The Language Neuroscience Podcast

Bilingualism, mind, and brain, with Ellen Bialystok

Stephen M. Wilson Season 6 Episode 36

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 1:07:20

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]

Lifespan Cognition And Development Lab

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]