The absence of a "Critical Period"

It is sometimes believed that there is a critical period – perhaps at the onset of puberty – after which it is biologically impossible to acquire truly native-like capacity in a second or foreign language. This essay points out some essential questions that have been neglected or little noticed in the research to date and indicates why the evidence would lead us to believe that no such critical period exists. A research project is also proposed that would, eventually, give us greater insight into this very difficult field of study. According to the author, there is no critical period after which it is not possible to acquire a native-like capacity in a second or foreign language.

A concept that appears regularly in popular form and is discussed from time to time by specialists in language pedagogy, is the concept of a critical period for language learning. Everyone accepts that adults can and do learn languages. However, a great meany people – including a number of specialists in psychology and/or in language teaching and learning -- believe that after the onset of puberty it becomes biologically impossible for any individual to acquire "native-like" competence in a second or foreign language. (See, for example, Scovel, 1988). In recent years, there has been an increase in the "complexity" of the types of competencies tested in this area: phonology, syntax, suprasegementals, rhythm, intonation and perception of each of the above. Nonetheless, there still remains a good deal to be done to refine tests in this area.

It is very important in this debate to separate out the concept of the acquisition of a "first" language very late in life from the concept of second-language acquisition. When, for some reason, a child is not raised in a normal community and is therefore never exposed to language until of 7, 8 or even 10 years of age, it is probably not possible to separate out the concept of "trauma" or "psychological deprivation" from the concept of "language acquisition". It is therefore difficult or even impossible to decide whether any long-term negative results have occurred in each case as the result of a biological impossibility or as a result of emotional and social deprivation. These cases do not form part of this particular debate about a critical period.

One reason for different results in language learning when we compare adults and children is that adults typically do not want to give up their accent. They very much prefer, on the contrary, to maintain the "foreign-sounding accent" that identifies them as coming from "their orginal culture". They prefer, in this way, to show constant pride in the culture in which they grew up.

When we draw on our own experience and observe immigrants to Canada or when we look at the literature on this subject, our observations as well as the results of a number of studies seem to indicate that there is always a language deficit if the person begins speaking a second language after puberty. There was, for example, a study of Italian immigrants arriving in New York in which the conclusion was that "everybody" had an accent but that the accent was progressively more pronounced depending on the age of arrival in the city (Oyama 1973). A study in the sixties showed that those who originally spoke Spanish never acquired a "perfect American accent" if they started their learning as an adult.

It can and MUST be noted in the neither study that two underlying questions were not asked:

"Did the participants in the study TRULY WANT to acquire a perfect accent in speaking English?"

and

"Were the participants willing to make a strong effort in order to 'sound' like Native Speakers of English?"

It is in fact quite certain that the participants, had they been asked, would have said that they did not even want to lose their Italian and/or Spanish accent. In the case of immigrants to New York, the Italian community into which they moved would have been highly structured. The concept of the "Pater familias" was alive and well as was the very strong concept of the extended family. Not only authority/power but also wealth tended to be concentrated in the hands of the older member or members of the family -- the "Godfather". This person was very strongly identified as "Italian" by his habits, his actions and, above all, by his accent. The underlying goal of younger people would be to "become like that powerful person" -- and HE had a very strong Italian accen. The fact that one had an accent in the community would be seen as very positive and not negative. Since there was no incentive to acquire a "perfect American accent", it is not reasonable to argue that these people were "unable" to acquire such an accent.

In the Spanish case, the facts are the same. Two questions were not asked:

i). "Do you TRULY WANT to acquire a perfect accent in speaking English ?

ii). Have you ever truly wanted to sound EXACTLY like an English-speaking American ?

If those questions had been asked, I would predict that the answer in both cases would have been a resounding "No. Of course not!"

Having worked for a long time in accent remediation for people who wish to speak French and/or English, I have frequently offered my services, free of charge, to people who might want to no longer have an accent. A typical case was a friend of mine from Russia (age 40). He spoke English reasonably well but with a marked accent. When I offered to give him a "perfect" English accent, free of charge and with less than 20 hours of work on his part, he thought about it for a few days and then politely declined. He made the following comment which I have heard many times since then:

And THAT is the crux of the matter. We continue to test people to see if they have acquired a perfect accent when THAT is not what they wanted to do in the first place. It is not logical to claim that human beings did not acquire a perfect accent because of some "critical period impediment" when they did not want to acquire that accent in the first place.

2). The "Packsack Problem".

This argument can best be illustrated by a story.

Once upon a time there were people who live in the mountains in very dangerous circumstances. When the children were very young, the society attached a "front-pack" to them in such a way that it could not easily be removed. In this front-pack were placed safety items like small axes and crampons that could be used for climbing, battery-packs that could be used for producing heat and so on. Because of it, an important number of the Mountain people had, through the years, avoided losing their lives. When each child reached puberty in this society, more items were added to the pack so that it eventually weighed about 80 pounds. For cultural and social reasons, it was never removed and it was considered very impolite to refer to it in any way. All members of the society were secretly very proud of the ubiquitous front-pack. Even at night and in bathing, it was implied that the pack was never removed -- but people were careful to never verify this very intimate matter and to scrupulously avoid any mention of the front-pack. In the valleys and on the plains however, people had no use for such a survival pack.

One day, as part of an intellectual study, the researchers in the valley decided to compare the abilities of people in the valley with those of the Mountain People. In particular they decided to use a "Backpack Test" and see if the Mountain People could be as agile as the "Plains-People" when both of them had to carry a pack on their back that weighed about 100 pounds. After a great many tests, it was clear that those people who had always lived in the plains consistently showed greater agility in jumping, running and climbing when they carried the test backpack of 100 pounds. It was therefore determined by these tests that living in the Mountains had affected the genetic makeup of the Mountain People and that they were "biologically unable" to equal the Plains People in carrying backpacks. It did not occur to any of the researchers to notice that the Mountain People always had a pack of 80 pounds on the front in addition to the pack that was put on their backs during the Backpack Test.

In our tests of "second-language acquisition capacity", we often forget about the "hidden front-pack" -- the extra burden that is caused by the presence of the first language. As in the story above, the person who has acquired a second language has TWO PACKS to deal with, and we have to account for this fact in our tests and in our results. Indeed, in the language field, the problem is even worse than in the Backpack tests of the parable above. Not only does the person being tested have TWO languages, these two language occupy the SAME space in the brain and constantly interfere with each other. (See Nicol 2000). It has been noted that the bilingual brain seems to handle a second language differently than does the unilingual brain its one language. Any other finding would be very surprising. Obviously, the brain had to acquire two ways of doing the same thing and somehow keep them from interfering too much with each other. The original set of nerve pathways are already taken -- a new and different set has to be set up for the second language while still maintaining the first set. It is not logical to expect exactly the same mental structures to exist in a unilingual and a bilingual brain. This does not mean, however, that a "difference" in processing implies that there is somehow a "deficit".

Any conclusions drawn from comparing people who carry only one Packsck -- i.e. one language, as opposed to those who carry TWO (those who are fluent in two languages) require a great deal of caution. The fact that some of these people, despite their handicap, are better than or equal to a very large number of "monolingual subjects" can perhaps be taken as a demonstration that the concept of a critical period for language learning and/or acquisition is not valid.

3). The Case of a Person Who Gave Up Her Front-pack – i.e. who acquired a second language and refused from then on to speak or even understand her mother tongue. (A Research Project Proposal).

The question could be asked: Would it be possible some day to actually make a valid comparison of "monolingual speakers" and "those who had acquired a second language as adults". The answer is a tentative yes. What we need are subjects who learned a second or foreign language and who chose, thereafter, to never again speak or understand their first language. They would be very strongly motivated to acquire a way of speaking the second language that did NOT allow other people to perceive that they originally belonged to a different language community. Could that exist? Possibly, as it did in the following case. And this story is true.

When I went to Grenoble to learn French in 1960, my roommate was from Paris. He told me that his mother was originally from England and that she had met and married his father in her early twenties. At a later date I went with him to Paris to visit his parents. At one point in the conversation, I said something to his mother in English. She replied in French and said, very politely, that she did not understand or speak English at all. During my time at the house, I noted that my friend's mother spoke impeccable French and that she even had accentual and vocabulary items that identified her clearly as having been raised, not in Southern England, but rather in the very street and "quartier" of Paris in which she lived. Later I asked my friend why he had said that his mother was originally from England. He said that, after the war, his mother had gone to Paris on a trip -- she was about 23 years old and, as far as he knew, had only a rudimentary knowledge of the French language. She had gone through her teen years at a time when travel to France would have been unthinkable. Her parents and her family were very "anti-French" so it is most improbable that she every studied French as a young person. In Paris, she met and fell in love with a young French person. He was Catholic, she was Protestant and their backgrounds were very different -- but they decided to get married anyway. When she went back to England to introduce to her family the man she loved very much, her family totally and absolutely rejected him and treated him literally like a lump of dirt. She was so deeply hurt and totally furious that she resolved never to use a word of English again and never, under any circumstances, to be identified with people who spoke English. It is my strong impression at the time that, even though she had passed the age of puberty, she was absoltely able to do as she desired. She had no accent whatsoever in French and she was never perceived as having come from any other community.

What we need is a group of people like "Anne" who can be studied for their relative competence in the second language and who have chosen, not only to abandon their first languages, but also to make every possible effort to avoid "sounding" like a person who came from the cultural community represented by their first languages.

When and if we find a significant group of people who have been in exactly the same circumstances as "Anne" and test them, I predict that there would be no difference between their language capacity and that of very similar native speakers. In the comparisons, one would have to control for intelligence and access to language factors, but there is no doubt whatsoever in my mind that there would be no evidence of a Critical Period.

A person I know very well, "Louise", began at that age of 22, to become "totally fluent" in French. She eventually completed a PhD degree in France, got married to a French national and became "more competent" in many areas of the French Language than native speakers. In a French-speaking community in France, everybody would, for a time, take her to be a native French-speaking person. However, even for her, there was always a tiny problem -- she did not have any trace of a "regional" accent in the language.

When we listen to a native speaker of a language we expect that person to have something in his or her accent that gives us clues as to the area in which he/she was raised and went to school. When we hear a Texas drawl or a Newfoundland phoneme we are inclined to believe that the person with that "accent" is a native speaker of English. The same is true in French. We expect to hear a Québec accent or a slight Parisian accent or some type of accent from Brittany or from Normandy or from Marseille. People are expected to have "come from somewhere". The adult learner of a second or foreign language, however, almost never acquires a regional accent. That fact alone can be sufficient for the native speaker to say" "He / she is not a native speaker". This does not mean that the learner has not acquired a native level of competence or that there is any deficit in his language competence; it merely means that he/she has not acquired any trace of regional accent. My friend's mother, "Anne", in Paris had understood that. Since she absolutely did not want to be perceived as "non-French" she took the trouble to acquire ways of speaking that people identified with a particular street or "quartier' in Paris.

When we ask a group of people to say whether such and such a participant 'sounds like a native speaker' it could well be that what is being identified is the lack of a regional accent and NOT a reduced level of linguistic competence. In these types of tests, it is precisely the non-specified elements of the evaluation that we are getting at: "Do you perceive that person to be a native speaker of Language X?" Since the complex set of "reasons" for that evaluation are not and cannot be specified, it can well be that a some of the factors that enter into the evaluation has nothing whatsoever to do with the "language competence" of the adult learner. It can be simply the lack of a regional accent is the problem and/or one or more subtle elements that have simply not been available to the adult learner because of the different learning environment in which the adult learner must live.

It is interesting to note that in the Asher test in 1969, two or more of the native speakers of English were identified as non-native. It is possible that even non-natives who have travelled a great deal and been subjected to many different accents can be perceived as "non-native". That is precisely what happens to "Louise" (above) from time to time. When she finished teaching a French class in the sixties, she tells me that one of the students came up and complimented her on the quality of her English, not realizing that English was, in fact, her mother tongue. She had lived and studied in Ireland and taken pains to acquire the accent of Dublin while she was there. the result of this fact coupled with influences from the French language was that she was perceived as a person whose native language was not English. This perception had nothing whatsoever to do with any "lack of competence" in the English language.

When we carry out our tests of "competence" it is essential that these and other considerations play a role. A "difference" in oral language production is not necessarily a "deficit". To date researchers have not clearly identified or come to grips with the many difficulties in this area -- "motivation", the "Frontpack Problem" and the Regional accent problem, among others. It is clear, as Thomas Scovel pointed out, that people tend to "underestimate" the degree to which they have an accent. (Scovel 1988). My friend "Louise", above, is absolutely convinced, as am I, that adults can and do acquire native-like capacity if they desire to do so. They do not learn in exactly the same way as children do -- since they already have a language in place with its complete set of habits, grammatical rules, structures and constraints.

All of the evidence I have seen points to two conclusions:

What we need is to use a set of very carefully designed tests and apply them to a group of people like "Anne" in Paris -- i.e. people who have been very highly motivated to use only the second language acquired as an adult, and who have access to a community which is prepared and able to identify and help correct problems as they occur (just as parents and peers do for children). We then need to examine differences that could still exist and determine what their effects might be. Language tests must evaluate both production and discrimination tests in a wide range of language competencies: phonology, suprasegmentals, syntax, and vocabulary. If and when the appropriate group of people like "Annie" have been found, adequate tests will be very difficult to prepare and administer. It is a truism to say that human language competence is remarkably complex. Even if a biological constraint exists, which I seriously doubt, it will be no easy task to demonstrate the areas of language competence that may play a role, and the effects that one might observe in the subjects that are being tested.

The conclusion of this essay is that the evidence to date does not support the idea of a biologically determined critical period after which the acquisition of a second and/or foreign language cannot reach or surpass that of a truly comparable "first-language learner".

**** **** **** ****