The Language Neuroscience Podcast
A podcast about the scientific study of language and the brain. Neuroscientist Stephen Wilson talks with leading and up-and-coming researchers about their work and ideas. This podcast is geared to an audience of scientists who are working in the field of language neuroscience, from students to postdocs to faculty.
The Language Neuroscience Podcast
Bilingualism, mind, and brain, with Ellen Bialystok
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
In this epidode, I talk with Ellen Bialystok, Distinguished Research Professor of Psychology at York University, about her case for the “bilingual advantage”, or to be more precise, the case that growing up in a bilingual environment reconfigures mind and brain for adaptability, effiency, and resiliance.
Bialystok E, Craik FIM, Klein R, Viswanathan M. Bilingualism, aging, and cognitive control: Evidence from the Simon task. Psychol Aging 2004; 19: 290-303. [doi]
Bialystok E, Craik FIM, Freedman M. Bilingualism as a protection against the onset of symptoms of dementia. Neuropsychologia 2007; 45: 459-64. [doi]
Schweizer TA, Ware J, Fischer CE, Craik FIM, Bialystok E. Bilingualism as a contributor to cognitive reserve: Evidence from brain atrophy in Alzheimer’s disease. Cortex 2012; 48: 991-6. [doi]
Berkes M, Bialystok E, Craik FIM, Troyer A, Freedman M. Conversion of mild cognitive impairment to Alzheimer disease in monolingual and bilingual patients. Alzheimer Dis Assoc Disord 2020; 34: 225. [doi]
Bialystok E, Hawrylewicz K, Grundy JG, Chung-Fat-Yim A. The swerve: How childhood bilingualism changed from liability to benefit. Dev Psychol 2022; 58: 1429-40. [doi]
Bialystok E. Bilingualism modifies cognition through adaptation, not transfer. Trends Cogn Sci 2024; 28: 987-97. [doi]
Related papers:
Alladi S, Bak TH, Duggirala V, et al. Bilingualism delays age at onset of dementia, independent of education and immigration status. Neurology 2013; 81: 1938-44. [doi]
Zahodne LB, Schofield PW, Farrell MT, Stern Y, Manly JJ. Bilingualism does not alter cognitive decline or dementia risk among Spanish-speaking immigrants. Neuropsychology 2014; 28: 238-46. [doi]
1
00:00:00,000 --> 00:00:08,120
Welcome to the Language Neuroscience Podcast.
2
00:00:08,120 --> 00:00:10,080
This is episode 36.
3
00:00:10,080 --> 00:00:13,880
I have an amazing guest today, but first quickly I have a favor to ask.
4
00:00:13,880 --> 00:00:17,600
I know that some people have used the Language Neuroscience Podcast for teaching.
5
00:00:17,600 --> 00:00:21,400
If you've done so, could you kindly shoot me a quick email describing how?
6
00:00:21,400 --> 00:00:24,600
smwilson@uq.edu.au.
7
00:00:24,600 --> 00:00:27,560
You know how academia is, sometimes you need to document things.
8
00:00:27,560 --> 00:00:32,640
Okay, my guest today is Ellen Bialystok, Distinguished Research Professor of Psychology
9
00:00:32,640 --> 00:00:35,360
at York University in Toronto, Canada.
10
00:00:35,360 --> 00:00:38,720
Ellen is a phenomenally productive and highly influential researcher
11
00:00:38,720 --> 00:00:42,280
who has spent five decades investigating how speaking two languages
12
00:00:42,280 --> 00:00:45,960
shapes cognitive processes and brain structure across the lifespan.
13
00:00:45,960 --> 00:00:49,240
She's looked from early childhood through aging and dementia.
14
00:00:49,240 --> 00:00:52,200
She's especially well-known for her research findings that bilingualism
15
00:00:52,200 --> 00:00:55,360
confers cognitive advantages, particularly in aging,
16
00:00:55,360 --> 00:00:57,320
and may be protective against dementia.
17
00:00:57,320 --> 00:00:59,360
We're going to discuss all of this today.
18
00:00:59,360 --> 00:01:00,840
All right, let's get to it.
19
00:01:00,840 --> 00:01:02,600
Hi, Ellen, how are you?
20
00:01:02,600 --> 00:01:05,080
Hi, Stephen. Thank you for inviting me.
21
00:01:05,080 --> 00:01:09,480
Oh, yeah, it's really great to have a chance to talk with you.
22
00:01:09,480 --> 00:01:11,800
So, let's get started.
23
00:01:11,800 --> 00:01:14,920
I often like to find out about people's background
24
00:01:14,920 --> 00:01:18,200
and what led them into the kind of researcher that they've become.
25
00:01:18,200 --> 00:01:20,120
And you're particularly interesting, right?
26
00:01:20,120 --> 00:01:23,040
You've made your career studying bilingualism.
27
00:01:23,040 --> 00:01:25,800
Made your name is the world's strongest proponent
28
00:01:25,800 --> 00:01:29,560
of the hypothesis that bilingualism confers a cognitive advantage.
29
00:01:29,560 --> 00:01:32,920
And you come from a very famously bilingual country.
30
00:01:32,920 --> 00:01:36,680
So I'm curious about your own experiences with languages.
31
00:01:36,680 --> 00:01:39,720
What led you to this area of study?
32
00:01:39,720 --> 00:01:45,320
Okay, so I think the assumption in your question is that people choose
33
00:01:45,320 --> 00:01:48,920
what they're going to study, and I'm not sure I agree with that.
34
00:01:48,920 --> 00:01:54,680
I don't-- I didn't ever choose this, but you go one step
35
00:01:54,680 --> 00:01:59,240
in front of the other, and the path takes you there.
36
00:01:59,240 --> 00:02:03,720
So, I was always-- as long as I could remember--
37
00:02:03,720 --> 00:02:08,440
as a little kid, I was interested in the relationship
38
00:02:08,440 --> 00:02:09,880
between language and thought.
39
00:02:09,880 --> 00:02:13,920
I thought this was just the most baffling question.
40
00:02:13,920 --> 00:02:20,440
And I remember thinking when I was a little kid or wondering,
41
00:02:20,440 --> 00:02:24,840
if you can't speak English, can you think?
42
00:02:24,840 --> 00:02:27,560
I thought, I mean, that was, I was little, OK?
43
00:02:27,560 --> 00:02:30,880
So that was the sort of thing that was always on my mind.
44
00:02:30,880 --> 00:02:36,280
And I didn't ever choose the particular path,
45
00:02:36,280 --> 00:02:41,480
but I did make small decisions that take you to where you are.
46
00:02:41,480 --> 00:02:43,440
So, for example, I entered university.
47
00:02:43,440 --> 00:02:44,960
I studied psychology.
48
00:02:44,960 --> 00:02:47,600
It was a good thing for girls to do.
49
00:02:47,600 --> 00:02:52,280
I was interested in language.
50
00:02:52,280 --> 00:02:57,400
My graduate work was all on language acquisition in children.
51
00:02:57,400 --> 00:03:02,480
But again, from the perspective of language and thought,
52
00:03:02,480 --> 00:03:05,640
as they're learning words, do they already
53
00:03:05,640 --> 00:03:08,720
have the concepts, or does learning the word
54
00:03:08,720 --> 00:03:10,960
create an opportunity for the concept?
55
00:03:10,960 --> 00:03:14,600
Really the same question I had when I was like six.
56
00:03:14,600 --> 00:03:17,080
And now I was studying it as a graduate student.
57
00:03:17,080 --> 00:03:19,800
I was interested in that.
58
00:03:19,800 --> 00:03:22,680
But bilingualism wasn't a field.
59
00:03:22,680 --> 00:03:26,480
And it certainly wasn't anything on my radar.
60
00:03:26,480 --> 00:03:31,200
And so, the way I got into bilingualism
61
00:03:31,200 --> 00:03:34,920
was more a series of accidents.
62
00:03:34,920 --> 00:03:37,680
And once I started looking at bilingualism,
63
00:03:37,680 --> 00:03:39,960
in fact, I didn't start with bilingualism.
64
00:03:39,960 --> 00:03:44,720
I started with second language acquisition.
65
00:03:44,720 --> 00:03:47,840
Okay. And then it's just accidents all the way down.
66
00:03:47,840 --> 00:03:48,760
It's not turtles.
67
00:03:48,760 --> 00:03:50,400
It's accidents.
68
00:03:50,400 --> 00:03:53,880
So, when I got my PhD, there were no jobs.
69
00:03:53,880 --> 00:03:56,720
There were no academic positions anywhere.
70
00:03:56,720 --> 00:04:02,440
I ended up getting a job as a project director on a study
71
00:04:02,440 --> 00:04:05,120
that was already funded.
72
00:04:05,120 --> 00:04:08,280
It had no room for input.
73
00:04:08,280 --> 00:04:12,880
I was supposed to manage the project.
74
00:04:12,880 --> 00:04:18,440
And the project was a study of high school students
75
00:04:18,440 --> 00:04:23,000
in Ontario learning French as a second language.
76
00:04:23,000 --> 00:04:24,640
That was the study.
77
00:04:24,640 --> 00:04:27,120
The methodology was determined.
78
00:04:27,120 --> 00:04:29,440
The measures were determined.
79
00:04:29,440 --> 00:04:32,600
I was just supposed to manage it.
80
00:04:32,600 --> 00:04:36,880
But I'm not very good at following instructions.
81
00:04:36,880 --> 00:04:40,200
And so, I really didn't like this project.
82
00:04:40,200 --> 00:04:42,560
I mean, I really didn't like it.
83
00:04:42,560 --> 00:04:48,360
And it was really the most formative thing that ever happened
84
00:04:48,360 --> 00:04:51,080
because I had never read any of the literature
85
00:04:51,080 --> 00:04:53,160
on second language acquisition.
86
00:04:53,160 --> 00:04:55,600
So, I had to absorb all that.
87
00:04:55,600 --> 00:04:59,680
I had to somehow fulfill the mandate
88
00:04:59,680 --> 00:05:02,480
of this funded research project.
89
00:05:02,480 --> 00:05:03,520
Yep.
90
00:05:03,520 --> 00:05:07,200
But in my spare time, I was doing side projects.
91
00:05:07,200 --> 00:05:09,480
So, I was running little experiments
92
00:05:09,480 --> 00:05:14,200
on second language acquisition as I was going through this.
93
00:05:14,200 --> 00:05:18,080
So that got me into the world of second language acquisition
94
00:05:18,080 --> 00:05:20,040
and applied linguistics.
95
00:05:20,040 --> 00:05:22,040
And then from there, you just move around.
96
00:05:22,040 --> 00:05:24,920
And things become interesting.
97
00:05:24,920 --> 00:05:27,040
And you follow that lead.
98
00:05:27,040 --> 00:05:31,040
And one thing leads to the next.
99
00:05:31,040 --> 00:05:31,800
Right.
100
00:05:31,800 --> 00:05:32,760
Oh, that's fascinating.
101
00:05:32,760 --> 00:05:35,440
Yeah, I didn't mean to imply that we
102
00:05:35,440 --> 00:05:37,160
choose our field of study.
103
00:05:37,160 --> 00:05:40,760
But yes, what's the path that gets us there?
104
00:05:40,760 --> 00:05:45,000
And everybody's got a unique path.
105
00:05:45,000 --> 00:05:48,280
So, you started out studying,
106
00:05:48,280 --> 00:05:51,600
did you use those high schoolers in your earliest experiments,
107
00:05:51,600 --> 00:05:52,480
In the…?
108
00:05:52,480 --> 00:05:53,120
Oh, yeah.
109
00:05:53,120 --> 00:05:54,440
Yeah, that's where it started.
110
00:05:54,440 --> 00:05:58,520
And I was trained as a developmental psychologist.
111
00:05:58,520 --> 00:06:01,440
So, when I did get an academic position,
112
00:06:01,440 --> 00:06:02,880
I cycled back to that.
113
00:06:02,880 --> 00:06:06,040
And my research became more firmly grounded
114
00:06:06,040 --> 00:06:10,640
in developmental psychology, child development,
115
00:06:10,640 --> 00:06:14,160
some language acquisition, but not only language acquisition.
116
00:06:14,160 --> 00:06:17,040
So, I did that for a long time.
117
00:06:17,040 --> 00:06:20,440
And then bilingualism just kind of crept in accidentally.
118
00:06:20,440 --> 00:06:24,800
I don't even know if I can recall the moment.
119
00:06:24,800 --> 00:06:30,160
I suddenly realized that you could just study bilingualism.
120
00:06:30,160 --> 00:06:30,760
Right.
121
00:06:30,760 --> 00:06:34,320
But your earliest major works in bilingualism
122
00:06:34,320 --> 00:06:37,600
were in the language acquisition space, right?
123
00:06:37,600 --> 00:06:38,680
Yeah.
124
00:06:38,680 --> 00:06:40,760
Well, they were,
125
00:06:40,760 --> 00:06:45,480
I mean, I never thought of myself as a language acquisition
126
00:06:45,480 --> 00:06:50,200
researcher in the usual sense, because I'm not a linguist.
127
00:06:50,200 --> 00:06:53,200
And I think you need to know a lot more about linguistics
128
00:06:53,200 --> 00:06:58,880
than I do, to do justice to those processes.
129
00:06:58,880 --> 00:07:04,000
So, I came at it from psychology, psycholinguistics.
130
00:07:04,000 --> 00:07:07,640
We can wave your hands at grammatical structures
131
00:07:07,640 --> 00:07:11,120
and not be expected to take them too seriously,
132
00:07:11,120 --> 00:07:14,160
as I say, psycholinguists.
133
00:07:14,160 --> 00:07:15,880
And so that was always my approach.
134
00:07:15,880 --> 00:07:21,760
Well, but increasingly, if I were doing this now,
135
00:07:21,760 --> 00:07:28,240
I think a much better firmer knowledge of linguistics
136
00:07:28,240 --> 00:07:30,080
would have been important.
137
00:07:30,080 --> 00:07:30,880
Oh, I don't know.
138
00:07:30,880 --> 00:07:32,920
Sometimes I feel like it can be an impediment.
139
00:07:32,920 --> 00:07:35,120
I mean, Roger Brown, Roger Brown, right,
140
00:07:35,120 --> 00:07:36,240
was a psychologist.
141
00:07:36,240 --> 00:07:40,520
And he did the most innovative studies
142
00:07:40,520 --> 00:07:43,200
on early language acquisition, syntax in particular.
143
00:07:43,200 --> 00:07:46,720
And I actually find when I read his stuff,
144
00:07:46,720 --> 00:07:49,040
it's almost his naivety that makes it good,
145
00:07:49,040 --> 00:07:51,880
because he comes at it as an amateur, and he's just like,
146
00:07:51,880 --> 00:07:54,080
oh, wow, grammar, like, oh, what's up with that?
147
00:07:54,080 --> 00:07:57,440
And he almost approaches it as a child himself,
148
00:07:57,440 --> 00:07:59,880
you feel when I read his book.
149
00:07:59,880 --> 00:08:02,200
And it led him to think, oh, what would a child actually
150
00:08:02,200 --> 00:08:04,080
need to get this?
151
00:08:04,080 --> 00:08:04,880
Do you know what I mean?
152
00:08:04,880 --> 00:08:06,600
Absolutely, absolutely right.
153
00:08:06,600 --> 00:08:09,400
But he never would have called himself a linguist.
154
00:08:09,400 --> 00:08:10,320
No, of course not.
155
00:08:10,320 --> 00:08:11,160
That's what I'm saying.
156
00:08:11,160 --> 00:08:13,320
He's in the psychology department.
157
00:08:13,320 --> 00:08:16,800
I mean, his work is so foundational.
158
00:08:16,800 --> 00:08:21,120
Everything we know about children's language acquisition
159
00:08:21,120 --> 00:08:25,960
somehow comes back to his work, and his students,
160
00:08:25,960 --> 00:08:29,560
and those three little kids who are learning English.
161
00:08:29,560 --> 00:08:30,400
Exactly.
162
00:08:30,400 --> 00:08:31,960
So yeah, I mean, he was a psychologist, right?
163
00:08:31,960 --> 00:08:32,800
Definitely not a linguist.
164
00:08:32,800 --> 00:08:35,360
So, maybe it's OK.
165
00:08:35,360 --> 00:08:37,960
So, you know--
166
00:08:37,960 --> 00:08:40,280
I don't know that he was a social psychologist.
167
00:08:40,280 --> 00:08:42,560
He wasn't even a cognitive psychologist.
168
00:08:42,560 --> 00:08:42,760
Yeah.
169
00:08:42,760 --> 00:08:48,440
He was a social psychologist, which is pretty incredible.
170
00:08:48,440 --> 00:08:50,800
Yeah, so I think it's OK that you came to bilingualism
171
00:08:50,800 --> 00:08:53,440
as a psychologist and not as a linguist.
172
00:08:53,440 --> 00:08:53,960
Thank you.
173
00:08:53,960 --> 00:08:58,280
But when you were probably starting out,
174
00:08:58,280 --> 00:09:01,560
certainly, in the earlier in the 20th century,
175
00:09:01,560 --> 00:09:05,000
there was a widespread belief that bilingualism was detrimental
176
00:09:05,000 --> 00:09:07,080
to kids.
177
00:09:07,080 --> 00:09:10,640
Was that view still prevalent when you got into this field
178
00:09:10,640 --> 00:09:13,880
or started to create this field?
179
00:09:13,880 --> 00:09:16,960
So that line, that particular line,
180
00:09:16,960 --> 00:09:21,120
I can track it pretty closely to what's going on
181
00:09:21,120 --> 00:09:24,200
and the people who were major players in that.
182
00:09:24,200 --> 00:09:26,920
People talk about the turning point
183
00:09:26,920 --> 00:09:31,160
being the 1962 study out of Montreal,
184
00:09:31,160 --> 00:09:35,440
where Wally Lambert and Elizabeth Peal
185
00:09:35,440 --> 00:09:41,600
tried to bring some methodological rigor to the question
186
00:09:41,600 --> 00:09:45,080
and see what's really going on.
187
00:09:45,080 --> 00:09:50,320
And their famous result is that the bilingual kids
188
00:09:50,320 --> 00:09:54,680
were better on both verbal and nonverbal tasks.
189
00:09:54,680 --> 00:09:55,880
That's so.
190
00:09:55,880 --> 00:09:58,560
There is no,
191
00:09:58,560 --> 00:10:02,200
the previous people in the first half of the century,
192
00:10:02,200 --> 00:10:06,640
they called it mental retardation from bilingualism.
193
00:10:06,640 --> 00:10:08,160
They talked like that.
194
00:10:08,160 --> 00:10:09,160
Yeah.
195
00:10:09,160 --> 00:10:12,000
Peal and Lambert said there's nothing of the kind.
196
00:10:12,000 --> 00:10:14,760
In fact, the bilinguals are better.
197
00:10:14,760 --> 00:10:15,840
So, what did Peal and Lambert...
198
00:10:15,840 --> 00:10:17,600
Except they're wrong, sorry.
199
00:10:17,600 --> 00:10:19,640
I was going to say, what did they do differently
200
00:10:19,640 --> 00:10:20,560
to reach that conclusion?
201
00:10:20,560 --> 00:10:21,920
Well, they did a lot that was different,
202
00:10:21,920 --> 00:10:23,480
but there were problems in their study.
203
00:10:23,480 --> 00:10:26,520
And in fact, I recently published a paper with a couple
204
00:10:26,520 --> 00:10:32,280
of my students where we did a deep dive into the Peal
205
00:10:32,280 --> 00:10:34,600
and Lambert paper.
206
00:10:34,600 --> 00:10:40,160
And their conclusions were not right.
207
00:10:40,160 --> 00:10:44,080
The interpretation of the study and its place
208
00:10:44,080 --> 00:10:47,800
in this history is not right.
209
00:10:47,800 --> 00:10:51,360
And yet, without them, we wouldn't be where we are.
210
00:10:51,360 --> 00:10:55,200
So, we can tolerate a few errors to get
211
00:10:55,200 --> 00:10:59,000
to this more elevated point.
212
00:10:59,000 --> 00:11:02,800
Here is what was going on.
213
00:11:02,800 --> 00:11:08,320
All of the research, all of the research up to and including
214
00:11:08,320 --> 00:11:14,640
Peal and Lambert was based on standard IQ tests.
215
00:11:14,640 --> 00:11:20,320
And their description, their definition of intelligence
216
00:11:20,320 --> 00:11:24,680
was, it's your score on an IQ test.
217
00:11:24,680 --> 00:11:29,680
So, up to that point, people were giving intelligence tests
218
00:11:29,680 --> 00:11:33,760
verbal and nonverbal to kids who were monolingual
219
00:11:33,760 --> 00:11:35,320
or bilingual.
220
00:11:35,320 --> 00:11:38,760
And the bilingual kids did more poorly.
221
00:11:38,760 --> 00:11:41,320
Lots of reasons they did more poorly.
222
00:11:41,320 --> 00:11:42,240
They were immigrants.
223
00:11:42,240 --> 00:11:44,640
They didn't know the language the test was in.
224
00:11:44,640 --> 00:11:47,240
There were socioeconomic differences.
225
00:11:47,240 --> 00:11:48,600
There were all sorts of reasons.
226
00:11:48,600 --> 00:11:52,120
They did more poorly, that were never documented.
227
00:11:52,120 --> 00:11:56,400
So, one thing Peal and Lambert were really careful about,
228
00:11:56,400 --> 00:12:00,760
they thought, was matching the monolinguals and bilinguals
229
00:12:00,760 --> 00:12:03,600
on all of these extraneous factors
230
00:12:03,600 --> 00:12:07,320
like socioeconomic status and so on.
231
00:12:07,320 --> 00:12:10,400
So, they gave these carefully matched kids.
232
00:12:10,400 --> 00:12:12,120
They were 10 years old.
233
00:12:12,120 --> 00:12:15,080
These tests, the bilinguals did better.
234
00:12:15,080 --> 00:12:22,640
Now, score on an IQ test is interesting.
235
00:12:22,640 --> 00:12:25,760
But it's really just a score on a test.
236
00:12:25,760 --> 00:12:29,760
What happened afterwards when I started working
237
00:12:29,760 --> 00:12:38,040
in the area, in the '70s was not measuring kids on standard IQ
238
00:12:38,040 --> 00:12:44,040
tests, but on their performance on cognitive tasks.
239
00:12:44,040 --> 00:12:47,080
So how do you get from standard IQ tests
240
00:12:47,080 --> 00:12:50,280
to cognitive tasks with bilinguals?
241
00:12:50,280 --> 00:12:51,800
It was The Bridge.
242
00:12:51,800 --> 00:12:54,360
And I was very involved in The Bridge.
243
00:12:54,360 --> 00:13:00,000
The Bridge was once it became acceptable,
244
00:13:00,000 --> 00:13:02,320
and there's a lot of reasons they used IQ tests,
245
00:13:02,320 --> 00:13:06,920
once it became acceptable to not just use IQ tests
246
00:13:06,920 --> 00:13:10,480
to measure kids, the next chunk of research
247
00:13:10,480 --> 00:13:13,240
was on metalinguistic awareness,
248
00:13:13,240 --> 00:13:17,200
which is the perfect bridge between the two worlds.
249
00:13:17,200 --> 00:13:19,920
My hypothesis made sense.
250
00:13:19,920 --> 00:13:23,680
If bilingual kids were going to be better at anything,
251
00:13:23,680 --> 00:13:27,560
they would have better understanding of intuition
252
00:13:27,560 --> 00:13:33,320
about and awareness of language, because they spoke two of them.
253
00:13:33,320 --> 00:13:37,480
So that was the bridge between the IQ testing, which
254
00:13:37,480 --> 00:13:41,200
essentially ended with Peal and Lambert,
255
00:13:41,200 --> 00:13:44,160
and the actual cognitive tasks that didn't begin
256
00:13:44,160 --> 00:13:46,960
until really the '80s.
257
00:13:46,960 --> 00:13:51,320
So, it was that metalinguistic transition
258
00:13:51,320 --> 00:13:58,600
that started showing actual performance advantages
259
00:13:58,600 --> 00:14:00,920
for bilingual kids.
260
00:14:00,920 --> 00:14:01,640
OK.
261
00:14:01,640 --> 00:14:05,000
And do you think that the main flaw of the IQ tests
262
00:14:05,000 --> 00:14:09,080
was that, they were too influenced by language differences?
263
00:14:09,080 --> 00:14:12,200
And you needed to get away from language,
264
00:14:12,200 --> 00:14:15,960
or was it that you actually needed to focus on specifically
265
00:14:15,960 --> 00:14:16,920
they were,
266
00:14:16,920 --> 00:14:18,760
what was different about their language?
267
00:14:18,760 --> 00:14:21,360
Well, I think the first thing is you
268
00:14:21,360 --> 00:14:25,800
got to understand why this was happening.
269
00:14:25,800 --> 00:14:28,160
And it's not pretty.
270
00:14:28,160 --> 00:14:32,240
Why were people in the first half of the last century
271
00:14:32,240 --> 00:14:34,400
giving IQ tests to everybody?
272
00:14:34,400 --> 00:14:37,160
And who was behind it?
273
00:14:37,160 --> 00:14:39,920
It was the eugenicists.
274
00:14:39,920 --> 00:14:47,880
It was all followed from these really 19th century racist ideas.
275
00:14:47,880 --> 00:14:52,240
So, the 19th century eugenicists, people like Francis
276
00:14:52,240 --> 00:14:56,340
Galton and Paul Broca, they were really
277
00:14:56,340 --> 00:15:03,200
committed to this race intelligence equation.
278
00:15:03,200 --> 00:15:05,240
And they spent their careers trying
279
00:15:05,240 --> 00:15:07,280
to show how to prove it.
280
00:15:07,280 --> 00:15:09,440
So, they had these crazy ideas.
281
00:15:09,440 --> 00:15:12,920
They would do things like measure head circumference
282
00:15:12,920 --> 00:15:17,000
because bigger heads have bigger brains and bigger brains
283
00:15:17,000 --> 00:15:18,040
are smarter.
284
00:15:18,040 --> 00:15:22,880
They did all of this crazy stuff to prove their point.
285
00:15:22,880 --> 00:15:29,080
And then, in the early 1900s, Lewis Terman comes up
286
00:15:29,080 --> 00:15:31,360
with a scientific tool.
287
00:15:31,360 --> 00:15:33,480
You don't have to measure people's heads.
288
00:15:33,480 --> 00:15:35,480
You just give them this test.
289
00:15:35,480 --> 00:15:37,360
And then you know how smart they were.
290
00:15:37,360 --> 00:15:40,960
And Terman was a fellow traveler.
291
00:15:40,960 --> 00:15:44,520
He was a eugenicist just like the rest of them.
292
00:15:44,520 --> 00:15:50,400
So, the first IQ test, the Binet-Stanford IQ test,
293
00:15:50,400 --> 00:15:59,640
was that it was just leapt upon as the tool
294
00:15:59,640 --> 00:16:06,560
that would allow these people to prove their horrible ideas.
295
00:16:06,560 --> 00:16:08,840
That's why kids were given IQ tests
296
00:16:08,840 --> 00:16:11,500
because they were immigrants because they
297
00:16:11,500 --> 00:16:15,960
were racially different, all kinds of reasons.
298
00:16:15,960 --> 00:16:18,120
So that's why they were given these IQ tests.
299
00:16:18,120 --> 00:16:21,560
And they were given these tests without any attention
300
00:16:21,560 --> 00:16:24,160
to whether or not they spoke the language
301
00:16:24,160 --> 00:16:26,920
in which the questions were written.
302
00:16:26,920 --> 00:16:30,720
So, there wasn't, I don't want to go on about this too much.
303
00:16:30,720 --> 00:16:33,280
I mean, I find this history fascinating.
304
00:16:33,280 --> 00:16:37,840
There was one woman who was at Columbia University,
305
00:16:37,840 --> 00:16:38,920
who got it right.
306
00:16:38,920 --> 00:16:41,920
Name is Natalie Darcy.
307
00:16:41,920 --> 00:16:47,680
And during the '40s and '50s, she was giving IQ tests
308
00:16:47,680 --> 00:16:52,360
to kids, fully understanding everything I'm telling you.
309
00:16:52,360 --> 00:16:57,480
And she published some data from her own PhD,
310
00:16:57,480 --> 00:16:59,400
but then wrote two major reviews.
311
00:16:59,400 --> 00:17:03,840
And her conclusion was that if you give these kids
312
00:17:03,840 --> 00:17:08,480
these standardized tests, there will be no difference
313
00:17:08,480 --> 00:17:14,640
between monolingual and bilingual kids on nonverbal tests.
314
00:17:14,640 --> 00:17:18,760
But monolinguals will do better on verbal tests.
315
00:17:18,760 --> 00:17:21,920
And she was exactly right.
316
00:17:21,920 --> 00:17:24,880
So, the Peal and Lambert study didn't show that.
317
00:17:24,880 --> 00:17:26,000
Why not?
318
00:17:26,000 --> 00:17:30,880
So, in this review I do, if the paper I pull out some possible
319
00:17:30,880 --> 00:17:33,280
reasons why they got different results.
320
00:17:33,280 --> 00:17:37,840
The right result is, no difference on nonverbal tests.
321
00:17:37,840 --> 00:17:40,840
But monolinguals do better on verbal tests.
322
00:17:40,840 --> 00:17:42,600
Yeah, oh, that's fascinating.
323
00:17:42,600 --> 00:17:45,520
I mean, I'm glad you went back to this history
324
00:17:45,520 --> 00:17:49,160
because I also am a huge history of science buff.
325
00:17:49,160 --> 00:17:52,520
So, I always love to start with those foundational studies.
326
00:17:52,520 --> 00:17:54,760
Yeah, shame about the Broca contribution
327
00:17:54,760 --> 00:17:56,840
that you alluded to there.
328
00:17:56,840 --> 00:18:00,280
But yeah, so, you're making this interesting point here
329
00:18:00,280 --> 00:18:03,000
that you're kind of well associated
330
00:18:03,000 --> 00:18:05,040
with the concept of a bilingual advantage.
331
00:18:05,040 --> 00:18:06,320
But you're saying here, actually there's
332
00:18:06,320 --> 00:18:10,320
a subtle bilingual disadvantage on verbal tests. Right?
333
00:18:10,320 --> 00:18:14,640
So, bilingualism is a very good thing in many ways,
334
00:18:14,640 --> 00:18:18,200
but it's not 100% a free lunch, right?
335
00:18:18,200 --> 00:18:19,840
Exactly.
336
00:18:19,840 --> 00:18:25,640
And I actually recoil at the phrase bilingual advantage.
337
00:18:25,640 --> 00:18:28,240
For years I've been going around insisting
338
00:18:28,240 --> 00:18:30,680
that I never said that, but I did. (laughter)
339
00:18:30,680 --> 00:18:32,360
And people showed me that I did. So, OK.
340
00:18:32,360 --> 00:18:34,080
I mean, I was--
341
00:18:34,080 --> 00:18:36,320
What would you like to call it today?
342
00:18:36,320 --> 00:18:38,280
What would you prefer to call it?
343
00:18:38,280 --> 00:18:40,760
Once you label it, once you label it,
344
00:18:40,760 --> 00:18:43,480
bilingual advantage, it becomes a thing.
345
00:18:43,480 --> 00:18:45,280
And once something is a thing, you
346
00:18:45,280 --> 00:18:47,000
can go out and look for it.
347
00:18:47,000 --> 00:18:48,600
So, let me look over here.
348
00:18:48,600 --> 00:18:49,920
Nope, not there.
349
00:18:49,920 --> 00:18:50,600
How about here?
350
00:18:50,600 --> 00:18:51,120
Is it here?
351
00:18:51,120 --> 00:18:52,120
Nope, not there.
352
00:18:52,120 --> 00:18:55,440
And when you don't find it, you say it doesn't exist. Right?
353
00:18:55,440 --> 00:18:58,000
So, you can't objectify it.
354
00:18:58,000 --> 00:19:03,200
It's an incredibly reductionist approach
355
00:19:03,200 --> 00:19:07,400
to a very complex set of abilities.
356
00:19:07,400 --> 00:19:14,080
So now what I would say is, from the moment of birth,
357
00:19:14,080 --> 00:19:19,280
bilingualism and just being in a bilingual environment,
358
00:19:19,280 --> 00:19:22,040
changes mind and brain.
359
00:19:22,040 --> 00:19:26,960
Children's brain development is instantly modified
360
00:19:26,960 --> 00:19:29,280
from being in a bilingual environment.
361
00:19:29,280 --> 00:19:33,040
Children's development from the earliest days
362
00:19:33,040 --> 00:19:37,000
are changed by being in a bilingual environment.
363
00:19:37,000 --> 00:19:41,800
And these adaptations continue throughout life.
364
00:19:41,800 --> 00:19:44,160
Now, are they advantages?
365
00:19:44,160 --> 00:19:46,520
On balance, they are advantages,
366
00:19:46,520 --> 00:19:50,800
because most of what's modified is better.
367
00:19:50,800 --> 00:19:55,960
And this experience leads to a more efficient and more
368
00:19:55,960 --> 00:19:58,040
resilient brain.
369
00:19:58,040 --> 00:20:00,360
So that's an advantage.
370
00:20:00,360 --> 00:20:03,800
But it doesn't mean that if you just pull a bilingual off
371
00:20:03,800 --> 00:20:07,720
the street, they're going to perform a Stroop test
372
00:20:07,720 --> 00:20:09,400
better than somebody else.
373
00:20:09,400 --> 00:20:11,080
That's not going to happen.
374
00:20:11,080 --> 00:20:13,560
It's not that kind of advantage.
375
00:20:13,560 --> 00:20:17,440
But it's a reconfiguration of mind and brain that leads
376
00:20:17,440 --> 00:20:21,480
to more resilience, better efficiency,
377
00:20:21,480 --> 00:20:24,800
and in the end, the real advantage
378
00:20:24,800 --> 00:20:31,280
is in older age, because cognition holds on better,
379
00:20:31,280 --> 00:20:36,440
because it's more adaptable to what this individual is
380
00:20:36,440 --> 00:20:38,080
trying to do.
381
00:20:38,080 --> 00:20:40,000
So that's the bilingual advantage.
382
00:20:40,000 --> 00:20:43,840
It's not the ability to do the Stroop test.
383
00:20:43,840 --> 00:20:49,160
It's a much broader based set of processes
384
00:20:49,160 --> 00:20:54,280
that are better tuned to the cognitive and brain challenges
385
00:20:54,280 --> 00:20:55,280
throughout life.
386
00:20:55,280 --> 00:20:56,040
OK.
387
00:20:56,040 --> 00:20:57,200
That makes a lot of sense.
388
00:20:57,200 --> 00:20:59,440
And we're going to talk in just a moment about, the evidence
389
00:20:59,440 --> 00:21:04,040
that you've published for that position
390
00:21:04,040 --> 00:21:08,240
about the advantages that emerge in older age.
391
00:21:08,240 --> 00:21:09,840
But before we quite get to that,
392
00:21:09,840 --> 00:21:14,800
I think that you're talking about essentially executive
393
00:21:14,800 --> 00:21:16,960
function advantages.
394
00:21:16,960 --> 00:21:19,200
But there's also these kinds of social advantages
395
00:21:19,200 --> 00:21:21,080
to being a bilingual that I actually
396
00:21:21,080 --> 00:21:24,720
think there's a much less contentious in the modern field.
397
00:21:24,720 --> 00:21:26,600
Do you think that's the case?
398
00:21:26,600 --> 00:21:27,200
I agree.
399
00:21:27,200 --> 00:21:29,920
And people don't really talk about that.
400
00:21:29,920 --> 00:21:32,040
Is it just because we all agree on that?
401
00:21:32,040 --> 00:21:33,480
I think we do agree with it.
402
00:21:33,480 --> 00:21:39,280
But I used to end talks and arguments by saying,
403
00:21:39,280 --> 00:21:42,160
even if everything I'm saying is wrong,
404
00:21:42,160 --> 00:21:45,160
even if there's no cognitive advantage,
405
00:21:45,160 --> 00:21:49,600
even if your brains aren't, even if all of that would be the case,
406
00:21:49,600 --> 00:21:54,600
which it isn't, a bilingual can speak two languages.
407
00:21:54,600 --> 00:21:56,400
And that's huge.
408
00:21:56,400 --> 00:21:58,680
They can travel to different places.
409
00:21:58,680 --> 00:22:00,000
They can read different books.
410
00:22:00,000 --> 00:22:02,040
They can communicate with different people.
411
00:22:02,040 --> 00:22:03,760
They can order different food.
412
00:22:03,760 --> 00:22:05,960
You can't take that piece away.
413
00:22:05,960 --> 00:22:08,080
And I think that's all advantage.
414
00:22:08,080 --> 00:22:09,040
But you're right.
415
00:22:09,040 --> 00:22:10,360
Nobody talks about it.
416
00:22:10,360 --> 00:22:10,880
Well, yeah.
417
00:22:10,880 --> 00:22:14,440
I think it's because it's a kind of a silent consensus.
418
00:22:14,440 --> 00:22:16,640
And I think from my perspective, I
419
00:22:16,640 --> 00:22:19,840
would just see it as being like, you have like,
420
00:22:19,840 --> 00:22:22,480
complete insight into the differing mindsets
421
00:22:22,480 --> 00:22:26,320
in two different human cultures, to the extent
422
00:22:26,320 --> 00:22:28,960
that mindset and language are interlinked in some way, which
423
00:22:28,960 --> 00:22:32,400
I kind of think they are, although I'm into Wolfian.
424
00:22:32,400 --> 00:22:35,160
But just having that, I think, would really enhance your
425
00:22:35,160 --> 00:22:36,640
understanding of what it is to be human.
426
00:22:36,640 --> 00:22:40,960
Because you've kind of got access to two ways of being human.
427
00:22:40,960 --> 00:22:42,280
Does that resonate with you at all?
428
00:22:42,280 --> 00:22:45,000
I agree.
429
00:22:45,000 --> 00:22:45,720
I agree.
430
00:22:45,720 --> 00:22:47,800
Now, I mean, I'm a cognitive psychologist, so I'm
431
00:22:47,800 --> 00:22:51,280
going to want to track down those cognitive changes too.
432
00:22:51,280 --> 00:22:54,520
But I mean, I think if you could just
433
00:22:54,520 --> 00:23:00,680
package that piece, you know, what do we mean to have that kind of access,
434
00:23:00,680 --> 00:23:06,000
that kind of perspective, I think that's an enormous benefit?
435
00:23:06,000 --> 00:23:07,440
Yeah, and we don't study that, do we?
436
00:23:07,440 --> 00:23:08,800
Like nobody's really studied.
437
00:23:08,800 --> 00:23:09,800
There are a few studies.
438
00:23:09,800 --> 00:23:14,360
Actually, there are a few studies that look on perspective
439
00:23:14,360 --> 00:23:15,080
taking.
440
00:23:15,080 --> 00:23:18,080
The ability to take a different position, a different perspective,
441
00:23:18,080 --> 00:23:20,720
and an argument, see things in another way.
442
00:23:20,720 --> 00:23:22,560
There's some studies.
443
00:23:22,560 --> 00:23:23,720
And it's all true.
444
00:23:23,720 --> 00:23:26,960
Bilinguals are good at that.
445
00:23:26,960 --> 00:23:27,480
Yeah.
446
00:23:27,480 --> 00:23:28,920
OK, great.
447
00:23:28,920 --> 00:23:31,560
So, let's now talk about the more contentious issue, which
448
00:23:31,560 --> 00:23:34,320
Is, the cognitive or executive advantages.
449
00:23:34,320 --> 00:23:36,520
I'm probably not going to say it in quite the way
450
00:23:36,520 --> 00:23:39,640
that you'd prefer to.
451
00:23:39,640 --> 00:23:43,920
But can we start with talking about your 2004 paper
452
00:23:43,920 --> 00:23:47,880
that I think was one of the very influential ones that
453
00:23:47,880 --> 00:23:52,840
made this case where you did use the Simon task,
454
00:23:52,840 --> 00:23:56,880
and you looked at older and younger bilinguals and monolinguals.
455
00:23:56,880 --> 00:24:00,560
Can you kind of run through what you found in that paper?
456
00:24:00,560 --> 00:24:01,080
Right.
457
00:24:01,080 --> 00:24:02,600
So, that paper came out.
458
00:24:02,600 --> 00:24:05,760
I'm going to go back to this, you do the next thing,
459
00:24:05,760 --> 00:24:06,680
and then the next thing.
460
00:24:06,680 --> 00:24:12,560
So, I'd been a developmental psychologist forever.
461
00:24:12,560 --> 00:24:16,560
I'd only ever done research with children.
462
00:24:16,560 --> 00:24:23,480
And I had kind of gotten as far as I could with bilingual children.
463
00:24:23,480 --> 00:24:26,680
I was giving them all these children,
464
00:24:26,680 --> 00:24:32,520
child-appropriate tasks, where they had to do certain things
465
00:24:32,520 --> 00:24:36,360
that looked like what we would call executive functioning,
466
00:24:36,360 --> 00:24:38,400
but these were little kids.
467
00:24:38,400 --> 00:24:41,360
So, I'd gotten as far as I could with that.
468
00:24:41,360 --> 00:24:50,200
And I got this grant that included release time.
469
00:24:50,200 --> 00:25:00,320
So, I used the grant to learn how to do adult cognitive research,
470
00:25:00,320 --> 00:25:02,480
because I've never really done it.
471
00:25:02,480 --> 00:25:05,680
And I spent two years at this wonderful institute,
472
00:25:05,680 --> 00:25:08,640
the Rotman Research Institute, working with colleagues
473
00:25:08,640 --> 00:25:16,080
there, trying to figure out how we would take my research
474
00:25:16,080 --> 00:25:19,240
with children, bilingual children,
475
00:25:19,240 --> 00:25:23,680
and re-instantiate it in a meaningful way with adults,
476
00:25:23,680 --> 00:25:25,240
because nobody had ever looked at that.
477
00:25:25,240 --> 00:25:29,720
That was complete Terran Cognita.
478
00:25:29,720 --> 00:25:31,280
So, we came up with this study.
479
00:25:31,280 --> 00:25:37,280
The first study, exactly the one you mentioned,
480
00:25:37,280 --> 00:25:39,040
where we had,
481
00:25:39,040 --> 00:25:41,080
and because I wasn't in a university,
482
00:25:41,080 --> 00:25:45,440
we didn't have access to an undergraduate research pool.
483
00:25:45,440 --> 00:25:48,480
So, there was no young adults, which is…
484
00:25:48,480 --> 00:25:51,080
What's a cognitive psychologist to do? (Laughter)
485
00:25:51,080 --> 00:25:52,320
Exactly.
486
00:25:52,320 --> 00:25:55,280
That's what all cognitive psychology is based on.
487
00:25:55,280 --> 00:25:58,800
19-year-old kids taking psych 100, right?
488
00:25:58,800 --> 00:26:00,080
That's it.
489
00:26:00,080 --> 00:26:02,720
But anyway, I didn't have access to that, because I was at this other place,
490
00:26:02,720 --> 00:26:05,840
which was a geriatric hospital.
491
00:26:05,840 --> 00:26:09,280
So, I had access to older adults.
492
00:26:09,280 --> 00:26:12,800
And so, we filled it in with what we call middle-aged adults.
493
00:26:12,800 --> 00:26:14,560
And they were just people in the community
494
00:26:14,560 --> 00:26:17,680
that we got largely by word of mouth.
495
00:26:17,680 --> 00:26:21,320
And we did this study, and the results were very dramatic.
496
00:26:21,320 --> 00:26:24,280
And to this day, I'm not convinced there
497
00:26:24,280 --> 00:26:26,160
Wasn’t a problem in this study.
498
00:26:26,160 --> 00:26:28,560
The results were almost too dramatic,
499
00:26:28,560 --> 00:26:32,240
because I've replicated that basic design many times,
500
00:26:32,240 --> 00:26:36,560
and the results have never been quite as large.
501
00:26:36,560 --> 00:26:39,920
We get significant results, lots of them.
502
00:26:39,920 --> 00:26:43,240
But that was particularly eye-popping.
503
00:26:43,240 --> 00:26:45,360
Can you share the results with our listeners
504
00:26:45,360 --> 00:26:48,480
who will not necessarily have read the paper?
505
00:26:48,480 --> 00:26:49,720
OK.
506
00:26:49,720 --> 00:26:54,800
We had monolingual and bilingual participants
507
00:26:54,800 --> 00:27:01,120
who were either like 45-years-old or 75-years-old.
508
00:27:01,120 --> 00:27:03,200
And they did a Simon task, which
509
00:27:03,200 --> 00:27:06,920
is a standard executive function measure
510
00:27:06,920 --> 00:27:09,800
in the cognitive literature.
511
00:27:09,800 --> 00:27:11,960
The task is very simple.
512
00:27:11,960 --> 00:27:19,320
You have to control your response by resisting a compelling queue.
513
00:27:19,320 --> 00:27:25,040
You look at a screen, and there is do we use arrows?
514
00:27:25,040 --> 00:27:27,200
Or I don't remember what the stimuli were,
515
00:27:27,200 --> 00:27:30,280
but they might have colored squares or something.
516
00:27:30,280 --> 00:27:31,400
And you get a rule.
517
00:27:31,400 --> 00:27:34,880
If you see a red square, press the right button.
518
00:27:34,880 --> 00:27:40,240
If you see a green square, press the left button.
519
00:27:40,240 --> 00:27:41,680
So that's easy.
520
00:27:41,680 --> 00:27:47,320
Except the red and green squares appeared on one
521
00:27:47,320 --> 00:27:49,160
or the other side.
522
00:27:49,160 --> 00:27:54,200
Either matching the response key or conflicting with it.
523
00:27:54,200 --> 00:27:59,600
And overriding that position cue is incredibly difficult.
524
00:27:59,600 --> 00:28:05,080
You need a lot of control because the stimulus flashes on,
525
00:28:05,080 --> 00:28:09,520
and you automatically want to respond to the flashing
526
00:28:09,520 --> 00:28:10,720
stimulus.
527
00:28:10,720 --> 00:28:14,720
It's a much faster response than stopping
528
00:28:14,720 --> 00:28:18,280
to decide what color it is and then which key you need to press.
529
00:28:18,280 --> 00:28:21,120
So, this is a well-known effect, the Simon effect.
530
00:28:21,120 --> 00:28:25,320
And the basic idea is that the difference
531
00:28:25,320 --> 00:28:29,840
between your reaction time to respond
532
00:28:29,840 --> 00:28:33,760
to the congruent case where it's a red box
533
00:28:33,760 --> 00:28:34,800
and it's on the right,
534
00:28:34,800 --> 00:28:40,880
and that's the right key versus a red box on the left,
535
00:28:40,880 --> 00:28:42,920
where you still have to press the right key.
536
00:28:42,920 --> 00:28:47,560
The difference in time is the Simon effect.
537
00:28:47,560 --> 00:28:53,280
So, we did this, and we found that in both groups,
538
00:28:53,280 --> 00:29:00,160
the Simon effect cost was much greater for the monolinguls.
539
00:29:00,160 --> 00:29:05,000
So, the bilinguals could resolve that conflict,
540
00:29:05,000 --> 00:29:10,120
avoid the pull of the position, and respond correctly
541
00:29:10,120 --> 00:29:12,200
for much longer.
542
00:29:12,200 --> 00:29:15,160
So that was the first time this had been reported.
543
00:29:15,160 --> 00:29:17,960
And it got a lot of attention.
544
00:29:17,960 --> 00:29:18,280
Yeah.
545
00:29:18,280 --> 00:29:21,080
And it's an extremely strong effect in your paper.
546
00:29:21,080 --> 00:29:23,040
And as you mentioned, like you wonder,
547
00:29:23,040 --> 00:29:26,440
if you didn't get quite a strong and effect later,
548
00:29:26,440 --> 00:29:27,640
what kind of issues do you think
549
00:29:27,640 --> 00:29:30,600
that could have been with that first paper that led
550
00:29:30,600 --> 00:29:32,480
to such a striking finding?
551
00:29:32,480 --> 00:29:32,720
Yeah.
552
00:29:32,720 --> 00:29:36,040
I think there was something about the parameters
553
00:29:36,040 --> 00:29:38,400
we were using in the design.
554
00:29:38,400 --> 00:29:39,520
I don't know.
555
00:29:39,520 --> 00:29:44,560
We've replicated the effect just not as large.
556
00:29:44,560 --> 00:29:45,920
Mm-hmm.
557
00:29:45,920 --> 00:29:48,200
I mean, were the groups, were the monolinguals and bilinguals
558
00:29:48,200 --> 00:29:50,200
like similar in other respects?
559
00:29:50,200 --> 00:29:51,280
Maybe not.
560
00:29:51,280 --> 00:29:55,440
I maybe, you know, there, anything you don't control carefully
561
00:29:55,440 --> 00:29:58,200
enough can influence the results.
562
00:29:58,200 --> 00:30:01,960
And so, you know, maybe one day I should go back
563
00:30:01,960 --> 00:30:06,280
and dig up those data and see if I can figure out why the effect
564
00:30:06,280 --> 00:30:08,280
was so large.
565
00:30:08,280 --> 00:30:11,440
But maybe if the effect hadn't been so large,
566
00:30:11,440 --> 00:30:14,280
it would not have caught the attention of so masny people (Laughter)
567
00:30:14,280 --> 00:30:15,880
Who knows… (Laughter)
568
00:30:15,880 --> 00:30:16,480
569
00:30:16,480 --> 00:30:18,640
So maybe it's like the Peal and Lambert study.
570
00:30:18,640 --> 00:30:20,080
Like even if it wasn't perfect,
571
00:30:20,080 --> 00:30:21,760
it set the field in a helpful direction.
572
00:30:21,760 --> 00:30:23,160
That's a very good point.
573
00:30:23,160 --> 00:30:25,800
Yeah, that's what you all wanted to say, right?
574
00:30:25,800 --> 00:30:28,160
So, yeah, I mean, there's been like, you know,
575
00:30:28,160 --> 00:30:31,560
20 years of work since then, many replications,
576
00:30:31,560 --> 00:30:34,480
some failed replications we can talk about.
577
00:30:34,480 --> 00:30:39,960
But it's not like it's been just left there to stand on its own.
578
00:30:39,960 --> 00:30:44,720
So next, I wanted to talk about the 2007 paper,
579
00:30:44,720 --> 00:30:47,360
because especially with respect to people who
580
00:30:47,360 --> 00:30:50,240
listen to my podcast, it's about language and the brain.
581
00:30:50,240 --> 00:30:54,120
So, this is like the brainiest of your papers
582
00:30:54,120 --> 00:30:57,840
because it's got people with dementia.
583
00:30:57,840 --> 00:31:01,840
So, can you tell us about that 2007 paper
584
00:31:01,840 --> 00:31:03,400
and what you found there?
585
00:31:03,400 --> 00:31:07,920
Yeah, so actually, it's a perfect segue
586
00:31:07,920 --> 00:31:10,200
from what we're just talking about.
587
00:31:10,200 --> 00:31:13,520
When the 2004 paper was published,
588
00:31:13,520 --> 00:31:19,320
I got non-stop calls from international press, hundreds
589
00:31:19,320 --> 00:31:26,600
and hundreds of reporters wanted to talk about that study.
590
00:31:26,600 --> 00:31:28,680
And I mean, this, like, you know,
591
00:31:28,680 --> 00:31:31,160
Reuters called me and it was in all their papers,
592
00:31:31,160 --> 00:31:33,360
AP called me, it was in all their papers.
593
00:31:33,360 --> 00:31:37,320
So, there was huge amount of press.
594
00:31:37,320 --> 00:31:45,440
And every single science writer I spoke to, asked the same question.
595
00:31:45,440 --> 00:31:49,480
Do you think this is helpful for dementia?
596
00:31:49,480 --> 00:31:54,920
And I had to say, like, 400 times, I have no idea.
597
00:31:54,920 --> 00:31:57,000
All of our participants were healthy.
598
00:31:57,000 --> 00:32:00,040
If they weren't healthy and cognitively OK,
599
00:32:00,040 --> 00:32:02,000
they wouldn't have been in a study.
600
00:32:02,000 --> 00:32:04,760
I have no idea.
601
00:32:04,760 --> 00:32:08,760
But science writers are very smart.
602
00:32:08,760 --> 00:32:10,920
And most of them are PhDs in science.
603
00:32:10,920 --> 00:32:15,560
It's a really highly educated group at the highest level.
604
00:32:15,560 --> 00:32:21,200
So, when 300 science writers ask you the same question,
605
00:32:21,200 --> 00:32:22,360
you've got to pay attention.
606
00:32:22,360 --> 00:32:25,440
Yeah, you're like, that's the next question to work on.
607
00:32:25,440 --> 00:32:28,440
Uh-huh.
608
00:32:28,440 --> 00:32:34,720
And so, since I was already at this geriatric hospital,
609
00:32:34,720 --> 00:32:43,920
and already had access to the clinic and people who knew how to do this,
610
00:32:43,920 --> 00:32:48,080
we set up the study with the clinic.
611
00:32:48,080 --> 00:32:53,800
And the first study in that 2007 paper was very simply,
612
00:32:53,800 --> 00:32:56,880
we went through the clinic records.
613
00:32:56,880 --> 00:33:01,400
And Toronto is very, very diverse.
614
00:33:01,400 --> 00:33:05,840
You know, you just dip your hand into the population,
615
00:33:05,840 --> 00:33:10,480
and half the people you grab will be bilingual, always.
616
00:33:10,480 --> 00:33:12,320
It's just that diverse.
617
00:33:12,320 --> 00:33:17,440
So, our first study, we thought all we're going to do
618
00:33:17,440 --> 00:33:19,720
is see if there's anything there.
619
00:33:19,720 --> 00:33:22,000
So, we went through the clinic records,
620
00:33:22,000 --> 00:33:29,640
and we pulled out files for people who had no or few,
621
00:33:29,640 --> 00:33:33,720
because there's no, everybody has co-morbidities.
622
00:33:33,720 --> 00:33:38,360
But we pulled out as much as possible clear diagnoses
623
00:33:38,360 --> 00:33:42,680
of dementia without a bunch of other stuff.
624
00:33:42,680 --> 00:33:50,760
We recorded education, background, socioeconomic stuff.
625
00:33:50,760 --> 00:33:53,200
You know, what was your occupation?
626
00:33:53,200 --> 00:33:57,160
And crucially, language history.
627
00:33:57,160 --> 00:34:01,880
And the only thing that we were interested in at that point,
628
00:34:01,880 --> 00:34:06,200
in that first study, was age of diagnosis.
629
00:34:06,200 --> 00:34:07,720
Yeah.
630
00:34:07,720 --> 00:34:12,680
That makes sense as a measure, because if bilingualism is protective
631
00:34:12,680 --> 00:34:16,440
in some sense, then you might expect
632
00:34:16,440 --> 00:34:19,240
that they would be diagnosed later.
633
00:34:19,240 --> 00:34:21,080
Before we talk about what you found,
634
00:34:21,080 --> 00:34:24,800
can you talk about, was it easy to characterize bilingual status
635
00:34:24,800 --> 00:34:28,240
based on the medical record? Was that information, well recorded?
636
00:34:28,240 --> 00:34:30,240
It was not easy.
637
00:34:30,240 --> 00:34:32,920
I can say that after our study, that hospital now
638
00:34:32,920 --> 00:34:36,000
collects full-language data for everything.
639
00:34:36,000 --> 00:34:39,320
So, that was a good legacy.
640
00:34:39,320 --> 00:34:41,800
It was not easy.
641
00:34:41,800 --> 00:34:45,320
And in fact, we went through hundreds of files
642
00:34:45,320 --> 00:34:49,640
that we could not use just because we weren't sure.
643
00:34:49,640 --> 00:34:56,080
So, we only ended up reporting the data from the files
644
00:34:56,080 --> 00:35:01,400
where we were really certain about the person's background.
645
00:35:01,400 --> 00:35:07,160
And the more ambiguous cases, we just didn't want that kind of mess,
646
00:35:07,160 --> 00:35:09,320
so, we didn't include them.
647
00:35:09,320 --> 00:35:15,280
So, these were clear cases of people we really believed were
648
00:35:15,280 --> 00:35:18,080
monolingual or bilingual.
649
00:35:18,080 --> 00:35:18,920
Uh-huh.
650
00:35:18,920 --> 00:35:20,880
And what did you find?
651
00:35:20,880 --> 00:35:25,040
We found that all else being equal,
652
00:35:25,040 --> 00:35:29,200
the age at which they were first diagnosed with dementia
653
00:35:29,200 --> 00:35:34,760
was four and a half years later in bilinguals.
654
00:35:34,760 --> 00:35:36,560
That's a very large effect size.
655
00:35:36,560 --> 00:35:37,680
It's huge.
656
00:35:37,680 --> 00:35:40,120
It's huge.
657
00:35:40,120 --> 00:35:44,600
And so, you immediately must have wondered,
658
00:35:44,600 --> 00:35:49,400
and I know because you looked at the analyses from the paper,
659
00:35:49,400 --> 00:35:52,280
what differences between your groups could potentially
660
00:35:52,280 --> 00:35:55,080
explain that, apart from the difference
661
00:35:55,080 --> 00:35:57,440
that you think it is, which is the bilingualism difference?
662
00:35:57,440 --> 00:35:58,960
So, what kind of things did you investigate?
663
00:35:58,960 --> 00:36:00,240
Well, so we looked at a lot of that.
664
00:36:00,240 --> 00:36:03,600
We looked at all kinds of things about their lifestyle.
665
00:36:03,600 --> 00:36:09,680
Right after that paper came out, a group in India
666
00:36:09,680 --> 00:36:14,280
where they have a much larger population, thought,
667
00:36:14,280 --> 00:36:17,560
oh, that's really interesting, but I wonder if it holds up.
668
00:36:17,560 --> 00:36:21,120
So, they just followed our methodology in their clinic
669
00:36:21,120 --> 00:36:23,040
and they had like 1,000 patients.
670
00:36:23,040 --> 00:36:25,040
Are you talking about Alladi et al., 2013.
671
00:36:25,040 --> 00:36:26,360
Alladi and Bak
672
00:36:26,360 --> 00:36:27,360
Yeah.
673
00:36:27,360 --> 00:36:28,760
Alladi and Bak.
674
00:36:28,760 --> 00:36:30,320
Suvarna Alladi.
675
00:36:30,320 --> 00:36:33,000
She was the neurologist in this clinic.
676
00:36:33,000 --> 00:36:35,720
Thomas Bak is the neurologist also,
677
00:36:35,720 --> 00:36:38,680
but he worked in Edinburgh.
678
00:36:38,680 --> 00:36:40,720
So, they collaborate on this study,
679
00:36:40,720 --> 00:36:42,960
and they had almost 1,000 patients.
680
00:36:42,960 --> 00:36:46,760
And they found in their sample that bilingual patients
681
00:36:46,760 --> 00:36:51,720
were diagnosed for 4 and a half years later than monolinguals.
682
00:36:51,720 --> 00:36:53,440
Replicated exactly.
683
00:36:53,440 --> 00:36:58,120
There's now like 20 studies from all over the world
684
00:36:58,120 --> 00:36:59,600
that have that result.
685
00:36:59,600 --> 00:37:04,400
So, it just replicates all over the place.
686
00:37:04,400 --> 00:37:06,600
Yeah, there's definitely…
687
00:37:06,600 --> 00:37:12,600
it's definitely a lot of people have run that basic design.
688
00:37:12,600 --> 00:37:17,480
So, what kind of factors did you look at in yours?
689
00:37:17,480 --> 00:37:21,840
Like you looked at education, immigrant status.
690
00:37:21,840 --> 00:37:24,120
How did your groups, do you remember how
691
00:37:24,120 --> 00:37:26,080
your groups compared on these kind of factors?
692
00:37:26,080 --> 00:37:29,800
Yeah, well, as much as possible, they were similar.
693
00:37:29,800 --> 00:37:35,600
In fact, I remember that the monolinguals had more education
694
00:37:35,600 --> 00:37:37,760
than the bilinguals.
695
00:37:37,760 --> 00:37:40,320
But I had a story about why that was the case.
696
00:37:40,320 --> 00:37:46,040
And the reason is that this hospital is a Jewish hospital,
697
00:37:46,040 --> 00:37:51,960
most of the older bilinguals arrived after the war.
698
00:37:51,960 --> 00:37:57,720
And they didn't go to high school because of life.
699
00:37:57,720 --> 00:38:00,760
So, they were stuck in war-torn Europe,
700
00:38:00,760 --> 00:38:04,320
or they survived the war, and then they immigrated.
701
00:38:04,320 --> 00:38:09,520
So, the number of formal years of education in that group
702
00:38:09,520 --> 00:38:12,360
was significantly less than the monolinguals,
703
00:38:12,360 --> 00:38:14,720
but not because they weren't smart people.
704
00:38:14,720 --> 00:38:15,240
Yeah, exactly.
705
00:38:15,240 --> 00:38:17,560
No, that was striking when I read your paper.
706
00:38:17,560 --> 00:38:20,520
I was like, OK, this potential education
707
00:38:20,520 --> 00:38:22,480
confound goes in the wrong direction.
708
00:38:22,480 --> 00:38:25,800
I mean, it goes in, it doesn't help.
709
00:38:25,800 --> 00:38:29,080
Yeah, it's not a confound that would explain it.
710
00:38:29,080 --> 00:38:31,640
But your bilinguals were mostly immigrants,
711
00:38:31,640 --> 00:38:32,840
and your monolinguals mostly non-immigrants.
712
00:38:32,840 --> 00:38:33,680
Yeah, right.
713
00:38:33,680 --> 00:38:36,880
So, immigrants, people latched onto that,
714
00:38:36,880 --> 00:38:39,680
or the healthy immigrant effect.
715
00:38:39,680 --> 00:38:42,240
But in India, none of them were immigrants,
716
00:38:42,240 --> 00:38:44,800
and in many of the other replications,
717
00:38:44,800 --> 00:38:46,920
none of them were immigrants.
718
00:38:46,920 --> 00:38:51,400
So, it is the case that our bilinguals are usually immigrants,
719
00:38:51,400 --> 00:38:54,440
and we try to control for that.
720
00:38:54,440 --> 00:38:59,120
We did subset analyses where we only looked at non-immigrants.
721
00:38:59,120 --> 00:39:03,560
So, native bilinguals and native monolinguals,
722
00:39:03,560 --> 00:39:05,400
we got the same result.
723
00:39:05,400 --> 00:39:06,240
So, we tried.
724
00:39:06,240 --> 00:39:09,640
We tried to see if there was a confound.
725
00:39:09,640 --> 00:39:10,640
Yeah, definitely.
726
00:39:10,640 --> 00:39:13,120
It's clear that you were addressing this question
727
00:39:13,120 --> 00:39:14,280
from the outset.
728
00:39:14,280 --> 00:39:16,400
I mean, when you say the healthy immigrant effect,
729
00:39:16,400 --> 00:39:18,720
just to flesh that out for people that haven't been
730
00:39:18,720 --> 00:39:23,000
thinking about it for 20 years, what's the idea there?
731
00:39:23,000 --> 00:39:25,800
Well, I'm only saying that because we
732
00:39:25,800 --> 00:39:27,720
were told that this is a thing.
733
00:39:27,720 --> 00:39:28,240
OK.
734
00:39:28,240 --> 00:39:32,120
I was never heard of it until someone published a critique
735
00:39:32,120 --> 00:39:34,240
of actually, our next paper.
736
00:39:34,240 --> 00:39:38,040
I think the next paper was a more important paper.
737
00:39:38,040 --> 00:39:40,680
It was a more careful study.
738
00:39:40,680 --> 00:39:45,080
So, we went into a different hospital.
739
00:39:45,080 --> 00:39:48,040
And here the question was, what's going on?
740
00:39:48,040 --> 00:39:50,640
How come these bilinguals are older?
741
00:39:50,640 --> 00:39:51,160
So…
742
00:39:51,160 --> 00:39:53,600
Sorry, just for my second keep track.
743
00:39:53,600 --> 00:39:55,600
Are you talking about Schweizer et al., 2012?
744
00:39:55,600 --> 00:39:56,120
Yes. Yes. You are good. (Laughter)
745
00:39:56,120 --> 00:39:56,280
746
00:39:56,280 --> 00:39:57,080
747
00:39:57,080 --> 00:39:57,680
748
00:39:57,680 --> 00:39:58,200
749
00:39:58,200 --> 00:39:59,200
750
00:39:59,200 --> 00:39:59,800
751
00:39:59,800 --> 00:40:00,800
You read this.
752
00:40:00,800 --> 00:40:04,160
So, Schweizer, he's a neuropsychologist.
753
00:40:04,160 --> 00:40:07,120
And working in a different hospital.
754
00:40:07,120 --> 00:40:11,800
And here we turn the question around.
755
00:40:11,800 --> 00:40:17,880
So, what we wanted to do was look directly
756
00:40:17,880 --> 00:40:23,240
at the Alzheimer's neuropathology in monolingual
757
00:40:23,240 --> 00:40:25,960
and bilingual patients.
758
00:40:25,960 --> 00:40:30,800
So, this study was really carefully controlled.
759
00:40:30,800 --> 00:40:38,520
We took Alzheimer's patients from their clinic
760
00:40:38,520 --> 00:40:41,800
at this different hospital, St. Michael's Hospital.
761
00:40:41,800 --> 00:40:49,560
And we matched them on chronological age.
762
00:40:49,560 --> 00:40:54,160
So, we're not looking for bilinguals to be older.
763
00:40:54,160 --> 00:40:56,080
We're looking for
764
00:40:56,080 --> 00:40:58,400
because brains change, right?
765
00:40:58,400 --> 00:40:59,360
Brains change.
766
00:40:59,360 --> 00:41:04,400
So, we want them to be the same chronological age.
767
00:41:04,400 --> 00:41:09,000
We also wanted them to be at the same level
768
00:41:09,000 --> 00:41:14,720
of Alzheimer's pathology, clinically, clinical domain.
769
00:41:14,720 --> 00:41:18,400
So, their clinical levels were the same.
770
00:41:18,400 --> 00:41:25,600
So, if these pairs of people presented to a neurologist,
771
00:41:25,600 --> 00:41:29,320
the neurologist would not see anything different between them
772
00:41:29,320 --> 00:41:31,560
in all of the behavioral measures,
773
00:41:31,560 --> 00:41:35,760
in all of the background measures, in the demographics,
774
00:41:35,760 --> 00:41:39,200
they were matched for education and SES.
775
00:41:39,200 --> 00:41:42,000
They were completely matched.
776
00:41:42,000 --> 00:41:43,960
And then we looked at their brains.
777
00:41:43,960 --> 00:41:52,120
We had CT scans of their brains.
778
00:41:52,120 --> 00:41:57,360
And if you, and these were early Alzheimer's patients,
779
00:41:57,360 --> 00:42:03,560
so, the main accumulation of the beta-amyloid
780
00:42:03,560 --> 00:42:07,600
and all of their problems are in the hippocampal area,
781
00:42:07,600 --> 00:42:11,720
In the temporal lobe, which is why memory goes first.
782
00:42:11,720 --> 00:42:17,760
And when we compared side by side the CT scans
783
00:42:17,760 --> 00:42:20,200
for the monolinguals and bilinguals,
784
00:42:20,200 --> 00:42:24,800
the bilinguals had significantly more deterioration.
785
00:42:24,800 --> 00:42:26,200
They had worse brains.
786
00:42:26,200 --> 00:42:26,700
Yeah.
787
00:42:26,700 --> 00:42:29,220
The pathology was more advanced.
788
00:42:29,220 --> 00:42:33,860
And yet despite the pathology being more advanced,
789
00:42:33,860 --> 00:42:39,380
they presented at the same level as these monolinguals
790
00:42:39,380 --> 00:42:42,260
with less advanced disease.
791
00:42:42,260 --> 00:42:42,760
Yes.
792
00:42:42,760 --> 00:42:43,100
OK.
793
00:42:43,100 --> 00:42:46,700
So that really brings home your interpretation, right?
794
00:42:46,700 --> 00:42:49,500
You're not saying that bilingualism stops the brain
795
00:42:49,500 --> 00:42:52,180
from undergoing pathological process of aging.
796
00:42:52,180 --> 00:42:55,220
You're saying that a bilingual brain is better
797
00:42:55,220 --> 00:42:59,260
positioned to handle the pathological changes of aging,
798
00:42:59,260 --> 00:43:00,320
yeah?
799
00:43:00,320 --> 00:43:01,580
Exactly.
800
00:43:01,580 --> 00:43:03,780
That's exactly right.
801
00:43:03,780 --> 00:43:06,260
And then that's why it's important
802
00:43:06,260 --> 00:43:13,260
to distinguish between age of onset studies
803
00:43:13,260 --> 00:43:17,660
like this and incidence studies.
804
00:43:17,660 --> 00:43:21,260
So, some people turn to incidence studies and say, look,
805
00:43:21,260 --> 00:43:24,260
bilinguals are getting Alzheimer's.
806
00:43:24,260 --> 00:43:26,020
Of course they are.
807
00:43:26,020 --> 00:43:31,780
It doesn't stop the pathology, but it gives resources
808
00:43:31,780 --> 00:43:36,660
for the individual to cope with the pathology
809
00:43:36,660 --> 00:43:43,620
and hold on to better cognitive functioning for longer.
810
00:43:43,620 --> 00:43:44,300
Right.
811
00:43:44,300 --> 00:43:45,180
Yeah.
812
00:43:45,180 --> 00:43:49,620
So, I think that it's probably a fair
813
00:43:49,620 --> 00:43:51,660
I'm not sure, correct me if I'm wrong,
814
00:43:51,660 --> 00:43:52,860
but my reading of the literature
815
00:43:52,860 --> 00:43:56,260
Is, there's quite a few solid replications
816
00:43:56,260 --> 00:43:57,740
that use a similar approach to you
817
00:43:57,740 --> 00:43:59,940
where you look at age of onset.
818
00:43:59,940 --> 00:44:02,900
And then there's several large longitudinal studies
819
00:44:02,900 --> 00:44:06,540
that often don't replicate your finding,
820
00:44:06,540 --> 00:44:07,900
and they don't find difference between
821
00:44:07,900 --> 00:44:09,620
monolingual and bilingual groups.
822
00:44:09,620 --> 00:44:12,420
And you've made the case that,
823
00:44:12,420 --> 00:44:14,940
I mean, you've basically pointed out
824
00:44:14,940 --> 00:44:16,820
flaws in those longitudinal studies,
825
00:44:16,820 --> 00:44:18,700
including what you just said.
826
00:44:18,700 --> 00:44:21,260
It's not that you're saying that you'll never get dementia.
827
00:44:21,260 --> 00:44:24,100
It's that it'll be later.
828
00:44:24,100 --> 00:44:26,860
But yeah, so I think like--
829
00:44:26,860 --> 00:44:30,300
to me, one of the most strongest longitudinal studies
830
00:44:30,300 --> 00:44:33,580
that I read was the Zahodne et al., 2014.
831
00:44:33,580 --> 00:44:36,180
I'm sure you read it because you've written about it.
832
00:44:36,180 --> 00:44:38,660
And I found that study interesting to talk about,
833
00:44:38,660 --> 00:44:42,340
because I think it's a good study that doesn't support
834
00:44:42,340 --> 00:44:45,140
your findings, but also does have some silver linings
835
00:44:45,140 --> 00:44:47,220
for your perspective at the same time.
836
00:44:47,220 --> 00:44:49,340
But this study totally supports it.
837
00:44:49,340 --> 00:44:50,700
OK, so tell us about that.
838
00:44:50,700 --> 00:44:51,200
Yeah.
839
00:44:51,200 --> 00:44:52,780
There are two things about that study.
840
00:44:52,780 --> 00:44:55,180
First of all, there were no monolinguals.
841
00:44:55,180 --> 00:44:56,940
They say there were, but they weren't.
842
00:44:56,940 --> 00:45:00,300
The study was done in New York City in Spanish Harlem.
843
00:45:00,300 --> 00:45:02,700
People lived there for 50 years.
844
00:45:02,700 --> 00:45:07,260
And in the study, they were designated as Spanish monolingual.
845
00:45:07,260 --> 00:45:10,460
So, after 50 years, you're going to pick up something.
846
00:45:10,460 --> 00:45:12,460
So, there really were no monolinguals.
847
00:45:12,460 --> 00:45:15,100
But more importantly than that, they
848
00:45:15,100 --> 00:45:18,180
designed it as an incidence study.
849
00:45:18,180 --> 00:45:20,700
And there was no difference in incidence.
850
00:45:20,700 --> 00:45:25,340
But deep buried in the paper is the fact
851
00:45:25,340 --> 00:45:30,140
that the bilinguals were older when they were diagnosed.
852
00:45:30,140 --> 00:45:36,340
Yeah, that's true in the simple model with no variance.
853
00:45:36,340 --> 00:45:37,260
They were older.
854
00:45:37,260 --> 00:45:38,340
And that's all I'm saying.
855
00:45:38,340 --> 00:45:42,500
They're going to be older, not that they're not going to succumb.
856
00:45:42,500 --> 00:45:46,780
But they're going to be older when it becomes diapparent
857
00:45:46,780 --> 00:45:48,380
and then diagnosed.
858
00:45:48,380 --> 00:45:51,460
Yeah, the study, just to fill in people
859
00:45:51,460 --> 00:45:56,660
that haven't read it recently, you would have read it
860
00:45:56,660 --> 00:45:57,180
a while ago.
861
00:45:57,180 --> 00:45:59,260
I've read it recently.
862
00:45:59,260 --> 00:46:02,740
So, the immigrants came from Latin America,
863
00:46:02,740 --> 00:46:05,300
and they were all native Spanish speakers,
864
00:46:05,300 --> 00:46:07,300
and they varied in their English speaking ability.
865
00:46:07,300 --> 00:46:08,300
And that was their contrast.
866
00:46:08,300 --> 00:46:10,740
It was people that said that they spoke English well,
867
00:46:10,740 --> 00:46:12,420
all the way down to people who said they spoke English,
868
00:46:12,420 --> 00:46:13,300
not at all.
869
00:46:13,300 --> 00:46:14,700
Your point is that, well, if you've
870
00:46:14,700 --> 00:46:17,580
been in America for like 30, 40 years,
871
00:46:17,580 --> 00:46:21,220
you probably don't speak English not at all.
872
00:46:21,220 --> 00:46:25,420
So yes, it's not the same kind of contrast that you've made.
873
00:46:25,420 --> 00:46:29,820
But yeah, there was definitely trends in their data
874
00:46:29,820 --> 00:46:32,340
that the more bilingual,
875
00:46:32,340 --> 00:46:33,700
that's kind of same more bilingual.
876
00:46:33,700 --> 00:46:37,500
The more bilingual individuals did better.
877
00:46:37,500 --> 00:46:39,460
Well, they definitely did better on memory and executive
878
00:46:39,460 --> 00:46:40,300
function, interestingly.
879
00:46:40,300 --> 00:46:41,100
Yeah, they did.
880
00:46:41,100 --> 00:46:47,060
I don't think their paper contradict anything that I've said.
881
00:46:47,060 --> 00:46:49,620
I think it, well, I mean, it wasn't statistically
882
00:46:49,620 --> 00:46:53,260
significant for the thing that you care about the most.
883
00:46:53,260 --> 00:46:56,260
But there was a trend in that direction.
884
00:46:56,260 --> 00:47:00,580
And there was, yeah, there were significant effects
885
00:47:00,580 --> 00:47:03,020
on memory and executive function in their bilingual.
886
00:47:03,020 --> 00:47:04,500
So, I think it had a lot of,
887
00:47:04,500 --> 00:47:06,300
I think it's an, I brought it up because I think it's
888
00:47:06,300 --> 00:47:10,100
an example of a non-replication that nevertheless has
889
00:47:10,100 --> 00:47:12,100
a lot of silver linings if you…
890
00:47:12,100 --> 00:47:14,700
I've never thought that was…
891
00:47:14,700 --> 00:47:17,980
I thought the way they wrote it up and pitched, interpreted
892
00:47:17,980 --> 00:47:22,060
their results, was problematic for what I'm saying.
893
00:47:22,060 --> 00:47:24,220
But I never thought their data were.
894
00:47:24,220 --> 00:47:25,940
Yeah, no, it's a well done study.
895
00:47:25,940 --> 00:47:28,340
And they just, yeah, they just took a different…
896
00:47:28,340 --> 00:47:29,580
yeah, they went a different direction
897
00:47:29,580 --> 00:47:30,380
with their interpretation.
898
00:47:30,380 --> 00:47:32,580
Yaakov Stern, he is the best, right?
899
00:47:32,580 --> 00:47:35,100
He knows how to do this stuff.
900
00:47:35,100 --> 00:47:40,380
But, so I think there's a lot of issues there that
901
00:47:40,380 --> 00:47:43,100
need to be taken into account.
902
00:47:43,100 --> 00:47:44,660
Yeah, no, for sure.
903
00:47:44,660 --> 00:47:48,180
I mean, yeah, so you brought up earlier
904
00:47:48,180 --> 00:47:51,620
like your, later, your current perspective
905
00:47:51,620 --> 00:47:53,340
on how this all works, right?
906
00:47:53,340 --> 00:47:56,740
So maybe you'd be great to circle back to that now.
907
00:47:56,740 --> 00:47:58,900
Like, so we've kind of, we're in this situation
908
00:47:58,900 --> 00:48:01,940
where you've made these seminal findings.
909
00:48:01,940 --> 00:48:04,420
They've been discussed.
910
00:48:04,420 --> 00:48:07,500
Obviously, people have got different perspectives.
911
00:48:07,500 --> 00:48:11,660
And then now you're in this recent TICS paper
912
00:48:11,660 --> 00:48:13,660
that you were mentioning before…
913
00:48:13,660 --> 00:48:15,980
I'm talking about before…
914
00:48:15,980 --> 00:48:18,780
you've kind of started to sharpen, I'd
915
00:48:18,780 --> 00:48:22,340
say, your theory of how bilingualism
916
00:48:22,340 --> 00:48:24,620
confers an advantage.
917
00:48:24,620 --> 00:48:26,700
Would you say that, is that like in response
918
00:48:26,700 --> 00:48:29,740
to the 20 years of literature, or is it just really
919
00:48:29,740 --> 00:48:32,900
your own thinking evolving over time?
920
00:48:32,900 --> 00:48:34,460
Oh, it's both, really.
921
00:48:34,460 --> 00:48:40,020
I mean, you know, you're always responding to what's out there.
922
00:48:40,020 --> 00:48:42,580
Well, can I just add one more thing
923
00:48:42,580 --> 00:48:45,300
before we leave the Alzheimer's piece.
924
00:48:45,300 --> 00:48:48,420
There was one other study we did that I thought
925
00:48:48,420 --> 00:48:51,860
was a really important piece of the argument.
926
00:48:51,860 --> 00:48:53,100
It's a study I did
927
00:48:53,100 --> 00:48:54,700
I don't remember when it was published.
928
00:48:54,700 --> 00:48:59,340
The first author was my student, Matthias Berkes.
929
00:48:59,340 --> 00:49:02,220
And what he showed, he was also done
930
00:49:02,220 --> 00:49:08,940
through records, he showed that monolinguals and bilingual.
931
00:49:08,940 --> 00:49:12,420
We've talked about how bilinguals hold on longer,
932
00:49:12,420 --> 00:49:17,540
diagnosed later, have worse brains for the same cognitive level.
933
00:49:17,540 --> 00:49:24,420
But then, when they inevitably decline, it's more precipitous.
934
00:49:24,420 --> 00:49:28,460
And that makes perfect sense if the idea is they've
935
00:49:28,460 --> 00:49:31,420
been holding back the dam.
936
00:49:31,420 --> 00:49:33,660
You can only do it for so long.
937
00:49:33,660 --> 00:49:38,140
So, what have bilinguals been doing before they were diagnosed,
938
00:49:38,140 --> 00:49:40,500
you know, all that stuff?
939
00:49:40,500 --> 00:49:46,660
Their minds and brains have been somehow compensating, coping
940
00:49:46,660 --> 00:49:49,700
in ways that they could only do up to a point.
941
00:49:49,700 --> 00:49:53,180
So that's the bilingualism piece.
942
00:49:53,180 --> 00:49:55,900
So, what did you guys find in that paper about the rate
943
00:49:55,900 --> 00:49:56,420
of decline?
944
00:49:56,420 --> 00:49:57,580
Can you just clarify that?
945
00:49:57,580 --> 00:49:59,980
Yeah, the rate of, we looked at,
946
00:49:59,980 --> 00:50:02,660
we took rate to be the time it took
947
00:50:02,660 --> 00:50:11,100
to go from a visual diagnosis of MCI to full-blown Alzheimer's.
948
00:50:11,100 --> 00:50:16,460
And it was significantly shorter for the bilinguals.
949
00:50:16,460 --> 00:50:22,420
So, once they're on the train, the decline was steeper.
950
00:50:22,420 --> 00:50:24,340
OK, that's fascinating.
951
00:50:24,340 --> 00:50:28,100
So why do you think the advantage didn't continue to haul?
952
00:50:28,100 --> 00:50:31,060
Like, it's almost like after the dam burst,
953
00:50:31,060 --> 00:50:32,340
there was no longer an advantage.
954
00:50:32,340 --> 00:50:33,780
And the monolinguals caught up.
955
00:50:33,780 --> 00:50:36,540
Or did they not fully catch up?
956
00:50:36,540 --> 00:50:38,740
I think everybody meets at an end point.
957
00:50:38,740 --> 00:50:40,020
Everybody meets an end point.
958
00:50:40,020 --> 00:50:43,540
But do the bilinguals remain preserved at…
959
00:50:43,540 --> 00:50:44,860
Up and up to a point.
960
00:50:44,860 --> 00:50:50,740
The early and mid-stages of the disease, they do.
961
00:50:50,740 --> 00:50:56,020
We tried at one point testing more advanced patients,
962
00:50:56,020 --> 00:50:57,300
but it was impossible.
963
00:50:57,300 --> 00:50:57,800
Oh, yeah.
964
00:50:57,800 --> 00:51:00,260
So, the prediction is that once you get
965
00:51:00,260 --> 00:51:04,260
beyond the sort of moderate disease level,
966
00:51:04,260 --> 00:51:09,420
if you would be able to continue giving them detailed tests,
967
00:51:09,420 --> 00:51:14,260
the bilingual gap would close down because they're catching up.
968
00:51:14,260 --> 00:51:15,740
That's the prediction.
969
00:51:15,740 --> 00:51:17,060
But it was terrible.
970
00:51:17,060 --> 00:51:20,420
I mean, we had one time a wonderful research assistant,
971
00:51:20,420 --> 00:51:23,020
and we sent her into people's homes,
972
00:51:23,020 --> 00:51:25,660
where there was a patient with Alzheimer's
973
00:51:25,660 --> 00:51:28,340
and she had all these tests she was supposed to give them.
974
00:51:28,340 --> 00:51:29,580
It was hopeless. Couldn’t do them.
975
00:51:29,580 --> 00:51:31,380
Yeah.
976
00:51:31,380 --> 00:51:32,860
I mean, I understand.
977
00:51:32,860 --> 00:51:35,580
I mean, I did my postdoc in a dementia center,
978
00:51:35,580 --> 00:51:37,900
and it was very eye-opening.
979
00:51:37,900 --> 00:51:43,820
And yeah, it's a very difficult population to work with.
980
00:51:43,820 --> 00:51:46,140
Right now, I work primarily with Stroke,
981
00:51:46,140 --> 00:51:50,540
which is much more oddly a happy population to work with,
982
00:51:50,540 --> 00:51:52,900
because they're tending to be on a positive trajectory.
983
00:51:52,900 --> 00:51:54,140
They're on a positive trajectory. Yeah, they recover. Exactly.
984
00:51:54,140 --> 00:51:59,140
And yeah, and dementia that's certainly not the case.
985
00:51:59,140 --> 00:51:59,660
OK, yeah.
986
00:51:59,660 --> 00:52:03,700
So yeah, thanks for sharing that piece of the puzzle as well.
987
00:52:03,700 --> 00:52:07,380
I think that's an important part of the story.
988
00:52:07,380 --> 00:52:10,060
So maybe the last paper I'd like to talk about,
989
00:52:10,060 --> 00:52:12,140
yeah, you already mentioned it earlier.
990
00:52:12,140 --> 00:52:15,780
It's this recent TICS paper where you talk about mechanisms.
991
00:52:15,780 --> 00:52:17,980
And first of all, I mean, I kind of just
992
00:52:17,980 --> 00:52:23,420
wowed by the fact that you're still writing theoretical,
993
00:52:23,420 --> 00:52:26,900
After so long in this field, you are just not,
994
00:52:26,900 --> 00:52:27,740
you can't hold still.
995
00:52:27,740 --> 00:52:30,900
You're still like revising your views and so on.
996
00:52:30,900 --> 00:52:32,220
That's very cool.
997
00:52:32,220 --> 00:52:35,580
So, in this paper, you kind of contrast
998
00:52:35,580 --> 00:52:39,220
a transfer view of what the mechanism could be
999
00:52:39,220 --> 00:52:41,500
versus an adaptation view.
1000
00:52:41,500 --> 00:52:43,340
And I know you will mentioned this earlier,
1001
00:52:43,340 --> 00:52:45,380
but can you kind of flesh out those two possibilities
1002
00:52:45,380 --> 00:52:49,140
and explain why you come down on one side rather than the other?
1003
00:52:49,140 --> 00:52:50,220
OK.
1004
00:52:50,220 --> 00:52:56,060
So, when the research started accumulating,
1005
00:52:56,060 --> 00:52:59,420
and there was some interest in figuring out why bilinguals
1006
00:52:59,420 --> 00:53:04,100
are doing these asks better, executive function tasks,
1007
00:53:04,100 --> 00:53:07,300
better than, why would being bilingual
1008
00:53:07,300 --> 00:53:10,100
make you better at a Stroop task or a Simon task?
1009
00:53:10,100 --> 00:53:13,300
I mean, it doesn't really make a lot of sense.
1010
00:53:13,300 --> 00:53:18,660
So, in the, I guess early '80s to mid '80s,
1011
00:53:18,660 --> 00:53:23,340
there was a huge amount of psycholinguistic research
1012
00:53:23,340 --> 00:53:27,340
showing that both languages in a bilingual mind
1013
00:53:27,340 --> 00:53:28,700
are constantly active.
1014
00:53:28,700 --> 00:53:30,780
There's no language switch.
1015
00:53:30,780 --> 00:53:35,020
So, there's a constant need to select,
1016
00:53:35,020 --> 00:53:36,900
because bilinguals don't make mistakes.
1017
00:53:36,900 --> 00:53:39,900
They don't slip into the wrong language.
1018
00:53:39,900 --> 00:53:43,020
There is a constant need to select.
1019
00:53:43,020 --> 00:53:47,780
And people talked about that as inhibition.
1020
00:53:47,780 --> 00:53:52,180
And this was the language that these researchers were using
1021
00:53:52,180 --> 00:53:57,020
that bilinguals inhibit the unwanted language.
1022
00:53:57,020 --> 00:54:00,580
So, I thought, it just kind of made sense to me.
1023
00:54:00,580 --> 00:54:04,220
All right, they're inhibiting the unwanted language.
1024
00:54:04,220 --> 00:54:06,860
I'm not a linguist, again.
1025
00:54:06,860 --> 00:54:10,540
But when I see the performance on these tests,
1026
00:54:10,540 --> 00:54:16,220
like the Stroop task, you have to inhibit reading the word
1027
00:54:16,220 --> 00:54:19,940
in the Simon task that I described earlier.
1028
00:54:19,940 --> 00:54:24,220
You have to inhibit responding to the side
1029
00:54:24,220 --> 00:54:26,220
that's flashing with the stimulus.
1030
00:54:26,220 --> 00:54:29,900
So, I thought, maybe what's happening
1031
00:54:29,900 --> 00:54:37,380
is that constantly having to inhibit the non-target language
1032
00:54:37,380 --> 00:54:41,620
just makes bilinguals better at inhibition.
1033
00:54:41,620 --> 00:54:44,300
So that was my first guess.
1034
00:54:44,300 --> 00:54:48,220
I didn't think it was a terrible guess, but turns out to be wrong.
1035
00:54:48,220 --> 00:54:55,060
But anyway, I thought inhibition somehow is boosted,
1036
00:54:55,060 --> 00:55:00,420
and it's then more available no matter what you need to inhibit.
1037
00:55:00,420 --> 00:55:04,940
But then lots of stuff challenged that story.
1038
00:55:04,940 --> 00:55:07,900
The story didn't hold up for lots of reasons.
1039
00:55:07,900 --> 00:55:10,660
And we take too much time to go through it all.
1040
00:55:10,660 --> 00:55:11,460
Just briefly, though.
1041
00:55:11,460 --> 00:55:14,260
I think the important reasons, yeah.
1042
00:55:14,260 --> 00:55:14,900
All right.
1043
00:55:14,900 --> 00:55:17,140
So, here's a couple of examples.
1044
00:55:17,140 --> 00:55:24,780
I mentioned earlier that bilingualism begins
1045
00:55:24,780 --> 00:55:30,500
to impact mind and brain from the moment of birth.
1046
00:55:30,500 --> 00:55:35,660
There’re cool, cool studies with infants, six-month-old infants,
1047
00:55:35,660 --> 00:55:38,500
eight-month-old infants.
1048
00:55:38,500 --> 00:55:41,140
Infants don't speak.
1049
00:55:41,140 --> 00:55:44,220
They're not inhibiting anything.
1050
00:55:44,220 --> 00:55:50,860
So, it isn't just you try to tap down French,
1051
00:55:50,860 --> 00:55:52,620
because we're speaking English.
1052
00:55:52,620 --> 00:55:54,260
So, inhibition didn't work there.
1053
00:55:54,260 --> 00:56:01,420
Second, a lot of the studies first with children
1054
00:56:01,420 --> 00:56:05,100
and then with young adults showed that inhibition
1055
00:56:05,100 --> 00:56:07,540
isn't one thing anyway.
1056
00:56:07,540 --> 00:56:10,860
And so, if you're looking for differences
1057
00:56:10,860 --> 00:56:13,260
between monolinguals and bilinguals
1058
00:56:13,260 --> 00:56:18,540
and your hypothesis is that it's inhibition,
1059
00:56:18,540 --> 00:56:20,660
you're not going to find it all the time
1060
00:56:20,660 --> 00:56:23,260
because there's different kinds of inhibition.
1061
00:56:23,260 --> 00:56:30,700
And so, we then started to zoom in on a more detailed understanding
1062
00:56:30,700 --> 00:56:36,380
of what sort of inhibition in a task is handled better
1063
00:56:36,380 --> 00:56:37,380
by bilinguals.
1064
00:56:37,380 --> 00:56:41,620
So, inhibition is a big thing didn't work.
1065
00:56:41,620 --> 00:56:44,220
Then you get to the older adults.
1066
00:56:44,220 --> 00:56:48,940
Why would inhibiting a language all your life
1067
00:56:48,940 --> 00:56:54,140
help preserve your cognitive level in older age
1068
00:56:54,140 --> 00:56:56,860
and then into dementia?
1069
00:56:56,860 --> 00:56:58,980
Again, it didn't make any sense.
1070
00:56:58,980 --> 00:57:01,900
So, the inhibition story didn't hold together,
1071
00:57:01,900 --> 00:57:07,420
but that was the one that became adopted by the field.
1072
00:57:07,420 --> 00:57:08,660
And it is really easy.
1073
00:57:08,660 --> 00:57:10,340
I mean, you got all these students out there
1074
00:57:10,340 --> 00:57:12,220
and they have to do experiments.
1075
00:57:12,220 --> 00:57:14,540
So, they say, I'm going to give monolinguals
1076
00:57:14,540 --> 00:57:16,860
and bilinguals an inhibition task.
1077
00:57:16,860 --> 00:57:21,300
Oh, look, there's no difference because it was the wrong prediction.
1078
00:57:21,300 --> 00:57:24,420
So, inhibition took on a life of its own
1079
00:57:24,420 --> 00:57:29,500
and I tried to understand what was going on.
1080
00:57:29,500 --> 00:57:31,020
And I just--
1081
00:57:31,020 --> 00:57:33,740
Sorry, can I just make sure I'm understanding, right?
1082
00:57:33,740 --> 00:57:38,300
So, this view is like that the driving effect
1083
00:57:38,300 --> 00:57:43,380
is like the transfer of some skill from being bilingual
1084
00:57:43,380 --> 00:57:47,060
and the skill and question that the transfer is inhibition.
1085
00:57:47,060 --> 00:57:50,980
And if that's your view, then if that's not the explanation,
1086
00:57:50,980 --> 00:57:53,380
then that might actually explain some of the no results
1087
00:57:53,380 --> 00:57:56,940
from people who've argued against an advantage
1088
00:57:56,940 --> 00:58:01,020
for bilinguals in aging if I'm phrasing it OK.
1089
00:58:01,020 --> 00:58:02,940
Because they might be looking at things
1090
00:58:02,940 --> 00:58:05,700
where you actually wouldn't necessarily even predict
1091
00:58:05,700 --> 00:58:07,980
that the bilinguals' advantage would show itself.
1092
00:58:07,980 --> 00:58:09,860
So, you're going to have a different concept
1093
00:58:09,860 --> 00:58:12,740
of where the advantage comes from, if this is,
1094
00:58:12,740 --> 00:58:13,820
Exactly right.
1095
00:58:13,820 --> 00:58:17,580
And if inhibition isn't the thing, then you shouldn't expect
1096
00:58:17,580 --> 00:58:20,740
bilinguals to transfer inhibition to everything
1097
00:58:20,740 --> 00:58:23,300
you throw at them that happens to require
1098
00:58:23,300 --> 00:58:24,780
what you want to call inhibition.
1099
00:58:24,780 --> 00:58:25,620
OK, so…
1100
00:58:25,620 --> 00:58:30,900
So, looking in the wrong place, I now describe that as looking
1101
00:58:30,900 --> 00:58:32,460
in the wrong place.
1102
00:58:32,460 --> 00:58:33,700
It wasn't inhibition.
1103
00:58:33,700 --> 00:58:35,780
But there was something.
1104
00:58:35,780 --> 00:58:39,220
And it really started to make sense when we
1105
00:58:39,220 --> 00:58:41,380
added the brain science.
1106
00:58:41,380 --> 00:58:44,420
Because bilinguals,
1107
00:58:44,420 --> 00:58:47,660
I mean, there's a lot of stuff on structural differences
1108
00:58:47,660 --> 00:58:50,420
between monolingual and bilingual brains.
1109
00:58:50,420 --> 00:58:54,660
But I think the more important stuff is the functional stuff.
1110
00:58:54,660 --> 00:59:00,900
So once neuroimaging was added to these studies,
1111
00:59:00,900 --> 00:59:05,740
it was very clear that even if the behavioral results
1112
00:59:05,740 --> 00:59:12,700
were identical, the functional connectivity was not.
1113
00:59:12,700 --> 00:59:18,260
Bunch of EEG studies that clearly show whether or not
1114
00:59:18,260 --> 00:59:22,100
there are differences in accuracy and reaction time
1115
00:59:22,100 --> 00:59:28,620
bilinguals are performing these tasks with less effort.
1116
00:59:28,620 --> 00:59:34,300
It was less effortful, even if the behavioral outcomes
1117
00:59:34,300 --> 00:59:35,780
were the same.
1118
00:59:35,780 --> 00:59:40,220
And then you've got other studies with connectivity
1119
00:59:40,220 --> 00:59:45,500
in fMRI, white matter tracts, functional connectivity
1120
00:59:45,500 --> 00:59:47,900
at rest, really interesting stuff.
1121
00:59:47,900 --> 00:59:51,300
bilingual brains were wired differently.
1122
00:59:51,300 --> 00:59:53,180
Yeah, thank you, Daniela Perani's
1123
00:59:53,180 --> 00:59:54,460
work there, probably.
1124
00:59:54,460 --> 00:59:56,460
Yeah, yeah, yeah, yeah, yeah, exactly.
1125
00:59:56,460 --> 00:59:56,980
Yeah.
1126
00:59:56,980 --> 00:59:57,460
OK.
1127
00:59:57,460 --> 01:00:01,220
So, if the brain is wired differently,
1128
01:00:01,220 --> 01:00:05,420
it has a different level of preparedness
1129
01:00:05,420 --> 01:00:09,180
for the tasks in which it will engage.
1130
01:00:09,180 --> 01:00:11,980
So, the crucial thing is no longer
1131
01:00:11,980 --> 01:00:14,220
who can do a Stroop task faster.
1132
01:00:14,220 --> 01:00:15,780
That's not important.
1133
01:00:15,780 --> 01:00:20,660
The crucial thing is what does the preparedness of the brain
1134
01:00:20,660 --> 01:00:22,260
look like?
1135
01:00:22,260 --> 01:00:26,780
And this, I think, is the key to understanding
1136
01:00:26,780 --> 01:00:30,420
effects in older age and dementia.
1137
01:00:30,420 --> 01:00:34,820
The bilingual brain has better connectivity
1138
01:00:34,820 --> 01:00:40,980
and a more active state at rest.
1139
01:00:40,980 --> 01:00:45,540
And when it's doing a task, the EEG tells us
1140
01:00:45,540 --> 01:00:48,100
they don't have to try as hard.
1141
01:00:48,100 --> 01:00:52,140
So, this is a preservation on the brain
1142
01:00:52,140 --> 01:00:56,620
that takes them further into aging without losing
1143
01:00:56,620 --> 01:01:00,140
cognitive position.
1144
01:01:00,140 --> 01:01:02,900
OK, so that's the adaptation account
1145
01:01:02,900 --> 01:01:05,140
that the brain is that the bilingual environment
1146
01:01:05,140 --> 01:01:08,900
creates a more efficient, intentional system
1147
01:01:08,900 --> 01:01:11,460
that plays out in many different contexts,
1148
01:01:11,460 --> 01:01:13,580
not just in inhibition tasks.
1149
01:01:13,580 --> 01:01:18,500
And does it matter then, what kind of bilingual you are?
1150
01:01:18,500 --> 01:01:20,460
Like, is it better to be a simultaneous
1151
01:01:20,460 --> 01:01:22,740
bilingual or a sequential bilingual, for instance?
1152
01:01:22,740 --> 01:01:24,340
Like, what do you think about that?
1153
01:01:24,340 --> 01:01:24,940
Yeah.
1154
01:01:24,940 --> 01:01:28,300
So, this is also a big deal.
1155
01:01:28,300 --> 01:01:35,820
Because bilingualism is not a categorical concept, right?
1156
01:01:35,820 --> 01:01:40,220
And in the early studies we pretended it was,
1157
01:01:40,220 --> 01:01:43,860
the current research is much smarter
1158
01:01:43,860 --> 01:01:50,220
and treats it as a continuum with lots of ways
1159
01:01:50,220 --> 01:01:54,660
of measuring what moves you along the continuum.
1160
01:01:54,660 --> 01:01:59,220
So, lots of factors make bilingual experiences
1161
01:01:59,220 --> 01:02:04,860
across different people, age of onset, level of proficiency
1162
01:02:04,860 --> 01:02:07,500
who you speak to, in what context.
1163
01:02:07,500 --> 01:02:09,500
You know, all that matters.
1164
01:02:09,500 --> 01:02:11,940
And we haven't even talked about multilingualism, right?
1165
01:02:11,940 --> 01:02:14,660
I mean, that's not going to make it.
1166
01:02:14,660 --> 01:02:15,820
Let's not.
1167
01:02:15,820 --> 01:02:16,340
Not today.
1168
01:02:16,340 --> 01:02:21,620
The bottom line is bilingualism is a continuum
1169
01:02:21,620 --> 01:02:26,420
and degree of bilingualism matters greatly, depending,
1170
01:02:26,420 --> 01:02:28,380
you know, however you measure it.
1171
01:02:28,380 --> 01:02:32,700
And studies in the last five or so years
1172
01:02:32,700 --> 01:02:36,740
look at regressions or correlations
1173
01:02:36,740 --> 01:02:40,020
between degree of bilingualism and the outcomes.
1174
01:02:40,020 --> 01:02:46,340
So, it's calibrated to degree of bilingual experience.
1175
01:02:46,340 --> 01:02:49,740
OK, so the particular bilingual environment certainly
1176
01:02:49,740 --> 01:02:53,100
is going to matter according to this theory.
1177
01:02:53,100 --> 01:02:57,020
OK, I have one last question for you, if I may.
1178
01:02:57,020 --> 01:03:00,100
Do you think that, you know, given your findings,
1179
01:03:00,100 --> 01:03:02,500
do you think that a prospective intervention
1180
01:03:02,500 --> 01:03:08,300
would be worth exploring? Could learning a language later in life,
1181
01:03:08,300 --> 01:03:10,100
Specifically, because you want
1182
01:03:10,100 --> 01:03:12,540
to stave off dementia?
1183
01:03:12,540 --> 01:03:15,300
Is that a clinical trial that you'd like to see done?
1184
01:03:15,300 --> 01:03:16,060
Right.
1185
01:03:16,060 --> 01:03:18,660
So, I have two answers to your question.
1186
01:03:18,660 --> 01:03:23,500
First, we did a little training study that was very cute.
1187
01:03:23,500 --> 01:03:26,460
This was with the colleague of mine, Jed Meltzer.
1188
01:03:26,460 --> 01:03:28,500
So, we had older adults who were monolingual
1189
01:03:28,500 --> 01:03:29,820
when we trained them on stuff.
1190
01:03:29,820 --> 01:03:33,860
And by the end of the training study,
1191
01:03:33,860 --> 01:03:41,100
the group who spent 16 weeks learning Spanish on Duolingo
1192
01:03:41,100 --> 01:03:42,580
performed better on these tasks.
1193
01:03:42,580 --> 01:03:44,420
So that's kind of a demonstration.
1194
01:03:44,420 --> 01:03:45,900
But my real answer is,
1195
01:03:46,340 --> 01:03:51,980
what's hard for your brain is good for your brain.
1196
01:03:51,980 --> 01:03:55,420
So, to stave off dementia, we know all the things we need to do.
1197
01:03:55,420 --> 01:03:57,260
Everybody knows the list.
1198
01:03:57,260 --> 01:04:01,500
You've got to be active, engaged, busy, you've got to read,
1199
01:04:01,500 --> 01:04:05,020
you've got to do cross-word puzzles, you've got to do all these things.
1200
01:04:05,020 --> 01:04:08,980
And the reason is you've got to keep your brain alive.
1201
01:04:08,980 --> 01:04:15,060
So, bilingualism, learning a language in older age, is hard.
1202
01:04:15,060 --> 01:04:19,660
So, learning a language in older age is good for your brain.
1203
01:04:19,660 --> 01:04:21,980
But it will not make you bilingual.
1204
01:04:21,980 --> 01:04:24,980
It will be good for your brain just because it's
1205
01:04:24,980 --> 01:04:29,820
one of those activities that you need to keep doing to keep
1206
01:04:29,820 --> 01:04:32,580
your brain alive.
1207
01:04:32,580 --> 01:04:33,420
Yes, but…
1208
01:04:33,420 --> 01:04:37,460
OK, so you don't see that as creating a bilingual environment
1209
01:04:37,460 --> 01:04:42,620
of the sort that you theorize creates that lifelong protection.
1210
01:04:42,620 --> 01:04:47,220
Well, people want to know, it will help.
1211
01:04:47,220 --> 01:04:47,820
I mean…
1212
01:04:47,820 --> 01:04:48,860
Yeah, I understand.
1213
01:04:48,860 --> 01:04:49,860
That is there.
1214
01:04:49,860 --> 01:04:52,660
But people ask me this question because they
1215
01:04:52,660 --> 01:04:54,900
want to become bilingual.
1216
01:04:54,900 --> 01:04:57,620
And so, I think it just kind of be a little more realistic
1217
01:04:57,620 --> 01:05:01,540
about what's possible, but absolutely learn another languages.
1218
01:05:01,540 --> 01:05:03,180
But it's a great activity.
1219
01:05:03,180 --> 01:05:05,660
Yeah, but you don't, yeah.
1220
01:05:05,660 --> 01:05:09,260
And I guess what I'm saying is like you think it has virtue
1221
01:05:09,260 --> 01:05:12,620
but because it's cognitively challenging, but it doesn't create
1222
01:05:12,620 --> 01:05:15,260
the kind of bilingual brain that you're talking about
1223
01:05:15,260 --> 01:05:17,980
that would have that long term resilience built in.
1224
01:05:17,980 --> 01:05:18,980
Yes, that is it.
1225
01:05:18,980 --> 01:05:24,100
OK, so it's too late for all of us that have failed to become bilingual
1226
01:05:24,100 --> 01:05:26,460
when we had the chance.
1227
01:05:26,460 --> 01:05:27,300
That's OK.
1228
01:05:27,300 --> 01:05:29,140
I can live with that.
1229
01:05:29,140 --> 01:05:31,220
It's still good for your brain.
1230
01:05:31,220 --> 01:05:32,740
Yeah.
1231
01:05:32,740 --> 01:05:35,300
It probably is.
1232
01:05:35,300 --> 01:05:37,980
All right, well, thank you so much for your time.
1233
01:05:37,980 --> 01:05:41,020
What was really fun to talk about these things.
1234
01:05:41,020 --> 01:05:44,140
Bilingualism is just, like whenever I ever tell anybody
1235
01:05:44,140 --> 01:05:48,020
in like civilian life that I work on language in the brain,
1236
01:05:48,020 --> 01:05:51,180
they almost always ask questions related to bilingualism
1237
01:05:51,180 --> 01:05:51,900
or multilingualism.
1238
01:05:51,900 --> 01:05:52,420
Oh, really?
1239
01:05:52,420 --> 01:05:53,700
For some reason, that's the most fascinating,
1240
01:05:53,700 --> 01:05:55,860
and that's the most asked topic that I get.
1241
01:05:55,860 --> 01:05:56,420
What do they ask?
1242
01:05:56,420 --> 01:05:57,660
What are the questions they ask you?
1243
01:05:57,660 --> 01:06:00,020
The most common question would probably
1244
01:06:00,020 --> 01:06:02,940
be like if I speak two languages, are they
1245
01:06:02,940 --> 01:06:04,860
processed in different parts of my brain?
1246
01:06:04,860 --> 01:06:05,860
I think
1247
01:06:05,860 --> 01:06:09,180
oddly enough, I think that's not a question
1248
01:06:09,180 --> 01:06:11,980
that we actually fully know the answer to.
1249
01:06:11,980 --> 01:06:14,180
But that's probably a topic for another day.
1250
01:06:14,180 --> 01:06:16,100
But yeah, people love this topic.
1251
01:06:16,100 --> 01:06:19,300
And I don't think I've done a podcast on it before.
1252
01:06:19,300 --> 01:06:20,300
So, I think people are going to be interested in this.
1253
01:06:20,300 --> 01:06:21,980
Well, you have had Cathy Price.
1254
01:06:21,980 --> 01:06:24,580
And she's done some really important bilingualism stuff.
1255
01:06:24,580 --> 01:06:27,340
Yeah, I think, but I don't think we necessarily
1256
01:06:27,340 --> 01:06:28,780
talked about that on the podcast.
1257
01:06:28,780 --> 01:06:32,180
But yeah, no, I mean, Cathy Price has worked on everything.
1258
01:06:32,180 --> 01:06:33,780
I could have a podcast where I'd only
1259
01:06:33,780 --> 01:06:36,260
talk to Cathy Price each week, and we'd
1260
01:06:36,260 --> 01:06:38,180
stay busy for a long time.
1261
01:06:38,180 --> 01:06:40,180
But yeah, I think people are going to really find this
1262
01:06:40,180 --> 01:06:41,540
interesting as I did.
1263
01:06:41,540 --> 01:06:43,260
So, thank you so much.
1264
01:06:43,260 --> 01:06:44,300
Thank you very much for inviting me.
1265
01:06:44,300 --> 01:06:45,420
I really enjoyed it.
1266
01:06:45,420 --> 01:06:46,060
Yeah, all right.
1267
01:06:46,060 --> 01:06:46,780
Well take care.
1268
01:06:46,780 --> 01:06:49,260
And I look forward to hopefully meeting you
1269
01:06:49,260 --> 01:06:50,980
in real life at some point.
1270
01:06:50,980 --> 01:06:52,020
OK, bye-bye.
1271
01:06:52,020 --> 01:06:53,300
OK, bye.
1272
01:06:53,300 --> 01:06:55,500
OK, well, that's it for episode 36.
1273
01:06:55,500 --> 01:06:57,260
Thank you, Ellen, for joining me on the podcast.
1274
01:06:57,260 --> 01:06:59,060
This was a really fun conversation.
1275
01:06:59,060 --> 01:07:01,460
I've linked the papers we discussed in the show notes.
1276
01:07:01,460 --> 01:07:05,980
And on the podcast website at langneurosci.org/podcast.
1277
01:07:05,980 --> 01:07:07,940
Thank you to Marcia Petyt for editing the transcript
1278
01:07:07,940 --> 01:07:10,340
of this episode and thank you all for listening.
1279
01:07:10,340 --> 01:07:12,420
Bye for now and see you next time.
1280
01:07:12,420 --> 01:07:15,420
[MUSIC PLAYING]
1281
01:07:15,420 --> 01:07:45,420
[MUSIC PLAYING]