Many people have argued that the evolution of the human language faculty cannot be explained by Darwinian natural selection. Chomsky and Gould have suggested that language may have evolved as the by‐product of selection for other abilities or as a consequence of as‐yet unknown laws of growth and form. Others have argued that a biological specialization for grammar is incompatible with every tenet of Darwinian theory ‐‐ that it shows no genetic variation, could not exist in any intermediate forms, confers no selective advantage, and would require more evolutionary time and genomic space than is available. We examine these arguments and show that they depend on inaccurate assumptions about biology or language or both. Evolutionary theory offers clear criteria for when a trait should be attributed to natural selection: complex design for some function, and the absence of alternative processes capable of explaining such complexity. Human language meets this criterion: grammar is a complex mechanism tailored to the transmission of propositional structures through a serial interface. Autonomous and arbitrary grammatical phenomena have been offered as counterexamples to the position that language is an adaptation, but this reasoning is unsound: communication protocols depend on arbitrary conventions that are adaptive as long as they are shared. Consequently, language acquisition in the child should systematically differ from language evolution in the species and attempts to analogize them are misleading. Reviewing other arguments and data, we conclude that there is every reason to believe that a specialization for grammar evolved by a conventional neo‐Darwinian process.
Tomasello, M., & Call, J. (2018). Thirty years of great ape gestures. Animal Cognition, 1-9.
Graham, Kirsty E; Catherine Hobaiter, James Ounsley, Takeshi Furuichi, Richard W. Byrne (2018) Bonobo and chimpanzee gestures overlap extensively in meaning. PLoS Biology
The main argument of this paper is a defense of the view that language, and more importantly the innate capacity of “generative grammar”, is indeed a product of evolution and natural selection like an eye is a product of evolution and natural selection. The argument is simple: “language shows signs of complex design for the communication of propositional structures, and the only explanation for the origin of organs with complex design is the process of natural selection”.
ReplyDeleteI found myself sensitive to the “intermediate steps” criticism section. Indeed, I think there is a legitimate question asked there, i.e. how can rules of symbols manipulation evolve if not in a “all-or-none” fashion (because symbols are themselves discrete, and not gradient like a leg or an eye could be). Pinker and Bloom offer an answer to that, stating basically that the rules of (universal) grammar are similar in their structure to rules of “other mental systems”.
Two comments from that answer:
1) Couldn’t the last sentence be used as an argument for viewing the mind as a “general thinking processer” rather than an amalgam of specialized sections (as proposed by the tenants of generative grammar claim)?
2) This put the burden back on the evolution of symbol-manipulation. I believe that Pinker and Bloom used the example of fruit flies having a leg where an antenna is to tell that mutation can account for some “abrupt” mutation that are kept by evolution. But whereas the antenna is not totally discontinuous from a leg, I am not sure to understand what a symbol-manipulation rule should mutate from. This seems to point back (again) to the symbol-grounding problem in the sense that it would be absurd for symbol-manipulation rules to evolve without the symbols it manipulates being connected to the environment the symbol-user interacts with.
Pinker & Bloom helped re-open interest in the evolution of language but they passed in silence over the only difficult problem in evolution: the evolution of Chomsky's Universal Grammar (UG). All the other features of language (vocabulary, ordinary grammar, even propositionality) are potentially explainable by adaptive scenarios, even if they are speculative ones. But the evolution of UG is another story, because of the "poverty of the stimulus" (a form of underdetermination):
DeleteUG is a complex and abstract set of rules. Children are never taught these rules, nor do they hear or speak UG violations (hence there is no error-correction through supervised learning -- nor can the UG rules be learned through unsupervised learning). Chomsky concluded that UG must therefore be innate because humans are born able to obey the rules of UG without learning them. This poverty of the stimulus looks to be as much of a problem for the evolution of UG as it is for the learning of UG.
But Pinker & Bloom pass over this problem in silence (even though Pinker had been the specialist in explaining to psychologists how and why UG is unlearnable based on what the child hears and says). P & B write only about the rest, which does not pose a problem in principle for evolution (even though we can only speculate on how language evolved).
UG is indeed an interesting topic. It is unlearnable to us today because of the "poverty of the stimulus", which leads to the conclusion that it must be innate to us, and also calls into question its development during evolution.
DeleteHowever, I'm still wondering whether there could be any chance that it was evolved in the same process which contributed to our sweet tooth. Though children today can never be taught with the UG rules and see UG violations and thus learn UG rules in any form of learning (supervised, unsupervised, or instruction), we can never be sure whether the poverty of the stimulus was historically a problem from the very beginning, when our ancestors first "invented" language. If they actually heard and spoke both UG-compliants and UG-violations, during which it turned out that UG-compliants apparently offered evolutionary advantages against UG-violations, they used less and less UG-violations gradually until it completely disappeared, and only usages well-formed by UG rules survived until today. This could give rise to the "poverty of the stimulus" situation we are facing right now. I don't know how to rule out this possibility since we can never know for sure how they have lived and what they have experienced in the past.
“ Second, it may be difficult to evolve a huge innate code.”
ReplyDeleteIn this part of the text, do the authors suggest that different languages exist because of lazy evolution? Evolution gave us the ability to make sounds and it ‘assumed’ that the same language would come about, or did it not matter if the same language came about as long as the sounds could be used in a sequence to fulfill their function?
This could be similar to imprinting where if a person is spoken to with a specific order of sounds that that individual can make, that individual will be able to pick up and to learn these sound sequences.
"In this part of the text, do the authors suggest that different languages exist because of lazy evolution?"
DeleteAccording to Pinker & Bloom, it is implausible that language (or rather sound-meaning-correspondences, as it is referred to as in the section of interest in the paper) is encoded in our genome. This makes sense, since evolving such a huge code would take far too much time, while the code itself (and with it our language ability) would be vulnerable to mutations. Rather, it is the *mechanism* for learning these sound-meaning pairs that seems to be innate. This ability enables us to learn any sound-meaning pair available in the speech we hear in our environment (i.e. “The genome can store the vocabulary in the environment”). I can imagine that this process of lazy evolution would have contributed to the development of different languages, given the diverse environments our ancestors lived in.
"This could be similar to imprinting where if a person is spoken to with a specific order of sounds that that individual can make, that individual will be able to pick up and to learn these sound sequences."
I think you’re referring to imitation or instruction here (imprinting is what we discussed in the context of ducklings that associate the first ‘shape’ they encounter in their environment with their mother – this is a prime example of lazy evolution).
Yes, the capacity to learn (whether by imitation, unsupervised or supervised induction, or instruction) the grammatical rules for producing the infinity of possible grammatically correct utterances) is an example of the laziness of evolution, just as all learning (whether of how to make all the legal moves of checkers or chess, or how to walk or dance, or to categorize or name things) are also examples of the laziness of evolution.
DeleteThe only tricky case is the one P & B do not discuss, which is the rules of Universal Grammar (UG), which are unlearnable by imitation, nor by unsupervised or supervised induction, and are not taught by instruction (because they weren't even known before Chomsky). UG is unlearnable because of the "poverty of the stimulus": children never produce or hear UG mistakes, hence they never get corrected. (To learn rules, you need positive and negative data: examples of what follows the rules and what violates the rules. Children only have positive data. This is the form of underdetermination Chomsky called "the poverty of the stimulus.")
Ordinary Grammar (OG) can be and is learned by imitation, supervised learning, and instruction. UG cannot not. Yet children speak UG-compliantly.
The capacity to learn OG is ordinary "lazy evolution," coding the capacity to learn rather than the things that are learned with it. UG must be an example too, but in that case the rules seem to have been coded into our genes and brains, rather than just the capacity to learn them. And the poverty of the stimulus is a problem for evolution too (because evolution too is a form of trial-and-error learning, supervised by the survival/reproduction consequences of "evolving the wrong code.")
Would universal grammar have come about by evolution then? If it cannot be learned and there is no positive or negative feedback that can be received, could it have come about by evolution? And I am a little unsure of how OG and UG differ? Is the difference simply that OG can be learned but UG cannot?
DeleteAsk me this question in class. Yes, UG would have to have evolved. But there is a problem with that too...
DeleteThe grammar you know is OG. You don't know UG unless you studied linguistics, and no one knew about it till Chomsky -- yet every speaker of every language obeys its rules (without knowing them), and they could not have been learned (because there is only positive evidence).
I have a question that is at the same time related and not related to the question of the evolution of language. I am wondering about the evolution of communication. I am not sure where to post, so I am posting here.
DeleteIndeed, communication is a non-individual skill and capacity. How is communication evolving among species? My reasoning is that you need a lot of apparatus to communicate (the ability to mind-read, the ability to know that other members have intentions, etc.). But doesn’t the evolution of communication apparatus need to be “synchronized” among different members of a species? By that I mean that not only I need to evolve it myself, but if I am the only member of a species that is able to communicate, well it is not useful in any circumstances since other members are not able to either understand or “answer” me (acting in a way that is influence by the communication I provided).
Is Baldwinian evolution enough to explain the evolution of communication? That if the capacity to learn to mind-read other members of our species is general enough, as well as the capacity to “express” ourselves through various means, then two or more members should be able to learn how to communicate together? But what about hard-scripted innate signals about other? How can they evolve if every member of a species must have the same signals?
“A major problem among even the more responsible attempts to speculate about the origins of language has been that they ignore the wealth of specific knowledge about the structure of grammar discovered during the past 30 years. As a result language competence has been equated with cognitive development, leading to confusions between the evolution of language and the evolution of thought, or has been expediently equated with activities that leave tangible remnants, such as tool manufacture, art, and conquest.”
ReplyDeleteIn this piece, Pinker & Bloom (1990) argue that language is a natural evolutionary development, like the echolocation of bats or stereopsis. The authors support the Chomskian nativist theory of language -- that humans are born with the innate capacity for language -- but elaborate in that this capacity is a result of genetic adaptation rather than other natural-physical forces (the non selectionist view). This view is supported by seemingly conventional evidence, through examining the canonical work of giants Chomsky and Gould, and finding flaws in their arguments in the face of widespread public/academic reception of their arguments.
Pinker has popularized his belief about the natural selection of language, notably through his many public appearances and incredibly popular books -- he is known for his staunch evolutionary perspectives and his inquiry on human nature, many of which, I do not agree with. However, this final paragraph, which warns against the conflation of language development and the evolution of cognition, I find particularly interesting. It is true that there exists a species without human language capacity that are capable of “tangible remnants” (most notably, tool use) -- leaving cognition as a non-evolutionary development implies the authors’ inability to provide their characteristically selectionist views on the “easy problem” of cognitive science while also implying that language is separate from the “easy (and hard) problem.”
What would it be to not equate evolution of language with the evolution of thought? And if a non selectionist view of language is as absurd as suggested in this paper, why is it possible to separate cognition from this argument?
Hi Gili!
DeleteI thought about how they interpret the relationship between language and thoughts also because, in part 3.1, Pinker claims:
“The vocal-auditory channel has some desirable features as a medium of communication: it has a high bandwidth, its intensity can be modulated to conceal the speaker or to cover large distances, and it does not require light, proximity, a face-to-face orientation, or tying up the hands. However it is essentially a serial interface, lacking the full two-dimensionality needed to convey graph or tree structures and typographical devices such as fonts, subscripts, and brackets.”
I realize that here, Pinker discusses a vocal-auditory channel, a subfield of language (not language as a whole); however, I think this gives insights about what it would mean not to equate the evolution of language with the evolution of thought.
Language, even though it has an atomic power. Any thought that comes to mind can be expressed in a language. Also, any expression of language X can be expressed in language Y even if you must use more words, as we discussed in the class. However, in their example, they talk about how the vocal-language is not the most straight forward way of expressing a given thought. “Euclidean relations: a picture is worth a thousand words.” You can think of a graph with Euclidean relations at the snap of a finger. It might be a process that language plays a small role, but seemingly it cannot be the only role player. Thinking about this graph was not equivalent to the production of the thousand words that described this picture.
This kind of analysis of how the mind works reminds me of the homunculus. Consider the mind to be equivalent to what homunculus does, (I realize that this wasn’t the way we discussed the homunculus, but bear with it just for the example). Homunculus, from all these examples, has the language as one of its tools of perception, calculation, and production. But, for sure, homunculus can do more than what language can do, in a short time, it can think of a graph even though it cannot express it in language instantly.
I think it follows from the above example that it is natural to not equate the evolution of language from the evolution of thought. Thoughts can be generated independent of language, and for instance, computations of a language are not always carried to thoughts. However, it is obvious that they interact with each other. More complex thinking of graphs can lead to the generation of structures in the language that allows for efficient/clear expressions. And, computations of language can yield to changes in thoughts.
In the end, though not equating the evolution of language and evolution of thought doesn’t tell us anything about how does possessing a language makes us different than the species that doesn’t possess a language, also with regards to how the thoughts are for formed.
Gili, Pinker & Bloom are certainly right that some of the capacities underlying language are learned and some are inborn (including the capacity to learn). That's ordinary "lazy evolution," not coding more than necessary and offloading the rest on the environment and learning.
DeleteBut it is puzzling that Pinker, of all people (the "messenger" who explained the unlearnability of some parts of grammar to skeptical -- but ignorant -- psychologists and philosophers) glosses over the only problematic case for both learning and evolution: Universal Grammar (UG), with its "poverty of the stimulus," because the child does not hear or make UG mistakes. Evolution would have the same problem too, unless UG is innate not only to the brain but to the Big Bang at the origin of the universe.
What does "non-selectionist" evolution mean? (Explain it in relation to nonsupervised and supervised learning and the selective effect of survival/reproductive success or failure.)
Pinker's "evolutionism" dates from this article, but he is not particularly grounded in evolutionary biology. (And his notion that we're becoming "better" because we kill a decreasing proportion of the human population than we used to is a bit deceptive: we may be killing a decreasing proportion of people but an increasing number of them -- and that's because we are increasing our reproducing faster than we are increasing our killing. [And that's just for killing people: with animals we are increasing both the proportion and the number that we are killing; the numbers (and means) are monstrous; and for an increasing number of species the proportion is 100%.])
(Chomsky thinks that the innateness of UG is not so much a constraint on language but on thought itself: Non-UG "thoughts" are unspeakable because they are unthinkable. Yet he thinks this constraint is not a formal logical one.)
Onur, yes, a picture is worth a thousand words, but a picture is not a proposition. It does not "say" anything. Some thinking and perceiving may be parallel. but its transmission is not. The visual modality is more parallel (in at least two spatial dimensions, at the fovea) whereas hearing is more serial (not entirely: consider polyphony and harmony, for example). But there is no doubt that speaking is better than gesturing (slower, invisible in the dark or at a distance; interferes with other uses of hands, etc.). Gesture (pantomime) is better only as a starting point, to allow iconicity to link gestures with their "referents." But once the proposition is born, formed of category names whose iconicity is now irrelevant, it's best to migrate to the oral modality. And our parallel processing capacities are limited in every modality -- sensory or mental.
By the way, writing, whether iconographic or alphabetic, is even slower than gesture.
We have discussed that the way we connect symbols to referents is arbitrary, just as Pinker and Bloom stated: “there is no reason to call a dog dog rather than cat except for the fact that everyone else is doing it.” This is the idea that the “coding protocol” can be arbitrary as long as it is shared. I agree with this point that they have made.
ReplyDeleteHowever, they add on to say that “But given that language must do one or the other, it is highly adaptive for each member of a community of speakers to be forced to learn to do it the same way as all the other members.” I am hesitant to agree with this, because I don’t think there is anything evolutionary about the capacity of language being passed on by generations through learning. Rather, I would argue that it is simply the capacity to learn as well as the feeling of understanding that allows for the shared “coding protocol”. As humans are born with the capacity to learn any language, it makes sense that we are able connect symbols to their referents through trial and error in the language that we “enculturate” to. But, I don’t really see how evolution explains how/why meaning = grounding + feeling.
What is arbitrary about symbols is their shape: they do not resemble their referents. How we connect them to their referents is through sensorimotor feature-detection. That's not arbitrary. though it's only approximate.
DeleteEvolution is lazy. It codes neither names nor conventions. But the adaptive value of language certainly depends on calling things by the same name rather than a different one by each speaker...
No explanation -- evolutionary or otherwise -- explains how and why organisms feel rather than just do.
I agree with Erica here! Why would language be adaptive instead of only learnable? Children are exposed to a certain environment since birth and imitate what they observe, hear , sense. Wouldn't this be enough to explain why children learn languages?
DeleteHi Andrea,
DeleteAll language cannot be learned. We know this is true through looking at Universal Grammar - a set of rules innate to all humans. Young children are not taught these rules through either supervised or unsupervised learning. Poverty of stimulus explains this; there is an absence of negative evidence for UG rules.
Hey William,
DeleteHere, I wasn't referring to UG but to OG rules that are learned by supervised and unsupervised learning. UG is innate. Here, like Erika, I don't get why the ability to learn OG would be transmitted in an evolutionary manner. I guess our confusion comes from Pinker's and Bloom's one and the fact that they completely mixed up OG and UG and gave OG examples when talking about UG.
The main goal of this paper was to argue that human language evolved by natural selection. The authors explain in depth the concepts of language design, language diversity, arbitrariness in language, cultural implications of language, and “lazy” evolution. They argue that there are aspects of language that are innate and aspects of language that are learned. Aspects of language that are innate include universal grammar, which the authors only touch on in a few paragraphs are the end of the article. Aspects of language that are learned include vocabulary, such as why one calls a dog “dog” and not “cat”.
ReplyDeleteOne of the main questions I would have like to see the authors address is: what makes language different from other forms of communication? I don’t believe the authors even attempt to explain this question, other than state that language can express propositions while other forms of animal communication cannot. Is this really the only difference between the two forms of communication? And if so, why hasn’t language developed in non-human animals if there is only a slight difference between their current communication and human language?
Hi Taylor,
DeleteI was reading the section Phyletic Continuity (5.4) of the paper and it got me thinking about your second question which concerns why language hasn't yet developed in non-human animals despite their communicative nature. Pinker and Bloom (P & B) are clear to point out that Homo Sapiens were definitely not the firsts to develop language, Australopithecines (a primate ancestor) for example, were the "first talkers". Although being primates and having 99% the same DNA as humans, chimpanzees do not have developed language. It is suggested by P & B that the genes that encode for universal grammar, are found within in that 1% of DNA that differs from humans because chimpanzees have one different base in their genetic code. A single base difference has a ripple effect on all of the subsequent pairs it forms, and thus the genes. From this conclusion, I think the authors would argue that other species likely don’t have language capacity because they also are lacking the correct bases that code for universal grammar.
Couldn't we argue that animals can make propositional statements though? Isn't a warning call for a certain predator a propositional statement "X is here" - the "is here" part is propositional, isn't it?
DeleteTaylor, It is not just that language can express propositions, but that if you can express any proposition, you can express any proposition. No other communication code has that power.
DeleteWhy didn't other species evolve language? Let's discuss that in class. The first question is how and why did our species evolve language...
Akhila, it's not known whether Australopothicines had language; besides, they were almost identical to us and interbred with us.
Pinker's speculations about the 1% difference are completely vacuous: I hope by now most of you will have detected that...
Erin, we can interpret a warning call as if it were a proposition (in fact, with our language-prepared brain, we can hardly stop ourselves from doing it), but the proposition is in our heads, not theirs (any more than our dog is saying "I want to go outside" when he scrapes at the door. He just wants you to open the door. Nor is a green light saying "You may go now." (Not even the sign that says "Vote for Trump" is making a proposition: it is the one who wrote that sign who is making the proposition.)
I’d like to try and attempt to answer Taylor’s initial question with an additional detail. I think the authors implicitly state that both its propositional component and its ability to provide learning through instruction (without stimulus involvement) are the two things that make language different from other means of communication that we might see in other species.
DeleteThe propositional nature of language is already discussed a bit in these comments, but the ability to learn through instruction seems to me (and correct me if I’m wrong) as an important by-product to the propositional nature of language. The ability to make statements about the state of the world (ex. “there is a tiger near that tree”) intuitively feels like the basis for instruction-based learning. Gestural communication doesn’t feel adequate for intentional instruction-based learning specifically for the reason mentioned above: gestural communication allows for interpretation, but not the communication of intentional materials (in that “primitive” world, screaming in response to a tiger isn’t you making a conscious effort to make people learn how to behave in a tiger attack).
I also remember reading somewhere (or maybe discussing in class) the idea that the dawn of propositional language might have led to cognitive differences in the way we interpret “information” we acquire. In that sense, maybe hearing your pal scream and thinking it’s because a tiger is nearby is different from your pal telling you “a tiger is nearby” not just in the conclusions you can logically make about what your friend’s intention is, but also in the way you use and conceptualize the information you either deduced or learned from your friend?
Instruction is transmitting propositions, just as all verbal (as opposed to just vocal or gestural) communication is. Presumably being able to produce propositions comes with the ability to understand propositions.
DeleteBut I'm not sure what you mean by "interpret": If I imitate an animal about to attack, presumably the similarity takes care of itself; I don't need a further interpretive step "Oh, that's an imitation of an animal about to attack." That would be to interpret an action propositionally.
In this paper, Pinker and Bloom discuss the evolution of human language as parallel to the development of biological systems -- both are driven by natural selection. They support this assertion by providing a myriad of examples which ultimately point back to one central concept: the complexity of human communication and the structures that support its complex design can only be explained through natural selection. Language, like an organ, evolves and adapts according to the species' needs.
ReplyDeleteI found many of their points quite interesting as they presented an opportunity to appreciate the incredible capacities and diversity of language. The section on inherent tradeoffs, in which the authors analyze the conflict of interest between speaker and hearer, reminded me about our conversation last class when we questioned differences in male and female reproductive capacities and goals; if a man can and (theoretically) should inseminate an infinite number of females, why doesn't he? Within individual languages, there is no reason to name a referent by a consistent name except for the reason that everyone else uses the same name. Pinker and Bloom pose that "any shared system of communication is going to have to adopt a code that is a compromise among these demands" which I believe may also relate to our reproductive behaviour. What underlies this seemingly innate capacity to collaborate?
I do not believe that Pinker and Bloom addressed their perspectives on the exact mechanism of natural selection, but I would like to posit that they are not arguing for a "survival of the fittest" paradigm. Considering the aforementioned point, as well as their evaluation of language's social value, Pinker and Bloom's position seems to align with a cooperative, rather than a competitive, form of evolution.
Just as a reply to the following: "I do not believe that Pinker and Bloom addressed their perspectives on the exact mechanism of natural selection, but I would like to posit that they are not arguing for a "survival of the fittest" paradigm. Considering the aforementioned point, as well as their evaluation of language's social value, Pinker and Bloom's position seems to align with a cooperative, rather than a competitive, form of evolution."
DeleteIt doesn't seem to me that competition as an evolutionary strategy precludes cooperation / that they are part of different paradigms. Cooperation within an in-group lends a competitive advantage against non-members, and cooperation as a species yields a competitive advantage over other species. I didn't get the impression that Pinker and Bloom are arguing for language as a product of natural selection through a different paradigm of evolution.
"Group selection" is a controversial notion in evolutionary biology because groups do not have genes -- individuals do. But if individual cooperation within a group favors individuals' genes (more than selfish non-cooperation), it can evolve in individuals.
DeleteLike we talked in our class, there is a way that competition and cooperation can work together. Altruism can exist in the world of evolution under two different theories that have been proposed. The first is the theory of Kin where shared genes are worth protecting, and therefore protecting those that share your genes is beneficial. The theory of reciprocal altruism is another theory where you benefit those that may benefit you in the future. This leads to rules that you see in the animal kingdom such as chimpanzees beating male chimps that don't play by the rules. These rules show that cooperations and competition can survive together. However, there is controversy in these theories as Harnad pointed out that in this case, there is no cheater test. Unlike how peacock tails have a cheating measure in place where if you are not fit, and you have a tail, you will most likely not survive, the altruism theories do not.
DeletePinker & Bloom argue that human language evolved like any other complex biological system in an organism: through natural selection. As they mention, much of the theories revolving around the evolution of language are merely speculations and not very empirical. It's a difficult thing to study empirically. It's clear that natural selection did play a role in the way our human brains have evolved, since it is an organ.
ReplyDeletePinker & Bloom state, "{Chomsky & Gould] have repeatedly suggested that language may not be the product of natural selection, but a side effect of other evolutionary forces such as an increase in overall brain size and constraints of as-yet unknown laws of structure and growth"
But why do increase in brain size & natural selection have to be mutually exclusive here? We already had some of the hardware to be able to use language, but perhaps it wasn't until we developed a symbol system that we began to use these parts of our brains for language and because of that, over time, we got better at using those parts. Pinker & Bloom state, "…natural selection is the only scientific explanation of adaptive complexity. "Adaptive complexity" describes any system composed of many interacting parts where the details of the parts' structure and arrangement suggest design to fulfill some function." I definitely agree with this.
We needed sensorimotor capabilities & object recognition to ground symbols for language. As they discussed (and we discussed in lecture), evolution couldn't have coded for every aspect of language, but it did code for the mechanisms and capacity to learn. As stated in other comments above, it is interesting they avoided addressing the case of Universal Grammar, they just mentioned it very briefly.
Most of P & B's remarks about evolution and about language evolution are banal platitudes for those who already understand biology. And, as you mention, the only potentially interesting part (the evolution of UG). they have nothing to say about. (Oddly enough, Pinker became a zealous (pop) "evolutionary psychologist" after this article, but had less and less of substance to say about language.)
DeleteWhen talking about the evolutionary sequence of universal grammar being possible, in 5.2.1 Nonshared Innovations, Pinker and Bloom notes that “One possible answer is that any such mutation is likely to be shared by individuals who are genetically-related. Since much communication is among kin, a linguistic mutant will be understood by some of his or her relatives and the resulting enhancements in information sharing will benefit each one of them relative to others who are not related.”
ReplyDeleteHere I am not sure how this mutant could be shared by individuals that are genetically related. How could this be achieved? Are they suggesting that the person and their relatives undergoing the mutation altogether at some point in their life? I don’t know if that sounds like a plausible explanation because I feel it has already been a pretty small likelihood for an individual to undergo mutation and it would be so much harder for the mutation to happen on the person and his relatives. Or do they mean that the person mutate first and pass on his gene to the upcoming generation? In that case, it would not be first and foremost an advantage to pass on, rather function more like a disadvantage. Since before he has kid, this grammar is not shared between him and his kin.
Yes, this mutant discussion is muddled. I think he means a mutation shared by kin might favor family communication, which is adaptive for the individuals who share the gene. But it's very vague about whether it is about UG. OG of course has no language-specific genetic basis other than general learning capacity.
DeleteI was also confused about the even more general answer they provided for the evolutionary sequence of language right after the quote pulled by Mingjuan:
Delete"But we think there is a more general answer. Comprehension abilities do not have to be in perfect synchrony with production abilities. Comprehension can use cognitive heuristics based on probable events to decode word sequences even in the absence of grammatical knowledge [...] When some individuals are making important distinctions that can be decoded with cognitive effort, it could set up a pressure for the evolution of neural mechanisms that would make this decoding process become increasingly automatic, unconscious, and undistracted by irrelevant aspects of world knowledge."
Are they saying that mutations for the production may have occurred first and then mutations for understanding effectively occurred in response? When they say "set up a pressure for the evolution of neural mechanisms" it is all so vague and lacking any situational grounding.
I think this passage about mutation would mean that the mutation is passed on generation by generation through genes - with every offspring of an individual inheriting the mutation that allows for understanding of such grammatical system. "Since much communication is among kin, a linguistic mutant will be understood by some of his or her relatives and the resulting enhancements in information sharing will benefit each one of them relative to others who are not related." I think it is meant that the mutation would favour family communication. This means that it would only be individuals who are part of that familial system who would understand the linguistic mutant. Additionally, the fact that they are able to understand benefits them in a ay that will favour them and their genes over individuals who are outside of that familial system and therefore, cannot understand the linguistic mutant. Perhaps this is the way that such a linguistic mutant was passed on and continued.
Delete[I]
ReplyDeleteI still remember the first day I learned about UG and how thrilled I felt to be born to be connected with all humans in this very special way, but then the more linguistics/psychology courses I take, the more doubts I have on “UG” – What really is UG? Why do I not feel the “U” when I learn about UG? Why are people claiming such phenomenon to be universal yet it is not? And later, it develops into “If what I am learning about really is ‘UG’, then I do not want to commit myself to this idea.” However, this thought confuses me a lot because I know there is something universal about language.
After reading Natural language and natural selection and attending the lecture, some of my doubts are clarified (especially on what is not UG), yet, still there are things that do not yet make sense to me. For example, if UG is unlearnable, parameter is learnable, yet (it seems that) “UG in different languages” (I put quotation mark for this phrase because I am not sure if I really understand what that means) varies in terms of parameter settings, then the parameters would have been there, born with UG, then how is “parameter” learnable? It seems that it is suggesting that when a child is born, UG is there, and a set of parameters is there, too, and to learn a language it is about learning the parameter setting in the language – not the parameter. However, does it really make sense, for example, for a Chinese monolingual baby to have a parameter for inflection on verb to indicate tense when, in Chinese, tense is only expressed by adverb? Or, does it really make sense for a kid to born with all the possible rules that would happen in a language and just “wait” for them to be “set”? I think it makes more sense if the parameters can be learned, but if parameter can be learned, what role does parameter play in UG?
You don't learn UG, but you do learn that in your language pronouns are expressed as separate words accomanying verbs, or instead as inflections (endings) on verbs ("I go" vs "go-I"): the "pro-drop" parameter. UG comes in several different forms, including a pro-drop and a non-prodrop form. It's a bit like an alternative notation for the same algorithm: Standard addition (1 + 2) or 'reverse polish notation" (1, 2 +). Mathematics is universal, but it looks different in standard vs reverse polish notation.
Delete[II]
ReplyDeleteAlso, what exactly is “UG for different language”? Is that just OG? (I think it is suggested that there is a difference but I am not sure.) It occurs to me that the some UG constrains that one has in the “UG for one language” are more like a constrain from parity, convention, conformity.
But even with all these doubts, I still believe there are constraints in all languages and there is UG. One statement from yesterday’s lecture that made most sense to me is that “UG is innate to the brain; it is not a constrain on language, but a constrain on thought.” This is what I have come to conclude as well after all. Language conveys thinking; if something is not thinkable, it will not be expressed in language. In the lecture, Professor Harnad mentioned the poem that made the famous sentence “colourless green ideas sleep furiously” sensible. But without the context, this sentence would still not be said because it is an “unthinkable thought”. (It does not explicitly violate any syntactic rule...But we also said that in terms of using a language to construct well-formed sentences, the symbol manipulation has to be about its meaning.) Another example, in the paper, a constraint that says you will never hear a kid put adjectives or determiners before pronouns and proper names – but why would a kid use adjectives or determiners to modify a proper noun in a normal setting? However, I would imagine if I show a picture of me to a kid and tell her “This is Xinyi”. Then, I show 20 pictures of me laughing and 20 pictures of me crying, I’d say if the kid captures this difference, “happy Xinyi” and “sad Xinyi” should be elicited. Same thing goes to “big” and “small”, yet it would be more absurd for a kid to see me being weirdly big and weirdly small because a proper noun like a name that refers to a person just does not usually exist in such states. But then, one would (I would) argue that “now you are treating Xinyi more like a common noun,” and yes, but otherwise, in what context would I need to modify or make clear “it is this [proper noun A] but not that [proper noun A]”?
I think now the question that I have is still to what extent a phenomenon would be considered as a UG phenomenon. And, I know it has been said that "it is not about logic", but the more I think about it, the more it is about "logic."
(it is not just about logic, maybe? I know people can still say illogical things... But just like the proper noun example I give, it would not "make sense" to put a determiner or an adjective before a proper noun, unless it makes sense when certain context is given.)
DeleteEvery language has an (unlearnable) UG as well as a (learnable) OG. Only UG is universal; OG varies from language to language. UG has parameters (see above) which are learnable and learned. Each parameter gives UG of a different "flavor." Countless languages are possible (because all of the OG possibilities), but the number of UG parameters is finite; each yields a different "flavor." Think of each parameter-setting as a (slightly) different formal type of language (or subtype of UG) notation. By the same token, languages can be gestural, oral or written. They may look different in form, but their UG structure is the same.
DeleteI think your proper-noun/common-noun distinction is conveyed by the determiner ("this is Xinyi" vs "this is a Xinyi"). You'll have to ask Xinyi or a linguist about how "null determiners" work. But there is no escaping technical analysis here: You can do a certain amount of UG-testing from the armchair intuitively (because you "know" UG innately), but why some things are UG-legal and UG-illegal cannot be settled by innate intuitions alone. (By the way, "colorless green ideas sleep furiously" is meaningful, just as "square circle" is. It's just logically contradictory. That's why we call it nonsense, not because it is just nonsense syllables like "squiggle squoggle." And even "color green idea sleep fury," cleaned of the cheating inflections and morphology, is not meaningless. But it's just a string of (meaningful) category-names, not a proposition (unless you cheat again to mark the verb)...)
Unless one has some technical knowledge of UG, one cannot intuit whether it is just logic. Till then, I'd accept the word of those who know -- that it's not.
Chomsky has often lamented that people who don't know UG technically but do know how to speak (at least one) language are confident in their assumptions (and criticisms) about UG whereas people who know how to count and add are not so confident about the higher maths that they do not know technically...
That said, the observation that "what is unthinkable is unsayable" is not very helpful when one is puzzling about the nature and adaptive function of UG.
Thank you for the clarification! I think I just have two more questions in terms of the relationship between UG, OG, and parameter: is the *setting* of parameters learnable (and learned) or the parameters themselves (or both)? Also, for “Each parameter gives UG of a different 'flavor'," are the “differently flavoured” UGs UG or OG? (You mentioned “subtype of UG” in the comment, so I guess they are UGs. Yet, what would be the difference between such UG and the corresponding OG?) (Or, in another word, is the universality of UG shown by the exact same abstract rules or the possibility of the existence of such rules?)
DeleteI think my proper-noun/common-noun distinction is not exactly the same as “This is Xinyi” vs “This is a Xinyi”. Just because Xinyi can be happy and sad does not mean that there could be multiple Xinyis (while if I accept “a Xinyi”, I would have to accept there could be Xinyis – well, there could be, but not in this sense. Even when Xinyi talks to herself, she is talking to “herself”, not another “Xinyi”). They are just some appropriate descriptions for the one Xinyi, which is why I would still argue that it is a proper noun. Just like how people sometimes say “poor you” – there is not a “poor you” that is distinctive from just “you”.
Although, in terms of the determiner, I do remember reading about an experiment on how children learn common and proper nouns. In this experiment, each child (2 yrs old or younger) was shown a pair of dolls. In the common noun condition, one of the dolls was introduced in the following way “Look what I’ve brought you. This is a zav”, and this doll was referred to as “the zav.” In the proper noun condition, the doll was introduced with “Look what I’ve brought you. This is Zav.” In both of the two groups, the second doll was never named and referred to as “the other one”. In the testing part of the experiment, the children were asked to feed the zav/Zav, dress the zav/Zav, etc., and it is how the children in different groups interact with the named dolls would indicate if they know the difference. The experiment showed that the distinction is made as young as 17 mos, but only for girls. This shows that they already have more structure than just NP; they must have DP as well, even though they are not producing determiners at that time yet. Why don’t children produce the category that are present in their grammars from the beginning? In this week's Pinker’s paper, he said “the child would set parameters on the basis of a few examples from the parental input, and the full complexity of a language will ensue when those parameterized rules interact with one another and with universal principles.” Using this experiment as an example, would the syntactic distinction between proper noun and common noun be a parameter or a universal principle?
Parameters vs. parameter settings if just terminological. UG is UG, whether it's made of wood, metal or faux leather "parameters." Arithmetic is arithmetic whether in standard notation 2+2 or Polish notation +(2 2).
DeleteOG would be different in an SOV and a VOS language, but OG differs from language to language anyway.
I am Stevan, but I am also one of many Stevans (much fewer than Stevens of Stephens), just as there are many New Yorks. Individuals vs kinds is not so much a linguistic matter as a metaphysical one (not relevant here). Even less relevant is the hard problem (of sentient individuals).
No surprise that young children can already distinguish individuals and kinds, and that determiners (a, the) can help.
“Third, arguments that language is designed for communication of propositional structures are far from logical truths. It is easy to formulate, and reject, specific alternatives. For example, occasionally it is suggested that language evolved as a medium of internal knowledge representation for use in the computations underlying reasoning. But while there may be a language-like representational medium -- "the language of thought," or "mentalese" (Fodor, 1975) -- it clearly cannot be English, Japanese, and so on.”
ReplyDeleteIn this paper, Pinker and Bloom argue that natural language evolved via natural selection, much like any other biological trait. In this passage, I don’t understand why they say that a language-like representational medium cannot be a natural language. I think they made this point in order to show that natural language must be designed for communication of propositional structures, as anything “language” with any other purpose is not a natural language. I’m still unclear as to why this is, though.
Also, I find it frustrating that they never address why it is that UG, specifically, became the internal “template” on which language would be formed. In fact, they barely address UG at all, which is interesting considering UG is central to the argument they claim to be making.
Thinking clearly isn't just propositional. Only language is propositional. There can be non-propositional thinking. (Only our language-biassed brains think thinking must be propositional!)
Delete"Representation" is a weasel-word. (P & B know no more than you or I do about the relation between language and thought.)
The theory that we may have already had words embedded in our genetic code seems to imply that categories are innate and symbols somehow come pre-grounded? How and why would we evolve to be equipped with knowledge of all the categories in advance, when we could instead learn them from direct, relevant experience and from others we interreact with via verbal definitions/explanations (based in concepts we learnt/understood already, i.e. grounded)? It makes more intuitive sense to me that categories first have to be learned via perception by interacting with them in the world and then labelling them with names.
ReplyDeleteLikewise, the possibility of having genetically encoded the meaning of every word used to refer to categories doesn’t account for the ability to name any new categories that we could encounter in the future.
It’s ridiculous to imagine that everyone on the planet would have or need to learn all the same categories and it would be miraculously serendipitous if everyone happened to label those categories with exactly the same name, given that names are arbitrary.
All of this pre-encoding also seems unlikely considering the “laziness” of evolution.
Lastly, the attempt at explaining the original adaptive value of language doesn’t seem well-reasoned because it begs the question: language would already have to be pervasive for “gifted orators” to have any persuasive powers among their peers and possess any evolutionary advantage.
I agree. It not only makes more sense intuitively that we would need to learn categories from direct experience, but it also makes no logical sense that we would come into the world with words already written in our genetic code. This couldn't make sense as there are things that come into existence during our lifetimes that require new names that couldn't possibly be accounted for by evolution. The only possible way we could have come into the world with innate categorical knowledge of all things would be if a Lamarckian theory of evolution were supported, whereby one could inherit acquired characteristics, including knowledge.
DeleteWhile I agree with you both, I think it's important to note that Lamarckian theory of evolution refers only to physical changes in organisms through their life time, but I see the point that you're trying to make. Technically, if language was encoded in our genes and passed down that would be physical on some level, but regardless I don't think that's what that theory refers to.
DeleteAlso, while I agree with both of you that it makes more sense that we learn categories, I don't totally agree with one thing that you said, Rachel. When you said "the possibility of having genetically encoded the meaning of every word used to refer to categories doesn’t account for the ability to name any new categories that we could encounter in the future," I disagree. Having categories and meanings of words genetically encoded would actually, I think, give us the ability to name new categories because we would still be able to alter what comes intuitively to us. My thinking behind this comes from me relating this to breathing. Breathing is something controlled by our brains that we naturally do, and it's encoded in us so that we can live. However, we're able to modify our breathing in new situations. For example, when we go swimming for the first time we don't automatically know not to breathe in, but we can modify what we know innately to the new situation or category. Although, I don't think that language is embedded in our genetic code, I think my example disrupts my opinion a little bit.
I agree with the points that Rachel stated above. I agree that it seems a little silly to believe that we already have words embedded in our genetic code. If this were true, as Rachel mentioned, it would seem implied that some categories are innate and that somehow some words would already be pre-grounded. This does not seem like a plausible thing considering the amount of time we spend teaching children which words to use when referring to specific referents. In addition, if we were born with words embedded in our genetic code without the pre-grounding, then this would be pointless. We all know that evolution is lazy, so why would it code something if it will not be useful?
DeleteWhen reading about the theory that words could already be embedded in our genetic code, I often think about how similar the human genome is to that of chimpanzees. Studies have shown that our closest living relatives share up to 96% of our DNA. If the genetic code of chimpanzees is so similar to ours, why don’t chimpanzees have language? I feel as though having words coded into our genetic code is not intuitive.
Rachel also touched upon categorization and learning new categories. I believe that even if we had all words already embedded in our genetic code, it would not be enough to be able to categorize. An important part about categorization and learning new categories is to learn the invariances of affordances. Affordances are features of objects that we learn by what you do with the object. To learn affordances, you need to interact with it using your sensorimotor capacities. The invariance is the stuff that is present in the all the members of a category and absent in all non-members of a category. These invariances are not only limited to one specific feature, they can be quite complex and consist of a list of features. Some category learning requires sensorimotor inputs! They cannot be known by simply having words pre-embedded in our genetic code.
"Also, contrary to what Lieberman implies, there does exist variation in grammatical ability. Within the range that we would call "normal" we all know some individuals who habitually use tangled syntax and others who speak with elegance, some who are linguistically creative and others who lean on clichés, some who are fastidious conformists and others who bend and stretch the language in various ways. At least some of this variation is probably related to the strength or accessibility of different grammatical subsystems, and at least some, we suspect, is genetic, the kind of thing that would be shared by identical twins reared apart."
ReplyDeleteI do not agree with the above suggestion that the proposed instances of unconventional sentence construction indicates that individuals are genetically predisposed to access certain grammatical subsystems over others. In that case, why is it that a proposition with tangled syntax and superfluous vocabulary can still be judged as being well-formed in accordance to the syntactic rules of UG? Rather, I believe that individual variations in how one produces propositions that they mention is OG that is learnt , corrected and acquired through the external input one experiences throughout their lifetime (especially during developmental years). I find it interesting that Pinker & Bloom do not differentiate between OG and UG in this passage or attempt to address how UG could have evolved at all (particularly since children do not receive negative examples). In my opinion this distinction seems more fitting to explain variations in language production as compared to their idea of genetically determined grammatical subsystems.
To my understanding, a proposition is categorization along with a truth value. For example, "the cat is on the mat." In this instance, "on the mat" is a category, and "cat" is a member of that category, as the cat is in fact on the mat. This final piece is the truth value: it is true that the cat is on the mat.
ReplyDeleteI understand some animals can communicate through sounds, but they cannot say everything that can be said, thus it isn't a language. What I'm still a bit fuzzy on is why the little they can communicate is not considered propositional. If a bird makes a sound indicated a snake is nearby, isn't that putting "snake" in the category of "nearby" with the truth value being true? The animal is incapable of categorizing anything through communication (language), but it can make those few alerts. I know we've gone over this several times in class, so my apologies.
In reading you response to Omur, I understand how a picture doesn't "say" anything, as you put it, but intuitively a warning call to another animal does say something, even if that's the only thing they can say. They don't have language, but they can make *one* proposition: a snake is near.
In response to Gabriel, I think that propositional languages, which are symbolic, should have the properties that symbolic representations have. These include systematicity and compositeness that allow you to combine and recombine any words to get any meaning that can be said or thought of. Animals clearly don’t have that.
DeleteComputation does have these properties. But something that animals have that computation does not is grounding. When animals emit signals, they do mean something-- whether these signals represent danger, food or pleasure. Moreover, it likely feels something to the animal that emit the signal to do so.
But our language, natural language, has both of these properties. It has the systematicity and compositeness of symbolic system, but it also has the grounding and feelings that allow us to truly understand anything. I think that animals lack the first component, while computational systems lack the other.
To answer Gabriel (and hopefully kid-sib), I would say that we have learned from class that with the ability to create a proposition comes the ability to create any proposition in the world. Thus, in the example that you give, we could not say that the however many sounds that animals are able to produce intentionally (for example, a distinct sound to warn peers from dangers that will be recognized and correctly interpreted), specifically because the reason you give (i.e. they do not have a language, they cannot communicate everything and be understood, just a certain number of things). Further, I would say that when they make a sound that signals danger, they probably do not have the intention of saying “a snake is near”, but probably just repeating a sound that they know will make others flee. To illustrate that, I doubt that they have a different way to communicate “a snake is near”, and any other predator is near, so they are not making a proposition, just telling other to get out of there.
DeleteRe: Chloe Savignac: “But something that animals have that computation does not is grounding.”
DeleteI disagree with your assertion that animals have grounding. To have grounding means connecting an abstract symbol to a referent, which I don’t think applies to animal communication. Animals aren’t saying “a snake is near,” they are screaming out in fear. My understanding is their communication is closer to pantomime but using sound, if that makes sense?
Also, to echo Marianne’s point, animals are not making a proposition, because a proposition is a subject + a predicate, and you only have a predicate here “danger.” Animals don’t have syntactic rules that make propositional speech possible.
Animal communicative signals can be intentional or automatic, innate or learned. All are categorization ("doing the right thing with the right kind of thing"). And every proposition is a categorization ("A is B" means "A is in B"). But not every categorization is a proposition.
ReplyDeleteA proposition is (a lot more) than categorization: A proposition is "portable": If you are really able to state and mean a proposition at all, you are able to state and mean any and every other proposition. You have a language.
Here is something a little more than an analogy: "Subitizing numerosity" (being able to perceive at a glance, and categorize, whether you are viewing 0, 1, 2, 3, 4 or "more" things) is not the same thing as counting. Counting is a recursive process you can apply to any number of things.
Counting is "portable," subitizing is not.
Gallistel, C. R. (1988). Counting versus subitizing versus the sense of number. Behavioral and Brain Sciences, 11(4), 585-586.
An utterance that is OG noncompliant is nevertheless interpretable (and, further, understood as functionally grammatical). There is nothing that prevents me from interpreting "apple red" as the proposition "the apple is red". This is not because the "variation in grammatical ability" is genetic, as Pinker and Bloom suggest. Rather it is because, underlying human language, there is a grammar (UG) that allows us to make non-OG-compliant utterances that are still semantically interpretable.
ReplyDeleteWe understand the evolutionary advantage of language (OG) to be that we can make propositions: that we can express whatever it is that we want to express.
Although they never explicitly make claims about UG, in discussing the evolution of language, Pinker and Bloom merge OG and UG into the same idea. Pinker and Bloom appear to confuse their discussion of the evolutionary advantage of language with a discussion of the evolutionary advantage of UG...
Although the OG noncompliant utterance of "apple red" is interpretable as "the apple is red" this is just one example. In this case, there aren't many other alternative interpretations as red is a benign physical quality and an apple is an object that doesn't do much, but a more complex OG noncompliant sentence may be less interpretable such as "bed Jimmy stove now".
DeleteThe fact that OG is learnable implies that there are right and wrong utterances, otherwise one could not learn it. So an OG noncompliant sentence would have to be incorrect and the only way to know that might be because of its semantic ambiguity
Yes the fact that OG is learnable implies that there are right and wrong utterances. However, UG is not learnable and thus, does not have right and wrong utterances. This is one of the reason why we can confirm that UG is innate and unlearnable - because there are no negative cases of UG or UG noncompliant statements.
Delete“Noam Chomsky, the world's best-known linguist, and Stephen Jay Gould, the world's best-known evolutionary theorist, have repeatedly suggested that language may not be the product of natural selection, but a side effect of other evolutionary forces such as an increase in overall brain size and constraints of as-yet unknown laws of structure and growth”
ReplyDeleteI am a bit confused about this assertion. Here, Pinker states that Chomsky sees language as an evolutionary byproduct, but in later sections discussing these “non-selectionist” components of evolution he does not cite Chomsky once. Obviously Chomsky sees UG as innate, therefore agreeing that it somehow emerged through evolution. However, it doesn’t seem as if Pinker has supporting quotes from Chomsky equating language to the spandrels analogy. Pinker does quote Chomsky in saying:
It is perfectly safe to attribute this development [of innate mental structure] to "natural
selection, `` so long as we realize that there is no substance to this assertion, that it
amounts to nothing more than a belief that there is some naturalistic explanation for
these phenomena. (1972: 97)
Evolutionary theory is informative about many things, but it has little to say, as of now, of
questions of this nature [e.g., the evolution of language].
It seems as if these quotes demonstrate that Chomsky (while not negating the possibility) has little to say on the evolution of language, as it would just be contriving a story. So, I am confused at why Pinker is arguing against Chomsky in this paper, if Pinker’s ultimate intention is to prove how language could have evolved through natural selection.
I understand that in this paper, Pinker and Bloom argue for the evolution of language as a product of natural selection, however how would they explain how we were able to connect a symbol to its referent and carry on the grounding through time ? Also, how did specific OG rules evolved in a uniform manner with the consensus of a whole population?
ReplyDeleteMore over at the end they state, " But such skepticism should not lead to equally unsupported assertions about the necessity of spandrels and saltations.
A major problem among even the more responsible attempts to speculate about the origins of language has been that they ignore the wealth of specific knowledge about the structure of grammar discovered during the past 30 years. As a result language competence has been equated with cognitive development, leading to confusions between the evolution of language and the evolution of thought"
How is thinking different from language? Isn't thinking language itself? How are they both separate? Could thinking have evolved without language ?
Thinking and language are mostly separate. It is hard for us to imagine what thinking without language would be like. However, take the case of a newborn infant. This child is clearly able to cognize, but does not know any language. Another case would be Helen Keller: before she learned language and could speak, Keller was most definitely cognizing. Last case: my dog does not know english, or any other language for that matter, though he most definitely shows signs of feelings (to know for sure would be solving the Other Minds Problem). So yes, thinking (cognition) is different from language. And thinking (cognition) certainly evolved without language.
DeleteHi William,
DeleteYou’re saying that children are able to cognize, but don’t know any language. However, children are strongly motivated to learn languages because it’s a need for human communication. Therefore, I don’t think we can think of thinking and language as two separate entities. Infants are able to understand the world around them. They do not use the words that we use, but they probably have a language on their own. They might have symbols on their own. Thinking is separate from language only to a certain degree, but they are highly connected.
The most interesting argument of this paper to me is the comparison of language to writing. On this, Pinker states that:
ReplyDelete"the ability to use a natural language belongs more to the
study of human biology than human culture; it is a topic like echolocation in bats or stereopsis in monkeys, not like writing or the wheel."
To me this draws an interesting divide between the methods with which we communicate today and what enables them in our being. If language is product of biology but writing is a product of human culture it stands to reason that there was some evolutionary decision that determined verbal language to be the most critical or effective means of communication for early humans. To me, this also implies that writing was a creation out of "luxury" or "convenience" for later humans to enable a more massive scale of communication differentiating us from other species.
In this paper, Bloom and Pinker aim to argue for the evolutionary basis of language. However, the basis for their argument seems confusing. They argue that if a cognitive process is complex and there are no alternate explanations for the complexity, then it should be attributed to natural selection. However, a process with complex design and lack of alternative theories could also indicate that we have not progressed enough in the knowledge domain to come up with alternate theories. The argument that “It is absurdly improbable that some general law of growth and form could give rise to a functioning vertebrate eye”, hence leading us to conclude that evolution led to the formation of the eye, is a proof by elimination. This can only work if we can list out all the possible theories. But we cannot be sure that we are not missing a possible theory. Also, the authors argue “it is not necessarily illegitimate to infer both special design and adaptationist origins on the basis of function itself. It all depends on the complexity of the function from an engineering point of view.” It seems like they are implying higher complexity necessitates it to be evolutionary in origin. But as we discussed in class we do not know the aim of evolution. The idea that evolution occurs to ensure survival and reproduction seems like one possible explanation, not a theory. It is entirely possible that both survival and reproduction serve another bigger goal that we cannot fully comprehend yet. So, this only means this is a higher probability, not a confirmed theory? The authors address these questions in the article by stating that “the entire theory of natural selection may be literally unfalsifiable”. However, they rightly argue that in the face of nonadaptionist alternatives, there is some merit to forming these theories to broaden our understanding and further research on the topic.
ReplyDeleteRegarding Chomsky’s arguments for non-evolutionary basis of language, the authors seem to imply that these arguments “are truly incompatible with the neo-Darwinian theory of evolution”. I struggle to understand how the argument for a sudden capacity for language due to increasing brain complexity, contradicts natural selection more than adaptive mutations that survived generations. I do not see an inherent contradiction between both views.
When the authors say “arguments that language is designed for communication of propositional structures are far from logical truths” – are they saying that language did not evolve for communicating propositions? Or not just for communicating propositions?
I like how the authors see the purpose of language to be “communication of
propositional structures over a serial channel”. The origins of language remain unknown as adaptionist views argue for the slow evolution of language, without evidence for how it evolved. The view of sudden emergence of the ability for language due to increasing brain complexity also does not explain how this change occurred.
Like the previous paper, this one strongly advises towards interdisciplinary work. In a field like linguistics and language acquisition where evolutionary research is already well established, involving experts and theories from other domains could only be beneficial. Without the risk of being taken off track or of the other domains taking over, new ideas and perspectives could be introduced this way. This would also give the opportunity to bring more modern research on the evolution of language into the domain and theories.
ReplyDeleteThis approach may also allow for re-examining the application of natural selection to language.
i. Pinker and Bloom’s extensive foray into the evolution of language was, as they admitted to in the preface to their argument, not entirely exciting. However, despite the density of the test, the discussion of the evolution of language and the diverging opinions which surround it was nevertheless important. Notably, like a spectre throughout the paper, Chomsky’s Universal Grammar hangs over the author’s commentary and is addressed only briefly. The “innate arbitrary foundation” of UG proves to be foundational for language as a whole, and exists, with the ability to learn, as the innate parts of human language evolution, essentially allowing for its limitless propositional powers. The authors argue simply that human language, like other biological adaptations, evolved through natural selection. In light of the progress we’ve made in the class since, this speaks to the ways in which evolution need not impose major physical adaptations to ensure the progression of a species and can offload it onto other, perhaps cognitive means, which can progress themselves.
ReplyDelete