Learning How to See Through Architecture
From Suzhou to Doha
从苏州到多哈：我如何从建筑中学习观看世界
Suzhoudan Dohagacha: Me’morchilik orqali dunyoni ko‘rishni qanday o‘rgandim
Zhao Guiyang
External Expert · Center for Retraining and Professional Development, Academy of Arts of Uzbekistan
乌兹别克斯坦艺术科学院教师再培训与专业发展中心 · 外籍专家
01 | From Suzhou to Doha: The Order Within a Joint
During the first few years of my career, I went to the Suzhou Museum almost every weekend.
After returning again and again for several months, I found myself drawn not only to the overall image of the building, but increasingly to details that were easy to overlook.
For a while, I became particularly interested in one of its joints.
It seemed to move from the exterior wall into the courtyard, across the roof, and into the interior ceiling, before appearing again in the joints of display cases and furniture. On the surface, it was only a line. But following that line revealed something else: the architecture, interior, landscape, and furniture had not been designed separately and then assembled afterward. From the building down to the furniture, the same sense of order continued across different scales.
Years later, when I visited I. M. Pei’s Museum of Islamic Art in Doha, I encountered something strangely familiar.
The interior palette of the museum is restrained. Dark grey stone covers the floor, warm beige defines the walls, and black lines appear between them. Yet once inside, it becomes clear that these are not the only colours of the building.
There are also the sky, the sea, and sunlight.
Light enters the central hall, reflections from the water fall across the stone, and the colours of the interior change over the course of the day.
Standing there, I thought again of Suzhou.
A Chinese garden does not need to contain all of its colours within itself. The sky, shadows of trees, reflections on water, and changing seasons are already part of it. Doha is certainly not Suzhou, and Pei was not reproducing a Chinese garden. But the way the outside world was brought into the architecture felt familiar to me.
The courtyard of the Suzhou Museum works in a similar way. From inside, one can look through the architecture toward water, stone compositions, and the landscape beyond.
So when I arrived in Doha, I kept wondering: faced with a completely different environment, how would Pei treat the view outside the window?
Standing in the central hall of the Museum of Islamic Art and looking outward, the answer was already there.
Instead of Suzhou’s pools and stone landscapes, there was the bay, fishing boats, and the skyline of the modern city.
The two landscapes could hardly be more different, yet the windows perform a similar role. A window is not simply an opening in a wall. It determines where we stand, what we see, and which parts of the world outside become part of the architecture.
This also made me pay closer attention to how Pei approached Islamic culture.
He did not cover the building with Arabic calligraphy or vegetal ornament, nor did he turn pearls, fish, or fishing boats directly into decoration. Rather than relying on easily recognisable local symbols, he seemed to be asking a different question: was there an underlying order within Islamic architecture that could be understood anew?
Layering, turning, repetition, and geometric progression became part of the building.
A square is gradually transformed through chamfered corners into an octagonal geometry, then continues to contract upward toward the dome. The transformation does not happen all at once, but unfolds layer by layer.
Similar relationships continue from the exterior into the interior. The main mass, central hall, dome, openings, stairs, stone joints, and architectural details operate at very different scales, yet they do not feel like isolated design decisions.
So when I looked at the semicircular elements near the top of the hall, I found myself less concerned with whether they were meant to symbolise pearls.
Standing there, the light had already entered the building. Through the windows, the fishing boats on the water and the city in the distance had entered as well.
Qatar did not need to be drawn on the wall.
It was already present in the architecture through light, material, and the real world outside the window.
Looking back years later, I understand more clearly why I spent so much time staring at that joint in the Suzhou Museum.
It was never really the joint itself that interested me.
What interested me was why a line appeared exactly where it did, why a window faced in a particular direction, and why one material ended at precisely that point. Decisions that seemed small were all pointing toward a larger whole.
Suzhou and Doha do not look alike, and I no longer need to search for visual similarities between them.
What felt familiar was the way space was organised beneath the surface of form.
01｜从苏州到多哈：一条缝里的秩序
刚参加工作的那几年，我几乎每个周末都会去苏州博物馆。
连续去了几个月以后，吸引我的已经不只是建筑的整体形象，而是那些很容易被忽略的细节。
有一段时间，我特别喜欢观察建筑里的一条缝。
它从外墙进入院落，经过屋顶，延伸到室内天花，最后又出现在展柜和家具的拼缝中。表面上看，它只是一条线，但顺着这条线看下去，会发现建筑、室内、景观和家具并不是各自完成以后再拼到一起的。从建筑到家具，同一种秩序一直延续到不同尺度。
多年以后，来到贝聿铭设计的多哈伊斯兰艺术博物馆，我又感受到了这种熟悉的东西。
博物馆内部的色彩很克制。深灰色石材铺在地面，墙面是温暖的米色，其间穿插着黑色的线。但真正进入空间以后，会发现建筑的颜色其实并不只有这些。
还有天空、海水和阳光。
光线进入大厅，海面的反射落在石材上，室内的颜色也随着一天中的时间发生变化。
站在那里，我又想起了苏州。
园林从来不需要自己拥有所有颜色。天空、树影、水面和季节，本来就是它的一部分。多哈当然不是苏州，贝聿铭也没有复制园林，但那种把外部世界纳入建筑的方式，我并不陌生。
苏州博物馆的中庭就是这样。人在室内，可以透过建筑看到水面、片石和远处的景物。
所以来到多哈以后，我一直很好奇：同样是一座博物馆，面对完全不同的环境，贝聿铭会怎样处理窗外的景观？
站在伊斯兰艺术博物馆的中央大厅向外看，答案其实就在眼前。
窗外没有苏州的水池和片石山水，取而代之的是海湾、渔船和现代城市的天际线。
两个地方的景观完全不同，但窗户承担着相似的作用。它不只是墙上的一个开口，也在决定人站在哪里看、看见什么，以及窗外的什么会成为建筑的一部分。
这件事也让我开始注意贝聿铭如何处理伊斯兰文化。
他没有在建筑表面大量使用阿拉伯书法或植物纹样，也没有把珍珠、鱼和渔船直接变成装饰。相比这些容易辨认的地方符号，他似乎更关心另一件事：伊斯兰建筑背后有没有一种可以被重新理解的秩序。
于是，层叠、转折、重复和几何递进进入了建筑。
方形经过切角逐渐向八角形转换，再继续向上收拢，最后进入穹顶。这个变化并不是突然发生的，而是一层一层完成。
类似的关系又从建筑外部一直进入室内。主体体量、中央大厅、穹顶、窗洞、楼梯、石材分缝和细部之间，看起来尺度不同，却并不是彼此孤立的设计。
所以，当我看到大厅顶部那些半圆形构件时，我反而不太在意它们究竟是不是在象征珍珠。
因为站在大厅里，光已经进来了；透过窗户，海上的渔船和远处的城市也已经进来了。
卡塔尔不需要再被画在墙上。
它已经通过光线、材料和窗外真实的世界进入了建筑。
多年以后再回头看，我才更明白，当初为什么会在苏州博物馆盯着一条缝看那么久。
我真正感兴趣的并不是那条缝。
而是一条线为什么出现在这里，一扇窗为什么朝向那里，一种材料为什么恰好在这个位置结束。那些看起来很小的决定，最后都指向一个更大的整体。
苏州和多哈没有相似的外形，我也不再需要寻找这种相似。
真正让我熟悉的，是形式背后那种组织空间的方式。
02 | I. M. Pei and Jean Nouvel: Two Ways of Confronting Complexity
While visiting the Museum of Islamic Art in Doha, I kept thinking of another building: Jean Nouvel’s Louvre Abu Dhabi.
Both museums stand by the sea. Both have to negotiate relationships between Islamic culture, the Arab world, and the contemporary museum. Yet they move in almost opposite directions.
The most powerful impression of the Louvre Abu Dhabi comes from its enormous dome.
Multiple layers of geometric patterns overlap. Sunlight passes through openings of different sizes and orientations, moving constantly across the ground, the walls, and the bodies of visitors. Walking beneath it, two associations easily come to mind at once: the geometry of Islamic architecture and the shifting patterns of light created when sunlight passes through palm leaves.
Beneath the dome is not a single, complete building.
Instead, a collection of low white structures stands apart from one another, creating streets, squares, and spaces along the water. Walking through them feels less like moving through a single museum building than passing through a small seaside city sheltered beneath an enormous dome.
Sea, sunlight, Islamic geometry, the traditional city, climate, and the museum’s narrative of world civilisations are all brought into the same spatial environment.
Nouvel does not rush to organise these elements into a simple answer.
He allows them to coexist.
The difference becomes even more apparent in Doha.
Not far from Pei’s Museum of Islamic Art stands Jean Nouvel’s National Museum of Qatar.
The former is stable in mass and gradually contracts inward and upward. The latter is composed of numerous discs of different sizes and orientations that intersect, cut across, and collide with one another. The design takes the crystalline formation of the desert rose as its point of departure, but standing in front of the building, the strongest impression is not simply that of an enlarged “desert rose.” It is the simultaneous presence of many relationships.
The boundaries between wall and roof become uncertain. Discs intersect with one another, while interior and exterior are repeatedly redefined.
Because the two museums stand so close to one another, the contrast is particularly striking.
One continually draws inward; the other continually unfolds.
But when I look at Nouvel’s other works together, another interesting question emerges.
The Fondation Cartier pour l’art contemporain in Paris, the Institut du Monde Arabe, the Louvre Abu Dhabi, and the National Museum of Qatar hardly share a single recognisable “Jean Nouvel form” that can simply be repeated from one project to another.
This made me wonder whether we are too accustomed to looking for an architect’s personal style through outward form.
Perhaps what remains consistent in Nouvel’s work is not a particular shape, but the way he responds to a place.
In Abu Dhabi, sunlight, the sea, the Arab city, and the idea of a global museum become part of the problem. In Qatar, the desert, geology, national history, and a rapidly changing city form another set of conditions.
When the place changes, he allows the architecture to change with it.
Looking back at Pei, it would be wrong to say that he ignored such specific conditions.
Suzhou, Doha, and Fragrant Hill in Beijing are, of course, entirely different places.
But when Pei encountered these differences, he seemed to keep asking another question: within all this complexity, is there something that can be distilled?
The Suzhou Museum does not simply reproduce a traditional Chinese garden.
White walls, courtyards, water, borrowed views, and roof forms remain, but they are reorganised. In Doha, Pei confronted a very different culture, yet he continued to search for more fundamental relationships between geometry, proportion, light, and space.
I gradually came to think that the most interesting difference between Pei and Nouvel is not whose architecture is more complex, or who is more modern.
It is the direction in which they respond to complexity.
When Pei encounters complexity, he tends to reduce, distil, and control, searching ultimately for a clear and stable whole.
Nouvel is more willing to preserve the conditions that coexist within a place, allowing history, climate, technology, materials, and human activity to enter into relationships with one another.
If I had to compress this difference into a single sentence, I would put it this way:
Pei’s strength comes from distillation; Nouvel’s comes from relationships.
This also makes me think about an interesting distinction between a classical spirit and a contemporary sensibility.
When I describe Pei as possessing a certain classical spirit, I do not mean that he used the styles of classical architecture.
Quite the opposite: he consistently worked through the language of modern architecture.
But behind his buildings there seems to be a persistent belief that even a complex world can be reorganised through proportion, geometry, centre, and order into a relatively coherent whole.
Nouvel seems less concerned with achieving that kind of completeness.
Contradictions can remain. Uncertainty can remain. Light changes every day, human activity continually enters the building, materials age, and cities continue to transform. Architecture does not necessarily have to control all of these things. It can also create a place in which they are allowed to happen.
I do not think these two approaches need to be ranked against each other.
Taken to an extreme, Pei’s approach could filter out some of the contradictions of reality through excessive control. Nouvel’s, if continually layered, could ask architecture to carry too many concepts and narratives.
What interests me is that they face an equally complex world and make fundamentally different choices.
And that difference eventually brought me back to a more basic question:
If what remains consistent in an architect is not a particular form, then what exactly is a personal language?
02｜贝聿铭与让·努维尔：面对复杂世界的两种方法
参观多哈伊斯兰艺术博物馆时，我不断想到另一座建筑——让·努维尔设计的阿布扎比卢浮宫。
两座博物馆都在海边，都需要面对伊斯兰文化、阿拉伯地域和当代博物馆之间的关系，但最后几乎走向了两个方向。
阿布扎比卢浮宫最强烈的印象来自它巨大的穹顶。
多层几何结构彼此交错，阳光从不同大小和方向的孔隙中穿过，在地面、墙面和人的身体上不断移动。人走在下面，很容易同时想到两件事情：伊斯兰建筑中的几何纹样，以及阳光穿过棕榈树叶以后留下的斑驳光影。
穹顶下面也不是一栋完整的大楼。
一组低矮的白色建筑彼此分开，中间形成街道、广场和临水空间。走在其中，与其说是在参观一座建筑，不如说是在穿过一座被巨大穹顶覆盖的海边小城。
海水、阳光、伊斯兰几何、传统城市、气候和博物馆关于世界文明的叙事，都被放进了同一个空间。
让·努维尔没有急着把这些东西整理成一个简单的答案。
他让它们同时存在。
这种差异在多哈甚至看得更直接。
距离贝聿铭的伊斯兰艺术博物馆不远，就是让·努维尔设计的卡塔尔国家博物馆。
前者的体量稳定、层层收拢；后者却由大量大小和方向不同的圆盘彼此穿插、切割和碰撞。设计的出发点来自沙漠玫瑰的晶体结构，但真正站在建筑面前时，最直接的感受并不是一个被放大的“沙漠玫瑰”，而是大量关系同时发生。
墙和屋顶的界线变得模糊，圆盘彼此穿插，室内与室外也被不断重新划分。
两座博物馆离得很近，这种反差因此格外明显。
一个不断收拢，一个不断展开。
但如果把让·努维尔的其他作品放在一起看，又会出现一个有意思的问题。
巴黎卡地亚当代艺术基金会、阿拉伯世界研究院、阿布扎比卢浮宫和卡塔尔国家博物馆，在外形上很难找到一个可以反复套用的“让·努维尔造型”。
这反而让我开始怀疑，我们是不是太习惯通过外形寻找一个建筑师的个人风格。
也许让·努维尔真正持续的东西，并不是某种形状，而是他面对地点的方式。
在阿布扎比，阳光、海水、阿拉伯城市和世界性博物馆成为问题；到了卡塔尔，沙漠、地质、国家历史和快速变化的城市又构成了另一组条件。
地点变了，他允许建筑跟着变。
再回头看贝聿铭，也会发现他并不是不考虑这些具体条件。
苏州、多哈、北京香山当然完全不同。
只是贝聿铭面对这些差异时，似乎总会继续往下追问：这些复杂的信息里面，有没有什么东西可以被提炼出来？
苏州博物馆没有直接复制一座传统园林。
白墙、庭院、水面、借景和屋顶仍然存在，但都经过重新组织。到了多哈，他面对的是另一套文化，却仍然在寻找几何、比例、光线和空间之间更基本的关系。
所以我后来觉得，两个人真正有意思的差别，并不是谁的建筑更复杂，也不是谁更现代。
而是他们处理复杂性的方向不同。
贝聿铭面对复杂时，会不断删减、提炼、控制，希望最后找到一个清晰而稳定的整体。
让·努维尔更愿意保留地点中那些同时存在的条件，让历史、气候、技术、材料和人的活动彼此发生关系。
如果一定要把这种差别压缩成一句话，我会这样理解：
贝聿铭的力量来自提纯，让·努维尔的力量来自关系。
这也让我想到“古典精神”和“当代意识”之间一种很有意思的差别。
我说贝聿铭具有某种古典精神，并不是因为他使用了古典建筑的样式。
恰恰相反，他一直在使用现代建筑语言。
但在他的建筑背后，似乎始终存在一种信念：再复杂的世界，也可以通过比例、几何、中心和秩序，被重新组织成一个相对完整的整体。
让·努维尔似乎没有那么急于获得这种完整。
矛盾可以留下，不确定性也可以留下。光线每天都在改变，人的活动不断进入，材料会老化，城市也会继续变化。建筑不一定需要把这些事情全部控制住，它也可以为它们建立一个发生的场所。
我并不觉得这两种方法需要分出高低。
贝聿铭的方法如果走到极端，可能会因为过度控制而把现实中的某些矛盾过滤掉；让·努维尔的方法如果不断叠加，也可能让建筑承受过多概念和叙事。
真正吸引我的，是他们面对同样复杂的世界，却做出了完全不同的选择。
而这种差异最后又把我带回一个更基本的问题：
如果一个建筑师真正稳定的东西不是某一种形式，那么所谓的个人语言，到底是什么？
03 | Not Searching for Form, but Developing a Way of Seeing the World
This question gradually changed the way I understood the idea of a “personal language.”
When looking at the work of great architects, it is easy to remember form first.
Pei’s geometry, Nouvel’s dome, a particular material, a particular window, or a memorable space. After seeing enough of these things, it is also easy to feel an impulse: this works beautifully—could I use it in my own design?
But form is already the end of the process.
Before it appears, an architect has already made many judgements: what to notice, where the real problem lies, what to preserve, what to remove, and what method to use to organise everything that remains.
Over time, I began to break this process down for myself.
For me, it involves roughly four things: sensing a problem, defining it, choosing a method, and eventually developing one’s own expression.
1. Sensing the Problem
Design often begins with a kind of sensitivity.
You arrive somewhere and sense that something is not quite right, or that a certain relationship deserves attention, even if at first you cannot explain exactly why.
How should tradition and modernity meet? How can a museum that addresses the world truly belong to a particular place? Is intense sunlight simply a climatic problem to be solved, or can it become part of the architecture itself?
Questions that can later be articulated clearly may begin as nothing more than a feeling.
I have come to think that this feeling matters.
Because if we sense nothing in the first place, later analysis can easily become a matter of searching for problems within answers that already exist.
2. Defining the Problem
But a feeling cannot become a design by itself.
A real project contains too much information: site, programme, history, culture, climate, materials, technology, budget, and, of course, the people who will actually use the building.
A designer cannot make everything the protagonist.
The difficult step is therefore deciding, among all these conditions: What am I actually trying to resolve?
Once that judgement becomes clear, many things that initially seemed tangled together begin to separate.
Some forms may be beautiful but have little to do with the real problem. Other things may appear insignificant yet ultimately determine the entire project.
Design, then, is not only about adding.
Very often, it is also about removing.
3. Choosing a Method
Once the problem becomes clearer, method comes next.
Pei and Nouvel, as discussed above, represent two very different approaches. But what matters is not deciding whether I should “choose Pei” or “choose Nouvel.” The more useful question is:
Why does this particular problem require this particular method?
If I choose a method simply because I am familiar with it, or because a certain form happens to be fashionable, design can easily become a habitual exercise in form-making.
An effective method should be able to transform an initially vague judgement into a space that others can enter, move through, see, and experience.
4. Developing One’s Own Expression
Seen from this perspective, “personal language” means something different to me than it once did.
It does not necessarily mean repeating the same curve, material, or geometric form from one project to another.
What remains consistent may instead be the way a person approaches a problem.
Some people continually search for order within complexity; others are more willing to preserve contradiction. Some rely on proportion and geometry to organise space; others place greater trust in light, material, time, and human activity.
The buildings may look completely different, yet the way of making judgements can still belong unmistakably to the same person.
Seen this way, I think the most important works can go one step further.
They do not merely answer questions. Sometimes they make us realise, for the first time, that there was a question there at all.
After experiencing a building, we may realise that sunlight is not merely a climatic condition to be blocked; it can also become an architectural material.
A dome does not have to remain simply an element of religious architecture; it can become the sky above a “city.”
A museum does not have to be a single enclosed building; it can be formed by streets, squares, water, and buildings together.
Tradition, too, does not have to survive through traditional forms alone.
It may remain in proportion, in light, or in the ways people move, pause, and look.
This is why, when I visit a building now, I am less satisfied with simply knowing what materials it uses, what form it takes, or what it “looks like.”
What I want to understand is:
What did the architect see at the beginning?
And how did the final form emerge, step by step, from that initial perception?
These questions eventually return to my own design practice.
What is the real problem in front of me? What method is appropriate to it? And is what remains in the final design there because it truly needs to be there—or simply because I happen to like it?
None of these questions can be answered for me by a master architect.
Forms can be imitated. Methods can be learned. But ultimately, each person has to develop a judgement of their own.
Writing this, I find myself thinking again of the Suzhou Museum many years ago.
I had only recently begun my career then. I did not yet have all these ideas about method, order, or personal language. I simply found that joint fascinating, so I went back to look at it again and again.
Years later, travelling from Suzhou to Doha, and looking from Pei to Nouvel, I gradually began to understand what had attracted me in the first place.
What was worth learning was never the joint itself, but why Pei chose to place that joint exactly where he did.
03｜不是寻找形式，而是建立观看世界的方法
想到这里，我开始重新理解所谓的“个人语言”。
以前看大师作品，很容易先记住形式。
贝聿铭的几何、让·努维尔的穹顶、某一种材料、某一扇窗，或者一个让人印象深刻的空间。看得多了以后，也很容易产生一种冲动：这个东西很好，我能不能把它用到自己的设计里？
但形式其实已经是最后发生的事情。
在它出现以前，建筑师已经做了很多判断：他注意到了什么，认为真正的问题在哪里，决定保留什么、舍弃什么，又选择什么方法把这些东西组织起来。
后来我开始把这个过程反过来拆解。
对我来说，它大致包含四件事：感受到问题，界定问题，选择方法，最后形成自己的表达。
一、感受到问题
设计的开始，很多时候只是一种敏感。
来到一个地方，会觉得某些东西不太对，或者某种关系特别值得注意，但一开始未必能够准确地说出来。
传统和现代怎样相遇？一个面向世界的博物馆怎样真正属于一个具体地点？强烈的阳光到底只是需要解决的气候问题，还是也可以成为建筑的一部分？
这些后来能够被说清楚的问题，最初可能只是一种感觉。
我越来越觉得，这种感觉很重要。
因为如果一个人什么都没有感受到，后面的分析往往也只能从已有的答案里寻找问题。
二、界定问题
但感觉不能直接变成设计。
一个真实项目里的信息太多了：场地、功能、历史、文化、气候、材料、技术、预算，还有真正使用建筑的人。
设计师不可能让每一件事情都成为主角。
所以真正困难的一步，是从这些信息里判断：我究竟在解决什么？
这个判断一旦明确，很多原本纠缠在一起的东西反而会慢慢清楚。
有些形式虽然漂亮，却和真正的问题没有关系；有些东西看起来不起眼，却可能决定整个项目。
设计因此并不只是增加。
很多时候，它也是舍弃。
三、选择方法
问题逐渐清楚以后，才轮到方法。
前面看到的贝聿铭和让·努维尔，采用的是两种不同的方法。但真正值得学习的，并不是“我应该选择贝聿铭还是让·努维尔”，而是继续追问：
为什么这个问题需要这种方法？
如果方法只是因为自己熟悉，或者因为某种形式现在流行，那么设计很容易变成一种习惯性的造型。
真正有效的方法，应该能够把最初模糊的判断，变成别人可以进入、行走、观看和感受到的空间。
四、形成自己的表达
这样再回头看“个人语言”，它的意思也和我过去理解的不太一样。
个人语言不一定意味着每个项目都出现同一种曲线、材料或者几何形。
真正稳定的，可能是一个人处理问题的习惯。
有人总是试图从复杂中找到秩序，有人愿意保留矛盾；有人依靠比例和几何控制空间，有人更相信光线、材料、时间和人的行动。
建筑最后可以长得完全不同，但背后的判断方式仍然属于同一个人。
这样理解以后，我反而觉得那些真正重要的作品还会再往前走一步。
它们不只是回答问题，有时也会让我们第一次意识到：原来这里还有一个问题。
看完建筑以后，我们才发现，阳光不只是需要遮挡的气候条件，它也可以成为建筑材料。
穹顶不一定只是宗教建筑中的构件，它也可以变成覆盖一座“城市”的天空。
博物馆不一定是一栋封闭的大楼，它也可以由街道、广场、水面和建筑共同组成。
传统也不一定要依靠传统样式才能存在。
它可能藏在比例里，藏在光线里，藏在人怎样行走、停留和观看之中。
这也是为什么现在再去看一座建筑，我越来越不满足于知道它用了什么材料、采用了什么形式，或者它“像什么”。
我更想知道的是：
建筑师最初看见了什么？
他最后留下的这个形式，究竟是怎样一步一步产生的？
这两个问题最后也会回到我自己的设计里。
我面对的真正问题是什么？什么方法适合它？最后留下来的东西，是因为它真的需要存在，还是仅仅因为我喜欢它？
这些问题没有一个可以直接从大师那里拿到答案。
形式可以模仿，方法可以学习，但一个人最终还是要建立自己的判断。
写到这里，我又想起很多年前的苏州博物馆。
那时候我刚参加工作，还没有今天这些关于方法、秩序或者个人语言的想法。我只是觉得那条缝很有意思，所以一次又一次去看。
多年以后，从苏州到多哈，从贝聿铭看到让·努维尔，我才慢慢意识到，当时真正吸引我的是什么。
真正值得学习的从来不是那条缝，而是贝聿铭为什么让那条缝出现在那里。