How to handle numbers in your target language?

[reineke]

Mathematical and Linguistic processing

Abstract "This study investigates the neuro-mechanisms underlying mathematical processing in native (L1) and nonnative (L2) languages"

"Many people experience the confusion doing calculations in a second language (L2). Even proficient L2 learners resort to their native language (L1)1 to perform mathematical operations..."

http://ilabs.washington.edu/kuhl/pdf/Wang_etal_2007.pdf

I would like to note that if you're having trouble recalling simple numbers such as 9, 23, 72, 92 in French, German etc. it's likely a matter of insufficient exposure and automatization.

[Ani]

I am still working on it but I do have a suggestion for what is helping me.
Since I am teaching my little ones math bilingually, I needed to get this down really fast. I still stink over 1000 and I am best within 100 BUT.. What I have been doing is using the abacus to drill myself really quickly for a few minutes a day. I randomly swipe over some beads, say number. Swipe some more, say number, add 5, say number, subtract a handful, say number, etc. It saves me the step of internally translating digits on paper (even if I try not to, the English comes first) and you can practice ~50 a minute.

[Tomás]

Another learning technique idea: I just started working through a Spanish MOOC in descriptive statistics. The math in the lectures is very easy, and so it is more of an opportunity to learn to put Spanish words and phrases to things I already know. Sort of like watching a movie you already know in your TL. Except with numbers....

[Alphathon]

In Dutch, the order of saying two-digit numbers is reversed from the order in English. For example, twenty four is vierentwintig (four and twenty). I found that learning to hear and say such numbers in Dutch was a whole lot more difficult than it would seem. It always caused a brief hesitation in either direction, which usually involved translation to English and re-ordering. You'd think that would be simple. But it took a lot of time for that to become more-or-less automatic.

I have read several articles (which I can't seem to find at the moment) that Dutch school children are disadvantaged at a certain stage in their development of mathematical skills because of this reversed number order. For example, large numbers such as 24,394, 837 are read in Dutch in order (243948 but then 7 and 30). That apparently upsets some basic math thinking at early stages of development, and you can understand why. So it is a small challenge to learn Dutch numbers if your native language doesn't reverse the last two digits. But Dutch is so much easier (and regular) in so many other important ways that I treat this as just an interesting characteristic of Dutch.

The same is true in German, but I don't really have a problem with it. To be fair it's one of the few things that I remember well from my high-school German so I suppose whatever we did made it stick. This also seems to have carried over into Dutch (as does so much else). My biggest problem with both seems to be parsing them when written out - I just don't see the boundaries in vierentwintig or sechsundachtzig. That said I have the same issue with medium-to-long compounds in German and is likely related to my dyslexia, so I think it's merely coincidental that it happens with numbers.

Gaelic is a different story. I'm fine with 0-10 and 11-20 aren't too bad either but anything higher than that is pretty complicated. For a start Gaelic has both vigesimal (base-20, like French) and decimal counting systems, and they don't work alike. Decimal is pretty straightforward: fifty-one cats translates as caogad is a h-aon cat, literally "fifty and one cat*".

Vigesimal on the other hand, in addition to being alien to English speakers, has peculiar word order: in vigesimal fifty-one cats translates as dà fhichead cat is a h-aon-deug, literally "two-twenty cat* and one-teen†". In vigesimal they also use a separate structure for 21-39: thirty-one cats is aon chat deug air fhichead, literally "one cat* teen† on twenty". (They also use leth-cheud (half-hundred) for fifty, presumably for brevity since dà fhichead X is a h-aon-deug is a bit of a mouthful.)

Needless to say this is a little difficult to wrap one's head around without a good deal of practice.

* In vigesimal plural marking is not used after aon (1), d(h)à (2) or multiples of f(h)ichead (twenty); it isn't used at all in the decimal system for numbers >20. Also, aon and dà lenite following nouns, that is they mutate the initial consonant: cat (/kʰaʰt̪/) becomes chat (/xaʰt̪/).
† Deug doesn't translate literally in many cases. It basically means "teen" – (a) trì-deug translates as thirteen, where trì means one. However, as you can see it is also used with aon and d(h)à (so "one-teen" and "two-teen"), and it also doesn't always occur alongside the unit number.

EDIT: Oh, and I forgot to mention, they also have a separate set of numbers for counting people. Luckily they only go up to 10.

[jeff_lindqvist]

In Gaston Dorren's Lingo: A Language-Spotter’s Guide to Europe there's a chapter on Breton numbers - a combination of the backwards system, vigesimals and "half-hundred".
Examples:
73 three-and-ten and three-times-twenty (13, 60)
52 two and half-hundred (2, 50)

(All Celtic languages work like that, more or less.)

[Tomás]

Another issue is colloquial use of numbers. For example, the ring road around my city is the 664. English speakers would call it the "six sixty four". Or, rarely, the "six six four". No one would ever call it the "six hundred and sixty four" -- even though that is how an ESL student would be taught to say that number.

[jeff_lindqvist]

Good point. Personally, I say numbers the fastest way or the way they make most "sense", which is often according to how they are grouped. Anything up to three digits are usually read as compunds (e.g. 108 is one hundred and eight), and not one zero eight), phone numbers are grouped in twos or threes, and read accordingly - e.g. ninety-nine sixty-eight eighteen (99 68 18), two hundred two hundred (200 200), sixty-nine thousand (69 000) and so on. Four digits may be read in twos: twenty-three fifty-four (if 2354 is a numerus currens, a code or similar) or even two-three-five-four. Years are usually like fifty-six (1956), zero-eight (2008) if you're sure that the others know which century you mean.

[Bao]

I don't agree that saying phone numbers as two or three digit numbers makes any sense at all. It takes more concentration to en- and decode a three digit number than three single digit numbers that are said with a pause before and after. The only time it makes sense to say two or three digit numbers is when memorizing and recalling a phone number, something we rarely do nowadays. Plus, people who memorized a phone number really well and can rattle it off are often too fast for the listener to write it down, let alone remember it without taking a note ...

[tommus]

Another issue is colloquial use of numbers.

And the English language use of O (as in oh) instead of zero, naught or nought. O is a letter and not a number. But it is very commonly used as a number. The main highway in this area is "the 401", pronounced four oh one. The "oh" has one less syllable than "zero", and flows better and faster in colloquial speech. Then there is James Bond, 007, pronounced "double oh seven" and never "double zero seven" or "zero zero seven". Zero also has the connotation of a "nobody", certainly not what you would want to associate with a debonair secret agent.

[Chung]

You only get better with practice as with any other linguistic feature. Just as the other lexical categories (somewhat inaccurately simplified as "parts of speech") have their respective characteristics or quirks, numerals have 'em too.

After a while, I've come to group the "problems" that I've had to grasp whenever dealing with numerals into the following.

1) cardinal vs. ordinal

This is for me means memorization, but sometimes it's a pain at first when the terms bear no similarity to those used in languages already studied as I had experienced when learning Turkish numerals.

2) conventions for grouping, basis, naming or reciting (e.g. decimal vs. vigesimal, separator placement in writing).

A recent problem in this area was for me to grasp the originally Chinese convention of higher units using myriad (ten thousand) that's also used in Japanese and Korean. I've had to spend a fair bit of time purging my westernized mind of one million to conceive it instead as 십만 (~ "sheep mahn") or "ten ten-thousand" in Korean. This is on top of the fact that in Korean the Chinese-based and native conventions coexist whose use is governed by the class of noun to be modified. Through a Pakistani co-worker, I've also been introduced to South Asian convention with its use of लाख / لاکھ (~ "lakh") or one hundred thousand to me (1,00,000 per their convention with numerals) and करोड़ / کروڑ (~ "crore") (1,00,00,000 per their convention) or ten million to me

3) agreement/concordance of elements modified by a numeral

For example, a noun modified by any numeral remains in singular in Hungarian, while elements modified by 5 or greater in most Slavonic lannguages decline in genitive plural but conjugate as 3rd person singular. From my Korean studies, I can also add the incremental burden of using counters that are lexically determined. In crude terms, it means that it's not correct to say "two people" and "two cups" as in English. Instead I have to remember to say "two [counter for people] person" and "two [counter for containers] cups" with the counters being distinct lexemes.

4) changes to the numerals' forms depending on their role in the sentence.

For example, "2" in Hungarian stands out from other numerals in that when it's attributive, it goes from its citation form of kettő to két. E.g egy, kettő, három... "one, two, three..." but ket kép "two pictures". This concept is applied further in Meadow Mari with most numerals having distinct attributive and predicative forms. E.g икте, коктыт, кумыт... "one, two, three..." but ик школ / кок школ / кум школ "one school / two schools / three schools". On a related note, certain numerals in most Slavonic languages are declinable as if they were adjectives. E.g. Slovak: jeden, dva, tri... "one, two, three..." but dva domy "two houses" (masculine inanimate, nominative plural), dve pivá "two beers" (neuter, nominative plural), o dvoch domoch "about/concerning two houses" (masculine inanimate, locative plural), pod dvoma pivami "under two beers" (neuter, instrumental plural).

All of the above is distinct from grammatical number.