> Fainali, xen, aafte sam 20 iers ov orxogrefkl riform, wi wud hev a lojikl, kohirnt speling in ius xrewawt xe Ingliy-spiking werld.
>> This text seems to be an example of a proposed spelling reform for the English language. The author has purposefully misspelled words to demonstrate how this reform would change the current spelling system. Here is a translation of the text into standard English:
>> "Finally, then, after some 20 years of orthographic reform, we would have a logical, coherent spelling in use throughout the English-speaking world."
>> The text suggests that after 20 years of implementing changes to the spelling system, the English-speaking world would have a more logical and coherent spelling system.
To be clear: the essay “The Awful German Language” is indeed by Mark Twain, while the short note “A Plan for the Improvement of English Spelling” is not.
As the saying goes, "Y is a crooked letter and can't be made straight.", although that has totally different meaning.
Similarly, the y in 'finally' makes an i sound (as in the i in 'machine'), while the y in 'tyrant' makes a different i sound (as in the i in 'finally'), and of course the y in 'year' makes yet another totally different sound. And then of course you have 'lye' and 'lie', which sound the same but with very different meanings.
That is what these jokes were about, simplifying so that each letter has only one sound and a spoken word can only be spelled one way while a written word can only be pronounced one way.
It doesn't work and is silly because letters take their sounds from context, combination, and position, as does the y in 'year'. And there aren't enough letters for all the sounds in English.
Actual source: A letter to The Economist (16 January 1971), written by one M.J. Shields (or M.J. Yilz, by the end of the letter). The letter is quoted in full in one of Willard Espy's Words at Play books. This was a modified version of a piece "Meihem in ce Klasrum", published in the September 1946 issue of Astounding Science Fiction magazine.
Fixing English consonants is the easy part. The hard part is consolidating the 25 or so different vowel sounds that exists in English. Given that we only have 5 vowel characters, this improved English would need a lot of vowel digraphs and/or accents.
To the contrary of most people, I think that English spelling is actually excellent because the spelling tells you less about actual phonology than etymology, and from there you can derive the pronunciation, thus the script is able to accomodate the accurate representations (phonologically) of every word from every language. There is no other language where people care about the correct pronunciation of words, because in most other languages the realm of possibility for the representation of sounds is extremely limited (take, for example, Hindi, with its supposedly phonetic script). Whereas Roman script can represent the accurate pronunciation of every word in Hindi, the Devanagari cannot even come close to representing all English words, not even all words written in Roman--which is probably why Roman script is becoming a dominant writing system in South Asia.
> Whereas Roman script can represent the accurate pronunciation of every word in Hindi
Isn't this only true if you already know how Hindi specifically is supposed to be pronounced? I don't think you could show someone a romanized Hindi text and expect them to even come close to an accurate pronunciation. For example, doesn't Hindi have like six ways of making the English "D" sound?
Same thing for e.g. Chinese, you can't represent tones with unaugmented Roman characters.
I was going to add a section about how the Hanzi script is the only other script which is actually able to carry information that Roman can't, which is why Roman script dominates everywhere but China and will probably become the default system of writing everywhere in the world aside from China. You can disagree w/ me, but if it was so awkward to use then you couldn't force people to adopt it on mass, the fact is that its so versatile that it really can be adapted to nearly any language.
No, as the "Ghoti" examples shows, contextual rules actually create infinite possibilities of phonological representations in the English script which is only exceeded by the IPA itself, since there are some edge cases, like "read" and "tear", where there is some confusion that wouldn't be clarified otherwise.
The Ghoti example shows just how poor English is at presenting sounds, relying on bizarre letter combinations and context. And it still cannot properly represent the large variety of sounds that exist in it, even if it reuses the same combination of letters for completely different sounds.
I don't know why being able to represent a great number of sounds with a relatively small set of characters is a bad thing, if English had to have separate signs for every single sound in the language, even if we didn't count loanwords, that would still be about 50 or 60 characters. And with it, well, why don't we just adopt the syllabary of the Japanese, which can almost never accurately represent loan-words phonetically. To my point, there is not a single script, aside from the IPA, that can represent more sounds than Roman script in English, because there is no direct correspondence between spelling and sound value.
> I don't know why being able to represent a great number of sounds with a relatively small set of characters is a bad thing
1. I didn't say it was a bad thing
2. English actually can't represent a grwat number of sounds using a small set: there are many, many sounds in English that are not actually represented by its set of letters. The most common of them, shwa, isn't in any shape, way or form represented by English language. It's a mess of different letter combinations in different contexts, none of which adequately express the sound.
To quote Wikipedia: "In English, schwa is the most common vowel sound". And yet, there's literally no way to represent the most common vowel sound in English. Instead, it's 6 different ways that heavily depend on context, using letters that have no relation to the sound to begin with.
3. You claimed that "in other languages the realm of possibility for the representation of sounds is extremely limited" which is, of course, untrue.
>"In English, schwa is the most common vowel sound". And yet, there's literally no way to represent the most common vowel sound in English. Instead, it's 6 different ways that heavily depend on context, using letters that have no relation to the sound to begin with.
I watched a video about this recently I found fascinating, I'll link to it below, but to summarize: the schwa isn't a letter because it occurs only in certain stress positions of a word, which changes depending on the surrounding words. But notice when you "talk slowly", that is, fixate on each and every word, you actually don't pronounce the shwa very often, if at all.
> I'll link to it below, but to summarize: the schwa isn't a letter because
Doesn't matter. You claimed English is so much better at representing sounds than other languages. And yet, it cannot represent the most common one. And of course, shwa is just one of them. As I mentioned before, English has significantly more sounds than letters, and of course cannot represent most of them even if those exist when you "soeak slowly".
More over, its so poor in expressing these sounds that it re-uses and repurposes completely different letter combinations for the same sound. Or the same letter for completely different sounds.
Because of course get/gist and use/you and fashion/mission/faction show us not the poor expressiveness of the language but how wonderfully rich it is in representing the different sounds.
English is way too irregular for this to work. Good examples are reign vs. sovereign and isle vs. island. The etymologies aren't linked and the modern spellings are basically due to scholars presuming the etymologies are linked and adjusting their spelling to reflect that.
this is actually explained by something called the Indo-European ablaut, or as the Indians who original came up with this call it, Guna and Vrddhi. Basically, there is a set of vowel grades in the IE languages: a -> (long) a -> (long) a; i -> e -> ai; u -> o -> au; (vowel) r -> ar -> (long) ar. The reason for the ablaut changes are fairly complex, but notice how in your example reign has the "e" as in pain, whereas "sovreign" has the "i" as in "bit", reign is actually the "guna" or first grade of the ablaut. English spelling doesn't generally represent these changes, nor does it have to, because people are aware of the sequence without knowing it, and those sound changes do not sound strange or out of place for them.
>scholars presuming the etymologies are linked and adjusting their spelling to reflect that
English actually has no official spelling for any words, its just that the King James bible was the first widely distributed printed book that people were able to read in the English speaking world, so people adopted the spellings for words in that book. Its quite arbitrary, which makes the elegance of the spelling system even better--the fact that no one got involved meant that we were able to create this beautifully complex system that could represent nearly an infinite number of sounds, because nobody tried to "standardize" and phoneticize English spelling.
This brings us to Sanskrit, which holds as its highest virtue the strict correspondence of spelling to sound, much to the dismay of learners who must at an early stage tackle the complex sandhi rules that enforce this.
The irony there is that Sandhi seemingly developed later, early Vedic Hymns, like Ancient Greek texts, have very relatively few external sandhi operations.
So how do we fix it? I remember Dutch having had a spelling reform at some point, and I'm pretty sure I've heard of similar things in both German and French. Does English (in any country or group of collaborating countries) have a standardization body or some volunteer group of people that might publish optional guidelines that people could follow?
Making it all consistent overnight would result in nobody being able to read (let alone write) the language anymore, but incremental tweaks that are gradual, obvious, and consistent could be doable.
I don't think we can, at least not while preserving its value. For one, english started as basically a pidgin or creole from the british isles being invaded by so many groups over centuries, so we're not starting off at the best place. Two, the other examples of spelling reform I'm aware of are either from highly centralized places (the French empire), or small groups of countries that border each other (German spelling reform). There's not really a way to get the US, UK, India, and parts of Africa to agree on how something is pronounced (and therefore agree on what the proper spelling should be).
The only way to "fix" it would be to have each country (or region within a country potentially), reform the language in a different way, which would destroy most of the value of having so many people in disparate places speak that language and be able to communicate.
It can be reformed. I specifically say that it can be.
My point is that, because you have so many disparate groups with completely different pronunciations (and therefore different target idealized spellings), reformation will destroy the majority of the value.
And nobody who actually speaks english has an interesting in destroying the value of understanding that language.
German (as well as French) is mostly 1 : n. There is typically only one possible pronunciation for a written word (exceptions exist). But for a spoken word there can be several spellings. You need to learn which one is correct.
English is n:m, everything is possible and you just need to learn it.
Finnish is >99% 1:1. A native speaker can should be able to write each word correctly. In practice spelling errors do exist. Partly because in colloquial speach / dialects not all words are pronounced fully according to the rules. Partly because some people are just unable to apply the rules. Not being able to pronounce a written word seems mostly impossible.
However, foreigners should not think Finnish is easy for that reason. There are many sounds foreigners cannot pronounce correctly. And several pairs of sounds that foreigners' ears cannot distinguish.
I learnt Finnish as an adult and have been speaking it for close to 30 years. Most Finns claim that I speak very well, but my daughter reminds me of pronunciation mistakes basically in every sentence. It's not about accent, it's just wrong.
In normal fluent speech I cannot hear the difference between short and long consonants. That does not hinder me to understand what people are saying. In 99% it's clear from the context. But I cannot produce the difference reliably either. When I try to make it short it might get long or the other way round. Normally I don't try anything and it often enough just gets wrong. If you cannot hear the difference yourself, it's of course difficult to get it right. I don't think many deaf people learn to speak well either.
While the wrong consonant length does not impact my understanding, native speakers are not at all tolerant to it. Many claim one cannot understand what is said. My theory is that bad Finnish is not a widespread concept. There are more speakers of bad English on the planet than native speakers of English. In the USA immigrants have been speaking bad English for centuries, nowadays it's practiced for business and tourism all over the world. Everybody has heard a wide variety of pronunciations. Immigrants have existed in Finland in slightly bigger numbers only for 20 years or less. And in the last ten years an increasing number does not even bother to learn Finnish. One can live with English these days. So native speakers are a rather homogeneous group. (Dialects exists but the differences are not huge and the more extreme one are rarely heard.)
Back to difficult phonemes. Probably most foreigners cannot pronounce "ä" correctly (it's different from the German "ä"). "äy" is probably even more difficult. The first word babys learn is "äiti" (mum, mother). It is surprisingly difficult for foreigners.
How is the only concrete possibility ranked at the very bottom of this subthread? Thanks! That at least gives some idea to how one might go about this and what effect it might have.
English is distinct in the sense that there is no one claiming authority over the language the way French Academy rules over French, or similar organizations in other languages. Simply put there is no right or wrong way to write English
Also, language is an evolving tool that we all co-own. If you want to see a change, modify how you write. People do this all the time. If it’s a good idea, maybe it’ll catch on.
There is nothing to be fixed; to imagine that there is some sort of Platonic Ideal English is absurd on its face - and in even asking the question of quote-unquote fixing it, you can really see the HN-stereotype penchant for a clear set of rules showing through. Linguistics moved past this a hundred damned years ago.
Natural languages are not programming languages. There is no such thing as "correct"; as with any language, there is only such thing as the many-varied ways in which native speakers speak.
Thə prablem with a sistəmatic, fōnetic chānj in Inglish speling is that it rāsəs thə questyən of wat acsent tū fāvər. Fər an Əmerican en Ōhīō, this transcriptshən prabablē māks səm sens, wuns yū get past thə shwa. I dout an Inglishman or ēvən ā Nū Yorkər wud fīnd it yusful.
But this is not phonetic. The Is in "Inglish" are hardly the same in any dialect.
And it doesn't solve the problem on the graphematic level. It ought to allow anyone to read the chosen dialect phonetically. But it suffers from an arbitrary choice of glyphs, so I have no idea how to pronounce eg. an (an actual problem I have pondered today is a-typical).
Conversely, if the spelling ought to be predictable for language learners, you are probably better of learning Mandarin (^_^)
Thr is anthr wy, whch ws dscvred ntrlly by early intrnt usrs: txt spk essntlly rduces englsh spllng to an abjad lk Arbc. Lk Arbc, Englsh is vry cnsnant hvy, and dialctl dffrnces are almst entrly vriashns n vwl qualty, so an abjad is the prfct wrtng systm fr us.
The perfect English abjad would use diacritics to mark vowels and some clever way to identify syllable boundaries (so you can distinguish ideal from idyll). But alas, we’re probably stuck with an alphabetic script forever.
Vowels were a later addition to many semitic languages as one example. Not sure that it would work so well in tonal or vowel-heavy languages but information is denser in consonants (no real surprise, I guess, given that they outnumber vowels).
Spanish has as much accent variation as English and has managed to develop a reasonable spelling where you know how to pronounce a vast majority of words just by reading them. The spelling doesn't have to be phonetic, just consistent. It should work for English, too.
Nope. Spanish dialects vary mainly by consonant sounds, for example, the lisped ("fervefa") vs non-lisped ("servesa") sound in words like "cerveza", and the standard ("yamas") vs Argentine ("shamas") sound in words like "llamas". The vowel sounds - "a", "e", "i", "o", and "u" - are extremely consistent throughout the Spanish speaking world.
English has a larger vocabulary and no diacritics (that anyone bothers to use), leading to more homophones as a consequence. Different spellings help to disambiguate.
"Thee proplem weth a systhemateck, fhoneteck chaigne en
Englush speleeng es thet et raeses thee questeon uv wut
aksent to faivor. Fer an Americon en Oheo, thees
transkriptshon probubly maeks sum sens, wunce yu get paest
thee skwaw. I daut an Engleshman or eaven a Nue Yorcer wud
fynd et ueseful."
please rewrite it but this time with a Southern drawl
"Th' proplem weth a systh'mateck, fhoneteck chaigne en
Aynglsh speln iz thayt et raeses th' questyun uv wut aksent
ta fayver. Fer a Amur'can down yonder in Ohayo, thees
transkripshun prahbly maeks sum sens, wunce y'all git past
th' skwaw. Ah daut an Engleshman or ev'n a Nue Yorcer wud
fynd et much ueseful."
- GPT-4 / https://sharegpt.com/c/Q5DW1wg
(prioritizing few "unforced" spelling changes or accent marks)
The problem with a sistematic, fonetic chainj in English spel'g is that it rases the question of wot axent to favor. For an American in Ohio, this transcription probably makes som sence, wonce yu get past the shwa. I dout an Englishman or even a Nu Yorker wùd fiynd it úsful.
Oh man, even if English is a pretty simple language having to learn the pronunciation of every word is tough…
I remember when I was mispronouncing recipe and no one had a clue what I meant (it wasn’t even in a cooking context). I was dumfounded when I found out how words like tomb or womb are pronounced. Or how about colonel? Or how nation is pronounced differently as a whole word vs when it’s part of national.
In contrast Spanish must be a bitch to learn as an adult due to all the verb tenses, but once you know the pronunciation rules you can read anything correctly even if you have no clue what you’re saying.
My wife has been in the US for about 25 years at this point, and we just had our talk about how to pronounce February, Wednesday, and turmeric yesterday.
The trick with French is not to think about it in terms of individual letters, but rather letter clusters, and then it becomes fairly regular. The individual letters can do wild things, but a given combination of letters generally has fairly clear rules about how it should be spoken. This is very different to English where, for example, "ough" can be pronounced in several different ways just by putting different letters in front of it. (oo, uff, ow, uh, oh, etc).
> Spanish must be a bitch to learn as an adult due to all the verb tenses
*adult Hungarian learner laughs*
You can memorize the Spanish verb tenses in a weekend. They're very regular and you essentially just have to memorize first, second and third person singular and plural for -er, -ar and -ir verbs. 3x2x3=18 endings. Using the regular form for irregular verbs is mostly understandable and even a common mistake for native speaking children.
Also, in addition to being a phonetically spelled language, it's also aspirated, which means that the sounds are pronounced very sharply and clearly and easy for ears to pick up.
It's entirely possible for an adult Spanish learner to become conversational in a couple of months.
Good luck learning all of the possible forms of Hungarian words. And though it's phonetically spelled, it's not aspirated, so the pronunciation sounds muted and mumbly, and it's hard to distinguish between sounds like D and N or P and B. Statistically speaking it takes 4-5 years of immersion for an English speaking adult to become conversational in Hungarian.
I've never tried to learn Hungarian but as a casual traveller there I deeply appreciated the phonetic spelling. The Hungarian alphabet can be memorized in an hour or two and then you can immediately pronounce any written Hungarian word and be understood (terrible accent I'm sure but the locals will get it).
By comparison, I've lived in Vietnam for several years now. The grammar is very simple. Every word is one syllable. A lot of common words are made from compounds of basic words (long horn animal = rhino type thing) so you can make educated guesses once you've learned a few words.
The problem? Nobody, including even my close friends, partner, and coworkers, can understand anything that I say in Vietnamese. Anything more complicated than "thank you", "hello", and a handful of numbers. And I cannot understand them either. If I pronounce a word as written but get the tone very very slightly wrong, it seems 100% impossible for Vietnamese people to infer the word I meant to say. It's incredibly frustrating and I've started and given up learning Vietnamese many times. I've learned some basic grammar, I've learned enough words to string a few sentences together. But what use is that when I go into a market and order something basic like "ice" ("nuoc da" , or just "da") and nobody has a clue what I'm saying?
Every conversation where I try to learn a new word in Vietnamese turns into ten minutes of "no, you're still not saying it right, the tone rises and falls at the end", "no, too much rise, say it like this" over and over until I give up. And that's not even mentioning the accents, differences in tone, and entirely different words used for basic things between the north and south of the country. Or the fact that locals are not in any way interested in trying to help a foreigner speak their language.
Learning Vietnamese is the most frustrating thing I've ever attempted in my life. I'm good at learning things. I'm sure I could sit down for a few days and memorize a hundred irregular verb conjugations. But I can't overcome this problem. How can you learn a language that you can't pronounce or hear?
I think Stuart Jay Raj has the best content on tones and making yourself understood in these tonal languages. Check out his Youtube channel: https://youtube.com/@StuartJayRaj
English has this problem too but with vowels. There are many different accents corresponding to different geographical locations. I can understand someone from Newcastle, or Birmingham, or Australia, or Texas, even though these accents have very different vowels and even numbers of vowels and vowel length, etc. Northern England and New Zealand English share the same vowels but mapped to different words[1].
But when a non-native person speaks English it can be difficult if their accent does not map to an existing named set of vowels. If they have learned a specific accent e.g. “International American” accent it makes life much easier. In many cases with minimal pairs, knowing what accent someone is speaking is crucial to comprehending what they are saying.
I suspect that something similar is happening in Vietnamese. It seems unlikely that everybody in Vietnam uses the exactly same set of tones, given the history of the country. Instead of learning Vietnamese tones maybe you need to learn the tones of a particular city and ignore speakers from outside that city?
Having traveled for many years, I've developed the ability to understand 95% of what people say in English even if their pronunciation is off. A funny thing about English is that actually the vowels aren't super important, except for in cases of ambiguity. Try it sometime. Yi cin riplic ill thi viwils in i sintinc with i singl viwil ind still indirstind. The problem is that most English speakers' ears aren't trained to expect unusual pronunciations, and it throws people off. This happens to me often when I speak Hungarian. Hungarians are not used to people learning Hungarian, so unnatural pronunciation can be really hard for them to follow.
English is really four languages (Germanic, French, Latin, Old Norse) that have been corrupted and morphed into an insanity of unnecessarily complex structures. The linguistic equivalent of four different cars taken apart and turned into an economy tractor race car used for pizza deliveries. Yes, it can go fast and deliver pizza to a farmer. But I wouldn't wanna learn to drive the bastard.
I figured out the gh at one point. Gh used to be its own phoneme. It slowly phased out of spoken english and got replaced by semi-random similar sounding phonemes. This g(h)/ch phoneme still exists in dutch and german though, and once you know about it you start finding a lot of cognates between the languages!
In fairness how many languages have letters that are always pronounced the same regardless of context? Even languages with highly regular pronunciation like Italian have letters that are pronounced differently depending on the other letters around them.
The o in women is really the only one that is completely irregular. The others follow.. not rules, but at least heuristics.
I'm not defending English pronunciation. It's clearly a mess. But this feels like a disingenuous way to demonstrate that. (Yeah I know it's clearly a joke.)
Hangeul very nicely reflects how Korean language's syllables change depending on what goes before and after them (phonetic context).
In comparison, most of irregularities with English spelling are either caused by extensive word loaning or by language changes that happened too long ago to make sense today.
This always irritates me somewhat. Let me link [0], as I do every time this comes up:
> Whenever the subject [of English spelling] comes up, someone is sure to bring up … George Bernard Shaw's ghoti -- a word which illustrates only Shaw's wiseacre ignorance. English spelling may be a nightmare, but it does have rules, and by those rules, ghoti can only be pronounced like goatee.
He goes on to substantiate the claim that it has rules, too, by explicitly stating and testing them!
> The purpose of this page is to describe those rules -- to explain the system behind English spelling, the rules that tell you how to pronounce a written word correctly over 85% of the time.
The way that's written gives the impression that the author thinks "over 85% of the time" is something to be particularly proud of - but I'm not so convinced it is. There's still an error rate in there of more than 1 in 7 words, and even this short paragraph I've written has 64 words. Following the rules should still result in 9 pronunciation mistakes, which seems like a lot.
jan Misali (or possibly others, but this is the first known occurrence to my knowledge) has argued that a much better example would be pfysche [1], which is within the observed English spelling rules and still hard to predict its pronunciation.
As a non native speaker I had a hard time with "scythe" which as far as I can tell is not matching all the rules in this page, and at the same time is relatively common.
I think rule 12 covers the vowel, and then rule 20 explains why it's sï+ not skï+ but I could be wrong about whether the vowel is a front vowel or not.
This irritates me immensely. Whenever somebody posts an article saying that this pronunciation doesn't make sense, somebody will assume that the article doesn't say that and then comment that it doesn't make sense posting yet another article. From the article:
> This fallacy arises from the incorrect application of the rules linking orthography to phonology1
I know that the pronunciation for this is all nonstandard and "wrong", and I still love it! So much so that I bought the ghoti.io domain name, and I'm currently writing a collection of C++ libraries under the collective project name "Ghoti.io". (.io actually means Indian Ocean)
First was a programming language called "Tang" ("Template lANGuage", also a type of fish). Project available on GitHub.
Second was a thread pool worker queue call "Pool". Also available on GitHub.
Now I'm working on an HTTP server named "Wave". For surfing the web, of course. Still working on it, not ready for public consumption.
Yes, they are bad puns and double meanings. No, I don't care. Yes, I'm enjoying it.
137 comments
[ 2.7 ms ] story [ 144 ms ] thread> What does this mean?
> Fainali, xen, aafte sam 20 iers ov orxogrefkl riform, wi wud hev a lojikl, kohirnt speling in ius xrewawt xe Ingliy-spiking werld.
>> This text seems to be an example of a proposed spelling reform for the English language. The author has purposefully misspelled words to demonstrate how this reform would change the current spelling system. Here is a translation of the text into standard English:
>> "Finally, then, after some 20 years of orthographic reform, we would have a logical, coherent spelling in use throughout the English-speaking world."
>> The text suggests that after 20 years of implementing changes to the spelling system, the English-speaking world would have a more logical and coherent spelling system.
Edit: wow, over twenty actually https://en.wikipedia.org/wiki/X#Other_languages
linguistics nerd flex
[0] https://faculty.georgetown.edu/jod/texts/twain.german.html
(it actually originates in a 1946 issue of Astounding Science Fiction. Sometimes the truth is stranger than astounding science fiction)
Similarly, the y in 'finally' makes an i sound (as in the i in 'machine'), while the y in 'tyrant' makes a different i sound (as in the i in 'finally'), and of course the y in 'year' makes yet another totally different sound. And then of course you have 'lye' and 'lie', which sound the same but with very different meanings.
That is what these jokes were about, simplifying so that each letter has only one sound and a spoken word can only be spelled one way while a written word can only be pronounced one way.
It doesn't work and is silly because letters take their sounds from context, combination, and position, as does the y in 'year'. And there aren't enough letters for all the sounds in English.
Actual source: A letter to The Economist (16 January 1971), written by one M.J. Shields (or M.J. Yilz, by the end of the letter). The letter is quoted in full in one of Willard Espy's Words at Play books. This was a modified version of a piece "Meihem in ce Klasrum", published in the September 1946 issue of Astounding Science Fiction magazine.
As read or read?
:-)
https://www.howtopronounce.com/reading-pennsylvania-2
https://en.wikipedia.org/wiki/Reading_Company
Isn't this only true if you already know how Hindi specifically is supposed to be pronounced? I don't think you could show someone a romanized Hindi text and expect them to even come close to an accurate pronunciation. For example, doesn't Hindi have like six ways of making the English "D" sound?
Same thing for e.g. Chinese, you can't represent tones with unaugmented Roman characters.
Interestingly this hasn’t stopped people from trying. https://en.m.wikipedia.org/wiki/Gwoyeu_Romatzyh
What.
Most other languages do not shy away from having more ways of presenting sounds native to that language than English.
There are 24 consonant and 25 vowel phonemes in Received Pronunciation alone. All inadequately represented by just 22 letters.
1. I didn't say it was a bad thing
2. English actually can't represent a grwat number of sounds using a small set: there are many, many sounds in English that are not actually represented by its set of letters. The most common of them, shwa, isn't in any shape, way or form represented by English language. It's a mess of different letter combinations in different contexts, none of which adequately express the sound.
To quote Wikipedia: "In English, schwa is the most common vowel sound". And yet, there's literally no way to represent the most common vowel sound in English. Instead, it's 6 different ways that heavily depend on context, using letters that have no relation to the sound to begin with.
3. You claimed that "in other languages the realm of possibility for the representation of sounds is extremely limited" which is, of course, untrue.
I watched a video about this recently I found fascinating, I'll link to it below, but to summarize: the schwa isn't a letter because it occurs only in certain stress positions of a word, which changes depending on the surrounding words. But notice when you "talk slowly", that is, fixate on each and every word, you actually don't pronounce the shwa very often, if at all.
Couldn't find the video but it turns out there are a million on the topic: https://www.youtube.com/watch?v=vp_WJAQ9fus
Doesn't matter. You claimed English is so much better at representing sounds than other languages. And yet, it cannot represent the most common one. And of course, shwa is just one of them. As I mentioned before, English has significantly more sounds than letters, and of course cannot represent most of them even if those exist when you "soeak slowly".
More over, its so poor in expressing these sounds that it re-uses and repurposes completely different letter combinations for the same sound. Or the same letter for completely different sounds.
Because of course get/gist and use/you and fashion/mission/faction show us not the poor expressiveness of the language but how wonderfully rich it is in representing the different sounds.
this is actually explained by something called the Indo-European ablaut, or as the Indians who original came up with this call it, Guna and Vrddhi. Basically, there is a set of vowel grades in the IE languages: a -> (long) a -> (long) a; i -> e -> ai; u -> o -> au; (vowel) r -> ar -> (long) ar. The reason for the ablaut changes are fairly complex, but notice how in your example reign has the "e" as in pain, whereas "sovreign" has the "i" as in "bit", reign is actually the "guna" or first grade of the ablaut. English spelling doesn't generally represent these changes, nor does it have to, because people are aware of the sequence without knowing it, and those sound changes do not sound strange or out of place for them.
>scholars presuming the etymologies are linked and adjusting their spelling to reflect that
English actually has no official spelling for any words, its just that the King James bible was the first widely distributed printed book that people were able to read in the English speaking world, so people adopted the spellings for words in that book. Its quite arbitrary, which makes the elegance of the spelling system even better--the fact that no one got involved meant that we were able to create this beautifully complex system that could represent nearly an infinite number of sounds, because nobody tried to "standardize" and phoneticize English spelling.
Ghoti - https://news.ycombinator.com/item?id=23581841 - June 2020 (239 comments)
Ghoti - https://news.ycombinator.com/item?id=2296927 - March 2011 (3 comments)
Making it all consistent overnight would result in nobody being able to read (let alone write) the language anymore, but incremental tweaks that are gradual, obvious, and consistent could be doable.
I don't think we can, at least not while preserving its value. For one, english started as basically a pidgin or creole from the british isles being invaded by so many groups over centuries, so we're not starting off at the best place. Two, the other examples of spelling reform I'm aware of are either from highly centralized places (the French empire), or small groups of countries that border each other (German spelling reform). There's not really a way to get the US, UK, India, and parts of Africa to agree on how something is pronounced (and therefore agree on what the proper spelling should be).
The only way to "fix" it would be to have each country (or region within a country potentially), reform the language in a different way, which would destroy most of the value of having so many people in disparate places speak that language and be able to communicate.
My point is that, because you have so many disparate groups with completely different pronunciations (and therefore different target idealized spellings), reformation will destroy the majority of the value.
And nobody who actually speaks english has an interesting in destroying the value of understanding that language.
Several. But for an easy grapheme-phoneme correspondence, you'd better look to Spanish. You can learn that in an hour or two.
> incremental tweaks
will slowly render older texts unreadable. Nobody (ok, almost nobody) can read old Dutch texts. Even 19th century Dutch is awkward to read.
English is n:m, everything is possible and you just need to learn it.
Finnish is >99% 1:1. A native speaker can should be able to write each word correctly. In practice spelling errors do exist. Partly because in colloquial speach / dialects not all words are pronounced fully according to the rules. Partly because some people are just unable to apply the rules. Not being able to pronounce a written word seems mostly impossible.
However, foreigners should not think Finnish is easy for that reason. There are many sounds foreigners cannot pronounce correctly. And several pairs of sounds that foreigners' ears cannot distinguish.
With previous experience of English, French, German and Spanish, when I look at https://en.wikipedia.org/wiki/Finnish_phonology I don't feel intimiated like I do by https://en.wikipedia.org/wiki/Standard_Chinese_phonology or get the sense of dispair that I get from https://en.wikipedia.org/wiki/Arabic_phonology
In normal fluent speech I cannot hear the difference between short and long consonants. That does not hinder me to understand what people are saying. In 99% it's clear from the context. But I cannot produce the difference reliably either. When I try to make it short it might get long or the other way round. Normally I don't try anything and it often enough just gets wrong. If you cannot hear the difference yourself, it's of course difficult to get it right. I don't think many deaf people learn to speak well either.
While the wrong consonant length does not impact my understanding, native speakers are not at all tolerant to it. Many claim one cannot understand what is said. My theory is that bad Finnish is not a widespread concept. There are more speakers of bad English on the planet than native speakers of English. In the USA immigrants have been speaking bad English for centuries, nowadays it's practiced for business and tourism all over the world. Everybody has heard a wide variety of pronunciations. Immigrants have existed in Finland in slightly bigger numbers only for 20 years or less. And in the last ten years an increasing number does not even bother to learn Finnish. One can live with English these days. So native speakers are a rather homogeneous group. (Dialects exists but the differences are not huge and the more extreme one are rarely heard.)
Back to difficult phonemes. Probably most foreigners cannot pronounce "ä" correctly (it's different from the German "ä"). "äy" is probably even more difficult. The first word babys learn is "äiti" (mum, mother). It is surprisingly difficult for foreigners.
Gud tu nou.
Yes, but this is true of literally every language. English is not special. This is Linguistics 101 stuff.
Friends don't let friends be prescriptivists.
Also, language is an evolving tool that we all co-own. If you want to see a change, modify how you write. People do this all the time. If it’s a good idea, maybe it’ll catch on.
Exactly: the the answer is an emphatic "No."
There is nothing to be fixed; to imagine that there is some sort of Platonic Ideal English is absurd on its face - and in even asking the question of quote-unquote fixing it, you can really see the HN-stereotype penchant for a clear set of rules showing through. Linguistics moved past this a hundred damned years ago.
Natural languages are not programming languages. There is no such thing as "correct"; as with any language, there is only such thing as the many-varied ways in which native speakers speak.
And it doesn't solve the problem on the graphematic level. It ought to allow anyone to read the chosen dialect phonetically. But it suffers from an arbitrary choice of glyphs, so I have no idea how to pronounce eg. an (an actual problem I have pondered today is a-typical).
Conversely, if the spelling ought to be predictable for language learners, you are probably better of learning Mandarin (^_^)
The perfect English abjad would use diacritics to mark vowels and some clever way to identify syllable boundaries (so you can distinguish ideal from idyll). But alas, we’re probably stuck with an alphabetic script forever.
This one took a moment. (It's "naturally", not "neutrally" or "entirely".)
The rest is surprisingly readable.
I remember when I was mispronouncing recipe and no one had a clue what I meant (it wasn’t even in a cooking context). I was dumfounded when I found out how words like tomb or womb are pronounced. Or how about colonel? Or how nation is pronounced differently as a whole word vs when it’s part of national.
In contrast Spanish must be a bitch to learn as an adult due to all the verb tenses, but once you know the pronunciation rules you can read anything correctly even if you have no clue what you’re saying.
Weddin just lies on the spectrum between Wodan and Odin.
https://en.wikipedia.org/wiki/Wednesday
When I noticed that first 'r' I questioned everything I thought I knew about the universe...
Except for the letter X which has 3 pronunciations, and some other weird stuff. Admittedly, nothing as bad as French or English.
Your comment actually articulates something I’ve always sensed but never explicitly acknowledged.
*adult Hungarian learner laughs*
You can memorize the Spanish verb tenses in a weekend. They're very regular and you essentially just have to memorize first, second and third person singular and plural for -er, -ar and -ir verbs. 3x2x3=18 endings. Using the regular form for irregular verbs is mostly understandable and even a common mistake for native speaking children.
Also, in addition to being a phonetically spelled language, it's also aspirated, which means that the sounds are pronounced very sharply and clearly and easy for ears to pick up.
It's entirely possible for an adult Spanish learner to become conversational in a couple of months.
Good luck learning all of the possible forms of Hungarian words. And though it's phonetically spelled, it's not aspirated, so the pronunciation sounds muted and mumbly, and it's hard to distinguish between sounds like D and N or P and B. Statistically speaking it takes 4-5 years of immersion for an English speaking adult to become conversational in Hungarian.
I've never tried to learn Hungarian but as a casual traveller there I deeply appreciated the phonetic spelling. The Hungarian alphabet can be memorized in an hour or two and then you can immediately pronounce any written Hungarian word and be understood (terrible accent I'm sure but the locals will get it).
By comparison, I've lived in Vietnam for several years now. The grammar is very simple. Every word is one syllable. A lot of common words are made from compounds of basic words (long horn animal = rhino type thing) so you can make educated guesses once you've learned a few words.
The problem? Nobody, including even my close friends, partner, and coworkers, can understand anything that I say in Vietnamese. Anything more complicated than "thank you", "hello", and a handful of numbers. And I cannot understand them either. If I pronounce a word as written but get the tone very very slightly wrong, it seems 100% impossible for Vietnamese people to infer the word I meant to say. It's incredibly frustrating and I've started and given up learning Vietnamese many times. I've learned some basic grammar, I've learned enough words to string a few sentences together. But what use is that when I go into a market and order something basic like "ice" ("nuoc da" , or just "da") and nobody has a clue what I'm saying?
Every conversation where I try to learn a new word in Vietnamese turns into ten minutes of "no, you're still not saying it right, the tone rises and falls at the end", "no, too much rise, say it like this" over and over until I give up. And that's not even mentioning the accents, differences in tone, and entirely different words used for basic things between the north and south of the country. Or the fact that locals are not in any way interested in trying to help a foreigner speak their language.
Learning Vietnamese is the most frustrating thing I've ever attempted in my life. I'm good at learning things. I'm sure I could sit down for a few days and memorize a hundred irregular verb conjugations. But I can't overcome this problem. How can you learn a language that you can't pronounce or hear?
But when a non-native person speaks English it can be difficult if their accent does not map to an existing named set of vowels. If they have learned a specific accent e.g. “International American” accent it makes life much easier. In many cases with minimal pairs, knowing what accent someone is speaking is crucial to comprehending what they are saying.
I suspect that something similar is happening in Vietnamese. It seems unlikely that everybody in Vietnam uses the exactly same set of tones, given the history of the country. Instead of learning Vietnamese tones maybe you need to learn the tones of a particular city and ignore speakers from outside that city?
[1]: https://youtu.be/3TfbtouLm14
The o in women is really the only one that is completely irregular. The others follow.. not rules, but at least heuristics.
I'm not defending English pronunciation. It's clearly a mess. But this feels like a disingenuous way to demonstrate that. (Yeah I know it's clearly a joke.)
Finnish does. But the Finnish language only gained a writing system in common usage relatively recently, in the 1800s.
So, relatively speaking, written English is just slightly worse than Hangul.[2]
[0]https://m.youtube.com/watch?v=btn0-Vce5ug
[1]https://en.m.wikipedia.org/wiki/Tibetan_script
[2]https://m.youtube.com/watch?v=K53oCDZPPiw
In comparison, most of irregularities with English spelling are either caused by extensive word loaning or by language changes that happened too long ago to make sense today.
So no, they're not the same.
> Whenever the subject [of English spelling] comes up, someone is sure to bring up … George Bernard Shaw's ghoti -- a word which illustrates only Shaw's wiseacre ignorance. English spelling may be a nightmare, but it does have rules, and by those rules, ghoti can only be pronounced like goatee.
He goes on to substantiate the claim that it has rules, too, by explicitly stating and testing them!
[0] http://zompist.com/spell.html
The way that's written gives the impression that the author thinks "over 85% of the time" is something to be particularly proud of - but I'm not so convinced it is. There's still an error rate in there of more than 1 in 7 words, and even this short paragraph I've written has 64 words. Following the rules should still result in 9 pronunciation mistakes, which seems like a lot.
[1] https://www.youtube.com/watch?v=TEsqY4MH40s&t=25s
Arguably the author concedes Shaw's point before moving on to more interesting stuff.
> This fallacy arises from the incorrect application of the rules linking orthography to phonology1
Fewer people seem to be aware that ghoti is silent:
gh as in right
o as in people
t as in mortgage
i as in business
https://www.aquarium.co.za/news/do-fish-make-sounds-yes-list...
First was a programming language called "Tang" ("Template lANGuage", also a type of fish). Project available on GitHub.
Second was a thread pool worker queue call "Pool". Also available on GitHub.
Now I'm working on an HTTP server named "Wave". For surfing the web, of course. Still working on it, not ready for public consumption.
Yes, they are bad puns and double meanings. No, I don't care. Yes, I'm enjoying it.
And I still love the word "ghoti"!
[1] https://en.m.wikipedia.org/wiki/Ghoti_people