مصاحبه با Niels Schiller

00:00 / 00:00


گزارش خرابی

Hi. Welcome to Leiden. Where we are filming today we're at
the Leiden University's Center for Linguistics for our interview for
the module language in the brain. Now unfortunately as you can see Inge
couldn't be here today but, of course, it doesn't mean that I
have to do it alone. It's an interview, after all. So I'm here with Professor Neils Schiller,
a professor of Psycho and Neuro Linguistics at the Leiden
University Centre for Linguistics. So thank you for being here. >> Sure, it's my pleasure. >> I'm sure that the people who
are watching this, and myself too, have many questions, but maybe we
can just talk about a few of them. So, the first thing that I want
you to ask is, in the course, in the lectures, we've talked about the so
called Western Bias. So the assumptions,
their conclusions, what we know about language in the brain is mostly
based on a very small set of languages. And also these languages
are like English and Dutch, so they are Western European
languages mostly. Now my question is, do you think that
by studying different languages from different places, will the conception of
language in the brain that we have change? >> That's a good question and my answer
is that it will maybe not change but certainly it will, it may qualify
the theories that have been established over the last maybe 30, 40 years or so. Because the theories in psycholinguistics
and in neurolinguistics that have been proposed, both in the area of language
processing, language comprehension, but also language production. They have been mainly based on
western languages as you say. Like English, maybe Dutch, maybe Dutch and
English, and a couple of other languages. But not the variety of languages
that we see in the world, not the diversity of languages. [COUGH] So I think that people would agree that the general processing of languages is, more or less,
the same for people all over the world, because our general cognition is more,
or less the same. However, particular languages have
particular features, and these may cause that particular processing steps in the
processing hierarchy that they may change. Or that they may be modified,
or may be different. >> Could you elaborate on this maybe so the, processing hierarchy,
what do you mean by that? >> I certainly can. So for
example my own area is speech production. Speech production what people
usually think is when we produce, let's say a word. Okay, let's take the situation that
we name an object like this cup or this white cup and
you want to say white cup, for example. How does this work? Well, first of all, you have to
recognize the object and you have to conceptualize it so that would be like
the meaning level, the semantic level. >> Then you would have to
look into your lexicon, whether you have a word
to name this object. And maybe there are also words that
have a slightly different meaning but you don't want to say, for example,
mug or glass or, you want to say cup. So you have to select the word
cup from your lexicon. And then you have to basically
phonologically, and phonetically encode that word because once you have
selected the word from the lexicon, you cannot just put it out, so to speak. Now, when you want to say white cup, for example in Dutch, Dutch is a language that has grammatical gender. >> Mm-hm.
>> So when you also want to say white in Dutch,
you have to make it congruent in terms of gender so
that the adjective and the noun, the head noun, go together,
which is called agreement, right? >> Yes, yes. >> Grammatical agreement. So you have to find the grammatical
gender features of the noun. Now, it is believed that the grammatical
gender features of the noun are processed, are stored also in the lexicon. And it has been shown recently,
relatively recently, in the last 10 or 20 years, that when we access
nouns that we also automatically access the gender information that we
needed, it's being selected as well. [COUGH]
What has been shown is that, what has not being shown so
far is how this works for other languages. So for example recently very
recently a PhD student of mine went to the south of Ethiopia to
study a language there called Konso. And in Konso it's the case that the grammatical
number feature, number system, excuse me. The grammatical number system and
the grammatical gender features, their gender system,
that they are intertwined. Now based on the experiments
that Mulugeta Seraw carried out with Konso speakers in the field. So he took a laptop and
a microphone with him to go there and carry out psycholinguistic
experiments in the south of Ethiopia. We think we are able to disambiguate cases
where we have to do with grammatical number processing and cases where we have
to do with grammatical gender processing. >> Yeah, so you're saying it's, this was a relatively recent
project that he carried out. Now, so there is an increase, at least, in other languages than
standard western languages. That is being, that are being studied. >> Yeah, and how far there is really an
increase, I mean, I think that people in the field of psycho and neurolinguistics
have started to also look at other languages that have different features
from the western European languages. And I think this is something
that has to be done more. However, there are some
practical constraints as well, because you have to carry out
the experiments mostly in the field. You have to get to the speakers. And therefore, you need mobile equipment. Now here at Leiden, for example,
we decided when we expand our lab in the future that we want to have
equipment that is also portable. Okay?
So that we can do mobile experiments basically. And it's interesting that
the producers of this equipment that we're using that they're
also going in the same direction. So they make their equipment smaller
such that this is portable and so forth. >> Yeah.
>> So I think this is a trend, and I think this trend will become
more important in the future. And may the, so and that the data that we gain on the basis
of these experiments that are being carried out in the field with different
languages on the studied languages. But they may qualify our
conceptions of language processing. >> Okay and so this was a PHD
project that you supervised? >> Yes that's right. >> And do you yourself also work in countries like Ethiopia
to do your own research? >> I haven't been doing
research in Ethiopia. I went there to look at the circumstances
and to give a little masterclass there, but I haven't been studying in
the field of Ethiopia myself. What I've been doing is studying
languages here in the lab. So, just to give you an example, for example over the last couple of years we
have been carrying out several studies where we looked at morphological
processing in the brain. Because it has been disputed whether the level of the representational
level of Morphology is independent of, let's say, the semantic processing level
and the phonological processing level. As I've just explained, the semantic level that we start
with when we want to name something. And then there's a phonological level
before we actually utter something. But the question is whether
there is something in between. I mean, we know that there are morphemes,
and we also know that morphemes have a representation in the brain. At least, we think there is. I mean, there are some speech
errors that suggest that. But there's very little experimental
evidence suggesting that so far. There have been some experiments in
Germany that have been done on German, and what we did was to carry out experiments,
both behavioral and ERP experiments on Dutch,
but also in English. And maybe it's easier, for example,
if I explain the experiment in English. So what people had to do in that
experiment was naming a picture. For example, a picture of the moon. So they had to say moon at some point. And before they saw
the picture of the moon, they had to read different prime words and
different conditions. So for example, one was moonlight. So they saw a compound and
they read that aloud where there was morphological overlap
between the prime and the target. >> Yeah, sorry, and prime,
that's a word that's used before the task? >> Exactly and it sort of preactivates as we think about
it certain representations in the brain. >> Yeah.
>> So when you, so the assumption was that when you read moonlight that the morpheme
moon may become preactivated such that when you name the picture of the
moon it's already been preactivated, and it's easier to access and therefore
you're faster to name the picture. >> Now when you say faster,
those are very small measures, right? >> That's right. That's also correct if you compare for
example, the naming of moon in a morphologically congruent condition
relative to a control condition, we are talking effect sizes of
about 20 to 30 milliseconds. >> Yeah.
>> So that's a 50th part of a second
that we are talking about. However, these differences are systematic,
and we can measure them, and we can do statistical analyses,
and prove that it's a real effect. Yeah, okay so when compare for
example the reading of the production of moonlight relative
to something like honeymoon. >> Yeah.
>> Which has nothing to do with moon, I mean,
at least it's not semantically obvious, that honeymoon is also something
that has to do with moon. You find effects for
both moonlight and for honeymoon. And the effects are basically
of the same effect size. So, what does this suggest? This suggests that presumably it
has nothing to do with a semantic level of priming, because if it were
a semantic level then you would expect differences for
the semantically transparent. And for the semantically opaque condition. And by adding other conditions as well, we were able to show that there is
basically a morphological processing level that seems to be independent of both
the semantic and the phonological level. Now you wanted to ask a question. >> Yeah.
So I was wondering but maybe you were going to say this as well. So you gave an example now of English and did you use this example with
English native speakers? >> Well that's the interesting thing. We did this with English native speakers
but also with bilingual speakers. >> Yeah.
>> Of Dutch and English, in this case. >> Yeah.
>> I mean here in the Netherlands in line, it's easy to find bilingual Dutch,
English, speakers. So what we did was we also looked
at bilingual speakers, and what we actually did was to see first of all
whether they established the same effect. And the answer, the short answer, is yes,
also bilingual speakers show that effect. Both in Dutch and in English, by the way. What has also been shown is that
when you present words in Dutch, like the prime word, sorry, the word,
the prime word in English and then you present a number of
other items in between in Dutch. So when they have to
switch to Dutch again, and when they then see the target and they
have to produce the target in English, so they have to switch back to English,
the effect size is exactly the same. So they don't suffer from having
been spoken in Dutch in between. >> Okay.
>> In between the naming of the prime and the naming of the target. Okay so when they then switch to another
language it doesn't destroy the effect. >> So could you say that then this
English is stored in a buffer to use it- >> Well, the way we think about it, it's a little bit different. We think that it shows that at least
switching to another language doesn't make use of that you have to
proactively inhibit the other language. Because if that were the case,
then you would expect that it's more difficult to switch back to
English when you have spoken Dutch. But that doesn't seem to be the case. So we know at least how it doesn't work. But we don't know exactly
how it really works. But we have excluded at least one
possibility in the corrective inhibition account. >> And of course that's good for the fields that there's still
a lot of work to be done. >> Yeah I mean the progress that we make,
there is progress but it comes in very little steps and
there's still a lot of work to be done. >> Yeah,
well I mean that seems to be a really tantalizing way to end this interview. Of course there is a lot more
that we could talk about. But for now, I would invite you
all to go to the forum if there's things that you would like to discuss
about the interview that we just saw. And I would like to thank
Professor Schiller for his time. >> Sure, my pleasure. Yeah, thank you. >> And well, see you in the next video.

دانلود با کیفیت بالا
دانلود با حجم کم