How do our brains process speech? - Gareth Gaskell

418,514 views ใƒป 2020-07-23

TED-Ed


ืื ื ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ืœืžื˜ื” ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ.

ืชืจื’ื•ื: Ido Dekkers ืขืจื™ื›ื”: Sigal Tifferet
00:06
The average 20 year old knows between 27,000 and 52,000 different words.
0
6814
7231
ื‘ืŸ ื”-20 ื”ืžืžื•ืฆืข ืžื›ื™ืจ ื‘ื™ืŸ 27,000 ืœ-52,000 ืžื™ืœื™ื ืฉื•ื ื•ืช.
00:14
By age 60, that number averages between 35,000 and 56,000.
1
14045
6008
ืขื“ ื’ื™ืœ 60, ื”ืžืกืคืจ ื”ืžืžื•ืฆืข ื”ื•ื ื‘ื™ืŸ 35,000 ืœ-56,000.
00:20
Spoken out loud, most of these words last less than a second.
2
20053
4277
ืื ืื•ืžืจื™ื ืื•ืชืŸ ื‘ืงื•ืœ, ืจื•ื‘ ื”ืžื™ืœื™ื ื”ืืœื• ื ืžืฉื›ื•ืช ืคื—ื•ืช ืžืฉื ื™ื”.
00:24
So with every word, the brain has a quick decision to make:
3
24330
4205
ืื– ืขื ื›ืœ ืžื™ืœื”, ื”ืžื•ื— ืฆืจื™ืš ืœืงื‘ืœ ื”ื—ืœื˜ื” ืžื”ื™ืจื”:
00:28
which of those thousands of options matches the signal?
4
28535
3700
ืื™ื–ื• ืžื™ืœื” ืžืืœืคื™ ื”ืื•ืคืฆื™ื•ืช ืžืชืื™ืžื” ืœืื•ืช ื”ื ื›ื ืก?
00:32
About 98% of the time, the brain chooses the correct word.
5
32235
4110
ื‘ืขืจืš ื‘-98% ืžื”ืคืขืžื™ื, ื”ืžื•ื— ื‘ื•ื—ืจ ืืช ื”ืžื™ืœื” ื”ื ื›ื•ื ื”.
00:36
But how?
6
36345
1130
ืื‘ืœ ืื™ืš?
00:37
Speech comprehension is different from reading comprehension,
7
37475
3640
ืชืคื™ืฉืช ื“ื™ื‘ื•ืจ ืฉื•ื ื” ืžืชืคื™ืฉืช ืงืจื™ืื”,
00:41
but itโ€™s similar to sign language comprehensionโ€”
8
41115
3260
ืื‘ืœ ื”ื™ื ื“ื•ืžื” ืœื”ื‘ื ืช ืฉืคืช ื”ืกื™ืžื ื™ื --
00:44
though spoken word recognition has been studied more than sign language.
9
44375
4486
ืœืžืจื•ืช ืฉื”ืฉืคื” ื”ืžื“ื•ื‘ืจืช ื ื—ืงืจื” ื™ื•ืชืจ ืžืฉืคืช ื”ืกื™ืžื ื™ื.
00:48
The key to our ability to understand speech
10
48861
2560
ื”ืžืคืชื— ืœื™ื›ื•ืœืช ืฉืœื ื• ืœื”ื‘ื™ืŸ ื“ื™ื‘ื•ืจ
00:51
is the brainโ€™s role as a parallel processor,
11
51421
3270
ื”ื•ื ืชืคืงื™ื“ ื”ืžื•ื— ื›ืžืขื‘ื“ ืžืงื‘ื™ืœื™,
00:54
meaning that it can do multiple different things at the same time.
12
54691
4000
ืžื” ืฉืื•ืžืจ ืฉื”ื•ื ื™ื›ื•ืœ ืœืขืฉื•ืช ื“ื‘ืจื™ื ืžืจื•ื‘ื™ื ืฉื•ื ื™ื ื‘ื•-ื–ืžื ื™ืช.
00:58
Most theories assume that each word we know
13
58691
2610
ืจื•ื‘ ื”ืชื™ืื•ืจื™ื•ืช ืžื ื™ื—ื•ืช ืฉื›ืœ ืžื™ืœื” ืฉืื ื—ื ื• ื™ื•ื“ืขื™ื
01:01
is represented by a separate processing unit that has just one job:
14
61301
4470
ืžื™ื•ืฆื’ืช ืขืœ ื™ื“ื™ ื™ื—ื™ื“ืช ืขื™ื‘ื•ื“ ืฉื•ื ื” ืฉื™ืฉ ืœื” ืจืง ืžืฉื™ืžื” ืื—ืช:
01:05
to assess the likelihood of incoming speech matching that particular word.
15
65771
5160
ืœื”ืขืจื™ืš ืืช ื”ืกื‘ื™ืจื•ืช ืฉื“ื™ื‘ื•ืจ ืžืชืื™ื ืœืžื™ืœื” ื”ืžืกื•ื™ื™ืžืช ื”ื–ื•.
01:10
In the context of the brain, the processing unit that represents a word
16
70931
4208
ื‘ื”ืงืฉืจ ืฉืœ ื”ืžื•ื—, ื™ื—ื™ื“ืช ื”ืขื™ื‘ื•ื“ ืฉืžื™ื™ืฆื’ืช ืžื™ืœื”
01:15
is likely a pattern of firing activity across a group of neurons
17
75139
4657
ื”ื™ื ื›ื ืจืื” ืชื‘ื ื™ืช ืฉืœ ืคืขื™ืœื•ืช ื™ืจื™ื” ื‘ืงื‘ื•ืฆืช ื ื™ื•ืจื•ื ื™ื
01:19
in the brainโ€™s cortex.
18
79796
1890
ื‘ืงืœื™ืคืช ื”ืžื•ื—.
01:21
When we hear the beginning of a word,
19
81686
1820
ื›ืฉืื ื—ื ื• ืฉื•ืžืขื™ื ื”ืชื—ืœื” ืฉืœ ืžื™ืœื”,
01:23
several thousand such units may become active,
20
83506
3780
ื›ืžื” ืืœืคื™ ื™ื—ื™ื“ื•ืช ื›ืืœื• ื™ื›ื•ืœื•ืช ืœืคืขื•ืœ,
01:27
because with just the beginning of a word,
21
87286
2066
ื›ื™ ื›ืฉืฉื•ืžืขื™ื ืจืง ื”ืชื—ืœื” ืฉืœ ืžื™ืœื”,
01:29
there are many possible matches.
22
89352
2180
ื™ืฉ ื”ืจื‘ื” ืืคืฉืจื•ื™ื•ืช ื”ืชืืžื”.
01:31
Then, as the word goes on, more and more units register
23
91532
4003
ืื–, ื›ืฉื”ืžื™ืœื” ืžืžืฉื™ื›ื”, ื™ื•ืชืจ ื•ื™ื•ืชืจ ื™ื—ื™ื“ื•ืช ืžื–ื”ื•ืช
01:35
that some vital piece of information is missing and lose activity.
24
95535
5131
ืฉื—ืกืจื•ืช ื›ืžื” ืคื™ืกื•ืช ื—ื™ื•ื ื™ื•ืช ืฉืœ ืžื™ื“ืข, ื•ื”ืŸ ืžืคืกื™ืงื•ืช ืœืคืขื•ืœ.
01:40
Possibly well before the end of the word,
25
100666
2460
ื›ื ืจืื” ื”ืจื‘ื” ืœืคื ื™ ืกื•ืฃ ื”ืžื™ืœื”,
01:43
just one firing pattern remains active, corresponding to one word.
26
103126
4964
ืจืง ืชื‘ื ื™ืช ืคืขื™ืœื” ืื—ืช ื ืฉืืจืช ืคืขื™ืœื”, ืฉืžืชืื™ืžื” ืœืžื™ืœื” ืื—ืช.
01:48
This is called the "recognition point."
27
108090
2738
ื–ื” ื ืงืจื "ื ืงื•ื“ืช ื”ื–ื™ื”ื•ื™."
01:50
In the process of honing in on one word,
28
110828
2820
ื‘ืชื”ืœื™ืš ื”ื”ืชืžืงื“ื•ืช ื‘ืžื™ืœื” ืื—ืช,
01:53
the active units suppress the activity of others,
29
113648
3070
ื”ื™ื—ื™ื“ื•ืช ื”ืคืขื™ืœื•ืช ืžืฉื‘ื™ืชื•ืช ืืช ืคืขื™ืœื•ืช ื”ืื—ืจื•ืช,
01:56
saving vital milliseconds.
30
116718
2120
ื•ื›ืš ื—ื•ืกื›ื•ืช ืžื™ืœื™-ืฉื ื™ื•ืช ื™ืงืจื•ืช.
01:58
Most people can comprehend up to about 8 syllables per second.
31
118838
4797
ืจื•ื‘ ื”ืื ืฉื™ื ื™ื›ื•ืœื™ื ืœื”ื‘ื™ืŸ ืขื“ 8 ื”ื‘ืจื•ืช ื‘ืฉื ื™ื”.
02:03
Yet, the goal is not only to recognize the word,
32
123635
3330
ื•ืขื“ื™ื™ืŸ, ื”ืžื˜ืจื” ื”ื™ื ืœื ืจืง ืœื–ื”ื•ืช ืžื™ืœื”,
02:06
but also to access its stored meaning.
33
126965
3450
ืืœื ื’ื ืœืงืฉืจ ืื•ืชื” ืœืžืฉืžืขื•ืช ืฉืœื”.
02:10
The brain accesses many possible meanings at the same time,
34
130415
3780
ื”ืžื•ื— ืคื•ื ื” ืœื”ืจื‘ื” ืžืฉืžืขื•ื™ื•ืช ืืคืฉืจื™ื•ืช ื‘ืื•ืชื• ื–ืžืŸ,
02:14
before the word has been fully identified.
35
134195
2680
ืœืคื ื™ ืฉื”ืžื™ืœื” ื–ื•ื”ืชื” ืœื’ืžืจื™.
02:16
We know this from studies which show that even upon hearing a word fragmentโ€”
36
136875
5143
ืื ื—ื ื• ื™ื•ื“ืขื™ื ื–ืืช ืžืžื—ืงืจื™ื ืฉืžืจืื™ื ืฉื’ื ืื ืฉื•ืžืขื™ื ื—ืœืง ืฉืœ ืžื™ืœื” --
02:22
like "cap"โ€”
37
142018
1280
ื›ืžื• "ืงืค"
02:23
listeners will start to register multiple possible meanings,
38
143298
3500
ืžืื–ื™ื ื™ื ื™ืชื—ื™ืœื• ืœื”ืคืขื™ืœ ื”ืจื‘ื” ืžืฉืžืขื•ื™ื•ืช ืืคืฉืจื™ื•ืช,
02:26
like captain or capital, before the full word emerges.
39
146798
5172
ื›ืžื• ืงืคื˜ืŸ ืื• ืงืคืงื™ื™ืง ืœืคื ื™ ืฉื”ืžื™ืœื” ื”ืžืœืื” ืชื•ืคื™ืข.
02:31
This suggests that every time we hear a word
40
151970
3150
ืžื›ืืŸ ืขื•ืœื” ืฉื‘ื›ืœ ืคืขื ืฉืื ื—ื ื• ืฉื•ืžืขื™ื ืžื™ืœื”
02:35
thereโ€™s a brief explosion of meanings in our minds,
41
155120
3360
ื™ืฉ ืคื™ืฆื•ืฅ ืงืฆืจ ืฉืœ ืžืฉืžืขื•ื™ื•ืช ื‘ืžื•ื— ืฉืœื ื•,
02:38
and by the recognition point the brain has settled on one interpretation.
42
158480
4811
ื•ื‘ื ืงื•ื“ืช ื”ื–ื™ื”ื•ื™, ื”ืžื•ื— ืžืชืžืงื“ ื‘ืคื™ืจื•ืฉ ืื—ื“.
02:43
The recognition process moves more rapidly
43
163291
2930
ืชื”ืœื™ืš ื”ื–ื™ื”ื•ื™ ืขื•ื‘ื“ ืžื”ืจ ื™ื•ืชืจ
02:46
with a sentence that gives us context than in a random string of words.
44
166221
4600
ื‘ืžืฉืคื˜ื™ื ืขื ื”ืงืฉืจ ืžืืฉืจ ื‘ืžื—ืจื•ื–ื•ืช ืืงืจืื™ื•ืช ืฉืœ ืžื™ืœื™ื.
02:50
Context also helps guide us towards the intended meaning of words
45
170821
4188
ื”ืงืฉืจ ื’ื ืขื•ื–ืจ ืœื”ื ื—ื•ืช ืื•ืชื ื• ืœืžืฉืžืขื•ืช ื”ืžื™ื•ืขื“ืช ืฉืœ ืžื™ืœื™ื
02:55
with multiple interpretations, like "bat," or "crane,"
46
175009
4000
ืขื ืคืจื•ืฉื™ื ืžืจื•ื‘ื™ื, ื›ืžื• "ืขื›ื‘ืจ" ืื• "ืคื—",
02:59
or in cases of homophones like "no" or "know."
47
179009
4000
ืื• ื‘ืžืงืจื™ื ืฉืœ ื”ื•ืžื•ืคื•ื ื™ื ื›ืžื• "ืขื˜" ื•"ืืช".
03:03
For multilingual people, the language they are listening to is another cue,
48
183009
4384
ืœืื ืฉื™ื ืจื‘ ืœืฉื•ื ื™ื™ื, ื”ืฉืคื” ืœื” ื”ื ืžืื–ื™ื ื™ื ื”ื™ื ืจืžื– ื ื•ืกืฃ,
03:07
used to eliminate potential words that donโ€™t match the language context.
49
187393
5313
ื”ืžืฉืžืฉ ื›ื“ื™ ืœื‘ื˜ืœ ืžื™ืœื™ื ืคื•ื˜ื ืฆื™ืืœื™ื•ืช ืฉืœื ืžืชืื™ืžื•ืช ืœื”ืงืฉืจ ืฉืœ ื”ืฉืคื”.
03:12
So, what about adding completely new words to this system?
50
192706
4000
ืื–, ืื™ืš ืžื•ืกื™ืคื™ื ืžื™ืœื™ื ื—ื“ืฉื•ืช ืœื’ืžืจื™ ืœืžืขืจื›ืช ื”ื–ื•?
03:16
Even as adults, we may come across a new word every few days.
51
196706
4000
ืืคื™ืœื• ื›ืžื‘ื•ื’ืจื™ื, ืื ื—ื ื• ื™ื›ื•ืœื™ื ืœื”ื™ืชืงืœ ื‘ืžื™ืœื” ื—ื“ืฉื” ื›ืœ ื›ืžื” ื™ืžื™ื.
03:20
But if every word is represented as a fine-tuned pattern of activity
52
200706
4403
ืื‘ืœ ืื ื›ืœ ืžื™ืœื” ืžื™ื•ืฆื’ืช ื›ืชื‘ื ื™ืช ืžื•ืชืืžืช ืฉืœ ืคืขื™ืœื•ืช
03:25
distributed over many neurons,
53
205109
2330
ืฉืžืคื•ื–ืจืช ืขืœ ื”ืจื‘ื” ื ื™ื•ืจื•ื ื™ื,
03:27
how do we prevent new words from overwriting old ones?
54
207439
4553
ืื™ืš ืื ื—ื ื• ืžื•ื ืขื™ื ืžืžืœื™ื ื—ื“ืฉื•ืช ืœืžื—ื•ืง ืžื™ืœื™ื ื™ืฉื ื•ืช?
03:31
We think that to avoid this problem,
55
211992
2330
ืื ื—ื ื• ื—ื•ืฉื‘ื™ื ืฉื›ื“ื™ ืœื”ื™ืžื ืข ืžื”ื‘ืขื™ื” ื”ื–ื•,
03:34
new words are initially stored in a part of the brain called the hippocampus,
56
214322
4763
ืžื™ืœื™ื ื—ื“ืฉื•ืช ื ืื’ืจื•ืช ืจืืฉื™ืช ื‘ื—ืœืง ืฉืœ ื”ืžื•ื— ืฉื ืงืจื ื”ื”ื™ืคื•ืงืžืคื•ืก,
03:39
well away from the main store of words in the cortex,
57
219085
3608
ืจื—ื•ืง ืžื”ืžืื’ืจ ื”ืžืจื›ื–ื™ ืฉืœ ื”ืžื™ืœื™ื ื‘ืงืœื™ืคืช ื”ืžื•ื—,
03:42
so they donโ€™t share neurons with others words.
58
222693
3370
ืื– ื”ืŸ ืœื ื—ื•ืœืงื•ืช ื ื™ื•ืจื•ื ื™ื ืขื ืžื™ืœื™ื ืื—ืจื•ืช.
03:46
Then, over multiple nights of sleep,
59
226063
3010
ื•ืื– ื‘ืžื”ืœืš ืžืกืคืจ ืœื™ืœื•ืช ืฉืœ ืฉื™ื ื”,
03:49
the new words gradually transfer over and interweave with old ones.
60
229073
5397
ื”ืžื™ืœื™ื ื”ื—ื“ืฉื•ืช ืžื•ืขื‘ืจื•ืช ื‘ื”ื“ืจื’ื” ื•ืžืชืžื–ื’ื•ืช ืขื ื”ื™ืฉื ื•ืช.
03:54
Researchers think this gradual acquisition process
61
234470
3520
ื—ื•ืงืจื•ืช ื—ื•ืฉื‘ื•ืช ืฉืชื”ืœื™ืš ื”ืจื›ื™ืฉื” ื”ื”ื“ืจื’ืชื™ ื”ื–ื”
03:57
helps avoid disrupting existing words.
62
237990
3364
ืขื•ื–ืจ ืœื”ื™ืžื ืข ืžื”ืคืจืขื” ืœืžื™ืœื™ื ืงื™ื™ืžื•ืช.
04:01
So in the daytime,
63
241354
1420
ืื– ื‘ืžืฉืš ื”ื™ื•ื,
04:02
unconscious activity generates explosions of meaning as we chat away.
64
242774
4530
ืคืขื™ืœื•ืช ืœื ื”ื›ืจืชื™ืช ืžื™ื™ืฆืจืช ืคื™ืฆื•ืฆื™ื ืฉืœ ืžืฉืžืขื•ืช ื›ืฉืื ื—ื ื• ืžืฉื•ื—ื—ื™ื.
04:07
At night, we rest, but our brains are busy integrating new knowledge
65
247304
5001
ื‘ืœื™ืœื”, ืื ื—ื ื• ื ื—ื™ื, ืื‘ืœ ื”ืžื•ื—ื•ืช ืฉืœื ื• ืขืกื•ืงื™ื ื‘ืฉื™ืœื•ื‘ ื”ืžื™ื“ืข ื”ื—ื“ืฉ
04:12
into the word network.
66
252305
1820
ื‘ืจืฉืช ื”ืžื™ืœื™ื.
04:14
When we wake up, this process ensures that weโ€™re ready
67
254125
3530
ื›ืฉืื ื—ื ื• ืžืชืขื•ืจืจื™ื, ื”ืชื”ืœื™ืš ื”ื–ื” ืžื‘ื˜ื™ื— ืฉื ื”ื™ื” ืžื•ื›ื ื™ื
04:17
for the ever-changing world of language.
68
257596
3100
ืœืขื•ืœื ื”ืฉืคื” ื”ืžืฉืชื ื” ืžื™ื•ื ืœื™ื•ื.
ืขืœ ืืชืจ ื–ื”

ืืชืจ ื–ื” ื™ืฆื™ื’ ื‘ืคื ื™ื›ื ืกืจื˜ื•ื ื™ YouTube ื”ืžื•ืขื™ืœื™ื ืœืœื™ืžื•ื“ ืื ื’ืœื™ืช. ืชื•ื›ืœื• ืœืจืื•ืช ืฉื™ืขื•ืจื™ ืื ื’ืœื™ืช ื”ืžื•ืขื‘ืจื™ื ืขืœ ื™ื“ื™ ืžื•ืจื™ื ืžื”ืฉื•ืจื” ื”ืจืืฉื•ื ื” ืžืจื—ื‘ื™ ื”ืขื•ืœื. ืœื—ืฅ ืคืขืžื™ื™ื ืขืœ ื”ื›ืชื•ื‘ื™ื•ืช ื‘ืื ื’ืœื™ืช ื”ืžื•ืฆื’ื•ืช ื‘ื›ืœ ื“ืฃ ื•ื™ื“ืื• ื›ื“ื™ ืœื”ืคืขื™ืœ ืืช ื”ืกืจื˜ื•ืŸ ืžืฉื. ื”ื›ืชื•ื‘ื™ื•ืช ื’ื•ืœืœื•ืช ื‘ืกื ื›ืจื•ืŸ ืขื ื”ืคืขืœืช ื”ื•ื•ื™ื“ืื•. ืื ื™ืฉ ืœืš ื”ืขืจื•ืช ืื• ื‘ืงืฉื•ืช, ืื ื ืฆื•ืจ ืื™ืชื ื• ืงืฉืจ ื‘ืืžืฆืขื•ืช ื˜ื•ืคืก ื™ืฆื™ืจืช ืงืฉืจ ื–ื”.

https://forms.gle/WvT1wiN1qDtmnspy7