10.4: Language Processing in the Brain
- Page ID
- 114835
\( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \)
\( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)
\( \newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\)
( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\)
\( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\)
\( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\)
\( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\)
\( \newcommand{\Span}{\mathrm{span}}\)
\( \newcommand{\id}{\mathrm{id}}\)
\( \newcommand{\Span}{\mathrm{span}}\)
\( \newcommand{\kernel}{\mathrm{null}\,}\)
\( \newcommand{\range}{\mathrm{range}\,}\)
\( \newcommand{\RealPart}{\mathrm{Re}}\)
\( \newcommand{\ImaginaryPart}{\mathrm{Im}}\)
\( \newcommand{\Argument}{\mathrm{Arg}}\)
\( \newcommand{\norm}[1]{\| #1 \|}\)
\( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\)
\( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\AA}{\unicode[.8,0]{x212B}}\)
\( \newcommand{\vectorA}[1]{\vec{#1}} % arrow\)
\( \newcommand{\vectorAt}[1]{\vec{\text{#1}}} % arrow\)
\( \newcommand{\vectorB}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \)
\( \newcommand{\vectorC}[1]{\textbf{#1}} \)
\( \newcommand{\vectorD}[1]{\overrightarrow{#1}} \)
\( \newcommand{\vectorDt}[1]{\overrightarrow{\text{#1}}} \)
\( \newcommand{\vectE}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash{\mathbf {#1}}}} \)
\( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \)
\( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)
The brain and language processing
Below are some updated graphics of the brain with respect to processing. They may be useful to you as you read the subsequent material.
Neurolinguistics: Syntactic Category Differences in the Brain, in Anderson's Essentials of Linguistics
Video Script
We’ve seen that we can group words into categories according to how they behave. We know that words within a particular syntactic category behave similarly to the other words in that category. They’re similar in their inflectional morphology, and in their syntactic distribution, that is, what positions they can occupy in a sentence. That’s some linguistic evidence that syntactic categories are real. There’s also neurolinguistic evidence that our brains respond differently to words from different categories.
Lorraine Tyler and her colleagues used functional magnetic resonance imaging (fMRI) to measure blood flow to different regions of the brain. The idea behind fMRI is that brain activity consumes oxygen, and when a particular area of the brain is active, then more blood flows there to bring it more oxygen. The researchers asked people to do an easy task. They showed them lists of three words and asked them to decide if the third word, the one in all caps, was related to the first two. So in this example, sparrows, thrushes, and wrens are all kinds of birds, so the participants would respond Yes. In this next example, hammer, wrench, banana, the first two are tools but the third one is a fruit, so it’s not related to the first two, so the participants would answer No.
Some of the words in this task were nouns, like the lists we just saw, and some were verbs, like these ones: eating, grazing, dining. All of those words are related to eating, so the participant would decide Yes. This is a pretty simple task, but what the researchers found in the fMRI was that there were several areas of the brain that showed greater blood flow for verbs than for nouns! Apparently, the brain was reacting differently to words from these two different syntactic categories, even though the task was the same for both categories.
We also see differences between nouns and verbs in the brains of people with aphasia. Aphasia is the name for any kind of language disorder that results from an injury to the brain, such as a stroke or a tumour. There are different kinds of aphasia that have different kinds of symptoms. Louise Zingeser and Rita Sloan Berndt found a dissociation between nouns and verbs in the speech of two different groups of people with aphasia.
The researchers asked their participants to do a few simple tasks. One was a picture naming tasks, where the researcher would show a line drawing and ask the participant to say what it was, like a fish or a car. In another task, they asked participants to describe how they would go about a particular action, such as making a birthday cake or attending a concert. And in the last task, they gave participants a picture book that depicted a well-known fairy tale but didn’t have any words in it and asked them to tell the story. So from all these tasks, they had a good collection of each person’s speech. For each person, they calculated the ratio of nouns to verbs.
It turns out that in people without any brain injury, the ratio of nouns to verbs is pretty close to one. That means there are about the same number of nouns as there are verbs in the average person’s speech for these tasks. But for people with agrammatic aphasia, verbs are very difficult to produce. These people had more than twice as many nouns as verbs. And for people with anomic aphasia, nouns are quite difficult. This group of people had fewer nouns than verbs.
Aphasia researchers call this kind of pattern a dissociation. We say that nouns are verbs are dissociated from each other because it’s possible to have verb production impaired while noun production is ok, or vice versa. This dissociation is consistent with the idea that verbs and nouns are processed differently in the brain.
So what does all this mean? We’ve seen that, in typical brains, a simple task with verbs involves greater brain activity than the same task with nouns. And in people with brain injuries, some people have an impairment of verbs but not nouns, while others have an impairment of nouns but not verbs.
All this suggests that our brains treat words differently depending on what category they’re in. Or in other words, different syntactic categories exist not just in language, but also in the brain!
Check Yourself
- Answer
-
"Humming, singing, whistling."
Hint: Think about the categories of each group of lexicon, and remember which took more bloodflow to recall.
- Answer
-
"Anomic aphasia."
Hint: Think about the word that the patient can't recall, and then think about its category.
- Answer
-
"Agrammatic aphasia."
Hint: Think about the word that the patient can't recall, and then think about its category.
Psycholinguistics: Traces in the Mind, in Anderson's Essentials of Linguistics
Video Script
We’ve been working with a theory that says that the operation MERGE generates a Deep Structure. For this wh-question, Who did Lucy invite to wedding, the Deep Structure looks like this. This wh-pronoun who refers to whoever it is that Lucy invited, and it is generated in this position in the complement of the Verb head, which is exactly where the noun phrase complement would be if we know who it was that Lucy invited. The preposition phrase gives us more information about the event of inviting, and it’s adjoined at the V-bar level. Because this is a wh-question, there’s both a [+Q] feature and a wh-feature in the C-head position.
Then the operation MOVE does its work.The wh-phrase moves up to the Specifier of CP, where it can support the wh-feature in C. Then do comes into the T-head in its past-tense form, did, then moves up to the C-head position.
One element of this theory that we’ve been taking for granted so far has to do with the trace that’s left behind when something moves. When we speak a sentence, we pronounce words in their Surface Structure positions, but we don’t pronounce anything in the Deep Structure position. But when we draw the tree, we show the deleted copy in that Deep Structure position, to suggest that, in the underlying representation, in our mental grammar, there’s something unspoken occupying that position.
There is some linguistic evidence for the existence of traces in our mental grammar. We’re claiming that there’s a trace in this position in the complement of invite. Notice that it’s not possible for any other phrase to occupy that position: if we try to put another noun phrase in the complement position, we can observe that each attempt is ungrammatical.
There’s also some psycholinguistic evidence for the existence of traces. The evidence comes from what’s called a visual world experiment. In this kind of experiment, a person’s eye-movements are measured using a device called an eye-tracker. The eye-tracker records where their eyes move while they listen to a spoken paragraph and look at a visual scene. The spoken paragraph goes like this:
This story is about a boy and a girl. One day they were at school. The girl was pretty, so the boy kissed the girl. They were both embarrassed after the kiss.
The idea behind a visual world experiment is that you look at what’s being mentioned. So when you hear the boy, your eyes move to the picture of the boy, and when you hear the girl, your eyes move to the picture of a girl.
At the end of this paragraph, one group of listeners heard a wh-question, Who did the boy kiss that day at school? A different group of listeners heard the same paragraph, but followed by a yes-no question, Did the boy kiss the girl that day at school?
These two sentences are similar in their structure, but they have a crucial difference. In the complement of the verb position, the yes-no question has an overt noun phrase, the girl. In that same position of the wh-question, there’s a trace, a deleted, unpronounced copy of the moved wh-word.
The researchers focused on this exact position in the spoken question: they observed where the participants’ eyes moved after the verb kiss. They compared how often the participants looked at the girl vs. how often they looked at the boy. In the yes-no question, when they heard the verb kiss, people looked at the boy 11% more often than to the girl, maybe because the boy is the one doing the kissing. But in the wh-question, when they heard the verb kiss, people looked to the girl 21% more than to the boy.
In both scenarios, the boy kissed the girl. But people’s eye movements differed in the two conditions. We know from previous studies that eye movements are quite closely synchronized to what’s being mentioned in the discourse. In this study, people’s eyes move to the girl not just when the sentence refers to her overtly, but also when the deep structure contains a trace that refers to her. The evidence from this eye-tracking experiment suggests that traces don’t just exist in tree diagrams, but also in our minds.
Check Yourself
- Answer
-
"What did Christina order what at Chipotle."
Hint: Recall what we studied earlier in Syntax with respect to movement, and then think about where that 'what' would have originated from.
- Answer
-
"*What did you eat sandwiches for lunch?"
Hint: Recall what we studied earlier in Syntax with respect to movement, and then think about where that 'what' would have originated from.
- Answer
-
"What did the fox chase ^ into the hedge?"
Hint: Think about the role that a Prepositional Phrase plays, which includes movement or location.
Neurolinguistics: Using EEG to Investigate Syntax and Semantics, in Anderson's Essentials of Linguistics
Video Script
When we started talking about semantics, we observed that a sentence’s syntax influences its semantics, because of the principle of compositionality. For example, we saw that a given string of words can have two different meanings if it has two different grammatical syntactic structures. And yet, we also observed that syntax is independent of syntax. A noun phrase that has the semantic thematic role of Agent often occupies the syntactic position of Subject, but not all Agents are Subjects, and not all Subjects are Agents!
The division of labour between thematic roles and grammatical roles is some evidence that syntax and semantics are represented differently in our minds. There’s also evidence from neural imaging to show that our brains process semantic information differently from syntactic information. This evidence comes from electroencephalography or EEG. Electroencephalography uses electrodes to measure electrical activity on a person’s scalp from which scientists can draw conclusions about the person’s neural activity. The particular EEG technique that gets used in neurolinguistics is ERPs or event-related potentials, which measure the timing of the neural response to a particular event, like a sound or a word.
When we’re observing ERPs, we always do so by comparing responses to different kinds of events, and the usual comparison is between events that are expected and events that are unexpected. For example, a sentence like, “She takes her coffee with cream and …” sets up a very strong expectation in your mind of what the next word will be. If the next word that arrives in the sentence matches your mind’s expectation, then the electrical response at your scalp will look something like this: the baseline condition. But if the next word that shows up violates your mind’s expectation, then compare your brain’s response: We observe a spike in negative voltage about 400 milliseconds after that unexpected word appears.
This response is called an N400. The N in N400 stands for a negative voltage, and the 400 indicates that this spike in negative voltage shows up, on average, about 400 milliseconds after the event. The N400 was first observed in 1980 by Kutas & Hillyard and has been replicated hundreds of times since then. It’s clear from all these studies that the particular kind of event that leads to an N400 response is a word that is unexpected in the semantic context.
The N400 is the brain’s response to an unexpected or surprising event, but not every kind of surprise will produce an N400. In other words, we have expectations about things besides the meanings of sentences. Think about a simple sentence like, “The bread was…” If that sentence finishes with “eaten”, that fits our mind’s expectation, and this is the baseline brain response. Now, what expectation do you have for this sentence, “The ice cream was in the..”? You probably expect a noun to come next, to follow the preposition and determiner. But if what comes next is not a noun but a verb participle, this violates your mind’s expectation. Notice that the word eaten is semantically consistent with ice cream, but is not consistent with the syntax of the sentence: determiners are followed by nouns, not verbs. So the brain’s response is a positive voltage about 600 milliseconds after that unexpected word: a P600.
When we’re using language in real time — either reading or listening — our mind sets up expectations about what’s going to happen next. If what happens next violates our semantic expectations, the brain’s response is an N400. And if what happens next violates our syntactic expectations, the brain’s response is a P600.
These two different brain responses give us further evidence that syntax is independent of semantics in our brains!
Check Yourself
- Answer
-
"N400."
Hint: Think about what is wrong about the indicated word, and the types of ERPs described above.
- Answer
-
"P600."
Hint: Think about what is wrong about the indicated word, and the types of ERPs described above.
- Answer
-
"P600."
Hint: Think about what is wrong about the indicated word, and the types of ERPs described above.
Neurolinguistics and Second Language Learning, in Anderson's Essentials of Linguistics
Video Script
As we’ve been talking about mental grammar, we’ve concentrated almost entirely on the mental grammar of your native language — the language you learned to speak in childhood, in your home. Linguists refer to your native language, your first language, as your L1. But many people in the world speak more than one language, and many of those people learned a second or third language in a different way from their L1. Any language that you learned after childhood, whether you learned it in school, using software, by travelling or immigrating somewhere, is called an L2 (even if it’s really your third or fourth language).
Learning an L2 is different from learning an L1 for a couple of different reasons. One is that, obviously, the language learner is not a child, so their cognitive processes might be different from those of a child. L1 learning happens by being immersed in a language environment, and most of the learning is unconscious, without overt teaching. L2 learning often happens with a lot of conscious effort: studying and memorizing and practicing.
But of course, the biggest difference between L1 learning and L2 learning is that when you start learning an L2, you already know at least one other language. The mental grammar of your L1 can influence the mental grammar that you’re developing for your L2: this is called transfer. Transfer can be helpful in L2 learning or it can pose a challenge. If your L1 includes a structure that’s similar to a structure in the L2, then you might experience positive transfer, which facilitates learning the L2: you can transfer what you know from L1 and apply it to the L2. But if the structures that you’re learning in L2 are different from those in your L1, then you might experience negative transfer: the knowledge from your L1 could make it more difficult to learn the new structures in the L2. And of course, you might experience both positive and negative transfer from your L1 to different parts of the grammar of your L2.
One theory of second language acquisition predicts that we would observe differences between native speakers and beginner L2 learners, but as the L2 learners become more proficient, their mental processes should become more and more native-like — that is the mental grammar of a fluent L2 speaker should look very similar to the mental grammar of an L1 speaker of that language. We can use the tools of psycholinguistics and neuroscience to learn about the mental grammars of L2 learners. Let’s take a look at some of the evidence.
Several studies have compared N400 effects in L1 and L2 speakers of a language. Remember that the N400 is an electrophysiological response that our brains show when a word is semantically unexpected in a given context. The brains of native speakers of English show a negative voltage about 400 milliseconds after a semantically unexpected word (socks), compared to an expected word. But what do the brains of non-native speakers show? What do we see in L2 learners?
A 2001 paper by Anja Hahne compared L1 speakers of German with L1 speakers of Russian who had moved to Germany in their 20s and had been living there and studying German for an average of six years. The experiment used fairly simple German sentences like these ones:
Die Tür wurde geschlossen / The door was being closed.
Der Ozean wurde geschlossen / The ocean was being closed.
Obviously, the word closed is a reasonable way for the first sentence to end but is a pretty unexpected way for the second sentence to end. So it’s not surprising that the native speakers of German showed an N400 in response to sentence 2 compared to sentence 1. The L2 speakers, the ones who had started learning German in their 20s, also showed an N400 to sentence 2. Hahne concluded that words that are semantically unexpected cause “essentially similar semantic integration problems in native participants and second-language learners” (Hahne, 2001: 263). In other words, the evidence from the N400 suggests that the lexical semantic component of an L2 learner’s mental grammar is not too different from that of an L1 grammar.
Now, we know that there’s a whole lot more to mental grammar than just the meanings of words. What can ERPs tells us about morphology and syntax? Remember that native speakers’ brains often show a P600 response to sentences that are syntactically unexpected. For many years, studies that looked at the P600 in L2 learners seemed to suggest that adult language learners never really approached native-like proficiency in their L2 morphosyntax: the P600 response to syntactic violations was significantly delayed or not there at all in these late learners. But some more recent research has suggested that maybe those earlier studies just didn’t give the learners enough time to learn their L2 — of course, their mental grammar wasn’t native-like if they hadn’t been learning the language for very long.
A 2013 study by Harriet Bowden and her colleagues looked at L1 English speakers who started learning L2 Spanish in university. They compared learners who had completed first-year Spanish to learners who had completed more than three years of university Spanish and had spent a year abroad. And they included a control group of L1 speakers of Spanish. The researchers presented Spanish sentences that violated syntactic expectations about word order like these ones. Sentence 1, “I have to run many miles this week” has the expected word order, while sentence 2, “I have to miles many run this week” is unexpected in its word order: the quantifier many comes after the noun miles, and that whole complement phrase comes before the verb run. Sentence 2 is ungrammatical in Spanish.
As you’d expect, the native speakers of Spanish showed a P600 in response to the ungrammatical sentence. But so did the advanced L2 learners: their ERP response was the same as that of the L1 Spanish speakers. It was only the beginning learners, the ones who had had only a year of Spanish, who showed an atypical P600: it was a smaller response and more diffuse. The researchers concluded that “University foreign-language learners who take L2 classes through much of college and also study abroad for one or two semesters …show evidence of native-like brain processing of syntax” (Bowden et al., 2013: 2508).
So this study suggests that one year of studying a language maybe isn’t enough to achieve native-like fluency, and three years of study including a year of immersion allows a learner to approach native proficiency, but the researchers also wondered whether the kind of language-learning makes a difference to learners. If you’re learning a language in university, you probably spend three or four hours a week in the classroom, and maybe two or three more hours each week studying. But that’s not the only way to learn a language.
A study in Montreal looked at university students who were L1 speakers of Korean and Chinese who were enrolled in a nine-week intensive English L2 course. These learners were studying, practising, using English at least 8-10 hours a day, five days a week, for nine weeks. The researchers tested the learners on sentences with morphosyntactic violations in the tense features on the verb, like these ones:
1a. The teacher did not start the lesson / 1b. The teacher did not started the lesson.
2a. The teacher had not started the lesson / 2b. The teacher had not start the lesson.
Notice that in 1b and 2b, the verb has unexpected morphology on it. A native speaker of English would show a P600 response to 1b and 2b in comparison to 1a and 2a. In this study, the researchers measured learners’ ERP responses at the beginning of the course and after the nine weeks, and they also asked the learners to judge whether the sentences were grammatical. At the beginning of the course, none of the learners showed P600s in response to the syntactically unexpected sentences, and they also weren’t very successful at deciding whether sentences were grammatical or ungrammatical. After the nine-week course, all of the learners showed P600 responses to the syntactically unexpected sentences, and the learners who scored highest on the grammatical judgments showed the largest P600s. This study suggests that even short-term, intensive L2 learning can help a learner develop a mental grammar that approaches that of a native L1 speaker.
And the results of all of these studies tell us that L2 language learners can achieve fluency that compares to that of a native speaker; it just takes lots of training to get there!
Check Yourself
- Answer
-
"Negative transfer."
Hint: Think about the structures of French and English, and think about whether they are more similar or more different.
- Answer
-
"Negative transfer."
Hint: Think about the structures of Russian and English, and think about whether they are more similar or more different.
- Answer
-
"Positive transfer."
Hint: Think about the structures of Russian and French, and think about whether they are more similar or more different.
Psycholinguistics of Word Meanings, in Anderson's Essentials of Linguistics
Video Script
In the last unit, we suggested that intensions for word meanings might be organized in our minds in fuzzy categories. Our minds construct categories of things based on our experience in the world: each time we encounter an extension of a word, we count it as an exemplar in that fuzzy category. There is some evidence from psychology and psycholinguistics that our mind really does represent a difference between prototypical category members and peripheral members. For lots of categories, we have some instincts about what kinds of exemplars are prototypical and what kinds are peripheral. When we give somebody the name of a category and ask them to name an exemplar, people from a given language community are remarkably alike in the first things they name as exemplars. If your mental grammar for English is like mine, then perhaps your prototypical bird is a robin, your prototypical fruit is an apple, and your prototypical tool is a hammer.
In a behavioural study of word recognition, participants saw a word appear on a screen and had to say the word out loud. This is called a rapid naming task. Some of the words referred to prototypical exemplars of their particular category and some of them referred to peripheral exemplars. The prototypical and peripheral exemplars were all mixed up in the experiment, but when the researchers measured how fast people had been able to name the word that they saw, the found that people were faster to name the prototypes than the peripheral exemplars.
The same researchers used these words in a lexical decision task. In this kind of task, a word appears briefly on a screen, and the person’s job is just to decide whether it’s a word or not, and say Yes or No. So if the word pants appears on the screen, you would say “Yes”, because it’s a real word in English. But if pfonc appears, you say “no”, because that’s not a word of English. What the researchers found in the lexical decision experiment was, again, that people are fast to make a decision about a word if it refers to a prototypical category member, and slower to make the decision if the word refers to a peripheral member.
These findings indicate that the process of recognizing a word is easier and faster if that word refers to a prototype. We can interpret these findings to mean that our intensions for categories are made up of exemplars and that prototypical exemplars have a privileged position in our intensions.
So that’s a couple of examples of psycholinguistic tasks we can use to observe how words are processed in our minds: a simple naming task, and a lexical decision task. There’s an additional task that we can combine with each of these, to allow us to investigate relationships between different words. That task is called priming. A primed lexical decision task works like this: First, a word appears on the screen for a very short length of time: that word is called the prime. The prime disappears, and then a second word appears on the screen. This word is the target, and the participant makes a lexical decision about the target. The prime word can have an influence on how quickly people make their lexical decision about the target word.
For example, in one condition, the prime might be doctor and the target nurse. In another condition, the prime could be apple and the target nurse. As you might expect, people are faster to make their lexical decision to nurse when it’s primed by doctor than when it’s primed by apple. When we observe this faster lexical decision, we interpret that to mean that these two words are connected to each other in our minds.
Over the years, psychologists and psycholinguists have conducted thousands of experiments on priming, and the results of these experiments show us how words are related to each other in our minds. The scientific literature has shown priming between words that are members of the same category, for words that are synonyms, antonyms, and even for words that describe things that share attributes. For example, an orange and a baseball aren’t members of the same category, but they’re both spheres, so they can prime each other.
Looking at all these and many other priming effects, we can conclude that those semantic relationships play an important role in how the meanings of a word are organized in our minds.
Check Yourself
- Answer
-
"Broccoli."
Hint: Think about the meaning of carrot, and the meaning of broccoli, and what they share in common.
- Answer
-
"Sad."
Hint: Think about the meaning of happy, and the meaning of sad, and what they share in common.
- Answer
-
"Month."
Hint: Think about the meaning of happy, and the meaning of sad, and what they share in common.