What happens when a language learned as a child is forgotten over time? Many adoptees and emigrants have no conscious memory of their native tongue, but a new study suggests at least some information remains in the brain. A team from the University of Bristol in England showed that English-speaking adults older than 40 who had spoken Hindi or Zulu as children were able to relearn subtle sound contrasts in these languages, but adults who had never spoken the languages could not—even though the childhood speakers had no explicit memory of the languages. Because memories are neuronal connections that get reinforced with regular access, the finding means that even connections that have not been reaccessed for decades do not disappear completely, as previous evidence had suggested.
This article was originally published with the title Once Learned, Never Forgotten.
See what we're tweeting about
9 Comments
Add CommentI could have told the reseraches this from my own and several others' experiences!
Reply | Report Abuse | Link to thisFunny! I could have told the researchers this based on my own experience and that of others I know
Reply | Report Abuse | Link to thisDoesn't post right away, right jjina??
Reply | Report Abuse | Link to thisI thought that I had read some report in, for example, in SciAm, many years ago that indicated that unused neural networks supporting language specific sounds were 'trimmed' very early, perhaps 18 months. This was the explanations, for example, that Japanese speakers had extreme difficulty with 'Rs' and 'Ls', etc., unless English had been learned very early.
Reply | Report Abuse | Link to thisIf my recollections are correct and the phoneme-trimming process does occur, it would quite reasonably follow that any neural connection supporting a language specific phoneme survived the initial trimming would still be available later in life for 'relearning'.
Perhaps of more importance, for example, might be if this process could be identified as a determining factor in the development of language skills in disadvantaged social groups. If so, it could follow that improved language instruction prior to the 'trimming process' could generally improve not only the speech of disadvantaged children but their early classroom learning experiences.
Back in the days after network TV but before cable TV, VCRs, video game and cell phone proliferation, when PBS broadcast Sesame Street to most very young children, there may have been some improvement in language skills among the disadvantaged that could be detectable in adults born from perhaps 1970 to maybe 1990.
So does it mean all memories, not specific to language faculty, are accessible somehow?
Reply | Report Abuse | Link to thisSo does it mean that all memories, not specific to language faculty, are accessible somehow?
Reply | Report Abuse | Link to thisSahar - n my opinion, the author is wrong to consider the ability to make and identify certain language phonemes despite not having done so for years as indication that memories of facts or events might also be retained indefinitely. I submit that these phoneme circuits are more akin to the ability to see colors: if one lived on a planet orbiting a red Sun, perhaps the ability to see other colors would be lost by about 18 months.
Reply | Report Abuse | Link to thisIt seems more likely that language phoneme neurons are firmware children are born with to aid in the development of language. Since they are specialized, unused connections can be eliminated once language is learned without affecting the ability to communicate in a native language. Having some special recognition firmware for the sounds not used in normal language usage may detract from optimal language skills.
That likely has no bearing on memory circuits. Functionally, memory learning has no early deadline: one does not lose the ability to form new memories at 18 months, for example. Certainly in the case of short term memory, the neural networks involved appear to be reused for new memories after some time period. Long term retention of memories seem to depend on whether or not the are transferred to long term memory by a (perhaps nightly) migration process that also clears short term memory.
There certainly may be conditions which make accessing long term memories more difficult (as seems to begin around 40 years of age), but this most likely involves deterioration of memory index connections used to relate memories to each other. In that sense memories may become inaccessible, effectively lost even though they are still retained. But that's just my personal opinion.
Dear jtdwyer,
Reply | Report Abuse | Link to thisFirst many thanks for your comprehensive answer to my question. Next, back to your first paragraph, after reading the original paper entitled "Preserved Implicit Knowledge of a Forgotten Childhood Language", I noticed the author doesn't mean phoneme circuits are kind of memory, on the other hand, the first sentence of the paper is totally written in agreement with your opinion about the critical period of language learning.
Regarding your note on memory circuits, I wonder whether the network involved in memorizing 'language' in later stages, like learning words, which are more complicated than phonemes, will be implemented on top of this early-embedded language circuit. And also I suppose this later network obeys the generic feature of memory networks you mentioned in last paragraph, such as deterioration of memory indices, like forgetting words.
Sahar - I admit that I haven't read the full article: I was primarily responding to the last sentence of this preview, which stated:
Reply | Report Abuse | Link to this"Because memories are neuronal connections that get reinforced with regular access, the finding means that even connections that have not been reaccessed for decades do not disappear completely, as previous evidence had suggested."
I suggest that the persistence of inherent phoneme circuit firmware, once the trimming process has occurred, regardless of any usage, does not apply to learned memories stored in 'volatile' memory circuits.
Again, my comments only apply to this brief summation, which may conflict with the full article. Thanks very much for bringing this to my attention.
Interesting question about the language learning process. Unlike mathematics or other applications of symbolic reasoning capabilities, spoken language most often requires little if any instruction or teaching, especially regarding specific, complex rules. It would seem that the fundamental capability to learn spoken language is also effectively 'hardwired'. This also seems to apply to other organisms that, through all sorts of mechanisms, effectively communicate among themselves.
I think that I demonstrate the affect of aging on memory, including my inability to think of that best word. It usually appears to me laterally (using peripheral vision?), after I've abandoned its direct pursuit... Very interesting - thanks.