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.
Saturday 2 January 2016
8a. Pinker, S. & Bloom, P. (1990). Natural language and natural selection
Pinker, S. & Bloom, P. (1990). Natural language and natural selection. Behavioral and Brain Sciences13(4): 707-784.
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.
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.
Subscribe to:
Post Comments (Atom)
The definition of Exaptation is as follows: Exaptation describes a shift in the function of a trait during evolution. For example, a trait can evolve because it served one particular function, but subsequently it may come to serve another.
ReplyDeleteDuring this reading I cannot help but wonder how we are currently ‘adapting’ to technological innovations and machinery. Our reliance on devices and digital technology to accomplish our daily tasks and lifestyle is so ubiquitous we barely give it any thought.
In the article, the vertebrate eye was discussed as ‘the classic example of adaptive complexity’. “It is impossible to make sense of the structure of the eye without noting that it appears as if it was designed for the purpose of seeing.”
When discussing exaptation, Pinker cites Mayr and says, “there must be an intermediate evolutionary stage at which the part can subserve both functions (Mayr, 1982) , after which the process of natural selection shapes it specifically for its current function.”
I am wondering whether the human eye, initially used more for seeing in the distance (hunter-gatherer lifestyle), now almost always glued to close-up screens, will slightly exapt in the way outlined above to avoid the current common problem AMD (age-related macular degeneration) and the need for most adults to wear reading glasses.
Aliza, there will only be a change in the distribution of genes if some genes enable their bearer to survive and reproduce better than others. If AMD is genetically based, and there is another gene that fixes it, and the AMD is bad enough to handicap survival and reproduction, then the non-AMD gene will out-compete the AMD gene in later generations. But is there any evidence that any of this is genetically based? (Calling it "exaptation" does not change any of these evolutionary basics.)
DeleteThe article had me nearly convinced that language had to be evolved by natural selection, was it not for two points that I was not so sure about. I think the authors overestimate how much of language has been shaped by evolution and has become innate. I believe that specific expressive abilities and some nuances of language are learned. Language might have started at the same time throughout different groups, but it makes sense for our ancestors to have devised rules similar to each other at that level of intelligence and capacity. Then, it makes sense that these were taught to their children. The human brain certainly must have adapted and evolved or allocated resources to general abilities that one needs for language production and comprehension, and there are areas specific to certain aspects of language. However, the rest must be taught. In one place in the article it says that "They must be programmed so that the mere requirement of conformity to the adult code, as subtle and arbitrary as it is, wins over other desiderata." However, I believe that children learn a lot of the principles through trial and error, and not all the grammatical principles are hardwired as the authors suggest. Although parents might not provide the rules of a certain language as one would if they were to learn a language later on in life, I believe that kids can figure it out since they hear it without any formal rule teaching, starting from the womb, to school age, every day.
ReplyDeleteAnother question this article raises for me is why no other animal has yet to come even close to acquiring a language? What made us deviate from the others? The fact that animals don’t even come close to learning basic language against the fact that there are some really intelligent ones, makes one question. So could it be that our enlarged brain structures and enhanced physical capabilities lead us to allocate these resources to language, instead of language leading to enlarged brain structures and enhanced connectivity? Thus, it makes sense for language to be kind of a by-product, unlike the authors would like to believe.
Hello Deniz! I liked your point on how evolution explains some but not of language. However, the article does not say that a specific language is innate. It is relating to universal grammar, the aspects of grammar that are innate. The reason why it is thought to be innate is because children do not make mistakes that violate UG when acquiring language, although they make other mistakes in learning a specific language when they set parameters incorrectly. So this means you are completely right! However, one part of what you said, namely, that children need adults to teach them language, I will have to disagree with. There are a few reported instances of new languages being created. For example, there is one case where deaf children who were isolated from each other and had no exposure to language when put together crated their own sign language. This was a completely new language that adheres to UG but was not taught!
DeleteThat is to say, certain conjugations of irregular verbs and lexicon and parameter settings must be learned. However, the ability to communicate through language that abides by UG does not. This is the aspect of language that is being referred to in the article!
I also wanted to make a comment on the part of the article that refers to “naïve adaptionism”. This maybe true that when looking at spandrels we have to consider that they are a byproduct of the limitations imposed by a dome. However, in biology and evolution how can we accurately go back and determine if language is the dome or the spandrels? We only have the final result, of acquired language, but we do not know the architectural limitations that are present. How do we figure out architectural limitations on something that we have not built?
Daniz, yes, of course, a lot of language is learned (vocabulary, pronunciation, ordinary grammar), because evolution is lazy and only codes as much as needed to survive and reproduce as well as or better than the competition. (I don't think P & B suggest otherwise.)
DeleteBut there is one controversial property of language that is not learned, and we will discuss it next week (Valentina already mentions it in her comment): Chomsky's Universal Grammar (UG), a complex set of (syntactic) rules that all languages follow, but that children do not learn, neither by trial-and-error induction nor by verbal instruction. They can't learn UG that way because they never make UG errors, and never get corrections. That's the "poverty of the stimulus." So they must know it already. (Nor is it evident how UG could have evolved.)
So UG is what makes the evolution of language controversial -- not the stuff P & B talk about, which is all relatively straightforward.
Why didn't other apes evolve language? I'm as perplexed as you are...
Valentina, you're right about UG, but the Nicaraguan sign language was not invented by a single child but jointly invented by bringing together deaf children who had lived only within their own (hearing) families and treated as if they were mentally retarded. They have normal, language-prepared brains (prepared by Baldwinian evolution, strongly predisposing them to learn language); it just happens that they an't hear. So when they came to teach them a sign language, the kids beat them to it by inventing a sign language of their own, just because it was the first chance they ever got to try to communicate with other deaf children through gestures.
You are right that "spandrels" is just a metaphor and does not really explain anything (except developments in iconography).
Valentina’s example of the Nicaragua children coming up with their own sign language draws attention to some important features of human capability of language. First, the Nicaragua children are evidence that there is high plausibility that gestural language existed before verbal manifestation of language evolved. Baldwinian evolution resulted in the development of innate UG. Evolution satisfices human’s language capacities, we are all born with the ability to learn language, evolution has resulted in a good enough solution that has granted humans the capability of creating propositions. Therefore, this explains why the Nicaraguan children were as a group, able to come up with new language that complies with universal grammar but most likely has very different syntax or grammar than the local spoken Nicaraguan language. These children were able to do this without any induction or negative feedback to distinguish propositions complying to UG and those that do not comply with UG.
DeleteSecondly, the Nicaraguan children example points to the fact that an invented language cannot be considered a language, unless it is understood by more than one person. Language cannot be invented by one person alone because there is no right or wrong feedback for categorization. This is comparable to trying to categorize an emotion ‘X’ that no one has ever felt before. The next time ‘X’ emotion is felt by the only individual who has felt it, there is no external feedback about what that emotion is called. The feedback I am discussing here is different and most definitely distinguishable from the poverty of stimulus feature of UG.
Nadia, considering that researchers have shown that chimpanzees are able to use nonverbal gestures to communicate among themselves, I think it is safe to say that gestural language existed before the verbal manifestation of language. Since some researchers believe that they have been able to tap into what some of what the gestures mean, it is interesting to consider if chimpanzees have the capacity to develop stronger language abilities and eventually verbal language abilities. This is quite a reach however, and the other-minds problem will always keep us from fully knowing if the gestures that have been interpreted do actually correlate with their interpretation.
DeleteEugenia, the problem here is that a chimpanzee’s gestural communication and the human gestural language are fundamentally different. Where human language is able to communicate propositions that allow for a subject to be communicated as part of a category, a chimpanzee’s gestural communication is limited to pointing. It’s true that gestural communication, as used by chimpanzees, existed before the verbal manifestation, but human gestural language, such as American Sign Language, came to exist after the verbal manifestation of language.
DeleteHi Eugenia and Mila,
DeleteI agree with both of you, I think the nuances that arise when describing 'gestural' language come about because throughout the course, we have used gestural language to mean different things. We have used gestural language to describe sign language AND to describe much simpler gestures that animals like Chimpanzees are able to make and use as a form of communication. Like you mentioned Mila, human gestural language, like sign language, can be used to form propositions and has the same characteristics as languages like English or French. Sign language follows the rules of universal grammar and with this form of gestural language, we have the capability to say any proposition. With chimpanzee gestural language, there is not this capacity to generate propositions, communication is achieved mostly through gestures that point to referents or gestures that symbolize referents.
A bit tangential to the course, but the above point comparing sign language to English or French is supported by research in language development that shows children learn sign language much in the same way that they learn other propositional languages (e.g. babies “babble” in sign language as they do when learning other languages, similar brain regions are engaged, etc.). I think we could see UG as the commonality underlying all human languages with OG being the differences between each language that we learn from birth (whether that be Chinese, sign language or Arabic).
DeleteSign language should certainly be distinguished from gestural communication, which is common in many animal species as well as humans. I think it is important not to conflate broader forms of communication with language, which we understand to represent a distinct form of communication via propositional structures.
Grammatical devices and expressive functions do not pair up in one-to-one fashion. For example, some languages use word order to convey who did what to whom; others use case or agreement for this purpose and reserve the use of word order to distinguish topic from comment, or do not systematically exploit word order at all
ReplyDeleteThe author seems to imply that the diversity of languages disproves a theory for a universal grammar. However, studies such as the ones performed by White & Juffs (1998) showed that native Mandarin speakers (who do not possess wh-movement rules in their native language) still showed that they did not violate the universal grammar rules for wh-movements when learning English for instance. In other words, their judgments of the rule was still accurate even though they were never explicitly told about it. Therefore, even if I agree that most languages do not “pair up in a one-to-one fashion “, there are still some grounds common to all languages.
Josiane, as Valentina mentions above, word-order is not really a part of UG: it's a parameter that we learn, the way the mole learns how big to grow. A parameter (Chomsky likens it to a knob on a radio, picking the station) may have several possible settings, and a first language can only have one of them. But some kids learn multiple first languages at the same time, or go on to learn a second or third language later, and the other parameter settings are still available for that. It's just within one language that you can't have more than one.
DeleteThis article was very clear and well written in my opinion, however it was also very long and so I feel like a short summary is in order to be able to discuss it. Pinker and Bloom discussed a straight forward topic: the fact that language must be partially innate and partially nurtured. As a human trait (the way wings are bird's trait, not exclusive to birds but a truly performant tool in birds with huge evolutionary advantage as compared to hens or penguins), it was shaped by natural selection, allowing individuals with "stronger" language skills to gain more reproductive advantage. Since it is both innate and learnt, language is made of universal grammar and ordinary grammar, with the addition of vocabulary and other language-specific characteristics. Universal grammar is an innate part of language (point agreed to by Chomsky, proved by the poverty of the stimulus regarding non UG-compliant sentence structures, and erroneously explained by the Spandrel metaphor), and the other features (ordinary grammar, vocabulary, pronunciation etc.) are learnt. It is pretty obvious that the innate part of language is shaped by natural selection (since language gives reproductory advantage), but is also said that the learnt portion. The part I do not understand quite as well is the distinction made between UG being innate, and UG being a product adaptation. It seems like they consider these two to be incompatible, whereas it seems to me like Universal Grammar could have evolved slowly whilst being innate, let's say the first human with language had Universal grammar from birth, it's just that their Universal grammar probably corresponded to 10% of what today's humans are born with. Please let me know if you think I misunderstood the problematic here.
ReplyDeleteWith UG, it is assumed that everyone starts with the exact same structure from birth- this is what makes it universal. To distinguish between languages we say that each person learns 'Principles and Parameters'. These PPs serve to restrict the ways UG functions within each person's grammar. It's as if you have a giant box of light switches, and nurture turns some on and some off. I would assume that the the authors are saying that UG is a concrete consistency but the PPs change throughout time.
DeleteJulie, good summary. But distinguish the evolutionary advantage of language itself (being able to learn and teach new categories through words: propositions) -- which is fairly straightforward, partly innate and partly learned -- from the evolutionary advantage of UG in particular, which is controversial because it is not clear either how UG evolved nor what advantage it conferred (could it have evolved gradually? how? why?). P & B skip over this problem completely, shuffling together what's learnable, what can evolve unproblematically, and UG, which is a special problem, not solved by the relatively obvious things they same about the evolution or learning of other features of language.
DeleteAdrian, P & B are saying nothing substantial at all about UG (which, to me, casts doubt on whether they even understand why it's a problem for evolution.)
Language learning is not programming: parents provide their children with sentences of English, not rules of English.
ReplyDeleteWhat about when language is learned by rules? Almost necessarily, the first spoken language that a child learns will be through hearing sentences, not learning rules, otherwise it would be difficult to find a medium through which to provide the rules (though one could make the argument of demonstrating rules through non-verbal strategies).
However, once a child has a basic understanding of one language, the opportunity arises to learn a second language via rule learning (with the rules being presented in their first language). Most children who are truly bilingual from a young age have actual language exposure from both languages, and not simply rules. However, it is possible to teach someone a language only through rule-teaching.
So would we then say that first-language learning is definitely not programming, but second-language learning could be viewed that way, if done in a certain way?
I would say that even second-language learning cannot be solely programming/learned by rules. Let's just invoke Searle's chinese room argument again, in the version where all off Chinese's grammatical structures are learned by the individual in the room, we still do not say that he is a Chinese speaker on any level.
DeleteFurthermore, I think language acquisition literature would point to the necessity of a certain linguistic context that does not only explicitly instruct second-language learners the rules of grammar. But rather, through surveying the language of native speakers, there is a re-adjustment of the parameters of universal grammar.
Dominique, yes, second languages are more likely to be learned through explicit verbal instruction rather than through trial and error induction, like first languages.
DeleteYi Yang, languages (both first and second languages) are not just syntax. There are words, with grounding and meaning, with grounding provided by learning categories (word meanings) by induction, and eventually instruction too.
This is a little off-topic, but Dominique’s comment made me think about the educational implications of learning a second language based on rules. As Dominique pointed out, first-language acquisition involves repeated exposure to the sentences of a language with very little rule learning (which seems analogous to unsupervised learning), whereas second-language acquisition usually involves less pure exposure and more rule learning in a classroom setting (which seems analogous to supervised learning). There have been a bunch of studies showing that language immersion programs, which basically involve immediate exposure to complex sentences of the target language, are typically much more successful than classroom based approaches in teaching a second language. Is this because they mimic the unsupervised learning conditions of first-language acquisition?
Delete@Dominique
DeleteReflecting on language acquisition, the learning of another language other than your own (i.e., second language, third language, etc) is accomplished through explicit verbal instruction. But, it does not answer how language was learned when explicit verbal instruction was not the norm. How did different early hominids communicate with one another, if not through trial-and-error induction?
@Olivia
Second-language acquisition would be analogous to supervised learning, in a classroom setting. However, there are refugees or immigrants who learn by repeated exposure (none or minimal rule learning). They are complete beginners in a foreign language and therefore the language of instruction is already in the foreign language. However, they obtain vocabulary or speech patterns that are not representative of the overall language competence. It may be that the foreign language is acquired in an informal setting which requires further investigation.
I think this point exactly illustrates why, for example, I learned English almost immediately with my parents, but have been studying French for 8+ years and can barely hold a conversation. Rule based language learning is very, very difficult and is why (at least in the US) no foreign language students can actually speak the languages they're learning. French was taught to me very systematically, yet I don't really have a sense for how to actually use it. This is why watching TV and reading are better than doing grammar exercises.
DeleteTo add my two cents to this thread full of interesting ruminations on second-language acquisition: it is an interesting point to me that the grammar of a language is not something that appears to be best learned and taught through unsupervised learning. That is, I think we’ve seen a pattern in which unsupervised learning, much like a second-language immersion program or a child picking up grammar by listening to sentences, is more useful than rule-based learning for acquiring instances of grammar. I wonder why this seems to be directly opposite to what we think of as category learning. In the first few generations of language, categories are named and learned through trial and error. But it is at the precise point where speech becomes propositional where we can begin teaching and learning through hearsay, or what we could perhaps consider supervised learning. Collectively, our linguistic canon includes all the categories ever delineated and named, and this is best learned through supervised learning. Paired with the learning of grammar that appears to be optimised for unsupervised learning, I think we can begin to get the sense that UG developed as an innate ability, best learned by observation, because it supported our ability to learn categories through supervised learning.
DeleteRegarding Dominique’s “Most children who are truly bilingual from a young age have actual language exposure from both languages, and not simply rules. However, it is possible to teach someone a language only through rule-teaching.
ReplyDeleteSo would we then say that first-language learning is definitely not programming, but second-language learning could be viewed that way, if done in a certain way?”
Your pondering of whether learning a second language learned strictly via rules (instruction) instead of by exposure (induction) reminds me of learning another skill, playing & learning music.
Almost all people in Western culture are exposed to music, and a minority of those people go on to play their own music in some respect. Most of these amateur music makers do not learn music in a formal setting and merely learn from what they hear. These people are able to play relatively well, and especially if they start young, they can become very talented and proficient. Alternatively, there are musicians who are classically trained from the beginning, with music theory being central to their playing. These musicians tend to play in a very clunky, almost robotic manner at first, and tend to struggle to make coherent, natural-sounding music at first, as compared to their induction-taught counterparts. Eventually though, the instruction learners can become fluent in their craft and usually match or surpass their induction-educated peers. The act of playing music can be looked at as a means of communication within a strict set of rules (Western music) and either method will end up with the ability to play music.
The instruction-educated musicians also have access to induction, even if it’s just the sound of their own instrument as they play, the sound of others playing around them, or the music that they listen to. The same can be said of the second-language speakers, even if they are in a language course. They will hear the language in the world outside the class or even in the class when others around them (including the teacher) speak.
At the end of the day, language and speech are categories, which means using the right words for the right situations or things. Whether one learns it from induction (first language) or instruction (second language), they both can create the same output after enough learning. Your analogy to programming seems to be a different way of looking at instruction learning, and either way it doesn’t seem like programming to me. When we execute a program, the code is the sole information besides a few basic concepts built into the machine, while humans are constantly bombarded with examples of how to speak or how to play/make music.
Karl, good points and good analogy.
DeleteInduction is mostly trial-and-error induction, with corrective feedback (i.e., reinforcement/supervised learning).
But there's unsupervised learning (mere exposure) too, both in music and in language.
Formal music training includes learning to read music; without that there is "only" imitation and improvisation.
Word referents are categories; syntactically well-formed/ill-formed utterances are categories.
But learning a language is not learning a category; it's lot's of things (pronunciation, syntax, vocabulary).
Does UG have an expiry point or is it operational throughout one's life? Pinker argues that there is some kind of UG but it is the ability of a child to look for combinatory rules when they hear language. He writes " From the very start of language acquisition, children obey grammatical constraints that afford them no immediate communicative advantage". He says that children say Big Dog but never Big Fred. Based on this argument any individual learning a second language after a certain age is learning language through the exact same process. They are adequately using the constraints and rules provided to them to learn a new language. This is clearly not unsupervised learning because UG is the innate category that allows for people to pick up the rules that govern language and through the rules same rules, learn the language. So according to Pinker there is no difference between UG and OG. How does he then account for the inability of second language users to correctly pronounce morphemes?
DeleteRE: “Symons' (1979) observation that tribal chiefs are often both gifted orators and highly polygynous is a splendid prod to any imagination that cannot conceive of how linguistic skills could make a Darwinian difference.”
ReplyDeleteThis quote in particular stood out to me as it appears that Pinker and Bloom are conflating an individual’s oratory or language skills (not one and the same as far as I understand) with an innate linguistic capacity that all humans are born having. Even if there is some validity to modern linguistic capacity having developed due to evolutionary adaptation, I don’t see how an akin natural selection would be behind one human being a better public speaker than another one. In that regard, their argument regarding natural selection acting as a ‘programmer’ that codes for optimal linguistic ability may have some merit, but the main issue I take with the aforementioned quote is that for all intents and purposes, general linguistic capacity is static (amongst healthy individuals) whereas one can improve their speaking style and efficacy over time. Furthermore, natural selection generally is referred to in a context where the attributes that are passed on over time are ones that facilitate the likelihood that an individual or animal will survive. Oratory aptitude is not something that directly determines whether one lives or dies.
I am not disputing that the linguistic skills humans have don’t make a “Darwinian difference”; linguists posit that the linguistic capacity encompassed by UG was a result of a single genetic mutation, one that improved chances of survival (linguistics professor Jessica Coon mentioned this today at NiRC). However, I think the example brought up in the quote is a simplistic one and detracts from other legitimate points made in the paper; one such point has to do with the motivational aspects of language (“within a group of interdependent, cooperating individuals, the states of other individuals are among the most significant things in the world worth knowing about”). The point about motivation is touched on more in 8b.
Liza, P & B's conjecture about oratory skills was evolutionary explanation of the silliest kind (as a hypothesis about why language evolved). But there is (genetic) variation in many traits within a population; so after language evolved, oratorical skills might have been useful; but I doubt that gene-based oratorical skilles were useful enough to influence survival/reproductive success enough to change the proportion of the genotype in the population.
DeleteThis paper argues that human language is evolved by natural selection and the authors used lots of supporting evidence. This reminds me of the the very first introductory lecture when professor Harnad talked about language being a new way of learning categories, and categories are mostly learned and not borned with. With language, instead of learning something by trial and error, we can learn the categories with the arbitrary language label. As such, surely there are a lot of evolutionary benefits associated with language, but I’m just wondering if language is not purely evolved by natural selection and there’s a learned counterpart, probably with the help of mirror neurons?
ReplyDeletePeihong, the disposition to learn language evolved (other apes lack it) but language itself (except UG) is learned.
DeleteThe paper argues that one of the reasons that language is evolved by natural selection was the reproductive success comes along with superior grammar. I do agree with the inborn nature of universal grammar that underlies all human languages posit by Chomsky. That is, these rules are not learned and all new-borns have the capacity to generate grammatically correct sentences. However, I wonder if this can also apply to sign language because I remember from other social psychology class, sign languages also follow universal grammar. As such, is it possible that only the universal grammar part of language is selected by evolutionary force and the rest (e.g. pronunciation, different adjectives, etc.) are highly malleable or learnable with interpersonal variations or genetic mutations as stated in the paper.
ReplyDeletePeihong, language was adaptive because it was language, but no special bonus for "superior grammar."
Delete"Stevan says" that language evolved first as sign language, and then migrated to the vocal modality.
If UG evolved, it is not clear how or why.
Yes, but I remember in the paper the author argued for the “special bonus” associated with superior grammar, and that’s why communication with correct grammar is important for, e.g. social economic status. However, if language itself is adaptive, why did different cultures develop different grammars building upon UG? In fact these grammars are not easy to learn for kids so they do not speak naturally according to these rules.
DeleteIn this article Pinker is arguing that language is evolved and learnt through natural selection, but he is implying that it was an “adaptation,” just like the human eye, rather than a by-product or noise. I find the eye analogy pretty different to language development, as the function of the eye, like the function of the heart is vegetative, whereas the function of language (to categorize and communicate) is so much more complex. Language gave us an “instructive” advantage which studies showed required less trials to learn but took longer to apply to a real-life situation. While apes have induction which in fact has a greater advantage when doing things as opposed to learning them. It seems to me then that language was not necessarily more “adaptive” for humans, and so I wonder why Pinker argues that.
ReplyDeleteSecondly, he talks about UG in a Fodorian sense of a “cognitive equivalent of the Big Bang.” But, UG is not about being born with grammar rules inside your head. That doesn’t really make sense. Instead, due to Baldwinian evolution (which says genes don’t code more than they need to), it is about have a genetic tendency to learn UG. So although Chomsky speaks of UG as “innate” he means it in the same way as Pinker sees it an evolved by natural selection.
Nimra, surely language was adaptive, since we are all able to learn it and other apes are not. (And we are at 7.5 billion, whereas nonlinguistic species (except the ones we purpose-breed for our "use") are all dying out.
DeleteNo, there is no genetic tendency to learn UG, because UG (because of the poverty of the stimulus: not errors or correction) is unlearnable. So it must be innate (though it is hard to see how it could have evolved).
• In my opinion P & B are convincing in their view that the only way to explain the development of language is through natural selection. They criticize Chomsky’s argument that physical laws of the brain can explain the emergence of language, which they think there is absolutely no support for. From their paper it is clear that this is a very controversial topic, but for me natural selection as an explanation is very appealing. I believe that it has become a necessary skill for the human species to survive. What I’m wondering is how there can be so much debate and skepticism around natural selection being the answer if the evidence is there in historical examples showing the variation of language, something has clearly been selected for and has evolved? Perhaps I’m looking at the problem far too simply, though, so I would be interested in hearing if anyone has arguments against natural selection as the explanation for language.
ReplyDeleteI think I would understand Chomsky's view in favor of language developing from physical laws or as a by-product of another adaptive function if language didn't have as big and as important of a purpose as it does. How does it make sense that language has transformed into this extensive, complex, and extremely useful skill as a by-product and not with an evolutionary purpose?
DeleteI am also convinced by P & B's argument, but I am not sold on the idea that language is/was a necessary skill for all of human survival. Assuming that the only major difference between us and the great apes is language (or, more specifically, propositions), they seem to be surviving just fine without it. Our dominance of our environment and other species is probably the most notable result of language.
DeleteHello Maya and Colin, like you both said I do agree that from reading P&B, the development of language through natural selection is seemingly evident. Its evolutionary purpose and development through natural selection is supported by its complexity and extensive patterns.
DeleteMaya, when you said that "it [language] has become a necessary skill for the human species to survive." I believe you are saying in our modern society - that we have adapted the social use of language and learned to be cooperative and cognitively interact via the tool of language to evolutionary accelerate. Such that it is now a necessary skills in this modern world for each individual to function as a part of a human social collective. Is this perhaps what you mean? Or do you literally mean in terms of basic survival?
If it is the latter, then I would agree with Colin: without language, humans are still very capable of surviving. Such as the cases of feral children. They do not communicate in language to humans and is difficult for them to integrate into society, but doesn't seem to have an implication on their survival interest.
I'm not sure if I agree with the example of feral children. P & B discuss a "language of thought" which, if I understood correctly, still implies some sort of language that is used by humans, even if they cannot vocalize. I think that although humans may not need speech to survive, it certainly has demonstrated to be evolutionarily advantageous by putting them at the top of the food chain, and I think that is enough to argue that it has been selected for.
DeleteI agree with Colin in Grace in that humans are still capable of surviving if they lack language abilities. I believe this is because they are still able to communicate with those around them in ways other than spoken language. However, as Lucy said, I'm not sure I agree with the example of feral children either. Feral children have usually been isolated from human contact from a very young age, meaning they lack many of the basic social skills required to communicate effectively (in addition to language) and behave appropriately. As such, I feel as though they would have difficulty cooperating and functioning as part of the collective, when introduced into society for the first time.
DeleteOn the note of language as a "necessary skill for human species to survive"- I agree with above comments that it's not necessary. However, this paper argues (I think quite well), that language can theoretically help improve chances of survival. This also reminds me of discussion of degrees of freedom, where language is not the only way that humans can survive or thrive, but that it is one way that proved itself. It is not necessary, but it is sufficient to aid in our survival.
DeleteTo Lucy, from what I understood from the paper, these authors do not agree with a conception of language of thought. At one point they say "language competence has been equated with cognitive development leading to confusion between the evolution of language and the evolution of thought," suggesting that thought is not language. I think this relies on a main argument that language is social and comes with communicative adaptive functions, whereas thought is only internal.
After reading this article, I sensed a contradiction regarding Chomsky and Pinker & Bloom's arguments. Pinker & Bloom argue that language shows adaptive complexity where any system composed of many interacting parts fulfill some function. Natural selection seems to be the only explanation that we can account for the evolution of language.
ReplyDeleteOn the other hand, Chomsky believes that language is a side effect of evolutionary forces and that language is directly tied to our cognitive abilities. A contradiction rises as Chomsky agrees that our brain capacities rose out of a series of adaptions from evolutionary forces such as an increase in brain size – so why is the language system not in any way adaptation? Is it cognitive capacities that are responsible for language or natural selection?
These two views seem necessarily incompatible. If natural selection is responsible for language and natural selection is also responsible for cognition, then shouldn't cognition create language? I personally am more convinced by Pinker & Bloom's argument as language is so unique to human existence and yet apes and chimps fail to learn it. Therefore it seems necessary for survival as compared to the non-linguistic species, we have longer life expectancies, brain development and knowledge and ultimately have evolved to become highly capable of dealing with the environment.
Re: 5.3.4. Social use of language and evolutionary acceleration.
ReplyDeleteI really like the points that are brought up in this section, which discuss how evolutionary processes naturally led humans to acquire language. Early humans depended on extended cooperation for survival, communicating about food, safety, nurturance, and reproductive opportunities. In hunter-gatherer societies, basic linguistic expression allowed people to understand each other instead of operating as individuals limited to egocentric tasks. For instance, if person A hunts for meat and person B picks berries, both people could trade items benefiting both parties. In economics, this is represented by the law of diminishing marginal utility, which is a law stating that as a person increases consumption of a product while keeping consumption of other products constant, there is a decline in the marginal utility that person derives from consuming each additional unit of that product. This opens more possibilities when more factors come into the equation, and evolutionarily humans utilized language to make such trades possible.
This section also alludes to a cognitive "arms race" whereby humans had an unusually rapid enlargement of the frontal lobes in comparison to other species. Interactions with other humans gave people an advantage for survival, creating greater demands on cognition. As time progressed, interactions naturally became more complex, leading to negotiations, learning of other people's desires, reproduction, and so forth. Such new possibilities equipped humans with an evolutionary advantage, one that still grows today.
While I do appreciate the authors’ point on the social uses of language and the implications on the evolvement of language, I think the authors fall short on giving a sufficient explanation for how and why grammar is involved. My question for them would be how do the social uses of language offer any explanation for the evolution of grammar? They have to bear the onus of showing how and why grammar is needed for the social uses of language. One can reasonably imagine ‘language’ without grammar for social uses – although this can be disputed.
DeleteThe authors bring up a compelling point about the evolutionary nature of language. Although language (vocabulary) itself is learned, the components that make it possible (vocal tract, UG) are innate. Interesting point about the cognitive "arms race" Neil. Although this still doesn’t tell us why humans had the need for complex interactions whereas apes did not? What made us special? Are apes able to communicative everything that their environment requires with gestures (and without grammar)?
DeleteI find the cognitive arms race explanation of human linguistic development compelling, because it is compatible both with Chomsky’s observation of a UG (which must be innate to be universal) and with natural selection occurring over time. The theory, as far as I understand it, is based upon the mutual benefits of living in groups.
DeleteAs opposed to living solo, groups provide mutual benefit so long as benefits are shared via a system of reciprocation (if you scratch my back, I’ll scratch yours). Thus, it makes sense that primates cooperate with each other: however, it is even more advantageous to be in a system of reciprocation if you can cheat (Here, come scratch my back, sorry I can’t return the favour because I’m late for cave painting) because you gain the benefit of others’ work in addition to your own. Chimpanzees love this by demanding instant reciprocation, but in a larger human (or human ancestor) society, increased social complexity requires more complex reciprocation. Cooperation only works if it is mutual, and therefore we need to be able to ‘detect’ if someone is going to cheat us. This ‘cheater detector’, in addition to social complexity, would have provided large survival benefits through collective organization. As a trade-off, this would have required larger, more resource-demanding brains, a-la modern human beings.
This works with natural selection, because it does not spring suddenly into action, it develops because of selection pressure for social intelligence over time. It works with UG because the innate language would form in the same way through evolutionary pressure. The UG connection is admittedly a bit tenuous, and I would love to hear what others have to say!
I also found that section particularly interesting. I agree with Elise in that, if the authors are relying on social factors to explain the evolution of language, they would need to explain why other social creatures, like apes, did not develop language as humans did. The other articles attribute the absence of language in apes to a lack of motivation. However, I am not convinced by that argument. Despite the fact that apes lack the vocal apparatus to allow for complex movement and coordination necessary for verbal language, apes communicate via vocalizations, facial expressions and gestures. This section suggests that what separated the two species seems to have been the differing demands of their social environment. Whereas for apes, their environment remained stagnate and simple enough to communicate via pantomime, human life grew too complex to be simply communicated via gestures.
DeleteExamples of social behaviors that account for the evolution of human language are the emergence of monogamy and bi-parental care in humans, which necessitated complex social interaction, coordination and synchronization between the two parents to increase survival of offspring (this is not a particularly satisfying argument because apes live in much larger, seemingly more complex family groups). Although I find the social explanations for language evolution very interesting, I don’t think they offer much explanatory power on their own for how and why grammar evolved, as Austin said.
RE: “Evolutionary theory offers clear criteria for when a trait should be attributed to natural selection: complex design for some function, and the absences of alternative processes capable of explaining such complexity”
ReplyDeleteThis article argues in favor of a neo-Darwinian explanation for language evolution. Their support for this claim is that language is far too complex to be a product of a series of random mutations. Thus, the only mechanism that could account for such an intricately designed trait is natural selection.
The issue, however, is that the authors conflate general language capacity (combinatorial categorization) with grammatical capacity. While language’s function is easy to see (one can bypass “show” and go right to “tell”), the function of universal grammar is not so clear. Therefore, by not defining UG’s function, the authors violate the criteria for attributing this trait to natural selection. Although they state that “functions of grammatical devices play an important role in evolution”, they never really discuss why that is.
I agree Manda. They state “But in fact there is such an account: the nature of language makes arbitrariness of grammar itself part of the adaptive solution of effective communication in principle.” I have trouble seeing from the reading how specifically grammar evolved with respect to adaptive complexity. “For universal grammar to have evolved by Darwinian natural selection, it is not enough that it be useful in some general sense. … There are no conclusive data on any of these issues.” First of all, to refute rebuttals is not to construct a justifiable position. Second, and more substantively, I agree with Bates: “What protoform can we possibly envision that could have given birth to constraints on the extraction of noun phrases from an embedded clause? What could it conceivably mean for an organism to possess half a symbol, or three quarters of a rule? (p. 3) ...monadic symbols, absolute rules and modular systems must be acquired as a whole, on a yes-or-no basis -- a process that cries out for a Creationist explanation. (p. 30)" Going back to my post on gradualism and exaptations, this is exactly the problem that is left unaddressed. There must be evidence that grammar evolved gradually with intermediates, otherwise, with non-gradual changes, the argument from adaptive complexity (based on the authors’ own words) does not stand.
DeleteThis comment has been removed by the author.
ReplyDeleteRe: Gradualism. “Thus transitional forms, even if evolving over very long timespans, are unlikely to appear in the fossil record until they reinvade the ancestral territory; it is only the invasion that is sudden (see, e.g., Ayala, 1983; Dawkins, 1986; Mayr, 1982; Stebbins and Ayala, 1981). In any case it is clear that evolutionary change.”
ReplyDeleteConceptually, a problem emerges here. If it is the case that transitional forms are unlikely to appear in fossil record, then it begs the question: what evidence exists to support the assertion that the invasion is sudden? A lack of evidence cannot be explained away with ‘sudden invasion’. What is the evidence for sudden invasion if not fossil record?
This has a significant implication on the argument for language as evolved. If it can be shown with evidence that language resulted from non-gradual change, then the explanation of adaptive complexity fails.
Re: Exaptations. “In response, Dawkins (1986: 81) writes: "An ancient animal with 5 per cent of an eye might indeed have used it for something other than sight, but it seems to me at least as likely that it used it for 5 per cent vision. ... Vision that is 5 percent as good as yours or mine is very much worth having in comparison with no vision at all. So is 1 per cent vision better than total blindness. And 6 per cent is better than 5, 7 per cent better than 6, and so on up the gradual, continuous series."”
More conceptual problems in this case as well. Again, referring to how language may be the result of non-gradual change, there must be evidence to show that it is or it is not. The problem of exaptations, again, cannot be explained away using arbitrary numbers. What evidence is there that 5% vision is “as good as yours or mind”? Furthermore, if 5% is used for “something other than sight”, then what is it? The scientific onus lies on both ends of the argument. A lack of evidence does not mean language is the result of non-gradual changes. However, without evidence to support that language is the result of intermediate evolutionary stages, then adaptive complexity is pure conjecturing.
In addition, it is not clear as to why these two issues are seen as independent of the role of selection in evolutionary change. Of course it matters since if language evolved in non-gradual changes, then it means that adaptive complexity falls apart.
Similar to the problem we have with evolutionary psychology, I find it difficult to to draw the line between reasonable theory and absurdity. This article is an example illuminating that boundary. In Pinker’s discussion of negative evidence as non-“crucial” (as Pinker himself calls it) for language acquisition, he draws on two points to support his argument; that parents seem to not give their children helpful feedback or any feedback at all, and that a case of a mute child (who could therefor not receive feedback on his verbal utterances) still had perfect grammar comprehension.
ReplyDeleteOn the first point; a parent’s choice to omit authority, of correcting their child on their speech, may very well be different in the lab for a variety of reasons like not wanting to embarrass their child. The results in the lab do raise reasonable suspicion but the fact is that parents do correct their children, it is not uncommon to observe this publically.
The second point seems irrelevant; a child does not necessarily need to receive negative feedback on a personal/direct level to learn grammar. Being mute is an exception to the rule, most children can speak, so the mute child can learn from other children being corrected by negative feedback.
The two points say little about whether negative evidence is necessary or not. They seem to be biased conclusions or, “just so” stories in favour of the argument of in-born capacities (which would be fine if the arguments were sense-worthy!).
What do those people arguing against the evolution of language through natural selection claim as the origin of language?
ReplyDeleteIt seems obvious that language must have evolved somehow, as the poverty of the stimulus argument shows that universal grammar cannot be learned and therefore must have an innate (i.e., genetic) basis.
It also seems obvious that language would confer a massive fitness advantage to those who have language over those who don't. Therefore language-users would be selected for over non-users.
So is the argument over how language initially appeared out of non-language? That's certainly an interesting question, but whether or not language is an evolved and selected-for trait seems like a given at this point, even before having read this article.
This comment has been removed by the author.
DeleteMichael, you are right that the evolution of language through natural selection seems most intuitive, especially considering Pinker and Bloom's undeniable arguments that "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," but because this argument does not necessarily answer our how/why question, I can see why this is still a topic of debate. The real question is about the origins of Chomsky's Universal Grammar, which professor Harnad points out is the 'harder problem in linguistics'. Our class discussion really put this how/why question into perspective when professor Harnad questioned why primates, a species so close to us, did not also evolve language capacities. Was it because they simply did not feel the need or was it some other process that occurred during evolution? These are the unknowns that evolutionary and language acquisition theories cannot explain.
DeleteIt seems clear that no one actually has a solid origins of language story. One of the many issues is that though language users would have had a significant evolutionary advantage, that doesn't tell you about how language came about in the first place. Evolutionary theory, strictly speaking, is agnostic about how traits come about, and is only concerned with what survives and doesn't survive (i.e it's "blind").
DeleteAs for UG, I'm still confused. The fact that there are no simple languages and no simple syntax shows us that language could not have evolved from simpler languages, which then leads people to argue that UG must be genetically coded. But that means that it must have come about in one go, through some kind of monster-mega-mutation, which seems highly improbable.
Why can't UG just be the expression of the convergence of multiple constraints, such as cognitive/computational demands on memory, constraints on transmission and efficiency, and other biological constraints? Isn't this more plausible than the genetics story?
Your comments reminded me of these specific parts of the text: “If a current theory of language is truly incompatible with the neo-Darwinian theory of evolution, one could hardly blame someone for concluding that it is not the theory of evolution that must be questioned, but the theory of language.” & “doubting whether an innate generative grammar could have evolved by natural selection.”
DeleteI am curious to know given that we cannot answer what has been the adaptive value of UG compliant languages, is it plausible to think that there will be an evolutionary account for UG but we just haven’t found it yet? If not should we change the framework of applying evolutional theories to UG or change our perspectives on UG as the required capacity for natural languages.
This was a very interesting and compelling paper written in a kid-sybly style with an excellent sense of flow. That being said, I feel like the paper could have been organized differently; it could have been divided into discussion of the evolution of language, and an exploration of how it is that we acquire language.
ReplyDeleteEvolution
Evolutionary theory states that when a design for a certain function is complex, for example, let’s take wings in insects. Wings in insects have two functions, garnering energy from the sun, and locomotion. Since garnering energy from the sun is a more complex function, we say this one has the evolutionary explanation, and locomotion is deemed an ‘exaptation’ (a spandrel to use Jacobs’ analogy.) Grammar is very complex; “grammar is a mechanism tailored to the transmission of propositional structures through a serial interface (Harnad)” Since grammar does not pertain to a more complex function than the function of communication, we say it’s existence has an evolutionary explanation. The only good explanation for the origin of language and other organs of complex design (like the eye) is the process of natural selection.
As to the question of whether grammar evolved along the lines of a “punctuated equilibrium" vs. more gradually, we can say that it probably was more gradual, because we have no event to point to that would cause this innate ability to come about. Just because we have learned to use our language abilities to a much fuller extent than cave people, does not mean that our potential for language has not been available to us for hundreds of thousands of years.
Language Acquisition
Convincingly, Chomsky’s thesis of Universal language states that we are born with an inborn mechanism to learn language. Since all human societies have language, and all of our languages are predicated on particular grammatical rules (mentioned in 3.1), we can deduce that our propensity for language should be treated as an innate ability, like echolocation in bats. Since there is no negative evidence in children, (children do not make enough grammatical errors to deduce that they are learning by induction alone), we infer that their inborn learning algorithms help them make sense of all the language data that surrounds them. Children learn sentences, not rules, so they must have some natural prosperity to learn the rules. Therefore, the ability to learn language should not be treated like something we invented on the Rift Valley and then passed on to future generations, but as an ability pre-programmed by nature.
Thanks for clarifying parts of the reading Lauren. However, you posit that grammar's evolution was probably gradual rather than according to "punctuated equilibrium", meaning grammar evolved in bursts. The Pinker and Bloom reading qualifies punctuated equilibrium a bit more: it says that these bursts often take place in smaller populations and then these populations invade larger ones. The fact that the initial burst that might have birthed grammar was small scale might be why we cannot pinpoint an event that caused grammar to come about.
DeleteYou cite Pinker and Bloom in that more complex features such as insect wings storing energy from the sun are more likely to show that its design indicates purpose. However, you claim that grammar is such a complex feature whose design and function might tell us about how and why it evolved. However, while this may be true, grammar might not be sufficiently complex and could be a byproduct of some other feature. Pinker and Bloom say that grammar is in fact poorly designed for expressing/communicating subtle emotional patters, tone, facial expressions. Moreover, even though grammar is useful for conveying geographic relations between objects, it is not well designed for precise Euclidean relations. To conclude, I think it would be premature to say that grammar evolved for the transmission of propositional structures because it might be a byproduct of some other feature.
I think that the argument here for punctuated equilibrium is a bit more compelling here, but, Shanil, your explanation veers a little too close to the idea that language started in one population and then spread, which P&B discredit in the very beginning of the article as something not even worth considering. I think that on a small scale, grammar evolving along the lines of a “punctuated equilibrium” makes sense for how something like language tends to spread throughout populations, even today. I would draw a parallel to the invention of things like the lightbulb, or the telephone, in which history has shown us that different inventors in different communities came up with the same invention independently from each other. We can probably attribute this to the two inventors having a common problem that the solution was meant to solve, along with having common tools with which to solve this problem. We can see the evolution of language in the same way.
DeletePinker and Bloom argues that human language is evolved by natural selection, because of its signs of complex designs for communication of propositional structures, and that the seemingly only explanation for the origins of organ with similar complexity is the process of natural selection.
ReplyDeleteThe paper first talks about the natural selection in the context of explaining the evolution of some traits. This is where they give convincing arguments about the development of the function eye analogous to the becoming of human language. This as the human trait is shaped by natural selection, allowing the development of language itself and the enhancing of the behavioural expressions to eventually gaining more reproductive advantage.
They also come to the conclusion that the human language is partly innate and partly nurtured/learned, such as being able to learn new categories via words and propositions. However, this is where their argument becomes a little bit questionable when they started to talk about Universal Grammar (UG) and its evolutionary advantage, without fully discussion how it would evolve.
Because of the poverty of the stimulus, without error or correction, we know that UG is not learnable but innate. So it seems like there is no genetic propensity and capacity to learn UG – going back to wonder how, then, it could be evolved. It seems that Pinker and Bloom didn’t really addressed this, or at least directly as they only discussed extensively about the language functions, what could be learned, and what could be evolved, and seemingly overlooking the problem of UG’s evolution.
So perhaps, as Steven says in class, that Pinker and Bloom did not fully understand the properties of UG and the problem it poses when talking about its evolution.
Pinker and Bloom’s paper in honesty, was not my favourite. Maybe it is because they prefaced their entire argument with the fact that it is boring. None the less, I can recognize the importance of evaluating the evolution of language and the diverging opinions regarding it. One thing I had trouble reconciling was Chomsky’s assertion that our cognitive capacity has enabled language as we know it and that our brain capacities arose out of a series of adaptations. On first glance, this does not seem to contradict Pinker and Bloom’s position in which the human language faculty can be explained by natural selection. Evidently, their divergent opinions are nuanced because Chomsky does not dismiss the mechanism of evolution completely when it comes to language. I am not sure if I am grasping this correctly but I think Chomsky is proposing an indirect evolutionary effect (by product) in that traits are favoured through natural selection which can affect our cognitive capacity (e.g. increased brain size). As such, cognitive capacity in general is responsible for the human language faculty, even though it (cognitive capacity) has been shaped by traits subjected to natural selection. Is this correct? Am I close?
ReplyDeleteRE: Inherent tradeoffs
ReplyDeletePinker and Bloom raise good points in discussing Slobin's writings on the constraints of serialized language production and the sorts of tradeoffs that necessarily occur. Slobin's example of the varying inflectional rules between the Serbo-Croatian and Turkish language systems adheres to that idea of a somewhat arbitrary set of language features (that adheres to the general needs).
The balancing between economical production (brevity) and comprehension is analogous to something of a predator-prey relationship graph (https://bhsscienceinquiry.wikispaces.com/file/view/PredPrey.gif/131125521/384x288/PredPrey.gif) where the acquisition of linguistic traits favouring reduced vocal effort accrues at the initial cost of the opposing trait (in this case, the traits favouring clarity) before the need for better comprehension becomes great enough to reverse the trend as a greater breadth of distinctive linguistic changes are incorporated.
It's no coincidence that this pattern of language change over time can be (loosely) likened to animal population charts, reflecting the same sorts of additional linguistic adaptations that may have been crucial for the development of language as a whole.
I think the spandrel analogy is really useful and cool to describe non selective evolution. It made me wonder what other aspects of cognition might be considered an evolutionary spandrel? For example, it seems hard to argue that something like “feeling happy” could ever not be an spandrel, as it does not appear to provide any evolutionary advantage. I don’t think however that this can be applied to all feelings in general, as fear seems to definitely be a trait that would be selected for. So if there are some feelings that are evolutionarily advantageous, it seems the ability to feel would be a trait that was selected for. I think this relates in an interesting way to the easy problem.
ReplyDeleteI might be completely mistaken in thinking this, but I believe that physiological states that underlie emotion, such as fear or of happiness, are clearly understood in that they are the result of environmental factors and modify an animal's behaviour in an advantageous way. As you mentioned, fear more obvious in it's purpose, such as causing an animal to hide from a predator, but it would also be easy to argue for the advantage of an animal being in a happy state. For instance, maybe they'll be more receptive to mating.
DeleteActing upon these emotions seems to fit well into an evolutionary framework, as well as the easy problem of how and why animals do what they do.
The actual "feeling" of fear or of happiness, the hard problem, seems a lot more difficult to explain through evolutionary selection. I think this sensation of feeling is very possibly where the spandrel would lie in this scenario (if there is a spandrel here at all!).
I agree Lucy, spandrels are super interesting to think about with regards to cognition and even emotions. We tend to want to come up with logical evolutionary explanations for feelings or behaviours (eg. men are more likely to cheat because infidelity supports passing on more genetic material to off-spring) but a lot of these explanations far oversimplify the deeply complicated and simultaneous process of natural selection.
DeleteKylar I think you're right in some regard that evolutionary psychologists have made efforts to pin down environmental factors that would lead to changes in behaviour, but a lot of these are theories and some contradict each other (eg. oxytocin as a panacea). But I like your point on the hard problem, sounds similar theories of cognition that present consciousness as an accidental symptom of cognitive processes. The idea that it "accidentally" arose would mean that there wasn't a specific mechanism for feeling, it didn't give evolutionary advantage, but came about because of other traits selected for. It's a nice theory I think, but kind of useless to actually answer any question about the hard problem. But, a nice way to come to terms with the apparent uselessness of consciousness.
Even though Pinker shrewdly dismantles his points, Gould’s argument that the mind is a single general-purpose computer has given me pause. “…If many of the specific things it (the brain) can now can do are the product of direct selection “for” that particular behaviour… once you build a complex machine it can perform so many unanticipated tasks. Build a computer “for” processing monthly checks at the plant, and it can also perform factor analyses …”
ReplyDeletePinker makes an excellent point in saying that this cannot be completely true, because someone (or in our case, nature) needs to reprogram the computer (or brain) to perform new complex functions, I still think Gould is tapping into something interesting.
I listened to an episode of Radiolab called “Batman”, where they tell the story of how a blind kid basically learned to echolocate via induction. Instead of going to a special school for the blind like most kids do, he taught himself to make clicks as he navigated the world to aurally sense his location relative to other objects. Via induction alone, he learned to ‘echo-locate’ so well that he could ride a bike. He said in the podcast that he has a full mental representation of the world around him, and navigates his world with ease. He currently helps blind children using his techniques, but has found that often children cannot learn “to see” this way once they have passed a certain age (the critical period.)
Gould’s general-purpose computer argument made me think of this story for a few reasons. First, it seems like we may have more inborn learning algorithms than we are aware of. If our environment is right, it channels our learning. Our environment is a programmer, in a sense. If we are blind, our auditory cortex takes over our visual, and we learn to navigate. If we are born into a speaking family and have the physiological faculties for verbal language, we learn to speak. If we are blind, and we go to a school that tells us we are handicapped, we will not learn to navigate our world independently. We are general-purpose computers with many learning programs available to us. Our environment, and sensory motor capabilities dictate which ones are most important to us.
Over time, we have adapted (via Baldwinian evolution or natural selection or both) to learn to quickly learn to use language. Over generations we have become so efficient at it that we have a universal grammar. While this is mystifying and amazing, I posit that if our environmental conditions had been different in our evolutionary past, different learning programs would have come to the fore.
I found the reading a bit too lengthy in addressing that human language is evolved by natural selection just like other specialized biological systems. I particularly like the part that mentioned the natural selection is the cause of the vertebrae eye. Even though it is quite clear to us that language and eye are both evolved by natural selection, the difference between eye and language, in terms of their structure and nature, are way too different in a way that people may not aware how similar the two could be, in related to natural selection and evolutionary significance. I think the reading did a great job in clarifying that.
ReplyDeleteRE: While there are no doubt aspects of the system that can only be explained by historical, developmental, or random processes, the most likely explanation for the complex structure of the language faculty is that it is a design imposed on neural circuitry as a response to evolutionary pressures.
ReplyDeleteAre Pinker and Bloom suggesting that the evolutionary pressures can only explain how the brain developed the “capacity” to learn/comprehend language rather than the development of a language itself. Based on lesion and stroke studies we know the brain processes discrete aspects of language in different areas. For example, patients with Broca’s aphasia exhibit deficits in their ability to produce syntactically sound sentences. Could this suggest language developed in discrete package (the ability to distinguish a warning grunt from an aggressive one) and gradually became more complex. Like Dawkins stated, 1% of vision is better than no vision at all.
Alex- I think Pinker and Bloom go further than just hypothesizing that natural selection gave rise to our capacity to learn language; if they had simply stopped there, then I doubt there would be as much disagreement between them and Chomsky as there is, since the latter concedes that our capacity for language could have come about vicariously through the evolution of other adaptive traits. Instead, Pinker and Bloom assert that language itself, as well as our capacity for language, both began through natural selection, refuting the idea that we’ve had language and language capacities all along but rather developed them.
DeletePinker argues that language evolved by natural selection by assuming that because of its complexity, the only plausible explanation for such complex design is natural selection. If natural selection was the case, there would have been intermediate steps leading from no language to the language in its present form but no conclusive data exists which would strongly bolster this Darwinian account of language. The fact that we cannot find any linguistic antecedents in closely-related species, or significant resemblance to human language or acquisition shows absence of evidence for the gradual continuity and evolving that would have had to accompany a biological explanation of natural selection. Therefore, although Pinker does a thorough job in creating a window of opportunity or chance possibility for natural selection to serve as an explanation, I don’t think he has succeeded in convincingly demonstrating how natural selection is the most plausible explanation due to his assumption that natural selection is the default explanation, and only the other accounts of language have the burden of proof.
ReplyDeleteI agree with your claim. I found that Pinker made a lot of conclusions based off of little/no evidence. The truth is human language is very complex, and we understand very little of it. The conclusions drawn in this article, I found, were more like a philosophical train of thought rather than scientific insight.
DeleteFor example this conclusion " the only explanation for the origin of organs with complex design is the process of natural selection." Not that the claim is baseless, there is just not enough evidence to make such a big claim.
Delete"The fact that we cannot find any linguistic antecedents in closely-related species, or significant resemblance to human language or acquisition shows absence of evidence for the gradual continuity and evolving that would have had to accompany a biological explanation of natural selection"
DeleteWhile it is undeniable that evidence of language in closely related species is absent, I do feel that the paper did do a good job in addressing why this evidence is absent. Pinker raises the fact that chimpanzees and humans shared a common ancestor 6 or 7 million years ago. This time span allows for hundreds of thousands of generations over which language could have evolved through small individual genetic variations.
Pinker and Bloom present a good overview of the discussions surrounding language evolution. In refuting the claims that language is incompatible with Darwinian theory and that language is merely an exaptation, P & B argue that human language’s “adaptive complexity” can only be attributed to natural selection.
ReplyDeleteIn the section, The Process of Language Evolution, P & B say that for UG to have evolved through a process of natural selection, “there must have been a series of steps leading from no language at all to language as we now find it,” with each step constitutive of an “intermediate grammar” that was selected for based on its reproductive advantage. The authors use immigrants, tourists, aphasics and children to support the existence of grammatical intermediates. The authors also mention that linguistic rules are categorical, in that a sentence is either grammatical or not – it is an all-or-none decision; yet these rules developed in a gradual process. On this point, I wonder if the authors may have misunderstood universal grammar, as they seem to have conflated UG with ordinary grammar (OG). Unlike OG, there is no variation between individuals because UG is an innate category, and any variation between individuals would be a consequence of learning via trial-and-error (a process that only affects OG). I guess I’m confused about the idea that, if language and UG are categorical, how can one reconcile that language is all-or-none with the idea that it evolved gradually (i.e. that there were once “intermediate” languages)?
I do not think that language has to be either an adaptation or a spandrel, because I don’t see these possibilities as mutually exclusive. It is almost certainly true that language abilities developed partially due to increasing complexity in brain functions, however it also seems unlikely that they simply arose by “chance,” due to the emergence of more neural connections. Although the brain has certain “language areas,” these are not solely devoted to the production of language, meaning that, if the abilities to understand and produce language are an independent adaptation for a specific function, all other functions that these brain regions participated in would have to be spandrels.
ReplyDeleteI also don’t agree with the argument that people vary in their grammatical abilities, intended to disprove Lieberman’s argument that syntax is acquired by learning abilities and not brain modules. I don’t think the proper debate is regarding whether language abilities are completely learned or completely innate, but whether there is a universal grammar that is innately present through some method other than natural selection. Although Pinker and Bloom provide examples of how people vary in their language abilities, it seems to me that they are only talking about the specific, learned components of a language (like vocabulary) or other abilities unrelated to language (like the use of creativity in speech), not the UG components like how a syntactic tree changes when a person produces a question rather than a statement.
Lastly, I think the presence of pidgins and creoles is an interesting argument for the presence of an innate universal grammar. I believe these would be useful to study, in order to learn more about humans’ innate language abilities. This is because they may be the only example of people using the rules of universal grammar, without any exposure to these rules (since the parents are not fluent speakers of this language). It begs the question whether children who became fluent in the new creole were also exposed to another language spoken fluently by a parent, or whether it is truly possible to use these rules without any learning of UG taking place.
Chomsky believes it is unlikely that language is a product of natural selection. While I was open to this opinion, I was left unsatisfied with his justification. He says that language could easily be a side effect or a by product of other evolutionary forces, such as an increase in overall brain size or unknown laws of structure or growth. If an increase in brain size is what allowed for the development of language, this still does not explain WHY language developed. As the human brain began to grow larger, there must have been a subset of humans living with greater brain volumes who had still not developed language. In this sense, Chomsky is saying that language might be a by product of an evolutionary process - which is most likely explained by natural selection. If my understanding is correct, then isn't Chomsky's explanation somewhat contradictory? Or am I missing something?
ReplyDeleteIn discussing how language evolved from gesturing and pointing to propositions and referents in humans, I feel like it’s fair to say that humans developed these capacities because of the requirements of more complex living arrangements. Hunting and collecting food, building huts, trading etc. all required humans to be able to cooperate effectively. Obviously, cooperation is made easier if communication is efficient. Essentially, language might have developed out of necessity. Is it possible that much like universal grammar is innate, apes and other non human species might have the capacity to develop language, but might be lacking the necessity to do so? Although, this theory does not necessarily address HOW humans developed language, or why non human species would not benefit for more sophisticated means of communication.
ReplyDeleteThere is a general consensus that language developed at a time when our ancestors really demonstrated aspects of much greater intelligence than those of the animal kingdom thousands of years ago. Though, alongside cerebral development was the development of the head and the jaw, allowing for the production of speech sounds. If we look at our close relatives, chimpanzees, there are many similarities between us, though a major difference is that of language.
ReplyDeleteMany believe that language has to do with the evolution of intelligence but, given that language is supposedly innate and that it’s what separates humans from animals, was language a by-product of evolution or could the evolution of higher level intelligence have been the result of language? If this was the case and language was not innate at some point, how did it come about? If it was always innate, why do those who miss the ‘critical period’ have such a deficit in some cases?
P&B’s article examines and then refutes the Chomskian notion that language is incompatible with Darwinian evolution, or that language is an exaptation/co-opted trait. They feel language shows: "Adaptive complexity (which) describes any system composed of many interacting parts where the details of the parts' structure and arrangement suggest design to fulfill some function”. Chomsky’s account of language is largely dismissive to evolution as a mechanism, which I find to be limiting and unjustifiable in ignoring a mechanism worthy of scholarly consideration. P&B refutation allows for serious consideration of the mechanisms of language past that of just ‘nature’. However, there is a lack of evidence on their part for the evolutionary predecessors of language like we see in other adaptive traits, like the vertebrate eye where its most basic component photoreceptors emerged over 600 millions years ago. So it seems clear that both ideas about the origin of language have merit, and yet neither seems to adequately address its emergence. I imagine there must be a way to reconcile statements that appear to be at the opposite ends of the spectrum, where language emerged in part through adaptive changes but also can potentially be attributed to co-opted physiological/behaviours.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteI really appreciated the spandrel analogy in this reading. Whether or not natural language arose from natural selection is a complicated argument, and I think Pinker and Bloom made it easy to understand. I especially appreciated the description of the human eye as an example of something that was clearly designed for the precise purpose of vision. The human eye is so complex, that vision could not be a by-product of some other purpose for the eye. The example of the engineering of a TV was also helpful to me. A television set could not have been engineered to do anything else besides to be a television. Yes, analyzing post-hoc whether or not something is a product of evolution can never be proven, but Pinker and Bloom’s detailed explanation of WHEN one can credit natural selection and when one should not made total sense to me. Natural language is so complex, that it could not have evolved for a purpose other than making propositions. I was convinced by their argument that UG could not have evolved by any other mechanism.
ReplyDeleteThe authors write “the genome stores the vocabulary in the environment.” I am not sure what was meant by this. So the ability to learn language is encoded in the human genome, but the vocabulary is learned through trial and error? Is this what is meant by offloading?
ReplyDeleteHi Natalia, I think that that is exactly what is meant by offloading. It's presented in a strange way, genomes don't "store" information in the environment of course, they only have data in their own genetic code. But off loading takes advantage of predictable patterns in the environment such that the genome doesn't need to have the stimulus hardcoded into itself. There is no gene that tells you how to detect apples, categorize, or use the word "apples" to denotes the red round fruit, but the gene is coded to allow you to learn that the word for apples indicates the category apple. Similarly the ability to speak is not hard-coded into our genes (or feral children wouldn't be mute). Human babies are raised by other humans who do speak, allowing us to make vocal formations from the audio information naturally presented to us in our environment. The subject-predicate formation ability is what is stored in the gene, the apples and verbal speech are "stored" in the environment.
DeleteTo follow up on Cassie's explanation, another example of offloading or 'lazy' genes is the fact that organisms don't encode or pass down what their mothers look like. If you think about it we aren't born with the innate capacity to pick out members of our species, rather that is learnt. We assume that the thing present at birth will take care of us and we learn over time that the thing is called 'mother'. We can see examples of this in cross species raising where a cat can raise babies or another species. Vocabulary is much like this as we are born with the capacity to learn language and make the necessary sounds but the actual words and classroom grammar are learnt by trial and error through the environment, not our genes.
Delete. I don't think that language is a spandrel or the side product of some other genetic adaptation. Language seems too specific and colossal in nature while it also a backbone that is common among all languages. I am kind of confused about the point of this article because I don’t think anyone ever doubted the importance of language’s effects on evolutionary acceleration, social interactions, technology, reproduction etc.
ReplyDeleteRE: "Languages are complex combinations of elegant principles and historical accidents"
ReplyDeleteI think something important to note in the evolution of language is the abundance of weird, inconvenient irregularities in them. When sounds or whole parts of words change in languages, it can lead to inconsistencies that don't make logical sense. These are then levelled (the odd word is somehow changed to make it fit into the pattern) to make the language clearer, but this interplay continues and all of the irregularities are solved. I think this accounts for why there is no one language, and why languages don't become more similar as they evolve. They aren't able to evolve into a "perfect" form of communication because whenever a change occurs it has a ripple effect making a bunch of other problems. This also accounts for the vast diversity of languages there are in the world.
(On the topic of expanding grammar rules, like saying cutted) - “Why do children eventually abandon this simple, logical expressive system? They must be programmed so that the mere requirement of conformity in the adult code, as subtle and arbitrary as it is, wins over other desiderata.”
ReplyDeleteI'm not sure I agree with this point, that this particular part of grammar must be programmed - more that it seems children are programmed to pick up on rules and use them overall, but when they notice that no one else uses those specific instances, they instead notice they are doing something wrong. This would mean that there would be a universal tendency to learn rules but also that social factors can override this. I do not agree that these social factors overriding is programmed, though. I think instead that humans have a tendency to want to fit in and conform to the behavior of others and that this is certainly not a linguistic-specific genetic coding.
This paper as a whole seems to base it's argument on the lack of evidence for its counter argument. For example, the authors say "A given feature of language may be arbitrary in the sense that there are alternative solutions that are better from the standpoint of some single criterion. But this does not mean that it is good for nothing at all!"
ReplyDeleteBut my problem with this type of argument is that - wouldn't you have to explain why it is good for something at least? this particular passage seems to avoid the question and just say "well everything must serve some function of some sort, but we won't get into the specifics." And again, more generally this argument relies on saying something along the lines of "well you can't prove that language came about from evolution, so it must come about from evolution." There is a tendency to take the absence of evidence as if it were positive evidence for their theory.
Pinker and Bloom argue that language is a trait that has evolved through natural selection because language fits the criteria for such traits. In doing so, Pinker and Bloom are contending with arguments from Chomsky that language is likely to have evolved as a result of other physical properties or evolutionary processes rather than one of natural selection. Ultimately I believe that Pinker and Bloom are right to reject alternative explanations and to propose language as having evolved through a natural selection mechanism of evolution. Language is a complex means of fulfilling the function of communication between individuals and involves the transmission of propositional structures. Since language is so specialized in its structure, similar to an organ such as an eye, is unlikely to have emerged randomly or spontaneously from genetic mutation. By rejecting other implausible non selectionist evolutionary explanations we must conclude that language evolved via natural selection.
ReplyDeleteOf course notably the fact that language having an innate, biological component does not negate that there many aspects of language which must be learned through interacting with the environment. Evolution does not encode all aspects of language and infants must be taught words and rules specific to the language spoken in their environment.
As well, I was happy to Pinker and Bloom contest ill-formed evolutionary explanations that are essentially “just so” stories (e.g. ignoring bilateral organization of human body and proposing we have two breasts because humans sometimes have twins).
"All we argue is that language is no different from other complex abilities such as echolocation or stereopsis, and that the only way to explain the origin of such abilities is through the theory of natural selection."
ReplyDeleteI’m somewhat confused by this. What are the authors referring to here – language itself or linguistic capacity? It seems silly to say that language itself is like (or could be developed like) echolocation or stereopsis – There is something fundamentally different.
Take the example of a child raised without adequate exposure to language. As far as we know, children like this simply do not develop language, and, as far as I am aware, they also don’t develop something akin to language or something language-like of their own invention to compensate for the lack of exposure to language. This child would still have stereopsis, but no language, despite having had the evolutionary 'natural selection' process as humans who do have language.
Or, take for example a whole family or group of people that somehow did not have exposure to languages that exist in the world. Would they develop a system of communication amongst themselves using universal grammar? We don’t know, but this strikes me as unlikely.
Ultimately, it seems like there is something environmental and learned about language that would not be the case for something like echolocation and stereopsis. I am not sure that explaining language through natural selection alone accounts for this.
• When looking at Darwinian evolution in regards to language, I wonder how it might explain the evolution of certain physiological features that allow for verbal language. Was some ancestral voice box just sitting in our ancestors’ bodies, not being used? Or was it always there but just being used to produce other noises, like shouts and grunts? I know that we don’t know and probably won’t ever know exactly what changed in our ancestry that allowed us, and not our cousins the apes, to develop verbal language and change that device to produce guttural sounds to produce what we know as words. Was it just some random mutation?
ReplyDeleteOn “mentalese”, I found this so interesting! I never really thought about it, but I know that people often say they have an inner monologue, to which many people attribute as our consciousness, but does it really “speak” in a language? You never truly need to finish a sentence in your head and you never really need to have actual words associated with your thoughts because they are your thoughts and you feel what your thought is. When you think about something, you can verbalize it if you choose, but accompanying those words is some feeling or some state of feeling whatever it is you are thinking. I think this felt state is also what allows you to not need to finish the sentence or verbalize it at all and still know what you think.
ReplyDeleteRE: “Evolutionary theory appears to have very little to say about speciation, or about any kind of innovation. It can explain how you get a different distribution of qualities that are already present, but it does not say much about how new qualities can emerge.”
ReplyDeleteI’m not entirely sure if I agree with this because why can’t new emergent qualities be the evolved conjunctions of variations of old qualities. If we want to look at this argument in the context of language, will the use of a symbol system without any iconic connections to referents be a new innovative trait?
We can find biological accounts for evolution of physiological capacities in speech production that have led to development of language but can we connect the biological accounts of genes to evolution of language as a capacity generated by our brains while we still have no answer to the easy problem? It seems to me that language is another thing that cognizers “do” and we need to solve the easy problem first before we can talk about language evolution.