one 二 three 四 five 六 seven 八 nine 十...

June 25, 2012 8:13 AM

Why is it so hard to quickly count from 1 to 100 alternating between 2 different languages on each digit? Any cognitive scientists want to explain this?

(Try it yourself) You speak two languages fluently. You can count in both languages forwards and backwards super fast. But when you have to switch between the two on each digit, it slows you down dramatically and you stumble along the way.

What is going on in our brains when this "switch" happens? And is there an actual word/term for this?

I would say this feeling is similar to those puzzles where you have to name the correct colour on a word but it tricks you by using a different colour (ex: blue text on the word "orange", which you may end up saying orange when you first see it and that would be incorrect)
 

posted by querty to Science & Nature (13 answers total)

I would guess that it's because when you count, you are reciting a memorized sequence, not actually associating each sound with a concept. Anything that forces you out of that sequence takes more time. (Not a cognitive scientist, but once embarrased herself in front of a hundred people and the Grandmaster of her karate style by totally flubbing "count backwards from ten" in Japanese.)
posted by restless_nomad at 8:17 AM on June 25, 2012 [10 favorites]

it might simply be interrupting a pattern. Humans are all about patterns, it helps us predict and assume things that might happen.

You know that two comes after one and three comes after two. But trying to say what is a completely different word from the one you normally associate in the pattern makes you grind to a halt. You'd have to memorize a completely new pattern with both languages mixed.
posted by royalsong at 8:18 AM on June 25, 2012

Yes, based on my experience just now when I tried it. It's like when I would make a mix tape of my favorite songs by XYZ-artist and listen to that over and over in the car. Then when I listened to the actual album, I would expect the next song to be the one on my mix tape.

I hereby dub this the Elemeno Problem.
posted by jeffamaphone at 8:25 AM on June 25, 2012 [8 favorites]

Yeah, I recite sounds when I do this, not numbers - not the thing the sounds represent. I could probably count to 20 in alternating English/French (native/non-native) if I practiced, the same way I could sing a choral piece in Latin or Italian or German with enough practice, even though I don't speak those languages.
posted by rtha at 8:33 AM on June 25, 2012

i'm not a cognitive scientist. i speak a bunch of languages. i think what you're referring to is 'code switching.' it's actually a whole new skill beyond simply learning a 2nd language- i.e. once you develop the ability to speak another language, it doesn't mean you can switch rapidly between them. that is a skill, that if you want to develop, is a skill in and of itself.
posted by saraindc at 8:40 AM on June 25, 2012 [1 favorite]

For the same reason that I can't recite my multiplication tables in English, only in Italian. I'm not doing math, I'm remembering a series of sounds.

Also, try saying any sentence alternating words in the two languages, you're not exactly going to do so smoothly.
posted by lydhre at 8:44 AM on June 25, 2012

I'm sort of a cognitive scientist. (Philosopher who works with them.) restless_nomad's answer is a pretty good rough description of the consensus on this sort of question. Something like the ordinal numbers we know in base 10 or the alphabet that we recite is really a sequence constructed out of features of numerous different cognitive domains - linguistic elements from speech production, associations with arithmetical processing - all wrapped up in an elaborate bit of procedural memory. As saraindc's answer also suggests, you store those procedures separately even if you're fluently bilingual. When you don't do so, you tend to see "bleeding" from one language into another, as when my Spanish speaking students in California would drop words from English into their Spanish sentences and vice versa. If you did start to meld those together more deeply and regularly, you'd actually find it harder to *just* do the numbers in English or *just* do them in Spanish.
posted by el_lupino at 8:49 AM on June 25, 2012

> those puzzles where you have to name the correct colour on a word but it tricks you by using a different colour (ex: blue text on the word "orange", which you may end up saying orange when you first see it and that would be incorrect)

This is called the Stroop effect. It's been widely studied in experimental psychology, including studies of bilingual individuals (the Wikipedia article links to one of these studies).

As an aside, I just tried to quickly count mixing two different counting systems, both in Korean, and utterly failed - Korean has two sets of number words, one which is native Korean, and the other using the Korean readings of the Chinese hanja numbers. I tried to count alternating the two and had to stop after 3 because I couldn't tell which word came next.)
posted by needled at 9:16 AM on June 25, 2012

I don't know. But here's some data: I started trying to do this in English (for odd numbers) and French (for even numbers). The first few numbers were hard but by 6 or so I was in a rhythm. But then I said

"eleven, douze, thirteen, quatorze, fifteen, seize, seventeen, dieciocho"

and decided to give up.
posted by madcaptenor at 9:39 AM on June 25, 2012

When I was a linguistics undergrad, one of my courses covered this. As I recall, the explanation was that numbers are not like other words—when you're using numbers, you're using a different part of your brain along with the language centers.

For the same reason, it is much harder to read numerals smoothly in a second language, while reading their names written out is no different than most reading in that language. For example, it is much harder for a non-native Spanish speaker to smoothly read this out loud:

Mi madre tiene 3 hermanas. 1 hermana vive en la avenida 72, y su número de teléfono es 353-1234. Una 2a hermana es de 27 años y tiene 4 hijos. La 3a hermana es muy aficionada a Douglas Adams, y ella utiliza el número 42 cada vez que puede.

than it is to read this:

Mi madre tiene tres hermanas. Una hermana vive en la avenida setenta y dos, y su número de teléfono es de tres cinco tres uno dos tres cuatro. Una segunda hermana es de veinte y siete años y tiene cuatro hijos. La tercera hermana es muy aficionada a Douglas Adams, y ella utiliza el número cuarenta y dos cada vez que puede.

I make no guarantees that my Spanish is any good at all.
posted by ocherdraco at 10:03 AM on June 25, 2012 [2 favorites]

Best answer: restless_nomad, el_lupino, and others are right - this is more about disrupting a memorized sequence than switching languages per se. It's kind of like trying to spell your name backwards. You know the sequence, but it takes effort to think about it. (Conversely, counting backwards is easy because we have lots of practice counting down to things.) Instead of just reciting a memorized chain of words/syllables (see also the memory term: "chunking"), you now have to THINK about the number, remember what language you're supposed to be using, then access the word for that number in your mental lexicon in that language.

I guess you could call this phenomenon code switching (sometimes called code mixing), but it's not really what linguists and psychologists mean when they talk about code-switching. Code-switching happens naturally for many bilinguals, and tends to be an effortless, often unconscious, process. It typically happens when you're speaking in language A, but can't find the right word/term for a particular concept, and so sub in a word or phrase from language B. Or, you might switch languages for entire utterances if you're speaking to one person vs. another in a conversation because you think it's easier/more appropriate for that listener (e.g., a bilingual teenager talking to her grandfather in language A, but her mother in language B, even though all 3 parties in the conversation know all 3 languages.)

What you're describing though, OP, isn't really the same thing. Here, you're forcing someone to break up a pattern in an unrealistic and regimented way. If you're reciting a chain of numbers to someone, chances are you'd stick to the same language throughout, because it's the same 'realm' of vocabulary knowledge and it's pretty unlikely that you would only know every 2nd number in language A. Further, your expectations of what is easier/more pleasing for the listener would be consistent throughout the task.

The Stroop Effect and the number-mixing task feel similar to you because both activities require you to inhibit a really strong and automatic behaviour (i.e., reading words and reciting numbers in sequence), and switch your attention/focus to something else (i.e., the colour of the font and thinking in another language). This inhibition is difficult. I suspect if you scanned someone's brain while they did these activities, you'd find a lot of frontal lobe activation for both tasks (the frontal lobe handles 'higher order' mental functions like inhibitory control, planning, working memory, switching from following one set of rules to another, etc. Collectively, these are called "executive functions".)

I don't know any specific name for your number-mixing phenomenon, but I would probably say that it requires inhibitory control and cognitive flexibility. Incidentally, there's good research showing that being bilingual protects against declines in executive-function abilities in old age, probably due in part to the fact that bilinguals get daily practice at inhibiting one language and selecting another. (Take a look at Ellen Bialystok's work, if you're interested in this.)
posted by miss_kitty_fantastico at 10:15 AM on June 25, 2012 [2 favorites]

*puts on educational psychology hat*

There's a number of theories that apply here. Behaviorism would say that rehearsing the sequence from 1-100 makes that task easy. Mixing languages creates a novel sequence that hasn't been reinforced.

One possibility from an information-processing perspective is that the compound words from 21-99 in English are individual cognitive units. "Thirty-two" is a single chunk. The exercise forces you to do more work to deconstruct that concept into two chunks, "thirty" and "two," and then hold one of them in working memory while you translate the other. You might also need to re-associate the two digits to make the connection to the next chunk, "thirty-three."

From a language-processing perspective, forcing the alternation of languages within compound words (rather than within an utterance) is likely a novel grammar. Grammatical rules allow us to unconsciously construct valid utterances within a language. Using the new grammar rather than one you're familiar with for constructing compound words for numbers demands more cognitive work.
posted by CBrachyrhynchos at 10:39 AM on June 25, 2012 [1 favorite]

I'm not a cognitive scientist, but I imagine my brain like a computer directory with information stored in folders with subfolders in them. When I want to access "Numbers" I automatically open the Numbers, English, Counting folder and begin. But when I want to alternate with Spanish, I have to consciously pull information from a different folder - a different source.

So basically when I want to alternate, I keep needing to Alt-Tab between the two folders to see what the next answer is. It causes a delay.
posted by tacodave at 2:24 PM on June 25, 2012