Beyond Flat Cutting: Debunking Hanfu as a 2-D tradition

N15 Restoration Team

Artifact N15 Restoration team (from left): 王鹏智 Wang Pengzhi, 柯宇鹏 Ke Yupeng, 鲁余栋 Lu Yudong, 琥璟明 Hu Jingming, 罗海波 Luo Haibo

Original Title: 华采衣兮若英—江陵马山一号楚墓(N15)复原 [Flamboyant and Illustrious Clothing Such As Petals – Restoration of Jiangling Mashan Chu Grave No.1 Artifact N15]
By: Hu Jingming, Hunan Normal University
Translated by: Juni Yeung  

Translator’s Foreword: This is a prime example of how the Hanfu movement challenges previous authorities on Chinese historical sartorial research. Whereas in the past visual analysis of actual artifacts and artwork alone made up the bulk of empirical research, the new generation takes to active analysis of the clothing’s structure as a livable, usable piece. Numerous pieces of knowledge regarding the history and structure of Hanfu itself have been debunked as myths thanks to these people’s work, and leads us to all the more in awe to the creativity and genius of our ancestors.  

Hu’s Foreword: As Duanwu arrives, our restoration research project draws to a successful close. This may be a new way of examining and savoring our history, as we try in the past to understand history through reading words, but now we analyze and organize historical accounts with archaeological data, and restoring a visual impression to actualize our point. Not only is this an effective supplement to textual information, but also a more sensual way of interpreting history. From restoring our ancient clothing, we also provide valuable reference material for our modern fashion design and development.  

This restoration was done by the Hanfu Research Team at Huangshi Polytechnic Institute Fashion major, consisting of Hu Jingming, Lu Yudong, Luo Haibo, Wang Pengzhi, and Ke Yupeng. In order to achieve as close a texture to the original artifact, we used silk as the fabric of choice, while the collar is real silk brocade. We combined hand sewing with machine sewing for our handiwork.  

Also, we would like to thank netizen Piaoguo Yinzhou 飘过沂州, as her research results have provided great help and guidance to our efforts. When our model put on the piece, I could feel this piece’s own unique aura and aesthetic, and I am nothing short of awed and inspired from our ancestor’s sense of beauty and creativity.  

In the modern context, I believe this clothing is really suitable as a wedding dress, but definitely not something for the streets. It must be worn on a clean, polished floor, or a carpeted, bamboo-matted ground, or of the sort.  

Let us begin.


In 1982, Chu Tomb No.1 in Mashan District, Jingzhou-shi, Hubei Province, a large batch of Warring States era (6~2c. BCE) silk textiles were unearthed. Not only were they vast in amount, the collection was complete, and in good condition. It was the first time such an amount of silken goods were found like this, so this tomb became known as a “treasure trove of silk”.  

Artifact name: Small rhombus-patterned Brocade-surfaced mian down robe 小菱形纹锦面绵袍  
(NOTE: Cotton did not exist in China at the time, so silk “down” (綿 mian2, silk radical) was used, and should not be mistaken for 棉 cotton (also read mian2, wood radical)  
Artifact Registry: N15
Artifact Origin: Jiangling Mashan Chu Tomb No.1
Era: Late Warring States Era (Approx. 340-278 BCE)
Dimensions: Robe Length 200cm, sleeve span 345cm  

Theoretically, this piece would just fit Yao Ming (the basketball player), but through investigations from the Science Academy, the owner of this grave was determined at 1.6m in height, around 40-45 years old at the age of death.  

Status of Owner: (A noblewoman from the state of Chu), and from comparison of other Jiangling Chu tombs, their family status should be slightly higher than Shi.  

Use of article: Comparing from other archaeological records, this type of robe is formal wear of Chu noblewomen.  

Let us look at some pictures of the restored piece now.  

The restored artifact, worn by a model of presumably the same figure as the original owner.

Note that you cannot lift your legs too high when walking, or take great strides, on risk from stepping on your own coattails. 

 Now some snippets of the recreation process:  

The silk fabric used for the restored piece. Quite a bundle.

Not only was N15 a sight to be worn, but quite the spectacle to be made. Someone asked me what I'm doing with all the hoarded silk, now they know!

Long roll of fabric.

Busy at work by the sewing counter.

Now let us compare the original with the recreated one:

The original N15 (left), and the replica (right).

The bottom hem was not placed in a straight line, because the piece is too big, and it can’t be laid flat. *headsmack*

If we allow even more liberal measurements with this piece, we can create this kind of wrapping around the body with ease. The broad waistsash in the drawing is still a mystery to us, as nothing resembling such a waist belt was found in any Chu tombs, hence it was not used in our restoration. 

Reference materials: 

Horse and Chariot Procession Drawing, Jingmen Baoshan Chu Tomb No.2 (Warring States era)


Dragon, Phoenix, and Person Drawing, from Chenjia-Dashan Chu Tomb (Warring States era).

This kind of lower hem is also found on other relics from the period:

Bronze Lampstands from the Warring States Period, Jingmen Baoshan Chu Tomb. Now in collection of Hubei Provincial Museum.

Now let us analyze the structure of the clothing itself:

Overview of the cloth panels for the N15 artifact.

We will find that most of the intricate workings are under the armpits and by the chest area – this is actually a great creation of our ancestors. Due to our lack of data in the past, some scholars have rushed to the conclusion that Chinese clothing is defined as strictly two-dimensional cutting. This actually is a misunderstanding – 2-D cutting is used here in opposition to 3-D tailoring, specifically in reference to the method involving pasting clothing panels onto a dressform or mannequin, which appeared from pre-modern times.

These scholars were attempting to argue that Chinese clothing lacked a 3-D structural basis, but from what we see now, this view cannot stand. This piece is a prime example, from the special shape treatment by the chest area and the use of a “small waist” (explained later), the upper torso cannot be laid flat, and the basis of the structural design was 3-D in mind. Therefore, although Chinese clothing is still largely based off of flat surfaces, that is not to say that the third dimension did not exist – rather, the concept of third-dimension tailoring may have preceded the West by over one millennia.

Now let us talk about the loop-around collar and lapel: 

If the waist line was at different angles, the results would be...?

If we want to bring the front and back lapels to our backs, we need two conditions: the first being the clothing sufficiently generous in space, and second being the collar sufficiently long. The Zhiju [straight-hemmed, Shenyi] from Mawangdui solved this problem by extending out with a panel at the end of the collar. There is another way, however, that is more economical on material – the method used in this Mashan Chu Tomb N15 Artifact. 

The first situation from the diagram is how we typically tailor our clothes today: With this method, we find that it does not satisfy the lengths for going around to the back. If we pull it by force, the clothing would not be a straight fit. 

In the second situation, we rotate the connecting seam by a certain angle, and our collar distance becomes longer. The seam itself (in red) also became longer. In actuality, the clothing itself became fatter. However, if we are to change the  lower shang to match the top yi angle, a new problem arises, that the outer contour (green line) will be tugged to one side, again making the piece not fall straight. 

The third, in adjustment from the second problem, we form the third diagram, where teh angle of the shang is opposite to the yi, and the excess volume now becomes the needed space to fill in the missing part from pulling the piece ot the back. This is why the panels on my diagram are drawn as such. The chest area cannot be laid flat as it is on a table, and will form many creases. However, when worn on a body, this extra volume will be filled up, and push the shang to one side, causing the piece to swirl around the body and create a blunt angle. This angle is used to create that swirling collar lapel, like the fourth part of the diagram. 

Let us look at the clothes again. The creases on top are there because of this very reason. The old diagrams drawn by the museum staff were obviously wrong, as it overlooks that extra volume overlapping in the front, hence the collar they have is by far shorter than the actual object. The collar on this here is about 200cm long. 

N15 diagram from the 1980's, from the Hubei Provincial Museum.

Another important point to note is that the museum staff were obviously fooled by the visual illusion of the clothing being laid flat on the table versus being worn by an actual person, paid no attention to the extra fabric creased over when laid flat, and simply drew out the positions where the seams were evident visually. At the same time, they were fooled by the size of the collar opening: On my N15 replica, the collar is 16cm wide, but because of how it is laid out flat and creased over by the chest, it looks as if it’s 30cm wide, which is theoretically impossible for any clothing. 

So here it is again, the N15 pattern by our revisions. The red arrows indicate fabric weave direction, while the green dotted lines indicate folds. With this, you’ll easily understand how this piece of clothing works. 

Hu Jingming's revised N15 diagram.

On the Quju [curve-hem] mis-recognition:

When we look at the back of the piece, we see a little triangle extending around from the front. This reminds us of the Quju, and is often the place and reason where we may often mistakenly identify historical artifacts as Qujus. When I was making the patterns of this piece, I discovered that this kind of clothing may be related, or even be the predecessor of the Quju. As I was saying before, the unique angle cutting at the front has made the angle of the lapels smaller, hence allowing that corner to shift to the back. It is a possibility that the Quju was created on this basis, but this design was not immediately rendered obsolete by the Quju, but rather coexisted.

Qujus’ extra triangle at the back belongs to an extended lapel, so when it comes around, it no longer needs the “small waist” assistance, nor the special treatment by the chest in front, in turn making it actually simpler than the N15 in design and even more economical on fabric. If we take a tally, all the Qujus from the Warring States period were essentially single-wraps around the body, including the one from Mawangdui. The only Qujus that made layers after layers around the body only occurred during the Han Dynasty [1c. BCE to 1c CE]. To this, I am beginning to suppose that the appearance of the Quju is actually from the economization of the N15 type design, but the N15 itself still belongs to the Zhiju [straight-hem] design.

The "little triangle" wrap-around collar end: (left) Warring States Sitting man Lampstand (Sanmenxia-shi Shuangchuanling, discovered 1975); (right) Terracotta Warriors from the Qin Emperor's Mausoleum.

This kind of wrap-around collar end is common to the era, and may have many variants. Some look similar to the N15, while others may look a bit different.

On the “small waist” piece and the Swallow-tail: 

As we previously discussed on the wrap-around collar, the panels may shift as a result. Other interesting situations happen at the same time. 

Still recall the image explaining the theory on the wraparound collar? The extra volume on the chest will push the shang to shift to one side, so the two corners of the shang will fall to gravity, forming a pair of swallow-tail ends. However, here we must rely on something to prevent the clothing from warping, and that is the “small waist” piece. Shen Congwen once argued that this piece was the “ren (衽)” in ancient texts, and had a literally pivotal position in clothing structure, being positioned at the intersection of the body, sleeves, and lower skirt. 

The small waist, found in mediaeval robe designs around the world, but Chinese artifacts are particularly early examples.

When laid flat, this small waist piece will warp and crease. In order to demonstrate it better, I will show a piece that uses the small waist, but is not the N15. The seams are marked by red lines, and the markings will help us in understanding how it is placed between the pieces.

If we pull the clothing to show the waist piece, the piece can be seen as a rectangular panel. Seam K (diagram 2) is the sleeve seam.

After we wear this piece, the volume will be filled up and spread out. The deep neck opening on the back will go up to the neck, and the back will be pushed forward to increase the volume out front, allowing more girth for wrapping around to the back. When the extra allowance from the chest is filled, the lower shang will be pushed to swirl, causing the corners to fall and cause the swallow-tails to form. Actually, the N15 will definitely form swallow-tails by design, but because it is simply too long, this feature is no longer visible. However, on shorter specimens, this situation will become evident. With the uplifted dug-in neck, the entire back panel of the clothing will be lifted up while excess fabric will go to the front, forming a long front and short back, exposing the swallow tail from under the back panels.

Both Mashan and Mawangdui clothing specimens are treasure troves, and it can explain many previously inexplicable designs we can only see on sculptures and figurines. To understand Han dynasty clothing images, studying the Chu tomb from Mashan becomes necessary, as Liu Bang himself lived in Chu lands prior to his rise, and Chu culture has distinct influence on Han imperial culture. Although Chu culture also belongs under the Huaxia civilization umbrella, in contrast to Central Plains culture, Chu culture has its unique touches in contrast, and the establishment of Han signified the nation-wide spread of Chu culture, forming contemporary Han culture along with Central Plains and Qin regional culture.

One Robe, many images:

One robe, many styles.

Another interesting situation to note is that the same robe can be worn in many ways to give off different images: Using the piece in my last example, I can create a bunch of different images by simply adjusting how tight or loose it is worn.

The latter two styles can be interpreted as the “swallow tails” having been moved to the front instead being wound to the back, but the key is still in whether the chest space allows it.

On “fish-tail” dresses in Western couture:

As long as any clothing is sufficiently long to drag on the ground, the lower hem will form a trumpet-opening commonly known as the “fish tail”. This term is problematic when used to refer to Hanfu, since the “fish tail dress” in Western couture is a 3-D tailored design created by cutting open the bottom hem of a skirt or dress and fitting new panels in, shaping it to the likeness of a fish’s tail. In Hanfu, however, the theory and shape are different from the fish-tail concept, so calling it such is quite problematic.

Various artifacts and modern reconstructions show that "fish tail dresses" and the Hanfu swallow-tail are quite different things.

As we see from artifact examples, such robe designs existed well into the Han, so we can suppose that robes like N15 existed well into the Han. Some may question this by saying that the clay figurines were simply made to look like so to stand better, but as [fellow researcher] “Piaoguo Yinzhou” suggested, such considerations need not apply in drawings, especially of sitting people (see painting in example). Also, the China Costume Restoration Team (Zhongguo Zhuangshu Fuyuan Xiaozu)’s earlier project further proves my point that as long as the robe is long enough, the shape of the hem can be shaped like so.

On comparison with the Japanese Juuni-Hitoe

The Japanese Juuni-hitoe (12-layer robe) may have a similar-looking features, but has entirely different construction.

There are more differences although having similar appearances between N15 and the Juuni-hitoe. The Japanese piece also has an opening out front, but the basis of this opening is different from that of the N15 (in that it does not have the extra chest space, nor does it cause the lower part to swivel to one side). The Juuni-hitoe’s opening is also very large, and cannot be closed up, evidently proving that it does not have the swiveling feature. The reason why they look similar: the extra length of the robes itself, dragging along the ground behind the wearer.

Afterword and random stuff:

And some more pictures before I finish up, pictures of me on field study in Jingzhou Museum:

Artifacts from Jingzhou Museum.


Hu Jingming sketching notes.

4 Comments (+add yours?)

  1. Charles Henry Wolfenbloode
    Jun 07, 2011 @ 04:44:40

    That robe is amazing…


    • Satsuki Shizuka 五月靜
      Jun 07, 2011 @ 15:48:21

      I think I’ve seen this “small waist” treatment in mediaeval robes across Europe ca. 7th~12th c. CE. What I’m interested in is whether the Hanfu technique has any correlation to Central Asian/East European designs of the period, or later. Afterall, the Silk Road existed prior to its own “discovery” and it’s hard to say who influenced who until we examined all possible clues…


  2. Satsuki Shizuka 五月靜
    Jun 07, 2011 @ 14:56:04

    The article has been revamped to fix an inherent WordPress bug that mixes up captions and photos that follow subsequent to each other. There should be more photos now, and captions should now be correct. I apologize for any misunderstandings in the previous version.


  3. Trackback: Han Dynasty Hanfu – Susanna Herst

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


June 2011
%d bloggers like this: