Saturday, 2 January 2016

8a. Pinker, S. & Bloom, P. (1990). Natural language and natural selection

Pinker, S. & Bloom, P. (1990). Natural language and naturalselectionBehavioral 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.


  1. I couldn't access the link provided but I think this is the same paper (albeit the unedited preprinted version)

    1. Thanks Jordan. There are lots of versions online. Not sure why the Harvard one disappeared. I've switched the link to one that works.

  2. “Why is there more than one language at all? Here we can only offer the most tentative of speculations. For sound-meaning pairings within the lexicon, there are two considerations. First, one might suppose that speakers need a learning mechanism for labels for cultural innovations, such as screwdriver. Such a learning device is then sufficient for all vocabulary items. Second, it may be difficult to evolve a huge innate code.”

    Regarding the second point: If we already had the necessary words in our genetic code, would this imply that categories are innate and symbols somehow come pre-grounded? Neither of those conclusions make any sense to me. However, “innate” in the sense that I’m speaking of here doesn’t match up perfectly with Fodor’s view. It strikes me that his position is that “innate” is somehow deeper than evolution. Either way, I think that if we had evolved an innate code, humans would then be quite poor at identifying and reacting to (essentially categorizing) new, ambiguous, or otherwise unfamiliar stimuli. Genetic adaptation works far too slowly to be able to efficiently and appropriately react to new stimuli one by one. Rather, a general language ability and a sensorimotor grounding ability would make a system much more able to survive and reproduce.

    “Language learning is not programming: parents provide their children with sentences of English, not rules of English. We suggest that natural selection was the programmer.”

    It has been demonstrated that growing up in an environment with a richer vocabulary (in which the parents use less common words) will give the child a richer vocabulary. The grammatical abilities of the child are not profoundly affected either way. Natural selection seems to have given us all the ability to use grammatical rules, but the content of language does not seem to be innate.

    Regarding the grammatical rules that underlie language, Chomsky’s poverty of the stimulus argument still holds up in the context of evolution by natural selection (which is a good thing, because there’s a lot more evidence for evolution than there is for the “poverty of the stimulus” argument). What comes next is how to interpret the poverty of the stimulus. I find Pinker’s interpretation that the rules for language construction are an evolutionary adaption more convincing than Chomsky’s “physical laws” explanation, which doesn’t actually explain much.

    “Furthermore, in a group of communicators competing for attention and sympathies there is a premium on the ability to engage, interest, and persuade listeners. This in turn encourages the development of discourse and rhetorical skills and the pragmatically-relevant grammatical devices that support them. 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.”

    Pinker lost me here. I feel like this argument is hardly any different than the one prof. Harnad presented in class from the evolutionary psychologist who attended singles bars. Besides, even if we agree to Pinker’s premise, wouldn’t that mean that the personality trait of Introversion (which is pretty consistently shown to be 50% genetic) would have largely disappeared by now? There’s also a cultural bias implicit in Pinker and Symons’ claim because many cultures don’t value “gifted orators” in the same we way do in the West.

    1. You're right, Joe, the "option" of genetically encoding the meaning of all the words naming categories that are actually named (or potentially nameable in the future) (and even coding their names!) is just so absurd that it is hardly worthy of thinking about! I have no idea what made P & B air that silly speculation.

      Categories first have to be learned, perceptually (most are nowhere near innate, and could not be), and then the named ones need to be labelled with names. Why would every human on the planet have (1) learned exactly the same categories and (2) given them exactly the same (arbitrary) name?

      Evolution does not "like" to code more than the absolute minimum in advance, if it can be picked up from exposure and learning. That's just as true for organisms without language. But language is probably the most dramatic example: Why (and how!) evolve in advance all the categories you can learn, both from direct experience and from (grounded) verbal definitions? Language itself is what evolved, and language was what then made it possible to learn an infinite number of potential categories through (grounded) words combined in subject/predicate propositions defining/describing new categories.

      -- We'll be talking about the poverty of the stimulus in Week 9. It means (for kid-sib): children learning their first language cannot learn some of the grammatical rules (called Universal Grammar [UG] rules) by trial and error and correction (the way they learn all the rest of grammar) because children only hear (and speak) sentences that obey the rules of UG: They do make trial-and-error grammatical mistakes, but not UG mistakes.

      -- UG-compliant sentences are a category. You can't learn a category or its rules from positive examples only (as in the "laylek" example I described, of an "uncomplemented category"). You need to sample both the members and the non-members (and you need corrective feedback on which is which, so that you (or your brain) can detect and abstract which features distinguish the members from the non-members. So it looks as if the rules distinguishing the category "UG-compliant sentences" from non-UG-compliant ones cannot be learned because the child never makes or hears UG mistakes, let alone get corrected for them. That's the "poverty of the stimulus." Chomsky concludes from this that UG must be innate.

      Pinker not only does not solve the problem of the poverty of the stimulus, but he completely begs the question by not distinguishing the problem of UG from the many other properties of language, most of them perfectly learnable (ordinary grammar rules, vocabulary, etc.)

      And yes, you're right, Joe, that Pinker's and Symon's "explanations" of the origin and adaptive value of language are vacuous -- and actually beg the question: You'd already have to have language for its persuasive and rhetorical powers to become any kind of advantage. But the question is, why and how did language start? Certainly not because "it" was useful for persuasion or rhetoric! (In fact, you first have to ask more carefully, just what is the "it," i.e., what is language? It's a capacity that humans have: A capacity for what, exactly? And how and why did that start?)

      Reading 8b suggests a hypothesis.


  3. While it was somewhat frustrating to read a lengthy paper in which the author concedes right off the bat that what they are arguing for is boring the paper did do a good job of explaining the debate and the background for it. I was in agreement with most of the paper following the arguments against why language is an adaption designed from evolution and found the example of diversity to be the best evidence for adaptation, I got a bit confused when he introduced some of the arguments for language being a spandrel. My understanding is that this view is simply we had other faculties selected for and those evolved over time until at some point these capacities were able to hold language and thus we developed language as a result of other faculties developing and not because language was being selected for. One of the arguments for this was that if the brain was perceived as a computer then at one point the computation system became powerful enough to hold the data and abilities required for language. The counter-point to this was that a computer programmed to do one thing won’t just suddenly start running other programs.

    My response would be that with today’s computer programs we have seen neural networks and deep learning machines that will take input and learn to create categories and do things with that input as if it were creating new programs to work with. I could just be ignorant in what these neural nets and such do but my understanding is that if the brain operated like one then it would use the sensorimotor input it receives as a child and make sense of it all to the point where early humans would function as they would even without language. Then this computational power grew through some gene mutation and their neural net now had the power to make more complex connections and allowed for tenants of language to be learned and thus language is born. I’m not saying this is necessarily true, but I find the answer in the article to be lacking and non satisfactory. My theoretical answer doesn’t help me come to terms with the fact that if we assume UG is true then this is potentially present before children even leave the womb so my neural net theory can’t necessarily explain that. Language would still have needed to evolve to create a UG but this is where I get confused and boundaries begin to blur…

    1. Language and UG are not the same thing. Language includes vocabulary, pronunciation, ordinary grammar, sentences, communication, etc. UG is special. It's special grammar rules that all languages share. We'll discuss it in week 9. But see the poverty of the stimulus in the reply above to Joe. UG rules differ from vocabulary and ordinary rules, which all can be and are learned. UG can't be learned on the basis of the data available to the child (i.e., what other people say, what the child says, and correction). It's the data that are impoverished. No neural net could learn UG from those data either. (UG can be learned by teams of linguists across decades, but not by the child between age 1 and 5! And they learn it by generating and testing a lot more data, including, most importantly, examples of sentences that violate UG, so they can figure out what rules distinguish UG-compliant and UG-violating sentences, just as in learning any other category from examples of members and non-members.)

      And it's UG that Chomsky says is innate. It's for explaining the origin of UG that the "spandrel" metaphor was suggested (but it doesn't work).

    2. Why does the spandrel metaphor not work? Also, what can be considered the differences between human language and animal communication? Is it known whether other species of the Homo line also had early forms of communication through tell and not just show?

      Also with regards to chimps and their early ability to acquire some rudimentary language (i am thinking of Gua and Kellogg ) is this a form of UG as well that the chimps have ? Or what are they tapping into and why is it unable to continue to evolve after a certain threshold even when it is being pushed for (by the experimenter in this case)?

    3. Hey Naima -

      I took a class in evolutionary psych a while back, and we discussed cases of animal learning where they appeared to have really amazing abilities to do things like speak and do math (

      At least in the case of Clever Hans, it was not so much that he was a horse who knew math but rather that he was a horse who was conditioned and had learned a response so well of body langauge that he could appear to do calculations. Maybe this is a similar explanation to Gua, because I do not think that UG is something that exists outside of humans (but maybe I am wrong).
      I think that what made me intuitively always think cognition is NOT computation is this *special* thing about humans - that we are able to communicate in this way.

  4. Pinker and Bloom argue that language shows adaptive complexity (“any system composed of many interacting parts where the details of the parts’ structure and arrangement suggest design to fulfill some function”) and natural selection is the only explanation out there that can truly account for the development of this complexity (similar to the evolved complexity of our organ systems).

    On the other hand, Chomsky believes that our cognitive capacity, that which is infinite, has given rise to the complex language system we use. But I feel a contradiction here, and wonder if someone could clarify this for me/correct me if I'm wrong: given that Chomsky agrees that our brain capacities arose out of series of adaptations, why is the language system -- which is an emergent property of the brain -- not in any way adaptational? Or is Chomsky's position not black in white, in that he believes that natural selection can simply not account for ALL of the intricacies of the development of language?

    1. Hey Maya, I had troubling reconciling this contradiction as well. But after reading the Introduction section a few times and researching what constitutes a “neo-Darwinian” position, I think I was finally able to understand what they were getting at underneath all the heated rhetoric. The way I see it, your last sentence hit the nail on the head:

      Or is Chomsky's position not black in white, in that he believes that natural selection can simply not account for ALL of the intricacies of the development of language?

      You’re completely right, Chomsky argues that language is directly tied to our cognitive abilities. However, he doesn’t completely dismiss natural selection and the role of adaptation, because he believes, as you put it, that “our brain capacities arose out of series of adaptations.” Ultimately, natural selection favored particular traits that were necessary for generating a system capable of cognition. As Pinker and Bloom put it, Chomsky suggests that:

      language may not be the product of natural selection, but a side effect of other evolutionary forces such as an increase in overall brain size and constraints of as-yet unknown laws of structure and growth (page 3)

      Essentially, our cognitive capacities may be formed by traits created through natural selection, but it is the cognitive capacities that are responsible for language, not natural selection itself. Had these capacities been generated another way, they would still be able to generate language, excluding the need for natural selection entirely.

      the mind is composed of autonomous computational modules -- mental faculties or "organs" -- and that the acquisition and representation of language is the product of several such specialized modules (page 3)

      These modules create a complex system generating cognition, and cognition is what permits language. It just so happens these modules (and therefore cognition) were created by natural selection, but they didn’t have to be. Regardless of how you get cognition, cognition is what gets you language.

    2. However, Pinker and Bloom believe this is direct contradiction to their position:

      In this paper we will examine this position in detail, and will come to a very different conclusion. We will argue that there is every reason to believe that language has been shaped by natural selection as it is understood within the orthodox "synthetic" or "neo-Darwinian" theory of evolution (page 3)

      Accordingly to Wikipedia, neo-darwinism states that “evolution occurs solely through natural selection, and not by the inheritance of acquired characteristics resulting from use or disuse” ( Apparently even Darwin conceded that an individual’s use of particular trait throughout their lifetime influences the way that trait is passed onto their offspring. But Pinker and Bloom disagree. They believe patterns of language use had absolutely nothing to do with its evolution. Instead, those who could use language to communicate were able to survive increasingly complex environments, while those who could not, did not. Simple as that. This is also the reason we see the “intermediate steps” they discuss in Section 5.2 (page 38). Different universal grammars arose because they were required to survive in different environments. It’s just natural selection, nothing else, so cognition is entirely irrelevant.

      For a while, I had trouble understanding why these two views were necessarily incompatible. Natural selection alone may be responsible for language if natural selection alone is responsible for cognition and cognition is what creates language. I guess I was suffering from the same problem Chomsky is, which Pinker and Bloom mention in their Conclusion section:

      As a result language competence has been equated with cognitive development, leading to confusions between the evolution of language and the evolution of thought (page 50)

      What they’re saying is we need to forget about cognition all together. Natural selection is exclusively responsible for the evolution of language. So even if you were able to somehow generate cognition without natural selection, you would still have not created language. Meanwhile, Chomsky, believes cognition is exclusively responsible for the evolution of language, so if you did generate cognition without natural selection, you would have created language as well. Hope this helps!

  5. This comment is regarding the discussion from last class.
    First concerning distal and proximal causes of behavior, evolutionary psychology is said to be underestimating the consequence of our current environment (ex: language and culture) and then overestimating the influence of our original environment. During the class discussion, adoption as a problem in evolutionary psychology was briefly mentioned. For example, what justify adopting a cat? To take care of someone and to know that their life depends on you. The distal cause of “passing on the genes” cannot be applied here obviously. When it comes to adopting a child, a member of our own species, this principle can also be applied. But I think there is something more to it. I think human grant enormous importance to culture, beliefs, tradition, language, manners and virtue. (We also attain a moment in science where we can alter genes; maybe it is down regulating their importance). Culture variance seems to be unique and intrinsic to human. A domestic cat will behave pretty much the same way independently on which country he is living. Why is it then that makes us so acculturated? Is it that we have a superior plasticity potential? Or is it that our instinct has being altered by our growing capacity for language and categorization?

  6. “Our noses were made to carry spectacles, so we have spectacles. Legs were clearly intended for breeches, and we wear them." ... Yet evolutionary biologists, in their tendency to focus exclusively on immediate adaptation to local conditions, do tend to ignore architectural constraints and perform just such an inversion of explanation.”

    I appreciate this notion that we use our adaptations in a way that makes sense in our current environment, but they did not necessarily evolve for us to use them that way. We did not evolve noses to hold glasses. I think that this is a huge problem with evolutionary psychology today in particular. This argument reminds me of the conversation we had in class about sensible vs. ridiculous conclusions. Many people use our current environment to try to come up with a theory of how it became this way. The problem comes when they do not take into account how our society and culture has changed and influenced us, and how it is impacting their theory. We do not speak so that we can pick up people in bars, but we have learned to use our ability to speak to pick up people in bars.

    The distinction between why we’ve evolved to have our adaptations, and the current use of our adaptations is blurry in some cases, like language. The main question here is similar to the noses question: Did we evolve to have the ability to create language, or have we used this ability for language because it made sense? From what I understand Gould is arguing that we evolved something else, like a larger brain, and this enabled us to use language. This article is arguing that we have evolved specifically to have this ability for language.

    “It is absurdly improbable that some general law of growth and form could give rise to a functioning vertebrate eye as a by-product of some other trend such as an increase in size of some other part.”

    I think that this argument makes sense when we discuss eyes, but not the brain. Many parts of the brain functioning together are responsible for language, and while certain areas like Wernicke’s area and Broca’s area appear to be more specific to language, the other necessary regions may have adapted for other purposes. Our brain may have already been in the process of evolving for other tasks like working memory, planning, inhibition, etc and these regions contributed to the evolution of language as well. If you are asking if certain neurons that fire when we comprehend language developed just for that, maybe they did, but they also could have developed for more basic functions such as understanding patterns in sounds, like birds do when they sing, and we use those regions, plus higher order regions to design a language. I think that this article would argue that the circuitry necessary to produce language from these multiple parts must have evolved, and perhaps that is true, but I think that we should be careful not to underestimate the impact of other necessary non-language actions on the evolution of core adaptations, which also happen to be used for language.

  7. “Since the mere appearance of some nonoptimal feature is inconclusive, we must examine specific explanations for why the feature exists. In the case of the nonselectionist position espoused by Piattelli-Palmarini, there is none: not a hint of how any specific aspect of grammar might be explained, even in principle, as a specific consequence of some developmental process or genetic mechanism or constraint on possible brain structure. The position gains all its support from the supposed lack of an adaptive explanation.”

    Doesn’t grammar exist because we pass it on through teaching from generation to generation? When grammar was first created it wasn’t necessarily due to an adaptation, but could have been the result of a larger brain and the want for a better understanding in communication. I agree with Piattelli-Palmarini here, that while communicative aspects of language may have had an adaptive advantage, I don’t think that grammar itself necessarily evolved. We could still understand the gist of what people are communicating without perfect grammar.

    I find the “tradeoffs of utility” theory interesting. It says that the speaker’s goal is to be as brief as possible to get the information out, and the listener’s is to understand with the greatest clarity. However, I think that this article is treating the evolution of language as if it were the evolution of man to create language. Language itself may have changed to reach these fundamental goals, but that does not mean that our brain had to change or that we had to adapt. Adaptations happen over a long period of time, and while I am not a history major or anthropologist, I do not believe that we have had grammar itself for a long enough period of time for adaptations to have evolved allowing for a more precise grammar.

    1. I think there needs to be a critical distinction between innate universal grammar (UG) as Chomsky describes (the formal constraints that allow humans to have the capacity to learn/speak a language, which are based on the poverty of the stimulus argument and the fact that all language speakers converge on the same grammar despite being exposed to varying input) and learned rules of grammar that differ from language to language and are passed down from generation to generation (e.g. irregular verbs, prescriptive vs. descriptive grammar). So in the case of universal grammar, I believe that it has a highly specific function making it tough to say it is a spandrel (i.e. a byproduct of evolution of another function). On the other hand, some grammar rules are learnt through formal rules in society and in this way, gets passed down from generation to generation. That being said, I do not believe Pinker and Bloom give an adequate explanation of the origin of universal grammar. Specifically, I have having trouble with universal grammar fitting the criteria for natural selection in which the retention across generations is due to “increased changes of survival and reproduction” (6). Even though universal grammar seems to serve a specific function, I agree with you that communicative aspects of language have served an adaptive advantage (through survival value), yet I am not sure the same can be said for UG.

  8. 8a. “words encode abstract general categories and only by contributing to the structure of major phrasal categories can they describe particular things, events, states, locations, and properties.”
    The authors describe the uncontroversial facts about substantive universals, unanimously agreed upon by all theories of universal grammar. Major phrase categories are introduced as a combination of words, that are composed of a major lexical item (the building blocks of language, categories like noun, verb, adjective) and other words that refer to something in the world. Words are said to “encode abstract general categories”, and the authors claim that only phrase categories can refer to things in the world. This makes sense when we look back to previous categorization lectures. The word “dog” could be argued to refer to the entire list of breeds, specific animals in specific locations- namely the general category. However, when I say “my dog” (a noun phrase) I am referring to a specific referent in the world or a “particular thing”. But now I turn to a more ambiguous example- in our discussion of categorization, we said Renuka is a category. Renuka can refer to a more general category of all of the instances of the robot from different angles on different days, in different sweaters etc., but we also discussed that Renuka can be a single unique instance of Renuka at one particular moment. This point in the article could potentially contradict the idea of the word “Renuka” referring to a single referent in the world. However, could it be that Renuka is a phrase category (as there is no determiner so Renuka constitutes a grammatical NP?). Does this reconcile the potential contradiction between our discussion and the point of categories only being encoded by major phrasal categories?

  9. I am a little disappointed with this paper, and although they warn us ahead of time, I thought it was quite boring…

    I think I am mostly disappointed with the fact that they barely touched upon Universal Grammar (UG). Being an article on natural language, I would have thought they would have spent much more time discussing this and incorporating this in their discussions. I mean one thing that is particularly unique about us is that we are said to have UG – so shouldn’t we try to understand why we specifically have it & when did it first appear? But maybe trying to explain the evolution of language and the evolution of UG are two different things? But since the article didn’t really explain what is language – I guess I am a little confused with this issue.

    Something else that was puzzling for me is that I thought both suggestions for the emergence of language to be a little absurd. On one end, you have people arguing that language is a spandrel “by-product of some other adaptation” & the other side saying that language evolved as an effect of “unknown laws of growth and form”. I have no idea what the latter means and I feel like they’re both arguing that language “just sort of happened” – but what does that explain? I think that if one wants to take the approach that UG emerges as a spandrel – one needs to explain which aspect of UG is a spandrel & how it evolved.

    Lastly, (again following up on UG), while reading the article I couldn’t help but think about the poverty of stimulus – “natural language grammar is unlearnable given limited data available to children learning a language & therefore has to have some sort of innate linguistic capacity”. I was disappointed that the article didn’t mention this once & couldn’t help to notice that this article did not touch upon differentiating between learned grammars and unlearned grammars, namely UG.

    This paper has also left me with a few other unanswered questions, namely what is language and why do we as humans have it but not others? What makes us different than those organisms? Surely knowing where it came from & how it started is helpful, and I don’t find this article was convincing in that area.
    Perhaps that was not the goal of this article and I just completely missed the point, which is highly probable…

    1. I think that perhaps one of the reasons why UG was not mentioned in the article is because the authors seem to be rejecting certain views Chomsky has on the evolution of language, and thus were trying to avoid using any Chomskian terms.

      “Something else that was puzzling for me is that I thought both suggestions for the emergence of language to be a little absurd. On one end, you have people arguing that language is a spandrel “by-product of some other adaptation” & the other side saying that language evolved as an effect of “unknown laws of growth and form”. I have no idea what the latter means and I feel like they’re both arguing that language “just sort of happened” – but what does that explain?”

      I think you might have just misread what the authors state in the beginning!

      The authors are arguing that language evolved as a product of natural selection, and thus evolved through random mutations wherein having a ‘greater language capacity’ was beneficial and thus resulted in higher fitness.

      The other side, argued by Chomsky and Gould, says that language is not a product of natural selection. Rather, they say it could have evolved from either a by-product of another adaptation OR unknown laws of growth and form. (Language as a “side effect of other evolutionary forces such as an increase in overall brain size and constraints of as-yet unknown laws of structure and growth”)

      You are totally right in saying that both reasons seem to be something that ‘just sort of happened’. As a by-product of another adaptation, language would be something that came about as a phenomenon coincidentally linked to or arising from other adaptations, rather than being something that directly increases fitness and was selected for. As a product of an unknown law of growth and form, I think it would just be something that arises from as of yet unknown, mathematically explainable physical processes (which I assume will also be linked to changes in size and structure of the brain, thus also kind of a by-product of other adaptations?) :
      “The answers may well lie not so much in the theory of natural selection as in molecular biology, in the study of what kinds of physical systems can develop under the conditions of life on earth and why, ultimately because of physical principles.” (quote from Chomsky, 1988 in the paper)

    2. I agree with you for sure though that the article doesn't really 'explain' anything. It seems, at least to me, that the entire point of the article was just to assert that language is a product of natural selection, rather than what Chomsky and Gould hypothesize it to be, and not anything further.

  10. - I find it hard to comprehend the part the author talks about that categorical rules could be innate and passed on, and a parent with no grammatical rules at all could arise from mutation. Rules have no meanings themselves, so wouldn’t that suggest that rules of language are learned rather than innate? I thought the principle behind UG is that we’re born with a capacity to learn rules, not have the rules already. Albeit, we have an understanding/feeling of sentences that sound weird if we randomly manipulate constituents in a sentence, but would that be attributed to our innate structures, or the way we have learned rules? The author argues that learnable connections can become innate due to selection pressures, displaying the Baldwin effect, based on evidence confirmed by Hinton and Nowlan, but do those connections equate to inborn rules?

    - I’m a little confused by the author’s argument against grammatical devices as general-purpose tools. He argues variations in language correspond to differences in the extent to which the same specific set of mental devices is put to use, but not differences in the kinds of devices that are put to use. I agree with that, but wouldn’t that mean mental devices are built to have general functions? Maybe not a generalized structure as linguists though, but the mental device itself would still be a general-purpose tool?

    - This is slightly unrelated, but when the author brought up “language of thought”, it got me thinking about robots. If robots are capable of a language of thought (whether we actually have it or not), would it be considered as thinking? Since it’s not taking in external input, but can produce or not produce an external output. It’s kind of like a robot being able to talk to itself inside its head?

    - His argument about the need for arbitrariness and standardization helps answer my question on blog 5 about the need for polysemy, symbols with multiple referents, among human language.

    1. Hi Oliver,

      I’ll try to clarify a couple of things about the Universal Grammar.

      “I find it hard to comprehend the part the author talks about that categorical rules could be innate and passed on, and a parent with no grammatical rules at all could arise from mutation . . . I thought the principle behind UG is that we’re born with a capacity to learn rules, not have the rules already.”

      The Universal Grammar (UG) is actually a set of innate rules, not merely the capacity to learn such rules. Chomsky suggested that we possess such an innate Grammar based on the fact that children succeed in producing grammatically correct language without going through a trial and error process of rule acquisition. Indeed, there is a ‘Poverty of Stimulus’ insofar as the stimulus children receive during language learning is insufficient to define the rules that they nonetheless consistently comply to when they produce language.

      “Rules have no meanings themselves, so wouldn’t that suggest that rules of language are learned rather than innate?”

      You are right that the development of the UG poses a problem for the evolutionary account of language learning. Internal rules would seem to be behaviorally inert without the linguistic capacities that rely on them. So how could such internal rules be selected for? As Harnad has pointed out above, the authors do not provide a satisfactory answer to this question, so we need a new theory. But we should not immediately dismiss the idea that UG did indeed evolve.

    2. Hey Oliver,

      Interesting question about the robots, however, I think the answer could be laid out as so... If a robot can talk to itself inside its head, and if a robot is just a bunch of machinery and connections (NOT neurons) just like a computer, but it is made to look like a human, then you would also be asking "can a computer talk to itself inside its head?" If something can talk to itself inside its head, then it is thinking right? Then that just brings us back to the question "Can computer/machines think?" which brings are directly back to the beginning of this course and to Turing. I think machines/computers/robots can take inputs and provide outputs, but I don't think they can actually THINK, at least the way that humans do.

    3. Thanks Timothy, for clarifying things up for me.
      Thanks Jordana, I understand my fault in that question now. It also seemed similar to a homunculus explanation, which was false so thanks for reminding me of the previous material.

  11. This comment has been removed by the author.

  12. In Natural Language and Natural selection, Steven Pinker and Paul Bloom suggest that the development of language was an evolutionary process. They compare the development of language acquisition to the development of echolocation in bats: this processes emerged as a product of natural selection and made animals possessing these properties more apt for reproduction and therefore more likely to transmit their genotype. The authors of this article give quite convincing arguments that language acquisition is an evolutionary process. However, I do not quite understand why if the development of language resulted from an evolutionary process, why does this make it so clearly distinguishable from ‘’writing or the wheel’’ that are processes that are thought, according to the authors to have developed from human culture rather than from human biology.
    I understand that in order to be able to write, you need a certain amount of stimulation.You need someone to show you the relationship between the written and the spoken work. You also need to memorize how to write words. However, it is not the same situation with language acquisition. Children need a certain amount of language stimulation in order to be able to learn how to speak. According to the poverty of the stimulus hypothesis, there is thought to be an innate ability to produce and interpret language as youngsters are not exposed to a sufficient amount of language and are also not told what are the wrong phrasal structures (according to the universal grammar) to produce the language that they produce. However, the same is true to writing, not all the words have been taught to every child that writes properly in his or her language. These children infer how to write certain words on the basis of how they write other similar words. Why cannot the ability to be able to write also be considered an innate process, just as the acquisition of language? It can also be said that the ability to better understand the grammatical rules required in spelling (it is important to note that it is not in all cases that knowing how to say something helps with knowing how to spell it).
    For the case of the wheel, it is often said that what allowed apes to become humans is the development of the ability to use tools. The invention of the wheel is the development of the ability to use an additional tool that could have also came from genetic evolutionary processes.

    1. Hi Anastasia,
      I think writing really cannot be amounted to being the same thing as speaking (and I think the authors agree with that?). The reason why language is so clearly distinguishable from writing is because one has to be thought, and the other doesn’t.
      With regard to language, it’s true that we require stimulus in order to acquire a specific language. But there is no “learning”, in the way you would learn a language in school. There is no need to practice speech, or an accent, or vocabulary… It all happens by itself, subconsciously; there is no such thing as learning a language. As you said, children aren’t systematically corrected when they produce a wrong structure. Yet they systematically end up knowing the correct one, without having learned it (and when they do get corrections they tend to ignore them). When you speak, you don’t need to think about how to structure your sentence, there is no wondering whether what you say will be grammatically correct or not, because it will be. No one “decides” what is right or wrong to say (except prescriptivists, but that’s another story), because there is no wrong way to speak. You can’t decide “not to speak”, it’s something that just happens.
      However when it comes to writing, it’s a completely different issue. As you said, writing has to be learned. I for one can remember elementary school where I would practice my cursive and try to make my writing look pretty. And spelling tests, where you’re told when you’ve written something the wrong way. Writing was something that was invented, that has rules that must be learned, and when they are learned then they can be applied in a logical way. This is how people can guess as to how something is spelled. It’s not an innate process, it’s reasoning and deduction, more like math. The fact that it’s related to language is irrelevant. Even today, as I write, I still wonder as to the spelling of some words that I have previously learned. But when I speak, there is no doubt or hesitation in sounding out a word. Moreover, writing is completely arbitrary. By this I mean that someone, one day, decided that a circle with a line going up was going to represent the sound /d/. Someone decided that the word “light” would be spelled as is, rather than “lait” (which, let’s face it, would make things a lot simpler). This is all visual representation of speech, thus must be learned.
      All of the writing related activities must be learned, practiced, inferred, and especially conscious. None of these things apply to language. The main and huge difference between language and writing is this: language is ACQUIRED. Writing is TAUGHT.
      (Also note that every non-impaired person can speak and understand speech, regardless of background, but not everyone is able to write.)

  13. “"Adaptive complexity" describes any system composed of many interacting parts where the details of the parts' structure and arrangement suggest design to fulfill some function. The vertebrate eye is the classic example […] organs of extreme perfection and complication”
    While this explanation and argument makes sense, I disagree with analogizing the eye and linguistics ability. Yes, the eye is an incredibly complicated structure, so much so that any object made to mimic some of its effects closely resembles its structure. Yes, the odds of this randomly happening are low, hence it supports the concept of natural evolution. I also agree that language is also incredibly complicated, so much so that as of this day the best programmed computer still cannot mimic it adequately. However, there is a core difference I feel the authors have ignored. And that is that while the eye and its mechanisms are all physical, language is not. The eye can be seen as a whole which is a sum of parts. Reproduce those parts, or find objects able to mimic the part’s properties, and you will get an eye/something able to “see”, e.g. a camera. This does not work for language. It is all in the brain. There isn’t a language equivalent of a “part that focuses light” and “part that registers colors”. We know that there are parts of the brain that are more focused on specific functions of language, e.g. Broca’s area for speech production. But first of all, it is not reproducible, and second of all, it does not inform in any way as which parts can tell the difference between a subject and a verb, or whether a subject is or isn’t dropped in a given language. All we know is the general area where language related activity occurs, but we know nothing about the physical complexity of the language mechanism. Therefore it’s functioning cannot be simplified to that of an eye (which has a straightforward complexity). One is physical, the other is not. In this way I don’t believe this argument adequately supports the hypothesis that language is solely the product of natural evolution.

  14. "While one might justifiably argue that an entire system of grammar must evolve in a gradual continuous sequence, that does not mean that every aspect of every rule must evolve in a gradual continuous sequence....No single mutation or recombination could have led to an entire universal grammar, but it could have led a parent with an n-rule grammar to have an offspring with an n+1 rule grammar, or a parent with an m-symbol rule to have an offspring with an m+1 symbol rule (Pinker & Bloom 2000).

    These questions are some of the more fascinating questions of language I've encountered. In PSYC 213, we learned a little bit about the poverty of stimulus argument and how universal grammar is innate. However, we were not exposed to the process. Grammar seems to be quite categorical in nature-the sentence is either grammatical or not. However, it is possible that rules do develop at a gradual pace. Later the authors mention children, immigrants and tourists to show that grammar expression does have some intermediates, and it is possible to have variation. Therefore, it is possible for natural selection to select for the best possible variants. I didn't quite understand protorules, but I imagine they are the critical leap that allowed the authors to formulate the natural selection for language.

    1. I do prefer the explanations proposed in the second article though. While this article states there are small selective advantages that can accumulate over time, I would argue that instruction had explosive effects. By increasing the efficiency of learning and many activities throughout life, I imagine language, which cuts out the middle man of direct sensorimotor experience (if the words used are grounded), would have provided a great many advantages that natural selection would have acted upon.

    2. Hey William!

      I was wondering if I might offer a slightly differing opinion.

      "Grammar seems to be quite categorical in nature-the sentence is either grammatical or not. However, it is possible that rules do develop at a gradual pace. Later the authors mention children, immigrants and tourists to show that grammar expression does have some intermediates, and it is possible to have variation."

      I took this talk about 'intermediate grammar expression' as a bit of a mistake on the part of the authors, in the sense that they (maybe unintentionally?) conflate UG (universal grammar) with OG (ordinary grammar). While there is certainly natural variation in one individual's proficiency with OG (as with children, tourists, and immigrants) this has no bearing on UG. All these individuals have UG already, even if their language is broken or governed by what Pinker and Bloom termed 'protorules'.

      In this sense I actually agree with your original sentiment: that grammar IS categorical in nature, in particular UG grammar, which probably (according to Prof Harnad) did not develop at a gradual pace because there is not real transition or 'protolanguage' between pantomiming and propositional language.

  15. While I thought the paper did a thorough job of discussing the more insubstantial claims against the evolution of language, there were a few key points where I was left unsatisfied or unconvinced. One that I really bothered me - Pinker’s failure to broach the question of the evolution of UG - was discussed in class today so is a bit of a moot point now unfortunately.

    Thankfully there was another issue I wanted to talk about. The authors suggest some probable initial selective benefits that language would have had, and conclude without hesitance that they make great sense and would have been sufficient for the explosion of language in the human species. However, I still sort of fail to see how - given the scenario that a mutation suddenly endowed an individual with some form of linguistic ability - it would have been a reproductive benefactor, precisely because a mutation by definition arises in just one human, so there would be no one to receive and understand the communication and to reproduce with the speaker as a result of it. Pinker suggests that perhaps the fact that the gene might be shared by family members would increase reproductive fitness, but I feel like the logic there is somewhat flawed. It’s obviously true that the gene might be passed down to offspring who could then use language to increase each others’ chance of reproduction, but even then we’re ignoring that it would have been selectively irrelevant for the initial mutant to have this completely unique ability. Maybe that’s splitting hairs a bit, or I’m missing a key point in their logic, but I just wasn’t convinced that a ‘mutation for language’ scenario is a viable model for the origin of language. I won't discuss it on this page, but I found the hypothesis in 8b much more satisfying and conceivable.

    On a lesser note, there were some other strained claims about selective benefits, particularly the observation that tribal chiefs are both gifted orators and highly polygynous. I found it bizarre that Pinker would mention something so similar to the more ridiculous EP hypotheses we’ve talked about (e.g. the singles bar/language postulation), and so the whole narrative re: the early reproductive advantages of language left a bit of a poor taste in my mouth.

    1. I also found this idea that there would be a single mutation in one person to be extremely unsatisfying and slightly crazy because that's saying also that one mutation is responsible for the capacity to use language. I don't really understand why they were so confident in saying this because it undermines the entire complexity of language capacity but maybe it's because they rely on the theory that complex change is reliant upon small mutations that might not even be detectable. Further, it seems impossible to guess what the first mutation may have granted in terms of grammatical capacity or categorical capacity etc... If you consider it for a moment, it seems honestly impossible. But I too think that the huge issue is the dismissal of UG and how that explains a lot of the issues I have with the idea of their mutation explanation. Also, in the conclusion of the paper the authors write "Our conclusion is based on two facts that we would think would be entirely uncontroversial: 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." The first clause makes perfect sense to me but then when they include the fact that "the only explanation for the origin of organs" is natural selection doesn't lead at all to language. This leads to organs, and it would necessitate that language has its own organ (even if you say this is the brain, the brain is also responsible for a lot of other things, so I'm unsatisfied with this) and it seems to just pull a really grand conclusion as if it's nothing.

  16. This comment has been removed by the author.

  17. Just a question I had in case anyone knows the answer (if there actually is an answer) - what on earth does Chomsky mean when he suggests that language is an "emergent property" due to undetermined "physical laws"? What is 10 physical laws apply when 10 neurons are placed in an object the size of a basketball, under the special conditions that arose during human evolution supposed to mean or to tell us about the origins of language? Does he actually have a concept/theory of what these physical laws might be or how they could result in language?

  18. I found this to be quite a dense reading that ultimately stated the obvious - human language
    evolved through the process of natural selection.

    Some interesting questions/thoughts that came to mind as I was reading this article by Pinker and Bloom were:

    1) Is Chomsky's theory about Universal Grammar causing stagnation in research on language acquisition in that many linguists neglect the fact that language could be developing in paralleling alongside social interaction see Social Interactionist theory) and that researchers try to "mold" their findings to best fit his proposed theory? Are we looking in the wrong places?

    2) Will there ever come a time where we will see natural languages converge to one globalized language? Could the process of natural selection lead to for example, a preference for the English language across the global population?

    3) When we are thinking or mentalizing inside our heads, are we "thinking" in natural language or are our ideas first formed without a direct link to natural language? Going back to lecture 6 on categorical perception, we talked about the Sapir-Whorfian (SW) hypothesis which says that mankind's perception of colour is reliant on the categorical structure of our language (the strong SW hypothesis essentially being linguistic determinism). So what I mean is, do I need to know the word for table in English to perceive a table when I am looking at it? Is our mental language different from our natural language in that it needs to be linked to it in the mental process going on inside our mind?

  19. I don’t think I fully understand Baldwinian evolution. If for example, an animal learns how to do something that helps its survival, and other animals imitate it/young learn from their parents how to do it, eventually those who were the fastest learners of this would have the best chance of survival. Thus, natural selection would select for those who could best learn this adaptive behavior. But I’m not sure what this has to do with the adaptive behavior itself – doesn’t it have to do more with the ability to learn quickly/well? If an offspring of these “good learners” was raised in isolation how would it have evolved this learned behavior or just the propensity to learn it well? Maybe this is what UG is – a circuitry in the brain that is a genetic propensity and was selected for because it allowed humans to learn language/ varying degrees of grammatical communication that was useful for their survival… Still, how this UG itself came to be seems really unclear. If we assume that UG is a neural circuitry encoded in the genes that sensitive to specific syntactic rules, could random mutations really have lead to it?

  20. First of all, I want to clarify the concept of Universal Grammar (UG). UG IS NOT a set of grammatical rules that we are automatically born with. UG is the capacity to acquire grammar - of any and all languages - despite limited input (poverty of the stimulus).
    If we look at human anatomy, we can identify an evolutionary function of each part of the vocal tract that is unrelated to language. Each component of the vocal tract serves an important survival function independent of the function(s) it contributes to our ability to produce speech. The article only talks about the structure of the vocal tract briefly in the introduction and then again here:

    "The vocal-auditory channel has some desirable features as a medium of communication: it has a high bandwidth, its intensity can be modulated to conceal the speaker or to cover large distances, and it does not require light, proximity, a face-to-face orientation, or tying up the hands."

    Here I think it is important to distinguish between language and communication. Of course language is a form of communication but not all communication is language. Most mammals communicate through some form of sound (ex. birds chirp, monkeys yelp, etc). So based on the structure of the vocal tract, we can say that yes, it was designed for sound to be produced and pass through it for the purpose of communication. But there is no part of the vocal tract that was developed specifically for language. The vocal tract exists for the functions of breathing, swallowing, chewing and producing sound. The vocal tract did not have to undergo any anatomical changes to accommodate language. In my opinion, this points to what Chomsky and Gould have argued about language being an evolutionary 'side effect,' rather than being the real product of natural selection.
    It simply happened to be the case that the structures already present in the vocal tract, which were there for survival, were useful for language, the structures did evolved specifically for language. This points to language being more of a 'side effect' than a selected trait. Of course language had to have involved some chemical/structural change at the neuronal level. But that change could then be explained as a side effect.

  21. “All languages are complex computational systems employing the same basic kinds of rules and representations, with no notable correlation with technological progress”

    When Pinker states that “languages are complex computational systems,” does he mean to say that language is computation? If so can this be true? I have a hard time reconciling this explanation due to languages strong tie to cognition. If we are suggesting that language is simply turning squibble to squabble, then the squibble in question is our thoughts (or cognition) and the squabble is the interpretable output of language (be it vocal, written or gesture). We have already noted that thinking (or cognition) isn’t just computation, thus my question is how can language, the medium to communicate our cognition, be just computation. On the other hand, is Pinker indicating that the translation of languages, from one to the other, is just computation? This is obviously correct as the language being used is independent of what is actually being expressed.

    “In fact the child's system has greater expressive power than the adult's.”

    I find P & B’s word choice to be poor here. “Expressive power,” relates to what someone can express and their ability to express it. With this in mind, obviously an adult has a better ability to express what they want and what they express is beyond the child’s scope due to its lack of experience. In the explanation that follows, P & B suggest instead that adults are constricted by the rules (grammar) of their own language and can no longer express themselves in the simple and “logical” forms that children do, which are technically “improper” in a grammatical sense. This is very different than saying that adults can express less than children can and that their ability to do so decreases with age.

    1. Hey Cristian, I wholly agree with your point about expressive powers of children and adults. Pinker and Bloom poorly describe how a child's system can have a greater expressive power than that of an adult. I think in light of rest of the paper, it seems they are getting at something more basic than how much a child or adult can express. An adult's "system" has less room for flexibility than that of a child because it is contained in the framework of learned, cultural grammar rules. The authors imply that before complete language acquisition, children are something of a blank slate, whereon any set of grammatical frameworks can be adopted. But once the first language has been learned, "adults" lose this plasticity of language.

    2. Hi Cristian,

      When you say “I have a hard time reconciling this explanation due to languages strong tie to cognition”, it brings us back to the computation/cognition argument. Pinker believes that “languages are complex computational systems” but this can still be cognition. Computation is a part of cognition but not wholly cognition. So if we consider that, then we can surely say that language is computation which is part of cognition.

      But then again if we take the symbol grounding problem into account discussed in previous classes, we see that there is more to language/words than just sense and reference. There may be a sense of ‘feeling’ involved. Here, I share your view and am not clear to what extent is language more than just computation.

      Additionally, there is at the argument of the innate structure and how we are designed to learn language but there is not much empirical evidence to back this claim.


      Your question “is Pinker indicating that the translation of languages, from one to the other, is just computation?” takes us back to the Chinese room argument. If it is just computation, do we really understand the language being translated?

      Furthermore, we can extend this argument to Pinker’s claim that language is computation. If such is the case, then do we truly understand language? Obviously we do and hence language can not just be computation.


      To conclude, I am inching towards the argument that “natural selection is the programmer”. Language can be just computation given that we have an innate structure to learn language. This structure can be the one that links language with more than just sense and reference (and in turn also take the symbol grounding problem into account).

  22. There are two aspects of the evolutionary argument I have trouble marrying. The first is that language development follows the development of thought, where children begin with babbling, to combining simple sentences, to a growing complexity of vocabulary and sentence structure with a growing capacity to think through these different statements as they exist in the real world.
    The second is the modularity of language, where an individual can be a “linguistic savant who is mentally [disabled]”, or a person who is “linguistically impaired with normal intelligence”. While modularity of language could grow in complexity as modularity of thought could become more complex, this does not fit with the example of individuals with disabilities confined discretely to linguistic, or non-linguistic areas.

    I will accept the theory that universal grammar exists due to mechanisms of natural selection, but I would not go insofar as to say that language is subject to natural selection.
    There is a creative aspect to language, inventing languages, changing through place and time
    Yes everyone has the same universal grammar and basic abilities in language, but some can express much more **ideas than others through their vocabulary - perhaps this has a relation to where complexity of thought matches complexity of lexicon?

    Did language allow us to have an increasingly complex culture, where we could use information from past generations to inform the decisions of the present, where !Kungs’ work could move their group forward by saving time from war expenditures, OR did an increasingly complex social culture require the development of “socially-relevant abstract information” that could be coded and decoded between in-group members? This is the chicken and the egg problem that I would like to have seen an attempt to answer in this article.

    Finally, a point not brought up in 5.3.4 Social Use of Language and Interaction, is the advantage of having a tool to understand others within the in-group, but being able to stay discreet when communicating in the presence of someone from an out-group. Humans our social, but tribal creatures where out-group conflicts have been common throughout history. Perhaps having universal grammar and an innate affinity to learn and seek to learn language encoded in the genome, but having the specifics of a language be environmentally depended served this in-group, out-group purpose of language.

  23. As I am also doing a degree in English literature, I find some parts of this article particularly interesting, especially Piattelli-Palmarini's argument:

    The arguments that language could not be an adaptation take two forms: (i) language could be better than it is, and (ii) language could be different than it is.

    I was initially convinced by especially the second point. The fact that language could be different than it is is interesting. Why can't we say: "Who did John see Mary and?" This definitely at first was nearly enough to convince me that grammar is not predictable as an adaptation to communication - it seemed so arbitrary. However, as the rest of the paper unfolded, I quite liked how they exalted arbitrariness. It is true - language is arbitrary, and so are the rules of grammar. Both UG (the aspects of communicative language shared by all languages) and grammar are arbitrary. This paper explained how arbitrariness is fine so long as it is shared because, once shared, the arbitrary rules of grammar become a common denominator with which we can communicate. This then led me to thinking about how this relates to computation... could the rules of grammar (the arbitrary rules and shapes and symbols, etc) be looked at kind of like a type of algorithm? How language ties into computation and how much of language IS computation is interesting - it also hearkens back to Searle's Chinese Room Argument and his reply to the Systems Reply.


  24. “Second, it may be difficult to evolve a huge innate code (18).”

    Goes on to talk about how genes are costly and you don't want to take too much space up in the genome. This seems contradictory to me, because the rest of the paper is trying to explain why languages probably did come about as a result of evolution and natural selection. I think that if you bring up the argument that there is no universal language because of how costly it would be to the genome you need to address the fact that the converse is a big hit to your argument – that if innate codes are costly, why should we have one at all?

    I liked when he talked about the “Tradeoffs of utility” within languages and the fact they are unavoidable (19). Similarly to how we can accept the weak Whorf hypothesis or agree that cognition is some computation, I think parts of language must have evolved or are still influenced by our evoltinoary past and our predisposition towards wanting to do the least amount for the highest yield. I still am not convinced by the article on the whole though – language is far too cmplex to be something that evolved! Also, the example of the eye as the organ bothers me – shouldn’t other species have language if it is so vital and important and successful and most of all evolutionary ?

    Gould says that language is a spandrel because of his frequently stated position that the mind is a single general purpose computer – he thinks that with increases in size we get an increase across the board in functionality (25). I think Pinker is smart to criticize this – building a computer that is really good at one task and making it more powerful doesn't mean it will suddenly be both a chess-playing machine and a payroll distributing machine.

  25. Living gregariously in an environment where there are threats, predators, or needs that must be met for the group (e.g. infant care, copulation), communication is necessary. This can take the form of body language, facial expression, or conspicuous physical traits such as plumage, pelage, and adornments like horns; I would consider these non-verbal modes of communication precursors to ‘formal’ verbal communication. And, just as is asserted in the paper, these non-verbal cues would evolve via the Darwinian trajectory, over long periods of time and via natural selection based on mate-preference. In my mind, language can evolve just like these traits.
    Focusing on the human lineage, we can bring in semiotics – the study of symbols and signs. For example, if we take an early hominin whose basic needs would involve hunting and thus cooperation, which would necessitate signaling to one’s hunting partner (this assumes group living, which archeology has shown was evolutionarily advantageous for our ancestors) that “MAMMOTH” is “THERE” – this refers to an abstract concept in an abstract location. Evolution certainly does not always (seldom) move in the direction of efficacy, but I believe over many hundreds of years this kind of referential system would need to involve in complexity to make one’s points clear and concise – thus increasing one’s ability to successfully catch prey.

    1. Hi Felicity,

      I got a little stuck on one of your sentences:

      "This can take the form of body language, facial expression, or conspicuous physical traits such as plumage, pelage, and adornments like horns; I would consider these non-verbal modes of communication precursors to ‘formal’ verbal communication.”.

      i understand that the listed physical traits and body language/facial expressions (non-physical traits?) are not verbal communication. But, I think that body language and facial expressions convey communication substantially more than physical traits can. Body language and facial expressions can assist in mind reading, but fur, feathers and horns cannot. One can attempt to communicate with another through body language or facial expressions, but one cannot attempt to communicate with another through physical traits. Could you clarify what you mean by communication through the listed physical traits?

  26. I thought this article was very interesting. However, there was something that bothered me in this article by Pinker and Bloom and that was when they referred to language as an organ. They stated "language shows signs of complex design for the communication of propositional structure, and the only explanation for the origin of organs with complex design is the process of natural selection”. I do not see eye to eye with this article and think that their comparison to language as an organ to be quite obscene. Especially if an organ is known to be part of an organism that has a specific function and is a distinct structure. However, language is more broad and involves many different structures such as the throat, mouth and speech box.

    What is the significance of the authors comparing the development of language to the development of the eye? Then, where does the concept of mutations come into play; can’t it significantly alter the evolution?

    Furthermore, the authors mention that language pass the test "when it is appropriate to invoke natural selection as an explanation for the evolution of some trait,”. Is this enough for an explanation/pass or is there more?

    1. There's nothing wrong with describing an inborn brain mechanism an "organ" even if it's anatomy is distributed. The hippocampus is an organ; perhaps so is the brain mechanism for language learning. Same for the eye and the rest of the brain's visual mechanism. It got to be what it is now through evolution, including mutations.

      The fact that humans are the only species that can learn language makes it very likely that this is because some evolved brain functions are unique to humans.

  27. The main argument behind Universal Grammar is that a child does not get enough input or feedback to construct the complexity and accuracy of grammar they are able to produce (‘’poverty of the stimulus’’ argument). Therefore, there must be a ‘language template’ in their brains allowing them to process the input they do receive in a directed manner. But how can this ‘template’ have evolved? It seems like a universal phenomenon that children are not corrected on their early utterances, or given enough clear grammatical input in their early years, so how can any human have been able to make a template from scratch? Did it evolve in adult brains as they were first constructing the language? How can it be ‘universal’ if humans have been so dispersed across the globe for much of evolutionary history? Surely different languages evolved in very different ways by very different speakers? Are any elements of Universal Grammar found in any of our closest primate relatives?

  28. “In sum, primitive humans lived in a world in which language was woven into the intrigues of politics, economics, technology, family, sex, and friendship and that played key roles in individual reproductive success. They could no more live with a Me-Tarzan-you-Jane level of grammar than we could.”

    Doesn’t this directly contradict the whole argument they have built up to this point, namely that incremental complexities of grammar could have – and in fact necessarily – benefitted our ancestors?

  29. I do not understand the non-selectionist view of the origins of language. I don’t know if Pinker convinced me so thoroughly or if their position is just hard to defend… From my understanding of Pinker’s portrayal, there are two main points of evidence for the non-selectionists:
    (1) Language is all or nothing, but natural selection works in gradual generational changes.
    (2) The theory that grammar is universal somehow runs contrary to the way that selection works.
    If language did not arise from natural selection, and it most certainly did not arise spontaneously from genetic drift, where could it have come from? To say that it emerges from brain complexity above a certain threshold is not an adequate answer to our fundamental cognitive question: why and how can we speak like we do?
    I must say, though, UG is a really strange phenomenon. Who would have thought that there would be so much in common between all of the languages on Earth...


  30. I think the main point of this article is made clear in the abstract. It is the idea that the best way to learn more about language is to use an evolutionary perspective. Another way to say this is that language evolved as an adaptation. Apparently, some writers have challenged the view that language-capacity has an evolutionary explanation. They say that other explanations are better, such as that language arose as a by-product of increased socialization, or something. The main argument seems to be that language is not a direct evolutionary adaptation. The authors point out quickly (again, in the abstract) that this argument is not grounded upon a valid understanding of what it means for something to be an evolutionary adaptation. If something has a complex design in order to carry out some function, and if we can’t explain its complexity with alternative mechanisms, then that something is a trait that can be attributed to natural selection.

    So the basic explanation of Darwinian natural selection is sufficient to dispel the non-selectionist argument for language-ability. The reason I keep talking about how it’s a “basic” explanation is because in the recorded lecture that accompanies this article, Stevan says that this is pretty much the only point we need to take away from this article. So I guess the authors realize that 50 pages is kind of a lot for what they can effectively establish with their paper.

    When Stevan says that the paper chooses to ignore the “propositionality” of language, I think this is an interesting point. To me, it seems that talking about that is somewhat beyond the scope of what the authors are trying to do: discuss language as natural selection. Language already seems complex enough, even before discussing the way in which propositions form the structure of language. The abstract of the article makes it clear that if language is complex, and can’t be better explained elsewhere, it’s going to be because of natural selection. So why bother with propositionality, at least for the authors and specifically in this paper?

  31. I thought this article was very well written, and the authors undoubtedly present compelling support for their arguments. That being said, I found it interesting that the authors mentioned the comparison of "unconvincing adaptionist explanations" to Kipling's "just-so stories", made by Gould and Lewontin. Kipling's just-so stories are the perfect example to describe the shortcomings of adaptionist theories that seem to rely on "common sense" explanations rather than empirical scientific data to lend credibility to their claims. Pinker and Bloom write "all we have argued is that human language, like other specialized biolgoical systems, evolved by natural selection. Our conclusion is based on two facts that we would think would be entirely uncontroversial: 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." Later the authors state “We think there is a wealth of respectable new scientific information relevant to the evolution of language that has never been properly synthesized. The computational theory of mind, generative grammar, articulatory and acoustic phonetics, developmental psycholinguistics, and the study of dynamics of diachronic change could profitably be combined with recent molecular, archeological, and comparative neuroanatomical discoveries and with strategic modeling of evolution using insights from evolutionary theory and anthropology.” I think that the authors’ claims that the credibility of their evolutionary argument for the development of language and grammar rests on “two facts we think would be entirely uncontroversial” seems a bit like a just-so story: as the authors mention, they draw on a dearth of scientific data in making their argument, but their argument is a scientific one at its root. Instead, they seem to draw on more “common sense” reasoning to back up their arguments, which is the fatal flaw of the just-so stories that they themselves state are so implausible. While I understand that the scientific material they mentioned should be integrated into theories on the evolution of language is not necessarily in the scope of the present paper, I think that it would have been beneficial to draw on some more concrete empirical scientific evidence, relevant to language in particular not just “the origin of complex biological structure” via natural selection, to support their main arguments.

  32. This comment has been removed by the author.

  33. “The fact that one can conceive of a biological system being different than it is says nothing about whether it is an adaptation (see Mayr, 1983). No one would argue that selection was not the key organizing force in the evolution of the vertebrate eye just because the compound eyes of arthropods are different.”

    In many of the psych and neuro classes that I have taken over the years at McGill, the topic of Chomsky’s theory of a universal grammar and the innate structures for language are discussed. However, most discussion is avoided around whether or not this grammar or the human ability of language comes as a result of specific adaptation and natural selection vs. a bi-product of some other mechanism.

    The fact that such big names in the field of Linguistics and Evolutionary Biology do not agree that language could be naturally selected for is kind of surprising to me. This type of thought seems to be the opposite of what is often encouraged in scientific discovery – to reach a level of normal underdetermination. The reoccurring argument seemed to be that there are alternative possibilities within language (ways it could be different/better) so we discredit its specific selection as plausible. I find this ironic for the reason they provide in the article – the fact that the vision mechanism of the arthropod is different than that of the human would not lead a single scientist to say that the eye and vision was not a capacity that was specifically selected for with many years of adaptation over time, because it is such a complex and intricate organ of the body. The idea that language shouldn’t be looked at through a similar lens seems silly. It is likely because language was adapted to contain ambiguous possibilities for manifestations of underlying structure that it is so easy for us to think of alternatives. The human eye, and every other thing that has been naturally selected for, is almost certainly also a product of the balance of compromises in needs that existed throughout the evolutionary process. However, the alternatives are not as easily identifiable. But to suggest that ambiguity as a quality in itself cannot be selected for seems like a short-sighted understanding of natural selection.

    I also think that, because natural selection and the concept of survival of the fittest is often looked at in terms of helping one individual survive, language as natural selection is overlooked, as its adaptiveness comes from collectively learning, which increases the chance of shared – rather than individual – survival. As the text mentions, sometimes the only reason for doing something is that everyone else is doing it. Thus, it would be very easy to manifest an ambiguous grammar differently in different places, simply because standardization would evolve differently with time. A perfect example of this in language is the difference between France French and Quebec French. It seems as though skin colour could also attest to this idea, but is more believable because we can create an evolutionary advantage for each colour based on geography. Standardization in language may have been something whose selection was simply the fact that the “environment” (community) was all dependent on it for its effectiveness.

    Ultimately, I think that saying language isn’t a capacity that is a result of natural selection reflects a narrow-view on what qualities of organisms need to be selected for. When looking at something like the eye, physical structure can easily be considered in selection analysis, while variation may exist in other areas. For language, the qualities that are selected for and areas of variation may not be consistent with physical human traits (the eye, bipedal locomotion), and as such, need to be considered differently to be understood in terms of the natural selection model.

  34. This comment has been removed by the author.

  35. Yes, I agree with the notion that "the ability to use a natural language belongs more to the study of human biology than human culture;" and to a large extent agree 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" because "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.
    However, reading this article was rather frustrating, one because it was very long and most of the points were covered in the abstract, and two, because (and I guess it's not the authors' fault) I was hoping the paper would articulate and elaborate more on the role natural selection played in humans' ability for language and the myriad types of languages and communication techniques exhibited (and unseen, unknown to us) used by animals. Stevan also mentions they did not elaborate on the "propositionally" of language, which I would've also liked to read, but I guess these critiques don't necessarily fall within the scope of the article. A more nuanced follow up would be very interesting to read.

  36. That was an incredibly long article!
    From the look of it, Pinker and Bloom were true to their word, discussing a topic that is seemingly obvious - that language is at least partially innate, and was forged by natural selection, while in the process strangely conflating universal grammar (UG) and ordinary grammar (OG).

    If I have this at least halfway straight, it would seem that language is made of UG, OG, vocabulary, etc. Because of the poverty of the stimulus (the absence of non-UG-compliant sentences that a child perceives when they're learning language) UG must be innate, a structure upon which the rest of language must fit, and the other aspects of language, (OG, vocabulary...) are acquired via category learning. This innate UG, and the learning capacities that allow for the acquisition of the non-innate aspects of language, BOTH evolved because of selective pressures.

    So I suppose while Pinker and Bloom have addressed the important issue that language arose because of natural selection (duh) some important subtleties are missing that muddy up the conversation.

    Indeed this muddying is what messes up a clear discussion of all this protolanguage business. Pinker & Bloom mischaracterize Chomsky's concerns about finding an evolutionary explanation for UG.

    Chomsky says: "[an innate language faculty] poses a problem for the biologist, since, if true, it is an example of true 'emergence' - the appearance of a qualitatively different phenomenon at a specific stage of complexity of organization."

    This, to me, seems to be a legitimate concern. How was the transition made from 'pantomime to proposition' (simple descriptions, to actual veridical propositions)? I guess this is a discussion for next week, but I think it's worth discussing in terms of the sort of claims that Pinker and Bloom make.

    They hand-wave Chomsky's concern by pointing to examples of so-called 'protolanguages' like Pidgin, which while having limited vocabularies, are in reality full-fledged, propositional languages. They claim that these are examples that language is not 'emergent' but develops slowly, when in reality, they've come nowhere close to explaining how that magical transition (pantomime to proposition) was made.