The psycholinguistics of bilingualism - neurolinguistics — Transcript
Full transcript
- 0:00hello there and welcome back to this
- 0:02course on bilingualism this is already
- 0:04the last video so in this session I
- 0:06would like to talk about neural
- 0:08linguistics and specifically these three
- 0:10questions how does the brain host two
- 0:12languages which brain structures are
- 0:15involved in processing the two languages
- 0:17that are at work in a bilingual speaker
- 0:19second does learning a second language
- 0:22somehow change your brain does it rewire
- 0:24your brain does it make your brain
- 0:26different from the brain of a
- 0:28monolingual speaker and third if we
- 0:30switch from one language to another does
- 0:33it mean that different areas of the
- 0:35brain are being activated as we're using
- 0:38one language and then another all right
- 0:41that's what we talk about the ideas and
- 0:43the experiments that I'll discuss are
- 0:45described in more detail in Chapter ten
- 0:47of course all and le's the psycho
- 0:49linguistics of bilingualism if you're
- 0:51interested in the details or you need to
- 0:53look up the references go to the chapter
- 0:55it's all in there right so I need to
- 0:58start by pointing out a few things about
- 1:00different brain areas that are involved
- 1:02in language processing many of you will
- 1:05know this but bear with me so two areas
- 1:08that will come back again and again in
- 1:10this video are Broca's area and
- 1:12Wernicke's area they are associated with
- 1:15different language functions and we know
- 1:18a lot about these areas because of
- 1:21specific language impairments that are
- 1:24associated with brain damage with
- 1:26lesions to these areas so if a speaker
- 1:30sustains brain damage to Broca's area
- 1:33that's a because of a stroke there is a
- 1:35specific set of symptoms that arises and
- 1:38that is called Broca's aphasia so
- 1:41aphasia
- 1:42that's a speech impairment and Broca's
- 1:44aphasia is characterized by slow
- 1:46effortful speech that's telegraphic so
- 1:49it relies on individual words typically
- 1:52high-frequency words and smaller
- 1:54function words the grammatical bits and
- 1:56pieces they are omitted more often than
- 1:59not so there's a lack of grammatical
- 2:01markers determiners prepositions
- 2:04pronouns these things tend to get
- 2:07permitted and also word finding is
- 2:10severely impaired so Broca's aphasics
- 2:12struggle to
- 2:14instruct their utterances and what they
- 2:16get out is typically the sequence of
- 2:18lexical high-frequency of items Broca's
- 2:21aphasics tend to be aware of their
- 2:23speech deficit so they know that
- 2:24something is not right and this can be
- 2:27very frustrating experienced and on the
- 2:30other hand comprehension comprehension
- 2:32side of language tends to be largely
- 2:34intact so Broca's aphasics understand
- 2:36what it is that you are saying to them
- 2:39but they have problems producing
- 2:42language so but also processing is
- 2:46impaired to some extent with regard to
- 2:48syntactic structure so structures that
- 2:50rely on these functional items that
- 2:53Broca's aphasics have difficulties with
- 2:55so for instance if you give a Broca's
- 2:58aphasic a sentence such as this is the
- 3:00cat that the horse jumped on a complex
- 3:02relative Clause construction and you ask
- 3:04them well who did the jumping they have
- 3:07more difficulties figuring that out than
- 3:09a person without Broca's aphasia so very
- 3:12specific set of symptoms associated with
- 3:15brain damage to a specific brain area
- 3:18with Wernicke's area it's a different
- 3:20set of symptoms that arises so
- 3:23Wernicke's effects have poor auditory
- 3:25and reading comprehension in severe
- 3:27cases they can't even understand single
- 3:29words that you say speech on the other
- 3:32hand tends to be relatively fluent so it
- 3:34sounds normal except that it is
- 3:37disconnected from meaning it tends to be
- 3:39semantically empty it tends to be weird
- 3:41and another characteristic of Wernicke's
- 3:44aphasia is the use of para phages that
- 3:47is the use of words that sound just
- 3:49about right they follow the phonotactics
- 3:52of the respective languages but they are
- 3:54not actual words unlike Broca's aphasics
- 3:58Wernicke's aphasia tend to be unaware of
- 4:00their speech deficit alright so this
- 4:02just as a general primer of Broca's area
- 4:05and Wernicke's area and the symptoms
- 4:07that are associated with damage to these
- 4:10areas when I talk about these areas it
- 4:14reflects the idea that language tends to
- 4:17be localized in the brain that tend to
- 4:19be language centers that are responsible
- 4:21for very specific language related tasks
- 4:24and this of course is true
- 4:26but then again the story is more complex
- 4:28than that we have new evidence pointing
- 4:30to a more important role of connectivity
- 4:33between different areas no language area
- 4:36does it alone they have to work together
- 4:38they have to be connected to make
- 4:40language work and this means that
- 4:41language is not fully localized not
- 4:44fully modularized but rather we have
- 4:46different areas that are involved and
- 4:49some of these are even non cortical that
- 4:51is they don't sit on the outer shell of
- 4:55your brain the cortex and another thing
- 4:59is that actual aphasic patients rarely
- 5:01are perfectly clear cases of any one
- 5:04aphasia so when I described Broca's
- 5:06aphasia and Wernicke's aphasia I was
- 5:09giving you the stereotypical case in
- 5:12actual fact note who aphasic patients
- 5:14are completely alike kind of like
- 5:17bilinguals you don't find any to
- 5:18bilinguals are completely alike with the
- 5:21physics the symptoms may overlap now but
- 5:25even in speakers with similar lesions
- 5:27there may be differences because of
- 5:30let's say in different languages that
- 5:32are involved so in the literature there
- 5:35has been a trend to talk about brain
- 5:36organization rather than brain
- 5:39localization okay let's come to our
- 5:43questions that are specific to
- 5:46bilingualism do bilinguals have a
- 5:48language switch that allows them to
- 5:50change from one language to another
- 5:53Hernandez and colleagues did a study on
- 5:56spanish-english bilingual and they were
- 5:58using a picture naming task where you
- 6:01show people a picture and then you ask
- 6:03them what's the word for this in one
- 6:06language or in another language so they
- 6:09see pictures of a puzzle of a ladder
- 6:11ladybug a lamp or a lion and they have
- 6:14to say the respective words in Spanish
- 6:17or in English and they have to switch
- 6:20between different languages what has
- 6:22been used in terms of neuro linguistic
- 6:24methodology here was an fMRI design fMRI
- 6:29stands for functional magnetic resonance
- 6:31imaging and if you were planning on
- 6:33doing an fMRI study in your living room
- 6:36later today after disappoint you
- 6:39only works with heavy machinery and
- 6:42expensive equipment so you see an fMRI
- 6:44machine here and what this machine
- 6:47allows you to do is to measure magnetic
- 6:50fields in the brain depending on the
- 6:54blood flow so where in the brain you
- 6:57have a flow of fresh blood it comes in
- 7:01with much oxygen and it can be
- 7:04contrasted with the positions of used
- 7:07blood that has relatively less oxygen so
- 7:11researchers talked about so-called bold
- 7:13signals blood oxygen level dependent
- 7:15signals and as people engage in
- 7:18cognitive tasks fresh blood with lots of
- 7:21oxygen flows towards areas in the brain
- 7:24that are active in those specific tasks
- 7:26so if you ask people to do a math
- 7:29problem or if you ask people to imagine
- 7:31the picture of a butterfly you get
- 7:35different responses you get fresh blood
- 7:37flowing towards different areas in the
- 7:39brain right so mmm the the crucial way
- 7:44to do this is that fMRI allows you to
- 7:46compare levels of activity during a
- 7:52resting state on the one hand and a
- 7:54specific task on the other so the
- 7:56resting state is kind of the average
- 7:58level of activity the baseline of
- 8:01activity that you have at any given
- 8:03moment so your brain is basically on
- 8:06standby as it's doing its thing all of
- 8:09the time and the resting state is
- 8:11measured in such a way that you have
- 8:13participants and you tell them to think
- 8:15about well nothing in particular now of
- 8:18course you can't help thinking about
- 8:19something in particular right so every
- 8:22participant thinks about something
- 8:24different when you tell them that one
- 8:26person thinks about the grocery list
- 8:28another person isn't sure whether
- 8:29they've locked the door another person
- 8:32is thinking about a friend they're going
- 8:34to meet later that day and so on and so
- 8:35forth but on the whole these different
- 8:38thoughts are going to cancel each other
- 8:39out
- 8:40yeah so everybody thinks about something
- 8:43different and that gives you the kind of
- 8:45baseline that you can compare something
- 8:47more specific against so the specific
- 8:49task would be all of your participant
- 8:52using their brains for the same thing in
- 8:54the same way and then this allows you to
- 8:57measure which brain areas are actually
- 8:58more active during the specific task as
- 9:01compared to this averaged out resting
- 9:04state here are a few pictures that show
- 9:07the results of fMRI studies importantly
- 9:10these pictures don't show the brain of
- 9:11any one particular person but they are
- 9:14overlays of all the participants that
- 9:17have been averaged out yeah so here we
- 9:19see an fMRI scan of participants
- 9:23speaking words and not surprisingly we
- 9:25have Broca's area active and we also
- 9:28have a good chunk of the primary motor
- 9:32cortex active the primary motor cortex
- 9:35controls or your your limbs and your
- 9:38muscles and your mouth and your
- 9:39articulate errs they are part of that a
- 9:42big part of that okay so that's speaking
- 9:45word so it's very different from seeing
- 9:47words when you show people words on the
- 9:49page well unsurprisingly parts of the
- 9:52visual cortex that sits right in the
- 9:54back of your brain is highly active in
- 9:56hearing words while the primary auditory
- 10:00cortex is active and also Wernicke's
- 10:02area which is associated with looking up
- 10:06words connecting words sounds to the
- 10:09entries in the mental lexicon and then
- 10:12looking up meanings of those words
- 10:14that's what we're in Kies area is
- 10:16responsible for and we see it's active
- 10:18here okay so fMRI has excellent spatial
- 10:24resolution it tells you more or less
- 10:25exactly where something happens but on
- 10:28the downside it's temporal solution is
- 10:30less excellent so the blood flow takes
- 10:34some time and it's not super discreet so
- 10:37we're not sure exactly when something
- 10:40happens but for a process such as
- 10:43speaking words well it takes a couple of
- 10:45hundred millisecond so that's close
- 10:47enough right coming back to Hernandez
- 10:51and the picture naming tasks what people
- 10:53did was they saw a picture so here we
- 10:57see a picture of a horse and that was a
- 10:58sound indicating which language they
- 11:00should use so if they heard say well it
- 11:05was a sign use English and people would
- 11:07say horse in another condition so the
- 11:11condition that actually interested
- 11:13Hernandez and colleagues was the
- 11:16condition where you had one language
- 11:17first and then another so let's say you
- 11:20hear the you see the horse picture and
- 11:22you hear say and then you see a picture
- 11:24of this and you hear diga I'm butchering
- 11:28Spanish I'm sorry so here you're
- 11:31supposed to say na teeth and again sorry
- 11:35and the question that Hernandez and
- 11:38colleagues wanted to answer is whether
- 11:39there's a brain region active when the
- 11:42language cue indicates a language switch
- 11:44and they measured that via fMRI here are
- 11:48the results and you see when you compare
- 11:49the alternating condition to the
- 11:52monolingual English conditions are two
- 11:54stimuli following each other both of
- 11:56them are English here we have two
- 11:58stimuli following each other both of
- 12:00them are Spanish you see that
- 12:01something's going on in the alternating
- 12:03condition there's heightened activity in
- 12:06the dorsal lateral prefrontal cortex and
- 12:09this could be interpreted with some
- 12:12goodwill as a language switch so there's
- 12:15something going on telling you okay it
- 12:18was English
- 12:19now it's Spanish let's go but of course
- 12:22it's more complicated than that
- 12:24so the area that you see active there
- 12:26the dorsal lateral prefrontal cortex is
- 12:29associated with something that I talked
- 12:31about in the last video namely executive
- 12:34control so executive control is
- 12:36something that's very general and
- 12:38involved in several different cognitive
- 12:41activities it is responsible for
- 12:44selective attention or inhibition of
- 12:46distractors it's response is involved in
- 12:49manipulating ideas in working memory and
- 12:52so this is not really a language switch
- 12:56it's something more general and in the
- 12:58last unit we talked about how my lingual
- 13:01have a heightened ability of executive
- 13:05control and switching between languages
- 13:08is one situation in which executive
- 13:12control is needed that's why bilinguals
- 13:14are more trained with regard to these
- 13:15things and in the long run become
- 13:17better at ignoring distractors or
- 13:20focusing their attention on something
- 13:22that's important and disregarding
- 13:24something that's unimportant so
- 13:26conclusion no language switch but we
- 13:28found corroborating evidence for the
- 13:31idea that executive control is
- 13:33heightened in finding ones another
- 13:36question do l1 and l2 use the same
- 13:39neural structures do they depend on the
- 13:41same neural circuits here's a study by
- 13:43Kim and colleagues who had people
- 13:45generate sentences of people were asked
- 13:48to describe their daily experiences in
- 13:51the l1 and in the l2 and they were doing
- 13:55that in an fMRI machine and the
- 13:57researchers measured blood flow in late
- 14:02learners and in early learners and
- 14:04looked at differences that emerge and
- 14:06you see here the picture of Broca's area
- 14:08and a late learner and you see those are
- 14:11sentence generations in the in the
- 14:15native language here we have the second
- 14:17language and you see those two are
- 14:18fairly distinct here we have an early
- 14:20learner and so bilingual who has
- 14:24competence in Turkish and in English and
- 14:27you see that there's a large common area
- 14:29so late learners tend to do things
- 14:33separately the Broca's area early
- 14:36learners have more of a common area an
- 14:39overlapping area so this points to
- 14:42different neural wirings processing
- 14:45language in early and late bilinguals
- 14:48particular late learners tend to use
- 14:50different parts of Broca's area for
- 14:52speech generation and you can imagine
- 14:55how this would be less efficient or more
- 14:57effortful than what it is that the early
- 15:00learners do so they have one system that
- 15:02sort of fixes everything and here we
- 15:04have two systems each of which is
- 15:06struggling to get things done here's
- 15:10another study this time Italian German
- 15:13bilinguals who were asked to perform a
- 15:16chromatic ality judgment test so it is
- 15:18this sentence here a good sentence of
- 15:21German the whole of Liberty visa well
- 15:24those who speak a little bit of German
- 15:26you know that well laughing is not the
- 15:28right verb form here
- 15:29should be left singular laughing that's
- 15:33plural and so no this doesn't work you
- 15:36should say no that's not a great
- 15:37sentence of German subjects were divided
- 15:40into three groups our early learners
- 15:42with high proficiency late earners with
- 15:45high proficiency and late learners with
- 15:47low proficiency and here you see the
- 15:49three different types of responses so
- 15:52the early learns with my proficiency
- 15:55late learners with high proficiency and
- 15:57the late learners with low proficiency
- 15:59and you see when you compare these three
- 16:01conditions there's heightened activity
- 16:04in the late learners regardless of
- 16:09proficiency so in both of these groups
- 16:12processing is characterized by more
- 16:15neural activity so this tells us that
- 16:18it's actually more effortful and this is
- 16:22quite common sensical right so the early
- 16:23learners for them it's easier to perform
- 16:26the grammaticality judgment task it
- 16:28involves less neural activation than
- 16:30what we see in the late ones mm-hmm okay
- 16:34there's not only fMRI there's also
- 16:36something called ERP that is frequently
- 16:39being used ERP are event-related
- 16:41potentials and they work like this so
- 16:46people have a set of electrodes on their
- 16:49scalp and these measure electrical brain
- 16:53activity and there are two signatures
- 16:56that we need to talk about one is called
- 16:58in 400 and 400 n stands for negative 400
- 17:02stands for 400 milliseconds so an N 400
- 17:06is a negative electrical peak that
- 17:09occurs 400 milliseconds after the
- 17:11stimulus namely a semantically odd word
- 17:15something that does not quite work out
- 17:17in terms of meaning so we have two
- 17:20waveforms here and one is semantically
- 17:25okay that's meaning number sentence
- 17:28number one the pizza is too hot to eat
- 17:30and semantically that's fine nothing
- 17:32special here sentence two is weird the
- 17:35pizza is too hot to sing so here we have
- 17:38the onset of the words eat and sing
- 17:42and sing the offending word triggers an
- 17:45N 400 400 milliseconds down the line
- 17:48that's where you realize well pizzas
- 17:50don't sing do they so this kind of
- 17:55thought process goes on that's the end
- 17:57400 meaning there's also something
- 18:00called a P 600 by now you can figure out
- 18:02what P 600 actually means P stands for
- 18:05positive 600 600 milliseconds after an
- 18:09offending stimulus so here again we have
- 18:12two sentences one that is okay the cat
- 18:15won't eat nothing out of the ordinary
- 18:17mortals and technically the cat won't
- 18:20eating yeah no that's not English that's
- 18:23word soup you know the cat won't eating
- 18:25so 600 milliseconds after eating there
- 18:29is a positive charge at the 600
- 18:33millisecond mark okay and this of course
- 18:35can be exploited for the analysis of
- 18:38language processing in bilinguals erp is
- 18:42sort of complimentary to fMRI it has
- 18:44excellent temporal resolution
- 18:46millisecond by millisecond but in terms
- 18:49of space it has poor spatial resolution
- 18:52so we don't know where exactly the
- 18:54electrical activity in the brain is
- 18:57happening so that's why ERP and fMRI
- 18:59tend to be used alongside each other
- 19:02okay here's a study on English Spanish
- 19:05binding rules that had to perform a
- 19:07grammaticality judgment test so we're
- 19:10concerned with grammar syntax rather
- 19:13than meaning so p600 is the name of the
- 19:16game here and they had native speakers
- 19:20of Spanish react to amorphous syntactic
- 19:21violations with a particular ERP
- 19:25signature they showed the p600 and the
- 19:28question was do L to learn some Spanish
- 19:30so show the same pattern of brain
- 19:33activity so do also the learners show
- 19:36this p600 signature when they process
- 19:39language and so the first type of
- 19:44violation that the first type of
- 19:46syntactic mistake here was a mismatch
- 19:49between the article of the gender of an
- 19:54article and now
- 19:55that is associated with so it should be
- 19:57una Fiesta not una Fiesta that's wrong
- 20:01that triggers a p600 in Spanish
- 20:04monolinguals and it actually also
- 20:08triggers a p600 or something similar in
- 20:12l2 learners so here you see different
- 20:15types of stimuli and averages over them
- 20:18so one wave the bolt the black one is
- 20:21the acceptable sentence and then we have
- 20:23an unacceptable variant of the same
- 20:26sentence all the black that has the
- 20:28right determiner and the gray as the
- 20:31wrong determinant with the wrong gender
- 20:33and you see that at the 600 millisecond
- 20:36mark the unacceptable sentences tend to
- 20:39be above the acceptable sentences so
- 20:42conclusion l2 learners of Spanish show
- 20:45sensitivity to around gender determiners
- 20:47that can be measured in terms of AP 600
- 20:51however the experiment actually returned
- 20:54a mixed bag of results
- 20:56there was another condition that tested
- 20:59not for article head word mismatch but
- 21:02rather in terms of gender but rather a
- 21:06mismatch in terms of number man so here
- 21:09we have another violation that's rather
- 21:11similar El Ninos well it should be
- 21:15Ella's seem to learn in uses plural
- 21:17that's a number of mismatch that doesn't
- 21:19work there should be a plural determiner
- 21:22yeah and here you see in the results
- 21:25that well strangely enough this didn't
- 21:29seem to offend as much as the gender
- 21:33mismatch so here learners of Spanish do
- 21:36not show a p600 sensitivity to wrong
- 21:39number
- 21:39determiner so what this indicates is
- 21:42that the grammatical characteristics of
- 21:45the morphosyntactic violation that
- 21:47actually matters and not all grammatical
- 21:51violations are registered in the same
- 21:53way by l2 learners here's another study
- 21:58this time with english-speaking learners
- 22:00of French and you were entering semantic
- 22:04territory so meaning is at issue here so
- 22:06the n400 is the measure
- 22:08that is being used here and an ERP was
- 22:13conducted in response to three classes
- 22:15of words so this was a lexical decision
- 22:19task that people had to do and at the
- 22:20same time they were well an ERP was
- 22:26conducted in three condition first a
- 22:30semantically related word that was
- 22:35related to a prime so people saw the
- 22:38word shell as a prime and then AdWords
- 22:40want to sha as the target and of course
- 22:44dog and cat they are semantically
- 22:46related so that's a semantic relation
- 22:49then there were pairs of semantically
- 22:51unrelated words in his own house swath
- 22:55thirst may be well when I get home I'm
- 22:59usually thirsty but the semantic
- 23:01relation is not as clear as with Shana
- 23:05and Shah so no semantically unrelated
- 23:08and then the third condition involved
- 23:11non words so no that's an actual word
- 23:13that means word and nezzie that sounds
- 23:16very French oh you wouldn't as you know
- 23:19but it's not a word mm-hmm so this
- 23:23should trigger a specific response and
- 23:26people who are looking out for and four
- 23:30hundred the signature that reflects
- 23:32semantic oddity and also lexical status
- 23:35and so you find a strong in 404 stimuli
- 23:38that are semantically odd and it's
- 23:40actually strongest for words that are
- 23:42not even words where you try to find a
- 23:44meaning but can't find one because it's
- 23:46not in the mental lexicon and there's no
- 23:48meaning associated with it at all right
- 23:51so what came out there were two groups
- 23:54non learners or friends of people who
- 23:57did not speak any French at all and
- 23:59French learners and they were subjected
- 24:02to this kind of experiment three times
- 24:07okay session one then a delay of several
- 24:11days session - same thing and finally
- 24:14session three and you see that for the
- 24:17non learners there is no difference
- 24:21between the three conditions and you
- 24:23wouldn't expect that so they don't know
- 24:25the meanings of shaman and shot so they
- 24:28don't see the semantic relation between
- 24:29them
- 24:30mizzle and suave same thing yeah to them
- 24:34it sounds French its French
- 24:38and so Nazi nothing special about that
- 24:41that could just sound like a French word
- 24:43even though it isn't let's look at the
- 24:46French learners so they expectedly show
- 24:50an n400 that is strongest for the pseudo
- 24:56word target so this is here the reaction
- 24:58to nausea that is clear offender and
- 25:01also the others shall some little
- 25:05activity but you see that that activity
- 25:08actually goes down for the related word
- 25:10target so when two meanings are strongly
- 25:14related then there's no need for this
- 25:16negative 400 because there's nothing
- 25:17semantically odd rather it's predictable
- 25:20that you would find this kind of
- 25:22semantic relation right summing up for
- 25:26today there are different brain areas
- 25:28involved in language there's
- 25:29localization to a certain extent but
- 25:31connectivity plays a strong role as well
- 25:35we've talked about different methods
- 25:37that can be used to study processing in
- 25:41bilinguals using the methods of
- 25:43functional magnetic resonance imaging
- 25:45and event-related potentials and that
- 25:49was it for this series believe it or not
- 25:51so glad you came along for the ride was
- 25:55a lot of fun and I hope you keep reading
- 25:58on bilingualism maybe go back to some of
- 26:00the studies in this book it's a really
- 26:03good book I can so recommend it so you
- 26:08really should go and buy it yeah alright
- 26:10that's it over I'll see you then and
- 26:13there will be you next time for
- 26:14something but I'm not sure what it is
- 26:16just yet okay bye
About this transcript
This page contains the full transcript of The psycholinguistics of bilingualism - neurolinguistics by Martin Hilpert, generated from the public captions YouTube serves with the video. The transcript has 3,894 words across 588 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.
What you can do with it
Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.
Free YouTube transcript tool
YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.