Has a remote Amazonian tribe upended our understanding of language? (2007)
newyorker.com
newyorker.com
Pointing that out made him really defensive; understandable, perhaps, but recursion has been found in literally every other language in the world. You can't be that surprised when people have serious reservations an argument that hinges on the validity of a tiny data set from a language that only you understand.
I stopped paying attention after that, though. Maybe there's other people doing work on it now.
All we are discussing is a biological system, and biological systems are things that may be investigated.
"Chomsky has replied that he considers recursion to be an innate cognitive capacity that is available for use in language but that the capacity may or may not manifest itself in any one particular language"
which makes me wonder - do they show an innate understanding of the concepts when it is discussed with them.
Alternatively, someone who speaks Piraha natively could learn English and discuss about this concept in English.
I don't really believe it, and most linguists haven't since the early 20th century, but that's the claim he made.
* https://ling.auf.net/lingbuzz/000411
* http://lingphil.mit.edu/papers/pesetsk/Nevins_Pesetsky_Rodri...
He finds that his interest in languages and linguistics exceeds that of spreading Christianity and makes a career change.
Then, either
A) He realizes that he’s the only one who speaks this language and concocts a scheme to seek fame and fortune by falsely claiming it contradicts the leading theory of linguistics because no one can prove him wrong. Mwahaha!
Or
B) As he learns about linguistic theories, he realizes that the language he learned as a missionary contradicts the leading theory of linguistics. This is academically interesting, so he writes a paper and gives talks describing what he found.
While A is certainly possible, it seems B is more likely.
[1] http://blog.thegrandlocus.com/2016/04/did-mendel-fake-his-re...
And if you're talking the scientific investigation of complex phenomena like language, fooling one's self is common. Indeed, unless a researcher is using tremendous care, their chance of at least partly fooling themselves is very high.
Scientific processes and checks and balances exist because of this. That Everett essentially ignored these and primarily argued his case in the press leads me being very dubious of him. But this still doesn't he's a "liar", that he believes what he claims is plausible but doesn't change anything else.
A true dichotomy would be more along the lines of: he is right or he is wrong.
Of course all possibilities subdivide, if he is wrong he is knowingly wrong and a liar or he made mistakes and too grandiose claims. Even if he is right, he could be right for the wrong reasons: his conclusions could be unsupported by his premises & evidence but still correct.
I remember growing up as an Evangelical Christian, I felt this argument was pretty decent. But eventually I came to realize that the argument hinges on the assumption that you can't be both a 'lunatic' in one part of your life, and functional or even an admirable person in other parts of your life.
Which of course is a silly assumption. The world is filled with genius crackpots, do-gooders who sucked in other parts of life, or people who are highly successful in one area and utter failures in other areas.
It's notable also the way Lewis cites honest people attesting to miracles, with again the false dichotomy that either the miracles happened as they described or the people were liars, when we have already a multitude of examples of honest, intelligent people thinking they've seen something that they haven't.
That may be my bias as a very casual reader who isn't particularly invested in linguistic theories. (I certainly like Chomsky's social/political criticisms.)
I don't think its wrong to swing for the fences, after all, the wider the claim the easier it is to disprove.
Language is a human capacity, so in fact there MUST be universal laws that govern it (to be more specific, this is true of human languages; aliens could have entirely different languages, but for any particular alien species, there will have to be some laws of the structure of that species' language).
How do you imagine language could develop (both evolutionarily and within each child that learns it)if there were no predetermined laws that restricted the set of possible word combinations? As Chomsky would say, if you think about it for a while, you will see that the alternative actually makes no sense.
That is not to say that the laws that Chomsky has identified must be correct. He could even be wrong about searching for universal laws at the level of grammar. But he has shown conclusively that there must be some universal structure that underpins all natural human languages.
Why should there be a predetermined law? I have never understood this argument and no, I don't "see that the alternative actually makes no sense". I would like to see some facts that support it, something its followers have failed to delivered for how many decades now?
Still looking for those facts, are we?
All of this ends up with a huge combinatory explosion of possible ways to assign meaning to language, requiring a huge amount of examples and counter examples to actually learn a human language. Just look at the training corpuses of something like GPT-X to see the amount of information that is required to learn elements of a human langauge if you start without any assumptions.
However, we know for a fact that human children get many orders of magnitude less examples of languahe use before learning a language - far too little to actually rule out all other possible structures if they had started with 0 assumptions.
Adding to this, there are certain elements of language structure that are not commonly thought about that happen to work identically in all natural languages that have been studied, sucha s the fact that order of words in a sentence matters, or the way certain words are omitted when constructing embedded sentences (e.g. 'can eagles that fly swim?' always having 'can' refer to 'swim', never 'fly').
Anyway, this discussion is going nowhere. Notify me if there is other evidence apart from "I cannot imagine how the human brain can do that". As other people have written in a similar thread some months ago, this is pseudo-science.
From my own readings, I thought that there is very good evidence adults don't bother to correct children in a lot of cases, and that children ignore the corrections completely anyway. They also don't seem to make the kinds of mistakes that you'd expect.
A great summary of some of some research I've been reading on this is the paper [1]. Also Steven Pinker's early work [2] is fascinating and easy reading.
I would love to bring myself up to date though, if these things have changed lately. Any help appreciated.
[1] Gleitman & Newport paper The Invention of Language by Children - https://www.sas.upenn.edu/~gleitman/papers/Gleitman%20&%20Ne...
[2] Steven Pinker - Why the Child Holded the Baby Rabbits - http://socsci.uci.edu/~lpearl/courses/readings/Pinker1995Ch5...
Couldn't agree more. It's barely even that, it's kind of just armchair philosophizing. (which is fun and often actually useful, but not sufficient to make claims about universal laws)
I disagree. There's a strong desire from a western scientific mindset that everything be possible to reduce and turned into laws and models, but that doesn't mean everything works like that. It's putting the cart before the horse.
> How do you imagine language could develop (both evolutionarily and within each child that learns it)if there were no predetermined laws that restricted the set of possible word combinations? As Chomsky would say, if you think about it for a while, you will see that the alternative actually makes no sense.
This is just armchair philosophizing, reminds me a lot of "praxeology". And even if it was true, which I don't believe at all, that doesn't mean that he or anyone else can successfully identify and articulate them.
> But he has shown conclusively that there must be some universal structure that underpins all natural human languages.
I don't see that at all and while I happily admit I'm not a linguist I'm sure there are countless actual linguists who would agree Chomsky is a bit of an emperor with no clothes.
I wouldn't think so either. That's what makes the claim, whether its right or wrong, interesting.
> And I guess I just don't see the point making any claims of universal laws.
Well it sets a goal post and leads to further questions. Are there exceptions? Why are these features universal and not others? Why is any feature universal? If there are exceptions and they're rare what makes them different from the other cases? Etc.
> "these are incredibly ubiquitous features and here's some useful abstractions and models"
If your model/abstraction doesn't differ from the ground truth as far as we know (or the proposer doesn't believe it does), what's the difference between that at claiming universality.
I'd claim that reality is just a model we don't yet know is wrong. For a long time we would say newtonian mechanics was universal. Now we say its a model that works at low velocity (relative to c). Was it wrong to claim universality back when we weren't aware of any counter examples?
Chomsky has been known to be wrong for decades. His real talent, if not genius, was pivoting away from linguistics while his reputation was still strong, and managing to preserve it despite his original claim to fame being thoroughly debunked.
You may or may not think Generative Grammar (or his more recent Minimalist program) is correct, but it's certainly not KNOWN to be wrong, and has yielded some insights. Has it changed over the years, of course, that's normal science. New data is found, new explanations are posited.
Further, Chomsky is still heavily and actively involved in linguistics. See his Feb 4th 2021 presentation [2] to eminent linguistics researchers in the field.
I'm not sure where this kind of comment comes from?
[1] http://www.ugr.es/~fmanjon/A%20Review%20of%20B%20%20F%20%20S...
Ah, but the above is a universal law, so it's self refuting!
Therefore, universal laws about human language are possible.
Checkmate, skeptics!
Anyone who actually thinks that is not just uncharitable, unproductively pedantic and arguing semantics, they're also wrong.
But OK, to satisfy them, we could say that, yes, universal laws about human language are indeed possible, and there is one such law. Let's label it "The Universal Law About Human Language".
It states
"The only universal law about human language is that there are no universal laws about human language (other than The Universal Law About Human Language)"
It even has recursion! (pun intended) So I guess us skeptics will also have to admit defeat that recursion is in fact universal...
I would happily grant there are more laws about human language. You could come up with all kinds of pointless truisms like "human languages facilitate communication" etc. No reasonable person would find much point in coming up with or debating "laws" like that. (but I'm sure there are plenty of "epistemologists" and "philosophers" who could debate them in minute detail until the end of time)
But Chomsky isn't claiming obvious truisms - he's claiming fairly detailed laws about universal grammar (and fairly detailed means through which they derive)
Critics point out such fairly strong claims need fairly strong, in your face evidence. The onus is on Chomsky to provide it. You don't get to turn around and shift the burden of proof on the critics and start arguing nonsense about how criticism of Chomsky's claims are "self refuting" or what have you.
Chomsky says there are universal, physiologically-based manifestations which are universal across humans. That is not really demonstrates but would not be surprising any more than the universality of the human visual system (as well as individual variation).
He claims to have identified some such universals, and that they are relatively high level. THAT is a bold claim, hard both to verify or falsify. But in principle not absurd.
I don’t believe it, but I no longer mock him with the phrase “grammar gland”. There is some insight there.
We can also see this when brain areas are recruited for tasks they were not specifically designed for. For instance visual processing areas can be recruiited for other tasks in the blind.
So while there is specialisation in brain tissue, rather than there being circuits designed at a low level for certain cognitive tasks, it is more that the connectivity and kinds of neurons within an area are tuned to excel at a specific class of problems.
It is softer and more diffuse organ than a "grammar gland". Rather it is a kind of soil within the mind that permits the growth of certain kinds of cognitive structures. Rather than being computer like and existing prior to the input of data, it is merely a capability expressed by the tuning of an overall system.
Chomsky's universals in this view become rather less important. It is merely the statement that brain tissue is tuned in such a way that it encourages, among other things, certain specific kinds of linguistic structure.
It's similar to noticing that the human hand is good stirring soup and then describing it as a soup stirring organ. Yes the hand has properties that allow it to be easily used as a soup stirring organ, but those same properties make it useful for some other things as well.
If only I had one more upvote.
Hasn't there been a lot of recent work showing similar circuits between songbirds and humans that seem to have independently evolved?
In its original form, it was fairly specific and made quite strong claims. One claim, for example, was that we humans uniquely have a specialized language organ/section of brain that evolved at some point in time. No evidence has emerged to support this idea as it seems language skills are not concentrated in a single area of the brain (unlike vision for example). And it actually seems implausible that a genetic mutation would produce a language "organ" as it would confer no advantage to an individual in a society where your peers had no language skills.
And you'd imagine that a "fundamental theory" of language would have some practical application but Chomsky's theories don't seem to have any. For example, you'd imagine that it would provide the basis for the study of language acquisition. I remember a friend with a doctorate in clinical child linguistics laughing when I enquired about how their work in therapeutics for children struggling with language problems used Chomsky's theories. Apparently this "fundamental theory" of language has absolutely no application in this area and earlier attempts to apply it led absolutely nowhere. She says that clinicians ignore it - it has no relevance.
Even in AI natural language processing it led nowhere when it was the foundation of AI research in this area from the 60s to the 80s. All breakthroughs in this area use statistical approaches.
I find linguistics fascinating particularly historical linguistics and I used to be quite surprised at how authors in the area seemed to tip-toe around Chomsky's theories - reading between the lines in Guy Deutscher popular book on the evolution of language, for example, he espouses a theory which seems fundamentally at odds with that of Chomsky's (and also seems far more plausible to me than the idea that language ability just sprang into existance through a random genetic mutation) but he seems to go out his way to credit Chomsky's brilliance.
Having observed the villification of Everett - far beyond what I think would have been reasonable - I now understand why linguists genererally either ignore Chomsky's theory or pay some sort of lip-service to give it credit rather than attack it outright. Doing the latter is academic suicide it seems.
As an amatuer in the area, it looks to me like the Emporer has no clothes. It's a theory which is defended vigourously but has no appparent application and is not truly falsifiable. Even when contradictory evidence emerges, the theory is "refined" not into something more useful - the way most scientific theories evolve - but into something vaguer and even less useful.
I also realize that I don’t even know what it means for a human language to be recursive. I get sub-clauses but I don’t see their existence as evidence of recursion being a fundamental feature of human language. In English for example, I doubt there are non-contrived examples of sentences which have more than 2 levels of sub-clauses. The concepts are certainly independent with computer languages - originally Algol had procedures (analogous to sub-clauses) without supporting recursion.
Procedures without recursion was a FORTRAN innovation. Recursion started in Algol. It would not be surprising if the first Algol effort lacked it.
What Chomsky did not understand, and still might not, is that his grammatical structures are fundamental to computation, and necessarily arise anywhere that symbolic processes of any complexity occur. So, their rise is not a product of culture, or of unique human biology, or of nerve biology, but of the architecture of the universe all operate in.
Chomsky's grammatical analysis was anticipated, in fully elaborated detail, in Sanskrit, millennia ago. The structures' importance in fundamentals of computation was understood (by a few -- Schonfinkel, Curry, Church) almost a century ago.[0]
It would not be at all surprising to find the principles to be important to DNA manipulation—particularly as part of immune response, and convergent in multiple independently-developed lineages.
[0] <https://writings.stephenwolfram.com/2020/12/where-did-combin...>
I think the interesting thing about recursion in language is precisely that we rarely see deeply nested examples yet we don't seem to have issues parsing deeply nested examples when they do occur (if they're not specifically constructed to be hard to parse). The fact that we probably didn't learn this by example might be an indication that our brains are somehow hardwired to look for regularity in grammar, including recursion.
This is a non-sequitur because you're implying that such a mutation would be binary: either a creature has it and can communicate with creatures that have the same, or it hasn't and communication is impossible. That's not how evolution works, specialization of organs can happen over thousands of generations. If a specific mutation is beneficial, it spreads through the population over time, and additional mutations in that area can further develop the organ.
Second, our speech apparatus isn't one "organ" in the narrow sense: our lungs/vocal chords/tongue/lips/cheeks/jaw all contribute to the richness of our speech. The breadth of our speaking ability is a specialization of fine muscle control in all those areas, coupled with a finely-tuned hearing apparatus to analyze and understand those sounds.
It doesn't seem like an extraordinary claim to me that our evolutionary path has enhanced our communication ability far beyond our peer species -- because it's clear to see how a rich vocubulary can aid a species' survival over time. I have no knowledge of Chomsky's specific claims, but I don't think you are representing his theory accurately.
The claim of Chomsky's UG is that there is a genetic component to the language faculty of humans. That's pretty much a truism given that we can speak and our closest biological relatives can't. Presumably this cashes out somewhere in our brain.
There's evidence that certain regions of the brain are involved in Language (Broca & Wernicke's areas) given evidence from brain injuries and their effect on language.
He does posit an LAD (language acquisition device) but this is a logical/theoretical construct, to aid in theory building. It seems very likely that we have this given the universality of language in our species and our ability to learn the language of our own culture. The LAD though has not been claimed to be physically located somewhere in our brains by Chomsky (again afaik). Though presumably if we have it, it is fairly likely to be somewhere within our skulls.
It could evolve primarily for internal thought first, or have some other benefit for enhancing existing systems of vocalized mimicry (for hunting, and maybe other uses).
>The basis of Chomsky's linguistic theory lies in biolinguistics, the linguistic school that holds that the principles underpinning the structure of language are biologically preset in the human mind and hence genetically inherited.[150] As such he argues that all humans share the same underlying linguistic structure, irrespective of sociocultural differences.[151] In adopting this position Chomsky rejects the radical behaviorist psychology of B. F. Skinner, who viewed behavior (including talking and thinking) as a completely learned product of the interactions between organisms and their environments.
that reminds a similar theory that stone axe making is a result of genetic transfer (like 'bird song') and not social transfer https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5066817/ . Curiously if one replaces "axe" with "language" in the article it would look a lot similar as how one would go proving Chomsky linguistic theory.
Languages That Could Change the Way You See the World (2015) - https://news.ycombinator.com/item?id=26224117 - Feb 2021 (50 comments)
Talking is throwing fictional worlds at one another - https://news.ycombinator.com/item?id=24468639 - Sept 2020 (42 comments)
A Simple Structure Unites All Human Languages - https://news.ycombinator.com/item?id=21030971 - Sept 2019 (53 comments)
‘Anumeric’ people: when a language has no words for numbers - https://news.ycombinator.com/item?id=14697453 - July 2017 (42 comments)
Grammar isn't merely part of language (2016) - https://news.ycombinator.com/item?id=13615273 - Feb 2017 (68 comments)
Chomsky, Wolfe and me - https://news.ycombinator.com/item?id=13378341 - Jan 2017 (33 comments)
Has Chomsky been blown out of the water? [pdf] - https://news.ycombinator.com/item?id=12511631 - Sept 2016 (36 comments)
Evidence Rebuts Chomsky's Theory of Language Learning - https://news.ycombinator.com/item?id=12450033 - Sept 2016 (95 comments)
Tom Wolfe tries to take down Darwin and Chomsky - https://news.ycombinator.com/item?id=12402713 - Sept 2016 (12 comments)
The Chomsky Puzzle - https://news.ycombinator.com/item?id=12367289 - Aug 2016 (78 comments)
There is no language instinct (2014) - https://news.ycombinator.com/item?id=10612447 - Nov 2015 (30 comments)
MIT claims to have found a “language universal” that ties all languages together - https://news.ycombinator.com/item?id=10016052 - Aug 2015 (94 comments)
Brazil's Pirahã Tribe: Living Without Numbers or Time (2006) - https://news.ycombinator.com/item?id=8772388 - Dec 2014 (30 comments)
What happens when a language has no numbers? - https://news.ycombinator.com/item?id=6569989 - Oct 2013 (53 comments)
Language may shape human thought - https://news.ycombinator.com/item?id=2812011 - July 2011 (25 comments)
Pirahã: a non-Turing-complete human language - https://news.ycombinator.com/item?id=1033741 - Jan 2010 (39 comments)
How does language shape the way we think? - https://news.ycombinator.com/item?id=804101 - Sept 2009 (1 comment)
Pirahã defies Chomsky's theory of universal grammar - https://news.ycombinator.com/item?id=740687 - Aug 2009 (6 comments)
Recursion and human thought - https://news.ycombinator.com/item?id=283678 - Aug 2008 (6 comments)
MIT-led team finds language without numbers (update to earlier studies on the Piraha) - https://news.ycombinator.com/item?id=246459 - July 2008 (2 comments)
Recursion and Human Thought: Why the Piraha Don't Have Numbers (quicktime video) - https://news.ycombinator.com/item?id=61237 - Sept 2007 (1 comment)
Btw, there are a ton more of these; the above are just the ones that got comments.
> The endgame is probably not to fully automate these lists but to have software generate a starter version and then give the community ways to edit it.
I agree. There has been a lot of recurring-interesting topics discussed on HN but there's no one place to tag them all. Relying on you/someone else to pop up and give the curated links is currently the only way for others to know what have been discussed before.
> The authors compared animal and human communication, eliminating the aspects of vocalization that are shared by both, and concluded that one operation alone distinguished human speech: recursion. In the course of working on the article, Fitch grew sympathetic to Chomsky’s ideas and became an articulate defender of the theory of universal grammar.
"(…) distinguished human speech(…)", if you think about it, that's a weird assertion/assumption to make. The idea, that there's a clear line between human and non-human communication, as if this was something binary.
We humans have this weird concept, that there's "something", that makes us fundamentally different from the rest of the animal life on this planet. Yes, we've got the largest impact on the biosphere, we can reason about it. But if you were to track back on the coginitive abilities of our evolutionary ancestors, I'd say it would be very difficult to put a hard threshold in the phylogenetic tree, where suddenly a whole subset of means of communication discretly vanishes.
Just my 2 cents…
Agreed, there is no hard threshold (biology seems to avoid those anyways) but subjectively there is some sum of fuzzy attributes in humans which allows for us to achieve the highest complexity of lifestyle(?) of all lifeforms on the planet.
A beaver might instinctively know how to build a dam because it's built into its DNA, but a human can read the writings of past dam builders, learn the abstract theory of dam building, come up with their own improvements, debate those changes with other people interested in dam building, communicate their design to dam builders, and publish their work to become a permanent part of dam building literature. For a beaver to change how it builds dams, it would require a change in their DNA to pass on to future generations.
I'm not denying that humans, in their current form, as a species are set apart from the rest of life on this planet.
Rather what I'm trying to get at is, that if you'd backtrace the evolution of humans, at no particular point in time you could make a clear desitinction of "this generation of pre homo sapiens species fundamentally differs in their linguistic capabilities from the next evolutionary step".
The linguistic capabilities of hominids and humans more likely than not developed gradually, just like every other feature that makes a distinct species. Eventually you'll be able to clearly tell them apart. But when applying a "derivative" operator on it, you'll find that evolutionary development is smooth and continuous.
And I think that also applies to linguistic capabilities. The proposition I'm making is, that the linguogenesis of homo sapiens can not be explained within the boundaries of that species. Rather I'd say that the roots of our language can be traced back far further than you'd presume by presence of certain vocal anatomical features alone.
Unless the hard threshold was created by an external force, such as humans killing off or interbreeding with every other ape species capable of high level language.
Currently, sure. Across all time, I'd go with blue-green algae.
I knew that someone were to point this out. Yes, of course across geological timeframes at the moment we're playing second fiddle at best. But over the course of, let's say, that past 10 million years or so, I'd say, we're a firm contender for first place.
I was only 19 then, but I remember liking it a lot.
- Bits are the quantum of information.
- Genes are the quantum of genetics.
- Memes are the quantum of culture.
- Action is the quantum of physics.
If "idea" is the wrong word for the "quantum of human thought", what would be?
I'd preemptively say "word" is a poor candidate as it is in reference to oral language. If there's a distinction for a "thought word", then that would seem appropriate, just as in computer architecture "word" is distinct. Anecdotally, I usually don't think in English words, but in my own "ideas". Related ideas carry meaning and compose thoughts. Those thoughts I then translate into oral sentences via words. Words are the quantum of speech, ideas are the quantum of thought.
But that wasn't what the OP was asking, or at least not how I interpreted the question. The question was whether thought should be considered the mechanism/substrate underpinning human language. My comment was about the relationship between thought and language, about which of those shapes the other, and to what extent they should be considered independent or universal.
Yes, when we (adults) reflect on how we talk, we tend to form opinions and meanings in our head before we produce sentences. So that might lead us to conclude that thought informs language, and not the other way around. Yet children learn to speak before they learn to think, at least that's the prevailing theory in child development studies. So then "the idea" cannot be the quantum of language because it is only true for adults who have already learnt both.
Then we also have evidence from cultures like the one in the article, which shows that when a language is missing certain concepts (numbers in this case), the adults who grew up in that language also lack the abstract mental capacity associated with it (counting in this case). The blanket statement "the idea is the quantum of language" implies that ideas are universal concepts, yet the evidence we have shows that ideas are shaped by one's childhood language.
So, to reiterate my point: language and thought are most likely two interdependent processes, each of which shapes the other. The more interesting question is whether we should really consider them as separate processes, or whether "thought" is just a manifestation of an internal dialog with ourselves, in which "ideas" are the langauge.