Ep. 323: Acquiring Language: Tomasello vs. Chomsky (Part One)

0:00 Mark: Foreign. To the Partially Examined Life, a podcast by some guys who at one point set on doing philosophy for a living, but then thought better of it. Our question for episode 323 is something like, how do children acquire language? And we read chapters 1, 2, 8 and 9 in Michael Tomasello's Constructing a Language, A Usage Based Theory of language acquisition from 2003. His 1995 article Language is Not an Instinct, as well as the Stanford Encyclopedia philosophy article Innateness and Language. For more information, please See partially examined life.com this is Mark Linsenmayer directing the attention of my various conspecifics in Madison, Wisconsin.

0:43 Seth: This is Seth Paskin in Anguilla, British Virgin Islands.

0:48 Wes: This is Wes Alwan pointing at you from Cambridge, Massachusetts.

0:52 Dylan: This is Dylan Casey sharing basic cognitive resources with many of you in Madison, Wisconsin.

1:00 Chris: And this is Chris. He's gesturing at my coffee on the table from Los Angeles, California.

1:06 Mark: Welcome back, Chris. You were on our Wittgenstein On Certainty episode. Wes had recommended Tomasello and I'm like, I think Chris brought up Tomasello. And in fact, I got Chris as my ringer because I was going to take my vacation. But then since Dylan couldn't make it, the day that I couldn't make it, the whole group took a vacation, basically, as people know, since you just heard two weeks of things that are not regular episodes before this. So here we are. How are you guys all doing?

1:35 Seth: Great.

1:36 Chris: Yeah, yeah, I'm great. How's the cruise, Seth? Amazing.

1:40 Seth: Yes.

1:41 Dylan: You're on a sailboat, right?

1:42 Seth: Yeah, we have friends. They have a sailboat. It's a 45 foot catamaran sailboat. And so we're staying with them.

1:51 Dylan: That's so awesome.

1:52 Seth: We were in St Martin and we went to St. Barts and now we're in Anguilla.

1:57 Mark: So I feel like this is a first for us for having someone. Actually, Dylan takes a lot of trips on boats, but has never actually tried to record something with us from a boat.

2:06 Dylan: I have not. Not even in port, which is probably where you are. You're surely not sailing?

2:12 Seth: No, no, we are. We're anchored off of a beach right now.

2:17 Dylan: Yes.

2:18 Mark: All right, well, if you vomit, I will assume that it is because of the bouncing of the waters and not because of the quality of the conversation.

2:24 Chris: Sure, yeah.

2:26 Mark: Wes, you introduced this topic.

2:28 Wes: Why? Well, part of it is I've just been interested in Tomasello for a long time. I've dipped into various of his books. I find the question of both ontogeny and I guess it's phylogeny I guess the question of how we first came to use language as a species, really fascinating. And then language acquisition in children, fascinating. And I think it's especially important because we talk about language a lot and philosophy. And in particular, you know, I think when we started talking about Wittgenstein and language or meaning as use, I felt like that was the first time I started to feel satisfied with the way we talked about language or felt like I was starting to have some insight into what language is.

But maybe the best way to put it is we talk about a lot about language and philosophy, but never to my satisfaction at any. I feel like I never know exactly what it is or how it works. It just seems like magic. And Tomasello's concept of joint attention, which I think is in itself actually quite a sophisticated idea and not unrelated to, say, the Hegelian recognition hall of mirrors and all that stuff. But that, for me is an aha moment. And I think if we want to talk about language as much as we do, we should have a better idea of what it is. Exactly.

And I think that's really strongly related to the psychology of its acquisition. I don't think you can get away from that.

3:54 Chris: Yeah, I had a very similar experience myself, going through philosophical readings. Obviously, I went through what I would call a Wittgenstein phase a bit ago, some number of years back, where I was reading everything, and it really clicked with me and sounded intuitive, his idea that meaning is use. Because I had obviously read the Frege and Russell stuff, and that was always just so abstract. But Wittgenstein was the first one. I was like, oh, no, this guy, he gets it. And so from there, I started thinking, well, I started reading what's the sort of discussion in linguistics today?

And so from there, you go to kind of Skinner, and then you get to Chomsky, and he has some really. For me, when I read Chomsky, I thought it was extremely insightful, especially his poverty of stimulus argument. And it all sounded very intuitive to me. And then, of course, from Chomsky, you kind of continue on and you see that there was kind of pushback on Chomsky. And, you know, again, reading Tomasello, I agree with Wes that there was an aha moment, especially with the understanding of joint intentional frames, intention reading, why you would think that language would evolve, and why it's species specific to us with our gestural sort of dispositions and stuff like that.

It's all just really, really fascinating. And I still feel in between them both, some of Chomsky's intuitions and Pinker's intuitions still pull me a certain way. And it was interesting to really delve deeper into Tomasello because I had done mostly just periphery research on him and read some of his papers, but I hadn't really delved into the book. So that was really cool to be able to investigate that. But I agree with Wes that it's really. I always felt like when I was studying or reading language and philosophy, I always felt like, kind of like there was this missing piece.

Like, oh, I don't really know how language works. So I feel weird making claims, strong claims about meaning and all these sorts of things. So it's cool to kind of get into the empirical work.

5:41 Mark: So we've been urged to read Noam Chomsky for a long time, whether his contemporary political stuff or even have him on the show. Apparently he's kind of promiscuous and where he will appear. So, you know, it's a possibility. But I was a little scared of that and I wasn't sure about. So his. In the Stanford article, it was saying, the first main thing here is that he wrote a review of Skinner's 1957 book, Verbal Behavior in 1959, but then he kept publishing about that, at least through the 80s is what Tomasello is reacting to. And he established this whole paradigm which then. Steven Pinker is very, very popular writer. So this article that we read by Tomasello, Language is Not an Instinct, is a review of Steven Pinker's book from the 90s, the Language Instinct, how the Mind Creates Language, which I've also read as well.

6:33 Chris: And I went through again during this, just to kind of go back over some sections and things. It's a good book. It's very well written.

6:39 Mark: Obviously there was an audio version of that. So I actually got through several chapters of that, just, you know, because it was easy. I wish there had been one of the Tomasello because this Tomasello, we assigned a lot of reading to ourselves and it was pretty redundant, I found. I don't know, it just depends how fascinated you were. And it is the kind of thing that, like, you know, I was able to get through of all of Darwin's Origin of Species because I listened to it as an audiobook. But actually just the way that I'm reading at this point, this is kind of a struggle for me.

But yeah, so Tomasello is. Even though Chomsky started this whole paradigm of universal grammar, generative grammar, the idea being that there's a lot that is innate about language use, obviously not learning the particular language, but learning just about everything that you would need to learn a particular language. Like as if you've got an operating system that we're all born with. And then it's just a matter of the particular software of English versus German versus whatever coming in and filling in the spaces. And so that is still a very prominent thing.

And Pinker is a representative, a proponent of, of that and is still very active now, I believe in that area. Tomasello is reacting against that, taking more of an evolutionary psychology point of view. Actually we don't need a universal grammar. We can through culture and through the fact that we are unique among species to have the ability to. We talked about in our first episode with Dr. Drew, we talked about theory of mind, which doesn't have anything to do with philosophy of mind. It's just that people have an idea that other people have intentions.

And it's the fact that apes don't actually really have that. They don't have that they can't see through another person's eyes or at least most animals cannot do this. And that's the thing that we pick up, you know, especially by about age 1. And so language. Tomasello thinks that that's sufficient actually to enable us to learn language in the way that we do. Whereas we, you know, the people Chomsky and others that he's disagreeing with think that they're has to be more baked into genetics than that.

8:34 Wes: Evolutionary psychology is typically associated with the type of thing that Tomasello is in fact opposing.

8:40 Chris: Well, he does have an evolutionary point of view on the origins of language. And he really, he kind of leans on that to sort of show us why a usage based theory makes just as much sense. Sometimes people call him a psycholinguist. Obviously Chomsky has his own evolutionary point of view about how it started, which is a bit peculiar.

8:58 Wes: Evolutionary psychologists, like, similarly to Chomsky, have a reputation for saying, oh, you know, why do we do this particular thing? It's a product of a particular brain module that must have had a selective pressure on it. And, and I used to argue against this sort of thing all the time on the show and say, well, no, this could just be a product of more.

9:16 Chris: Yeah, it could be a spandrel general.

9:19 Wes: Yeah, Cognitive capacities. That. And that's the type of argument that Tomasello is actually making here. We don't need grammar built into the brain or some sort of universal basic grammar that allows us to learn language. We can learn language based on this capacity for joint attention and being in Joint attentional frames on the one hand, and then on the other, on our other cognitive capacities and including our ability to detect patterns and things like that.

9:45 Dylan: A fundamental aspect of this to me is really the argument about what does innate mean? Because in the end, I think they're both saying that it's innate. It's just a question of where you draw the line. What's the equipment that you bring to the party such that you would say that it's inside your structure or you as an entity.

10:05 Wes: How specific is the innate equipment?

10:08 Chris: Right, yeah. Chomsky did change the paradigm and such, at least from the Skinner's disposition, where he thought, well, there's some sort of cognitive capacity or constraints involved in children acquiring languages. Not just empirical sort of hypothesis testing. They take in some data the same way that for sure, that's still the case regardless of whether you're a Tomasellian or a Chomskyan, and you still think there's innate cognitive capacities that contribute to language acquisition. It's just whether. How explicit are the rules that end

10:37 Wes: up allowing for, well, how specific are those capacities? Are they general purpose capacities, multi purpose, and they can be used for language as they are for many other things.

10:46 Chris: Right.

10:47 Wes: Or is there something very language specific and in fact grammar specific built into the brain?

10:52 Seth: It feels like maybe we just take a minute to talk about. Tomasello takes great pains to just sort of be very clear about the three arguments against, you know, the three arguments for the criteria that I think will help pull out the distinction that you were just talking about between.

11:09 Wes: So let's start with his review.

11:11 Seth: Let's start with the review. So essentially, I'll just give a quick summary. Skinner was the famous behavioralist. Right. So knowing language is just having a set of dispositions to do or say something in response to the world and others, that it's a matter of conditioning. You're conditioned to have some kind of behavioral action or some kind of response to the way other people say and do things or what happens in the world.

11:36 Mark: So a kid tries to say something, the mom smiles or says, I don't understand you. There you go. That's, you know, conditioning.

11:44 Seth: Yeah. The idea that this kind of harkens back to the old notion of ostension, which has a different term in here. But you point and say dog, you know, and you point and say dog, and eventually you say dog enough times, pointing at enough similar things, and the kid says, oh, yeah, concept of dog

12:01 Wes: association is also very important there. Right. Just being able to associate.

12:05 Seth: And so what Chomsky says in response to this is, look, language is historically unbound and it's stimulus independent. There's not. Parents don't train children in language. That's not the way language teaching or language learning works. And it's not that you can only say things that you've heard before, right? Like, it's not like children come up with, they say new things that they've never learned, or they put together concepts and words that they've never learned. So he's like, look, language acquisition is not training. Children are not trained. And more importantly, the stimulus that they get, the linguistic stimulus they get, it's too passive and there's not enough there to actually build out a language that they use. We can't look at languages of stimulus and say that there's enough there that you could actually come up with the language that you learn as a child.

12:57 Wes: Grammar is very complicated and there's just not enough there for them to build out grammar.

13:03 Seth: There's not enough in the data and you're not taught that grammar. There's not enough in the grammar and you're not taught that grammar. So, you know, basically what he said is there has to be more involved in the concept. And his point is there has to be something innate in the children that they bring to the table in order to develop and acquire the language because the training is not enough. The stimulus of linguistic data is not enough. The complexity of language, that's the ultimate outcome can't be gotten out of a combination of those things.

And so he proposes that there's something innate that children have. And I guess later on that becomes known as the universal grammar. And a lot of people pick up that research paradigm and try to figure out what that might look like and how it goes. But the point is that to what we were saying earlier was part of the Chomskyan argument, right? Is that the notion of the way that we philosophically conceive of using our reasoning capacities to build concepts and to do inference and so forth isn't rich enough, can't explain language in particular.

So language acquisition is a special thing. It's a special instance or a special faculty of the mind that isn't a part of it, just a general reasoning capability.

14:20 Mark: Could I throw in an example by Pinker? Just on that first anti Skinner point that I think is right on and addresses the way people think about language today is that parents today feel like I have to be constantly talking to my kid. You know, it's like, I need to play them Mozart. They're Going to be left behind. We need to, you know, give them all the stimulus so that they can catch up. And Pinker, following Chomsky, thinks that this is just not necessary. Other cultures, some of the parents like don't talk to their pre linguistic kids. Like the pre linguistic kid is not going to understand me.

Why would I even bother to talk to them? And they still develop just fine. And he compares the two in some other cultures they're like, it's important for my kid to be able to sit and to stand. So we'll put this kid on the ground and we'll bunch sand around them to put them in a sitting position before they're able to do it by themselves. And wow, eventually they learn to sit by themselves. And we know that the sand is not necessary. And you know, so this other stuff is just as not necessary for language. Kids will start talking as long as they're around people who are talking.

You don't have to try to teach them language.

15:19 Dylan: What I love about that example is, is it's exactly that kind of example of like learning to walk as a emergent feature of our biology. That Tomasello will point out is exactly how we learn language.

15:32 Chris: A fun little tidbit too, on Chomsky has a thing he writes on reflections on language. He relates, I know I'm not supposed to mention other readings, but he relates the problem and his mark just did it. So sure, yeah, I'll just, I'll blame it. He relates his issue with Skinner. He says that he came up with it thinking about Plato's dialogue.

15:50 Dylan: You know, Chomsky said this.

15:52 Chris: Chomsky said that? Yeah, he said that he refers to the language problem as one of Plato's problems, which I just thought that was really cool. But it's spot on really.

15:59 Dylan: I have this note about this where exactly this thing in my notes on the language is not an instinct, where it sounds like. It's just like the myth of recollection. The Chomsky's response is we have everything in our brains already and we're just going to recall. We're going to be triggered to recall how to use R. And that's what learning is. It's unearthing something that we already have.

16:25 Chris: Yeah, I always enjoyed that little connection.

16:27 Wes: It's worth pointing out in this context that we went through all of this with Locke versus the rationalists, right? What's innate, what's not. Even with Locke, we concluded, well, there has to be some kind of general cognitive capacity that's innate. Even though he's not a nativist. Right. And that's why he talks about relations and all that stuff. So very, very similar conversation that we've had probably several times on this podcast in the context of epistemology, but here, more specifically, it's going to happen in the context of language.

16:55 Mark: Even arguably for Kant, you don't have specific knowledge of, like this thing will cause this thing, or specific knowledge of everything has a cause. It's just that when you end up perceiving things, when you're old enough, you know, to have object permanence, you will perceive things causally. And so does that mean you have the innate belief in causality? Sort of. You have an innate template for, you know, things to fit in. And so that's the kind of thing

17:20 Wes: Chomsky's claiming causality is a good example because it's relational. Right. You get two objects and then they get fit into this frame and a similar type of thing. Maybe we should try to give some examples from Chomsky. What are some items in the universal grammar that correspond to, say, a Kantian category of causality or something?

17:39 Chris: Recursion would be his biggest one. The recursive element.

17:42 Wes: Yeah, so say something about that.

17:44 Chris: So Chomsky really puts a lot of weight into being able to engage in recursion. It's sort of a computational mechanism. He thinks that we, in mathematical thinking and our linguistic thinking, we're engaging in recursion. So one of the easiest examples, I guess highlight recursion is just embedded clauses. So language is infinite in such a way that you could say, you know, Mary likes soup. Mark thinks Mary likes soup. Jeff said, Mark thinks Mary likes. And you could do this infinitely. And this ability to use sort of discrete elements to create an infinite amount

18:17 Seth: of needs is sort of.

18:18 Chris: He thinks that that's essential to universal grammar into language and it creates a kind of constraint for it as well. He thinks that's across languages. All languages engage in this recursive element or have the ability to. Which separates it from.

18:31 Dylan: And he would argue that that recursiveness is also something that can't happen without it being innate.

18:37 Chris: Yes. Yeah. He thinks that that's an unobservable sort of grammatical rule that he thinks people can't come to just by hearing enough data or something like that.

18:47 Wes: Tomasello mentions a few things. Phrase structures, long range dependency relations, closed class elements, lexical categories. I'm not going to try and define all of those, but lexical categories, that seems like a very basic one to focus on. Like. So, for instance, What a noun is, what a verb is that's supposedly built in. Correct me if I'm wrong on any of this, Chris, because I. I didn't do a lot of looking into. Into Chomsky, but that's true.

19:13 Seth: He.

19:13 Chris: He thinks that there as noun phrases and verb clauses, you're already set up to parse those.

19:18 Wes: And then critically. Right. Those are. I think many of us might think, well, you know, language just kind of maps on to reality and their objects and their actions. And so the semantics, you know, and the way the world is, it's. That kind of grammar kind of arises in response to that. And so we might think that we could build that out. And I think that's what Tomasello is saying. One of the things he objects to is the idea defining nouns and verbs purely in terms of syntactic distributions or defining them in these very abstract ways that Chomsky and others do, and saying they're built into the brain when they're something more closely related to the way the world is and. And to the semantics and the sort of things that come out of joint attention.

20:10 Dylan: So again, this is like, where's the line for innateness come? And an argument about whether or not human beings basic potential abilities that are not contentful, whether use of those activities is rich enough to generate the ability of us to have language. And Chomsky seems to be saying, no, it's just not possible. And Tomasello's arguing, yes, of course it's possible. And then there's the basic features of the Language is not instinct. Tomasello reviews in Pinker's article, Pinker's book. What is the evidence for this? And there's like four or five different areas that are supposed to be the categories of evidence that we need this generative grammar, this innate ability in order to account for the fact that we have language, which Tomasello takes up specifically.

And his book is also sort of an extension of that, taking up these arguments specifically to say, well, either it's a misinterpretation or it is an unrequired sophistication of the theoretical account, at least

21:21 Seth: as far as the book is concerned. Because I went through and read the stuff that Wes had originally. The specifics of the Chomskyan position. It's not generally speaking that language acquisition requires this. He's talking about grammar, and he's saying grammar is not something that can be acquired empirically or through experience. Or maybe more precisely, there are aspects of the Chomskyan account of grammar that would necessitate some kind of a priori or, or innate knowledge. And what Tomasello is going to argue in his book is. Well, no, he's actually going to say two things.

One is he's going to use a notion of grammar that is not the exact same one that Chomsky's putting forth. And then he's going to say, and I'm going to give an account of how that grammar can in fact come through experience combined with all the various faculties or capabilities, you know, like just basic. The ability to have basic attention spans and create this joint frame of attention and so forth. But one of the things Tomasello does point out, as he says, you know, the Chomskyan position, is that grammar is somehow either prior to, or somehow symbols, symbolism and syntax.

Wes, you're going to have to help me here. It's. What's the right way to say this?

22:42 Wes: It's syntax and semantics.

22:44 Seth: Or syntax and semantics is somehow a subset of syntax or syntax. And Tomasello's saying no, in fact, we have the semantic experience first and we are able to generate, so to speak, syntax out of our semantic experience. That to me is a really interesting point about his argument.

23:04 Dylan: That's what it means to say we build language by use.

23:08 Seth: Yeah.

23:08 Chris: Chomsky's model is sort of a grammar first model and Tomasello's words first or semantics first and sort of grammar sort of either emerges or falls out of.

23:17 Seth: Yeah. And what's really interesting, it's not just words first, it's symbolism first. So that's the thing that's really interesting.

23:23 Dylan: It's all about meaning and communication, which is why the fundamental linchpin for him, as Wes mentioned earlier, is the recognition of intent and the ability to process intent, which is the fundamental piece of communication.

23:38 Wes: And the recognition of a particular kind of intent.

23:42 Dylan: Yes.

23:42 Wes: So not just any intention.

23:44 Chris: Communicative intent.

23:45 Wes: Yeah, communicative intent. Because it's an intent that's directed towards our intent.

23:49 Dylan: Yeah. So it's this triadic intent. Right. It's not just a processing of signals, which would be dyadic between two entities, but that we understand the intent of our parents and we process their intention about some third thing as opposed to just being told or experiencing.

24:08 Wes: Well, also we are sharing that attention in the sense that it's not just that they're attending to something. And then we. Because we do, according to Tomasello, we do some of this with language where we're adults are engaged in these goal directed situations and we understand those goals and then language gets used in context and so we can put ourselves in the position of the person who is trying to accomplish that goal. So, for instance, if you use a word in the context of a. Someone trying to achieve a goal and they fail, the child will still understand what the word means and use it correctly.

They won't simply think the word referred to the failed thing. So it's. Teleology is inherent, but for something to be symbolic beyond that, there has to be the idea on the part of the child that the adult is interested in modifying their own intentions, their own mental states. Right. So that, you know, I think maybe we should wait until we get into the book and we can say very, very specifically how all that works. But for when it comes to this general sketch in Tomasello's review, this capacity for symbolization, as you guys have been pointing out, is very important.

And then he points to a few other capacities, right? Just the. Our vocal capacities, our conceptual and perceptual capacities, our capacities to categorize things and to do pattern recognition and ultimately to construct these abstract symbolic schemas. Right. So all of this is about how does a child build out grammar? How do they, how do you learn grammar? How do they abstract grammar from this torrent of data that they're given, which in the beginning is just right. The way we speak language, it's not even segmented in a friendly way. Right.

They just get this stream of speech. And then the first thing you have to do is to say, what are the patterns in this, you know, stream of noise? And how do I. And it turns out children are very good at. Even young infants are very good at detecting those patterns. Primates as well, actually.

26:11 Dylan: That list that you just gave, I just want to comment that not all of them are species specific. There's only a couple of those that are species specific, but other ones are part of how other mammals work or primates work. And there's some of these things we've been talking about that you would say that have to be true about the cognitive capacities of all kinds of animals, especially pattern recognition.

26:37 Wes: Yeah, Critically, what's missing is the symbolic thing and the joint attention thing is the most.

26:43 Mark: So we had this discussion on the philosophy in a new key by Langer Langer, where we sort of went through this back of the hand evolutionary take on how we think people developed language. And, you know, we didn't have these many citations right at, you know, up to date citations at our fingertips to say, you know, one of the things I really enjoyed in the Tomasello, where he's talking about and this parallel some things we were saying in the Langer discussion about, like, alarm calls. So it seems like when a certain kind of monkey, and it says they have three kinds of alarm calls, one for snakes, one for eagles, one for something else, and that would determine, like, what the other monkeys are supposed to do. If it's snake, then what? You go in the lake or whatever. If it's a ground animal, you go in the tree.

27:30 Wes: That's the rule of thumb. And it rhymes. And that's how they remember the lakes. Snake, gun lit.

27:37 Mark: You go away from the predator. But that seems like it's referring to something that is absent. Right. You, the hearers, who are somewhere off in some other tree, cannot hear this thing. And it's not merely, you know, I'm trying to signal to you. It seems like it's just not a reflex response. It seems like this means snake, but its purpose, I guess, is to incite, you know, the behavior. So it's not just like, hey, you know what? Snakes are pretty cool. It's not just referring to the snake.

28:07 Chris: Yeah. It's never disinterested in that way.

28:09 Wes: Yeah.

28:09 Mark: If it's just supposed to be pragmatic to bring about some action, then it's not purely symbolic. And so that is the line that humans.

28:17 Wes: They're not trying to do. So these are vervet monkeys. And it's not symbolic or referential in the sense they're not trying to manipulate the attentional.

28:25 Dylan: Yes.

28:26 Wes: Or mental states of others when they do it. This is a hardwired behavior. This is a hardwired signal, Right?

28:34 Seth: Yeah.

28:34 Wes: Evolutionarily. It's not passed on culturally. There's no vervet monkey culture. They will do it even if they're

28:40 Chris: raised by other monkeys, they will give

28:41 Wes: off these calls if they're raised. Yeah. By humans with no contact with other vervet monkeys.

28:46 Dylan: I just want to clarify something. Right. Because what you're saying is that the evidence is that if you had vervet monkeys that were not raised with any other vervet monkeys at all, and you took a recording of the three different calls, they would execute those same three things.

29:03 Wes: They will do the behavior that's associated. And this is not uncommon. There's a lot of this in the animal kingdom, these types of animal and

29:10 Chris: birds who will build nests without seeing other birds build nests, Things like that.

29:13 Dylan: Yeah.

29:14 Wes: So straightforwardly instinctual. I think that in the Stanford article, we get more. There's more detail on this whole vervet monkey thing, because it's clear that they will do it. It's clear that they're just doing it instinctually and they're not thinking, oh, I'm going to communicate this with everyone else so I can save my friends. It's not. It's just. And chickens do the same sort of thing. Right. This is a point that Tomasello makes in the book. And in the context of there's no other domain in which they're using gestures or joint attention.

So anyway, Tomasello's argument is that those sorts of calls, you can interpret them symbolically. And in fact, they're not where you want to begin. If you want to compare primate communication with human communication or point to the sort of beginnings of human style communication, primate communication, you would have to actually start thinking about gestures, not that sort of hardwired stuff.

30:04 Seth: Just going to quote here from page eight, which is the first page in the book. And by the way Langer used the term, she talked about signaling versus symbolic, representing human linguistic communication. A sign versus assigned versus a symbol. Yes. So human linguistic communication differs from the communication of other animal species in two main ways. First, and most importantly, human linguistic communication is symbolic. Linguistic symbols are social conventions by means of which one individual attempts to share attention with another individual by directing the other's attentional or mental state to something in the outside world.

Other animal species do not communicate with one another using linguistic symbols, most likely because they do not understand that conspecifics have attentional and mental states. To oversimplify, animal signals are aimed at the behavior and motivational states of others, whereas human symbols are aimed at the attentional and mental states of others. It is this mental dimension that gives linguistic symbols their unparalleled communicative power, enabling them to be used to refer to and to predicate all kinds of diverse perspectives on objects, events, and situations in the world.

So, yes, he's reinforcing. Tomasello is reinforcing the idea that we do something special with language. But he's tying it very specifically to the idea that the purpose of symbolism is where we philosophically traditionally think of symbolism as generalization. And so it's natural for us to think about the acquisition of symbols as something that happens through induction. And that's a very traditional kind of philosophical. And then you extend that to the notion that your sensory perception and your empirical experience, if you will, reinforces through repetition a certain thing, and that that's how you do it.

What he's saying is. No, no, no, no, no. The purpose of symbolic representation is not to generalize a concept. It's to actually get somebody else to pay attention to something you want them to pay attention to, which theoretically you would not be able to do without a linguistic symbol. You would not be able to refer to anything or put their attention to anything outside of the immediate experience. So in essence, all you would be able to do if we didn't have symbolic representation with language would be to point to something in the immediate experience or drag somebody's.

Turn their head physically and force them to look at something. To me, that was the mind blowing moment of this, of this book. It's like, oh, my gosh. Like, I think that's a transformative and really interesting idea.

32:33 Wes: Yeah. I've tried to explain what I'm reading to various friends and they're always confused about it because they know that they've seen the pictures of monkeys using sign language and cooperating and doing all sorts of other stuff. And then I try to explain it and I never feel like I'm giving a satisfactory explanation. And it's always because I, I feel like I don't. Haven't really thought through the magic of what joint attention actually means. There's a video, I sent this to you guys, of one of Tomasello's lectures where, you know, he has some, some of his own video footage of some of his experiments.

And one of them involves sitting with a young infant who's reading a book. And you're both paying attention to the book and the infant is there, at least eight months old, maybe a little older, will be aware that they're sharing this object with you and will look up back at you to see that you're looking at the book and to share your reaction, to say, oh, isn't this cool? Isn't this the third thing outside, outside of us that we're sharing right now? And you can never get another animal to do that. Right. You can't do it with a chimp. You can't do it with any other animal.

If any animal showed you something, glanced at something as like, you know, pointed to something in a referential way, your mind would be blown. They wouldn't have to start talking at you. Wouldn't have to be one of those animated movies where the animals can talk or something like that, a single referential act that involved some sort of human like movement of their gaze that was referential or something like that. That would be like, holy crap.

34:04 Seth: I think that's why the Planet of the Apes movies are so unsettling.

34:08 Chris: Because they're sharing joint attention with us.

34:09 Seth: Yeah, because they're sharing joint attention.

34:11 Dylan: Dr. Zaius that's how it's done in animated movies. Right. Is that kind of those kinds of gestures going on? I wanted to talk through an example that isn't given but would seem to. Have you ever seen how retriever dogs work? Okay. And on the face of it, I'm just trying to think it through. You are giving signals to the dog and you throw up a giant decoy or you shoot a bird out of the sky. You can train the dog, the dog will go get it and bring the bird back to you. Right. And so why is it that that is not us both jointly understanding it? There is a third thing there. Right. But I guess the critical piece is that the dog is paying attention to what you're telling it to do. You're saying, go get the bird.

35:00 Wes: It does look much more like the Skinner account, I think ultimately of an association between certain signs and certain behaviors and conditioned responses.

35:09 Dylan: Yeah. What's confusing about it, right. Is if you watch it, right, the dog will be clearly aware of contextually of hearing some or seeing some kind of signal or wait for a signal. Right. And then it will go and act on that signal and will make decisions, what are clearly decisions about getting from one place to another and choices about that.

35:32 Chris: They also sort of point, right. With their sort of gestures. Yeah. So you have this kind of gestural situation. Dogs are a very interesting example for all this.

35:39 Wes: Yeah.

35:39 Chris: Because they do seem to, I think, of any animal, they watch our face more than something like primates and things like that. And so that could have a lot to do with their evolutionary history and us breeding them.

35:48 Dylan: They bind with us. I mean, they're evolutionary. I mean, dogs have co evolved with human beings as dogs as opposed to wolves and stuff like that. They form elements of our pack. So their social interactions with us are very much like we are another entity within their cultural circle.

36:05 Seth: That's an interesting point because you're talking, Dylan, about one of the points that Tomasello makes is, you know, he's very clear about saying we are so closely related to chimpanzees, like 90, whatever, nine point whatever percent of like. And so there has to be some biological, some evolutionary explanation for why we do language and why we have this ability to direct attention and chimpanzees don't. And in the videos that Wes sent, you know, a lot of them, a lot of the Examples are showing 9 month old children and 9 month old chimpanzees or 12 month old children and 12 month old chimpanzees side by side.

Doing the same tasks, showing how they function differently in either social settings with adults or with other children, other infant chimpanzees. And the dog example is interesting precisely for the reason that you mentioned, is that they have evolved in a social construct. So dogs definitely do try to get your attention in certain contexts. Right. And there are ways in which you try to get a dog's attention. So the question of, have dogs somehow developed some kind of abilities to direct attention using linguistic symbols to human beings? And I think the answer there would be no.

But it is an interesting corner case in that respect. But for Tomasello's point, he's trying to argue specifically about talking about the difference between chimpanzees and human beings.

37:33 Wes: If a dog, you know, if you're ever taking a walk with your dog, and the dog gives a turn of its head, like a nod of its head towards something that it wants you to look at, then you could say your dog is a symbolic species, right? Dogs want your attention, can get your attention. I mean, it's. Chimps can do that too, but it's not about. And they understand you're inanimate being. They understand you're a being with intentions, but they don't. And the same thing with chimps, but it's about shared. It's actually about this ability to share intention for this to be this third referential object out there.

38:05 Mark: So we're wondering, though, can they have concepts even if they're not linguistic concepts? And I think Tomasello is open to the idea that, you know, we're trying to communicate things before we have explicit words for it. In fact, it's the whole expression, like Wittgenstein says in the Tractatus, it's the sentence that's the basic unit. It's not the individual word. So whether you're using a gesture, whether using. You're trying to get something done or express something, I'm wondering, does a dog have a concept of its bowl? Or, of course, the family member who's not there right now, clearly, you know, we have a training thing of like, where's mommy?

And then, you know, if the dog is actually doing it, it's supposed to, like, look around the house. And this could be just a stimulus response thing. But, like, it also works if Mommy's not in the house at all. Yes, you know that. Clearly they know that that word goes with that person. And can they express that thing to you? Well, not exactly. You know, I would wonder if a dog there, probably somebody has a record of some. You say, where's mommy, maybe the dog runs and brings mommy's hat to you or whatever. Like is that then they recognize that this smell, you know, that you are talking about with them, the joint absent object of your attention.

39:20 Dylan: To me, this sits right at the place where just being cognitively rich isn't the only requirement for language. So to me, the, you know, dogs are an example. But plenty of other species demonstrate cognitively rich interactions with the world and experiences where they have memory. It's not just extinction. They have memory, they have concepts. They have a cognitively rich experience.

39:47 Mark: If they have concepts, do they have symbols? No, Dylan says no. I mean, Wes says no. Why?

39:52 Dylan: And they can figure things out. But that doesn't require symbols, right? It just requires concepts.

39:57 Mark: I mean, concepts are symbols.

40:00 Wes: What concepts? No, not necessarily. They're at bottom, procedural. There are exactly, you know, thinking of meaning as use.

40:08 Dylan: Right.

40:08 Wes: A dog knows the difference between its bowl and a chair. It doesn't have words for that. And it's not as abstract. And its behavior will illustrate that difference, you know, and it's not object specific, obviously. Right. If you change the bowl out, it's not going to be confused.

40:23 Dylan: And dogs figure things out. Like they'll go and not having seen an activity performed, they'll figure out how to get under the fence. Or chimpanzees. I was reading in a book by Franz de Waal how you present them with a puzzle that they've never seen before. There's food hanging and there's just artifacts around. They'll gather up those artifacts and stack them on top of one another or use them in unique ways to try to get access to something. Right. So exercises of memory, exercises of understanding, use and function. Those don't seem like they're the same thing as symbols, even if they're concepts. And maybe we're getting into a fine

41:08 Mark: distinction, but it's an important distinction because, so Pinker, at least, and this is the view that I'm looking at Chomsky through, would consider that notion that the dog has of mommy or the chair or the bowl or whatever as a symbol. It is not a symbol that it is consciously toying with in the way

41:24 Chris: that we do, but it's a.

41:26 Mark: It's a symbol that is manipulated. And so you could talk about just like a computer that a computer doesn't understand, but it uses symbols to manipulate. And so by your procedural, you're saying, Wes, you know, that part of what the procedures are, are not things necessarily just baked into the muscles of, I don't know, what exactly makes us able to pedal a bicycle successfully. But it might even be that there are things in the brain that we could call symbols. Maybe there are groups of neurons, however they're instantiated. But we, you know, then start this cognitive science project of figuring out what the chains of symbols, how the logic actually works.

And this is the whole paradigm that Chomsky helped to kick off in terms of trying to figure out. And my. One of my ongoing questions here, I know we're reaching the end of part one here, so I just wanted to throw this out here, is whether Tomasello is just providing a different way into linguistics, but you could still do all the cognitive science stuff, or whether he's rejecting that wholesale. To say that this sort of processing model is much too, a priori is much too. I don't care how humans actually do it. I just want to figure out schematically what the procedures are for getting from A to B.

You know, just like if you're trying to program a computer to understand, understanding quotes, language to give the proper output and that that's not how humans do it.

42:44 Wes: Just want to say this something about concepts versus symbols. I say concepts are procedural. Again, it's about use and it's about how we navigate the world, right? Because we can cut up the world into bits and pieces and categorize things and do make causal inferences and all that stuff. And then we do it with language as well. You know, one of the things we learned in Wittgenstein is that to have a concept or something is not just to have this image or something like that in our heads. It's about what we can do with it linguistically. There's a certain know how involved, and all of that is ingrained at a procedural level.

It's like knowing how to play tennis or something like that. Even before you get to the inability of an animal to actually use and symbols and to, you know, manipulate them in larger constructs. There are certain criteria for being symbolic that go above and beyond the having of concepts, in my opinion. And so. And in Tomasello, it seems like. So if you look at page 12, you know, he gives basically a list of criteria for human linguistic symbols. So, right. They're socially learned, they're intersubjective, which I think is another important aspect of this, the referential.

So the intersubjective part, which I think is really important, is that you. Maybe we'll get into this in the next part. But basically everyone understands that they're both a producer and a comprehender. They know that their interlocutors share the convention. And there is an other for you that you're keeping in mind. You're not just producing a signaling behavior that's meant to get someone to do something else. You're also inside the perspective of another

44:21 Seth: when you use a symbol, not an abstract other. Right. One of the other criteria later on, as he says, it's perspectival that the use is done in a context where you recognize there's another and you're both in a particular frame, so to speak, which he's going to talk about later.

44:36 Wes: Yeah, they have their perspective and you know, your perspective is different from theirs and you have to bridge that divide in order to communicate with them. You can't just assume they know everything you know, Although some people do assume that.

44:48 Chris: But that bears on the question about what, when we were talking about dogs and their joint intentional frame is. They're not participating in that intersubjectivity like that where they think that they're manipulating or bringing forth affecting our mental state or intentional state. They're just affecting our behavior or brain. So they're not an interlocutor in that way.

45:08 Seth: They're also not using the things indirectly in utterances referentially.

45:14 Mark: All right, well, if any dogs want to email us in response to that last point, then feel free. Pel or partial examination life.com we're going to wrap up part one here. If you are a Partial Examined Life Supporter, Part 2 will probably already be in your feed or if you support us through Apple, it'll be up next week. For everybody else, you know, you don't have to wait. Just become Partially Examined Life Support at partially examined life.com support. Thanks everybody. Talk to you soon.