Why are German numbers backwards? | Hacker News Reader

Why are German numbers backwards?

german.stackexchange.com

288 pointstosh419 comments

German here... I hate how we say numbers. Even after 36 years I still have problems with it. If I have to dictate phone numbers I'm saying each digit separately because everything else is just confusing and very often leads to swapped numbers on the other end. (sadly, most Germans say phone numbers as sets of two, and not as single digits) It just makes no sense and I very much prefer English, it is much more logical.

Some people have founded the association "Zwanzigeins" (look it up, they have a web site) where they try to push for another way of saying numbers in German and teaching them at school. But even they admit that the chances are very slim we change the way we say numbers.

Actually I rather like it. In my first language it works like that too. In practice when I count, I say the full word up to 20, and then start saying “one”, “two”, until I get to thirty to save time. This feels more natural given that the full word for 21, 22, etc is “one-twenty”, “two-twenty”, etc, rather than “twenty-one” etc.

You can do this in English or many other languages too if you want to.

>Actually I rather like it. In my first language it works like that too.

Yeah, it's easy for you since you grew up with that system, but as an expat in Germany it is a monumental pain when someone is dictating you long numbers (telephone, social security, insurance, etc.) in groups of two over the phone and you gotta scribble them quickly on a piece of paper since you tend to write the first digit you hear, but that's actually the last of the pair you gotta write so numbers get easily mixed up.

Example, dictating and writing down 23.45.67.89 in pairs over the phone, would sound like "3 ... and twenty", "5 ... and fourty", "7 ... and sixty", "9 ... and eighty" which is difficult to not fuck up and swap them when under pressure of writing quickly, if you don't count the same in your own language/culture, and you haven't agreed over the endinanness with the other party before the dictation starts.

So you're left with 2 choices if the other party uses this system, either you write the first digit you hear, which is actually the last, and leave a blank space in front, so you can write the "x_ties" number when it comes up, but that only works on paper but not on a dialing pad or keyboard as the cursor keeps moving too the right, or, the other option, you wait to hear each number pair before you start writing them down, then you start writing, but that can also causes mixups in your brain during the decoding of the reverse order from hearing to writing if the other party dictates the pairs quickly.

Or, you just throw in the towel and ask the other party to dictate it digit by digit and call it a day.

So, apologies, as I have to disagree with you. It may work well if you're counting incrementally to keep track of something, but for transferring non-sequential numbers over the phone, this is a stupid numbering system that causes more problems than it solves.

Sounds like you just have to get used to the endianness. It’s actually more consistent; in English, you say four-teen but also twenty-four. In German, they picked the way that is most logical for counting, and stuck with it throughout.

>It’s actually more consistent

In theory, yes, yet my adult brain cannot process correctly decoding this reversed order quickly, under pressure, over the phone in writing, even though I learned to be fluent in German. I guess you have to grow up with this system so it imprints on your subconscious from an early age, else, if you grow up with another system, and need to switch later in life, it's game over.

Learning this number system is easy, but under pressure over the phone, this reverse pair system falls apart quickly as you tend to write the first digit you hear instead of waiting for the full pair, which is why it's not used in military/critical radio transmissions, because it opens the gates to many errors and proves the system is broken for anything else than casual private use.

The system is not broken. It works for tens of millions of people, including over the phone. You’re just not wired for it. Welcome to living in your second language!

I will say that even an adult brain can adapt to foreign ways. It does get easier, though in the process you lose something of your original language.

In case you're not aware, your comments come off a bit condescending.

>It works for tens of millions of people, including over the phone.

Except it doesn't work well, as proved by the fact that this system is not used in the military since even top comment in this thread where a German agrees that even he gets confused by numbers in pairs over the telephone and as proven by the fact that Norway transitioned from the "German" way to the "English" way precisely to fix this issue.

I see you're very defensive about your culture/way of doing things, but just because some linguistical quirks exist to date in some languages, is in no way poof that they are good or that it works well, it's just proof that inertia is very strong as these issues get grandfathered in over time since transitioning to something better is too expensive for entire countries to make (look at why the imperial system is still used even though it's inferior to metric).

And for some countries/cultures, introducing certain linguistical quirks on purpose and keeping them was, and still is, a matter of national pride and differentiation between their culture and other very similar cultures (see French vs Belgian French vs Canadian French vs Swiss French, or German vs Austrian German vs Swiss German), so changing something for the better would be admitting something was wrong all along in their culture and would definitely face backlash from conservatives and purists, though Norway did the change successfully from the "German" way to the "English" way of speaking pairs of numbers in order to fix the confusion issues I mentioned.

I can say that struggle is not universal: I learned German as an adult, and don't struggle with writing down numbers I hear spoken. In fact, I'd never even thought about it being hard.

Every once in a while I say a number backwards (like once or twice a year), but I usually catch myself half-way after spitting it out.

We say ‘fourteen’ but not ‘four and ten’. Fourteen comes out as one word, like eleven. If ‘fiveforty’ were a word it would be easier to process as one word instead of five and forty which tends to be processed as two words.

Vierzehn and Vierundfünfzig are single words in German, but separate words in English. "Fourteen" (four and ten) being a single word is actually strange in English because the language normally splits words like these.

English has decided to use single words up to 20. Other European languages stop at 100. Both are arbitrary and right or wrong in their own way.

The English word would be "five and forty" because "fiveforty" would probably mean 200 going by traditional English (in the same say "four score and seven" means 4*20+7, not 24+7).

I’d say that the teens having a different structure in English is also weirdly out of place, not that the German counting makes sense. Having a structure where you always count from highest to lowest, and in the order you would write down the number, just makes more sense. It’s always been weird to me that we have unique words for the tens numbers in English.

>since you tend to write the first digit you hear, but that's actually the last of the pair you gotta write so numbers get easily mixed up

Somewhat reminds me of typical hexdump representation, where even if data has little-endian bytes, nibbles inside each byte are still ordered big-endian.

But it's the same in English up until the number 20. 16 for example, six-ten. The English just count differently after 20. But I could imagine "four and seventy" for example.

And it's not just German either. If you take the word 175 for example:

German: einhundertfünfundsiebzig (one hundred five and seventy)

Slovenian: sto petinsedemdeset (one hundred five and seventy)

Which is weird when you look at all the other neighbouring languages:

Polish: sto siedemdziesiąt pięć (one hundred seventy five)

Czech: sto sedmdesát pět (one hundred seventy five)

Slovak: sto sedemdesiat päť (one hundred seventy five)

Hungarian: száz hetven öt (one hundred seventy five)

Italian: centosettantacinque (one hundred seventy five)

Croatian: sto sedamdeset pet (one hundred seventy five)

Serbian: same as Croatian but in cirilic

You get the idea.

Given that, I'm holding you Germans responsible for our also stupid number system.

Sincerely, a Slovenian.

Dutch: honderdvijfenzeventig (hundred five and seventy)

Ah yep, seems like there is actually one more:

Danish: hundrede femoghalvfjerds (hundred five and seventy)

And the rest I've checked now:

Romanian: o sută șaptezeci și cinci (hundred seventy five)

French: cent soixante quinze (hundred sixty fifteen)

Swedish: hundra sjuttiofem (hundred seventy five)

Finnish: sata seitsemänkymmentäviisi (hundred seventy five)

Norweigan: hundre og syttifem (hundred seventy five)

Spanish: ciento setenta y cinco (hundred seventy five)

French should translate to "hundred sixty fifteen" which is another level of aberration altogether (I'm French)

Ah right I remember hearing somewhere that you guys don't have words for 70, 80, and 90 and do this odd sum of two thing. I suppose there are worse ways than the reverse German :D

The French language has such words, but Frenchmen don't use them. For example they prefer to say the old fashioned "quatre-vingt-dix" (4 - 20 - 10) instead of the perfectly fine "nonante" that French speakers in Belgium use.

It's the same in Switzerland, which makes an order of magnitude more sense IMO:

Soixante

Septante

Huitante

Nonante

Cent

Does the cognitive energy expended by French to do basic counting conditions their brain from early childhood for mathematical proficiency resulting in so many great mathematicians whose native language was French? </end_of_joke>

What I always wonder, do French programmers generalize this numbering scheme to pronounce 0x4B as quatre seize onze?

More precisely, French (cent soixante quinze) is actually: hundred sixty fifteen. Seventies, eighties (quatre-vingt = four twenties), and nineties (quatre-vingt-dix = four twenties and ten) are a mess in most French dialects.

the Danish is actually a little more complicated

the word for 60 in Danish is tres the word for 50 in Danish is halvtreds - so basically half 60 (I guess cause the original counting system in the Nordic region was based on 20s?), and since Danes don't pronounce the d and the halv is quick sometimes you get confused in what is being said.

But then the word for 80 is firs, fee-es with a partially swallowed r sound in there somewhere. and 70 is halvfjerds - half firs.

The word for 90 is halvfems - half fives.

a Dane speaking quickly can confuse others really quickly with these numbers as to whether it was said 50,60,70,80,90 and then you put the second number in 'backwards' as said, so

92 is to og halvfems - toe oh hellfems and so forth, but said very quickly with a tendency to not fully pronounce all of a word.

> the original counting system in the Nordic region was based on 20s?

No other Nordic language is like that.

It's probably not a coindicence that the same system the French use. Apparently French was the coolest language you could speak in the 1700s and all the nobility did it.

Only the Danish swalllowed the "twenty" part of the it, so it's no longer possible to deduce any meaning from hearing the word. Add that to the fact that "half" has a universally accepted meaning too, but should be understood here as "ten-less-than".

So I think Danish wins the most bizarre counting system over the French. And the French is far more so than the German. All they're guilty of is being careless with the ordering of numerals.

>> the original counting system in the Nordic region was based on 20s?

>No other Nordic language is like that.

ok, I was just guessing, hence the question mark.

But I guess Boris Jensen described the reason https://news.ycombinator.com/item?id=29369172

The system is actually based on scores, 20, which is called a snes in older Danish, so halvtreds is short for halv tredje snes, the half third score, and 60 is tres, short for tre snese, i.e. three scores and so on. So for the tens between 50 and 90, we count scores, and if it's not a whole number of scores, we name it the half of the score that we are into. It's also preserved in a very infrequently used variant word for 80, firsindstyve, which is just 4 score, more explicitly (tyve is the modern word for twenty). In conclusion: Yes, the Danish number system is relatively silly.

Danish is in fact slightly more complicated. They have a vigesimal system with a base of 20, with halvfjerds, or halffourth, meaning 3½ times 20. So rather hundred five and three-and-a-half score.

Norway has an alternative that is the same as the Germans. (175 - hundred and five and seventy)

It was more popular in the past, but is still used in many dialects.

I grew up with both the old one and the new one so I sometimes say it the old way and I am almost happy that my kids don't understand it immediately so I have to correct myself.

Fun fact: it was actually decided in Stortinget (the supreme legislature of Norway) in November 1950 and implemented in July 1951, as far as I know the only time a matter of how to pronounce something has been decided at that level.

Norway have had at least half a dozen language reforms in parliament that dictate the written language, and so indirectly pronunciation.

E.g. the 1907 reform removed a lot of soft consonants in favour of harder ones (e.g. "løb" -> "løp", "kage" -> "kake").

Most of them, incidentally, reducing the similarity with Danish...

Norwegian changed via a language reform a few decades ago. "Fem og sytti" used to be the norm (we inherited some of the Danish rules with the reverse numbers, but not the "halvfjerds" bit (which is effectively "half and four times 20")), and was still common well into the 80's-90's. I learned the new form at school, but picked up the old form from my parents.

If you want something completely insane check out standard French. 97 is "quatre-vingt-dix-sept" which translates directly to "four-twenty-ten-seven". Quebec French does this sanely though at least.

This is so sweet. J’adore le français.

Also, Belgians do that sanely too.

70: soixante-dix (FR) "sixty-ten". Septante (BE) is literally seventy (seven decades), which is much better.

And so on and so forth for 80 (octante in BE) and 90 (nonante).

Edit: fixed seventy

In Belgium (belgian french) we don't say "octante" for 80 (said in Switzerland maybe) but "quatre-vingts" (four twenty. 4-20). So 87 is said "quatre-vingt-sept" (four twenty seven). For 70 and 90 it is right.

Depends on the canton in Switzerland. To my knowledge most say “huitante” except for Geneva which says “quatre-vingt”.

Most people I know (from various Romand cantons) actually use both a bit randomly, I assume due to the strong French influence. What's funny is that they don't realise it until you point it out.

Octante as far as I know has been dead for a while and is not used anywhere, in Switzerland or elsewhere.

If your French is bad like mine you can get away with those even in France.

I hope my french is okay.

> 80 (octante in BE)

Nope. That’s switzerland, belgians completely illogically have kept the 20-based naming here.

Thanks the for correction.

So do Belgians do it 100% like the french? Or so they mix a bit of FR and CH?

The latter.

Sorry, but it’s also “quatre-vingt-dix-sept” here in Quebec. We do manage haha (it’s something you get used to / absorb as a native speaker, although it is of course a barrier for those learning the language)

You’re probably thinking of Belgium and Switzerland where 97 would be - as far as I understand - “nonante-sept”

See this other StackExchange on the topic (and geographical exceptions to the octante/huitante (!)/nonante usage): https://french.stackexchange.com/questions/187/quelles-parti...

Same in Catalan.

80 - vuitanta

Four score and seven-ten years ago :)

No, it translates directly to "ninety-seven" or to however you spell "97" in any other language.

> Literal translation, direct translation or word-for-word translation, is a translation of a text done by translating each word separately, without looking at how the words are used together in a phrase or sentence.

https://en.m.wikipedia.org/wiki/Literal_translation

I only knew term #1 and term #3, thanks.

Let's prove that English is insane by making a literal translation of that Wikipedia definition in French:

> Littéral traduction, direct traduction ou mot-pour-mot traduction, est le traduction de un texte fait par traduire chaque mot séparément, sans regarder à comment le mots sont utilisé ensemble dans un phrase ou phrase.

I'm not sure what you're trying to show there. Literal translations are always janky but can still be used to demonstrate things.

A similar literal translation from english to english of 97 would be nine-tens-seven. English has some irregularities in the teens but it's decimal all the way through. French switches from decimal to vigesimal.

> I'm not sure what you're trying to show there. Literal translations are always janky but can still be used to demonstrate things.

I'm trying to show that using a method that can only produce nonsense out of a given source is not the best way to demonstrate/prove _anything_ about the source. At best, it proves that literal translation is garbage, which is something everyone knows (except at tech giants, apparently).

There is only one way to translate "quatre-vingt-dix-sept" in English while preserving its meaning: "ninety-seven".

Pointing out the decimal <-> vigesimal switch (and the historical debates around it, for example) or the fact that the two systems cohabit would be far more useful and interesting than constructing a strawman in order to conclude that a language is "insane".

A literal translation is the easiest way to show what happens. I don't see the problem.

If you perform the same literal translation to 57, it doesn't look "insane". When you say the method "can only produce nonsense", I think you're just flat-out wrong.

If you’re a native speaker I imagine that’s true, but if you learned the language later in live it remains a pain. I feel comfortable discussing love, art and politics in French but I still dread writing down a phone number!

That’s about as odd as the danish ”syv og halvfems” which is “seven and half five” meaning “seven and 4.5 twenties”, so 97.

Yes, the "halvfems", "half five", could be translated as "half of the fifth twenty", or even "halfway of the fifth twenty (from the full fourth one)".

Luckily it is only numbers 50 to 99 that work that way. 31 is simply "enogtredive", as in "one and three tens". A hundred is "hundred", not "fems", as it could be (five twenties).

German swapped numbers are also possible and correct in Czech language "sto sedmdesát pět" is the same as "sto pětasedmdesát".

Yes, but it sounds archaic, if you use the swapped version.

Uh, it does not (unless you're 12 or so, I suppose).

Slightly older... older enough, that I remember Vlasta Burian's movies being aired in the tv, where it would fit.

I use it basically just for tram and bus numbers. It's more fun to say you're taking the dvaadvacitka instead of dvacetdvojka.

It is archaic but not in all contexts. i.e. If you talk about 125 ccm motrbike, it is always "german" way. Also it is used for human age or dates.

In Czech, both variants are possible. The German one is less frequent, though.

Fortunately nobody's going to use the reverse variant when dictating phone numbers.

To add to the list, Arabic also counts the same way.

مائة وخمسة وسبعون (One hundred and five and seventy)

I suspect that the German way of speaking comes directly from arabic and the fact that we have adopted the whole numbering scheme, digits as well the name for 'digit' (ziffer) sounds like 'zero' which is the key innovation of the number scheme.

You put so much effort into that and then totally missed French? four-twenty-ten-seven … yup, 97, of course. Multiplication and addition required.

That being said. I wonder how much of those kinds of complications lead to higher IQ and higher education test scoring. Simplification also dumbs down, stunts the mind.

Chinese Mandarin isn't quite as challenging, but they too essentially use multiplication 40 is "4 tens"

Chinese (and derivatives) basically just count like English, but without the inconsistencies.

4444 = Four thousand, four hundred, four ten, four.

Though it is interesting that they group by powers of 10k instead of powers of 1k.

> Chinese (and derivatives) basically just count like English, but without the inconsistencies.

Not really. "One thousand five" is 1,005 in English, but it's 1500 in Mandarin. For 1005 you'd need to say "one thousand zero five".

Could that be where “3V3” style notation came from? In electronics, 3V3 means 3.3V, not 3/3V or 3x3V or 3.003V, and likewise 1R5 means 1.5 Ohm. It’s handy but took me a while to get used to.

Don't know; Chinese usage seems unlikely to have been influential in the relevant time period.

There is another oddity in Chinese numbers which requires a bit of grammar explanation:

Chinese requires measure words when applying numbers to nouns. English has count nouns and mass nouns ("three crackers", where "cracker" is a count noun, versus "three loaves of bread", where "bread" is a mass noun); Chinese has only mass nouns. [1] Thus:

三个人 "three (三) people (人)", with 个 being a measure word appropriate for people

一只狗 "one (一) dog (狗)", with 只 being a measure word appropriate for animals

一首歌 "one song (歌)", with 首 being a measure word appropriate for poetry

Most nouns use 个.

The oddity is that 半 ("one half") occurs before the measure word when it represents the total amount, but after when it's a modification.

一个小时 "one hour (小时)"

两个小时 "two (两) hours"

半个小时 "half an hour"

一个半小时 "an hour and a half"

This also occurs with money, where it's probably the same grammatical rule:

三块 "¥3"

三块二 "¥3.20"

But for this to be fully consistent, I'd expect 零个半小时 "zero and a half hours" where in reality 半个小时 is used.

[1] Some people have argued that since e.g. "one day" 一天 has no measure word between 一 and 天, 天 must be a noun that requires no measure word. This is wrong; it is a measure word that requires no noun. An easy way to see this is that reduplication carries the same meaning that generally applies to reduplicated measure words, and not the meaning that applies to reduplication of nouns -- 天天 means "every day" in the same way that 个个 means "every [one]"; it does not mean "cute little day" in the same way that 狗狗 means "doggie".

Can you just say the measure words like you do in English then? Like "I'll have 3 loaves please". I guess it would be like 三个 or similar?

Also thanks for saying that they're equivalent to things like "Schools of fish" or "loaves of bread". They make way more sense to me now!

> Like "I'll have 3 loaves please". I guess it would be like 三个 or similar?

That's it exactly, and it's very common. Any time the noun is clear from context, you can leave it out. (You shouldn't leave out the measure word though - where in English you might have "I'll take three", in Chinese you'd still want 三个.)

If you walk into a restaurant, someone will ask 几位 "how many?". 几 is a question word for small numbers, and 位 is a (formal, polite) measure word for people.

Japanese is like that too (probably got it from Chinese), with the added fun that it has two sets of numerals: the indigenous Japanese one and the borrowed Chinese one. So you not only need to memorize the counting word but also which kind of numeral to use.

No, that's just pragmatism. The origins happen to be US/American:

It's the most compact, non-ambiguous representation, and avoids symbols that print poorly or are not available everywhere.

Ah, yes, shorthand, which the language loves. So the shorthand here is that if you don't add a unit qualifier then by default it's the next largest power of 10. So, "1,500,000" can be shortened from "1 million 5 hundred thousand" to "1 million 5", based on the theory that you're much more likely to use "1,500,000" in real life than "1,000,005". So then of course, to handle the latter use case, you now have to add the qualifier "1 million 0 ten 5", and then we shorten that of course to "1 million 0 5".

> "1,500,000" can be shortened from "1 million 5 hundred thousand" to "1 million 5"

Is this actually true? There is no "million" form in Chinese. Powers-of-1000 separators don't make any sense in the language either[1]. 150,0000 is, fully, 一百五十万 "one hundred five ten 10,000". And you obviously can't shorten that to 一百五 "one hundred five" because that's 150, not 150,0000. You could try saying 一百五万, but I don't think that makes much sense.

[1] The incompatibility between the Chinese number system and the English number system has resulted in Chinese speakers in America innovating the unit-of-salary "K". 工资几十K？ One K is, of course, USD 1000.

That's more of an analogous example. In Chinese you would shorten 一万五千 to 一万五.

Chinese is sweet and consistent until 10,000. They introduced a new word for it, wan(万) instead of 10 x 1000.

Why is that inconsistent? There are also separate words for 1, 10, 100, and 1,000, so why not 10,000 ?

If your number separators are every 3, then it feels weird. If it's every 4, then it does not. Change numbers into 1, 10, 1000, 1'0000 and then it doesn't feel like it's going against your writing habits.

Dunno about China, but Japan is a weird mix of two systems:

- Groups of 4 digits separated by the corresponding power-of-10000 kanji, e.g. 1億2345万6789. Makes sense, I would say it's actually better than just separating the digits, as it makes easier to recognize the size of the number.

- Groups of 3 digits separated by commas, e.g. 123,456,789. I asked around how they do to mentally regroup digits by four, and everybody told me that is quite difficult and don't understand why they don't group digits by four. Then again, they have the most difficult writing system in the world and a ridiculous calendar, maybe they just like challenges.

>That being said. I wonder how much of those kinds of complications lead to higher IQ and higher education test scoring. Simplification also dumbs down, stunts the mind.

AFAIK the net effect is that languages with complicated number representations do worse on math tests.

>Chinese Mandarin isn't quite as challenging, but they too essentially use multiplication 40 is "4 tens"

Can you say that the same about English? ie. four-ty = 4 × 10

edit: https://www.bbc.com/future/article/20191121-why-you-might-be...

>Can you say that the same about English?

It's similar but I don't think it qualifies as the same since we do not say "four tens". Forty is a concept in itself, just like suffix -s for plural is a separate concept from singular. Suffixes and prefixes are modifiers. We don't ordinarily say, for example, "many apple", we say "apples".

When I say eight-ten in Chinese, my mind is thinking of the singular concept '80', not 8 tens.

The same as when I say eight-y in English.

There was some theory that because the first ten digits in Chinese are very short phonetically that it is easier to keep numbers in your head.

The first numbers are short in every language. That doesn't distinguish Chinese in any way.

Taking some salient examples, in English 9 out of 10 of those numbers are single syllables and 7 is two. In French, all ten are single syllables.

It does distinguish Chinese. It's quicker to count to 10 in Chinese than in most other languages.

Malcolm Gladwell did some good research ('Outliers' is a great book) in this area.

Chinese are generally better at math than other ethnicities precisely because of their language.

Take a look at the following list of numbers: 4,8,5,3,9,7,6. Read them out loud to yourself. Now look away, and spend twenty seconds memorizing that sequence before saying them out loud again.

Gladwell points out that the English speakers have about 50 percent chance of remembering that sequence perfectly, but the Chinese are almost certain to get it right every time. He explains, "Because as human beings we store digits in a memory loop that runs for about two seconds. We most easily memorize whatever we can say or read within that two second span. "And Chinese speakers get that list of numbers—4,8,5,3,9,7,6—right every time because—unlike English speakers—their language allows them to fit all those seven numbers into two seconds," Gladwell adds.

https://gineersnow.com/students/best-explanation-asians-good...

Well if we're just speculating here, I'll add that since Chinese is tonal, Chinese speakers will remember the tune of sequence, not just a list of values. It's easier to remember a melody than a phone number.

Even non-tonal languages have melodies, for example English people use a lot of intonation compared for to my native Polish. When I was in UK everybody sounded like they were talking to infants or dogs :)

You don't normally use it for numbers but you certainly can, see: https://www.youtube.com/watch?v=ab8GtuPdrUQ

Another useful mnemonic that for me works even better than melody is rhythm. I noticed that I have about 20-notes buffer for last-heard rhythmic phrase even if I wasn't paying attention at the time. So for example after I ran down a flight of stairs I can count them by remembering the rhythm of my steps and adding them. My friend who has way better short term memory than me can't do this, but he can see the image he was looking at recently. Now that's cheating :)

> Even non-tonal languages have melodies, for example English people use a lot of intonation compared for to my native Polish. When I was in UK everybody sounded like they were talking to infants or dogs

It might just be more obvious since the English patterns are unfamiliar.

One of the more surreal experiences I've had was watching an English-language news broadcast in China. The presenter was speaking English and had obviously put in a lot of effort trying to learn what natural English sounded like. The general pattern of intonation over her sentences was quite realistic for English.

What made it surreal was that the intonation didn't match the words. Everything she said, it was like she was using the intonation pattern of some other sentence and applying it to a completely different sentence.

Polish has a lot of intonation—which is also kind of unique. I can instantly recognize people speaking Polish, even in larger groups of people talking all kind of languages, just by the very typical sentence melody that sticks out.

But it's indeed quite different to other languages, even the other Slavic ones.

Germanic intonation and sentence melody (for example like in German or English) is completely distinct from the Polish one. And this melody seems to be something sticky as you can always recognize Polish people just by their intonation even when they speak otherwise perfect German or English. That's not the case for for example Russian, or Czech, or Slovak people.

That seems to me like a lack of imagination on his part even assuming he has some grounds for the "2 second" rule.

How does he know that people remember it via "reading out loud to themselves"?

Maybe they visualize it instead.

Maybe people chunk it into a 3 digit and a 4 digit number, like a phone number.

Why should "reading out loud to yourself" be limited to the speed of actual speech anyway?

I’ll be damned, I just tried that and it was exceptionally easier to do in Mandarin, a language that I have to think to count in, than in English.

> Malcolm Gladwell did some good research

This is not something you hear often.

In Arabic, numbers from 1-10 are waaHid, ithnayn, thalaatha, arba:a, khamsa, sitta, sab:a, thamaaniya, tis:a, and :ashara. No monosyllabic numbers, and 8 has four syllables. And even these are short compared to the numbers in Inuktitut.

This is kind of a tangent, but I understand that the native title of the Arabian Nights is 'alf layla wa layla, the book of "a thousand nights and a night".

What is the "one" night in that title? Any chance wa is related to waaHid?

No. “wa” in Arabic corresponds to “and” in English.

>any chance wa is related to waahid?

No, wa (usually in basic sentences) means 'and'. Layla alone means one night. Alf is a thousand, and for certain numbers the singular is used over the plural, which is why it may seem confusing.

Interestingly, in Algerian Arabic, while other numbers are similar, two is different. It's zouj (one syllable). Except when counting e.g. twenty two, where it is similar to ithnayn (more like t'nin)

BTW, it's similar to German in that regard, because it's two-twenty.

Also interestingly, the way 8 sounds in Algerian Arabic would be 2 syllables. Although take it with a grain of salt because it's third-hand information. I learned this from my father, who's not native (but has lived in Algeria in his childhood)

Majority of French speakers say 80 as "4 × 20":

https://en.wiktionary.org/wiki/quatre-vingts#French

Not "4×20", but "4 20", and in a single word that means "80" without thinking about 4 and 20.

When you say eighteen, you think just "18" and not "8 teens". (Similarly, when you say "backwards" you think of the direction, not of "back wards")

As a French speaker, you don’t tend to see those as multiplication

You associate “quatre vingt” as meaning 80. In your head it’s 80. You don’t think four times twenty. So it’s not as complicated as it looks. I don’t see kids really getting that wrong.

As someone who only had basic high-school "French as a third language" and was never good at it, I'd still agree with that. it's one "symbol" so to speak for mental parsing.

Yup, a friend of mine learning French a few years ago asked me how I, as a native French speaker, deal with this problem. I didn't understand what he was talking about because I had never in my life even noticed it. Learned how to count before I learned how to multiply, after all.

It's even funnier when you're learning multiplications and divisions. We still have to think when we do 4 times 20, and every time we realize it's right there in the name.

I get that. It's the power of the brain's user of generalizations, i.e., patterns or classes, to represent things. The brain clearly also handles disambiguation far better than we consciously know to do. It seems like the brain essentially has a class named quatre vingt and it has a pattern of 4*20 that resolved to the concept of 80 which means 20+20+20+20.

It clearly comes from a lack of having a separate term for 80 or even 90 for that matter the way that German and English do; which I find peculiar too, considering that French a Romance language (not the heart romance), while the people are largely Germanic in origin, i.e., the Franks. It makes sense when you consider how the roman numeral system functions and that the Franks were in far closer proximity to Rome than the Germans, including the ones that moved to the British isles and became the English, i.e., Anglos and the Saxons, Germans. It seems that those interplays and intersections with the cultures are what determined how French language numbering worked based on when and where and what they had contact with.

80 maybe, but 91-99 are properly ridiculous. Sure its easy to get it, but it highlights deeper issue I've had since I've started learning french - its not elegant nor easy language, rather a 'spaghetti code' one, a mess of rules and tons of exceptions, and many things defy logic and are there 'because its like that and you have to memorize it'. You can have great talk on B1 level for example in English or German, with French you are still often lost quickly unless everybody else tries hard to dumb it down for you.

There is an institute in France hose sole purpose is to guard language, I wonder why they didn't find the motivation to clean it up a bit. It would make it much more attractive for outsiders and make it more global.

And its not just me, literally everybody I speak to who attempted to learn french has similar experience. Either they suck it up, face often humiliation from native speakers from their mistakes (its quite something to see senior banking colleagues laughing like little kids and pointing finger at you on a project meeting because you mixed gender of a noun) or often just give up.

This was cleaned up in the Belgian variant of French: https://en.wikipedia.org/wiki/Belgian_French.

- 70 = septante (versus soixante-dix, which is 60-10)

- 80 = huitante (versus quatre-vingt, which is 40-20), *EDIT*: wrong, see below.

- 90 = nonante (versus quatre-vingt-dix, which is 40-20-10)

The article also mentions something interesting I didn't know:

> The use of septante for "seventy" and nonante for "ninety", in contrast to Standard French soixante-dix (literally "sixty-ten") and quatre-vingt-dix'("four-twenty-ten"). Those former words occur also in Swiss French. Unlike the Swiss, however, Belgians never use huitante for quatre-vingts ("four twenties"), with the use of octante in the local Brussels dialect as being the only exception. Although they are considered Belgian and Swiss words, septante and nonante were common in France until around the 16th century, when the newer forms began to dominate.[4]

*EDIT*: This doesn't appear to be true, Belgian French speakers also say quatre-vingt for 80.

It’s the Swiss who fixed it and say huitante. The Belgians say quatre-vingt like the French.

I don't think "cleaning up" a language talked by so many people in the world is reasonable because some people have trouble learning it.

> Either they suck it up, face often humiliation from native speakers from their mistakes (its quite something to see senior banking colleagues laughing like little kids and pointing finger at you on a project meeting because you mixed gender of a noun).

I'm not sure if it's a language problem or a people problem. I often encounter people that mix the gender of nouns, and I don't really care about it. It's a lot to learn and not very important. Just like some people don't have a great accent, that's how it is, it doesn't stop people from communicating. Same for the people that I know, unless we're asked we wouldn't bother correcting someone that "le table" is actually "la table" because tables are female.

On my side, I find the pronunciation of English to be very hard to learn and to master, and am scared of sounding stupid whenever I talk English, so I avoid it, and end up not being good at it, so I can understand the sentiment.

If it’s any consolation, I find English pronunciation very difficult, and I’m a native speaker. It surprises me that in my thirties I still regularly encounter situations where I want to use a word and realise I’ve never heard it spoken before, so have no idea if the pronunciation I use in my head when reading it is correct.

I also often hear others mispronounce words; friends, colleagues, even on TV.

I guess my point is that if you’re mispronouncing English words you’re speaking it like a native!

I’m French and, at least in my circle, I’ve never seen a native French « humiliating » a non-native trying to speak French.

And to me there is two reasons : - we know our language is difficult to learn - we are really bad when it comes to speak any foreign language

As*oles are totally a thing (especially in the banking/financial sector) but most French people are admirative of anyone who speaks more than one language. Because most of us can’t.

Got off a train from Brussels at Gare du Nord and asked at the kiosk for “une cafe sil-vous-plait”, guy behind the counter responded in English

I don’t bother now. Germany, Italy, Greece, sure, but not France.

My guess is that you didn't say "bonjour" first. Major faux pas.

I'm an American living in Japan and the this happens quite a lot - even among friends who I'm pretty sure mean no harm.

It seems like a lot of people have this very strong "foreigner = English" mapping in their head. As soon as you're outed as "foreigner" for any reason, you get the English (even if you're proficient in the local language).

I definitely think it's disrespectful to switch languages on someone like that, but a lot of people just don't seem to think about it. Multi-lingual conversation manners are not commonly talked about!

Most likely he responded in English because he felt it would be easier for both of you, not because he wanted to humiliate you.

On the receiving side it comes off as "just stop trying and buy something"

Which is really quite reasonable if you are at the front of a queue of people waiting to buy something, or if the cashier has something else to do like make coffee for the people who already ordered. The intent is very unlikely to be to humiliate the foreigner.

Maybe it's a cultural difference but I have a hard time thinking that this is impolite. I'd have do the same and it wouldn't mean that your french is not perfect but only that I would like to ease your life.

Brussels is increasingly English-speaking in general, and folks who work in the train areas in particular have to be fully fluent in French, Dutch and English, where the latter is just the default communication choice if you detect someone is not a native Dutch/French speaker. As well the area around the Noordstation is particularly unwise for a tourist to visit for cafés, so it is extra important for them to get across where you ought to go.

An amusing version of this happened to me in Porto recently at a bookshop. Many moons ago I was fluent in Brazilian Portuguese, but today that's retained mostly in reading, I can't speak well nor can I easily follow European Portuguese speakers. I was browsing the Portuguese literature section when the shop keeper asked me question (which I didn't catch), and then upon ascertaining I spoke English she rather forcefully explained where the English section was, but very confusedly watched as I politely kept where I was. To make matters probably more confusing I ended up buying a linguistically avant-garde book that I soon learned really is best suited for native speakers (Guimarães Rosa, the author, is basically a Brazilian James Joyce)

The Swiss French speakers found a way around that: nonante. Ninety, the Swiss way, not bothered by the French Academy.

You might be having difficulties speaking/understanding the language with actual French people not because it's difficult/random (and it is, I believe that most teenagers can't write a full page of faultless French !) but because spoken and written French are quite different : contractions ( j'sais pas instead of je ne sais pas), using 'on' instead of 'nous', the simple past is never used, slang, etc. Since foreigners tend to learn written French, things get difficult quite quickly when in the field. Obviously, you're not allowed to write spoken French.

Interestingly enough, if you listen to people talking in the 50s (old films, radio, etc) , they speak something which is a lot closer to written French. I'm under the impression that there's a growing gap between spoken and written French, and the French académie is not helping by essentially preventing the written language from evolving.

As to cleaning it up, well, lots of people agree with you, but always end up facing extremely strong opponents who accuse them of destroying the very nature of French culture!

The English Eighty comes from proto-Germanic and means eight groups of ten.

Would you call 'four hundred' multiplication? That's a strange way of looking at it for me.

Long ago “forty” in English may have begun as “four ten” which most likely became “fourteen” but four tens could have maybe become forty.

"Fourteen" (four and ten) would have been from "scoring numbers", where you get to twenty (a score), keep track of the scores separately, and start over. Up to twelve, we used a duodecimal/dozenal system (a separate word for each number). That was also common in other non-Germanic Indo-European languages, notably the Brythonic Celtic languages (and various versions of Brythonic scoring numbers are still used in parts of Britain, depending on the pre-English dialect spoken in the area and changes over time, especially in children's games). French numbering still shows signs of "scoring", especially in the 60/70 and even moreso in the 80/90 region.

As a native Mandarin speaker, I don't tend to think of the "ten" as a ten. If anything, for me it conjures up an image of the number of 0s. So, the "hundreds" part in "four hundreds" would just means four followed by two zeros, etc. I may even have been taught this as a child; can't remember. Anyway, no arithmetic involved, at least not explicitly.

I don't know that this generalizes -- other Mandarin speakers may have a different experience. I'm really curious how Chinese-speaking people thought about these things before Arabic numerals, but not sure we can ever have a clear answer to that question.

> That being said. I wonder how much of those kinds of complications lead to higher IQ and higher education test scoring. Simplification also dumbs down, stunts the mind.

I would say more that simplification makes the boring bits easy and allows the mind to concentrate on more interesting higher-level concepts. As someone who has attended a range of educational institutions for the same courses, one thing that really stood out to me about the "top-level" ones was that there was none of this "life must be hard" attitude. For the core material, the teaching was excellent and designed to make it as easy to learn for students as possible to learn. Then while "lesser" universities were examining students on those core materials (often with questions they'd seen before), the top universities asking novel questions on material that hadn't even been explicitly covered, but which the students could reasonably be expected to answer on because they had a really solid grasp of the core stuff.

"life must be hard"

Ah yes, this is very german.

If it is not a grind, it is not really work you are doing.

And it is still kind of a honor badge to moan about how little you sleep, as this shows how hard you are working all day and the ones sleeping the most less, are the hardest.

(But I do see some healthy change in that regard)

> the ones sleeping the most less

Good god, this is beautiful. Is this a German idiom?

Not literally, as far as I know (I am also still wondering, of whether it was correct grammar), but there are plenty of:

"Morgenstund hat gold im Mund"

morning time is gold

(to which I agree at times)

"Der frühe Vogel fängt den Wurm"

The early bird catches the worm.

(to where I say, maybe the worm should have slept in that day)

I think the poster you are responding to is referring to the “beauty” (sarcastically, I assume) of the phrasing “most less” instead of “least.” It would be right in line with the confusing way numbers are spoken.

Yeah, I feared as much.

I guess it is just an english grammar issue Germans often make (less, lesser, least and so on for single syllable words and x, more x, most x for the rest). The german equivalent of least is also just one word: wenigsten.

I don't do sarcasm (and it's against the HN guidelines, btw).

"Sleeping the most less" is a beautiful turn of phrase, even if ungrammatical. Or maybe because of it.

French is definitely weird in that it introduces multiplication, but the addition operands are still in descending order as with English.

> I wonder how much of those kinds of complications lead to higher IQ and higher education test scoring

Probably not much. I doubt francophones are doing multiplication when they think of the number 80 any more than anglophones do addition when we think of the number 14. Rather, speakers presumably both memorize the names of each number and move on with life.

A trivial aside: English used to have something similar in a "score" being 20. So "four-score and ten" would be 90, not too dissimilar from the French.

English still has it. Nobody took it away. It's still in use.

How? What is a common modern day use?

I think it's still used a bit in British English, along with other things that strike the American ear as archaic like 'stones' and 'fortnights'.

The same meaning of a group of twenty. "There must have been a score of cars at the drive-thru." It hasn't changed.

It's barely ever used nowadays though, whilst 30 years ago you'd hear it occasionally, similar to a gross, or things being referred to in yards. It's changed IMO.

I think it's part of how we've tended to remove dialect and en-gb terms as we've more intra-UK mixing and more non-UK born residents?

You might not hear it often, but that doesn't mean it's barely used. English is spoken by more than just English people. I reckon I hear "score" used in the sense of twenty of something than I hear words you'd probably describe as more common, such as "laden" or "bereft."

Might I enquire the context and locality where you hear 'score' used?

You are right, I mistakenly assumed we were talking about en-gb.

I live in the southeastern United States. It's common for conveying the size of a group, for instance. If a lot of people showed up to an event but not on the scale of hundreds, it's common to hear "scores of people."

It's not common in my job to talk about scores as a metric because 20 isn't a natural unit for my work product, but I wouldn't be surprised to hear it like that, either. We still use "dozen" and "gross" for other non-decimal groupings much more commonly, e.g. "a dozen doughnuts."

As someone who has lived in the west and northeast, I’ve never heard the word score used to mean 20 outside of Lincoln’s address. In most dialects in the US, scores has just been relegated to another synonym for many and doesn’t get used outside of it’s plural form. Interesting to hear that there are places that still use it occasionally, much like I occasionally hear “gross” used as you mention.

Obviously Americans know it from the Gettysburg Address where Lincoln referred to the US being founded "four score and seven years ago", but he was being intentionally poetic, but you may not count 19th century as "modern" (even though from the linguistic perspective it is), but people often say things like "there are scores of movies where the protagonist finds out he is a prophesized hero" even today.

That usage (“there are scores of x”) is figurative and doesn’t literally mean “20 * n”.

My dictionary says it's "archaic".

You are, of course, technically correct but I can’t say I’ve ever heard it used outside of old books and plays, or fiction set in the past.

No, it is a word like ananas or worm, an image in your head.

Nobody in France would do any multiplication, it is just a word.

We have a weird language ('eaux' is 'o', imagine that?) but this is not one of the crazy things.

It's easy to pick on the weirdness of french numbers, but honestly "quatre-vingt" ends up just being a word like "eighty" in its own right. No French speaker is multiplying 20s in their head.

Probably the only true weirdness is the 70s and 90s because they use the teen words like douze and treize, but that's honestly where the weirdness ends, and larger numbers follow very consistent rules.

> four-twenty-ten-seven … yup, 97, of course

As others said, French speakers parse "four-twenty-ten" as ninety. Nobody thinks about this in term of 4 * 20 + 10. Although, I remember that it confused me a little bit when I learned how to count.

That being said, I'd be in favour to switch to the Swiss/Belgian/Canadian way and replace "quatre vingt dix" by "nonante".

Sad fact: French Canadian don't use the (much better) Swiss/Belgian septante/octante/nonante and the awareness in Quebec just isn't high enough to hope for a switch.

It's a shame because soixante-dix, quatre-vingt and quatre-vingt-dix are confusing to write (I probably made a mistake somewhere).

Note that Belgians ans Swiss don't agree on 80. It's huitante on one side and octante on the other (but I don't remember which is which)

Interesting, I knew Swiss said octante and my Belgian friends all say octante as well. Perhaps it's a regional thing?

Swiss person here. This "octante" idea needs to die. In my 30 years in French speaking Switzerland, I've never heard or read "octante" ONCE. It might be used in Belgium, but I'm not even sure about that.

Not to be that guy that argues with a Swiss about how Swiss talk but we get broadcasts from "RTS Un" (I think that's what it's called) and the announcers most definitely say "octante". I also followed a few Swiss YouTubers a few years ago and they all said "octante" as well. Maybe it's a regional thing?

EDIT: Actually now that I dug up a bit more stuff it does seem like "huitante" is the one? Maybe I am just remembering wrong, feel free to ignore my ramblings.

I will indeed feel free to ignore your ramblings, because you're thinking of "huitante" :-P

> French Canadian don't use the (much better) Swiss/Belgian septante/octante/nonante

Sorry for the mistake (I did check on wikipedia before adding Canada do the list but got it wrong).

> the awareness in Quebec just isn't high enough to hope for a switch.

Same in France. I believe most of us regard "septante/octante/nonante" as amusing and exotic sounding. Sadly, I've never heard anyone advocating for a switch.

I'm a french canadian and I've never in my life heard someone use "septante/octante/nonante". I understand the difficulty while learning French but, as a native speaker, we don't even think about it.

> As others said, French speakers parse "four-twenty-ten" as ninety.

But only 90 and 97-99 have "quatre vignt dix". 91-96 are written as "four-twenty-eleven" through "four-twenty-sixteen".

So you can't really avoid the complexity, because only half of the numbers in the 90s have '''90''' as part of the word.

>>Simplification also dumbs down, stunts the mind.

Citation needed?

Simplification is itself an act of intelligence. Removing complexity is difficult. Einstein, Feynman, Newton, and innumerable others are lauded for simplifying enormously complex ideas to the point of comprehension by the masses.

Oversimplifying is bad, because it implies lost information. Simplification itself is a form of sophisticated articulation.

More efficient representations of numbers are generally associated with better performance in math, historically. Roman numerals being a prime example of unnecessary complexity, compared to the maths being done by Indian people, and so on. Civilization tends to abandon conventions that are superseded in advantage.

Making things more difficult to formulate for structural reasons unrelated to the problem at hand is inefficient.

Having inefficient numbering in language is wasted energy at best.

> Simplification is itself an act of intelligence.

For the people who simplified things (Einstein, Feynman, Newton). For anyone who never had to learn or have anything to do with the more complicated system before it was simplified, it most certainly isn't an act of intelligence.

Indeed!

"I wanted to write you a shorter letter, but I didn't have the time"

> Simplification is itself an act of intelligence

Obfuscation is an act of counterintelligence.

>Chinese Mandarin isn't quite as challenging, but they too essentially use multiplication 40 is "4 tens"

This is exactly what forty means, except it goes back to the Proto-Germanic fedwōr tigiwiz and so the meaning is obfuscated. Much simpler to use the same word, no?

Interestingly, the dialect of French spoken in Belgium has a name for 90 - nonante.

They still use quatre vignt for 80 though.

> 40 is "4 tens"

so is "forty", optimized for quick pronunciation, a respelling of four tens.

I've never understood how Russian 40 is sorok, nor have i ever gotten an explanation, including from Russians, whereas the rest works like english, etc.

To be fair, it's a matter of language development and evolution. "Four score and seven years ago" is essentially the same thing, just a couple of hundred years ago, the French just kept it in the language.

Even more loyal than the French to the ancient vigesimal counting system are the Basque[0] and the Welsh[1].

Traditional Welsh has constructions as:

- 16: un ar bymtheg ("one on five-ten") - 18: deunaw ("two nine") - 41: deugain ac un ("two twenty and one") - 71: un ar ddeg a thrigain ("one on ten on three twenty")

[0]: https://en.wikipedia.org/wiki/Basque_language [1]: https://en.wikipedia.org/wiki/Welsh_numerals

French: cent soixante quinze (hundred sixty fifteen)

My understanding is that was the way in Serbian-Croatian but it died out. I personally knew people born in early 1900s talking like that. But I can be wrong: it could be just Autro-Hungarian influence.

Any real data on this?

You are also forgetting that you are comparing two totally different language trees (Germanic and Slavic) … ignoring Italian for the moment.

You essentially listed German and several dialects of the same language. If you had listed several of the German language dialects that also slightly vary how they say the number in the same German format/order you would have had a list of equal if not greater number of support for the German format.

I think that may also provide a bit of a clue as to why the order/format is different since it must have happened some time after English formed from the German language, possibly when/because the British adopted the format/order of the Romans. But that's just speculation/hypothesis on my part. I suspect there are people who have a better insight into how that separation happened.

taking the opportunity to say that the most voted answer in stackexchange is wrong for Greek, in Greek for example 175 is εκατόν εβδομήντα πέντε (one hundred seventy five)

> And it's not just German either. If you take the word 175 for example: > Croatian: sto sedamdeset pet (one hundred seventy five) > Serbian: same as Croatian but in cirilic

As a German speaker with Ex-Yugoslavian roots, I'd like to point out that you have a mistake in your list.

In Serbo-Croatian (former official language of Yugoslavia) 175 (sto sedamdeset pet) is actually the order in which the number is written. Only between 10 and 20 the pronounciation is somewhat the other way around. But like the French colleague in this thread it could easily be argued that the numbers between 1 and 20 have their own words because it's not tri-deset but trinest for 13. This could be because e.g. tri-deset is actually used for 30, which sounds like three times ten. It seems Slovenian counting is more similar to German, while Serbo-Croatian is more similar to English.

Serbo-Croatian counting examples: 1 - jedan

2 - dva

3 - tri

4 - četiri

5 - pet

6 - šest

7 - sedam

8 - osam

9 - devet

10 - deset

11 - jedanest

12 - dvanest

13 - trinest

14 - četrnest

15 - petnest

16 - šesnest

17 - sedamnest

18 - osamnest

19 - devetnest

20 - dvadeset

21 - dvadeset jedan

32 - trideset dva

43 - četrdeset tri

54 - pedeset četiri

65 - šesdeset pet

76 - sedamdeset šest

87 - osamdeset sedam

98 - devedeset osam

100 - sto

101 - sto jedan

111 - sto jedanest

121 - sto dvadeset jedan

212 - dvesto dvanest

222 - dvesto dvadeset dva

...

Very close to Polish.

My wife ist Polish. This might explain it.

It would be rather jedanaest, dvanaest, trinaest (at least in Croatian) - your version sounds how it is shortened in pronounciation in some regions. Also šezdeset with "z".

I have been living in Austria all my life. I might not have misspelled some things :)

Hi, yes, I assumed it would be the case. And it is not always easy to explain word terminations/cases. Especialy if you do not use it regularly or is limited to family interactions. I would guess your pronouncian might come from either some parts of Lika/Dalmatia or Bosnian part closer to Croatia. Is it the case?

Might be slavic influence, is the same in Romanian.

I’ll just add Danish: et hundrede og femoghalvfjerds (one hundred five and four scores where the fourth score is a half)

Sincerely, a Dane :)

"Hmm, having the most difficult to pronounce/hear phonetics in the world wasn't hard enough, we should also mess with the numbers" - the Danes, I presume

Stupid is debatable here. Computer processors also sometimes tend to use little endian numbers instead of big endian numbers. Germans and us Slovenians just seem to prefer attention to detail and put the most significant digit of a two digit number on the second place.

ZRC-SAZU might have some etymologycal answers.

On that note I notice that I usually misspell two digit numbers in Slovene. For example when writing a number, I usually write the right digit before the left when writing from dictation. Sometimes when I am thinking about a number I tend to say it the other way around, petindevetdeset instead of devetinpetdeset, even though I am a native speaker.

Then you also have Slovenian dialects, where the number order is different again i.e. Prekmurscina - stou sendeset pet.

The Dutch get numbers “backwards,” too. My poor daughter makes mistakes with writing numerals all the time. Like, writing “27” for tweeënzeventig. Sigh. She will learn eventually. I’m sure the mental challenge just makes people strong here, like the bicycling in the freezing rain.

In Dutch it is “honderd vijf en zeventig” (one hundred five and seventy). So the same as in German. Do we actually know the origin or reason?

One hundred Five and Seventy is middle endian, neither big nor little.

The only other example I can think of is the american date system

Same for Dutch.

Honderd-vijf-en-zeventig. Hundred five and seventy.

Do they all say the "one" explicitly? In Dutch, it's "hunderdvijenzeventig", (hundred five and seventy), without the "one". That "one" is slightly more likely to be used with thousands, and a lot with millions.

> I'm holding you Germans responsible for our also stupid number system.

do so and especially blame https://en.wikipedia.org/wiki/Frederick_II,_Holy_Roman_Emper... to whom Fibonacci dedicated his liber abaci on those fancy new numbers back then.

They must have messed it up introducing that to german day-to-day use. They did it in parts arab-ordered (spoken), half reading order, where sensible would have been right-to-left one digit after the other. So you wouldn't even have to count upfront.

I'm surprised Hungarians don't say 7 times 5 squared,

Do you have some kind of cognitive impairment like dyslexia?

We have the same way of saying numbers and I can't imagine anyone being confused by it. Its normal to just say individual digits for large numbers in any language.

> Do you have some kind of cognitive impairment like dyslexia?

Nothing I know about, no. It's not that I don't understand it or that my head explodes. I cope with it. I grew up with this way of saying numbers, and yeah - it's the way it is, it is normal. But I think it requires a tiny little bit more brain activity than it needs to be. For me (as a software developer) I tend to prefer easier and more logical systems. And the way of saying numbers is one thing that is just more logical the way it is in English or other languages.

Similarly, I don't like the way we write dates (28.11.2021) and much more prefer ISO8601 (2021-11-28). But I think this is a format people more agree on globally sooner or later with all its advantages.

Ok, so in a CS sense, why would prepending be better than appending for numbers? If there is a difference (imo there isn't because of the way we chunk thinking), but for counting appending is probably better as the significant part is first and non significant last? Same with dates, isn't it better to see the more significant info upfront? You are more likely to be confused about which day it is than which month, and about which month than which year it is.

ISO-date-format is only better for sorting. But for writting, the german format is far better, because it's written in order of priority and optionality. This is of course less relevant with computers today, but even for reading it still applys.

I do wonder why you care about optionality and priority and what those even mean and how it's better in any way. I don't think it's better at all. It's different.

You remind me of a website I found way back for "learning French as a German". The site was actually pretty decent. But then it started teaching you the numbers and the clock and it started off with how the French way of saying numbers and the time is so much more logical and better than the German way. I closed the site immediately and never opened it again and I did not continue learning French at that time. Stopped right then and there.

Priority and optionality do not help with parsing written dates in an internationalized context. And that is true before computers as well.

2021-02-03 is easy to parse as the 3rd of February 2021 because there's no country on earth that uses this date format to mean the 2nd of March 2021, otherwise it wouldn't help at all.

I'd say that they both depend on context. Let's imagine the two of us are talking about "going camping this month". Year and month are optional. If we're talking about "going camping later this year" the year and day are optional "let's go in February". Let's say we're trying to figure out whether to "still go camping this year or next year". Now day and month are optional.

Your last paragraph is a strong argument for the American system of month/day/year. Days lose most of their relevance unless they are in the current month, so month-first is much more logical and better, because it gives the mind the necessary accuracy without the useless precision. And furthermore…

Just kidding. Month-first is as crazy as camping in February. I’m only used to it because I’m American. Getting us to switch to day/month/year seems more confusing than switching to year-month-day, because the latter is different enough to remove all ambiguity when reading. 06/08/2021 could be June 8 or August 6, but 2021-08-06 is clear since (to my knowledge) no one has ever used “year/day/month.”

As you say, this all really applies to full written dates only, since conversation relies much more on context anyway. You are forgiven if you stopped reading this comment before now :)

When I was a kid we had German as mandatory language to learn. I remember that when learning numbers we thought that the teacher is making it up and is incompetent. It took a lot of explaining that it is actually for real. Anyway, due to these things I never got to learn this language, my brain just refused to memorise these rules :/

Every freaking time a German dictates a number they do it in a sane way for half the number then do the backwards way for the rest which totally trips me up. I hate it.

Similar to the Zwanzigeins movements, in Malagasy, we have people who'd wish to reverse the counting pronunciation, although in the public sphere it is virtually unheard of. I remember debating on forums on how practical that would be. But IMO people are so lazy they just resort to counting in French instead. Madagascar has so much other worries as of current that it's totally understandable in a way.

> If I have to dictate phone numbers

The problem also exists in English: https://www.youtube.com/watch?v=fVPZAXMCasI&t=154s

In Dutch it is the same way. We even have another word for billion. Billion in dutch means 1000x more than the English version. Compounding is translated as combined interest.

English has million and milliard [0], but American English preferred the short scale and that has had more influence over the language. The UK only officially switched over to the "American" system in 1974.

Many European languages have the long scale, English is the odd one out here, as is Brazilian Portugese if you'd still classify that as a European language.

[0]: https://en.m.wikipedia.org/wiki/Long_and_short_scale

As a Dutch speaker, I think of having the numbers "backwards" as a neat feature. You can give someone a phone number (12345) and then verify it by saying it differently (twelve, thirty four, five). If you verify by repeating the same numbers then there's a decent chance of introducing the exact same error the second time.

English itself isn't all that simple either, because they still follow strange rules before reaching 20 like many other West European languages. French even stuck to its base in 20, unlike English (though "four score" is still often used to say 80 in the famous quote). The word "million", from "mille" meaning 1000, is used to express a thousand thousands. The American system also switched to the short system (million, billion, trillion instead of million, milliard, billion) and UK English has made the same switch relatively recently but only because of American influences.

I don't think there's any natural or logical way of saying numbers per se. If there was, we wouldn't have been doing it "in reverse" for hundreds of years in Europe.

I can't feel strongly enough about it to be for any change but forcingeeveryone to change their habits is annoying and probably costly. You can't force a change in language, language changes by itself.

>As a Dutch speaker, I think of having the numbers "backwards" as a neat feature. You can give someone a phone number (12345) and then verify it by saying it differently (twelve, thirty four, five). If you verify by repeating the same numbers then there's a decent chance of introducing the exact same error the second time.

I don't understand how that's specific to backwards numbers

It forces you to stop and parse the numbers because you need to invert them in your head. For me, it's the same effect as writing something down because your brain needs to process it.

But I mean in English you do this exactly as shown

> You can give someone a phone number (12345) and then verify it by saying it differently (twelve, thirty four, five)

So how does it change?

Telling time in Dutch breaks my brain. Saying “it’s ten for half five” means it’s 4:20. (I think?) I’m really not sure I’ll ever have a solid understanding.

Why can’t we just say the numbers? Why must we dance around them? In a game of tell me the time without telling me the time the Dutch will win every time.

I'd translate it as ten before half five, but apart from that, yeah that's 4:20 (inclusive or 16:20).

I think in the UK they use half five as 5:30? Half past five basically. In NL it's half way towards five, maybe the Dutch are forward looking?

There are so many ways to say this in German and we mix it all the time, though some ways are more prevalent in certain areas. I'm leaving out the 'regular' version of just saying the numbers and such and there's also the fact that depending on situation (or how you feel that very second) you'll just say 4:20 or 16:20.

    4:05: 5 past 4
    4:10: ten past 4
    4:15: quarter past 4
    4:15: quarter 5
    4:20: ten to half 5
    4:20: 20 past 4
    4:30: half 5
    4:35: 5 past half 5
    4:40: 10 past half 5
    4:40: 20 to 5
    4:45: quarter to 5
    4:45: 3 quarters 5
    4:50: ten to 5
    5:00: "full"

I'm sure I missed some from parts of Germany I've never lived in/been to.

Sometimes the actual hour is implied in a question/conversation and you just want to say that it's the full hour you're talking about and just say "Voll" or "Um". Same works with "Halb" and "ten to half" if the hour is not important or implied by context, which you can't do if you just say the numbers.

EDIT: speaking of forgetting some. While it's customary to say "10 past 4" usually nobody says "15 past 4" and instead uses "four fifteen" (actually "vier Uhr fuenfzehn") or "quarter past 4" and then at 4:20 it's "20 past 4 again".

In some regions it's (4:15) "Viertel nach vier" while it's also "Viertel fünf" because 4:45 is "Dreiviertel fünf", while in those some regions it's then "Viertel vor fünf".

(Personally, I only use Viertel, halb and Dreiviertel, otherwise it's just "siebzehn Uhr zehn" or something.)

I used an English "translation" instead of the German words for the audience here to understand better. What you mention is true and part of my list already e.g.

    4:15: quarter 5 = Viertel fuenf

    4:30: half 5

Note that to an English-speaking person, this is wrong, as "half 5" means 5:30. I once tried to explain that logic to a few Brits, in that the German "half 5" means "half [of the hour from 4 to] 5" instead of "half [past] 5", but to no avail.

Well this happens if you try to show what Germans say in another language ;)

So "halb fuenf" is "half 5".

Same with the "full" for "voll" and for "um" I gave up. No idea how to say that "in English". Or for that matter "4 Uhr 5" for 4:05. "4 o'clock 5" doesn't quite do it, though I guess it's the closest one might come lol!

Dutch is my second language, my kids are Dutch, but they still sometimes struggle with the time.

"Tien voor half vijf" (ten before half five) is indeed 4:20.

"Tien over half vijf" (ten after half five) is 4:40.

Then when it's 4:45, it's "kwart voor vijf" (quarter before five).

I always have to think about it before I say it.

> (sadly, most Germans say phone numbers as sets of two, and not as single digits) It just makes no sense and I very much prefer English, it is much more logical.

German and English are very closely related. Grouping numbers into sets of two is common in English; it would be completely normal to vocalize 2514 as "twenty-five fourteen".

Presenting numbers below 100 in little-endian order was also normal in English, though that is no longer true of modern English.

> German here... I hate how we say numbers

We have three German+English bilingual kids and maths homework can get a bit soul-destroying when you can see your child knows the numerical answer to a problem and yet instead of saying "64" says "46" (or vice versa).

Our six year old even asked me - just last week - [in English] "Daddy, why do we say the numbers backwards in German?". Me: "Umm...."

But isn’t using the “zwanzigeins” notation prone to error as well?

Zwanzigeins could mean 20 1 or 21. The only thing that differentiates “20 1” from “21” is the duration of the delay between 20 and 1…

There's a nursery rhyme,Sing a Song of Sixpence, which has "Four and twenty blackbirds baked in a pie", so English wrote out numbers the same was German does.

The King James version (1605) consistently writes out numbers in this way, e.g.

Leviticus 12.4: And she shall then continue in the blood of her purifying three and thirty days ...

Genesis 11.16: And Eber lived four and thirty years, and begat Peleg ...

Genesis 11.12: And Arphaxad lived five and thirty years ...

What about more recent? In David Copperfield (1850) by Charles Dickens, we find:

"About five-and-twenty boys were studiously engaged at their books when we went in ..."

German is an LSB language.

There's a nursery rhyme, Sing a Song of Sixpence, which has "Four and twenty blackbirds baked in a pie", so numbers were once written out in English the same way they are in German.

I checked the King James version of the Bible (1605), which consistently writes out numbers in this way, e.g.

    Leviticus 12.4: And she shall then continue in the blood of her purifying three and thirty days ...

    Genesis 11.16: And Eber lived four and thirty years, and begat Peleg ...

    Genesis 11.12: And Arphaxad lived five and thirty years ...

What about more recently? In David Copperfield (1850) by Charles Dickens, we still find:

    About five-and-twenty boys were studiously engaged at their books when we went in ..."

    ‘Six-and-thirty years ago, this day, my dear,’ said my aunt, as we walked back to the chariot, ‘I was married.

IIRC, the original Hebrew also writes most numbers this way.

Beat me to the blackbirds, but could it have a French source?

I guess America didn’t go with “Seven and four score years ago...” though

I was at school in the north of England (Yorkshire) in the late 1970s/early 80s and there were a few schoolteachers, and some old folks, who still spoke this way.

My grandmother occasionally spoke like that too (not Yorkshire)

I couldn’t figure out why she said it backwards occasionally

Even after having lived in the US for almost 15 years and only speaking English 99% of the time, dictating numbers in two-digit pairs throws me off in English because I'm still traumatized growing up with this problem.

> I very much prefer English, it is much more logical

First time I ever heard someone say that about English!

Opened this thread to say exactly this. It wasn't a problem for me before I started using English regularly, but in the past few years, I've been getting German numbers wrong more and more often. It's just so confusing and I have to consciously think about it every time.

I was a bit unsatisfied by the top answer which mostly seemed to be a reaction to the connotations of the word ‘backwards’ rather than a discussion of the history which was tacked on at the end.

I think the answer is that languages didn’t traditionally have base-10 systems of counting words (e.g. in English you see things based on scores with irregular number names below 20 persisting, and you see systems based on the dozen and gross, and money and measuring had other counting systems). When Hindu-Arabic numerals arrived (via Fibonacci et al) and were adopted, languages adapted more towards base-10 systems to match the written numbers, and English ended up with a regular left-to-right system for numbers above 19 and German ended up with irregularities up to, I think, 99 (French and Dutch also have weird systems up to 99 I think). So the fundamental point is that the reason number systems are so similar (and therefore the reason this seemed like a sensible[1] question) is that they were redeveloped based on the new arithmetic system and people don’t really notice the vestiges of the old systems much.

[1] I don’t want to say that the question is bad but rather that without the historical context it seems like a question more specific to German than something like “why do adjectives come before the noun in English and after it in French” which ends up with an answer that is roughly general history plus “that’s just how it happened”.

I personally liked the reply further down the stack that said the German order is more useful in context of counting. You put the digit that changes with every count first and the one that stays the same for a while second (or mention it only when it changes). Because you don't typical count methodically like this if the number of things counted is large, this system only obtains for two-digit numbers. This is just a hypothesis, but I thought it was the most interesting answer.

I do the same thing in English, especially when counting to estimate time. I'm not thinking "twenty, twenty-one, twenty-two" I'm thinking "twenty, -one, -two'.

I think I prefer English's ordering for that. When I say "twenty, -one," it sounds like twenty-one, but as it's the twenty-first item that's not terribly confusing. Were it "neun-, zwanzig" it sounds a bit more like "neunundzwanzig."

I suppose that problem just moves to English if you're counting down, though. shrug

It wouldn't be a Stack Exchange site if the question and asker were not insulted.

"Closed as too stupid, you bloody idiot."

All the top answers are similarly awful, just smug bleating and ignoring the question. If someone doesn't know the etymological history, they shouldn't say anything!

Used to be like that in Norway. Some older people will still say "two-and-forty", "eight-and-seventy", instead of "forty-two", "seventy-eight", etc.

In 1950, the gov. decided that it was time to standardize things - and the catalysator for this was actually the phone switching centrals/boards, that argued having one standard method would decrease errors in the manual patching. Remember, back in the day you had human operators that would operate the switching boards.

This change was called "The new counting method", and describes how numbers between 20 and 100 are counted/pronounced.

Sound a bit inconsistent - why weren't 10 - 20 also changed? Would have been great to have a language that's consistent all the way through as far as counting is concerned :)

I have nowhere enough knowledge in linguistics to properly explain this, but numbers between 10 and 20 have their own unique pronouncement which sound quite incorrect if inverted. Not too different from English, 10,11,12 have their own endings, while 13 to 19 end with a "ten" - similar to the English "teen". But saying "three-and-ten","four-and-ten" etc. doesn't sound right at all, in our language.

It's after this that you get "twenty-one, twenty-two, ... " and up to "ninety-nine" - which can also be pronounced "one-and-twenty, two-and-twenty, ..." up to "nine-and-ninety".

I wish we just said 'ten-four' etc. instead of 'fourteen' which sounds almost exactly like 'fourty'. Especially over the phone someone could easily mistake one for the other

Fun fact: Hungarian doesn't have special names for the numbers 11-19.

Back when the Spanish had driven the Moors south monks were picking over the wonderful libraries they had left behind, one of the treasures they discovered was what we now call arabic numbers - but they screwed up, they took the numbers as they saw them whole into their writing system. They took numbers meant to be written in a right to left writing system into a left to right system without reversing them.

Writing numbers smallest digits first is particularly useful in business - when you add numbers together the result can be written in order, you don't have to guess and leave enough space for the answer to fit into.

But it's also screwed us over down the generations - it's the cause for our computers' big-endian vs. little-endian sillyness - took us a generation and we finally have decided that, well, the original arabic way of doing it was right

On the other hand, network byte order is big endian, so now we typically have a little endian devices converting to and from big endian to talk to each other.

Yes, we're stuck with that, at least for old protocols

Counterpoint: when talking or skimming text, the exact number is often not especially important to most of the audience, but the most significant digit or two are.

If there are 123 new covid cases in the cycling club, I'm not going to the meeting; it's not because of the 3 or the 20.

>If there are 123 new covid cases in the cycling club, I'm not going to the meeting; it's not because of the 3 or the 20.

Are you skipping because bicycles are SUPER popular during the pandemic and you don't want to fit into the crowd?

To know the magnitude of the most significant digit, you have to scan the whole number anyways. Looking for this info at the end of the number would be just as natural if you were used to it.

Yes, number representation should be floating point with the magnitude written at the front.

I'm no historian, but that explanation doesn't make sense to me, for two reasons:

* Pre-decimal numeral systems in use at that time (Roman/Greek/whatever numerals) were already written biggest-left to smallest-right, and had been so for more than a thousand years. * Arabic numerals were invented in India, and Indian languages are written left-to-right.

Well, it's not really backwards, that would be too easy. Example:

3 482 975 is "dreimillionenvierhundertzweiundachzigtausendneunhundertfünfundsiebzig"

which is in pseudo English:

"three million four hundred two and eighty thousand nine hundred five and seventy"

Also beware of (German -> English):

Million -> million

Milliarde -> billion

Billion -> trillion

Billiarde -> quadrillion

etc.

Oh, the joys of middle-endian notation.

But who are we to laugh at them, the American date notation (M-D-Y) is just as weird.

I recently realized that M-D-Y makes a lot more sense than we give it credit for. Date formats needn't be about the cardinality of the units, in this case it's about spoken language. The M-D-Y format simply follows the order of the words in (EDIT: American) English. We say January 4th, 1970. Try reading D-M-Y in (EDIT: American) English and you will quickly realize that it doesn't make sense. I suspect most other spoken languages use 4th January, 1970 so D-M-Y feels more natural to them.

EDIT: I'm sorry if I offended you by equating American English with English. Point taken. I misspoke. The point remains.

I always like the visual look of D-M-Y in writing; a word bounded by two numbers (1 January 1970).

If it's going to be all numbers, 1970-01-01 seems like the right way to do it. (RFC3339 is the best time format.)

Yes, in England we’d typically say “(the) 4th of January, 1970”, in line with D-M-Y. I have no idea if the correlation is causation, or which caused the other.

Sorry, but English speaking countries that use DMY simply say "4th of January", and it in no way doesn't make sense.

I would say “[the] 4th of January 1970” in English. I think saying “January 4th” is an Americanism. It feels to me that in some sense 4th is acting as an adjective coming after the noun it modifies, which is unusual in English.

Interesting that Americans commonly say "The fourth of July". Is that a carry over from 1700's American parlance?

> We say January 4th, 1970. Try reading D-M-Y in English and you will quickly realize that it doesn't make sense.

Beware your unconscious ethnocentricity! Presumably you are American. I am English. English is the language I use.

The D-M-Y format simply follows the order of the words in English. "We" say 4th of January, 1970. Try reading D-M-Y in English and you will quickly realise that it does make perfect sense... D-M-Y feels more natural to us.

Good. Then we both agree that date formats tend to reflect spoken language, not unit cardinality. FWIW English is my second language. I'm sorry if my ethnocentricity offended you.

> We say January 4th, 1970

I'd say "the fourth of January". Also, quick question - what do you call Independence Day in the USA? :)

Both ways are valid and can sound fine, it just depends on the locale and context.

It is true that this is common in American English. This goes somewhat to explain why patriotic Americans get together every year to celebrate the most American of holidays, "July Fourth" :)

You could be Canada. MM-DD-YYYY? DD-MM-YYYY? Unless you're in Quebec, you never have a clue. Government officially recommends YYYY-MM-DD due to this, but in the real world it's wild west.

I'm a German in an international work environment. Since I'm also fed up with this endless confusion, I'm sneakily forcing YYYY-MM-DD on everyone by quietly editing every instance of a date that I find in the wild, and it's progressing rather well.

Oh you deviant, you! I'm inspired. Keep it up.

Spanish guy working in Japan in an international environment here. Little by little I'm forcing YYYY-MM-DD for file and directory names, and DD-Mon-YYYY as the display format for applications and for the documents I write.

Time for a rant. I recently had to fix a web application where the guy who originally installed it set all the fields to free form text, and dates were a mix of:

- D/M/Y: good, European order.

- Y/M/D: also good, CJK order.

- M/D/Y: who the hell did it, if we never had an American in the project????

- Years with two or four digits. Every time I see a date written with only two digits for the year, I want to kill the motherfucker who did it.

- Months written as numbers, full names, or abbreviations.

Well, at least nobody used Japanese era years, or Roman numerals for months, but any other combination you can think of was there. 600 records, took me several hours because the software is not exactly user friendly and I had to convert each date to its internal representation by hand.

After that, I force every single web form to use date pickers for date fields. But of course text fields are still a mess, maybe I will adopt your approach too ;)

Probably an unpopular opinion, but despite its stupid order, I think M-D-Y makes complete sense for (American) English.

Some languages (or dialects of English) say "twelve/the twelveth (of) January 2022", so 12-01-2022 is the obvious notation. Many English dialects (the majority, given that it's how America does it?) would say "January twelveth, 2022", so 01-12-2022.

It's not that the notation is spun around, some versions of English just pronounce the dates backwards! This is the exact opposite of the question asked about numbers, where English (usually, mostly) follows the "logical" order.

As language is subjective, I don't think there's a right or wrong way to pronounce and order things. The West doesn't use lakh and crore, but there's no reason why the short scale is any better. People just decided to say things one way and stick to it.

So, you’re argument is that US English makes two mistakes, so it’s fine?

I think that, in some cases, some systems are objectively better than others because they are simpler. YYYY-MM-DD is such a case. Given that it hasn’t ‘won’ globally over others, it seems it is not that much better, though.

hh:mm:ss is one, too, not because no objectively better ones exist, but because none of them get used (using mixed-base base 24, base 60, base 60, plus, when using milliseconds, base 1,000 doesn’t make sense, but we didn’t manage to replace it (https://en.wikipedia.org/wiki/Decimal_time), so we’re stuck with it.

That may have to do with the fact that it breaks down for larger units, anyways. We could have a 10-day week, but for months and years, it’s impossible to avoid things getting messy.

The way a language work is not a "mistake". It's simply how language works. People are not robots and they will exhibit illogical behaviours if that's what feels better to them.

YYYY-MM-DD is probably the clearest way to express dates for computers only because it's not ambiguous and it's easy to sort.

Decimal time was a mistake. We cannot group our 365.24 days given by nature in a round, base 10 number so any attempts to make the entire system decimal is foolish. We could divide the day in 10 hours with each 100 seconds, but that would mean redefining all SI units to use this new second as a base. Distance is defined as the time it takes for light in a vacuum to travel a certain distance, we'd need to change that distance and all other units OR have the "scientific second" and the "metric second".

A nitpick, 365.2425. Don’t forget 400 year cycle!

But I wouldn’t mind to have a lunch at ‘43’ (from 43200) and go home at ‘65’, without redefining SI.

Isn't the last part:

nine hundred and seventy five?

The point is that the way you say "seventy five" in German is "five and seventy"

in pseudo English (meaning a direct translation) it is "nine hundred five and seventy", in real English it is "nine hundred and seventy five"

I somehow missed you wrote pseudo english. Not sure how I managed that.

No worries!

> Milliarde -> billion

Yeah, and you also see e.g. news anchors making mistakes in translation saying things like (Dutch) "Ze investeerden een biljoen". Turning a billion into a trillion.

And I also heard that Biden wants to pump a "triljoen" in the economy (quintillion)

This is the short-scale / long-scale naming convention for powers of ten. See: https://en.m.wikipedia.org/wiki/Long_and_short_scale

Million (10^6) / milliard (10^9) / billion (10^12) is, or at least was, common in British English, where American English is million / billion / trillion.

Numbers above 999 million were typically written as, e.g., "thousand million (10^9), million million (10^12), etc.

Long scale also has billiard and trilliard.

American usage seems to be dominant now, the UK officially converted in 1974.

Norway used to have the same system, but the government introduced a reform in the 50s to simplify. Many still use the "old" way.

The "correct" modern way is the same as in English.

In UK English a billion used to mean a million million up until about 50 years ago.

Big-endian vs. little endian.

In general, languages differ in syntactic structure on the level of word order. For example, in Latin, word order a bit free form, things could be moved around quite a bit and still make sense to the listener. Probably allowed for some interesting creativity and poetics.

In English, I would say about a book, "the book on the table". In another language, it may be "the on-the-table book."

There's even a slight semantic difference where the later example somewhat directly gives the book the property of being "on-the-table". IIRC from my school days, that example resembles Japanese construction, but please someone correct me if I'm wrong, it's been a long time since my comp ling course work.

German numbers are actually mixed-endian when they get to three digits and above. “136” is “one-hundred six-and-thirty”.

Interestingly Norwegian used to spell out numbers in the same order as German, but reformed this in the 1950s when telephone numbers became widespread:

https://no.wikipedia.org/wiki/Den_nye_tellem%C3%A5ten

I guess having a dedicated Språkrådet to oversee the development of the Norwegian language and a single broadcasting service(NRK) made the roll-out of this possible.

Reminds me of an episode of "King of Queens" when Arthur dictates a phone number: https://youtu.be/e_B40_WXDoQ

In German, it's even funnier because of the 'wrong order': https://youtu.be/gjPmUUCdLHw (hats off to the translators!)

There are still plenty of people using the old way, young and old.

I would say I use the old way in all situations except when I read out phonenumbers.

Actually, this is a modern thing. In the old German language (pre-1800) the numbers were spelled in the correct order in the areas where Hochdeutsch was spoken, with the single digit being at the end.

Also some early books from the era directly after the 30 jaehrige Krieg still use the different way of spelling numbers (e.g. from Eva Hartner comes to mind).

Einhundert-Zwanzig-und-Eins is still a number everybody understands, and it is also accepted in written form on bank transfer checks.

So I'd say just go ahead and use it this way :) the more people use this way of spelling numbers, the more likely it is that the language will adapt and change (back).

“Four and twenty blackbirds baked in pie” is a line from English nursery rhyme. Also two score and ten meaning 50 is not that old. At some point fairly recently English must have changed to the current system. Until the 1960’s there were 12 pence in a shilling and 20 shillings in a British pound sterling (240 pence).

My grandmother would often use that form of words in regular speech.

Yeah, I'm reading Pride and Prejudice right now (1813) and in dialogue they speak numbers "backwards" like that.

Australia's largest meat pie company is named after this. Although I doubt most people realize it actually means "24".

https://en.wikipedia.org/wiki/Four%27n_Twenty

It is because little endian-ness in speech allowed market makers to trade faster. If you have a commodity that's trading at around 24 or 25, there's no point in waiting to hear about the 20, you just care about the 4 or the 5. This allowed the HFTs of the old world to trade super fast and the rest of society adopted it as a result.

Just kidding.

Number systems in different languages get pretty weird. I still have people asking my why in Danish, 50 (halvtreds), 70 (halvfjerds), and 90 (halvfems) seem to have the word "half" in them, and it's half of 60 (tres) or 80 (firs) but not 100. The reason is the number system in the top half of the hundreds actually counts in 20s (snes) but that old word is basically never used anymore. So 70 is half a 20 to having four 20s, which got shortened (fire snes -> firs). Similarly 90 is half of a snes towards 5 20s, which we prefer to call a hundred.

There is some hope though. Swedish and Norwegian are reformed, despite also being closely related to German.

I did read a popsci piece about the effect on numbers on kids learning times tables. Chinese numbering seems much more sensible. 15 is just ten-five, 52 is just five-ten-two (Vietnamese as well). In that way perhaps it directly encodes the place value system that kids need to learn, whereas naming it "two and half of the score on the way to the third" is just confusing. Personally I sometimes do times tables in Cantonese, it seems to recall a lot faster than doing it in English and certainly Danish. If you think about it, the ten in the middle is just a constant, so you are only remembering two sounds. Also there's no converting between the tens version of the number (fifty) and five. The whole table is just combinations of the basic 1-9, with nothing in the units if it's divisible by 10.

The Danish case (at least for 50) was explained by Tom Scott in a numberphile video, I believe. https://youtu.be/l4bmZ1gRqCc starting around the two minute mark.

The half in the Danish 50 is derived from an abbreviation of an abbreviation, originating from "half away from three, times twenty" ((3-½)20 = 2½20 = 50) using some nice, outdated numbering orders. He doesn't talk about 70 and 90, but your explanation makes sense. That would mean there are two ways the "half" made it into the names for tens!

> The question, why German numbers are "backwards" is naive in many ways.

What a terrible way to start an answer.

I find that in German you often have to wait for the end of a complete sentence, often including a couple of subordinate clauses, to have a clue what is going on. Let's just say that backwards numbers are the least of my problems living in Bavaria.

The German word for that is Bandwurmsatz (“tapeworm sentence”, because it goes on and on and on). Native speakers likely have a higher threshold for calling a sentence that than foreigners.

Great word. Will go into my lexicon along with Sitzfleisch, Pimperlgsund and other gems I have picked up along the way.

I was mystified by the parenthetical in one of the answers: (Please note that "hundred" once meant 120.)

Seemed unlikely, apparently the ground truth has moved more through the ages than I expected. Lo, the "long hundred": https://en.wikipedia.org/wiki/Long_hundred.

It's still no match for French numbers: quatre-vingt-dix (four-twenty-ten) for 90.

Or Danish where 90 is halvfems or (5 - 1/2) * 20

... which is especially embarrassing because they got rid of the problem in Swedish and Norwegian.

With one small exception in Swedish. We still do it backwards for some numbers in the 10-19 range. 13 = tretton (three ten), 14 = fjorton (four ten), and so on.

95 could be spoken as "fifteen and four score" in English, and I'm told my great grandfather did this, though people thought he was old fashioned.

It would not be the Danish "five and half the fifth score" though. This is crazy.

Yeah but I'd never blame someone for telling me "nonante", the Belgian (and more correct latin) variant.

French is not actually that weird; it's just weird from 80-99. From 10-79, it's just like English handles it. They just never came up with single words for 80 and 90.

It's also weird for 50-59. English does not say "fourty twelve" ("quarante-douze").

Neither does French, you're confusing 50 with 70. https://frm.wiktionary.org/wiki/quarante-douze

Edit: 50, not 60.

At least in Belgium we fixed that with "septante"

But making things completely logical would have been too much for us so we kept quatre-vingt.

It's weird form 70-79 as well! To say 79, you need to say sixty-ten-nine!

But they actually did! Septante, huitant/octante and nonante are real French words. Only the French don't want to use them.

Note that you cannot use heptante for septante.

I think various french speaking regions outside France do use single words for those. E.g. in neighboring Switzerland I learned 70, 80, and 90 as septante, huitante and nonante if I remember correctly (I'm from the german speaking part)

Yes, they do. Belgian French has "nonante", for instance, which would be better, but which just never caught on in "standard" French. It appears in my Cassell's French Dictionary, but it's marked "dial." for dialect.

> Four score and seven years ago our fathers brought forth, upon this continent, a new nation

nonante!

There’s a joke in French:

What is 20*4?

Answer: 80 because multiplication is commutative.

Yet, Commutative Algebra is a German invention.

Be happy it's only backwards from 10 to 99.

Have you checked Malagasy? We read the numbers totally backwards although we've switched to the Latin alphabet like 200 years ago... It's sometimes so inconvenient for everyday life, especially for large amounts, that we end up counting in _French_ instead. For numbers 11 to 99, we also use casual abbreviations like (I translated) "one with two" for 21; or "seven with three" for 37, but reading large numbers from 10,000 upwards (mostly Malagasy currency) with the left-to-right writing system is tedious.

As the comments point out, our counting system in English does precisely the same sort of thing, often: thirteen, fourteen, fifteen, sixteen, etc., all name the ones digit before the tens. And because of logic and ease of counting relatively small numbers of things. You can tell this is the reason since, once you get over 20, all this reverses: twenty-three, twenty-four, twenty-five, etc.

It's the same pattern in French. Under 20: douze, treize, quatorze, quinze. Over 20: vingt et un, vingt-deux, vingt-trois, vingt-quatre, etc.

For some reason Italian flips at 17: sedici (16, six-ten), diciassette (17, ten-seven).

I think many English speakers would admit 11 through 19 are a bit of a weird case. It would probably make more sense if we went ten, oney-one, oney-two, etc. I imagine most English people subconsciously treat 13-19 as individually named as 0-12 are, rather than comprehending them as a composition of two numbers as 20-99 are.

That said, based on the other commenters here, the mixed endian nature of German counting seems very strange to me, being able to generally read numbers the same way they're serialised on paper seems useful.

So do eleven ("one left over") and twelve ("two left"), with a bit of stretch.

https://www.etymonline.com/word/eleven

https://www.etymonline.com/word/twelve

Arabic same as German. Urdu takes gold. Every number from 1 to 100 has a different name

> Urdu takes gold. Every number from 1 to 100 has a different name

Urdu/Hindi numbers (20-99) are mostly small-digit–big-digit combos, with special combination rules. One doesn't really have to remember eighty different numbers because of the rules. Plus, the distinction between how single-digit numerals are pronounced and how these combined digits are pronounced helps avoid confusion regarding order of the digits.

Well I somewhat disagree. The big number small number combination is not trivial to remember. Sometimes many of them don't even have any relation to previous numbers. Even to this day 40yrs on, I have to think hard before uttering any number. Eg: try remembering Unasi (79) and navasi (89), sarsath (67), etc.

Yes, it's not utterly consistent. Which is why I said "most" above.

EDIT: I too have to pause and confirm sometimes. But then that's only for the exceptions; which I sometimes rebel against and use the rule shamelessly for anyway (e.g., niña-van instead of un-sath). Then there are cases I always break the exception in favour of the rule, like saadhe-ek and saadhe-do instead of dedh and dhai.

I find it more curious that the languages I know best tend to have special words for 11 and 12 that don't follow the same logic as the rest. Eleven and twelve instead of one-teen, two-teen or something. And that even leads to things like a teenager being age 13+. In German it's the same elf + zwölf and then it continues with dreizehn, vierzehn etc. My guess is that it is somehow related to the fact that a dozen is a thing but I'm curious where it comes from. In French it goes all the way up to 16 (onze, douze, treize, quatorze, quinze, seize) before we end up with dix-sept for 17. French has always been the most peculiar to me as there's stuff like 82 being quatre-vingt-deux (4*20+12). And then there's languages like Vietnamese that happily start with 10+1 from the get go (mười một, mười hai, mười ba). Fascinating stuff :)

I was curious and found this rationalisation: https://www.mentalfloss.com/article/76007/why-it-eleven-twel...

It's pretty hand-wavey but still interesting speculation.

Historically english also spoke numbers in the same way as in German. For example “four and twenty blackbirds”.

As in the nursery rhyme:

https://en.wikipedia.org/wiki/Sing_a_Song_of_Sixpence

That form of counting is still understood, recognised, but viewed as archaic in England.

I guess it got lost in the evolution of Old English to Middle English, and the interplay of Old Norse plus the subsequent influence of Norman French, all of which bashed the Germanic core of of English in to its modern form.

Numbers in multiple Sherlock Holmes books are spelled out this way, so that switch must have happened much more recently.

I know elderly people who still speak this way. It's changed in the last hundred years.

> There are many more languages that speak or read (some of) their numbers "backwards", among them Greek, Latin (both directions possible), Celtic languages etc., and of course languages that actually read right to left like Arabic, where our written numbers come from

Ironically, in Arabic numbers are written left to right, just like in the west, reflecting that they were borrowed from India, whose indigenous writing systems are also ltr. It goes to show that not only is there not a "correct" order to express numbers, but the spoken order need not reflect the written order.

I am sure it has been said, but it is not any more backwards than "thirteen, fourteen, fifteen, sixteen, seventeen, eighteen, nineteen" -> three and ten, four and ten, five and ten...

In fact, in German it keeps being consistent, while in English and other languages you change the direction after 20... quite arbitrarily...

When you speak German they as backwards to you, as thirteen is to you when you speak English... Not backwards at all...

And then there is eleven and twelve... Like not even consistency inside one decade... We in Finland have decency two say "one-second ... nine-second"

Yes, we say "yksitoista" (one of second), "kaksitoista (two of second) and so on until twenty, where we revert to "kaksikymmentäyksi" (two tens and one; ten in actual plural). But in some old books from the early 20th century, the reversal goes on beyond 20: "kaksikolmatta" (two of third) for 22 and so on. I do not know how that archaic usage deals with 100+.

I often see a similar problem in code that deals with coordinate system transformations. But there is a simple solution! It boils down to properly naming your variables.

When using transformation matrices to transform from one coordinate system "In" into another coordinate system "Out" you have two options to name the matrix: M_{In,Out} or M_{Out,In}, which can be read as "In to Out" and "Out from In". Unfortunately what I often see is the first notation. It looks simpler at first, but it is actually backwards, similar to how the german numbers are backwards.

When you chain the transformations you get this weird forth and back:

M_{A to C} = M_{B to C} * M_{A to B}

M_{A,C} = M_{B,C} * M_{A,B}

Compare that to the (not backward) alternative:

M_{C from A} = M_{C from B} * M_{B from A}

M_{C,A} = M_{C,B} * M_{B,A}

Note how the Bs line up, C is the most left in both sides of the equation and A is on the right. When you transform a vector from coordinate system A to, lets say C, it looks like this: "vec_C = M_{C,A} * vec_A". Everything lines up and is in the natural order the transformations are taking place in. Compared to this, the backward notation is really confusing.

I see people make mistakes when dealing with transformation chains all the time because they use this weird M_{From,To} notation, or worse, no notation involving the coordinate systems at all.

Like take this example: "M = M1 * inv(M3) * M2", what does that even mean. Find names for the involved coordinate systems and name the transformations accordingly. That could be: M1 transforms to A from B, M3 is C_B, M2 is C_D. Then M is A_D:

B_C = inv(C_B)

A_D = A_B * B_C * C_D

My native language is Dutch, but I've been living in an English speaking country for almost 10 years now. It's annoying to having to change my mindset whenever I visit friends and family in my native country.

In Belgium, we say 25 as "five and twenty" (same as German), which makes sense, but if you've been saying "twenty-five" for 10 years, it does throw you off.

Not to mention that "half five" means 17:30 in most English speaking countries but 16:30 in Dutch.

In the US, I don't think most people would know what to make of "half five".

(I bet something like 20% of people would think hard and then decide you meant 2:30)

and let's not forget our masterpiece "ten to half five", one of the most straightforwardly intuitive ways of saying "16:20"...

I have noticed that I somewhat simplified my vocabulary in that regard, previously I would say "half five", but now I just go with "four thirty".

I've personally opted to never use the expression "half five" in English because it seems able to cause any amount of confusion and misunderstanding among people from different linguistic backgrounds.

I's ambiguous in British English, at least it used to be. Half five can be short for either half of five (4:30) or half past five (5:30).

Really? Where in Britain would that be ambiguous? I've grown up here and I've never heard someone say "half X" to mean "half an hour to X".

In some dialects of German they go even further, using “quarter five” for “16:15” and “three quarter five” for “16:45”.

At least that makes sense, telling time in Dutch switches orientation halfway through the hour, towards the closest half hour. German does the same thing, I believe.

"five past five" is 5:05, then of course "ten past five" and "quarter part five". Then comes "ten to half six" which would be 5:20. Then half six, ten past half six, quarter to six, ten to six and five to six.

It's interesting to see how Dutch and German time telling is clearly oriented at half hours while English is oriented at whole hours.

Now that the world is ruled by digital clocks, many people will just use digital (24 hour) time. "Eighteen hour four" would be the current time in this notation, which is a lot simpler. It's funny how the tool we use to tell time dictates the way we pronounce things!

I’m American and I wouldn’t know what to make of “half five” in English (I speak some German and it’s the same as Dutch there). To say 17:30 you’d say “half past five”.

After reading the answer, I've got some questions:

1. I like little-endian systems because adding or subtracting small numbers is easier in those systems byte-wise. Would the same benefit apply to human languages?

2. Why did the Arabic numeral system choose big-endian in the first place? It could easily have been little-endian, even including zeroes, likes 001 meaning one hundred. Who made the choice?

Arabic is written left to right, so aren't the number little-endian in their Arabic form ?

I was going to correct your mistake, but then I understood that you just wrote "right to left" right-to-left.

Ooops. Yes, I meant right to left. Not enough karma to edit and fix my mistake. Thank you.

Not sure about Arabic, but in Hebrew numbers are written from left to right.

In arabic numerals , but in hebrew numerals (a alphabetic numeral system) its right to left.

I prefer big-endian systems because you more quickly get an idea about the magnitude of a number as the bits come in.

It's quite important for programmers especially to decouple the idea of numbers from the notation and representation of numbers. Start with Knuth, chapter four. There is only one integer that we call "one", but there are infinitely many names for it: "un", "ein" etc. You could even make up your own name if you like. There is no right or wrong way to say or write numbers, there are just more or less convenient ways, depending on the application.

When it comes to natural language, the answer to "why is X the way it is?" is always ultimately the same: because it works. It is sometimes interesting to learn that this word came from that root and so on, but ultimately we're just making sounds with our throats and mouths that go into other people's ears. Everyone around the world is talking about the same stuff, there are just so many ways to say it we often settled on different ones.

Tbh, as a German I never really noticed this.

Thinking it trough, it also makes me wonder why it ain't consistent and breaks down after more than two digits? 21 is einundzwangig, but 121 is einhunderteinundzwanzig, 1121 is eintausendeinhunderteinundzwanzig, and so on. So there it's not really backwards anymore.

But once you’re at tens of thousands, it rears its ugly head again, einundzwanzigtausend.

It's the same in Dutch... But French is even weirder : 96 becomes four-twenty sixteen...

Danish is the worst.. “six half-fives” (6 + (5 - 1/2)*20)

This is just the etymology. These words are just words, both in Danish and French. It's not like speakers of that language do the maths before saying it.

It's only backwards for the last 2-digit numbers. Once you go into the hundreds and beyond, it's read from left to right. I think most Germans probably don't even realize it. They just recall all 2-digit numbers from memory.

In French, to an outsider, quatre-vingt-dix-neuf (99) literally sounds like 4 * 20 + 10 + 9... But most french people will think of 'quatre-vingt-dix' as a single word (90) not thinking of quatre (4), vingt (20) and dix (10) as 3 different words. This is especially true because a lot of French words and names have a hyphen in them so French people see the hyphen more as an integral part of the word rather than a separator.

> It's only backwards for the last 2-digit numbers. Once you go into the hundreds and beyond, it's read from left to right.

Well, no, because “twenty three thousand” will be “three and twenty thousand”. Ditto for millions.

I guess my knowledge of German is still too basic. I know in the hundreds it goes left to right so I assumed that it would continue until infinity. That is extremely confusing.

French and English are very simple and consistent by comparison; you just need to memorize the first 10 to 20 numbers, every 10th number from 0 until 100, every thousandth number from 0 until infinity (or however high you want to know). All the other numbers can be derived using the same simple rules.

In one of the stack responses they say:

> and this is also done in modern spoken Arabic with no strange reversals as in German

That is not fully correct. Arabic numbers are spoken in a similar way to German. For example, 42 is "2 and 40" (إثنان و أربعون in fus7a)

It is somewhat natural to read a number like that in Arabic. As you are reading right-to-left you come across a number where the least significant digit comes first.

In other words in English if you see 123 you think "one hundred twenty three" as you read left-to-right. That same number is written ١٢٣ (the 3 least significant digit is to far right) and is read "one hundred and three and twenty".

Now I'm curious, are there any cultures on earth that are fully little-endian? I know RTL languages like Hebrew and Arabic put the least-sig-digit on the right and are therefore read as little-endian, but afaik when they read them aloud they follow the Germanic approach of "Three hundreds and four and twenty", not starting from the least-sig-digit.

Also, Unicode does this crazy thing where it switches typing direction to LTR when you start Latin numerals even if you're in an RTL language, I think. I worked on an i18n project and the behavior of Unicode with RTL languages confused the hell out of me.

Byblical Hebrew has numbers in "two and twenty" format (including things like "fifty and hundred"), but in modern Hebrew they are pronounced similarly to English, including special exceptions for 11-19 and using Million->Billion (Milliard allowed as well)->Trillion rather than Million->Milliard->Billion->Billiard; And not including the "thirty five hundred" oddity, in Hebrew that's "Three thousands five hundred".

Here's a guess: It could be because of an emphasis on counting. Meaning, when counting, the significant, changing part of the numbers is the least significant figure, while the higher portions (potencies of the numeric system) become soon redundant. There seem to be a cultural differences in languages regarding the range in which this seems to matter: while it's just the tens in English, this is the entire range up to 100 in German. However, even in German this counting range ends at 100 at which point we're probably dealing with numbers that had been accumulated previously.

Interestingly, I have a German girlfriend who, when speaks English, sometimes reads numbers backwards mistakenly. For example, she reads 73, thirty-seven before quickly fixing it.

If you think German numbers are strange … don't even look at French that require multiplication and addition. Four-twenty-ten-seven? … yup, you guessed it 97.

https://jose-lesson.com/lin/2017/01/16/nonaginta-septem/

That's just the etymology. It's a word. French speakers aren't multiplying in their heads...

The answer goes through great lengths explaining why it’s just historical and everyone did it so it’s not backwards. That sounds backwards to me.

Not necessarily only in German. A lot of Indian languages also say number in this way.

In Bengali, we don't use full words for the 1's place, but a derivative of the root.

31- ektrish- ek (eka- 1) trish (30) 32- batrish- ba 33- tetrish- te (tin) 34- choutrish- chou (char) 35- pointrish- poin (panch)

and so on.

Now that I mention this, I also remember that Sanskrit words for number are like this, as well-

35- panchatringshati (pancha- 5, tringshati- 30)

Arabic/indian numbers, when written in arab writing direction from right (lowest significant decimal) to left (most significant) can just be read logically without scanning ahead but in order small to big.

Maybe german borrowed this RTL reading order for the 2-digit numbers.

German borrowed the word 'Ziffer' (= 'digit') from the arab word for 'zero', so there's some connections.

As a German who nowadays speaks French most of the time, I long for my simple German numbers...

99 in French is quatre-vingt-dix-huit (four-twenty ten-eight)

Not that it matters too much, but what you speller out is 98, not 99

Fascinating. I studied Finnish for three years and it never occurred to me that the numbers 8 and 9 (kahdeksan and yhdeksan) mean 2-from-10 and 1-from-10. Though 10 isn't "deksan", it's "kymmenen". Still the "yh" and "kah" should have been clearly seen as akin to "yksi" and "kaksi" (one and two).

I think modern theory is that stem is something like "-eksa" meaning without. And link to indoeuropean ten is just coincidence. Likely more believable to me looking at rest of the small numbers, which are pretty local.

It may have to do with the fact that, in German, by default the first syllable of a word is stressed, and in two-digit numbers the least significant digit is actually the more distinctive one in ordinary usage (in terms of wanting to express an exact number). Note how in English, the stress is usually on the least significant digit as well ("twenty-three").

If you think about it we say numbers "backwards" in english in the teens: "four-teen, fif-teen, six-teen" and so on.

Me as a foreign mother tongue find it really annoying, so I count in my head in my mother language.

There also has been a study, no matter if forward or backwards.. the first language you learned to count will be the language you use for counting in your head for the rest of your counting life

Isn't that cool? Never noticed, now Always pay attention to :)

I also do most math in my native language, Spanish. When I'm silent reading text in English, most of the times I "say" the numbers in Spanish, but now always. Anyway, I have more intuition about numbers in Spanish, so even if I read it in English I may translate a few milliseconds later if I must think about it.

I can do some elementary calculations in English, like 2+3=5. But for more complicated stuff like the second derivative of x^3 I must switch to Spanish and translate the result.

I’m split about 50/50 with German (native and residence) and English counting. But when I count in German and write numbers, I tend to make more mistakes because of them being "backwards" :(

Yes, it's a bit naive to ask "why they were initially conceived backwards"

But why it remains as such is just anachronistic ;) English inverted them back, so it is consistent with the rest of the > 100 numbers. (ok, the 0-20 range has exceptions in many languages, so it's fine)

> (ok, the 0-20 range has exceptions in many languages, so it's fine)

(ok, the 0-99 has exceptions in many languages, so it's fine)

It is worse than "backwards". It is mostly forwards except for the special case of the second digit that is transposed with the first digit. For example, 123 becomes "one-hundred-three-and-twenty". It was driving me nuts for ages.

Another confusing scenario to me is that Arabs who invented the numbering system write the digits left-to-right in a right-to-left writing system:

رقم هاتف والدتي هو 8765309

It won't let me paste vertical Japanese text or I would demonstrate that monstrosity too.

Most likely because Arabs did not invent the numbering system: https://news.ycombinator.com/item?id=29370553

Yeah let's not do original research in HN. This can have millions of reasons.

“Why are numbers “backwards”?? What a naive question! So ignorant! What’s backwards to you is forwards to somebody else! Anyway, when Germans studied numerals they decided to write them backwards in keeping with their written text”

NO language is free from long-standing sources of annoyance.

If you doubt that, look at inconsistencies in English verb tenses, or pronunciations in French, or (English again) the spelling or words. Or even the myriad of alphabets out there.

Netherlands calling. Same problem over here. Perhaps I would find inversion a little bit less annoying if it were at least applied consistently but no "123" = "one-hundred-three-and-twenty"!

My parents emigrated to the Netherlands when I was young. In my native tongue we say numbers in the same way as they do in English. My 8 year-old brain struggled to understand this new way saying numbers when I started school in the Netherlands. I got there in the end but I recall it feeling like it took forever.

It's not weird, the problem is you learned to communicate in English and learned another way of vocalizing numbers.

Given that in English, as well as in German, contrary to the roman letters, the numbers are arabic signs, I wonder in which direction arabs spell their numbers.

Does anybody know?

It's confusing.

For everyday usage we use a mix of the two; thus, 1998 is read as "a thousand and nine hundred and eight and ninety". So:

1998 >><<

But according to the traditional way (you hear it used in the news for example), it's read as "eight and ninety and nine hundred and a thousand". So:

1998 <<<<

However, this is so rare that many of the younger native Arabic speakers aren't even aware of it.

So in written Arabic, the digits go from least significant to most significant which means a number looks the same when written least-significant first right-to-left as it does in e.g. written English, most-significant first left-to-right.

Absolutely.

While I don't know for sure, some blame the new (mixed) way of reading numbers on the proliferation of colonial languages (English and French); i.e. people trying to mimic how the numbers are in these languages, but only doing it half way through.

The problem though is that this mixed way is used across the region, so it developing independently each time seems a bit way too improbable to be a solid cause.

Isn't arabic in general written and read from right-to-left? That way the traditional way seems very consistent.

If you think about it, arabic numerals are a right-to-left script.

2,021: one and twenty and two thousand.

To make my school life miserable. We switched language of instruction for maths and science between them at various grades

This is not unique to German. In Urdu, numbers higher than 20 are pronounced: 1-2 for 21 (aik-ees), 5-4 for 45 etc.

Try French... 95? Quatre vingt quinze. That's four times twenty fifteen.

German at least makes some kind of sense.

German numbers are basically written as:

10ⁿ 10ⁿ⁻¹ … 10⁴ 10³ 10² 10⁰ 10¹

Descending powers of 10 then the old switcheroo at the end.

The switcheroo happens in every three-digit group: When you read the number 123,123,123 out loud, you say "drei(3)-und-zwanzig(20)" three times. (I've used comma as the thousand separator here.)

Taken to its extreme, numbers would be orally expressed in binary digits or pick your base-n!

Why is it thirteen instead of ten-three?

Because lanuages are illogical, with lots of historic baggage.

I am wondering why the English have it only backwards for the numbers 13 to 19.

Yan tyan tethera methera pimp... dick... bumfit... giggot.

Which is to say, who are we to judge German numbers when there are weirder ones:

https://en.wikipedia.org/wiki/Cumbric#Counting_systems

confuses me a lot. I'm German but tend to speak numbers in English in my mind when writing them down.

also impressive is the number of cases:

eins eine einer ein einem einen eines

probably forgot a few

Turkish seems to be the most systematic than:

11 => On Bir (Ten One)

12 => On iki (Ten two)

…

21 => Yirmi bir (Twenty One)

Etc

Yet, the names of tens don't seem to follow any logic:

10: on (no connection to 1: bir)

20: yirmi (no connection to 2: iki)

30: otuz (no connection to 3: üç)

40: kırk (no connection to 4: dört)

50: elli (no connection to 5: beş)

60: altmış (sounds like 6: altı)

70: yetmiş (sounds like 7: yedi)

80: seksen (sounds like 8: sekiz)

90: doksan (sounds like 9: dokuz)

To this day, I can't figure out how "yedi" becomes "yetmiş", or whether they're related at all. Are they borrowed from Farsi?

No, Persian has words that are vaguely derived from the basic numerals.

It comes from Chuvash language. İnş forms counting numbers, and language researches think they used to count in tens, as they didn't need precise numbers.

yedi/seven (m) inş -> yētiminş -> yetmiş/seventy

altı/six (m) inş -> altıminş -> altmış/sixty. see: https://www.nisanyansozluk.com/?k=altm%C4%B1%C5%9F (in Turkish)

90 is tokuz-on (dokuz-on, nine-ten), 80 is formed that way too.

40, 30, 20 have specific names in old Turkish. Which means they used those numbers a lot, but no evidence why. No connection to 4, 3, 2 whatsoever (old Turkish for 4 ist tört, for example).

50 (elli) comes from the same root as the word which means hand (el). You can perhaps guess why.

Long story short: You just need to memorize them if you are speaking Modern Turkish because "historical reasons". Turks started using numbers a bit too early and early weirdness was never fully "corrected" :)

What I like in Turkish is that you don't use the unnecessary "one" before hundred/thousand. 100 -> yüz, 200 -> iki yüz. you don't say "bir yüz" for 100, just "yüz".

I know it's an irrational pet peeve, but i'd be happy if the German's (and others) would just stop using comma as decimal point.

To me is a bit weird, since I learned it that way.

But the dot as decimal point also makes sense.

Now, using comma as a thousands separator? Nuh-huh. Doesn't make any sense to me

2'000? Fine. 2_000? Fine. 2,000? It's just awful

I'd be happy to compromise so that Europeans ban decimal comma and Americans start using metric system.

At the international meridian conference of 1884, the French allowed the resolution for using the Greenwich meridian[1] to go through on the condition that another resolution promoting the use of the metric (or ‘decimal’) system (including decimal time!) also went through.

[1] or as the French called it, “Paris mean time, retarded by 9 minutes and 21 seconds”

Reply on news.ycombinator.com