By the method of practice he will merely reproduce by imitation the sentence _elle est allée_, which sentence he will have had occasions of learning or of reading. In both cases there is the possibility of error; the theory may be imperfectly known and one or more links in the chain of reasoning may be weak, or the sentence may have been badly memorized.
First attempts at imitation are sometimes inaccurate; our initial attempts at reproducing are occasionally unsuccessful; we wish to produce a foreign sound we have just heard, but utter a native sound instead; we wish to produce a foreign sentence, but construct it wrongly; we wish to express a certain thought, and fail to hit upon the right word. For this reason the method of practice is often termed _the method of trial and error_. We are told that some of our efforts will be successful and others unsuccessful, and that in the course of practice we shall gradually eliminate the unsuccessful ones. We are even told that it is only by making mistakes that we learn not to make them. Although we may admit a modicum of truth in this somewhat hyperbolical dictum, we would suggest that it is liable to misinterpretation. Nor is the term _trial and error_ an ideal one from the language-teacher's point of view.
In both cases there is an implication that all successful attempts must necessarily be preceded by unsuccessful ones, which is not only untrue but unsound pedagogy. The literal interpretation of the term and doctrine may induce a sort of fatalistic attitude, and the principle of _accuracy_ (which we shall deal with later) will suffer in consequence. It may provide teacher and student with an easy pretext to condone careless work. The student may say, "Since it appears that error must precede perfection, I will not unduly strive towards accuracy." The teacher may say, "Since psychologists tell us that error is inevitable, I will allow my pupil to do inaccurate work."
It may be true that some forms of inaccuracy in certain conditions tend to eliminate themselves in the course of time and practice, but it is certainly true that errors also tend to become habitual, and no psychologist has ever maintained that the forming of _bad_ habits is a necessary step towards the acquisition of good ones. We would lay stress on this point, for there seems to be a real danger in the misapplication of such terms as 'trial and error,' 'the selection of the successful and the rejection of the unsuccessful efforts,' 'practice makes perfect,' etc. Misunderstanding on this point has caused many teachers to encourage, and many students to acquire, pidgin-speech, and to consider it as the inevitable or even indispensable prelude to normal speech.