Post written by François Grosjean.
One of the most intriguing questions relating to bilingualism is whether a language acquired in very early childhood, and then forgotten, is in fact still present. In a first post on this topic, I mentioned the well-known example of little Stephen who had acquired Garo in India during his first years but who later forgot it when his parents returned to the United States. His father, anthropologist Robbins Burling, reported on the case and asked whether hidden deep in his unconscious he had not retained a remnant of his lost language which could be reawakened if he came into contact with the language again (see here).
In a second post, I reviewed a number of studies that seem to give contradictory results. Research conducted in France with adults who had been born in Korea, and had been adopted as babies by French families, seemed to show that the adoptee's first language, Korean, had indeed been lost.
However, a study undertaken in Sweden seemed to show that if reexposure to the first language takes place over a certain period of time and is intensive, then remnants of a seemingly lost language are more likely to be retrieved. Basically, a first language is not totally lost (see here).
This fascinating topic will certainly continue to be studied for many years to come. Those who believe that early first language representations are not overwritten, even though the language appears to be lost, can now refer to the results of a new, and very compelling, study. Lara Pierce and her coauthors, all based at McGill University and the Université de Montréal in Canada, sought to find neural evidence of a first language in participants who had forgotten it and had no conscious recollection of having known it.
They tested three groups of young people whose mean age was 13 years old. One group was made up of adoptees who had come from China to Quebec when they were one year old. Since then they had grown up speaking French only and had no longer been exposed to Chinese. The second group was composed of Chinese / French bilinguals who had continued speaking Chinese at home and had used French everywhere else. And the third group was made up of monolingual French speakers who knew no Chinese whatsoever.
The brain activation of the participants was examined using functional magnetic resonance imaging (fMRI) while they were doing a Chinese lexical tone discrimination task. In this task, they were asked to listen to pairs of phrases containing three syllables that were either pseudowords (e.g. da-shao-fa) or nonspeech hummed versions of the same syllables. Both pseudowords and their hummed versions contained tone information. The two elements of the pair were either identical, or the final syllable varied on tonal information only. Participants were asked to respond with a button press indicating whether the final syllable was the same or different in the pair.
What is interesting in this study is that it used tone information which is processed differently depending on whether one knows a tonal language or not. Thus, the Mandarin word "ma" can mean "mother", "hemp", "horse" or "scold" to listeners of Chinese depending on the tone used whereas the meaning remains the same ("mine") for listeners of French despite the change in tone. In the case of French listeners, it is the right hemisphere's frontal and temporal regions that process acoustic frequency information that are activated when listening to lexical tones, whereas it is the left hemisphere language regions that are activated in Chinese listeners since tones are linguistically relevant in their language.
How did the three groups of participants react to the stimuli they heard? In the monolingual French speakers, only the right hemisphere was activated, more precisely the right temporal regions which process complex, but nonlinguistic, auditory signals. In the bilingual participants, who knew and used both Chinese and French, the largest peaks of activation were in the left hemisphere (left temporal regions) showing thereby that they were using their linguistic knowledge of tones to process what they heard. What about the adoptees who had functionally "lost" their Chinese and had no conscious recollection of it? Their neural patterns matched those of the bilinguals showing thereby that they maintained their early neural representations over time even though they had received no Chinese language for some twelve years! The authors concluded that these representations acquired early in life are indeed present and can be revealed if the right procedure is used.
Of course, as more research is conducted in this domain, the story will become more complicated. Others factors will be taken into account and some may be found to have some importance. For example, the Canadian researchers found that the children who were adopted at later ages showed more activation in a part of the left hemisphere, the planum temporale, than those who were adopted earlier. So the amount of input that a language received before being forgotten will probably play a role. Other questions will be asked such as what exactly was acquired linguistically before attrition occurred, how sensitive is the research task used, and is perception restored more quickly than production? To these should be added factors that concern the type of reexposure needed to "reawaken" a language and how long it should last.
This said, the answer to the question in the title of this post, "Can a first language be totally forgotten?", now seems to be clearer than ever before: No, not totally!
Lara J. Pierce, Denise Klein, Jen-Kai Chen, Audrey Delcenserie & Fred Genesee (2014). Mapping the unconscious maintenance of a lost first language. Proceedings of the National Academy of Science (PNAS), doi: 10.1073/pnas.1409411111
"Life as a bilingual" posts by content area.
François Grosjean's website.