Hacker News new | past | comments | ask | show | jobs | submit login
Languages That Could Change the Way You See the World (2015) (nautil.us)
62 points by ColinWright 11 days ago | hide | past | favorite | 50 comments

The way the author described the case of Yélî Dnye ("A Language Where Colors Are Metaphors") makes it seem a bit underwhelming to me, and unfortunately the link to the research paper is broken. When she says:

>For example, to describe something as red, islanders say “mtyemtye,” which is derived from “mtye,” or “red parrot species.” Another example is “mgîdîmgîdî,” which can be used to say something is black, but is directly derived from the word for night, “mgîdî.”

...all I get is the impression that the names of colours in Yélî Dnye are derived from known objects which happen to be of that colour—an interesting bit of trivia, but nothing shocking. Then:

>Not only that, writes Levinson, but the islander’s grammar reinforces this metaphorical slant, saying, “The skin of the man is white like the parrot,” rather than “He is white.”

...I am confused by what is supposed to be that surprising in this way of expressing themselves. Does "is white like the parrot" stand for a literal mention of both the word for white parrot and a word derived from it? Or is it a stand-in for the fact that they don't refer to things being a colour but rather being like specific coloured things? I feel like something has been lost in translation, both literally from Yélî Dnye and metaphorically from the field of linguistics.

I was confused by that too. Do they find the word "orange" equally metaphorical, given that it was named after the fruit?

The link to the 2001 article they base it from is broken, but it sounds like all their color words are just derived from nouns, maybe with some affix indicating 'of the color like'. The description is too vague to make a lot of sense.

There's lots of really weird and interesting things about Australian languages that would make better subjects for an article, like the avoidance register in Warlpiri. Hard to see why they chose that particular one.

From my understanding the difference lies in the object/animal chosen to describe the colour wrt. to the situation.

You’d use the red parrot if I may say to describe an angry wordy person becoming red of anger.

And you’d use another object/animal to describe the exact same colour in another context. Say red frog for your shining new prince or whatever.

The Matses language in the article is interesting, where the verb form indicates how you know the information you are imparting. Something like that would be very useful for HN, making people tag statements to indicate their reliability.

Imagine the benefit of verb tenses such as "I am an expert in this area", "I read something about this topic a while ago", "I know nothing about the topic and am confidently speculating", or "I didn't read the article and am commenting off the title." Edit: the key point is that this verb tense is required; you can't say anything without stating your confidence.

I would be skeptical of how seriously to take the interpretation in the article.

There are huge differences between English and Matses, such that naively talking what's encoded in the verb morphology is misleading at best. If you look at the section on verbs in the thesis on Matses, the author has this to say in the first paragraph:

>Epistemic modality (e.g., -chit 'Uncertainty') and evidentiality (e.g., -nëdac 'Distant Past: Inferential', -ash 'Recent Past: Conjecture' and -denne 'Remote Past: Experiential') are coded in Matses by derivational suffixes or as part of the meaning of inflectional suffixes, while English uses adverbs and modals like perhaps, evidently, really, must, etc.

Matses actually encodes the same information as English, it just does so through a series of particles attaches to the verb, which is not surprising in a highly synthetic language. Mind you, I haven't read all 1200+ pages, but I'd doubt the article's take on things.

This depends on information I don't have access to.

In Chinese, one may of course indicate the gender of a third party, but this isn't mandatory, where in Romance languages, it is.

If evidentiary particles are mandatory grammar, and that is the claim, then uttering a sentence without including those particles would just sound wrong, the way the Russian tendency to leave off the article of a noun when speaking English sounds wrong to us. Native speakers simply wouldn't do so.

> In Chinese, one may of course indicate the gender of a third party, but this isn't mandatory, where in Romance languages, it is

Romance languages have gender as a grammatical category, and it exists on all nouns. Semantically, gender is arbitrary, except in the case of natural gender on animate nouns. The need to express the gender of a noun is a purely syntactic requirement, which Chinese does not have. It seems like a valid comparison, but it's not really.

> If evidentiary particles are mandatory grammar, and that is the claim

And that's probably not really true. Instead, there's a conflation of epistemic mode and tense (and maybe aspect because why not) the verb system, and that's all tangled up with the other affixes. It's not clear from the article, (or the paper I don't have time to wade through) if there's a default mode/tense that speakers use, and then only clarify when necessary.

I get your point, but I think the actual facts are probably much less stark or interesting than the article makes out.

The linguistic term for this, as I recently learned from the Lingthusiasm podcast, is "evidentiality" and apparently a number of languages have it as a feature. Some languages get really granular with it, having different forms of expression depending on whether you've seen something, done something, overheard it, etc.



I don't see why I would trust grammatically-required self-assessments of expertise any more than I already do when someone tags one of those phrases (not much).

I think the effect would be greater than zero, but not as much as some might think.

Some people have a pattern of speech that implies a level of authority that they don't actually feel.

It's obvious when some people need to 'police their tone', less obvious for others.

I'll conjecture that languages will invariably end up with a more neutral option, even when such options seem required. Sometimes it is useful to the speaker to be ambiguous, whether it is to understate something, imply something, or just take a mental load off.

Ancient Greek words did that. For example, the word "know" has various forms:

Verbs: Ginosko, oida, epiginosko, proginosko, epistamai, sunoida, agnoeo, gnorizo

Nouns: Gnosis, epignosis, agnosia


Leaning a new language might change your perception of the world in different ways too.

If you live in a country where several languages are spoken, you might realize some media will happily say two different things on the same subject in different languages! Politicians too!

When I learned about the cardinal language folks a few years ago I thought it was so cool that I’ve tried to always keep in mind which way is north ever since.

Whenever I learn enough of a new language to speak sentences I feel as though the person speaking isn’t quite “me” but rather a remix of me. We all have facets and speaking a different language brings that out.

I never worry much about cardinal points when I'm home. But then I spent some months in two countries in which the addressing system does not use street names or house numbers but points of references and cardinal directions [1].

After about a month there, keeping track of the cardinal points became second nature because it was a necessary part of functioning in those countries. Since directions in conversations usually are "Just walk two blocks then one East"

However when I'm home, I stop keeping track where the North is, because it is not that necessary. And while we also speak Spanish in my country, we tell directions in a different way, using words like "Turn right, the left, downhill, uphill.

[1] Addresses in Nicaragua and Costa Rica are very unusual for the rest of the world. For example "From the Cowboy's Sour Milk, Two Blocks West, Right hand" is a completely valid address in Managua.

Just to flag that these are not necessarily consensus analysis, I can't speak for the details but there were linguistic papers discussing that these people actually use different directions - cardinal when outside, but simple relational inside. So they are not necessarily all day aware of every direction (which in any case is easier when you live in a small hut and normally are in areas you know well).

I discovered this in the US. It was either N Main Street, or within a building (west corner).

I traveled extensively to the US for 15 years but never managed to subconsciously keep track of the cardinal directions.

I hadn't heard these before and I find that even less exotic languages change they way you think about things. I didn't understand grammar like indirect objects until I studied Spanish. In Vietnamese, every word is a single syllable and you chain them together to make more complex descriptors. They also have no tense. So to say "I went to the store yesterday", you say "I go store yesterday" The form of go does not change depending on the time period. Languages are so cool and if I could make as good a living just studying them I would.

> In Vietnamese, every word is a single syllable

Would you consider "sân bay" to be two words but "airport" to be one? If so, then that's just an artifact of the writing system. After all, it could just be "sânbay" and "air port". There are no spaces in speech.

Yes but what he's really talking about are morphemes. English has many multisyllabic morphemes where as Vietnamese has few.

I remember (vaguely) my year of Mandarin as an undergrad was similar - though to be fair English is also capable of a lot of word-order tricks for a Subject-Verb-Object language, we don't have any trouble with "I went to the store yesterday" vs. "Yesterday I went to the store" or even "I went, yesterday, to the store".

Quite agree -- I really enjoyed my linguistics classes in college, and if I hadn't discovered programming, might have become a linguist.

> Interestingly, the Pirahã don’t seem to have a very high opinion of outsiders. They are monolingual, preferring to stick with their own lexicon rather than borrow words from English or Spanish, and they call all other languages, “crooked head.”

A quibble with something in the article -- why would the Pirahã borrow words from Spanish? They're in Brazil. And as it turns out, they do have a few loanwords from Portuguese.

Somewhere I got the notion that if you're doing translations for Brazil you had to support Portuguese and Spanish, which is why you need internationalization instead of localization. It turns out that there are maybe half a million Spanish speakers in Brazil, but plenty of other languages, especially in the western half.

> Brazil’s immigrant languages include Catalan, Dutch, Japanese, Korean, North Levantine Spoken Arabic, Turoyo and Vlax Romani, as well as more mainstream European languages like German, Italian, Polish and Ukrainian.

> you had to support Portuguese and Spanish

I wonder about that (maybe internationalization for Latin America?). Portuguese, as far as I know, is very dominant in Brazil (spoken by 98% of the population). Spanish might be spoken as an L2 in the border states, but I don't know that there's any significant Spanish internationalization in Brazil. Most Brazilians I know say that whenever they have to communicate with Spanish speakers, they generally muddle through with Portuñol, a made-up pidgin of Portuguese with a few Spanish words thrown in to ease comprehension. Ironically, Spanish isn't a very popular language in Brazil (only 4% of the population speak it).


Come to think of it, I've seen more signage in Spanish in the United States than I've ever seen in Brazil. That said, I've only ever been to Rio and São Paulo. Would be interesting to get thoughts from Brazilians.

Nah, if you go into a borderline city, odds are almost 1 that you'll signage in Portuguese at the Spanish side of the border. Brazilian speak Portuguese, there's no details to add.

Those 2% (that feels way too high, I don't trust this number) are insulated immigrants (probably speaking German, Japanese, or Italian) or native communities speaking their own language. Those are very small and fringe communities, most first generation immigrants and natives speak Portuguese too (but as a second language).

Various sources cite Ethnologue's figure of 460,000 Spanish speakers in Brazil, which is 0.2% of Brazil's 211m population, so yes, much lower than the 4% figure cited in Wikipedia.

In fact, it seems there are many more German speakers in Brazil (about 4.5m), especially in the southern states.

The wikipedia numbers do not add to 100%, so I guess they accept multiple languages. Still, if that's the case, 7% for English is a bit too low, it should be closer to 15%.

Spanish is spoken by more than 4% of the US. It is an official language of New Mexico, the dominate language of Puerto Rico, and has a significant number of native born speakers in states bordering Mexico (which at one time owned those states). There are also a lot of immigrants (mostly from Mexico - both legal and illegal) who speak Spanish.

If you squint hard enough Portuguese is a dialect of Spanish, as is Italian. Native speakers will disagree with that statement, but most non-native speakers who have learned Spanish and one of the other agree that once you know Spanish it isn't a big deal to learn the other.

> If you squint hard enough Portuguese is a dialect of Spanish, as is Italian. Native speakers will disagree with that statement, but most non-native speakers who have learned Spanish and one of the other agree that once you know Spanish it isn't a big deal to learn the other.

I think Portuguese and Castilian Spanish have common roots (they're on a continuum of West Iberian). Italian is also related but belong to a different branch of the linguistic family. Here's a nice diagram:


It is possible they or I got our wires crossed about the fact that South America > Brazil. This story has been retold so many times that it is starting to become myth. Typically I use Switzerland as my go-to example of "Even if it's one country you can't assume one language." But Quebec is also next door - and to be perfectly cynical, you can count on people who only care about English and don't give a damn about anybody outside their bubble to have a politically incorrect Quebecois joke pop into their heads. They get it, even if they don't like it.

That makes sense. From my perspective, having worked in Brazil (and having been a student of the Portuguese language) for some years, I've come to understand there's actually a strong pride in the Portuguese language in Brazil and in being distinct from the Spanish-speaking countries around it, and this has resulted in Spanish being a tiny minority language in Brazil. The other contributing factor I think is that most Brazilians have little reason to actually learn Spanish because the languages are similar enough to enable simple interactions (at least on the surface -- they diverge significantly upon diving deeper). As a result, the presence of Spanish in Brazil is actually very small. Hence my original comment that it would be very unusual for Pirahã words to have the possibility of Spanish as an influence.

The French situation in Canada is a little different (speaking as a former Quebec resident). French is a co-official language in Canada spoken as an L1 by 20% of the population (and L1 speakers aren't only in Quebec, but also Alberta, Manitoba, Ontario and New Brunswick). Also, internationalizing for French isn't just a good idea, it is the law. Retailers in the U.S. know that consumer products without French labeling cannot be sold commercially in Canada.

As a layman with a slight interest in linguistics, I keep coming accross the Pirahã language as an example that refutes the Chomskyan idea of universal grammar. As far as I understand it, this is not actually true and the professional community mostly sees Everett as a charlatan. Is this correct? Does anybody know if there are professional linguists who take his theories/observations seriously?

Chomsky and most of his students call Everett a charlatan. For example, David Pesetsky, whose doctoral advisor was Chomsky, criticized Everett's work in the June 2009 issue of the Journal of the Linguistic Society of America, Language. Chomsky and Pesetsky may not be the most objective critics, though.

I read Everett's popular book on the Pirahã, "Don't Sleep, There Are Snakes", and found his descriptions believable, but I'm definitely not a professional linguist.

There's a nice interview with Everett in https://www.edge.org/conversation/daniel_l_everett-recursion... which includes details on the controversies with Chomsky's ideas.

Given Chomsky's refusal to admit he was wrong in his political judgments about Cambodia and Sudan, which some observers go so far as to label hypocrisy, I'm tending to cheer for Everett and hope he is correct.

Depending on how you define student, everybody who ever took a linguistics class is a student of Chomsky. But you raise a good point about objectivity.

So who/where are the disinterested supporters of Everett's work? What do they believe that the Chomskians are getting wrong about Everett's work?

The key word in my query here is disinterested, because almost all criticism of Chomsky is (implicitly or explicitly) based on his political views, rather than his linguistic scholarship.

He’s not taken too seriously.

My experience: several years ago I watched Everett give a talk on Pirahã, with only a quick skim of his paper (the first one that got him attention, forgot the name). He gave a basic talk about his work, with examples and translations, etc. I wasn’t convinced by his argument, and some of his answers to fairly reasonable questions were defensive.

Part of the problem is that (at the time, maybe still) all of the data on the language was collected by Everett (maybe with help). I believe he did the the translation, too. It’s reasonable to be skeptical about the data, particularly given the claims based on it. Also, he’s asserting a negative - that Pirahã lacks syntactic recursion, and you can’t prove that adequately with limited data that comes only from you (logical difficulties aside).

Thanks for the answer. Indeed, the fact that he was working only off of his own data felt like a red flag to me.

An interesting thing to follow would be if there are other people who went into the field to work with the Pirahã; and if not, why not, given the extraordinary nature of the claims?


A bit more insight here: https://news.mit.edu/2016/data-amazonian-piraha-language-deb...

> In the other two cases, Everett’s translations of the sentences differ from those of Sheldon. Everett believes there is no conjunction in those sentences, although Sheldon’s original translations suggested there was.

For instance, a Pirahã sentence transliterated as “ti xaigia ao ogi gio ai hi ahapita” was interpreted by Sheldon to mean, “Well, then I and the big Brazilian woman disappeared.” The conjunction of “I and the big Brazilian woman” could be an example of recursion. But Everett believes a better translation is: “Well, [with respect to me], the very big foreigner went away again.” And that sentence has no recursion.


> a graduate student at the University of British Columbia, Raiane Salles, has conducted recent fieldwork among the Pirahã unearthing potential evidence of “possessor recursion” — which creates phrases that could expand infinitely, such as “the foreigner’s parent’s dog” and “Migixoi’s husband’s mother’s clothes.”

So it seems there are already some signs of possible recursion in the language. On top of this, there's always the possibility that the Pirahã can understand recursion, but they just don't use it in their language. In the end this would be the more important thing to establish - do these people have a different mental apparatus from the rest of mankind? or do they just make use of the same thing differently...

Side note: I'd like to point out one of the very few science fiction novels about linguistics: The Languages of Pao by Jack Vance [1]. The novel is clearly inspired by the Sapir-Whorf concept that language shapes thought [0]. BTW: Another candidate would be 1984. Though not the main these, that novel introduces Newspeak, a language in which dissent cannot be expressed, thus becoming another tool of oppression.

Minor spoiler: In Vance's said novel, the different segments of the population of a planet are taught three separate languages in order to develop and specialize economically, industrially, and militarily.

[0]. https://en.wikipedia.org/wiki/Linguistic_relativity

[1]. https://www.google.com/books/edition/The_Languages_of_Pao/ke...

It's not really a main plot point, but Iain M. Banks's Culture uses a constructed language that de-emphasizes or lacks easy descriptions of the concepts of possession and ownership, dominance/submission, and aggression.

Babel 17 by Samuel Delaney is another classic in this genre: https://en.wikipedia.org/wiki/Babel-17

I've been unable to find a linguistics SF story based on a theory other than Sapir-Whorf

i think that Anthem by Ayn Rand also deserves an honorable mention in this list

Learning about (Ithkuil)[https://en.wikipedia.org/wiki/Ithkuil], a constructed language that incorporates a few of the ideas in the article along with many others has changed the way I see language and the effect it can have on thought.

How much of how we construct ideas is constrained by historical accident? This is the Sapir–Whorf hypothesis and a well known question. But Ithkuil makes me wonder, what thought is possible if we build a language with expressiveness that goes beyond anything we have naturally?

There is a SF novel that taps into these ideas:

The Player of Games, by Iain M. Banks

Is there a language with no word for money?

My first thought was some of the gift cultures of First Nations people, but the most obvious group that came to mind is the source of the word 'wampum', so that probably kills that theory.

I think we downplay how much 'international trade' went on, and for how long. The Hopewell civilization seems to have flirted with Egyptian levels of long-distance trade. The Mississippi river valley is a hell of a highway system.

You can't infer anything from anyone who was in the Americas before Europeans. They didn't have a written record to work with, and between disease killing vast numbers; and the introduction of guns and the horse (other European technology too, but those are the big ones): their cultures were all in major change as we observed them.

They probably didn't need much money then. Between being a barter society in general, and the affects of a small population relative to what the land could support there just wasn't as much to trade with each other. When someone did want to trade the Europeans had much better goods so even more of their old money wouldn't be used much anymore.

Or in different words it is a lot easier to give gifts when almost all your relatives just died each leaving you a large inheritance of things. We can guess that some amount of selfishness would have come [back] to the gift giving cultures if things would have been allowed to return to "normal". That isn't to say gifts wouldn't have been important to the culture, only that they would be relatively rarer as there wasn't as much excess anyone could afford to give.

Robin Wall Kimmerer (Onondaga) talks a good bit about gift economies, and she ties it all the way back to their oral histories. I doubt very much that you would rewrite your parables to fit the new normal. While that is a very Western thing to do, projecting your own baggage onto others is usually a mistake.

Oral history is well known to be unreliable and changes a bit with each retelling.

I didn't say they didn't have a gift culture before - those exist in other places as well. I suggested that at the time we were able to observe them there were a lot more gifts than normal, and had been for a hundred years or so - just long enough to seem like normal, but not really normal. Before and after the disruption there would be gifts - many people like to give gifts and it is a way to show your love for someone which is important. However in normal times there is a large scarcity limit to how many gifts you can give that doesn't apply.

Applications are open for YC Summer 2021

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact