There is so much talk these days about untargeted input. “Just speak the language with your students.” “You don’t have to pre-plan what language you will teach.” “They’ll get enough repetitions over the course of the year.

It’s interesting to me to think about the distinction between targeting defined as “planning in advance that you will include a word or phrase” and targeting as “repeating a certain word or phrase during input”. It’s particularly interesting in asking whether circling was used in demonstrations done at a recent conference that purported to promote “untargeted input”. (It was.)

The first definition of “targeting” really makes no difference whatsoever to anything. No one can see into your head. Students don’t have a video feed into your planning space to see what word you’re going to use tomorrow. Nor is there any evidence at all to prove (or even suggest, really) that using a word by design “disappoints” or “saddens” the students, or makes them feel that they’ve lost some precious and inviolate right of self-determination. The fact is that teachers who DO target in the sense of planning language to use also (obviously!) provide other language that is not pre-planned. There are two extremely obvious reasons for this: first, you can hardly spend an entire lesson simply repeating one word, and second, personalization (getting ideas for content from students) requires that the teacher include language that has not been preplanned, because just as the students have no video feed into the teacher’s planning space, the teacher doesn’t have a crystal ball to predict what random suggestions students might make during a class.

The second definition of “targeting” is the one that is far more germane to acquisition. If we define “targeting a word or phrase” as “providing more repetition of that word or phrase than would have occurred simply by having a normal conversation with a fluent speaker of that language about the same topic and information”, then anyone who does not target has hardly a hope of students acquiring that language. The precise number of repetitions a learner needs to acquire a word or phrase is not known, and indeed varies between individuals and even across days or topics or specific words and phrases for the same learner. But we know that in the vast majority of cases, one exposure to a word or phrase that is understood is not enough to provide the student with the means to output that language naturally and correctly at will, without conscious thought, in the future.

So to get acquisition in the artificially short time available in the classroom, we repeat bits of that language at artificially high rates. There is nothing to be ashamed of in that statement. The mission of the classroom teacher is to drive acquisition, and do it in the very artificial environment (limited time, high numbers, other restrictions) of the classroom. Saying that teachers should simply speak the language with their students, even if everything can be understood by the students, is to turn one’s back on the potential for optimizing the input to fit the situation.

Natural input works for natural acquisition in natural situations. It works for artificial situations, too. It just doesn’t work as well as optimized input — turbocharged natural input. Dense input. Input that provides many opportunities for the brain to make the match between the incoming language and the meaning, not just a few.

So I could not care less whether the specific words or phrases are selected before class or during class, nor by whom they are selected. That’s immaterial. What is crucially important is that acquisition requires repetition. CPR input — comprehended, personalized, repeated — is by far the most effective for a formal school environment. It doesn’t matter if you set your phrase three months in advance or focus on a random phrase your kids generate two minutes into class, as long as that phrase is repeated and incorporated into rich language input and presented in unpredictable ways.

So targeting as in “we don’t plan in advance” is really simply “stop planning in advance”. It has nothing to do with engineering input to benefit acquisition. Targeting as in “we repeat”, on the other hand, has everything to do with it.