Pairwork. Mmm.  Soooo good (say the administrators and observers). Students working together, taking responsibility for their learning, in a student-centered environment. (Whether or not these are really true is up for grabs. But we digress.) It’s like that cup of coffee your mouth is watering for right after I typed the words “cup of coffee”. Delicious. It might keep us up all night, we know caffeine isn’t good for us, but…soooo hard to resist!

In a discussion about what types of pairwork would fit with TPRS/CI teaching, a suggestion was made to make pairwork “okay”:

Use 3 students! 50% more likely that someone will say it right.

Seems logical, right? If people are objecting to the chance that students will make mistakes outputting language that is not yet acquired (because why would we be doing pairwork with language that has already been acquired, thus wasting input time in class) then we should simply minimize the damage by diluting it with more kids, right?

Maybe. But the other side of this coin is, using 3 students means that more students hear the wrong forms. And unless you happen to have a native speaker in that triangle, who is confident enough to say, “No, that’s not how you say it,” the best we can hope for is either a vote (and if you reference even “professional” translator forums and look at the terminology discussions, you’ll see that numbers of votes do not always equal correctness) or having one more kid hear the wrong forms than would have in the first place, with only two in the pair. Even if the students eventually “decide” on the right form, they’ve repeated and agonized over the wrong form to get there — and that, my friends, is Linguistics, not using the target language.

Why is this important? Who cares if they hear a few wrong sentences? Well, if we believe that language is acquired through comprehensible input, and that the input should be as correct as possible, so as not to have students acquire wrong language, then there is no point in the wide world to setting up any situation where there is pretty good chance students will be unable to output a very high percentage (if not all) of the language correctly when they’re talking to each other on their own. Student output is a kind of input. Not good input, but still input. I call it “throughput” — often mistakes spreading from one student to the next, like those little creepy-crawlies that jump from head to head. They’re not easy to get rid of. Mom has to go and buy a comb and maybe cut off all your hair. And all that time and expense and hair-pulling could have been used in fixing your originally-pretty hair into a nice style, instead of trying to repair something that wouldn’t have been broken had you not been put with the kid with the little bugs walking up and down the part in the hair to do that pairwork exercise on celebrity birth dates in French.

But what about little kids? They hear lots of wrong language, don’t they, like at daycare when they talk to each other?

And mistakes aren’t the end of the world. In a room full of 2 and 3 year olds speaking L1, there are lots of language mistakes going on but no one cares. They are experiencing the language and communicating and hopefully having fun.

No one cares because that is natural immersion. TPRS and CI teaching are not natural immersion situations. The 3 year olds speaking L1 at daycare have had 10,000 hours of input in that language already. The couple of wrong things they will hear from their agemates will be more than offset by correct input from their native-speaking or fluent parents, caretakers, elder siblings, and so on. That doesn’t usually happen in the lives of TPRS students. Sure, some of them may have speakers of their target language in their homes or community, but the vast majority do not. The only input they get in the language we are teaching is during class. And that means about 100 hours a year (in a typical middle- or high-school program), not 10,000 or even 1,000.

Clearly, the price of pairwork, like a Starbuck’s super-duper latte, is just too high.

But what if you have a AdMeddleStrator who just has to dictate what you do in your classroom (despite not being a) a teacher for the past 10 years, b) a foreign language teacher in the first place or c) a person with a clue about language acquisition)?

We need to decaffeinate that pairwork. And we do that by removing not the caffeine, but the unacquired language.

If you must do pairwork (and by “must” I mean there is an authoritative outside force that is demanding it), then do it — but only using language that you are completely certain has already been acquired.

In the worst of pinches, that’s the shared fluent language (usually English, in the US). A snappy “retell the story in pairs” in English may satisfy your admin’s thirst for a big cuppa Joe-Outputs-in-Spanish-Class.

Next up, design the output the students will be doing to be language they have already internalized to an enormous extent. Jump back a couple of stories and do something with that language. Anyone who’s taught legacy methods can come up with a million and six pairwork activities, or the Internet is full of them. Just don’t include anything that hasn’t reasonably probably been acquired by everyone already.

Or in a language and at a time when they can reliably read phonetics (for Chinese, this will be a good ways into the year, since we are teaching them to recognize, not necessarily sound out, Pinyin at the beginning), have them use structure you are pretty sure is very solid, and just drop a couple of nouns out there. They can easily (and correctly) “slot” those nouns into the places where they go in their very-solidly-acquired structures using the Super Seven or other things from early in the year.

The administrator can lead you to pairwork, but he can’t dictate the content (in the vast majority of cases). So have some go-to pairwork ready to go if you have any reason to believe your building might have this little problem. It’s cheaper than the five or six lattes it would take you to unburden yourself to your best friend if you get caught without it. With or without caffeine.