and an experimental result. Okay, so that's what abduction, I think is about. It's just about that step, that piece. But then there's another piece of the puzzle, which is that scientists and philosophers immediately go to, and it's the question, but what about H prime? What about this other rival hypothesis? What are you going to do about that? And ibe tries to answer that question. And Harmon said, well, H is simpler than H prime, and H has a better mechanism than H prime, and H has broader scope. Abduction, as I understand it, doesn't address that. So ibe is a bigger package. It tries to cope with two issues. Why, you know, what's the relationship between H and E? And then why H rather than H prime? H prime, prime. H prime, prime, prime. Okay, so I'm just working with a segment. I'm working with a little tiny. I'm working with a fragment of what ibe is about. So this is again, part of what falls into the rubric of a granular approach. I'm not looking at all, you know, all the sorts of considerations that scientists might use to choose to. To choose among competing hypotheses. I don't have an account of that. That's a different story. And in the literature, I mean, usually it's like you go for the most robust hypothesis or something like that? Well, maybe, but that's not what I'm trying to. To explain. I'm just doing the. The one simple experiment. I'm trying to give you an account of what is going on in, say, page 450 of Hodgkin Hudsley, 1952. It's just a small story.

B (31:38)

Right, but it can't be that small if it implies that the way philosophers of science have thought about these things is kind of misguided.

C (31:51)

I do think it's misguided. I think there are. It is misguided, and I think they're just maybe doing something else. I think maybe what. What the IBE folks are trying to do is not. They're not trying to assess one experiment, but they're trying to maybe assess the whole of the corpus of Hodgkin and Huxley's experimental work in 1952, something like that. It's more plausible as an account of that. Maybe. But, you know, if you read, for example, Lipton's book, he really doesn't make a choice. Abduction. You know, IBE is just everywhere. It just. It's a theory of psychology. It's a theory of how you interpret individual experiments. It's a theory of theory choice. So it's a lot of different things for Lipton, but it's not for me. For me, it's just this one little thing on how you interpret individual experiments in the experimental literature in certain areas. It's a really tiny, tiny thing.

B (32:53)

Right. So it's. It's like, you know, how. Or. Yeah, yeah. What are. What is the reasoning going on in this particular, you know, passage as. As explained in this particular paper about this particular experiment or series of experiments?

C (33:07)

Yes.

B (33:09)

And in each of those, they're not doing any comparison of, you know, what. This hypothesis with another hypothesis. They're just saying, oh, here, here's. Here's. Yeah, okay, fine.

C (33:22)

So some. There are disconfirmations. Right. So sometimes they. I mean, if. If you read Hodgkin Huxley, they go by very quickly when they say things like, well, we know it's not this rival. And I have a theory of that, but it's. But it's a theory of how we know it's not that hypothesis.

B (33:38)

Yeah, well, I mean, that. That kind of raises an issue that I hadn't. Hadn't thought of, which is, just because the weighing of other hypotheses is not explicit in a particular research paper, can you conclude that they. That there. That there hasn't actually been a contrast and a You know, sort of process of elimination going on, maybe, you know, below the hood.

C (34:07)

Sure. So again, by, in 1952, they, they, they had already ruled out that it was just potassium. So the theory in 1939 was the action potential is due to potassium, let's just say, just to simplify. And they did an experiment and one of the consequences of that hypothesis was shown to be wrong. But I have a theory of that interpretation, that experimental. It's a theory of. There's a theory of disconfirmation. So, but, so, and they clearly had that in mind. But that story, I mean, they had in the back of their minds in 1952, but it wasn't what they were doing on page 450 of 1952. They're separate things. But I'm, I'm. But what I'm telling you is I've got a theory of one experiment and a theory of the other experiment, but I don't have a theory of how you combine those two experiments. That's what I've is mostly about is how do you combine the results of two experiments? And they say, well, one's simpler than the other or something like that. So, I mean, John Bickle, he did a commentary on my book at the Deep South a couple of months ago, and he says, look, by the time Hodgkin and Huxley were done in 1952, they had ruled out all the rival hypotheses. So you have inference to the only explanation. And I said, well, okay, yeah, if you want to talk about all the experiments, then maybe the result of all the experiments is inference to the only explanation, if that's what you want to call it. But I'm not trying to do that. I'm trying to tell you what was going on in their minds in 1939. And I've got a theory of that. And then I'm also have a theory what was going on page 450 of Hodgkin Hudson, 1952. So I'm, I'm, I'm. Mine is a theory of the interpretation of individual experiments, not the interpretation of a collective of experiments, which is what I think I've is most charitably interpreted to be about.

B (36:03)

Okay, okay. And so then, so if we're talking about individual experiment, individual instances of reasoning. Right. Individual instances of performing an abduction.

C (36:16)

Yes.

B (36:16)

Explaining, using, using this, you know, sort of, you know, background information. Right, right. As part of your abduction. Right, yes. How do, how are there multiple ways to generalize from these particular. Where do scientists go from the particular instances that you are analyzing very closely. What are the various directions that one goes from there?

C (36:45)

Now, see, I don't know. I don't have a theory of that, but I've worried about it, right? So let's go, let's go to, let's go to the summer of 1949. And again, Axon number 15. So let's say, you know, they, they spent their morning dissecting out Axon 15. And after about three hours they get their apparatus up and running. And then let's say they get, they run and get 100 action potentials. Okay? Now that's a sample, a hundred action potential. Then there's all kinds of ways to generalize, right? They might say, okay, all of these action potentials were stimulated by depolarization. But then they might say, and so all the depolarizations of this axon were going to result in action potential. So they could generalize from the sample of the sample on neuron 15 or axon 15 to everything about axon 15. But then they could generalize in all kinds of ways. They could generalize to, well, what about other neurons in the squid, not the squid giant axon, and what about the squid giant axon of other squid? And then what about the squid giant axon in other species? So there is one species in Plymouth UK off the, off the coast of Plymouth in the, in the United Kingdom. And there was another species of, of squid off the Marine Biological Laboratory in Woods Hole, Massachusetts. And they were, there was a striking passage where they said, well, maybe, maybe the, the squid in Plymouth off Plymouth are not like the squid off Woods Hole. So there's all these questions about how you generalize. And there's the question of, okay, can you generalize from squid to humans, squid to other animals? So there's all kinds of, of questions about how to generalize. But I don't have any theory of that. I have no idea. But I do have a theory. I mean, I do have a theory of, of abduct, singular compositional abduction. And part of what I'm doing is, is teasing out all these different, you know, teasing out abduction from all these other kinds of considerations. So I'm, I'm, I'm bracketing generalizations and I'm also bracketing, as I, I was, I was saying a few minutes ago, how do you choose between H and H prime? How do you choose between rival hypothesis? That's a different kind of reasoning. It's not compositional abduction. Say one hypothesis is simpler than another. That doesn't look like Compositional abduction. You see what I'm saying, that there's different kinds of reasoning and generalization is important, but I don't have any philosophical insight into that. But I think I do have philosophical insight into this little segment of what is going on in science that I think is that everyone is missed. Everyone. So, yeah, I, I don't have a. I don't have a grand theory of scientific reasoning. And I. I guess I say this in the book, but I have a. The most detailed account of singular compositional abduction out there. Just this little tiny thing that's really fairly well worked out, I think.

B (39:51)

Okay. Okay, good. So. Well, that was helpful.

C (39:55)

Thank you.

B (39:57)

Well, let me, Let me just. I mean, I sort of raised this before, but one of the things that you mention, you know, repeatedly, you know, through the book and often in response to philosophers, you know, who you think are misguided in some way, not necessarily wrong, maybe just, you know, doing something different. I mean, that's fine.

C (40:16)

Yes, I think that they're doing something different. They're not doing what I'm doing.

B (40:20)

Right, okay. And I'm, I'm very sympathetic to that. But so you. So in the science, first, you know, let's look at. Exactly. You're not just like how they manipulate, but how they describe what they did, you know, in your typical paper. Right. You know, published paper has the, you know, introduction result, you know, methods, results, blah, blah, blah. Okay. And that's where you're, you know, that's what you're looking at. You're not in the lab. You're. You're at the journal level of description. Okay. Which is, you know already. You know, I mean, I, you don't really discuss that jump from what's in the lab and what's in the paper. But, you know, that's, that's maybe reason a whole bunch of other issues.

C (41:07)

Oh, it does. It definitely does. That's why I just. I, I say that's why I bracket them, because it's. It's a. It's a different animal.

B (41:14)

Okay, so here's the question. You make very clear that you're describing their reasoning in each of these sort of specific cases of compositional event or something like that. Right. You're describing these. Each one, and you're not. So it's descriptive, as you put it. It's not so prescriptive in any way. And so I was wondering, doesn't it have any sort of normative implications if not for the scientists themselves? This is how you ought to do this kind of reasoning. I'll leave that up to you. But also for the philosophers of science who've been discussing science and how scientific reasoning goes. I mean, there's some sort of normative implications for philosophers of science.

C (42:19)

Well, yes, so I mean, my examples of cases is not arbitrary. So Hodgkin and Huxley did win the Nobel Prize. So the scientists think this was good science. And the fellow Van Hoft I told you about, he's now, I guess, lionized as one of the heroes of stereochemistry and many people. And Watson and Crick won the Nobel Prize too. So a lot of the cases I'm picking are cases that are respected by scientists. So it looks as though most historians and philosophers of science want to say here's what the scientists do and here's what we should do. But I think they often get wrong the descriptive part. So I think, for example, say Purse, he's got the wrong descriptive account of how scientists use abduction. But he could say, oh, well, they're just irrational, you know, they're just irrational for thinking that they get confirmation out of abduction. You could do that. Well, I mean, I just don't want to have that fight. It strikes me as is not very interesting. I don't have any insight into that. But I mean, I mean, really, you want to say that that Van Toft was wrong, Nobel Prize winner later, he didn't win the Nobel Prize for that. Watson and Crick were irrational. Hodgkin, Huxley were irrational, you know, Is that what you really want to say? That seems like a tough road of hope. I mean, I'm not, I'm not going there, but I think my examples have been chosen with malice aforethought.

B (43:53)

Yeah.

C (43:55)

So, yeah, you can't do everything in a book. And that's just another thing that I chose not to do.

B (44:00)

Okay, but the, yeah, the implication from, you know, especially, you know, when you're talking about the Persians and Gil Harmon's account of Ibe and Peter Lipton and We'll get to the new Mechanists in a moment, very, very quickly. But the basic idea is you have to. You're not looking closely enough at the actual scientific reasoning in these particular cases, and you should do that.

C (44:28)

There is some value in doing that, I think. Yes, I think there is. I mean, maybe that's not what they're. Maybe they're not interested in how scientists actually do things. I mean, I've had people say, I don't care how they do it, I want a theory of how they should do it. But I would think if, where are you going to get a theory of how they should do it if they know if they don't do it that way? If, you know, if, if how can you be abducted against abduction if they use it? That seems like a difficult proposition.

B (44:54)

Yeah, I mean, it's sort of the scientists, I mean, can just say, well, we don't care what you do, you know. Yeah, yeah. You know, So I guess, you know, the, the, the alternative is, is like should, should scientists care about your account?

C (45:12)

I think they should. So here, here's where I think. Here's where I'm, I'm thinking, like, think about. I think abduction is just the thing in psychology. Psychology would not exist essentially without abduction. And B.F. skinner rejected abduction. So I think the big 20th century debate, part of the big 20th century debate about behaviorism versus cognizantism was tied up with the legitimacy of abduction.

B (45:40)

Really?

C (45:41)

Yes. So if you read the early pages of Verbal Behavior, Skinner basically says, we cannot study these ideas because they are unobservable and we are not going to entertain hypothetical entities. And he's, what he's doing is rejecting abduction, which is what the empiricists did. But by contrast, what Chomsky did, he was willing to postulate all this innate structure to explain rapid language acquisition. I mean, it's, to me, it's like the central, you know, the central unsung, pivotal part of 20th century psychology. To me, it's just huge.

B (46:23)

So, so is that rejecting abduction or is it rejecting, I mean, Skinner, whatever. Abduction or rejecting unobservables. Those don't seem to be the same to me.

C (46:36)

Well, it's, it's a part and parcel because abduction is how you get at the unobservable world. It's how you get at the deep past. It's how you get at microscopic entities. You can't see them. Right. And think about empiricists. Empiricists have a preference, at least for what they can see, smell, taste in here, but you can't see, smell, taste or hear ions or electrons, for example, to use a famous 20th century case. So it was all tied up, you know, so in. So here's where I think scientists really should care, is that they want to, that lots of sciences have, have embraced abduction. Chemistry has, I think neuroscience has. But, but behaviorists have dragged their feet on this, and many working on mental representations have dragged their feet on embracing abduction. But that's, it's the same thing, I think.

B (47:30)

Well, there's a, there's a conflict there, you know, Which I mean, certainly methodologically there's resistance to, you know, I mean, there's a certain, you know, perhaps too dogmatic adherence to empiricism. Yes, right. As part of establishing psychology as a serious science, not based on introspection. Right, right. That's part of the history. And so you can kind of understand why you would get to maybe the extreme of Skinner. But it's certainly a basic hypothesis, at least methodologically today is you can't posit anything without having some observable consequence.

C (48:26)

Right. That's the, that's, that's. But that's what abduction requires. There has to be something. You're explaining. There has to be an explanandum. But I mean, think about the enormous influence that Dennett continues to have. He resists abduction. He thinks that what's going on is you're not postulating mental representation to explain behavior. He's got this elaborate story how that's not really what is going on, that's alive and well.

B (48:48)

No, I know.

C (48:50)

So, yeah, yeah, right. You know, it's alive and well. Right. So. So you sort of asked earlier, should scientists care about this? Oh, I think they should because if you're going to. So they've got this kind of, they're kind of in a bind. It seems like you, if you're going to accept that Hodgkin and Huxley's reasoning, abductive reasoning was okay and, and Watson and Crick's abductive reasoning was okay and Van Hoff's reasoning, abductive reasoning was okay. What's wrong in them in the mental representations case? Now there could be something. But you gotta, you, you. Now you have all these acceptable cases of abduction. So what's wrong with other cases that are unacceptable? And I think there could be answers. But I think here revisiting what abduction is could be really helpful to practicing scientists and philosophers of psychology, for example.

B (49:38)

Interesting. Okay, let me. We're, we're getting close to, you know, running out of time here. So let me just. This might be getting opening up too big a weed section. But you mentioned, so I'll just go back. In the very beginning, you kind of started talking about the new mechanist and Karl Kraver and Stuart Glennon and Markovich's. Well, you actually didn't mention the matched interlevel experience. You Talked about Interlevel vs Intra. Can you just say as briefly but comprehensively as you can, how does your expand that criticism of either the mutual manipulability or their general sort of matched interlevel experiments account. And how does your account compare to theirs?

C (50:40)

Sure, there is an easy story. So let's go back to the action potential. Right? You have an X, you voltage clamp the axon. That's an X. And then you measure a current, that's a Y. And then you postulate a Z. Whereas they say, no, what you do is you. You manipulate an X, you measure a Z, you manipulate the Z and you measure the Y. And that's just not. That's the difference in the picture that for them, there is no hypothesizing unseen, unmeasured things. It's. You measure X and then. So you measure. So what they're. What they. Here's what set me off. In the very beginning, they said, you voltage clamp an axon, and then you measure an ion moving. And then you manipulate it. You move an ion, and then you measure a current. That just doesn't happen. That's the. That's the nub of the difference. That is what set me off six years ago, that picture that. You can't. You cannot manipulate an individual ion. You can't measure an individual ion. You just can only postulate them. And if you. Maybe you don't like the ion case, but you can't think about the set changes in membrane permeability. They couldn't measure the changes in membrane permeability. They just could only postulate. That's the big difference. That to me, to my mind, the manipulationists are empiricists. All they want is what you can manipulate in your measure, what you can see. And they just think. Sometimes that's what I think is wrong.

B (52:06)

So they just. Are they just, like, not. Yeah. Okay.

C (52:11)

I think they aren't looking closely enough at what Hodgkin Huxley actually did. Hodgkin and Huxley did. Right? I mean, they know textbook stuff, right? I mean, Craver has forgotten more neuroscience than I will ever know. But I don't think he has. You know, I don't think he's ground through, you know, Hodgkin Huxley, the second paper of that series, as many times as I have it. For me, it was just brutally hard work. But he's not. He's not going at it that way.

B (52:39)

Well, in a way, they've kind of put. I mean, if they're. If they think that you're manipulating ions, then clearly they have posited them. Correct.

C (52:49)

That's why I would think yes.

B (52:51)

So are they sort of taking your bit for granted?

C (52:56)

I don't know. I don't think they've actually thought through this case. Right. So if you read the mutual manipulability story, you know, in Kraeber's book, it's kind of set apart from actual experimental work. You know, it's like it's in its own little world, you know, of the Woodward framework.

B (53:19)

Right. Which we did not, unfortunately, did not get to. Yeah. So I do wonder. Their concerns just might be post postulation of the unobservable. And your concern is postulation of the unobservable.

C (53:38)

Yeah, but I mean, they still can't. I mean, they still can't do this stuff with Hodgkin Huxley. You still can't measure there. There is nothing like a, an an ion radar gun that, that allows you to clock individual ions moving. I mean, so, I mean, if we're, if we're waiting for mutual manipulability, we've still got a few years to wait. I think.

B (53:57)

Oh, I should. Yeah. Okay. All right. Well, I think, I mean, we could, you know, I'm. If we had more time, I would surely go into the, you know, relation the data, you know, Bogan and Woodward and so. But that, you know, that, that will have to, you know, I'll leave that as a, as something for readers to, to, to read. But I think you've given us a very good explanation of, you know, kind of what's, what is going on in the book. And I'm, I'm really, I'm really pleased about that.

C (54:33)

Well, thank you for your questions. I think, I mean, you can see my enthusiasm for this book. I believe in this book like nothing I've ever done.

B (54:41)

So, so what is, what is. What is next for you? What is, what are you working on now? Are you, are anything connected to this book or have you turned to something different?

C (54:50)

Oh, no, no, I think I've just begun. So I think that I want to tell a bigger story of, of the role of abduction in, in say, the empiricists versus the rationalists. One of the great divides is the empiricists didn't like abduction, whereas the irrationalists did. And so really, the battle over abduction, you know, the legitimacy of abduction, is a multicenter, multicentury debate. And so I want to talk about that. And I want. So the next book, I want to talk about that. But then I also show you to bring it to bear on the topic of mental representation and maybe cognitive ontology, because again, I think that philosophers of psychology, all of them probably have heard the slogan that mental representations are postulated in order to explain behavior. But that's just abduction. That's just what I mean by compositional abduction. So I think in general, this can be a legitimate form of inference, but I think it is an under theorized form of inference. So what I want to do is flesh out how it works. But there's an all I'm also, I think I want to work on. I think I spent a lot of time, you know, saying, okay, it's not Ibe, it's not purse. But I didn't spend enough time, I think, fleshing out how it all fits together. Right. So there, I think abduction is, it tells you we need something that, to fill this job description. And then another task, a separate one actually, is to figure out what things in the brain fit that job description. So they're really, I, I, I think in the case I've looked at, I've not pulled apart these two parts of abduction. One is postulating things that do a certain job and then finding the things that do the job. So you knew early on that, that when Hodgkovsky knew early on that to explain these currents, you gotta have some things moving across the membrane, they knew that. And then much of the hard work then was to find out that it was sodium ions rather than anything else. So those, but those are two separate sorts of inference. So I want to develop a new theory that recognizes that abduction has these two parts and they've never been distinguished, but I think they should be distinguished. And I think that would help reconcile what's going on in, say, philosophy of mental representation and neural representation. So that's kind of what I want to go. I think, I think there's just so much more to be done on this. It's just entirely untouched. But you can see I'm just fired up about it.

B (57:24)

Yeah, well, it's great to talk with somebody who is so fired up and it's very, I look forward to seeing because the whole mental representation debate probably needs a shot in the arm, I think, from a very different direction. And I think you would give it that.

C (57:47)

Well, thank you. Thank you. I hope so. I hope people will listen.

B (57:51)

But in any case, we are out of time. So I just want to thank you for a great interview and I wish you luck with the stuff that you're working on.

C (58:03)

Thank you.

B (58:04)

Okay, bye.

C (58:06)

Bye.

B (58:07)

You've been listening to an interview with Kenneth Aizawa, professor of philosophy at Rutgers University in Newark. We've been talking about his new book, Compositional Abduction and Scientific A Granular Approach, which is just out from Cambridge University Press. I'm Carrie Figdor. This is New Books and Philosophy, a podcast with the New Books Network. I hope you enjoyed the podcast, and thank you for listening.

D (58:46)

Reggie, I just sold my car online.

C (58:48)

Let's go, Grandpa. Wait, you did? Yep.

D (58:51)

On Carvana. Just put in the license plate, answered a few questions, got an offer in minutes. Easier than setting up that new digital picture frame.

C (58:59)

You don't say.

D (59:00)

Yeah, they're even picking it up tomorrow. Talk about fast.

C (59:03)

Wow. Way to go. So, about that picture frame.

D (59:06)

Ah, forget about it. Until Carvana makes one, I'm not interested.

B (59:10)

Car selling made easy on Carvana.

C (59:14)

Pick up fees may apply.