[SOUND] So what about age of acquisition and, and the brain, right? There are a couple of measures that have been used to look at brain activity. Specifically at where in the brain there might be activity for one type of task versus another or when comparing across subjects. In the case of bilingualism, we can look at what happens when we compare an early and a late bilingual, what happens in their brain as they are processing some stimulus. So this is work that's been done for many years. Originally, studies were done in the mid to late 90s. And one original study done by Daniela Perani looked at early and late bilinguals who were asked to listen to speech and found that, in fact the late bilinguals showed more activity compared to the early bilinguals. Now in a subsequent study, she thought about about this, and she said well if I take early and late bilinguals, but I don't match them in how well they speak the second language, then I'm really, it's a mismatch. It could be that it's proficiency. How well they speak the language that matters, and in fact, in subsequent studies, she found that that was the case. That when she matched the language proficiency, then the age didn't matter any more. In one of the most famous studies from that time by Kemital, they asked a group of early and late bilinguals to say to themselves what they did the day before morning, afternoon or evening. And then they scanned them using functional magnetic resonance imaging. What they found was that when they looked within Wernicke's areas, so the temporal areas involved in language, there was no difference between early and late bilinguals in the brain activity observed, basically, when you looked at both languages, it was essentially as if there was one big glob within Wernicke's area. Now, when they looked at Broca's area, they found a different effect. So for the early bilinguals, they found overlapping activity between the two languages. The two languages looked like they were laid upon each other. But when they looked at the late bilinguals, they found that there were two different areas, one for each language. Now this got a lot of press, it was a big deal at the time. It's highly cited, and there's a lot of controversy about this study. Partly because of the methodology, partly because there hasn't really been a replication of that study. And secondly because it was a, an odd task of telling yourself what you did the day before could involve lots of different things. Episodic types of memories, memories about what happened in our lives at a particular times of the day. It could be that they were speaking to themselves, we don't know exactly why they observed this pattern. And so there's a lot of controversy about this study. But it does suggest that there might be a difference between early and late learners. So let's go back to this idea of grammar, right? So what ow and oh and colleagues found was that the over heres could not pick up errors in grammar, even though they had heard the language very early in life. And that suggests that there's something about grammar that's sensitive to when a language is learned, right, and how long it's used. So there've been a series of studies looking at grammatical processing and age of acquisition. One of the most well-known ones was done by Weber-Fox and Neville. And in this study what they did is they asked a group people to listen to sentences that had grammatical violations. They had violations that indicated incorrect grammar, and they took a group of monolinguals and then a group of bilinguals who had learned English at different ages. And then they looked at their brainwaves, to look at what would happen if there was a grammatical violation or a semantic violation. Now semantic violations, what we mean by that is something like I take my coffee with milk and dog, right, so when the dog, when you heard dog, what happens is there's an N400, there's this wave that occurs, it starts at about 200 milliseconds after a word. it, it gets bigger and peaks about 400 milliseconds, that's why it's called a N400, and it's negative going wave. This is an EG wave that's averaged across many different words. Now the N400 is very robust there's a lot of literature on the N400, and it's thought to signify many different things. Words that occur more often in the language show smaller N400s, those that occur less often or less frequently show larger N400s. If you have a sentence like I take my coffee with milk and dog, the dog will give you a larger N400 because it's a semantic violation, it doesn't fit. Right there sugar would give you a reduced N400 because it fits. So lots of things that predict the N400, one of them being whether it fits in a sentence or not. So Weber-Fox and Neville looked at this N400 and looked at whether there would be a change in that N400 depending on when somebody had learned a language. And they also wanted to look at how well they spoke it. Weber-Fox and Neville also looked at grammatical violations, and grammatical violations can give different types of ways. One is a left anterior negativity, and the second is a P600, a positivity that occurs at 600 milliseconds. And this occurs much later, with the idea that when there's a violation, people are trying to repair that violation, and in repairing that violation there's this reconstruction of the sentence, and so the P600 occurs late, indicating this extra work that needs to be taken on, to reconstruct the sentence. So they looked at grammatical and semantic violation and looked at what happened across age. Now with the N400, they didn't find much sensitivity with age except for the latest group, the group that learned much, much later in life showed larger N400s. Relative to the group that learned early in life, but this was only the latest group, which was between 11 and 16, began to show differences in semantic violations for the N400. Now for the late anterior negativity, what's called ELAN and the P600, this positivity occurring at about 600 milliseconds, they found differences much earlier. And for some grammatical violations, as early as age of acquisition two. They found differences between those who had acquired English at age two, two to four, and native speakers who heard English from birth. So what this indicates, at least, is that grammatical processing and the brain waves associated with grammatical processing are very sensitive to when a language is learned. And the violations of semantics and meaning are much less sensitive to when a language is learned. And, and we could ask a couple of questions of why that might be, and if we think about the types of sentence that are used that are very academic types of sentences, they're involving general knowledge. Then we realize that these are things that can be transferred more easily from one language to the other, right? A dog is a dog. A chair is a chair. There are slight difference across different countries. I'm not trying to say there aren't any differences, but in terms of the standard academic types of concepts that are tested in these experiments, those transfer quite easily and are adapted quite easily across languages. But in terms of grammar, those don't seem to transfer quite easily. Hence, they're quite sensitive to age of acquisition. Now at this point, you might be wondering, well, so, so is it that I have to be exposed to a language early in life to have native-like you know, processing in my brain of, of a grammatical error? And there's a lot of controversy about this, right, studies are not all lined up in that direction that, that seems to be a general trend that grammar is sensitive to age. But there are other studies that have found that in fact at very high levels of proficiency, when speakers are quite good, they begin to show more native-like patterns, work done by Sonia Rossi and Angela Forizzi in Germany. Worked on by Eric Pakulak and Helen Neville in the U.S. and Oregon have also found similar types of findings. And also worked by and in Spain. In these studies ,they've found that you can find some features of brain activity that approximate those of native speakers in later learners who have very, very high proficiency. Work using functional magnetic resonance imaging, right, so as opposed the bent related potentials that look at the electrical activity in the brain and are very sensitive to timing, but not very sensitive to where in the brain a signal is coming from. Functional magnetic resonance imaging is not very sensitive to time, because it relies on blood oxygenation and hemodynamic changes in the brain. But it is very sensitive to where in the brain activity changes. Isabell Wartenburger, again with Daniela Perani and other colleagues looked at what would happen if they compared early and late acquirers, who were high and low proficiency, who spoke a language well or did not speak it as well. And their findings support some aspects of Weber-Fox and Neville studies. Specifically, that there's something about grammar that seems to be more sensitive to age of acquisition and when a language was learned. And there's something about errors in meaning or semantics that's less sensitive to age of acquisition. Again, their results are quite complex, and there are some pieces that suggest that proficiency may play a role as well. And we'll talk about that in subsequent sections, where we talk about language proficiency and, and expertise, and what that means. But there seems to be some idea, at least in literature, that grammar is sensitive to age of acquisition. It's not a consistent, exact finding, but it is a general theme that arises from the literature.
