Hacker News Reader

The problem is that "more irregular" is not well-defined.

Language works[1] in stacked layers a bit like the OSI model.

* Raw sound is interpreted as phonetic units

* These are then reinterpreted as phonemes according to phonological rules. For example, you perceive "butter" as being pronounced "b - uh - t - r", but, in American English at least, that "t" is pronounced more like a Spanish "r" than anything you would recognize in isolation as a "t".

* Phonemes are organized into morphemes (units of meaning, e.g. "happier" has two morphemes, "happy" and "er").

* Morphemes are combined into words according to morphological rules.

* Words are combined into sentences (or other types of utterance) according to syntax. Popularly when people say "grammar" they mean morphology + syntax.

You could talk about "regularity" or "irregularity" completely separately for any of these layers and even then you would have to carefully define what you meant and how to measure it. And it's not clear that English is particularly bad in any of them.

The phonology example I gave might sound weird but all languages have stuff like this and English is far from being the most confusing[2]. For morphology you could propose a standard like "frequency of verbs with irregular conjugation" (or nouns with an irregular declension) and point to the prevalence of verbs like "sing/sang/sung" in English, but these are not English specific; most Germanic languages have roughly the same set of these "strong" verbs. (In Icelandic, "to sing" = "syngja" but "he sung" = "hann söng", rather than something like "hann syngði" which you would expect regularly and which doesn't actually exist[3]). For syntax I don't have a clue how you would even begin to measure regularity or irregularity.

Another thing "irregular" could refer to is the writing system. But first of all, writing and [spoken] language are kinda separate things (and a huge chunk of languages have no writing system at all). We could start writing English with a totally different alphabet tomorrow and it would still be recognizable as English and not fundamentally different. This has actually happened to several languages in the past; for example, Korean was written with Chinese characters and now it is written with the totally unrelated Korean alphabet, but it's still Korean either way. Serbs today may write Serbian with either the Cyrillic or the Latin alphabet, but either way, it's still Serbian.

But even if we accept for the sake of argument that talking about writing systems is meaningful, English is still far from the most irregular. In Japanese for example, it is basically completely impossible to even begin to predict how to write a word if all you know is its pronunciation. And then there is diglossia: people in Egypt, for example, would believe that what they are speaking and what they're writing are both "the Arabic language", but the written form and the spoken form are so dramatically different in every way that from an objective perspective they are separate languages, a bit like how during the Middle Ages people spoke French, Spanish, etc. but wrote in Latin.

[1]: this is all a rough oversimplification

[2]: This Quora answer is pretty interesting: https://www.quora.com/Which-language-has-the-most-complex-ph...

[3]: I think, anyway. But I don't really know Icelandic so someone can feel free to correct me.

"In Japanese for example, it is basically completely impossible to even begin to predict how to write a word if all you know is its pronunciation"

Do you mean using the Kanji characters, or is there something else (other than the silent "u" that hides) that would trip you up when trying to write something in the hiragana or katakana letters?

I'm referring to kanji, which is what most words are written with.

Kanji you are right, but I’m not sure most words are written with it (happy to be corrected, I only did a tiny bit of Japanese study and was only in Japan for a week or so). Plus it is possible to write a word using either of the other two alphabets even if you don’t know the kanji.

Mandarin, Cantonese and likely other Chinese languages are maybe a better example. I don’t think they have a phonetic system that you could fall back on

You could in principle write anything you want in any script you want. But if you write a word that is typically written in kanji in hiragana instead, you are not writing according to standard Japanese writing conventions.

This is a non sequitur, it is like saying "Chinese writing is not irregular because you could always just write everything in Cyrillic"

Well not really - if a Japanese speaker write a word to another Japanese speaker they could conceivably write the word out through either of these alphabets - they're mutually understood and they are able to perfectly (iirc) represent all of the words used in the language.

I don't think anyone would suggest that it's possible to do this with Chinese and Cyrillic characters. I actually think you'd be the first to suggest this is the same thing. I'm really confused why you'd say this.

Russian stress is unpredictable, not even per word, but among different inflections of the word. That holds for both nouns and verbs.

For the nouns there's an attempt at classification that uses both numbers and letters[0] and even then, a lot of commonly used words fall out of it. English, on the other hand, only has half a dozen umlauted plurals (foot-feet, tooth-teeth, ..), a tendency to keep some plurals invariant (one fish, two fish), and just one suppletive plural (person-people)

The verb has many forms, some of which like the past passive participle, are unpredictable especially with regard to stress. This compared to English, which has the simplest non-trivial conjugation you can imagine (just add an -s), and about hundred verb roots that are semi-irregular (semi- because there's a lot of patterns)

0: https://ru.wiktionary.org/wiki/Викисловарь:Использование_сло...

And that's not even mentioning things like genders, which in most languages follow no rules at all, which makes them all "irregular". English closest cousin, Dutch, has 3 genders which for are completely random for all but compound words and derivations.

The claim that English would be the most irregular language in the world boggles my mind. What even is irregular about English? How can you even make such a claim without deep knowledge of many languages from many different language families.

>What even is irregular about English?

I'm guessing it comes both the inconsistent pronunciation and irregular verbs. People then assume non-English languages have more consistency and regularity. It's a non-rigorous intuition that English is messier than every other language.

There are several poems that poke fun at it. It would be interesting to see equivalent "meta" poems about Russian, Spanish, etc.

(1) irregular pronunciation:

http://www.wordhord.com/humor/english-pronunciation-poems/

(2) irregular verb tenses:

Tense Times With Verbs by Richard Lederer[1]

  The verbs in English are a fright.
  How can we learn to read and write?
  Today we speak, but first we spoke;
  Some faucets leak, but never loke.
  Today we write, but first we wrote;
  We bite our tongues, but never bote.
  Each day I teach, for years I taught,
  And preachers preach, but never praught.
  This tale I tell; this tale I told;
  I smell the flowers, but never smold.
  If knights still slay, as once they slew,
  Then do we play, as once we plew?
  If I still do as once I did,
  Then do cows moo, as they once mid?
  I love to win, and games I’ve won;
  I seldom sin, and never son.
  I hate to lose, and games I lost;
  I didn’t choose, and never chost.
  I love to sing, and songs I sang;
  I fling a ball, but never flang.
  I strike that ball, that ball I struck;
  This poem I like, but never luck.
  I take a break, a break I took;
  I bake a cake, but never book.
  I eat that cake, that cake I ate;
  I beat an egg, but never bate.
  I often swim, as I once swam;
  I skim some milk, but never skam.
  I fly a kite that I once flew;
  I tie a knot, but never tew.
  I see the truth, the truth I saw;
  I flee from falsehood, never flaw.
  I stand for truth, as I once stood;
  I land a fish, but never lood.
  About these verbs I sit and think.
  These verbs don’t fit. They seem to wink
  At me, who sat for years and thought
  Of verbs that never fat or wrought.

[1] https://www.ncra.org/files/MCMS/2ED22AD5-E440-4ECB-B7B4-9701...

Calling the pronunciation irregular is putting the cart before the horse. The spoken language is the fundamental thing. It's the spelling that is irregular; the pronunciation just is what it is.

Writing isn't the same thing as language. Language is an innate human activity; whereas writing is a piece of technology that some cultures use and others don't.

>Calling the pronunciation irregular is putting the cart before the horse.

Spelling sometimes precedes the pronunciation. That's why Detroit is not pronounced "day-twah" like the French would say it. Because English borrows so much from German, French, Latin, Greek, with some Scandinavian and native Indian thrown in, Americans often look at the spelling and derive an Americanized English pronunciation. Put another way, a lot of English sounds the way it does because of the way it is written. The spellings originated from different parent languages but the pronunciation did not tag along with it. This adds to the inconsistency.

As far which causal direction (pronunciation->spelling vs spelling->pronunciation) has the higher quantity in English, I don't know.

I agree that in the large scheme of evolution from apes to humans, spoken language is more fundamental and will come before writing.

It's true that writing can influence language (as can lots of other things). I wasn't claiming otherwise. The magnitude of each direction does not matter and is not related to the point I was trying to make.

The point I was trying to make is: an inconsistency between writing and language, which are two different things, is not evidence of an internal inconsistency in the language.

It is like saying the Linux syscall API is "irregular", because it is incompatible with the Windows one.

The fact that "bough" and "trough" are spelled the same is not a fact about the English language so it doesn't prove anything about whether English is irregular or not. It is a fact about the writing system commonly used to represent English.

Edit: with your comment about apes to humans, you make it sound as if writing was invented in some earlier stage of evolution (but after language). Actually this is not true. Writing was invented by biologically and behaviorally modern humans indistinguishable from you and me. It is still an optional part of the human experience and not used by all cultures. The complete opposite is true of [spoken] language.

>The fact that "bough" and "trough" are spelled the same is not a fact about the English language so it doesn't prove anything about whether English is irregular or not. It is a fact about the writing system commonly used to represent English.

I understand your separation of concepts here. Let me try to go meta and explain what many people are talking about:

umanwizard : English _aural_ is totally separate from English _visual_ ; in this division, "inconsistency" makes no sense; (English aural would be like Linux and English visual is like Windows)

others: "English language" is _both_ the aural+visual together. The aural & visual evolved together and influenced each other. Therefore, inconsistencies _can_ be identified. It doesn't matter if there was an "English spoken" before "English writing". The context for many people saying "English is inconsistent/irregular" is to say that "English aural+visual is inconsistent".

I'm not saying you must agree that "English language=aural+visual". I'm informing you that the combined unit is what they're talking about. The combined concept enables those meta poems to exist. It also means that when you tell a linguist that "English is inconsistent", the linguist knows what you're talking about. Likewise, when a linguist mentions that "English is inconsistent", _we_ know what he's talking about.

>, you make it sound as if writing was invented in some earlier stage of evolution

No, that wasn't my intention. Just repeating the commonly held assumption that vocal grunts and pointing fingers at objects to communicate happens before systematic writing.

At the point where I had reached the end of your first sentence, I was about to respond with your second sentence. As Lederer's poem shows, the irregularity exists in the spoken language, and for the most part, the written language merely reflects that, at least in approximately phonetic languages.

People being amused by irregular verbs is, in my opinion, a sign of how much regularity there is in English. For an example of irregularity in Russian there is a short story by M. Zoschenko, "The fire iron" (Кочерга, М.Зощенко [1]). The plot of the story is about some office workers who are trying to write an order for five fire irons but don't know how to write "five fire irons" so they toss around ideas to order three and two in different orders or request "a fire iron, five items of" etc. It makes no sense in English, obviously, but in Russian the issue is that a noun, when used with quantity, has to be in different cases depending on the quantity and for quantity of 5 the case is genitive plural[2], which is rarely used naturally with "fire iron" so most people don't know how to make one.

1. http://fanread.ru/book/7894628/?page=1

2. https://en.wikipedia.org/wiki/Russian_declension#Numerals

How about the examples of Ket, from TFA?