From Atlas: The Archaeology of an Imaginary City

DUNG Kai-cheung 董啟章 is an award-winning Hong Kong fiction writer, playwright and essayist. Among his 23 titles in Chinese are ‘Androgyny: evolution of a nonexistent species (1996), The Atlas: archaeology of an imaginary city (1997; forthcoming in English translation as The Atlas: Archaeology of an Imaginary City), The double body (1997), The rose of the name (1997), Visible cities (1998), The catalog (1999), A brief history of the silverfish (2002), P.E. period (2003), Works and creations (2005), Histories of time (2007) and The age of learning (2010). He also teaches university courses in creative writing and Hong Kong literature at various institutions in Hong Kong.


The extract below is from a book conceived in the early 1990s, as the transfer of power over Hong Kong from Great Britain to The People’s Republic of China was nearing. Ostensibly the work of archeologists documenting with the help of a full scientific toolkit the vanished archipelago-metropolis Victoria-- which also seems to have been Hong Kong in some recent past-- Dung’s Atlas is organized in four sections –“theory,” “the city,” “streets,” and “signs.” Through the scrim of vocabularies of cartography, historiography, semiology, geology and many others besides, the reader squints to glimpse the contours of a scattered, possibly ancient, water-island-city that at once depends on and completely escapes naming and description. In the hands of the unnamed archeologists, fragments of data, evidence, maps, are selected, weighed, translated, paraphrased, subject to critical scrutiny of uncertain if witty precision.

On the literary side, Dung Kai-cheung is obviously in delirious conversation with other travelers in hyperreality, from Marco Polo to Milorad Pavić, passing by Franz Kafka, Italo Calvino, Umberto Eco and Jorge Luis Borges, and looped though Foucault as he famously borrows from the Argentinean fabulist’s ChInese Encyclopedia. Then again, post-modern fiction, working with the pleasantly philosophical language game of naming and identity sometimes only plays catch-up to the historical realities of a place that has changed hands repeatedly between regimes, political and linguistic as much as symbolic. Jerusalem, one imagines, would be another such ‘mis-place,’ long overdue for a novelistic dig and remapping.

If Hong Kong’s fate for its last 150 years was to become decolonized from Britain and return to the fold of the mainland, is it really its Chineseness, that makes up its bedrock identity, or rather its collective memory of having for so long been not--Chinese ? And how come that though the English word-scrim has now been erased and Hong Kong was rewritten in Chinese character, these are now being read in a new Mandarin rather than the old Cantonese? In rising such questions, Atlas is as much a political satire as it is a tour de force of narrative gaming.


                                                                              --Nataša Ďurovičová

from Part
One: Theory


When we
study ancient maps, we find repeatedly that places with the same name appear in
different forms. These places lumped together under one name are not in fact
the same place but common places. Although they are not the same place, they
have something in common. This is how the term “commonplace” is defined.

Examples of commonplaces are numerous. Take, for example, a place called
Hung Heung Lou Shan (literally, Red Incense-Burner Mountain). There is a small
island called Hung Heung Lou shown on a map of San-on county (an area roughly
corresponding to the Pearl River Delta) in the 1819 edition of the San-on County Gazetteer. Here the
island is situated at the near southwest of the Kowloon peninsula, to the north of Yeung Suen
Chau and Kap Shui Mun. In an anonymous map drawn before 1840, entitled “A Map
of the China Coast,” however, Hung Heung Lou Shan has been moved to the south
of Yeung Suen Chau and its distance from Kowloon increased fivefold. This
island is long and narrow, lying crosswise from the northwest to the southeast.
Another “Map of San-on County” in the 1864 edition of the Guangdong Provincial Gazetteer shows a regularly-shaped island called Hung
Heung Lou Shun smack in the middle of waters south of Kowloon.

Given the similarities in their names and their overall relationship
with landmarks in the general vicinity, it is safe to conclude that Hung Heung
Lou, Hung Heung Lou Shan, and Hung Heung Lou Shun are commonplaces.
Nevertheless, we must be on our guard against taking it for granted that they
are the same place, for no place on any map can ever be the same place as any
other place on any other map. Every map has its own set of places, and every
place belongs exclusively to its own map. Therefore, no one single place could
ever transgress the map to which it owes its existence and become one with
another place. If similar configurations appear on different maps, it is
because of the fact that these places are commonplaces to each other. The Red
Incense-Burner of 1819, 1840, and 1864 cannot be the same Red Incense-Burner,
but each of them can only be the Red Incense-Burner of the maps labeled “1819,”
“1840,” and “1864” respectively.

As a matter of fact, these Red Incense-Burners are commonplaces to the
place called Hong Kong at a later age (or in another time and space), so that we come to the conclusion
that Hong Kong is also a commonplace. It follows that when every place has its
commonplaces, each of these places loses its distinctive character and becomes
simply a common place. No place can transcend itself to attain an eternal and
absolute state. When each and every place reiterates its existence through
common means, replicating one another’s commonality and vainly attempting to
raise this commonality to the highest degree, its repetitive self-affirmation
may end up as a stale convention. This is the reason that modern maps of high
precision lack imagination.

By making people forget that
places can only relate to one another as commonplaces, these conventions fool
us into believing that any place has always been the same—forever fixed and



In the map
in the 1819 edition of the San-on County Gazetteer, Tuen Mun Shun (Garrison Gate Highwater) is situated among a group of
islands in the sea to the west of Kowloon Shun, standing next to Pui To
Mountain (Cup Crossing Mountain). On the “Map of San-on County” in the 1864
edition of the Guangdong Provincial Gazetteer, however, Tuen Mun O (Garrison Gate Bay) appears among the mountains on
the eastern side of the mainland, to the north of Ma On Shan (Saddle Mountain),
facing Pui To Mountain from afar. Further, if we consult the 1897 edition of
the Guangdong Provincial Gazetteer, we discover that Tuen Mun Shun has been relocated to the western side
of the mainland, inside a bay called Tuen Mun (Regiment Gate), written with a
different character for Tuen.

There are two questions that
concern us here: firstly, the misrepresentation of the location of the place
signified as Tuen Mun; secondly, the misrepresentation of certain locations on
the maps as Tuen Mun. These two points imply that a misplaced place will always
deprive another place of its correct representation, resulting in a double
misreading. That is to say that, first of all, Tuen Mun is not where it “should
be,” and secondly that Tuen Mun occupies a place where it “should not be.”
Therefore, the prefix “mis” in “misplace” carries both the meaning of “wrongly
taking one thing as another one” as in “mistake” or “misunderstand,” and the
meaning of “should not be” as in “misbehaviour.” As for the concept of “place,”
in this school of thought, it can be understood as “representation” from the
perspective of production, or as “reading” from the perspective of reception.
In fact, “representation” and “reading” are just two sides of the same coin.

We can, for convenience’s sake,
call this school cartocentric, since its members do not believe in any
objective reality outside maps. Cartocentric scholars are totally unconcerned
about the correct location of Tuen Mun and even deny the legitimacy of such
questions. Their investigation is wholly preoccupied in how the “place” called
Tuen Mun is being represented and read. According to this view, all
representations of places are simultaneously both right and wrong: in whatever
place Tuen Mun appears, it cannot be invalidated by factors exterior to the
map. By the same token, anywhere that Tuen Mun appears is destined to be wrong.
From this is derived the thesis that “all places are misplaces, and all
misplaces are misreadings.” The map is regarded as the only operational field
of spatial senses.

Investigations from this angle
suggest that Hong Kong is also a misplace. Its appearance and evolution in the
history of cartography inevitably imply meanings of mistakes, misunderstanding,
and misdoing. However, it is also owing to this very inevitability and
actuality that it earns legitimacy and correctness, at least literally so.

It is evident that the passion
of the cartocentrics in rejecting and rebutting empirical knowledge does not
necessarily elevate them above other schools of thought. It remains but one of
many competing theories, all perhaps motivated by the desire to control the
object of knowledge by seizing the ultimate power of interpretation.

Scholars, in truth, are no different from suspicious and possessive
lovers whose derangement only increases the more deeply they probe, since
lovers always fix their eyes on misplaces.



If we open
a traditional textbook on cartography and look up the section on map reading,
we will find the following explanation: “Topographical map reading, or
topographical map interpretation, is a method of gaining knowledge about the
objective geographical environment through topographical maps. The process of
map reading requires the reader to recognize signs on maps and to allow the
received information to interact with the original spatial images registered by
the cognitive faculties of the cerebrum, and in so doing, transforms the
reading of signs into the knowledge of the geographical environment.”

However, for a map-reader like you or me, the ultimate aim of reading a
map is no longer knowledge of the actual geographical environment. With the
increase in knowledge of the geographical environment held by human
civilization, there is no longer any place on the surface of the earth left
unknown, so that the sense of amazement and exultation in discovering virgin
lands in the great age of navigation is a blessing that our age is forever
denied. Our age is crammed with so much knowledge that no space for the imagination
is left. In the foreseeable future, the totality of all scientifically produced
maps will allow us to know all knowable land on the surface of the earth. There
is only one place that is forever beyond the reach of our knowledge: the
entrance to the land of
the Peach Blossom Spring.

Therefore, we forlorn beings of
the contemporary world have no choice but to set out on a search for this land
of non-being. Yet, we no longer hold out hopes of finding it in forests,
valleys, the wilderness, or desert islands, nor do we have faith in the ability
of airplanes, ships, cars, or even our own legs to take us to this undiscovered
spot. Only on a map can we find a land that has never been trodden and never
will be. For a map-reader with an adventurous spirit, reading a map amounts to
the art of navigation. Amid thunder and lightning, surging waves and torrential
downpours, and tempests that disturb magnetic fields and distort compasses, the
map-reader dreams of a brave new world a thousand leagues in the depths of the

Such people are mocked by
orthodox cartographers as utopian map-readers, or utopian mad-readers. As a
self-professed scholar of cartography who has also joined the ranks of these
map-readers, I am always beset by contradictory and confused feelings. Can
academic research also be subjective, or even consist of imaginative
projections and speculations? Can there be a kind of personal scholarship, as
Roland Barthes proposes in Camera Lucida: a utopian “impossible science of the unique being”?

I have been searching for the
entrance to the land of non-being in old maps endowed with ancient charm and
wisdom. If maps can harbor secrets, I’d imagined that they would have to be
excavated from fragmented, moth-eaten documents rather than from so-called
scientifically rendered modern charts. In the course of these cartographic
ramblings, mistakenly regarded by others (and even myself) as academic
research, I stopped at various names, just as boats seek moorings in harbors.
In maps, names usually have more referential power than symbols. We cannot
imagine how a map with no names could refer to an actually existing place;
names are the only guarantee of referentiality. However, names frequently
possess the greatest imaginative ambiguity. They conjure up a series of
landscapes in front of your eyes more vivid than can be drawn by contour lines
or use of colors for different kinds of vegetation. Eventually I came to
understand that the entrance was not to be found in the hidden depths of
forests nor in the boundless expanse of deserts, but in names.

Two names brought me to the land
of non-being. The first is Chun Fa Lok (Spring Flowers Falling), which appears
on “A Coastal Map of Guangdong” in A
Comprehensive Account of Guangdong Province
, drawn by
Guo Fei in the late sixteenth century. This map adopts a panoramic perspective
from the land to the sea and from north to south, and has a small island named
Chun Fa Lok in the waters facing Kwai Chung (Sunflower Waterway) and Tsin Wan (Shallow
Bay). As a name for an island, Chun Fa Lok could hardly be more entrancing.
When its spring flowers wither and fall, they would form a vista as of a land
covered with red azaleas. What kind of a place could it be, where flowers fall
all year round? Would it be a memory of a magnificent spring, or a lament for
the withered flowers that follow? Compared with the Peach Blossom Spring, it
would be less bustling and more desolate, less hopeful and more elegiac. Chun
Fa Lok is a fallen Peach Blossom Spring, a soft-lens version of paradise lost.
But it isn’t without a faint trace of joy and promise, for when petals fall and
berries ripen, summer will be a time of fruitfulness. A land of ever-falling
spring flowers is a place that has never existed.

The second name is Fanchin Chow,
which appears as the name of an island in “A Chart
of Part of the Coast of China, and the Adjacent Islands from Pedro Blanco to
the Mizen,” drawn by Alexander Dalrymple, Scottish geographer and first
Hydrographer to the British Admiralty, in the years 1760 to 1770. The
topological features of several places on this marine chart remain uncertain,
while the coastline of the mainland and islands is also incomplete, leaving
many blank spaces open to speculation, and place names are given only in romanized
form, without Chinese characters. What could “Fanchin Chow” refer to? What was
it originally called in Chinese? Is it possible that it does not have a Chinese
name? Could it mean “boating in a shallow craft” (fan tsin chao)? Isn’t boating in a shallow craft an action
associated with the Peach Blossom Spring? I seem to have ventured deep into the
source of a mystical waterway aboard this shallow craft, arriving at the land
of non-being, secluded from the human world.

The land is called non-being because
it does not exist. All it is exists only as a name on the map.

According to later researchers,
Chun Fa Lok is the place subsequently known as Tsing Yi (Green Garment) Island,
while Fanchin Chow is Hong Kong Island. But these findings are beside the point
for a map-reader like me.


from Part Two: The City

Mirage: Towers in the Air

A description of
Victoria can be found in the book The Shipwreck of Kino from Bishu.

In August 1841, a Japanese seaman by the name of Kino set off from Toba
for Edo on the ship Eiju Maru, a vessel that could carry a load of 1000
piculs and had sails measuring 74 feet. A storm at a far-off sand bank in
mid-journey broke the mast so that the ship drifted out to sea. After a hundred
days adrift on the open sea, the entire crew of thirteen men were rescued by
the Spanish ship Esperanza and taken to California in the United States.
After many twists and turns, Kino was finally able to get passage to Macao on a
Portuguese vessel and then left Macao by way of Hong Kong to make his way north
along the coast to return to his own country. Kino arrived back in Japan in
1846, but, on suspicion of having violated the policy of national seclusion, he
was detained and interrogated. He was finally released and returned to his home
district of Bishu. The frontier defence authorities then sent officers to make
a second investigation who made a record of what he had seen and heard in
foreign countries. This became the text which is now included in the Edo
Anthology of Stories of Shipwrecks

Here follows an excerpt from The Shipwreck of Kino from Bishu concerning
Victoria. Implausible passages in the text may be due to Kino’s blurred memory
or the result of the investigators’ fertile imagination.

To the east of Macao is a barren island occupied by
the British, named Hong Kong. Hong Kong is over 20 ri square. Bare hilltops rise up to 1800–1900 meters. The
harbor faces north and is wide with a narrow entrance. It was afternoon when we
sailed into the harbor, and a spring mist obscured the surroundings and created
an air of loneliness which made us think that it was truly right to call it a
barren island. The ship moved slowly and we could catch indistinct glimpses of
several hundred large ships moored in the harbour, including Chinese and
foreign merchant vessels and British naval vessels. Everything was still, as if
it had been there hundreds of years. As we approached the southern shore of the
harbour, the mist was torn apart, there was a great sound of steam whistles,
the waves surged, and the ships moved back and forward busily, like suddenly
awakened sea creatures. In the swirling fog along the shore, buildings thrust
upwards and several thousand residences lined the harborside, row upon row of
houses of all shapes and sizes. Pushing aside the
waves, the slowly emerging buildings reached high up on the steep hillsides. As
we drew closer, the seawater seemed to recede from the shoreline, so
that an orderly arrangement of streets was revealed, with endless streams of
hawkers and fishermen. Everywhere hills had been levelled and large-scale
construction was taking place with stone masons, bricklayers, plasterers,
carpenters, and coolies numbering in their thousands. The work and activity
were a confused and noisy turmoil, like ants building an anthill. Berthing and
stepping ashore on that land born of the sea made us feel as if we had been set
adrift. The vast space before us created a dream-like scene like someone lost
at sea.

The approximate
time of Kino’s passage by Victoria would have been March 1845.


Ice House Street

Ice House Street was a hillside road located in Central. Its upper end
was connected to Lower Albert Road, and it intersected with Queen’s Road at the
bottom. Originally there had been an ice warehouse on this street, established
in 1845, which imported natural ice blocks from America for the consumption of
foreigners living in Hong Kong during the summer and to provide cold storage
for foodstuffs. It also supplied ice free of charge to local hospitals. Since Queen’s
Road ran alongside the harbour in those days, it was convenient for ships
transporting the ice to unload their cargo for storage at the harborside. The
Ice House’s profitability was threatened when a merchant set up a factory for
manufacturing ice in Spring Garden in Wan Chai in 1866, and it eventually
stopped importing natural ice in 1880. The Ice House and its competitor, the
Hong Kong Ice Company, were both taken over by Dairy Farm in 1918.

As it happens, the Chinese name of the street does not
correspond exactly with its English name: pronounced suet chong in Cantonese, it literally means “snow factory.” The
Chinese word for “ice” is bing, but
at that time, Hong Kong people were in the habit of referring to “ice” as “snow”
(suet). Another thing is that the “snow
factory” was not actually a factory (chong)
but a godown (fu). In short, the
correct translation for the English name would have been bing fu.

There is an
argument to the contrary, however, according to which the term “snow factory”
is an accurate description of the company, or, at least, that manufacturing
snow was one of the Ice House’s side-lines. According to this claim, the “snow
factory” in its early days was investigating how to manufacture snow, as well
as supplying ice to local residents. This suggests that snow might actually
have been manufactured, that is, creating the effect of a mock snowstorm in
imitation of the weather conditions in the expatriates’ home country. The idea
was both to ease the discomfort caused by the summer heat (to which the
foreigners found it hard to adjust) and to dispel their homesickness at the
lack of a true winter. In this sense, “snow factory” is not a mistranslation
but a wholly appropriate term for the enterprise’s other function and purpose.

advantage of “snow factory” as a term, compared with “ice house,” is that it
actually comes closer to revealing the true nature of colonial society. We have
no way of knowing now whether the plan to manufacture snow eventually
succeeded, but it was at one time a very agreeable (although secret) custom to
go to the Ice House to experience the joys of a chilly European winter. There
was supposed to have been a fully furnished Victorian-style living-room in the
cellar, where ladies and gentlemen could sit at their leisure and enjoy the
warmth and comfort of afternoon tea around an open fireplace.

In the
colony’s early days there was a story that circulated widely among the local
Chinese that the first act of each new arrival from Britain was to head
straight for the Ice House in Central to put their memories and dreams into
cold storage in the cellar, lest they rot in the cruel sub-tropical climate.

Hong Kong
people call ice-cream “snow cake” (suet
). Dairy Farm, the company which took over the Ice House, afterwards
became a major manufacturer of ice-cream. Its depot used to be in Ice House
Street, next to the Ice House, and the building afterwards became the premises
of the Fringe Club, a place where artists could perform their dreams whether
sweet and creamy or icy cold.


fom Part Four: Signs

Geological Discrimination

a geological map of Hong Kong completed in 1986 by the Geotechnical Engineering
Office, we may find clues about the exploration
of indigenous culture that were supposed to have been ardently pursued in Hong
Kong in the 1980s and 1990s. In this geological map (identified by serial
number 11), whose scope covers Hong Kong Island and the Kowloon peninsula, the
geological formations shown in the map can roughly be divided into four kinds: 1)
igneous rock (indicated on the map by colors in shades of crimson), which is
concentrated in central and northern Hong Kong Island and the main part of the
Kowloon peninsula, and consists mainly of semi-developed hill slopes, with only
the original shore of Central and Sheung Wan and the area of Tsim Sha Tsui
developed into central urban areas; 2) volcanic
rock (indicated on the map by shades of green), which is concentrated in
western, southern, and eastern Hong Kong Island, with its center in Victoria
Peak, and consists mainly of mountainous land unsuitable for development; 3)
sediment (indicated on the map by shades of light grey), which includes
sedimentary rock, gravel, silt, and clay, which is distributed along the
harborside of northern Hong Kong Island and the foreshores of the Kowloon
peninsula, and has been mainly developed into urban areas; and 4) rubble and landfill (indicated on the map by irregular
crossed lines), the main formation along the shores of urban Hong Kong Island
and Kowloon, occupying almost one third of the total urban area.

As pointed out in a PhD thesis in
the geography department at the University of Hong Kong at that time, “From Post-colonial
Theory to Post-geology” (representing the peak of agitation for indigenous
cultural exploration), cartographical circles, in the light of the chaotic
condition of the local land surface created by rapid urban development, brought
forward a proposal for getting to the source by
laying bare the intrinsic nature of geography. The timely completion of
geological maps of Hong Kong has indeed helped us to discover the channels
hidden underneath urban formations and to reflect on the implications of
developmental tendencies. Cartography thus turned into a brand-new category of
cultural studies, and the art of map-reading became a competitive skill in
study and practice among cultural studies researchers. Obvious among them is
the suitability of geological maps for making all kinds of comparisons that are
ingenious, exaggerated, and yet not lacking in internal logic, since the term
“indigenous” in Chinese means literally “native soil,” which in turn signifies
roughly the “substance of the earth” or “geology” in Chinese.

The arguments
of the indigenous chauvinists are focused on the distribution and relative
position of different geological strata. From this perspective, the
establishment of the City of Victoria, that is, the growth of local culture,
was not built on igneous rock and
volcanic rock from the distant past (from the Jurassic to the Cretaceous
period) but on comparatively recent sediment, even reclaimed land less than a
hundred years old. Igneous rock and volcanic rock were
distributed on plane surfaces at the margins of urban areas and on elevated rough ground belonging to the
category of wasteland, unsuitable for housing and not much good either for
cultivation. Earth used for reclamation, in contrast, although it is not a
naturally occurring substance, can give rise to a unique ecosystem out of
artificial materials; this is a phenomenon that cannot be lightly written off.
Since the constituents of reclamation material are heterogeneous, including all
kinds of organic and inorganic materials such as soil, gravel, and refuse,
indigenous chauvinists like to stress its hybridity, claiming it to be a
special characteristic of being “indigenous.”

However, the meaning of signs can
actually be reversed following changes in reading strategies. At the end of the
1990s there appeared a group of cultural studies map readers known as the
Granite School, which upheld the preservation of tradition, namely, the
long-standing and well-established historical value of granite (a kind of
igneous rock). Targeting the flat surface viewpoint of indigenous chauvinism,
they emphasise a vertical and historical excavation, presenting rock strata in
three-dimensional sections and thereby exposing granite
as the vast foundation at the deepest
underground level. As well, the solidity and density of granite and the glittering
crystals it embeds gain their unreserved praise.
The rock cover on the earth’s surface, which is loose, fragile, and subject to
erosion, will eventually collapse and disappear under the unrelenting onslaught
of time, while granite will fearlessly stand its ground, impervious to wind and
rain. In contrast to granite, as mighty as mother earth in all her majesty, the
little heap of piled-up earth along the shore will turn into margins within
margins even less negligible than negligible. Not even a product of some
geologic period, it is just a junkyard formed out of dumped waste material
within just a hundred years. What kind of “roots” could we hope to find beneath
this “native soil”?

In the indistinct rumbling at the
earth’s deepest strata, in the underfoot vibrations of faint enquiry,
indigenous chauvinist map readers still diligently trace the chronological
growth rings of urban shoreline reclamation on geological maps: 1985, 1982,
1964, 1904, 1863. 


From  Atlas:Archaeology of an Imaginary City (forthcoming from Columbia University Press in 2012).  

                Translated from the
Chinese by Bonnie McDougall and Anders Hansson

Bonnie S. McDougall is visiting professor of Chinese at the University of Sydney and professor emeritus at the University of Edinburgh. She has also taught at Harvard University, the University of Oslo, the Chinese University of Hong Kong, and the City University of Hong Kong. She has translated works by Bei Dao, Ah Cheng, Chen Kaige, Mao Zedong, and Leung Ping-kwan, among many others. Her recent books include Imaginary Audiences: Modern Chinese Literature in the Twentieth Century (2003) and Translation Zones in Modern China: Authoritarian Command Versus Gift Exchange and Fictional Authors (2011).

Anders Hansson holds a Ph.D. in History and East Asian Languages from Harvard University. He has been the cultural attaché at the Swedish Embassy in Peking, taught Chinese language and culture at the University of Edinburgh, and was the editor of the translation journal Renditions in Hong Kong. At present he is the chief editor of publications at the Macau Ricci Institute. He is the author of Chinese Outcasts: Discrimination and Emancipation in Late Imperial China (1996), and the co-editor of The Chinese at Play: Festivals, Games and Leisure and Chinese Concepts of Privacy (both 2002). He has also written on Chinese popular music.