On a noted foreign language teachers’ list, a recent discussion asked about the use of “crutches” by students. Interestingly, I think that CI teachers are much less opposed to crutches, simply because we do not have to fear that the use of a vocabulary list or “cheat sheet” will cause output to break down. Since our students are acquiring (not learning) the structure of the language, using a “crutch” is just a quick memory aid to bring up a lexical item to quickly drop into a slot in a sentence. It’s not like a Math student using a calculator — a device which applies a set of rules (also known in language terms as “grammar”!) to some items to give a result.

Students who have acquired the language have the calculator in their brains. It operates without them having to think about it, churning out the results of the application of sets of rules in the correct order (“Grammar”).  Students who are learning a language through memorization and application of rules rather than acquisition don’t have the calculator in their heads. They have to rely on a stack of punchcards giving them instructions (Okay, Suzy, we want to say “they speak”. That’s an “ellos” form on our chart. Take off the -ar ending. Now, replace it with an -an.)  The CI-taught student skips these steps as the rule for how to form this has been internalized, and requires no thought at all.
People have commented, “But students need to be able to judge whether or not the “crutches” are working properly! And who will make those “crutches” if no one can do the functions without the use of the “crutch”?  I am impressed by my own sense of restraint in not responding, “Well, fortunately there are more and more CI-taught students who have acquired the languages and are competent to judge and to make crutches for the traditionally-taught students”…
But it’s a fair question. How can a person judge a crutch (such as an on-line translator) if they do not already have the language acquired? What basic knowledge is needed for a person to be able to evaluate such tools?
Acquiring the basic structure of the language and a limited but high-frequency vocabulary is the “basic knowledge’ corresponding to the ability to tell North from South to see if your GPS is acting up, or know one’s multiplication tables to know that your calculator is spitting out garbage results. It can be applied to other “crutch” inventions to tell whether or not they are working.
So, to “crutches” for language users. For people who have acquired some language, these are usually lists of unfamiliar words — a memory aid. Even a technological memory aid isn’t the same as a calculator that performs a calculation. The calculator is providing the rule-governed structure, while the memory aid is simply bringing lexical items up when needed. When the lexical item has been activated enough times by being brought up this way, it may well be acquired.
Many of us CI-based types strongly discourage editing (pausing to consciously apply a rule in output) in beginners, because we want students to acquire the structure and be able to use it unconsciously. They would not have all the vocabulary they need,  even if they had acquired all the structures (who ever does? 😉 ) since structures are effectively a finite set but vocabulary is, in practical terms, infinite.  Using a “crutch” is editing in one of its most extreme forms, in one sense, so the underlying foundation of language has to be absolutely firm to begin with to allow the student to use a “crutch” and still appear competent. If the student has not internalized the structure of the language, but has to rely on conjugating in his head (let’s see, take off the -ar, and add an -an to the end) that rule will also be a candidate for a “crutch”, because it has to be applied consciously and output will be paused (even for a very short time) to allow its application. This sort of basic structure cannot be output based on rules if the speaker is to be fluent enough to effectively utilize a “crutch” to shore up only the low-frequency items — the purpose of a benign “crutch”.
Certainly the interpreter using a term manager in the booth has to be able to comprehend and speak correctly at high speeds in the first place. The term manager is only an aid to help insert or decode lower-frequency words in crisis situations (they are needed instantaneously and there is no way to anticipate what the speaker might say next). It also requires the diversion of attention from the ordinary tasks of comprehension and output, which means those functions have to be even more automatic and unthinking, because otherwise there will be a breakdown if attention resources run out.
This is an extreme example, and I realize that relatively few people become interpreters, but maybe in one way the language load on a person doing SI in an unfamiliar field sort of models the language load on a lower-level learner who has acquired the basics but has not expanded his vocabulary widely enough to deal with a particular situation. It’s still a means of providing on-demand lower-frequency items. It would not be, in either case, a means of conjugating verbs or providing grammatical agreement or whatever. Those are things that MUST be automatic (i.e., acquired) for any crutch to work.
CI teachers don’t fear crutches. They know that their students won’t be using them to limp around a room; CI-taught students use crutches only as aids to knock boxes off the highest shelves in the grocery store. And they know how to catch them before they hit the floor, too.