kor_sent
stringlengths
1
5.4k
โŒ€
eng_sent
stringlengths
1
17.5k
โŒ€
source
stringclasses
4 values
similarity
float64
0.12
1
โŒ€
from
stringclasses
5 values
__index_level_0__
float64
0
1.78M
โŒ€
์—ฌ๋Ÿฌ๋ถ„์ด ํ…Œ๋Ÿฌ๋ฆฌ์ŠคํŠธ๋ผ๋ฉด ์ˆ˜๋…„๊ฐ„ ์‹คํ—˜์‹ค์— ๋ฐ•ํ˜€ ์—ฐ๊ตฌํ•ด์•ผ ํ•˜๊ณ  ๋˜ ์„ฑ๊ณตํ•˜๋ฆฌ๋ผ๋Š” ๋ณด์žฅ๋„ ์—†๋Š” ์—„์ฒญ ํž˜๋“  ์—ฐ๊ตฌ๋ฅผ ํ•˜์‹œ๊ฒ ์–ด์š”?
So if you're a terrorist and have to choose that will require years of meticulous lab work and still might not pan out, or just blowing stuff up?
IWSLT2017
null
null
null
์•„๋‹ˆ๋ฉด ๊ทธ์ € ํญํŒŒ์‹œ์ผœ๋ฒ„๋ฆฌ๊ฒ ์–ด์š”?
You'll probably choose the later.
IWSLT2017
null
null
null
์•„๋งˆ๋„ ํญํŒŒ์‹œํ‚ค๊ฒ ์ฃ . ์ด๋ก ์ƒ์œผ๋กœ ์—ญ ์œ ์ „์ž ๋“œ๋ผ์ด๋ธŒ๋ฅผ
This is especially true because at least in theory, it should be pretty easy to build what's called a reversal drive.
IWSLT2017
null
null
null
๋งŒ๋“ค๊ธฐ๋Š” ์‰ฝ๊ฑฐ๋“ ์š”. ์ฒซ ๋ฒˆ์งธ ์œ ์ „์ž ๋“œ๋ผ์ด๋ธŒ๊ฐ€ ๋ณ€ํ˜•ํ•œ ์œ ์ „์ž ํ˜•์งˆ์„ ๋ฎ์–ด์“ฐ๋Š” ๊ฑฐ์ฃ .
That's one that basically overwrites the change made by the first gene drive.
IWSLT2017
null
null
null
๊ทธ๋ž˜์„œ ์œ ์ „ํ˜•์งˆ ๋ณ€ํ™” ๊ฒฐ๊ณผ๊ฐ€ ๋ง˜์— ๋“ค์ง€ ์•Š์œผ๋ฉด ์ทจ์†Œ์‹œํ‚ค๋Š” ๋‘ ๋ฒˆ์งธ ์œ ์ „์ž ๋“œ๋ผ์ด๋ธŒ๋ฅผ ๋ฐฉ์‚ฌ์‹œํ‚ค๋ฉด ๋ฉ๋‹ˆ๋‹ค. ์ ์–ด๋„ ์ด๋ก ์ƒ์œผ๋กœ๋Š” ๊ฐ€๋Šฅํ•ด์š”.
So if you don't like the effects of a change, you can just release a second drive that will cancel it out, at least in theory.
IWSLT2017
null
null
null
๊ทธ๋ ‡๋‹ค๋ฉด ์šฐ๋ฆฌ๋Š” ์–ด๋–ป๊ฒŒ ๋ ๊นŒ์š”?
OK, so where does this leave us?
IWSLT2017
null
null
null
์ด์ œ ์ธ๊ฐ„์€ ๋œป๋Œ€๋กœ ์ƒ๋ฌผ ์ข… ์ „์ฒด๋ฅผ ๋ฐ”๊ฟ€ ์ˆ˜ ์žˆ๋Š” ๋Šฅ๋ ฅ์ด ์žˆ์Šต๋‹ˆ๋‹ค.
We now have the ability to change entire species at will.
IWSLT2017
null
null
null
๊ทธ๋ž˜์•ผ ํ• ๊นŒ์š”?
Should we?
IWSLT2017
null
null
null
์ด์ œ ์‹ ์˜ ์˜์—ญ์— ๋„๋‹ฌํ•œ ๊ฒƒ์ผ๊นŒ์š”?
Are we gods now?
IWSLT2017
null
null
null
์ €๋Š” ๊ฑฐ๊ธฐ๊นŒ์ง€๋Š” ์ž˜ ๋ชจ๋ฅด๊ฒ ์Šต๋‹ˆ๋‹ค๋งŒ
I'm not sure I'd say that.
IWSLT2017
null
null
null
์ด ์ ์€ ๋ง์”€ ๋“œ๋ฆด ์ˆ˜ ์žˆ์–ด์š”. ์ผ๋ถ€ ๋˜‘๋˜‘ํ•œ ์‚ฌ๋žŒ๋“ค์ด ์œ ์ „์ž ๋“œ๋ผ์ด๋ธŒ๋ฅผ ๊ทœ์ œํ•  ๋ฐฉ์•ˆ์„ ๋…ผ์˜ ์ค‘์ด๊ณ 
But I would say this: first, some very smart people are even now debating how to regulate gene drives.
IWSLT2017
null
null
null
๋˜ ๋‹ค๋ฅธ ๋˜‘๋˜‘ํ•œ ์‚ฌ๋žŒ๋“ค์€ ์œ ์ „์ž ๋“œ๋ผ์ด๋ธŒ๊ฐ€ ์ž‘๋™ํ•˜๊ณ  ๋ช‡ ์„ธ๋Œ€ ํ›„์— ์ž๊ธฐ๊ทœ์ œ๋ฅผ ํ•˜๊ณ  ์†Œ๋ฉธ๋  ์ˆ˜ ์žˆ๋Š” ์•ˆ์ „์žฅ์น˜๋ฅผ ๊ฐœ๋ฐœํ•˜๊ธฐ ์œ„ํ•ด ๋…ธ๋ ฅ์ค‘์ด์ง€์š”.
At the same time, some other very smart people are working hard to create safeguards, like gene drives that self-regulate or peter out after a few generations.
IWSLT2017
null
null
null
ํ›Œ๋ฅญํ•œ ์ผ์ž…๋‹ˆ๋‹ค.
That's great.
IWSLT2017
null
null
null
ํ•˜์ง€๋งŒ ์ด ๊ธฐ์ˆ ์— ๋Œ€ํ•ด ๋” ๋งŽ์ด ๋…ผ์˜ํ•ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.
But this technology still requires a conversation.
IWSLT2017
null
null
null
์œ ์ „์ž ๋“œ๋ผ์ด๋ธŒ์˜ ํŠน์„ฑ์ƒ ์ „์„ธ๊ณ„์ ์œผ๋กœ ๋…ผ์˜๊ฐ€ ์ด๋ฃจ์–ด์ ธ์•ผ ํ•ด์š”.
And given the nature of gene drives, that conversation has to be global.
IWSLT2017
null
null
null
์ผ€๋ƒ๋Š” ์œ ์ „์ž ๋“œ๋ผ์ด๋ธŒ์— ์ฐฌ์„ฑํ•˜๋Š”๋ฐ ํƒ„์ž๋‹ˆ์•„๋Š” ๋ฐ˜๋Œ€ํ•˜๋Š” ๊ฒฝ์šฐ
What if Kenya wants to use a drive but Tanzania doesn't?
IWSLT2017
null
null
null
๋ฐฉ์‚ฌํ• ์ง€ ๋ง์ง€ ๋ˆ„๊ฐ€ ๊ฒฐ์ •ํ• ๊นŒ์š”?
Who decides whether to release a gene drive that can fly?
IWSLT2017
null
null
null
์ €๋Š” ๊ทธ์— ๋Œ€ํ•œ ๋‹ต์€ ๋ชจ๋ฅด๊ฒ ์Šต๋‹ˆ๋‹ค.
I don't have the answer to that question.
IWSLT2017
null
null
null
์šฐ๋ฆฌ๊ฐ€ ํ•  ์ผ์€ ์œ„ํ—˜์š”์†Œ์™€ ํ˜œํƒ์— ๋Œ€ํ•ด ์†”์งํžˆ ์–˜๊ธฐํ•˜๊ณ  ์„ ํƒ์— ๋Œ€ํ•œ ์ฑ…์ž„์„ ํ•จ๊ป˜ ์ง€๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
All we can do going forward, I think, is talk honestly about the risks and benefits and take responsibility for our choices.
IWSLT2017
null
null
null
๊ทธ ์„ ํƒ์€ ์œ ์ „์ž ๋“œ๋ผ์ด๋ธŒ๋ฅผ ์‚ฌ์šฉํ•˜๊ฒ ๋‹ค๋Š” ์„ ํƒ ๋ฟ ์•„๋‹ˆ๋ผ ์‚ฌ์šฉํ•˜์ง€ ์•Š์„ ์„ ํƒ๋„ ์˜๋ฏธํ•˜์ง€์š”.
By that I mean, not just the choice to use a gene drive, but also the choice not to use one.
IWSLT2017
null
null
null
์ธ๊ฐ„์€ ํ˜„์ƒ์œ ์ง€๋ฅผ ์œ„ํ•ด ๊ฐ€์žฅ ์•ˆ์ „ํ•œ ์„ ํƒ์„ ํ•˜๋Š” ๊ฒฝํ–ฅ์ด ์žˆ์Šต๋‹ˆ๋‹ค๋งŒ ํ•ญ์ƒ ๊ทธ๋Ÿฐ ๊ฒƒ์€ ์•„๋‹™๋‹ˆ๋‹ค.
Humans have a tendency to assume that the safest option is to preserve the status quo.
IWSLT2017
null
null
null
์œ ์ „์ž ๋“œ๋ผ์ด๋ธŒ์—๋Š” ์œ„ํ—˜์š”์†Œ๊ฐ€ ์žˆ๊ณ  ์ด ์ ์„ ๋…ผ์˜ํ•  ํ•„์š”๋Š” ์žˆ์Šต๋‹ˆ๋‹ค.
But that's not always the case.
IWSLT2017
null
null
null
ํ•˜์ง€๋งŒ ๋ง๋ผ๋ฆฌ์•„๋Š” ํ˜„์žฌ ๋งค์ผ ์ฒœ ๋ช…์˜ ์ƒ๋ช…์„ ์•—์•„๊ฐ€๊ณ  ์žˆ์–ด์š”. ์šฐ๋ฆฌ๊ฐ€ ๋ง๋ผ๋ฆฌ์•„๋ฅผ ํ‡ด์น˜ํ•˜๋ ค๊ณ  ๋ฟŒ๋ฆฌ๋Š” ์‚ด์ถฉ์ œ๋Š”
Gene drives have risks, and those need to be discussed, but malaria exists now and kills 1,000 people a day.
IWSLT2017
null
null
null
์–‘์„œ๋ฅ˜, ์กฐ๋ฅ˜ ๋“ฑ ๋‹ค๋ฅธ ์ƒ๋ฌผ ์ข…์— ์‹ฌ๊ฐํ•œ ํ”ผํ•ด๋ฅผ ์ค๋‹ˆ๋‹ค. ์•ž์œผ๋กœ ์—ฌ๋Ÿฌ๋ถ„์€ ์œ ์ „์ž ๋“œ๋ผ์ด๋ธŒ์— ๋Œ€ํ•œ ์†Œ์‹์„
To combat it, we spray pesticides that do grave damage to other species, including amphibians and birds.
IWSLT2017
null
null
null
๋ถ„๋ช…ํžˆ ๋“ฃ๊ฒŒ ๋˜์‹ค ๊ฑฐ์˜ˆ์š”. ๊ทธ ๋•Œ ์ด ๋ง์„ ๊ธฐ์–ตํ•ด ์ฃผ์‹œ๊ธฐ ๋ฐ”๋ž๋‹ˆ๋‹ค.
So when you hear about gene drives in the coming months, and trust me, you will be hearing about them, remember that.
IWSLT2017
null
null
null
โ€œํ–‰๋™ํ•˜๋Š” ๊ฒƒ์ด ๋‘๋ ค์šธ ์ˆ˜ ์žˆ๋‹ค. ํ•˜์ง€๋งŒ ๋•Œ๋•Œ๋กœ ํ–‰๋™ํ•˜๋Š” ๊ฒƒ์ด ํ–‰๋™ํ•˜์ง€ ์•Š๋Š” ๊ฒƒ๋ณด๋‹ค ๋‚ซ๋‹ค.โ€
It can be frightening to act, but sometimes, not acting is worse.
IWSLT2017
null
null
null
์—ฌ๋Ÿฌ๋ถ„์ด ์–ด๋–ค ์žฅ์น˜๋ฅผ ๋ฐœ๋ช…ํ–ˆ๋Š”๋ฐ ์ €์˜ ๊ธฐ์–ต๊ณผ ๊ฟˆ, ์ƒ๊ฐ์„ ๊ธฐ๋กํ•ด์„œ ์—ฌ๋Ÿฌ๋ถ„์˜ ๋‡Œ๋กœ ๋ณด๋‚ผ ์ˆ˜ ์žˆ๋‹ค๊ณ  ์ƒ์ƒํ•ด ๋ณด์„ธ์š”.
Imagine that you invented a device that can record my memories, my dreams, my ideas, and transmit them to your brain.
IWSLT2017
null
null
null
์ •๋ง ํš๊ธฐ์ ์ธ ๊ธฐ์ˆ ์ผ ๊ฑฐ์˜ˆ์š”.
That would be a game-changing technology, right?
IWSLT2017
null
null
null
๊ทธ๋Ÿฐ๋ฐ ์šฐ๋ฆฌ๋Š” ์ด๋ฏธ ์ด๋Ÿฐ ์žฅ์น˜๋ฅผ ๊ฐ€์ง€๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์ธ๊ฐ„์˜ ์˜์‚ฌ์†Œํ†ต ์ฒด๊ณ„์™€ ํšจ๊ณผ์ ์ธ ์Šคํ† ๋ฆฌํ…”๋ง์ด๋ผ๊ณ  ๋ถ€๋ฅด๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค.
But in fact, we already possess this device, and it's called human communication system and effective storytelling.
IWSLT2017
null
null
null
์ด ์žฅ์น˜์˜ ์ž‘๋™ ์›๋ฆฌ๋ฅผ ์ดํ•ดํ•˜๋ ค๋ฉด ๋จผ์ € ๋‡Œ ์•ˆ์„ ๋ด์•ผ ํ•ฉ๋‹ˆ๋‹ค.
To understand how this device works, we have to look into our brains.
IWSLT2017
null
null
null
๊ทธ๋ฆฌ๊ณ  ์งˆ๋ฌธ์„ ์‚ด์ง ๋ฐ”๊ฟ”์•ผ ํ•ฉ๋‹ˆ๋‹ค.
And we have to formulate the question in a slightly different manner.
IWSLT2017
null
null
null
์ถ”์–ต๊ณผ ์•„์ด๋””์–ด์™€ ์—ฐ๊ด€๋œ ๋‡Œ ์†์˜ ๋‰ด๋Ÿฐ ํŒจํ„ด์ด ์–ด๋–ป๊ฒŒ ๋‡Œ๋กœ ์ „๋‹ฌ๋˜๋Š”์ง€ ๋ฌผ์–ด์•ผ ํ•ฉ๋‹ˆ๋‹ค. ์šฐ๋ฆฌ๋Š” ์˜์‚ฌ์†Œํ†ต์„ ๊ฐ€๋Šฅํ•˜๊ฒŒ ํ•˜๋Š” ๋ฐ ๋‘ ๊ฐ€์ง€ ์š”์†Œ๊ฐ€ ์žˆ๋‹ค๊ณ  ๋ด…๋‹ˆ๋‹ค.
Now we have to ask how these neuron patterns in my brain that are associated with my memories and ideas And we think there are two factors that enable us to communicate.
IWSLT2017
null
null
null
์ฒซ์งธ๋กœ, ์—ฌ๋Ÿฌ๋ถ„์€ ๋‡Œ๋Š” ์ง€๊ธˆ ์ œ๊ฐ€ ์—ฌ๋Ÿฌ๋ถ„์˜ ๋‡Œ๋กœ ๋ณด๋‚ด๋Š” ์ŒํŒŒ์™€ ๋ฌผ๋ฆฌ์ ์œผ๋กœ ๊ฒฐํ•ฉ๋˜์–ด ์žˆ์Šต๋‹ˆ๋‹ค.
First, your brain is now physically coupled to the sound wave that I'm transmitting to your brain.
IWSLT2017
null
null
null
๋‘˜์งธ๋กœ, ์šฐ๋ฆฌ๋Š” ์„œ๋กœ ์†Œํ†ตํ•  ์ˆ˜ ์žˆ๊ฒŒ ํ•ด์ฃผ๋Š” ๊ณตํ†ต์˜ ์‹ ๊ฒฝ ์กฐ์•ฝ์„ ๋งบ์—ˆ์Šต๋‹ˆ๋‹ค.
And second, we developed a common neural protocol that enabled us to communicate.
IWSLT2017
null
null
null
์–ด๋–ป๊ฒŒ ์•„๋ƒ๊ณ ์š”?
So how do we know that?
IWSLT2017
null
null
null
ํ”„๋ฆฐ์Šคํ„ด์— ์žˆ๋Š” ์ œ ์—ฐ๊ตฌ์‹ค์—์„  MRI๋กœ ์‹คํ™”๋ฅผ ๋“ฃ๊ฑฐ๋‚˜ ๋งํ•˜๋Š” ๋™์•ˆ์˜ ์‚ฌ๋žŒ๋“ค์˜ ๋‡Œ๋ฅผ ์Šค์บ”ํ•ฉ๋‹ˆ๋‹ค.
In my lab in Princeton, we bring people to the fMRI scanner and we scan their brains while they are either telling or listening to real-life stories.
IWSLT2017
null
null
null
์ €ํฌ๊ฐ€ ์“ฐ๋Š” ์ž๊ทน์ œ๊ฐ€ ์–ด๋–ค ๊ฒƒ์ธ์ง€ ๋ณด์—ฌ๋“œ๋ฆฌ๊ธฐ ์œ„ํ•ด ์ €ํฌ๊ฐ€ ์‚ฌ์šฉํ•˜๋Š” ์ด์•ผ๊ธฐ์ธ, ์œ ๋Šฅํ•œ ๋™ํ™” ๊ตฌ์—ฐ๊ฐ€ ์ง ์˜ค๊ทธ๋ ˆ์ด๋””๊ฐ€ ๊ตฌ์—ฐํ•˜๋Š” ์ด์•ผ๊ธฐ๋ฅผ 20์ดˆ๊ฐ„ ๋“ค๋ ค๋“œ๋ฆฌ๊ฒ ์Šต๋‹ˆ๋‹ค. ์ง ์˜ค๊ทธ๋ ˆ์ด๋””: ๋‚ด๊ฐ€ ์žฌ๋ฏธ์žˆ๋Š” ์ด์•ผ๊ธฐ๋ฅผ ๋“ค๋ ค์ฃผ๊ณ  ์žˆ์—ˆ๊ฑฐ๋“  ๊ทธ๋Ÿฌ๋‹ค๊ฐ€ ์‚ด์„ ์ข€ ๋ถ™์ด๊ธฐ ์‹œ์ž‘ํ–ˆ์–ด. ์ข€ ๋” ๊พธ๋ฉฐ์„œ ๋ง์ด์•ผ.
And to give you a sense of the stimulus we are using, let me play 20 seconds from a story that we used, told by a very talented storyteller, Jim O'Grady: So I'm banging out my story and I know it's good, and then I start to make it better -- by adding an element of embellishment.
IWSLT2017
null
null
null
๊ธฐ์ž๋“ค์€ ์ด๊ฑธ "ํ—›์†Œ๋ฆฌ๋ฅผ ์ง€์–ด๋‚ธ๋‹ค"๊ณ  ํ•˜๋”๋ผ๊ณ .
Reporters call this "making shit up."
IWSLT2017
null
null
null
๊ทธ ์„ ์„ ๋„˜์ง€ ๋ง๋ผ๊ณ  ํ•˜๋ฉด์„œ ๋ง์ด์•ผ.
And they recommend against crossing that line.
IWSLT2017
null
null
null
๊ทธ๋Ÿฐ๋ฐ ์–ผ๋งˆ ์ „์— ๋‚ด๊ฐ€ ๊ถŒ์œ„์žˆ๋Š” ํ•™์žฅ์ด ํŒจ์ŠคํŠธ๋ฆฌ๋กœ ๋•Œ๋ฆฌ๋Š” ๊ฑธ ๋ดค์–ด.
But I had just seen the line crossed between a high-powered dean and assault with a pastry.
IWSLT2017
null
null
null
๊ทธ๋Ÿฐ๋ฐ ๊ฝค ์žฌ๋ฏธ์žˆ๋”๋ผ๊ณ .
And I kinda liked it."
IWSLT2017
null
null
null
์šฐ๋ฆฌ ํ•œ์Šจ: ์ด์ œ ์ด๋Ÿฐ ์ด์•ผ๊ธฐ๋ฅผ ๋“ค์„ ๋•Œ ๋‡Œ์—์„œ ๋ฌด์Šจ ์ผ์ด ์ผ์–ด๋‚˜๋Š”์ง€ ๋“ค์—ฌ๋‹ค๋ด…์‹œ๋‹ค.
Uri Hasson: OK, so now let's look into your brain and see what's happening when you listen to these kinds of stories.
IWSLT2017
null
null
null
์ฒญ์ž ํ•œ ๋ช…๊ณผ ๋‘๋‡Œ ํ•œ ๋ถ€๋ถ„์œผ๋กœ ๊ฐ„๋‹จํ•˜๊ฒŒ ์‹œ์ž‘ํ•ด๋ด์š”. ๊ท€๋กœ ๋“ค์–ด์˜ค๋Š” ์†Œ๋ฆฌ๋ฅผ ์ฒ˜๋ฆฌํ•˜๋Š” ์ฒญ๊ฐ ํ”ผ์งˆ๋ถ€ํ„ฐ์š”.
And let's start simple -- let's start with one listener and one brain area: the auditory cortex that processes the sounds that come from the ear.
IWSLT2017
null
null
null
๋ณด์‹œ๋‹ค์‹œํ”ผ ๋‡Œ์˜ ์ด ์˜์—ญ์—์„œ ์ด์•ผ๊ธฐ๋ฅผ ๋“ฃ๋Š” ๋™์•ˆ ๋ฐ˜์‘์ด ์œ„์•„๋ž˜๋กœ ์›€์ง์ž…๋‹ˆ๋‹ค.
And as you can see, in this particular brain area, the responses are going up and down as the story is unfolding.
IWSLT2017
null
null
null
์ด ๋ฐ˜์‘์„ ๊ฐ€์ง€๊ณ  ๋‹ค๋ฅธ ์ฒญ์ž์˜ ๋™์ผํ•œ ๋‡Œ ๋ถ€๋ถ„์—์„œ ์ผ์–ด๋‚˜๋Š” ๋ฐ˜์‘๊ณผ ๋น„๊ตํ•ด ๋ณผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
Now we can take these responses and compare them to the responses in other listeners in the same brain area.
IWSLT2017
null
null
null
๊ทธ๋ฆฌ๊ณ  ์ฒญ์ž๋“ค ๊ฐ„์— ๋ฐ˜์‘์ด ์–ผ๋งˆ๋‚˜ ์œ ์‚ฌํ•œ์ง€ ๋ฌผ์„ ์ˆ˜ ์žˆ์ฃ .
And we can ask: How similar are the responses across all listeners?
IWSLT2017
null
null
null
์—ฌ๊ธฐ 5๋ช…์˜ ์ฒญ์ž๊ฐ€ ์žˆ์Šต๋‹ˆ๋‹ค.
So here you can see five listeners.
IWSLT2017
null
null
null
๊นœ๊นœํ•œ ๊ณณ์— ๋ˆ„์›Œ ์ด์•ผ๊ธฐ๊ฐ€ ์‹œ์ž‘๋˜๊ธฐ๋งŒ์„ ๊ธฐ๋‹ค๋ฆฌ๊ณ  ์žˆ์„ ๋•Œ๋ถ€ํ„ฐ ์ด๋ฏธ ๋‡Œ๋ฅผ ์Šค์บ”ํ•˜๊ธฐ ์‹œ์ž‘ํ•ฉ๋‹ˆ๋‹ค.
And we start to scan their brains before the story starts, when they're simply lying in the dark and waiting for the story to begin.
IWSLT2017
null
null
null
๋ณด์‹œ๋‹ค์‹œํ”ผ ๋ชจ๋‘์˜ ๋‡Œ ๋ถ€๋ถ„์˜ ๋ฐ˜์‘์ด ์œ„์•„๋ž˜๋กœ ์›€์ง์ž…๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๋ฐ˜์‘๋“ค์€ ๋งค์šฐ ๋‹ค๋ฅด๊ณ  ์ผ์น˜ํ•˜์ง€ ์•Š์ฃ .
As you can see, the brain area is going up and down in each one of them, but the responses are very different, and not in sync.
IWSLT2017
null
null
null
ํ•˜์ง€๋งŒ ์ด์•ผ๊ธฐ๊ฐ€ ์‹œ์ž‘๋˜์ž๋งˆ์ž ๋†€๋ผ์šด ์ผ์ด ์ผ์–ด๋‚ฌ์Šต๋‹ˆ๋‹ค.
However, immediately as the story is starting, something amazing is happening.
IWSLT2017
null
null
null
์ง: ๋‚ด๊ฐ€ ์žฌ๋ฏธ์žˆ๋Š” ์ด์•ผ๊ธฐ๋ฅผ ๋“ค๋ ค์ฃผ๊ณ  ์žˆ์—ˆ๊ฑฐ๋“  ๊ทธ๋Ÿฌ๋‹ค๊ฐ€ ๋‚ด๊ฐ€- ์šฐ๋ฆฌ: ๊ฐ‘์ž๊ธฐ ๋ชจ๋“  ์ฒญ์ž๋“ค์˜ ๋ฐ˜์‘์ด ์ด์•ผ๊ธฐ์— ๊ฝ‚ํ˜€์„œ ๋ชจ๋“  ์ฒญ์ž๋“ค์˜ ๋ฐ˜์‘์ด ์œ ์‚ฌํ•œ ๋ชจ์–‘์œผ๋กœ ์œ„์•„๋ž˜๋กœ ์›€์ง์ด๋Š” ๊ฑธ ๋ณผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
JO: So I'm banging out my story and I know it's good, and then I start to make it -- UH: Suddenly, you can see that the responses in all of the subjects lock to the story, and now they are going up and down in a very similar way across all listeners.
IWSLT2017
null
null
null
์ œ๊ฐ€ ๋งํ•˜๋Š” ๊ฑธ ๋“ค์„ ๋•Œ ์—ฌ๋Ÿฌ๋ถ„์˜ ๋‡Œ์—๋„ ๋˜‘๊ฐ™์€ ์ผ์ด ์ผ์–ด๋‚˜๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค.
And in fact, this is exactly what is happening now in your brains when you listen to my sound speaking.
IWSLT2017
null
null
null
์šฐ๋ฆฐ ์ด๊ฑธ "์‹ ๊ฒฝ ๋™์กฐํ™”"๋ผ ๋ถ€๋ฆ…๋‹ˆ๋‹ค.
We call this effect "neural entrainment."
IWSLT2017
null
null
null
์‹ ๊ฒฝ ๋™์กฐํ™”๊ฐ€ ๋ญ”์ง€ ์„ค๋ช…๋“œ๋ฆฌ๊ธฐ ์œ„ํ•ด์„œ ๋ฌผ๋ฆฌ์  ๋™์กฐํ™”์— ๋Œ€ํ•ด ๋จผ์ € ์„ค๋ช…๋“œ๋ฆฌ๊ฒ ์Šต๋‹ˆ๋‹ค.
And to explain to you what is neural entrainment, let me first explain what is physical entrainment.
IWSLT2017
null
null
null
์—ฌ๊ธฐ ๋ฉ”ํŠธ๋กœ๋†ˆ์ด 5๊ฐœ ์žˆ์Šต๋‹ˆ๋‹ค.
So, we'll look and see five metronomes.
IWSLT2017
null
null
null
์ด ๋ฉ”ํŠธ๋กœ๋†ˆ์„ ๋‡Œ 5๊ฐœ๋ผ๊ณ  ์ƒ๊ฐํ•˜์„ธ์š”.
Think of these five metronomes as five brains.
IWSLT2017
null
null
null
์ด์•ผ๊ธฐ๋ฅผ ๋“ค๋ ค์ฃผ๊ธฐ ์ „์˜ ์ฒญ์ž๋“ค์ฒ˜๋Ÿผ ๋ฉ”ํŠธ๋กœ๋†ˆ์€ ๋˜‘๋”ฑ๊ฑฐ๋ฆด ๊ฒ๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ์„œ๋กœ ๋‹ค๋ฅด๊ฒŒ ๋˜‘๋”ฑ๊ฑฐ๋ฆด ๊ฑฐ์˜ˆ์š”.
And similar to the listeners before the story starts, these metronomes are going to click, but they're going to click out of phase.
IWSLT2017
null
null
null
์ด์ œ ๋‘ ๊ฐœ์˜ ์›ํ†ต ์œ„์— ๋ฉ”ํŠธ๋กœ๋†ˆ์„ ์˜ฌ๋ ค๋†“์•˜์„ ๋•Œ ์–ด๋–ค ์ผ์ด ์ผ์–ด๋‚˜๋Š”์ง€ ๋ณด์„ธ์š”.
Now see what will happen when I connect them together by placing them on these two cylinders.
IWSLT2017
null
null
null
์ด์ œ ์›ํ†ต์ด ๋Œ๊ธฐ ์‹œ์ž‘ํ•ฉ๋‹ˆ๋‹ค.
Now these two cylinders start to rotate.
IWSLT2017
null
null
null
ํšŒ์ „์œผ๋กœ ์ธํ•œ ์ง„๋™์€ ๋‚˜๋ฌด๋ฅผ ํ†ตํ•ด ์ „ํ•ด์ ธ์„œ ๋ชจ๋“  ๋ฉ”ํŠธ๋กœ๋†ˆ์„ ํ•˜๋‚˜๋กœ ๋งŒ๋“ค ๊ฒ๋‹ˆ๋‹ค.
This rotation vibration is going through the wood and is going to couple all the metronomes together.
IWSLT2017
null
null
null
์ด์ œ ์†Œ๋ฆฌ๋ฅผ ๋“ค์–ด๋ณด์„ธ์š”.
And now listen to the click.
IWSLT2017
null
null
null
์ด๊ฒŒ ๋ฐ”๋กœ ๋ฌผ๋ฆฌ์  ๋™์กฐํ™”์ž…๋‹ˆ๋‹ค.
This is what you call physical entrainment.
IWSLT2017
null
null
null
์ด์ œ ๋‹ค์‹œ ๋‡Œ๋กœ ๋Œ์•„๊ฐ€๋ด…์‹œ๋‹ค. ์‹ ๊ฒฝ ๋™์กฐํ™”๋ฅผ ์ผ์œผํ‚ค๋Š” ๊ฑด ๋ฌด์—‡์ผ๊นŒ์š”?
Now let's go back to the brain and ask: What's driving this neural entrainment?
IWSLT2017
null
null
null
๋‹จ์ˆœํžˆ ํ™”์ž๊ฐ€ ๋‚ด๋Š” ์†Œ๋ฆฌ์ผ๊นŒ์š”?
Is it simply the sounds that the speaker is producing?
IWSLT2017
null
null
null
๋‹จ์–ด๋“ค ๋•Œ๋ฌธ์ผ ์ˆ˜๋„ ์žˆ๊ฒ ์ฃ .
Or maybe it's the words.
IWSLT2017
null
null
null
์•„๋‹ˆ๋ฉด ํ™”์ž๊ฐ€ ์ „ํ•˜๊ณ ์ž ํ•˜๋Š” ์˜๋ฏธ์ผ ์ˆ˜๋„ ์žˆ์–ด์š”.
Or maybe it's the meaning that the speaker is trying to convey.
IWSLT2017
null
null
null
๊ทธ๊ฑธ ์‹œํ—˜ํ•ด๋ณด๊ธฐ ์œ„ํ•ด์„œ ์ด ์‹คํ—˜์„ ํ–ˆ์Šต๋‹ˆ๋‹ค.
So to test it, we did the following experiment.
IWSLT2017
null
null
null
๋จผ์ € ์ด์•ผ๊ธฐ๋ฅผ ๊ฑฐ๊พธ๋กœ ํ‹€์–ด์คฌ์–ด์š”.
First, we took the story and played it backwards.
IWSLT2017
null
null
null
์ด ์‹คํ—˜์—์„œ ์›๋ž˜์˜ ์ฒญ๊ฐ์  ํŠน์ง•์€ ๋งŽ์ด ๋ณด์กด๋˜์—ˆ์ง€๋งŒ ์˜๋ฏธ๋Š” ์ œ๊ฑฐ๋˜์—ˆ์ฃ .
And that preserved many of the original auditory features, but removed the meaning.
IWSLT2017
null
null
null
์ด๋Ÿฐ ์‹์œผ๋กœ ๋“ค๋ ธ์Šต๋‹ˆ๋‹ค.
And it sounds something like that.
IWSLT2017
null
null
null
์ง: ์‚ฌ๋žŒ๋“ค ์‚ฌ์ด์— ๋งค์šฐ ์œ ์‚ฌํ•˜๊ฒŒ ๋ฐ˜์‘ํ•˜๋Š” ๋‡Œ์˜ ๋ถ€๋ถ„์„ ํ‘œ์‹œํ•˜๊ธฐ ์œ„ํ•ด ๋‡Œ ๋‘ ๊ฐœ์— ์ƒ‰์„ ์น ํ–ˆ์Šต๋‹ˆ๋‹ค.
JO: And we flashed colors in the two brains to indicate brain areas that respond very similarly across people.
IWSLT2017
null
null
null
๋ณด์‹œ๋‹ค์‹œํ”ผ ๋“ค๋ ค์ฃผ๋Š” ์†Œ๋ฆฌ๋Š” ์†Œ๋ฆฌ๋ฅผ ์ฒ˜๋ฆฌํ•˜๋Š” ๋ชจ๋“  ์ฒญ๊ฐ ํ”ผ์งˆ์—์„œ ๋™์กฐํ™”๋ฅผ ์œ ๋ฐœํ–ˆ์Šต๋‹ˆ๋‹ค. ํ•˜์ง€๋งŒ ๋‡Œ ์†์œผ๋กœ ๋” ๊นŠ๊ฒŒ ํผ์ง€์ง„ ์•Š์•˜์Šต๋‹ˆ๋‹ค.
And as you can see, this incoming sound induced entrainment or alignment in all of the brains in auditory cortices that process the sounds, but it didn't spread deeper into the brain.
IWSLT2017
null
null
null
์ด๋ฒˆ์—” ์†Œ๋ฆฌ๋ฅผ ๊ฐ€์ ธ๋‹ค๊ฐ€ ๋‹จ์–ด๋ฅผ ๋งŒ๋“ค์–ด ๋ด…์‹œ๋‹ค.
Now we can take these sounds and build words out of it.
IWSLT2017
null
null
null
์ง์ด ๋งํ•œ ๋‹จ์–ด๋“ค์„ ๋’ค์ฃฝ๋ฐ•์ฃฝ์œผ๋กœ ์„ž์–ด๋ฒ„๋ฆฌ๋ฉด ๋‹จ์–ด๋ฅผ ๋‚˜์—ดํ•œ ๊ฒƒ์ด ๋ฉ๋‹ˆ๋‹ค.
So if we take Jim O'Grady and scramble the words, we'll get a list of words.
IWSLT2017
null
null
null
์ง: ๋™๋ฌผ... ์—ฌ๋Ÿฌ ๊ฐ€์ง€ ์‚ฌ์‹ค๋“ค...
JO: ... an animal ... assorted facts ...
IWSLT2017
null
null
null
๋งž์•„... ํŒŒ์ด... ๋‚จ์ž... ์ž ์žฌ์ ์œผ๋กœ... ๋‚ด ์ด์•ผ๊ธฐ... ์šฐ๋ฆฌ: ๋ณด์‹œ๋‹ค์‹œํ”ผ ์ด ๋‹จ์–ด๋“ค์€ ์ดˆ๊ธฐ์˜ ์–ธ์–ด์  ์˜์—ญ๋“ค์—๋งŒ ๊ตญํ•œ๋œ ๋™์กฐํ™”๋ฅผ ์ผ์œผํ‚ต๋‹ˆ๋‹ค.
and right on ... pie man ... potentially ... my stories UH: And you can see that these words start to induce alignment in early language areas, but not more than that.
IWSLT2017
null
null
null
์ด์ œ ๋‹จ์–ด๋ฅผ ๊ฐ€์ ธ๋‹ค๊ฐ€ ๋ฌธ์žฅ์„ ๋งŒ๋“ค์–ด ๋ด…์‹œ๋‹ค.
Now we can take the words and start to build sentences out of them.
IWSLT2017
null
null
null
์ง: ๊ทธ ์„ ์„ ๋„˜์ง€ ๋ง๋ผ๋”๊ตฐ.
JO: And they recommend against crossing that line.
IWSLT2017
null
null
null
๊ทธ๊ฐ€ ๋งํ•˜๊ธธ, "์ง, ์žฌ๋ฐŒ๋Š” ์ด์•ผ๊ธฐ์•ผ. ์„ธ๋ถ€์‚ฌํ•ญ๋„ ์ž˜ ์‚ด๋ ธ๊ณ .
He says: "Dear Jim, Good story. Nice details.
IWSLT2017
null
null
null
๊ทธ๋…€๊ฐ€ ๋‚˜๋ฅผ ํ†ตํ•ด์„œ๋งŒ ๊ทธ๋ฅผ ์•Œ๊ณ  ์žˆ๋Š” ๊ฑฐ ์•„๋‹ˆ์—ˆ๋‚˜?"
Didn't she only know about him through me?"
IWSLT2017
null
null
null
์šฐ๋ฆฌ: ๋“ค์–ด์˜ค๋Š” ์–ธ์–ด๋ฅผ ์ฒ˜๋ฆฌํ•˜๋Š” ๋ชจ๋“  ์–ธ์–ด ์˜์—ญ์˜ ๋ฐ˜์‘์ด ๋ชจ๋“  ์ฒญ์ž๋“ค ์‚ฌ์ด์— ๋น„์Šทํ•ด์ง€๋Š” ๊ฑธ ๋ณผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
UH: Now you can see that the responses in all the language areas that process the incoming language become aligned or similar across all listeners.
IWSLT2017
null
null
null
ํ•˜์ง€๋งŒ, ์žฌ๋ฏธ์žˆ๊ณ  ์ผ๊ด€์„ฑ ์žˆ๋Š” ์ด์•ผ๊ธฐ ์ „์ฒด๋ฅผ ๋‹ค ๋“ค๋ ค์ค„ ๋•Œ๋งŒ ๋ฐ˜์‘์€ ๋‡Œ์˜ ๋” ๊นŠ์€ ๊ณณ์œผ๋กœ ํผ์ ธ์„œ ๋” ๊ณ ์ฐจ์›์ ์ธ ๋ถ€๋ถ„๊นŒ์ง€ ๊ฐ‘๋‹ˆ๋‹ค. ์ „๋‘์—ฝ๊ณผ ๋‘์ •์—ฝ์„ ํฌํ•จํ•˜๋Š” ๋ถ€๋ถ„์— ๋น„์Šทํ•œ ๋ฐ˜์‘์„ ๋ณด์ด๋„๋ก ๋งŒ๋“ญ๋‹ˆ๋‹ค.
However, only when we use the full, engaging, coherent story do the responses spread deeper into the brain into higher-order areas, which include the frontal cortex and the parietal cortex, and make all of them respond very similarly.
IWSLT2017
null
null
null
์ €ํฌ๋Š” ๊ณ ์ฐจ์›์ ์ธ ์˜์–ต์— ์ด๋Ÿฐ ๋ฐ˜์‘์ด ๋‚˜ํƒ€๋‚˜๊ฑฐ๋‚˜ ์ฒญ์ž๋“ค ๊ฐ„์˜ ๋ฐ˜์‘์ด ๋น„์Šทํ•ด์ง€๋Š” ์ด์œ ๊ฐ€ ํ™”์ž๊ฐ€ ์ „๋‹ฌํ•˜๋Š” ์˜๋ฏธ๋•Œ๋ฌธ์ด๋ผ๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค. ๋‹จ์–ด๋‚˜ ์†Œ๋ฆฌ๊ฐ€ ์•„๋‹ˆ๋ผ์š”.
And we believe that these responses in higher-order areas are induced or become similar across listeners because of the meaning conveyed by the speaker, and not by words or sound.
IWSLT2017
null
null
null
์šฐ๋ฆฌ๊ฐ€ ๋งž๋‹ค๋ฉด ์ œ๊ฐ€ ๋˜‘๊ฐ™์€ ๋‚ด์šฉ์„ ์—ฌ๋Ÿฌ๋ถ„๊ป˜ ๋‹ค๋ฅธ ๋‹จ์–ด๋ฅผ ์‚ฌ์šฉํ•ด์„œ ๋“ค๋ ค์ฃผ์–ด๋„ ์—ฌ๋Ÿฌ๋ถ„์˜ ๋‡Œ์˜ ๋ฐ˜์‘์€ ๋˜‘๊ฐ™์„ ๊ฑฐ๋ผ๋Š” ๊ฐ•๋ ฅํ•œ ์˜ˆ์ธก์„ ํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
And if we are right, there's a strong prediction over here if I tell you the exact same ideas using two very different sets of words, your brain responses will still be similar.
IWSLT2017
null
null
null
๊ทธ๊ฑธ ์‹œํ—˜ํ•ด๋ณด๊ธฐ ์œ„ํ•ด ๋‹ค์Œ์˜ ์‹คํ—˜์„ ํ–ˆ์Šต๋‹ˆ๋‹ค.
And to test it, we did the following experiment in my lab.
IWSLT2017
null
null
null
์˜์–ด๋กœ ๋œ ์ด์•ผ๊ธฐ๋ฅผ ๊ฐ€์ ธ๋‹ค๊ฐ€ ๋Ÿฌ์‹œ์•„์–ด๋กœ ๋ฒˆ์—ญํ–ˆ์Šต๋‹ˆ๋‹ค.
We took the English story and translated it to Russian.
IWSLT2017
null
null
null
๊ทธ๋Ÿผ ์™„๋ฒฝํžˆ ๋˜‘๊ฐ™์€ ์˜๋ฏธ๋ฅผ ์ „๋‹ฌํ•˜๋Š” ๋‘ ๊ฐœ์˜ ๋‹ค๋ฅธ ์†Œ๋ฆฌ์™€ ์–ธ์–ด ์ฒด๊ณ„๊ฐ€ ๋‚˜์˜ต๋‹ˆ๋‹ค.
Now you have two different sounds and linguistic systems that convey the exact same meaning.
IWSLT2017
null
null
null
์˜์–ด ์ด์•ผ๊ธฐ๋ฅผ ์˜์–ด ์‚ฌ์šฉ์ž์—๊ฒŒ ๋“ค๋ ค์ฃผ๊ณ  ๋Ÿฌ์‹œ์•„์–ด ์ด์•ผ๊ธฐ๋ฅผ ๋Ÿฌ์‹œ์•„์ธ์—๊ฒŒ ๋“ค๋ ค์ค๋‹ˆ๋‹ค. ๊ทธ๋ฆฌ๊ณ  ๊ฐ ์ง‘๋‹จ์˜ ๋ฐ˜์‘์„ ๋น„๊ตํ•ฉ๋‹ˆ๋‹ค.
And you play the English story to the English listeners and the Russian story to the Russian listeners, and we can compare their responses across the groups.
IWSLT2017
null
null
null
์ด ์‹คํ—˜์„ ํ–ˆ์„ ๋•Œ, ์–ธ์–ด์  ๋ถ€๋ถ„์˜ ์ฒญ๊ฐ ํ”ผ์งˆ์—์„œ ์œ ์‚ฌํ•œ ๋ฐ˜์‘์ด ๋‚˜ํƒ€๋‚˜๋Š” ๊ฑธ ๋ณด์ง€ ๋ชปํ–ˆ์Šต๋‹ˆ๋‹ค. ์–ธ์–ด์™€ ์†Œ๋ฆฌ๊ฐ€ ๋งค์šฐ ๋‹ค๋ฅด๊ธฐ ๋•Œ๋ฌธ์ด์ฃ .
And when we did that, we didn't see responses that are similar in auditory cortices in language, because the language and sound are very different.
IWSLT2017
null
null
null
ํ•˜์ง€๋งŒ ๋” ๊ณ ์ฐจ์›์ ์ธ ๋ถ€๋ถ„์—์„œ๋Š” ์—ฌ์ „ํžˆ ๋‘ ์ง‘๋‹จ์ด ๋น„์Šทํ•œ ๋ฐ˜์‘์„ ๋ณด์ด๋Š” ๊ฑธ ๋ณผ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.
However, you can see that the responses in high-order areas were still similar across these two groups.
IWSLT2017
null
null
null
์ €ํฌ๋Š” ๊ทธ ์ด์œ ๊ฐ€ ๊ทธ๋“ค์ด ์ด์•ผ๊ธฐ๋ฅผ ๋งค์šฐ ๋น„์Šทํ•œ ๋ฐฉ์‹์œผ๋กœ ์ดํ•ดํ–ˆ๊ธฐ ๋•Œ๋ฌธ์ด๋ผ๊ณ  ์ƒ๊ฐํ–ˆ๊ณ  ํ…Œ์ŠคํŠธ๋ฅผ ๊ฑฐ์ณ ์ฆ๋ช…๋˜์—ˆ์Šต๋‹ˆ๋‹ค.
We believe this is because they understood the story in a very similar way, as we confirmed, using a test after the story ended.
IWSLT2017
null
null
null
์ €ํฌ๋Š” ์ด ๋™์กฐํ™”๊ฐ€ ์˜์‚ฌ์†Œํ†ต์— ํ•„์š”ํ•˜๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
And we think that this alignment is necessary for communication.
IWSLT2017
null
null
null
์˜ˆ๋ฅผ ๋“ค์–ด, ์—ฌ๋Ÿฌ๋ถ„๋„ ์•„์‹œ๊ฒ ์ง€๋งŒ ์ €๋Š” ์˜์–ด๊ฐ€ ๋ชจ๊ตญ์–ด๊ฐ€ ์•„๋‹™๋‹ˆ๋‹ค.
For example, as you can tell, I am not a native English speaker.
IWSLT2017
null
null
null
๋‹ค๋ฅธ ์–ธ์–ด๋ฅผ ์“ฐ๋ฉด์„œ ์ž๋ž์ฃ . ์—ฌ๊ธฐ ๊ณ„์‹  ์—ฌ๋Ÿฌ๋ถ„๋“ค ์ค‘์—๋„ ๊ทธ๋Ÿฐ ๋ถ„์ด ๋งŽ์ด ์žˆ์„ ๊ฑฐ์˜ˆ์š”.
I grew up with another language, and the same might be for many of you in the audience.
IWSLT2017
null
null
null
๊ทธ๋Ÿฐ๋ฐ๋„ ์šฐ๋ฆฌ๋Š” ์†Œํ†ตํ•  ์ˆ˜ ์žˆ์ฃ .
And still, we can communicate.
IWSLT2017
null
null
null
์–ด์งธ์„œ์ฃ ?
How come?
IWSLT2017
null
null
null
์šฐ๋ฆฌ๋Š” ์˜๋ฏธ๋ฅผ ๋‚˜ํƒ€๋‚ด๋Š” ๊ณตํ†ต๋œ ์•”ํ˜ธ๋ฅผ ๊ฐ–๊ณ  ์žˆ์–ด์„œ ์†Œํ†ตํ•  ์ˆ˜ ์žˆ๋‹ค๊ณ  ์ƒ๊ฐํ•ฉ๋‹ˆ๋‹ค.
We think we can communicate because we have this common code that presents meaning.
IWSLT2017
null
null
null
์—ฌํƒœ๊นŒ์ง€ ์ €๋Š” ์ฒญ์ž์˜ ๋‡Œ์—์„œ ๋ฌด์Šจ ์ผ์ด ์ผ์–ด๋‚˜๋Š”์ง€๋งŒ ๋งํ–ˆ์Šต๋‹ˆ๋‹ค. ์ด์•ผ๊ธฐ๋ฅผ ๋“ค์„ ๋•Œ์˜ ์—ฌ๋Ÿฌ๋ถ„์˜ ๋‡Œ๋งŒ์š”.
So far, I've only talked about what's happening in the listener's brain, in your brain, when you're listening to talks.
IWSLT2017
null
null
null
๊ทธ๋Ÿฐ๋ฐ ์—ฌ๋Ÿฌ๋ถ„๋“ค๊ป˜ ๋ง์„ ํ•  ๋•Œ ํ™”์ž์ธ ์ €์˜ ๋‡Œ์—๋Š” ๋ฌด์Šจ ์ผ์ด ์ผ์–ด๋‚˜๊ณ  ์žˆ์„๊นŒ์š”?
But what's happening in the speaker's brain, in my brain, when I'm speaking to you?
IWSLT2017
null
null
null
์ €ํฌ๋Š” ํ™”์ž์˜ ๋‡Œ ์†์„ ๋ณด๊ธฐ ์œ„ํ•ด์„œ ํ™”์ž๋ฅผ ์Šค์บ๋„ˆ ์•ˆ์— ๋„ฃ์—ˆ์Šต๋‹ˆ๋‹ค. ๊ทธ๋ฆฌ๊ณ  ํ™”์ž์˜ ๋‡Œ๋ฅผ ์Šค์บ”ํ•ด์„œ ์ด์•ผ๊ธฐ๋ฅผ ๋“ฃ๋Š” ์ฒญ์ž์˜ ๋‡Œ ๋ฐ˜์‘๊ณผ ๊ทธ์˜ ๋‡Œ ๋ฐ˜์‘์„ ๋น„๊ตํ–ˆ์Šต๋‹ˆ๋‹ค.
To look in the speaker's brain, we asked the speaker to go into the scanner, we scan his brain and then compare his brain responses to the brain responses of the listeners listening to the story.
IWSLT2017
null
null
null