Deb Roy: The birth of a word

譯者: Wenjer Leuschel 審譯者: Joyce Chou

Imagine if you could record your life --

想一想，如果你能記錄下你的生命－

everything you said, everything you did,

你的一言、你的一行

available in a perfect memory store at your fingertips,

指頭點按一下，即可從完美的記憶體取得

so you could go back

那麽你便能回顧

and find memorable moments and relive them,

重溫值得回憶的時刻

or sift through traces of time

或是在過去的時光中篩選

and discover patterns in your own life

並發現自己的生命中

that previously had gone undiscovered.

先前沒有注意到的模式

Well that's exactly the journey

那就是我的家庭

that my family began

在過去五年半以來

five and a half years ago.

經過的歷程

This is my wife and collaborator, Rupal.

這是我太太也是同事Rupal

And on this day, at this moment,

就是在那一天的那一刻

we walked into the house with our first child,

我們帶著第一個孩子

our beautiful baby boy.

漂亮的兒子進入家門

And we walked into a house

我們走進的房子是一棟

with a very special home video recording system.

裝配了非常特別的錄影系統的房子

(Video) Man: Okay.

(影片) 男聲：好了

Deb Roy: This moment

Deb Roy：這個時刻以及

and thousands of other moments special for us

其他無數個對我們而言是特別的時刻

were captured in our home

都在我們家被記錄下來了

because in every room in the house,

因爲在房裡的每個房間

if you looked up, you'd see a camera and a microphone,

如果往上看就會看到攝影機和麥克風

and if you looked down,

如果往下看

you'd get this bird's-eye view of the room.

就能鳥瞰這房間

Here's our living room,

這是我們的客廳

the baby bedroom,

這是嬰兒房

kitchen, dining room

廚房、飯廳

and the rest of the house.

還有房子的其他地方

And all of these fed into a disc array

這些都儲存到一組設計來

that was designed for a continuous capture.

攫取連續影音的硬碟中

So here we are flying through a day in our home

現在快速看一下我們家的一天

as we move from sunlit morning

從太陽升起的早晨

through incandescent evening

到點亮燈火的夜晚

and, finally, lights out for the day.

到最後熄燈就寢

Over the course of three years,

這整整三年裡

we recorded eight to 10 hours a day,

我們每天紀錄

amassing roughly a quarter-million hours

8到10個小時

of multi-track audio and video.

積聚了大約25萬小時的影音

So you're looking at a piece of what is by far

大家現在看到的是有史以來

the largest home video collection ever made.

最長的家庭影集的一小部分

(Laughter)

(笑聲)

And what this data represents

這些資料所代表的

for our family at a personal level,

對我的家庭而言，在個人的層面上

the impact has already been immense,

已有巨大的影響

and we're still learning its value.

我們仍須進一步了解其中的意義

Countless moments

有無數的時刻

of unsolicited natural moments, not posed moments,

無預期、不造作的自然時刻

are captured there,

都被記錄了起來

and we're starting to learn how to discover them and find them.

我們還在學習如何發現找到那些時刻

But there's also a scientific reason that drove this project,

不過這個專案也有其科學上的目的

which was to use this natural longitudinal data

爲的是要利用這個縱向紀錄的資料

to understand the process

來了解一個小孩

of how a child learns language --

如何學習語言－

that child being my son.

那個小孩就是我的兒子

And so with many privacy provisions put in place

在設置隱私保護的條件下

to protect everyone who was recorded in the data,

讓每個在影片中的人都受到保護

we made elements of the data available

我們讓我信任的MIT

to my trusted research team at MIT

研究團隊取用這些資料

so we could start teasing apart patterns

於是我們可以開始從這巨大的

in this massive data set,

資料集裡拆解出其中的模式

trying to understand the influence of social environments

試圖了解社交環境

on language acquisition.

對語言習得有哪些影響

So we're looking here

我們看到這裡

at one of the first things we started to do.

這是我們首先進行的

This is my wife and I cooking breakfast in the kitchen,

這是太太和我在廚房做早餐

and as we move through space and through time,

當我們在時空中移動

a very everyday pattern of life in the kitchen.

會出現一條在廚房裡的日常生活軌跡

In order to convert

爲了轉換

this opaque, 90,000 hours of video

這九萬小時的影像

into something that we could start to see,

成爲能辨識的東西

we use motion analysis to pull out,

我們利用動作分析

as we move through space and through time,

汲取我們的移動軌跡

what we call space-time worms.

我們稱之爲「時空蟲」

And this has become part of our toolkit

這是我們的工具之一

for being able to look and see

用來查看

where the activities are in the data,

資料中所發生的活動

and with it, trace the pattern of, in particular,

再利用此法去追蹤，特別是

where my son moved throughout the home,

我兒子在家中活動的軌跡

so that we could focus our transcription efforts,

好讓我們能聚焦在解讀

all of the speech environment around my son --

有關我兒子學說話的環境－

all of the words that he heard from myself, my wife, our nanny,

所有他聽到的我、我太太和保姆說的字詞

and over time, the words he began to produce.

然後經過長時間他開始說那樣的字詞

So with that technology and that data

利用那樣的科技和資料

and the ability to, with machine assistance,

再利用機器輔助

transcribe speech,

便能轉譯他說出的言語

we've now transcribed

我們現在已經轉譯完成

well over seven million words of our home transcripts.

足足超過700萬個家中言談的字詞

And with that, let me take you now

我們利用這樣的轉譯來

for a first tour into the data.

瀏覽一下這些資料

So you've all, I'm sure,

我相信各位一定

seen time-lapse videos

都看過時間推移影片

where a flower will blossom as you accelerate time.

加快時間推移就可以看到花朵綻放

I'd like you to now experience

我現在想讓各位體驗一下

the blossoming of a speech form.

言語的花朵是怎麽綻放的

My son, soon after his first birthday,

我兒子過第一個生日後不久

would say "gaga" to mean water.

會說「gaga」來表示「水」

And over the course of the next half-year,

再接下來的半年時間裡

he slowly learned to approximate

他慢慢學會說出接近

the proper adult form, "water."

成年人說的正確的「water」

So we're going to cruise through half a year

我們現在要用40秒

in about 40 seconds.

快步瀏覽半年的歷程

No video here,

這裡不播放影片

so you can focus on the sound, the acoustics,

這樣各位便能聚焦在聲音的

of a new kind of trajectory:

這種新的軌跡變化

gaga to water.

從gaga到水

(Audio) Baby: Gagagagagaga

(語音) 嬰兒：Gagagagagaga

Gaga gaga gaga

Gaga gaga gaga

guga guga guga

guga guga guga

wada gaga gaga guga gaga

wada gaga gaga guga gaga

wader guga guga

wader guga guga

water water water

water water water

water water water

water water water

water water

water water

water.

water

DR: He sure nailed it, didn't he.

DR: 他的確說中了吧

(Applause)

(掌聲)

So he didn't just learn water.

他不僅學到「水」這個字詞

Over the course of the 24 months,

在24個月的歷程裡

the first two years that we really focused on,

頭兩年我們專注在這事上

this is a map of every word he learned in chronological order.

這裡有張圖列出他學到的所有字詞的時序

And because we have full transcripts,

由於我們有完整的轉譯

we've identified each of the 503 words

我們辨識出他在第二個生日前學到的

that he learned to produce by his second birthday.

這503個字詞的每一個

He was an early talker.

他算是早說話的

And so we started to analyze why.

所以我們開始分析原因

Why were certain words born before others?

爲什麽有些字詞來得早？

This is one of the first results

這是最早發現的結果之一

that came out of our study a little over a year ago

大約一年前的研究結果

that really surprised us.

真的很讓我們吃驚

The way to interpret this apparently simple graph

這張圖看來簡單，但解讀起來

is, on the vertical is an indication

在垂直線上有一顯示

of how complex caregiver utterances are

從言語長度來看，照顧者的話語

based on the length of utterances.

是很複雜的

And the [horizontal] axis is time.

垂直軸線表示時間

And all of the data,

我們將所有的資料

we aligned based on the following idea:

根據下述的想法排列：

Every time my son would learn a word,

每當發現我兒子就要學會一個字詞

we would trace back and look at all of the language he heard

我們就回溯查看他先前聽到出現

that contained that word.

那個字詞的所有言語裡

And we would plot the relative length of the utterances.

我們就繪製出那個比較長的言語

And what we found was this curious phenomena,

結果我們發現這麽一個奇特的現象

that caregiver speech would systematically dip to a minimum,

照顧者都會有系統地把字詞減降到最少

making language as simple as possible,

讓語言儘量變得簡單

and then slowly ascend back up in complexity.

然後又逐漸升回到複雜

And the amazing thing was

令人訝異的是

that bounce, that dip,

那個升、那個降

lined up almost precisely

幾乎與每個字詞

with when each word was born --

誕生的時間恰恰吻合－

word after word, systematically.

一詞接一詞，很有系統

So it appears that all three primary caregivers --

因此看來所有三個主要的照顧者－

myself, my wife and our nanny --

我自己、太太和我們的保姆－

were systematically and, I would think, subconsciously

都有系統地，而且我認為是下意識地

restructuring our language

重新建構我們的語言

to meet him at the birth of a word

好迎接一個字詞的誕生

and bring him gently into more complex language.

然後讓我兒子慢慢學習更複雜的語言

And the implications of this -- there are many,

這之中蘊含了許多意義

but one I just want to point out,

但我現在想指出的一點是

is that there must be amazing feedback loops.

這過程中必有令人驚異的反應循環

Of course, my son is learning

當然，我的兒子正在

from his linguistic environment,

從他的語言環境中學習

but the environment is learning from him.

但他所處的環境也會跟著有調整

That environment, people, are in these tight feedback loops

環境和人也都在這緊密的反應循環裡

and creating a kind of scaffolding

彼此建立起某種

that has not been noticed until now.

以往沒有注意到的橋梁

But that's looking at the speech context.

不過，那是從言語情境來看

What about the visual context?

若從視覺情境來看又是如何呢？

We're not looking at --

現在看到的還不是－

think of this as a dollhouse cutaway of our house.

想像這是我家剪下來的玩具屋

We've taken those circular fish-eye lens cameras,

我們使用環狀連結的魚眼攝影機

and we've done some optical correction,

然後再做一些光學修正

and then we can bring it into three-dimensional life.

因此我們可以做成3D影像

So welcome to my home.

歡迎來到我家

This is a moment,

這是其中一個時刻

one moment captured across multiple cameras.

透過多重攝影機攫取的一個時刻

The reason we did this is to create the ultimate memory machine,

這樣做是要創造出最高端的記憶機器

where you can go back and interactively fly around

可在其中以互動方式前後快速地搜尋

and then breathe video-life into this system.

從而爲此系統的影像注入生息

What I'm going to do

我現在要讓各位

is give you an accelerated view of 30 minutes,

看的是壓縮30分鐘的加快影像

again, of just life in the living room.

這次也只在客廳裡

That's me and my son on the floor.

那是我和我兒子在地板上

And there's video analytics

加上了影片分析

that are tracking our movements.

追蹤我們的動作

My son is leaving red ink. I am leaving green ink.

我兒子留下紅色軌迹，我留下綠色軌迹

We're now on the couch,

我們現在在沙發上

looking out through the window at cars passing by.

看著窗外駛過的汽車

And finally, my son playing in a walking toy by himself.

最後我兒子獨自玩會動的玩具

Now we freeze the action, 30 minutes,

我們在此停格，這段有30分鐘

we turn time into the vertical axis,

我們把時間放到垂直軸上

and we open up for a view

然後從中來看一下

of these interaction traces we've just left behind.

剛才留下來的互動軌迹

And we see these amazing structures --

我們看到這個令人訝異的結構－

these little knots of two colors of thread

這些兩種顔色的小結點

we call "social hot spots."

我們稱之爲社交熱點

The spiral thread

這個蜿蜒交纏的點串

we call a "solo hot spot."

我們稱之爲單一熱點

And we think that these affect the way language is learned.

我們認爲這些熱點對語言學習有影響

What we'd like to do

我們想要做的是

is start understanding

開始理解

the interaction between these patterns

這些模式與我兒子接觸的

and the language that my son is exposed to

語言之間的互動關係

to see if we can predict

看看是否能夠預測

how the structure of when words are heard

聽到字詞時的結構

affects when they're learned --

如何影響到字詞的學習－

so in other words, the relationship

換句話說就是

between words and what they're about in the world.

字詞與現實世界之間的關係

So here's how we're approaching this.

這就是我們的解讀方法

In this video,

在這個影片裡

again, my son is being traced out.

同樣是追蹤我的兒子

He's leaving red ink behind.

他留下紅色的軌迹

And there's our nanny by the door.

門旁的是我們的保姆

(Video) Nanny: You want water? (Baby: Aaaa.)

(影片) 保姆：你要喝水？(嬰孩：Aaaa.)

Nanny: All right. (Baby: Aaaa.)

保姆：好的 (嬰孩：Aaaa.)

DR: She offers water,

DR：她問他要不要水

and off go the two worms

然後兩條時空蟲

over to the kitchen to get water.

開始蠕動到廚房拿水

And what we've done is use the word "water"

我們用來標示那個時刻

to tag that moment, that bit of activity.

那個活動的字詞是「water」

And now we take the power of data

我們現在有龐大的資料

and take every time my son

從中汲取我兒子何時聽到

ever heard the word water

「水」這個字以及

and the context he saw it in,

他看到水的情境

and we use it to penetrate through the video

我們利用來這些透析整個影片

and find every activity trace

找出每一個與水相關

that co-occurred with an instance of water.

同時發生的活動的軌迹

And what this data leaves in its wake

這些資料留下了

is a landscape.

一幅風景

We call these wordscapes.

我們稱之爲「字詞風景」

This is the wordscape for the word water,

這就是water的「字詞風景」

and you can see most of the action is in the kitchen.

各位可以看到大多在廚房發生

That's where those big peaks are over to the left.

即是在左邊那個大尖峰所表示的

And just for contrast, we can do this with any word.

做個比較，也可以爲別的字詞做個風景

We can take the word "bye"

比方說「good bye」裡的

as in "good bye."

這個「bye」

And we're now zoomed in over the entrance to the house.

我們現在放大房子入口部分

And we look, and we find, as you would expect,

我們查找也發現，各位看得出來

a contrast in the landscape

可以作爲對比的風景

where the word "bye" occurs much more in a structured way.

在那兒「bye」的頻率建構出清楚的風景

So we're using these structures

所以我們利用這種風景結構

to start predicting

開始進行預測

the order of language acquisition,

語言習得的先後順序

and that's ongoing work now.

那就是在我們目前的工作

In my lab, which we're peering into now, at MIT --

我在MIT的研究室－即是現在看到的－

this is at the media lab.

那是在媒體實驗室裡

This has become my favorite way

從影片汲取任何空間影像

of videographing just about any space.

已經成爲我最喜歡的方法

Three of the key people in this project,

這個專案有三個主要人物

Philip DeCamp, Rony Kubat and Brandon Roy are pictured here.

即是影像裡的Philip DeCamp、Rony Kubat和Brandon Roy

Philip has been a close collaborator

Philip是大家看到的

on all the visualizations you're seeing.

影片製作的同事

And Michael Fleischman

還有Michael Fleischman是在我實驗室裡的

was another Ph.D. student in my lab

另一位博士生

who worked with me on this home video analysis,

他和我一同分析這支家庭影片

and he made the following observation:

他說了以下的意見：

that "just the way that we're analyzing

他說「我們分析

how language connects to events

語言如何與事件相關聯

which provide common ground for language,

以作爲語言的共通基礎

that same idea we can take out of your home, Deb,

同樣想法也可帶到你家之外，Deb

and we can apply it to the world of public media."

我們可以用它來分析外面世界的公衆媒體」

And so our effort took an unexpected turn.

結果我們的研究有了料想不到的轉折

Think of mass media

想到公衆媒體

as providing common ground

提供共同的基礎

and you have the recipe

那就可以將我們的

for taking this idea to a whole new place.

想法帶到嶄新的境地

We've started analyzing television content

於是我們開始採用相同的原則

using the same principles --

分析電視的內容－

analyzing event structure of a TV signal --

分析電視訊號的事件結構－

episodes of shows,

播出節目的分集、

commercials,

商業廣告、

all of the components that make up the event structure.

構成事件結構的所有元件

And we're now, with satellite dishes, pulling and analyzing

結果我們現在用衛星碟抽出並分析

a good part of all the TV being watched in the United States.

相當一大部分在美國被觀看的電視節目

And you don't have to now go and instrument living rooms with microphones

現在不用再到各個客廳去裝設麥克風

to get people's conversations,

來取得人們的談話

you just tune into publicly available social media feeds.

只要收聽公衆能取得的社交媒體訊息就行了

So we're pulling in

於是我們每個月抽取

about three billion comments a month,

大約三十億則電視評論

and then the magic happens.

然後美妙的事發生了

You have the event structure,

這當中可以找到事件結構

the common ground that the words are about,

那些字詞內容的共同基礎

coming out of the television feeds;

從這些電視訊息裡透露出來

you've got the conversations

我們取得了關於

that are about those topics;

那些話題的談話

and through semantic analysis --

再經過語意分析－大家現在看到的

and this is actually real data you're looking at

確實是來自於我們進行

from our data processing --

資料處理的真實資料－

each yellow line is showing a link being made

每一條黃線顯示一則評論

between a comment in the wild

在外間造成的連結

and a piece of event structure coming out of the television signal.

於是從電視訊號逐漸顯出一點事件的結構

And the same idea now

同樣的想法現在

can be built up.

可以用來建構關聯

And we get this wordscape,

於是我們得到了這個「字詞風景」

except now words are not assembled in my living room.

只不過這些字詞並非在我家客廳裡組造出來的

Instead, the context, the common ground activities,

而是情境，即共同基礎的活動

are the content on television that's driving the conversations.

即電視的內容在推動談話

And what we're seeing here, these skyscrapers now,

我們現在看到的這些高聳的結構

are commentary

都是電視評論

that are linked to content on television.

在電視內容上有相互關聯

Same concept,

同樣的構想

but looking at communication dynamics

但請看它在另一個

in a very different sphere.

非常不同的空間所造成的溝通動態

And so fundamentally, rather than, for example,

而且深入根本，舉例來說，

measuring content based on how many people are watching,

與測量收視率所得的結果極爲不同

this gives us the basic data

此研究讓我們得到了

for looking at engagement properties of content.

用來檢視內容「佔用特性」的基本資料

And just like we can look at feedback cycles

如同我們可以檢視一個

and dynamics in a family,

家庭裡的反應循環和動態

we can now open up the same concepts

我們現在可以利用同樣的構想

and look at much larger groups of people.

檢視更大的人群

This is a subset of data from our database --

這是從我們資料庫來的一個子集－

just 50,000 out of several million --

只是透過公衆媒體來源取得的

and the social graph that connects them

幾百萬則訊息中的五萬則

through publicly available sources.

以及其間互相關聯的「社交圖」

And if you put them on one plain,

如果把它們放到一個平面上

a second plain is where the content lives.

另一個平面是內容活躍的地方

So we have the programs

於是我們有了節目

and the sporting events

和體育運動事件

and the commercials,

以及商業廣告

and all of the link structures that tie them together

還有所有把它們綁在一起的連結結構

make a content graph.

形成了一個「內容圖」

And then the important third dimension.

然後是重要的第三個面向

Each of the links that you're seeing rendered here

大家在這裡看到的每個連結

is an actual connection made

是某個人說了某東西

between something someone said

和某一件內容之間的

and a piece of content.

真實關聯

And there are, again, now tens of millions of these links

這裡又有這些關聯的幾千萬條連結

that give us the connective tissue of social graphs

這讓我們看見「社交圖」的「關聯組織」

and how they relate to content.

以及它們是如何與內容相關的情況

And we can now start to probe the structure

我們現在可以開始用

in interesting ways.

有趣的方式來探索這個結構

So if we, for example, trace the path

例如，如果我們追蹤

of one piece of content

某一件內容的路徑

that drives someone to comment on it,

那個內容讓某個人對它評論

and then we follow where that comment goes,

然後我們隨著那個評論的走向

and then look at the entire social graph that becomes activated

然後檢視整個啓動的「社交圖」

and then trace back to see the relationship

然後又回頭追蹤查看那個「社交圖」

between that social graph and content,

和內容之間的關係

a very interesting structure becomes visible.

於是顯現出一個非常有趣的結構

We call this a co-viewing clique,

我們稱之爲「共看集團」

a virtual living room if you will.

要說是一個虛擬客廳也可以

And there are fascinating dynamics at play.

這裡頭上演著引人注目的戲劇

It's not one way.

這不是單向的

A piece of content, an event, causes someone to talk.

一件內容，一個事件讓某個人說了話

They talk to other people.

這讓其他的人有感

That drives tune-in behavior back into mass media,

那就驅動了大衆傳媒的收視行爲

and you have these cycles

於是出現了這樣的循環

that drive the overall behavior.

驅動了整體的收視行爲

Another example -- very different --

另一個例子－情況很不同

another actual person in our database --

我們的資料庫裡有一位人士－

and we're finding at least hundreds, if not thousands, of these.

其實我們可以找到成千上百個例子

We've given this person a name.

我們給這位人士一個名字

This is a pro-amateur, or pro-am media critic

這是一位專業業餘者，或專業美國媒體評論員

who has this high fan-out rate.

此人有高度的粉絲收視率

So a lot of people are following this person -- very influential --

許多人追隨這位人士－他很有影響力－

and they have a propensity to talk about what's on TV.

那些追隨者傾向於在電視上說話

So this person is a key link

那麽這位人士是關聯大衆傳媒

in connecting mass media and social media together.

和社交媒體的一個主要連結

One last example from this data:

這份資料的最後一個例子是：

Sometimes it's actually a piece of content that is special.

有時確實是一件特別的內容

So if we go and look at this piece of content,

因此我們現在來檢視這一件內容

President Obama's State of the Union address

才幾個星期前的歐巴馬總統

from just a few weeks ago,

國情咨文演講

and look at what we find in this same data set,

再檢視我們在這組資料中發現些什麽

at the same scale,

用同樣的尺度來衡量

the engagement properties of this piece of content

這件內容的「佔用特性」

are truly remarkable.

真是令人驚奇

A nation exploding in conversation

整個國家頓時同步

in real time

爆發了談話

in response to what's on the broadcast.

那是對廣播的訊息所做出的反應

And of course, through all of these lines

當然，所有這些連結線也

are flowing unstructured language.

也流動著缺乏結構的語言

We can X-ray

我們可以在「社交圖」上

and get a real-time pulse of a nation,

透視一下

real-time sense

在不同的圈子裡

of the social reactions in the different circuits in the social graph

這個被這件內容啓動的國家

being activated by content.

有怎樣的即時脈動和即時官感

So, to summarize, the idea is this:

總結來說，我們的想法是這樣的：

As our world becomes increasingly instrumented

正當我們的世界變得越來越工具化

and we have the capabilities

我們有能力搜集

to collect and connect the dots

並在人們說了些什麽

between what people are saying

和他們說話的情境之間

and the context they're saying it in,

將那些點連結起來

what's emerging is an ability

那麽呈現的將會是洞悉

to see new social structures and dynamics

社交結構和社交動態的新視野

that have previously not been seen.

那將是前所未有的能力

It's like building a microscope or telescope

這像是製造了麥克風

and revealing new structures

和望遠鏡而顯現了

about our own behavior around communication.

我們的溝通行爲的新結構

And I think the implications here are profound,

我認爲其中隱含深遠的意義

whether it's for science,

無論是對科學而言

for commerce, for government,

對商業而言，對政府而言

or perhaps most of all,

或也許最重要的是

for us as individuals.

對我們個人而言

And so just to return to my son,

那麽我們回到我的兒子

when I was preparing this talk, he was looking over my shoulder,

我在準備這個談話時，他就在我身後看著

and I showed him the clips I was going to show to you today,

我讓他看今天給大家看的短片

and I asked him for permission -- granted.

也徵求他的准許－他同意了

And then I went on to reflect,

然後我開始醒思

"Isn't it amazing,

「這不是很令人訝異的嗎？

this entire database, all these recordings,

這整個資料庫，所有這些錄影紀錄

I'm going to hand off to you and to your sister" --

我把它們交給你和妹妹」

who arrived two years later --

妹妹晚了兩年出生

"and you guys are going to be able to go back and re-experience moments

「你們倆將能夠回顧並重溫

that you could never, with your biological memory,

你們的生物記憶可能

possibly remember the way you can now?"

不會記得的那些時刻」

And he was quiet for a moment.

他沈默了半响

And I thought, "What am I thinking?

我想「我想到哪裡去了

He's five years old. He's not going to understand this."

他不過才五歲，不會懂的」

And just as I was having that thought, he looked up at me and said,

我才剛這麽想，他抬頭看著我

"So that when I grow up,

說：「那麽，我長大了

I can show this to my kids?"

可以讓我的孩子看這個？」

And I thought, "Wow, this is powerful stuff."

我想「哇，這說得可真好」

So I want to leave you

那麽，我要給各位

with one last memorable moment

留下最後一個

from our family.

我們家值得回憶的時刻

This is the first time our son

這是我兒子第一次

took more than two steps at once --

走出兩步以上的情況－

captured on film.

拍攝在影片裡

And I really want you to focus on something

我真的希望讓大家看的時候

as I take you through.

要注意到其中一點

It's a cluttered environment; it's natural life.

周遭有點吵，這是自然的生活環境

My mother's in the kitchen, cooking,

我媽在廚房做飯

and, of all places, in the hallway,

就在走道上

I realize he's about to do it, about to take more than two steps.

我感覺到他就要走出兩步以上

And so you hear me encouraging him,

因此各位聽到我鼓勵他

realizing what's happening,

感到有事就要發生

and then the magic happens.

然後美妙的事發生了

Listen very carefully.

請仔細聽

About three steps in,

大概在走三步後

he realizes something magic is happening,

他感到發生了美妙的事

and the most amazing feedback loop of all kicks in,

這時最令人訝異的反應循環全都作動了

and he takes a breath in,

他鬆了一口氣

and he whispers "wow"

輕輕地說了「哇」

and instinctively I echo back the same.

我直覺反應地也說了同樣的話

And so let's fly back in time

我們現在重回那個時光

to that memorable moment.

回到那個難忘的時刻

(Video) DR: Hey.

(影片) DR：喂

Come here.

來，過來

Can you do it?

你辦得到嗎？

Oh, boy.

啊，孩子

Can you do it?

你辦得到嗎？

Baby: Yeah.

嬰孩：可以

DR: Ma, he's walking.

DR：媽，他走路了

(Laughter)

(笑聲)

(Applause)

(掌聲)

DR: Thank you.

DR：謝謝大家

(Applause)

(掌聲)