Cognition
AP Psychology Topic 2 17:54 English narration · English + 中文 subtitles burned in
Chapters
Transcript
Look at a sentence with a letter missing and you read it anyway.
看一句少了一个字母的句子，你照样能读出来。
Nobody had to tell you what the letter was.
没有人告诉你那个字母是什么。
Your brain supplied it, instantly, without asking you.
你的大脑瞬间就把它补上了，都没有问过你。
That is the whole of this unit in one moment.
这一瞬间就是整个单元的内容。
Perception is built, not received — and so, it turns out, is memory.
知觉是建构出来的，不是接收来的——记忆也一样。
Unit one ended with sensation, detecting a stimulus.
第一单元讲到感觉，也就是探测刺激。
Perception is what the brain does next: it organises and interprets.
知觉是大脑接下来做的事：组织并解释。
And it runs in two directions at once.
而且它同时朝两个方向进行。
Bottom-up processing starts with the stimulus: the raw marks on the page arrive and the brain builds upward from them.
自下而上的加工从刺激开始： 纸面上的笔画传进来，大脑从它们往上建构。
Top-down processing starts with expectation: what you already know about words tells you what the marks must be.
自上而下的加工从预期开始： 你已有的关于单词的知识告诉你这些笔画一定是什么。
Both are running whenever you perceive anything.
只要你在知觉任何东西，两者都在运作。
The exam asks you to say which one the scenario shows, so look for the clue.
考试要你说出情境展示的是哪一种，所以要找线索。
If the person needed prior knowledge, it is top-down.
如果那个人需要先前的知识，就是自上而下。
If they built the whole thing from the sensory detail, it is bottom-up.
如果他完全从感觉细节建构出整体，就是自下而上。
Before anything reaches you as a percept it has passed three filters.
在任何东西成为你的知觉之前，它已经通过了三道过滤器。
A schema is a mental framework built out of everything you have met before, and it tells you what to expect.
图式是由你以前遇到的一切建立起来的心理框架，它告诉你该期待什么。
A perceptual set is a readiness to see one thing rather than another, which is why a radiologist finds the shadow you cannot.
知觉定势是准备好看到某一种东西而不是另一种， 所以放射科医生能发现你看不出的阴影。
Context and culture do the same job from outside: the very same marks read as a letter on a page and as a number in a column.
情境和文化从外面做同样的事： 完全相同的笔画，在一页文字里读作字母，在一列数字里读作数字。
Perception is built, not received, and that is why two honest witnesses can watch one event and report two.
知觉是建构出来的，不是接收来的， 所以两位诚实的目击者可以看着同一件事，报告出两件事。
Look at four small demonstrations and watch your own eye do the work.
看四个小演示，注意你自己的眼睛在做什么。
Closure fills the gaps: four separate arcs and you see a circle that is not there.
闭合会补上缺口：四段分开的弧，你看到的却是一个并不存在的圆。
Figure and ground separates an object from its background, and you did it before you decided to.
图形与背景把物体从背景中分开，而你在决定之前就已经分好了。
Proximity groups things that are close together, so six dots become three pairs.
邻近性把靠得近的东西归成一组，于是六个点变成三对。
Similarity groups things that look alike, so the same grid reads as columns rather than rows.
相似性把看起来像的东西归成一组，于是同一个网格读起来是列而不是行。
Gestalt psychology is the name for all of this, and its claim is a strong one: the whole arrives before the parts do.
格式塔心理学就是这一切的名字，它的主张很强： 整体先于部分到达。
Attention then decides which whole you get.
然后由注意决定你得到哪一个整体。
Attention is limited, and that limit shapes everything after it.
注意是有限的，而这个限制决定了它之后的一切。
Selective attention picks one input and lets the rest go, which is what you are doing right now with this voice and not the room.
选择性注意挑出一个输入，把其余的放走， 你此刻正在这样做——听这个声音，而不是这个房间。
The cocktail party effect is the proof: your own name comes through from a conversation you were not following, so something unattended was still being processed.
鸡尾酒会效应是证据： 你没有在听的那段对话里，你自己的名字却传了进来， 所以未被注意的东西仍然在被加工。
Inattentional blindness is failing to see something in plain view because attention was elsewhere.
无意视盲是明明在眼前却没看见，因为注意在别处。
Change blindness is failing to notice that something changed between two looks.
变化盲视是两次看之间东西变了，你却没有察觉。
All three say the same thing: what attention misses never becomes a perception at all.
这三件事说的是同一句话：注意错过的东西，根本不会变成知觉。
Your retina is flat, and the world is not, so the brain reconstructs depth from cues. Binocular cues need both eyes.
你的视网膜是平的，世界不是，所以大脑要用线索把深度重建出来。
Retinal disparity is the small difference between the two images, and the bigger that difference, the nearer the object.
双眼线索需要两只眼睛。 视网膜像差是两幅图像之间的微小差别，差别越大，物体越近。
Convergence is the inward turn of your eyes when something comes close.
辐辏是有东西靠近时你双眼向内转。
Monocular cues work with one eye alone: relative size, relative clarity, texture gradient, linear perspective, and interposition, which is simply a nearer object blocking a further one.
单眼线索只用一只眼就行： 相对大小、相对清晰度、纹理梯度、线条透视， 还有遮挡，也就是近的东西挡住远的东西。
Here is the test that sorts them in the exam. Cover one eye.
考试里分辨它们的办法很简单：捂住一只眼。
Whatever still works is monocular.
还起作用的，就是单眼线索。
Two more things the brain does without asking.
大脑还会不打招呼地做两件事。
Watch a door swing open.
看一扇门被推开。
The shape falling on your retina goes from a rectangle to a thin sliver, and you go on seeing a rectangular door the whole way.
落在视网膜上的形状从矩形变成一条细缝， 而你整个过程中看到的都是一扇矩形的门。
That is perceptual constancy: the object stays the same even though the image does not.
这就是知觉恒常性：图像变了，物体在你看来没变。
The reverse trick is apparent movement, where you see motion although nothing moved.
反过来的把戏叫似动，就是明明没有东西动，你却看见了运动。
Show three still pictures in a row and something travels across them, which is all a film has ever been.
连着放三张静止的图片，就有东西从上面走过去， 而电影从来就只是这样。
Both are the brain preferring a stable world to an honest one.
两者都是大脑宁愿要一个稳定的世界，也不要一个诚实的世界。
Now turn from perceiving to thinking and problem-solving.
现在从知觉转到思维与问题解决。
Concepts are the units of thought: bird, justice, prime number.
概念是思维的单位：鸟、正义、质数。
A prototype is the best example of a concept, and it is why a sparrow feels more like a bird than a penguin does.
原型是一个概念最典型的例子， 所以麻雀比企鹅更"像"鸟。
Schemas grow in exactly two ways, and the exam wants the right one.
图式的生长恰好只有两种方式，考试要你选对那一种。
Assimilation fits a new fact into the schema you already have.
同化是把新的事实塞进你已有的图式。
Accommodation changes the schema so the fact fits.
顺应是改变图式，好让这个事实装得下。
A child who calls a whale a fish is assimilating; learning the word mammal and rebuilding the category is accommodating.
一个孩子把鲸叫作鱼，那是同化； 学会"哺乳动物"这个词并重建这个类别，那是顺应。
When a scenario asks which one happened, ask whether the schema itself had to change.
当题目问发生的是哪一种时，问问图式本身是不是必须改变。
There are two ways to solve a problem and they trade against each other.
解决问题有两种方式，它们互相权衡。
An algorithm tries every possibility.
算法把每一种可能都试一遍。
It is guaranteed to find the answer and it is far too slow for most of life.
它保证能找到答案，而对生活中大多数事来说又太慢。
A heuristic takes the shortcut.
启发式走捷径。
It is fast, it is usually right, and when it is wrong you cannot feel that it is wrong.
它快，通常也对，而当它错的时候，你感觉不到它错了。
Two shortcuts are named in the syllabus.
考纲点名了两种捷径。
The representativeness heuristic judges by how well something matches a prototype, which is how a quiet person who reads a lot gets guessed as a librarian rather than a salesperson, even though there are far more salespeople.
代表性启发式按"和原型有多像"来判断， 所以一个安静又爱读书的人会被猜成图书管理员而不是销售员， 尽管销售员多得多。
The availability heuristic judges by how easily an example comes to mind, which is why shark attacks and plane crashes feel common.
可得性启发式按"例子有多容易想起来"来判断， 所以鲨鱼袭击和空难感觉很常见。
Thinking is fast because it takes shortcuts.
思考之所以快，是因为它走捷径。
A heuristic is a mental shortcut that is usually right and sometimes badly wrong.
启发法是一种心理捷径，通常是对的，偶尔错得很离谱。
The availability heuristic judges how likely something is by how easily an example comes to mind, which is why plane crashes feel more common than they are: crashes are reported and safe flights are not.
可得性启发法用"例子有多容易想起来"来判断某件事有多可能发生， 所以空难感觉比实际更常见：坠机会被报道，平安落地不会。
Framing means the same information changes a decision depending on how it is worded.
框架效应是指同样的信息，说法不同，决定就不同。
A treatment where ninety percent survive is chosen more often than one where ten percent die, and those are the same treatment.
"九成存活"的治疗被选中的次数，多于"一成死亡"的治疗， 而这两者是同一种治疗。
Confirmation bias is looking for the evidence that fits what you already believe — and the person genuinely does not notice the rest.
确认偏误是只去找符合自己已有信念的证据—— 而当事人是真的没有注意到其余的部分。
Thinking also gets stuck, in four named ways.
思维也会卡住，有四种有名字的卡法。
A mental set is the habit of reaching for the method that worked last time, even when this problem is not that problem.
心向是习惯性地去拿上一次管用的方法，即使这道题根本不是那道题。
Functional fixedness is seeing an object only in its usual use, so the box stays a box and never becomes the step you needed to stand on.
功能固着是只按平常的用途看一个物体， 于是盒子一直是盒子，永远变不成你需要踩的那一级台阶。
The gambler's fallacy is believing a run must end: red six times, so black is due.
赌徒谬误是相信连续必须结束：连着六次红，所以该出黑了。
The wheel has no memory.
轮盘没有记忆。
The sunk-cost fallacy is carrying on because of what you have already spent rather than what is left to gain, and it is the reason people sit through films they stopped enjoying an hour ago.
沉没成本谬误是因为已经花掉的东西而继续， 而不是因为还能得到的东西， 这就是为什么人们会把一小时前就不喜欢的电影看完。
Executive functions are the mental controls that let you generate, organise, plan and actually carry out a goal, and they sit above everything we have just said.
执行功能是让你生成、组织、计划并真正完成目标的心理控制， 它凌驾于我们刚才说的一切之上。
Creativity has two halves and they run in opposite directions.
创造力有两半，方向正好相反。
Divergent thinking opens outward and generates many possible ideas.
发散思维向外打开，生成许多可能的想法。
Convergent thinking closes inward and narrows to the one answer.
聚合思维向内收拢，缩小到唯一的答案。
Creativity needs the diverging half, and an exam question almost always needs the converging half, which is a large part of why a creative student can find exams hard.
创造力需要发散的那一半， 而考题几乎总是需要聚合的那一半， 这在很大程度上就是为什么有创造力的学生会觉得考试难。
Memory has three stages, and forgetting can fail at any of them.
记忆有三个阶段，遗忘可能发生在任何一个阶段。
Encoding is getting information in.
编码是把信息放进去。
Storage is holding it.
存储是把它保存住。
Retrieval is getting it out.
提取是把它取出来。
This matters because the fix is different each time.
这很重要，因为每一种失败的补救办法都不同。
A name you never really attended to was never encoded, and no amount of trying will retrieve it.
一个你根本没有真正注意过的名字，从来就没有被编码， 再怎么努力也提取不出来。
But the name on the tip of your tongue was encoded and stored perfectly — that is a retrieval failure, and a cue will often release it.
但是那个"就在嘴边"的名字，编码和存储都好好的—— 那是提取失败，一个线索往往就能把它释放出来。
Memory systems are divided by how the information is processed, so memory is not one thing.
记忆系统是按信息被加工的方式来划分的，所以记忆不是一样东西。
Explicit memory is what you can put into words, and it splits in two: episodic memory holds the events of your own life, semantic memory holds facts you know without remembering learning them.
外显记忆是你能用语言说出来的，它又分成两种： 情景记忆装着你自己生活中的事件， 语义记忆装着你知道、却不记得是什么时候学会的事实。
Implicit memory is the kind you cannot describe.
内隐记忆是你说不出来的那种。
Procedural memory lives here, which is riding a bike or touch typing, and trying to describe it makes you worse at it.
程序性记忆住在这里，比如骑车或盲打， 而试图把它说清楚只会让你做得更差。
Prospective memory is a different job again: remembering to do something later, which is why you walk into a room and stand there.
前瞻记忆又是另一回事：记得以后要做某件事， 所以你会走进一个房间然后愣在那里。
Underneath all of it is long-term potentiation: a synapse that is used again and again gets stronger, and that is memory in biological terms.
这一切下面是长时程增强： 反复使用的突触会变强，这就是生物学意义上的记忆。
Three models of memory are examinable and you need all three.
有三个记忆模型要考，三个你都需要。
The multi-store model is a flow: sensory memory, which holds a visual trace called iconic memory and a sound trace called echoic memory, then short-term memory, then long-term memory.
多重存储模型是一条流水线： 感觉记忆，它保存一份叫图像记忆的视觉痕迹和一份叫声像记忆的听觉痕迹， 然后是短时记忆，然后是长时记忆。
The working memory model says the middle stage is not a waiting room but a workshop, with a phonological loop for sound, a visuospatial sketchpad for images, and a central executive directing both.
工作记忆模型说中间那一站不是候车室，而是车间， 里面有处理声音的语音回路，处理图像的视空间模板， 还有指挥两者的中央执行系统。
The levels of processing model ignores the boxes and asks how deeply you processed the material: shallow by how it looks, then by how it sounds, then deep by what it means.
加工水平模型不管这些盒子，只问你把材料加工得有多深： 浅层看样子，再深一点听声音，最深是看意思。
Deeper wins, every time.
越深越好记，每一次都是。
If encoding is the problem, the answer is a named strategy, not working harder.
如果问题出在编码，答案是一个有名字的策略，而不是"更努力"。
Chunking groups items so that a long string becomes a few units: a phone number is easier as three chunks than as ten digits.
组块把项目分组，让一长串变成几个单位： 一个电话号码分成三组，比十个数字容易记。
Spacing spreads the practice over days instead of one long session, and it beats the long session every time.
分散练习把练习摊到几天里，而不是一次长时间地学， 而且它每次都胜过那一次长时间的学习。
Elaborative rehearsal links the new material to something you already know, which is why an explanation sticks and a repeated word does not.
精细复述把新材料与你已经知道的东西联系起来， 这就是为什么一个解释能记住，而重复一个词记不住。
In an exam answer, name the strategy and say what it does.
在考试答案里，要写出策略的名字，并说明它做了什么。
Saying a student should study more earns nothing.
写"学生应该多学习"是拿不到分的。
Give someone a list of twenty words and they will remember the beginning and the end.
给一个人二十个词的清单，他会记住开头和结尾。
That U shape is the serial position effect, and both halves have names.
这一段讲的是编码与储存，而你怎么做这两件事，决定了什么能留下来。
Encoding and storing are what this beat is about, and how you do them decides what survives.
那个U形就是系列位置效应，两边各有名字。
Primacy is the advantage of the first items, which had time to rehearse into long-term memory.
首因效应是最前面几项的优势，它们有时间被复述进长时记忆。
Recency is the advantage of the last items, which are still sitting in short-term memory.
近因效应是最后几项的优势，它们还留在短时记忆里。
Now the strategies.
现在说策略。
Chunking groups items into meaningful units, so ten digits become three chunks.
组块把项目归成有意义的单位，十个数字变成三块。
Mnemonic devices such as the method of loci hang each item on a place along a route you already know.
像位置法这样的记忆装置，把每一项挂在你已经熟悉的一条路线上的某个地点。
The spacing effect says spreading practice over days beats massed practice, which is cramming.
间隔效应说把练习分散到好几天，胜过集中练习，也就是临时抱佛脚。
Elaborative rehearsal, linking new material to what you already know, beats maintenance rehearsal, which is just repeating it.
精细复述，也就是把新材料和你已经知道的东西连起来， 胜过保持性复述，也就是单纯重复。
Finally the extremes: a rare few people show highly superior autobiographical memory, while Alzheimer's disease and infantile amnesia limit storage from the other side.
最后是两个极端： 极少数人表现出超常自传体记忆， 而阿尔茨海默病和婴儿期遗忘从另一边限制了储存。
Getting information back takes two forms.
把信息取回来有两种形式。
Recall means producing it with no options in front of you, which is the essay question.
回忆是在没有选项的情况下把它产生出来，那就是问答题。
Recognition means picking it out of a set, which is the multiple-choice question, and it is easier because the answer is on the page acting as its own cue.
再认是从一组里把它挑出来，那就是选择题， 它更容易，因为答案就在纸上，本身就是线索。
Retrieval improves when conditions match the ones at learning.
当条件与学习时相符，提取会变好。
Context-dependent memory works better in the same place.
情境依存记忆在同一个地点更好。
State-dependent memory works better in the same physical state.
状态依存记忆在同样的身体状态下更好。
Mood-congruent memory pulls up material that matches the mood you are in now.
心境一致性记忆会调出与你此刻心境相符的材料。
And one finding beats all of these for practical use: the testing effect.
还有一个发现在实用上胜过以上全部：测试效应。
Testing yourself beats rereading, every time it has been measured.
自我测试胜过反复阅读，每一次测量都是如此。
Two kinds of interference block retrieval, and students lose the mark by mixing them up.
有两种干扰会阻碍提取，学生常常因为搞混而丢分。
Proactive interference is old material acting forward onto new: you changed your password last week and the old one keeps arriving.
前摄干扰是旧材料向前影响新材料： 你上周改了密码，旧密码还一直冒出来。
Retroactive interference is new material acting backward onto old: you learned Spanish this year and now the French you learned before is harder to reach.
倒摄干扰是新材料向后影响旧材料： 你今年学了西班牙语，之前学的法语现在更难想起来。
The trick is to learn the pair by direction, not by definition.
诀窍是按方向记这一对，而不是背定义。
Pro means forward. Retro means backward.
"前"是向前，"倒"是向后。
Say which way the material is pointing and the answer follows.
说清楚材料指向哪个方向，答案就出来了。
The forgetting curve shows the shape of the loss: it is steepest immediately after learning, not weeks later, which is exactly why reviewing on the same day is worth so much.
遗忘曲线显示出损失的形状： 学完之后立刻最陡，而不是几周以后， 这正是为什么当天复习价值这么大。
But forgetting is really four different failures wearing one name.
但遗忘其实是四种不同的失败共用一个名字。
Encoding failure means it never got in.
编码失败是它根本没进去。
Interference means other material is in the way, proactive when old learning blocks new, retroactive when new blocks old.
干扰是别的材料挡在路上： 前摄干扰是旧的挡住新的，倒摄干扰是新的挡住旧的。
Retrieval failure means it is in there and the cue is missing, and the tip-of-the-tongue phenomenon is this one, felt from the inside.
提取失败是东西在里面而线索不见了， 舌尖现象就是这一种，从里面感觉到的样子。
The tell is what a cue does.
判断的关键是线索起了什么作用。
If a hint brings it straight back, nothing was ever lost; it was only unreachable.
如果一个提示马上就把它带回来，那就从来没有丢过，只是够不着。
One more property, and it is the one with consequences.
还有一个性质，而它是有后果的那一个。
Memory is reconstructive, not a recording: every time you use a memory you rebuild it, and anything available at the time can get built in.
记忆是重构的，不是录像： 你每一次使用记忆都是在重建它，而当时手边的任何东西都可能被建进去。
The misinformation effect is what follows.
错误信息效应就是由此而来。
Information you meet after an event changes what you remember of the event, and the altered memory feels exactly like the original.
事件之后你遇到的信息会改变你对事件的记忆， 而被改过的记忆感觉起来和原来的一模一样。
Source amnesia is the smaller cousin: you keep the fact and lose where it came from, so a rumour and a textbook end up feeling the same.
来源遗忘是它较小的表亲： 你留住了事实，却丢掉了它的出处， 于是一条传言和一本教科书最后感觉是一样的。
This is why eyewitness confidence is such a poor guide to eyewitness accuracy.
所以目击者的自信心，是目击者准确性的一个很差的指标。
Let us do the real question, because the marks are for the shape of the answer.
我们来做那道真题，因为分数是给答案的结构的。
In the twenty twenty-five Article Analysis Question, participants read a summary containing twenty, fifty or eighty percent misleading sentences and then answered questions about a video.
在二〇二五年的文章分析题里， 参与者读了一段含有百分之二十、五十或八十误导性句子的摘要， 然后回答关于一段视频的问题。
Correct responses fell from seventy four percent to sixty three.
正确率从百分之七十四降到了百分之六十三。
Students were asked to explain how a finding supports or refutes the misinformation effect.
学生被要求解释某项发现如何支持或反驳错误信息效应。
First, quote the data: sixty three percent against seventy four.
第一步，引用数据：百分之六十三对百分之七十四。
Second, name the concept: the misinformation effect.
第二步，说出概念：错误信息效应。
Third, make the link: more misleading detail lowered accuracy, so the finding supports it.
第三步，把两者连起来：更多误导性细节降低了准确率，所以这个发现支持它。
A restated definition with no number does not answer explain.
只把定义重述一遍、不给数字，回答不了"解释"。
The mark is for the link between the number and the term.
分数是给数字和术语之间那个连接的。
Psychologists still disagree about what intelligence even is.
心理学家至今还在争论智力究竟是什么。
Spearman's g proposes a single general ability underneath every specific one.
斯皮尔曼的一般智力因素认为，在每一种具体能力下面有一个总的能力。
Other theories propose several.
其他理论认为有好几个。
Two of those you must know.
其中两个你必须知道。
Fluid intelligence is reasoning your way through a problem you have never seen, and it declines with age.
流体智力是靠推理走通一道你从没见过的题，它随年龄下降。
Crystallised intelligence is everything you have accumulated, and it does not.
晶体智力是你积累下来的一切，它不下降。
Early tests divided mental age by real age to give an intelligence quotient.
早期的测验用心理年龄除以实际年龄，得出智商。
Modern scoring does something different: it places you on a normal curve of the whole population, with the middle set at one hundred.
现代计分做的是另一件事： 把你放在整个人群的正态曲线上，中间定为一百。
Two findings finish the topic.
两个发现给这个话题收尾。
The Flynn effect is that scores have risen worldwide across generations, which points at society rather than biology.
弗林效应是说，几代人以来全世界的分数都在上升， 这指向社会原因而不是生物原因。
Stereotype threat is that fearing you will confirm a negative stereotype about your group lowers the score itself.
刻板印象威胁是说，害怕自己会印证关于自己群体的负面刻板印象， 本身就会拉低分数。
Intelligence testing brings two words that sound alike and mean different things.
智力测验带来两个听起来很像、意思却不同的词。
Reliability is consistency: the same person takes the test twice and gets about the same score.
信度是一致性：同一个人做两次测验，得到差不多的分数。
Validity is whether it measures what it claims to measure.
效度是它是否在测量它声称要测量的东西。
A test can be reliable and not valid, and that is the case the exam wants.
一个测验可以信度很高却没有效度，而考试要考的正是这种情况。
A broken scale that reads five kilograms too heavy every single time is perfectly reliable and completely wrong.
一台每次都多称五公斤的坏秤，信度完美，却完全错了。
Any assessment has to meet the psychometric principles, and three names come with them.
任何测评都必须满足心理测量学原则，随之而来有三个名字。
Standardization means the test is given the same way every time and scored against a comparison group.
标准化是指测验每次都用同样的方式施测，并对照一个比较团体计分。
Construct validity is whether it measures the trait it names, and predictive validity is whether it forecasts what it should.
构念效度问的是它有没有测到它所声称的特质， 预测效度问的是它能不能预测它应当预测的结果。
Test-retest reliability checks the same person later, and split-half reliability checks whether the two halves of the test agree.
重测信度是隔一段时间再测同一个人， 分半信度是看测验的两半是否一致。
Five things to take with you.
带走五点。
One: bottom-up starts with the stimulus, top-down starts with expectation.
第一：自下而上从刺激开始，自上而下从预期开始。
Two: a heuristic is fast and sometimes wrong, and framing changes a decision without changing the facts.
第二：启发法很快，有时是错的；框架效应在不改变事实的情况下改变决定。
Three: name which memory stage failed before you explain the forgetting.
第三：解释遗忘之前，先说清楚是哪个记忆阶段出了问题。
Four: proactive means old acting forward, retroactive means new acting backward.
第四：前摄是旧的向前，倒摄是新的向后。
Five: reliable is consistent, valid is correct.
第五：信度是一致，效度是正确。
Now try the practice questions below.
现在去做下面的练习题。