Negative Impact of English on Math Learning | Hacker News Reader

Negative Impact of English on Math Learning

wsj.com

38 pointsghosh41 comments

The same should then also apply to the German.

French is even worse with their sixty ten (70) and four twenty (80) and four twenty eleven (91).

This problem is actually obvious to any parent whose native language has natural naming for numbers after 10 but whose kid gets math education in any of the mentioned languages.

Some languages have separate but still unified names for the teens (compared to the other tens). I think that it would actually really make sense to also unify these names.

I also think that kids should be introduced to some other numerical system beside decimal in their early years.

This should be easy to test with French speakers, compare those living in France with those in Switzerland. French speaking Swiss use septante (70), huitante (80) and nonante (90).

The same goes for Belgium.

On the other hand, for other purposes, German is a far more regular and logical language than English. Generally, complicated nouns and verbs are composed of simpler nouns and verbs. Quite often, the German word for a thing is a pretty bald description of what it is, whereas the English word often is some sort of one-off, where it is not a loan-word stolen from some other language:

Panzerkampfwagen = Armored battle wagon

Tank = (British code word that stuck for some reason)

Schreibtisch = writing table

Desk = ?

True that. Enjoy those genders and plurals!!

The french have it easy. Danish for ninety - halvfems - is essentially "(5 - 1/2) * 20". Doesn't seem to hurt their maths though.

I remember (in the uk) doing different bases when I was 9 or 10 and we did binary at 14 in the computing class

I haven't read the scientific article article in question, so I cannot challenge the statistics (if any) or the methodology (since I don't know it), but I'm curious if there's any guess about the underlying mechanism here. Because it sounds a little fishy to me.

My native language is one of those that would supposedly make it easier to count beyond 10 (i.e. if you go back long enough in the linguistic history, 11 would be translated as "one [unit] after/on top of ten"), but the generalization is awkward: 21 is roughly "two tens and one". You have to remember about as many rules as in English (eleven and twelve are awkward, the rest aren't. My whole touted advantage would be in only two numbers.).

Furthermore, the positional system often means that, when you're computing "11 + 13" as a kid, you don't even say the words eleven and thirteen while performing the addition.

Besides, if complexity of the numerical system would be an indication, no one in France should know math. Quatre-vingt-dix-neuf is a frickin' lesson in arithmetic all by itself. I think 90% of the awkward conversations I've had in French were "please show me the numbers on the screen because I have no idea what you're mumbling there" at the supermarket.

I agree. It’s less about numbering system in different language and the numbers from eleven to nineteen is really just one tenth of the numbers in 0-100. It is simply that Chinese children spend much more time practicing those maths, because of many historical and culture reasons. They actually never think about any words when calculating 11 + 13, they just know the result right after they see them.

However, in terms of multiplication, the problem of English is not ..teens, but the pronounciation of numbers. Chinese or Korean has one simple syllable for each number, so only two syllable (at most 3 if you count “ten”) is needed when you shout out 32. Chinese children spend a whole year’s morning reciting multiplication table, which I could never imagine how terrible it would be if recite it in English. Therefore, every grade 3 children would shout out simple multiplication results immediately.

However, on the other hand, I believe most of the advanced mathematical concepts have nothing to do with numbers in a language.

> the problem of English is not ..teens, but the pronounciation of numbers. Chinese or Korean has one simple syllable for each number

I once found when mentally tallying points while grading papers, instead of saying "and a half" in my head, if I say the Chinese "ban" instead (e.g. "four-ban", "seventeen-ban") it becomes far easier to do.

Hearing numbers in German gives me always a pause. "Dreiundzwanzig" (23) - wait, let me swap the digits in my head so it would make sense for me and "Zwan" is not even "Zwei".

Also "Dreizehn" and "Dreißig" - ok, the first one is actually 13 and the last one is 30.

So there is a large research field awaiting. /s

I agree that it does sound strange to an English ear, but that's just because we're used to our system. In English, after 11 and 12, we have three-ten, four-ten, etc. Just aged for seasoning.

If you had read the article, you would have seen that that's what it's about.

Do you have a link to the research article. The press article has too much hand waving and very little details. The most interesting paragraphs is:

> The negative impact of English is apparent in a 2014 study comparing 59 English-speaking Canadian children from Ottawa, Canada, with 88 Turkish children from Istanbul, ranging in age from 3 to 41/2 years. The Turkish children had received less instruction in numbers and counting than the Canadians. Yet the Turkish children improved their counting skills more after practicing in the lab with a numbered board game, according to the study, co-written by Jo-Anne LeFevre, director of the Institute of Cognitive Science at Carleton University, Ottawa, Ontario. Turkish students learning to count in their native language "mastered it more quickly" than the children learning in English, Dr. LeFevre says.

If I read this literally, they are generalizing from a group of 59 students. Were all of them from one school in Canada and one school in Turkey? Did they repeat the experiment? Other languages?

"Mastered it more quickly" means that the number of turns to get a perfect score by the Turkish children was smaller than the number of turn to get a perfect score by the Canadian children? Or nobody get a perfect score and the Canadian children always got better scores than the Turkish children, but the improvement was bigger in the Turkish children than the Canadian children?

They have children between 3 an 4.5 y.o. (or 41/2 = 10.5?) How did they test all of them. The difference between 3 and 4.5 is very big (and with 10.5 is huge). I'm not sure that my daughter understood numbers as big as 10 when she was 3 years old.

> The Turkish children had received less instruction in numbers and counting than the Canadians. Yet the Turkish children improved their counting skills more after practicing in the lab with a numbered board game, according to the study...

To me, that almost sounds like there's a skill level ceiling that's as far as the numbered board game can take you, and the Turkish children improved more because they were further from the game's ceiling to begin with, because they had received less previous instruction in numbers and counting.

Malcolm Gladwell talks about this in his book Outliers, which may be a better explanation for you.

Here is an excerpt from that section: http://gladwell.com/outliers/rice-paddies-and-math-tests/

Maybe it's just me but I never struggled with number names nor math. I imagine even when naming numbers isn't perfectly sensical (always in some strict base 10 naming), this disadvantage resolves itself as students age. The more you experience mathematics the more you disengage with strict number naming and the more you engage with strict number writing (which is extremely sensical). Saying 'seventeen minus twelve' could give pause while mentally translating but seeing '17 - 12' really won't.

> Chinese has just nine number names, while English has more than two dozen unique number words.

Wut. Chinese has zero through nine, then ten, hundred, thousand, ten thousand, one hundred million. That's fifteen. Certainly better than English, still.

http://www.paulgraham.com/submarine.html

Submarine for game company featured in the story.

The story names the companies and games for sale, but doesn't name the research papers.

I don't know, even if this effect exists, its weight on the overall math performance is really weak, there many other factors at play. Just compare the PISA results [1] of countries that use the same language:

FRENCH

France: 495

Switzerland: 531

Luxembourg[2]: 490

SPANISH

Spain: 484

Chile: 423

Mexico: 413

Peru: 368

PORTUGUESE

Portugal: 487

Brazil: 391

This is just a personal anecdote, but I have studied maths in 3 different languages: Basque (with a counting system similar to the French), Spanish and English. And I can't really say that I felt more comfortable using one or the other, but the three use pretty similar systems though.

[1] http://www.oecd.org/pisa/keyfindings/pisa-2012-results-overv... [2] At secondary school, in general, until the 9th class every subject is in German, except mathematics and sciences (in French) https://en.wikipedia.org/wiki/Multilingualism_in_Luxembourg

PISA Maths exams results from 2012:

1. Shanghai (China) 2. Singapore 3. Hong Kong (China) 4. Taiwan 5. South Korea 6. Macao (China) 7. Japan 8. Liechtenstein 9. Switzerland 10. Netherlands 11. Estonia 12. Finland 13. Canada 14. Poland 15. Belgium 16. Germany 17. Vietnam 18. Austria 19. Australia 20. Ireland 21. Slovenia 22. Denmark 23. New Zealand 24. Czech Republic 25. France 26. United Kingdom 27. Iceland 28 Latvia 29. Luxembourg 30. Norway 31. Portugal 32. Italy 33. Spain 34. Russian Federation 35. Slovak Republic 36. United States

Asian dominance shows, but you have to bear in mind that China "cheats" by posting separate cities, if other countries did the same ignoring less wealthy regions - they would have better results too.

Also there are many English-speaking countries much higher than USA, and countries with much more complicated numerals (like Baltic or Slavic countries) that are higher than USA. And German numerals are more complicated than English ones too (125 is 100 + 5 + 20).

I wouldn't attribute the US education problems to the language.

This author draws conclusions like a BC astronomer; they are good given the area the article is restricting itself to, but this restriction is innately wrong. Numbers are weak because natural languages don't intrinsically use them and they are taught post-linguistic development to a fractured foundation, e.g. natural languages form like slang. Lets arbitrarily call 205-chi and 210-foe, how is one to find the name for 207; there is no interconnected phoneme structure that numbers and words can jointly draw upon. Yes, numbers are a naming system, but they become just another irrelevant layer when there is not a common core to draw upon. It's funny, with so many brilliant minds continuously rewriting coding languages no one has stopped to make human language less of a symantic ambiguity. Maybe Ithkuil is our only hope.

>It's funny, with so many brilliant minds continuously rewriting coding languages no one has stopped to make human language less of a symantic ambiguity.

People have tried. Check out "In The Land of Invented Languages" [0] if this interests you. Specific to your point, I would refer you to Lojban (a derivative of Loglan).

[0]http://www.amazon.com/Land-Invented-Languages-Adventures-Lin...

The simplest thing English speakers could do is get rid of the spoken words for 10 - 19 entirely. But what would you replace them with?

Tenty-one, tenty-two, ect? No. It sounds two much like twenty-one, twenty-two, and so on.

Replacing the tens with "onty" words makes sense because "on" is closer "one". There's no translation step needed in thinking "Onty-eight minus Onty-two" and coming out with six.

It would be a simple solution to the tens problem but hard with respect to getting people to adopt it.

It also solves only half the "problem", since each decade still has its own name.

Most decades have the same name. The only ones different are "twen" for "two", "thir" for "three", and "fif" for "five", which all sound very similar anyway so there's not really a problem.

If we follow existing practice of thir-teen, four-teen, etc. why don't we just use teen as the base? Teeny-one, teeny-two, teeny-three...

No link to the original research, or any input from linguists? Puhleeeze.

Japanese has different words for numbers based on what is counted. For each number 1-9 there's a Chinese-derived word and a native Japanese word. Numbers >10 always use Chinese pronounciations, except for numbers like 20 or 30 which might have a special word for them (if you're counting days or life years). Also number 4 is unlucky and has to be skipped when counting rooms or floors.

I love a lot of this new reasearch on the relationship between language and cognition, but.

What I hate about this article is the way that it plays on American parents' educational anxiety, implying that these very mild linguistic effects on cognition may impair Anglophone children relative to Asian children unless American parents invest in all these educational board games. The article specifically frames it as a problem of English linguistic deficiencies, which is incredibly naive. With math, children are essentially learning a new and unfamiliar symbolic language.

All of these correlations between language and cognition are interesting, but if you're interested in Sapir-Whorf, you're still looking at overall effects that are weak and small, overall. This correlation between linguistic/mathematical behavior may be stronger in a language like Piraha, but Daniel Everett would argue that what's going on there is less linguistic determinism and more of a cultural bias against the abstract or non-concrete. http://www.slate.com/blogs/lexicon_valley/2013/10/16/piraha_...

And look here at this paragraph in the OP: "The U.S.-Asian math-achievement gap—a sensitive and much-studied topic—has more complicated roots than language. Chinese teachers typically spend more time explaining math concepts and getting students involved in working on difficult problems. In the home, Chinese parents tend to spend more time teaching arithmetic facts and games and using numbers in daily life, says a 2010 study in the Review of Educational Research by researchers at the Hong Kong Institute of Education and the University of Hong Kong." Yet overall outcomes are taken as evidence of English's "linguistic disadvantages".

East Asian languages DO handle base-10 math in a simple, direct way that is easy to learn. This is especially true in Chinese, but begins to fall apart in Japanese. Japanese uses simpler words for higher numbers: ni-juu-hachi (two-ten-eight) for 28, for example. But Japanese also has two series of words for numbers under 10: 1 may be ichi or hitotsu, 2 may be ni or futatsu. These combine with a large series of counter words that makes counting in Japanese much more complex. Might this impact Japanese children's acquisition of numeracy? Plausibly, but it's easier to just elide over the differences in Asian languages handling of numbers.

There are just a plethora of linguistic differences between Asian languages and English, that necessarily impact education. Take the writing system, where Japanese has probably the most complex written system in the world, that takes over a decade to read proficiently even for native speakers. English has a terrible orthography, whose irregularities and inconsistencies make it difficult for children to learn and exacerbate economic inequality and linguistic privilege.

We really could spend a lot less classroom time on learning reading and writing the differences between plough, dough, and cough and more time on STEM if we improved the English orthography, but I don't exactly see the WSJ reporters calling for spelling reform =/ http://www.theatlantic.com/education/archive/2015/02/how-the...

Wen u sa u wont too improov thee Englush orthografee doo u meen speling foneticulee? Beecuz ay think that wood bee silee.

Instead, why not teach the English language properly to students? You know, where the words came from, what changes they suffered through the years... That's how I learned it as a non-native speaker and I can tell you I had a blast doing it, not to mention that now I literally can't forget how most words are spelled (I remember vaguely the derivation from Latin/Greek/Middle English/Anglo Saxon).

In other words, folks everywhere through the times were able to pick up this beautiful language, but all of a sudden we are having problems teaching it to kids these days? Is it the fault of the language, or the educators?

Mostly I think alternate spellings should be tolerated, and there should be fewer irregularities. I like a lot of Masha Bell's suggestions to simply make English spelling less irregular. http://phonetic-blog.blogspot.com/2010/09/luv-u-giv.html

Keep a silent <e> at the end of a word if it makes a vowel "say its name", but drop it otherwise. Then we could make a spelling distinction between a political moderat who wants to moderate political discourse, for example. Take off the inkhorn <p> at the beginning of "ptarmigan", spell philosophy with an <f> at the beginning like they do in most of the Romance languages. The only reason night has a <gh> in the middle is because a thousand years ago it was pronounced with a velar fricative in the middle that standard English hasn't used for centuries. Why not just rite nite rite?

Etymology and historical sound change is fascinating, but it's an irregularity that makes literacy harder for grade schoolers. The spelling is not the language. And is it really the case that it's more important to spend classroom time teaching children centuries-obsolete spellings rather than deep reading skills or math?

Do you really think preschoolers/kindergartners are going to get anything out of that? I certainly wouldn't've cared.

I don't believe that kids should be in school before 7 or 8 years of age.

Semi-relatedly, I've wondered why coders haven't adopted something like the tonal system for hexadecimal. http://panglott.blogspot.com/2010/01/nystroms-hexadecimal-nu...

Obligatory fleventy-five reference: https://panglott.wordpress.com/2015/01/27/fleventy-five/

My oldest son found math extremely hard to learn. Parts of this article mirror comments he has made to me about his lack of understanding in k-4 and how poorly things were explained. He was so desperate to learn math, he lied to me and his teacher both for an entire year in order to sneak his math work home and have me explain it to him.

> he lied to me and his teacher both for an entire year in order to sneak his math work home and have me explain it to him

sounds pretty smart

Isn't this a form of the Sapir-Whorf hypothesis ? https://en.wikipedia.org/wiki/Linguistic_relativity

Isn't that ironic (the title that is) as there is I am fairly sure more than one math :-)

sounds more like an excuse.

Reply on news.ycombinator.com