Why Navajo is one of the most difficult languages (2023)
bigthink.com
bigthink.com
Someone else clarified this [0] but got flagged to death for being a bit too harsh, but I do think it's worth mentioning because this is used mistakenly as a prominent example of Navajo's difficulty: the raw numbers of consonants and vowels are not particularly extreme in Navajo, and the numbers that TFA cites for English are wrong.
They seem to have done a relatively simple analysis of the alphabet, but that's not how linguists think about phonology. The alphabet is our way of representing a much more complex system that underlies English, and English's vowel system in particular is far more complex than the alphabet makes it appear.
Our total number of consonants sits at about 24—the alphabet alone would make you forget about some of our affricates (ch as in charge) and some of our fricatives (sh as in shush, zh as in vision, th as in theme, dh as in though), plus a nasal (ng as in ring).
Our vowel system, meanwhile, has anywhere between ~16 in American English to ~25 in Received Pronunciation. Native speakers don't tend to realize just how complicated our vowels are, but getting them wrong is one of the most common ways to get recognized for having an accent.
All told, English has between 40 and 49 phonemes (again, depending on accent), which is quite comparable to Navajo's 45.
I've no doubt that Navajo is harder to learn for most people than English is, but I don't think the phonology plays a big role in that. English's phonology is plenty difficult.
English is hard to learn also because the alphabet doesn’t convey this complexity. In other languages one can easily tell the precise pronunciation of a word just from how it was written because the alphabet of those languages contain all the consonants and vowels. This isn’t the case with English.
Sometimes I wonder about how unfortunate perhaps it may be that a language like English became the “default one” for the world. The language is so sloppy with accents and different pronunciations that people can use tiny differences to “other” people based on social status, geography, and so on.
It arguably became the default international language of trade because it's so sloppy, and more precisely because the language is a mish-mash of other languages, readily borrowing words from everywhere, and this is why it seems so sloppy. Some other, highly regular language like German would never have become the international language of trade because those languages aren't sloppy, and so aren't very good for non-native speakers to use daily, and aren't products of a culture that got rich from trade.
As for tiny differences in pronunciations being used to discriminate against people, try learning German and moving to Germany and see how the natives treat you. You'll always be an outsider there, because you're never going to learn the language or pronunciation to that level as an adult, and the culture is known to be pretty insular. English-speaking cultures are generally far more welcoming to immigrants and people who don't speak the language like a native, and it's pretty easy to understand "broken English", which again is part of why it works so well for trade.
Spanish isn’t widespread in the Central and South Americas because of the language, nor Portuguese in Brazil.
It’s not a coincidence that stores on the Champs Élysées speak Mandarin. Languages are used because they’re useful.
As for being treated as an outsider, I actually moved and lived to the US for close to a year. I made tons of effort to try and speak good English. On the face of it, the locals were very appreciative - French accent is so romantic, am I right? In reality, they made absolutely zero effort to understand me when my pronunciation differed slightly from the correct one, on countless occasions. And believe me, I tried.
(Failing to realise the L in would, should and could is silent, so would and wood are pronounced identically, is half the French accent.)
I found this quite interesting about Japan. I went on vacation there recently, and especially in more rural places many people do not speak English. I don't speak Japanese, just picking up a few words while there. The people there put an immense amount of effort trying to understand what you're saying, e.g. somehow we managed to explain to a taxi driver the location of our hotel using only some elementary words like water, station, train tracks, up/down/left/right, yes/no, please/thank you. They are also very appreciative and often find it quite funny/interesting when you try to speak Japanese, even just simple words. As another example we also had an incident where a random passerby noticed we were trying to find our way, so (I assume) he asked us where we were trying to go, and then led us to the right bus stop. IME any similar situation in the US would more likely lead to people avoiding eye contact or pretending to be busy than trying to help, certainly not proactively without being asked to
Basically the same reason why Latin America speaks Spanish and Portuguese. It's not because the languages were "better" in some way, but because it was conquered by speakers of those languages.
And a side effect of being a global language is that you will have to get used to everybody from India to Australia to Kenya speak your language. English speakers are expected to meet widely different dialects even at the most formal business meetings.
And it doesn’t just win officially. I’ve seen groups without a common language default to English with people translating back and forth to other languages as needed even when more than one language was known to several members of the group. It’s not absolute, but there does seem to be an inherent advantage to English when translating across multiple languages.
Broken English works better than broken French / German. Not sure about Spanish, but something is going on beyond English winning because it was already ahead.
French later became one of many lingua francas, but that was after the term was already in common usage to refer to any language used as a bridge.
[0] https://en.wikipedia.org/wiki/Mediterranean_Lingua_Franca
lolinder has already pointed out that this is untrue, but I should note that it's also logically impossible - the Latin phrase referring to French would be lingua Latina, the Latin tongue. If you interpret lingua Franca as Latin, you get "the Frankish tongue", which is a variety of Germanic.
Which, while I'm here geeking out about linguistics anyway, allows me to break out one of my favorite tidbits about the history of English:
"Broken English" is super easy to understand in large part due to the fact that Modern English was forged as a sort of pidgin at the intersection of Old English, Scandinavian, and then French. The crazy admixture of peoples in England between 800 and 1200 meant that most of the complicated grammatical bits of the language got sanded down to the essentials. English lost grammatical gender, its full case system, noun classes (mostly), most declensions, and on and on.
What replaced all that is actually a relatively simple grammatical system (on the scale of Indo-European systems) that relies a lot more on adding words than on modifying suffixes. And adding words is a game anyone can play—you may not pick the same modifiers as a native would, but you'll get your point across!
Most of our words are content words, and the near-duplicates are, as you say, the result of borrowing from various other languages. These words don't serve a grammatical role and we could make do without them just fine.
The words that serve to replace grammatical features that exist in other Indo-European languages are a very small subset of our words—words like "will", "may" the prepositions, and similar. Non-natives will reach for other words to substitute for these grammar words when they're unsure which grammar word to use, but I don't think that's why those other words exist.
English, like many other languages learnt widely as a trade language by non-native speakers, has lost a lot of complexity, notably case system and gender. (Indonesian is perhaps the canonical example of this.) Unfortunately its spelling was standardized just before it underwent a huge change in pronunciation, and its habit of borrowing words from other languages without changing their spelling, has left it with fairly chaotic orthography. A few phonetic features - unusual consonants like th, difficult to learn rules around stressed syllables, nasty consonant clusters (squirrel!) - do make it tricker than it could be for second language learners.
It didn't become the default international language of trade due to any features of the language itself (it did that thanks to the British Empire and in particular its former colony in North America) - but many features of modern English are the result of so many people learning and speaking it.
Because it is so widely spoken by so many people, it does have a large vocabulary and a lot of idioms. But the grammar is generally not too difficult - no grammatical gender, no case aside from a few pronouns, outside of a handful of irregular verbs conjugations are pretty easy ("put an 's' on the third person singular"), and so on.
Yes it is, and you say so yourself:
>Unfortunately its spelling was standardized just before it underwent a huge change in pronunciation, and its habit of borrowing words from other languages without changing their spelling, has left it with fairly chaotic orthography.
>A few phonetic features - unusual consonants like th, difficult to learn rules around stressed syllables, nasty consonant clusters (squirrel!) - do make it tricker than it could be for second language learners.
Most other languages don't borrow words so profligately, and have much more regular relationships between their written form and their pronunciation.
>but many features of modern English are the result of so many people learning and speaking it.
Right, and that makes the language "sloppy", which was my point before. You can't have a language shared by so many different people and cultures as a common trade language and have it stay as orderly and precise as German, for example.
>But the grammar is generally not too difficult - no grammatical gender, no case aside from a few pronouns, outside of a handful of irregular verbs conjugations are pretty easy
Yes, and these features all make the language much easier for people to adopt as an international language, despite the sloppiness. I have a hard time imagining Russian or German having had this success in a parallel universe. As an example, look at Dutch: the Netherlands is another country that got rich from trade, and was also a colonial power. But who outside of the Netherlands speaks Dutch? South Africa is the only place the language actually went anywhere. Meanwhile, the former colonies of Spain all still use Spanish.
Look at the international languages before English - Latin and Greek - both beasts of languages to learn. Vulgar Latin's simplified offspring (French, Italian, Spanish) are largely the result of second language learners casting off the bits of Latin they couldn't or wouldn't learn.
I don't believe that English borrows more profligately than any other language. Historically it was hugely influenced by Norman French but nowadays I'd suggest more languages borrow from English than English does from other languages. The eternal fight between the Académie française and the people who actually speak French over borrowed words vs "native" ones is a good example of this.
If Latin was tbe second language and Spanish/Italian was not there yet, what was their main language then?
"Orderly and precise" doesn't mean simple. A Rube Goldberg machine can be orderly and precise, while still looking like a ridiculous amount of unnecessary complexity.
That has more to do with the way the Dutch administered their colonies than anything else. Take Indonesia for example, they didn't push Dutch that much and instead let Bahasa Indonesia stay the lingua franca. While some people could speak Dutch and there were some schools, it's not been pushed on the population. See this answer on reddit discussing this https://www.reddit.com/r/AskHistorians/comments/hg6sz3/how_c...
On the other hand, French is also a frustrating language with complicated grammar yet a lot of people in Africa still speak the language. That's due to the different philosophy of French colonialism which was the so-called "mission civilisatrice" (as an example see Jules Ferry's speeches among many others). A part of that was enforcing the use of French within its colonies.
Some other languages. Just considering the Latin alphabet: French pronunciation is so far from spelling that spelling bees used to be (may still be) prime time television. Words in Spanish and Portuguese may be written in a deterministic fashion, yet pronounced quite differently in different countries.
And sure, German spelling does reflect pronunciation…because the spelling has been changed from time to time to match shifts in pronunciation…specifically the pronunciation used by a small number of people in Niedersachsen. Elsewhere in Germany (much less Switzerland) the local pronunciation can be incomprehensible. (This was the case in the UK and France until the advent of radio and then especially TV).
And that’s just four examples using basically the same alphabet. This phenomenon is common to all living languages. Even dead languages, for that matter: the Latin and (classical) Greek pronunciation I had to use in high school was a UK version different from that typically used in the USA or (based on one person I met), Sweden!
And do you really think the same thing doesn’t happen with Arabic, Hindi et al?
English ain’t so special.
- nouns have no gender !!!!
- conjugation is extremely simple (only one variation for 3rd person singular for most verbs!!)
- tenses are limited !
- vocabulary is 40% French-Latin and 60% Germanic, most of Europe and America can use the subset they prefer
- the writing is latin, simple to learn, complete enough to describe well the sounds
- the native community is immense, with a large interesting cultural body of work, making it interesting to learn
- it's easy to start communicating but it can be extremely rich in higher levels of language: you start talking quickly and then can move to beautiful structures later: there is a quick reward and a long term reward
There is nothing unfortunate it became default, there is simply no better lingua franca today. We could try Japanese or Chinese, Hindi or Spanish, but they all have way more quirks and way less qualities !
I strongly reject people saying that English became default by historical chance, it could have stayed French, moved to German or Spanish, it SHOULD be Chinese, but all of them are too hard for adults to learn as quickly as English and never as rewarding. Just think of the time it takes to learn an ideogram-based writing system, all that to end up able to read only communist propaganda or very old poetry for instance...
Both the United States and Australia could very easily have been French speaking, and if they had been, I suggest we'd be having this discussion sur internet en français
I feel it's hard for me to make spelling mistakes in English, which is not my native tongue, but VERY EASY to make them in French, that I studied and used for decades: there is an inherent beauty to English, in its simplicity, we should not dismiss. (yes I'm re-reading three times to make sure I didn't spell anything wrong lol)
The reason Mandarin is not an international language and English is, is because learning English as your L2 gives you the ability to speak to people in the United States and England and Australia and New Zealand and India and lot more besides, where learning Mandarin gives you the ability to speak to people in ... China (and even then, historically, only one small part of China).
The ridiculous writing system is indeed my problem with it - and the ridiculous variations in tones too but they swear it's easy if you're born in it... which is ofc the problem: a lingua franca works if adult merchants can learn it.
English nouns have a person/nonperson gender distinction that requires agreement by relativizers.
Compare:
1. The girl who I met yesterday
2. *The burger who I ate yesterday
This gender distinction is hazily applied to the third-person pronouns he (person), she (person), and it (nonperson), which is why different people may see referring to babies with it as normal usage, or as grossly offensive.
But in German the girl you met yesterday is neuter, and the burger you ate yesterday is masculine, and you need to know that to form a grammatically correct sentence about them.
You failed to distinguish the concepts even in your own terms; a girl is a person, and a burger isn't, and you need to know that to form a grammatical sentence about them.
Even though a girl is a person the gender of a girl (das Mädchen) is neuter in German. You need to know that grammatical gender to form a grammatically correct sentence - you do in fact need to say "it" ("er") in your sentence referring to the girl you met yesterday.
A girl is person and a burger isn't, but in German a girl is an "it" and a burger is a "he".
Taking sex-based gender distinctions for our example, we can observe that Chinese speakers aren't able to correctly choose between the English pronouns he and she, despite the fact that this distinction isn't arbitrary in the slightest (and Chinese speakers routinely make it themselves in other contexts!). The reason is simply that their native language doesn't make the distinction, not that the distinction is theoretically difficult to make.
All of this is secondary, the real deal is more knowledge, especially useful knowledge is generated, and has been generated in the English language for a while now.
If you want access to the knowledge ecosystem of the world, want to learn skills, trade, want jobs, want to do business you have to learn english.
The other cultures just never caught on with this one simple thing. You have to make rationality, scientific temper and coherence the very foundational basis for your society. And then you need run these processes for long. Decades to centuries for your language to now become the carrier of knowledge. Until then other arguments are irrelevant.
But there's something strange with English knowledge: it's never quite deep. I find French materials are always a bit deeper, a bit more complex, we don't try to explain to 5 yo all the time, we don't always insist so much on the high level introductory stuff - we have a way to explain extremely complex concepts to more people I find. Just a feeling: when I want to explore I use English but when I want to dig I find more interesting stuff in French.
It's very frustrating because I live abroad now, in Hong Kong, and only use English with my friends, partners etc. And sometimes I want to explain to them, give them something to read, but it's only in French because no English speaker ever tried to think to dig deeper, to give more perspective, to ask the right questions etc :D
I speak a few languages (French, English, Japanese, German, Spanish, Chinese), some fluently, some less so and, honestly, in all of these languages, I can use tiny differences to other people based on social status, geography etc... It's easy in Mandarin despite me not being able to speak it that well. Japanese has a lot of variation between regions. I can quickly tell the level of education of someone speaking French (my native language) as well as guess where they're likely to be from, etc... I don't think that English is any different from any other languages on those points.
In the grand scheme of things, it's unfortunate that English lazily conquered the internet without any kind of preventive maintenance rooted in critical thinking. It's not nutty (maybe just futile) to argue that English desperately needs fixing. The distributed nature of English should be encouraging for experimentation.
Most native English speakers just roll over and succumb to Stockholm Syndrome.
Every would-be/aspiring spelling reformer has mentally composed a manifesto of righteousness.
As a trivial example: "hat" and "bat" => /h/ and /b/ are phonemes.
A more complicated example: "teeth" and "teethe" => /θ/ (voiceless th) and /ð/ (voiced th) are phonemes.
We have many other sound variants—phones [0]—which are important to get right if you want to not have an accent but which aren't actually used to distinguish one word from another. Pairs of these are called allophones [1], and probably the easiest example to think of in English is all the many ways we pronounce /t/.
What this definition means is that you can't get it under 40 if your goal is to be understood 100% of the time. We have 40-49 sounds (depending on accent) that are used to distinguish words, and if you don't have one of those in your toolbox then you will occasionally be misunderstood.
You could probably select a few that make it unlikely to come up often (/θ/ and /ð/ are good candidates), but you could also likely do the same for any other language, Navajo included.
That isn't really right; you can go below 40 without making any detectible dent in how often you're understood. That would cause problems if you only ever spoke in single-word utterances, but if you did that, no one would understand you anyway.
People blame a lot of communication problems on pronunciation when they actually arise from other sources. I once asked a student at an English-immersion school whether a particular class was happening that day, and was confused by his answer of "poss". When he clarified "stop", I was able to guess what he meant, and he got upset and asked whether the difference between "poss" and "pause" was really so important to English speakers.
It isn't; the problem is that you can't use "pause" to refer to the one-off cancellation of a periodic event. If you're using the word appropriately and you can't pronounce /z/, no one is ever going to be confused, not least because there is no word /pɑs/ to be confused about. But if you can't pronounce English and you can't speak it either, people will be confused when you try to say things. Trying to work on your pronunciation is absolutely the wrong approach to that problem; if you get it up to being flawless, most of the confusion you were struggling with will still occur.
I meant 100% literally. You can drop below 100% and still not have a detectable dent in understanding, and I even suggest a few phonemes that would make it extremely unlikely to come up.
The fact remains that phonemes are definitionally sounds that can cause confusion if you combine them or drop them, and to arrive at the set of phonemes we already dropped dozens of other sounds that a native speaker would use in certain contexts but that are understandable if you mix them up.
And the fact also remains that if you apply the drop-phonemes-and-see-if-people-understand-you technique to other languages you can reduce their number of phonemes as well, which makes the whole exercise kind of pointless in the context of trying to judge relative difficulties of phonologies.
The definition of phoneme that we use isn't some sort of natural category, but it is what it is because other definitions would be even more arbitrary and less useful.
Combining /θ/ and /ð/ will mean that it never comes up, not even once, unless you intentionally force the issue by using a sentence like "we saw her teeth" that positions the word in a context where it's not clear whether it's a noun or a verb. In most cases that can't be done, as witnessed by the fact that English nouns regularly zero-derive into verbs - for the zero-derivation to work, you have to be able to tell the difference between a noun and a verb even though they're pronounced identically.
On the other hand, you can't actually have meant 100% literally, because that's not an achievable goal. No native speaker can reach a 100% rate of being understood.
But a foreign speaker can easily achieve a rate of being understood equal to that of a native speaker while dropping some of the English phonemic inventory.
> And the fact also remains that if you apply the drop-phonemes-and-see-if-people-understand-you technique to other languages you can reduce their number of phonemes as well
That's true, but the question I saw was just "how many phonemes could you drop from English while still being understood?". And the answer is that English, like every other language, features a lot of redundancy; many things can go wrong before you stop being understood, but every time you do get something wrong, your odds get a bit worse. The amount of slack you have in terms of phonemes depends on how good you are with the rest of the language.
Okay, what about “cot” and “caught”? In some English dialects those are distinct phonemes, while in other dialects they are not. Yet someone who speaks a dialect without the cot-caught merger can still understand someone who speaks a dialect with the merger, even in the rare case that they use the words “cot” and “caught”, probably because you would never sleep on a caught or say that you cot a cold. Surely the “caught” vowel is one of the nine phonemes between 40 and 49, but I’m arguing that if you can drop nine whole phonemes and still remain mutually intelligible, you could probably drop a couple more and still be understood. For instance, while it isn’t any standard native dialect, Elmer Fudd’s “r/w” merger, while rendering him an object of ridicule, doesn’t seem to impede his intelligibility.
I'd assumed you were trying to figure out if English really had about as complicated a phonology as Navajo, and fiddling about with dropping phonemes isn't a useful approach for that because you could apply it to Navajo just as easily if you knew Navajo. If you get American English down to 35 I'd assume you could also get Navajo down to 40 by the same processes.
But yes, you can drop phonemes and be understood from other context as long as there is other context (which there usually is).
- English does not have 5 vowels; depending on the variety that you speak and counting diphthongs it has about 20
- "phenology"?
- tonal languages, while challenging to learn for someone who doesn't speak one natively, are not that uncommon, and 4 tones (assuming they're counting correctly) isn't that many. Mandarin has 4. Cantonese has (depending on whom you ask) up to 9. There are American languages with 10-20
- English has 24 consonants by most counts (confusing the Latin alphabet with English phonology again), but there are African languages with famously many, many more than that
- "like Spanish, Navajo is a verb-centric language in which syntax centres on actions" - I'm not sure what that even means. Spanish, like Italian, is a pro-drop language which means you can drop the redundant pronoun if you want because the verb inflection tells you everything you need to know. Navajo is (depending again on who you ask) an agglutinative/fusional/polysynthetic language which, while weird to English speakers, is not that unusual (e.g. Turkish, Finnish)
Also, of course, comparing the number of sounds to the number of English alphabets makes absolutely zero sense.
1. https://www.sciencedirect.com/science/article/abs/pii/S00243...
Eh, I’m not sure. There are probably ways to infer whether or not e.g. Mandarin Chinese is harder for Chinese children to learn than English is for American children to learn. In fact I’ve seen a fairly convincing argument that this is the case!
Kind of like you may hear a language, and recognize what it is, but no idea what they are saying.
The Navajo language was not written at the time, and it’s very different than others and subtle differences in tone can make all the difference. Their written notes of what they heard were probably generally inaccurate which made breaking the code impossible (vs enigma which was exactly copied, and the output is German, they just then needed to break it)
But even that wouldn't have worked.
There were Cherokee, Choctaw and Lakota code talkers in World War I.
1. https://www.cia.gov/stories/story/navajo-code-talkers-and-th...
He discussed how much-used languages have lots of adults learning it, badly. This smooths the rough edges away over time.
But tiny languages can be as hard as our brains can accommodate, which is very hard.
I'm sure they meant 2000...
The article references a nytimes article for this data.
There was a surge in enrollment in the tribe during the pandemic due to the way aid benefits were set up.
(Of course to officially enroll in a tribe you have to prove you are a member of that tribe)
Ouch that hurts, Mr Brinkhof
Of course, you could then make the argument that these vowels are encapsulated in English (as the same vowels in English have different sounds), but I think the article makes it clear in the sentences around your quote that these extra letters exist because Navajo has much more complex tonal sounds than English, not because they are making something implicit in English explicit in their language.
English has 21 letters which we identify as consonants and 5 letters which we identify as vowels, which are used to write ~24 consonants (depending on accent, but generally clustered around there) and a large number of vowels that is impossible to neatly catalog because it varies widely by accent—it's somewhere in the neighborhood of 14-25, depending on accent.
Chinese and Japanese have arguably some of the most complex "alphabets" (yes, Japanese has hiragana/katana, but they are more recent and Chinese only uses pinyin/bopomofo since the last century) and they have some of the least complex sounds of all languages.
The number of phomenes in (Mandarin) Chinese is indeed not low, but it has only ~1200 different syllables (including those differentiated by only tone differences), which is pretty low compared to other languages.
English has 26 letters which represents 44 phomenes, modern Chinese has ~8000 characters for 1200 syllables, I don't think you can find a system in there.
"Edit out swipes."
- words for half-empty/full - Eskimo-Aleut has 50 (?) words for different forms of snow
- tonality adds complexity; mandarin (4 tones + neutral), cantonese (9? tones)
- and various patois, not just in other languages, but also english (my exceedingly limited source being Graham Norton's youtube videos with the British/Scottish and Americans haha)
This one is actually a very good example of the problem that OP is describing of non-linguists writing linguistics pieces.
The "50 words for snow" story is one of the first linguistics cautionary tales they'll teach you in a college linguistics course these days, on the same level as the Hyatt Regency walkway collapse is for engineering.
The short of it is that it's based on a misunderstanding of the word "word". English is a very analytic language [0] on the scale of things, meaning it doesn't tend to augment its words to create new meanings. Instead we add adjectives and adverbs and prepositions to adjust a word's meaning.
Inuit languages, meanwhile, are highly agglutinative [0], meaning instead of independent adjectives they attach morphemes (meaningful-chunks-of-speech, think prefixes and suffixes) to the word itself to alter its meaning. That's not really the same thing as having X words for Y, because the words are constructed from parts that have their own, independent meaning—these morphemes are essentially a set of adjectives, adverbs, and prepositions that happen to get attached to a word instead of just placed nearby.
https://languagelog.ldc.upenn.edu/nll/?p=3497
https://languagelog.ldc.upenn.edu/nll/?cat=37
https://snowclones.org/2007/05/31/if-eskimos-have-n-words-fo...
And a book:
https://www.amazon.com/Eskimo-Vocabulary-Irreverent-Essays-L...
If they are using AI, it's not a good one. ChatGPT says: "English has about 14 to 20 vowel sounds, depending on the dialect. Standard American English typically has 14 distinct vowel sounds, while Received Pronunciation (RP) in British English has around 20. This variation occurs because different dialects have different vowel inventories."