Demo of the 3D Model of Mount Tianzhu by Gao Chao

A major research aim of Altergraphy is to contextualize the inscriptions, that is, replacing them in the landscape at various scales—the scale of the territory, the scale of the mountain, the scale of the rock formation or cliff—and temporalities/viewers—the immutable zenithal eye or as the crow flies, the ferretting(?) eye or as the fox runs, the remote eye of literary narratives, or the hyperrealistic eye of a scanner treating each asperity of the rock as if it were a micro-landscape…

To gather this complex and refined palette of ways to record and visualize/render the four inscribed mountains by ZDZ during fieldwork, we had several members of the team assigned to each scale of enquiry:

• The reconstruction of mountains through drone photogrammetry
• The mapping of mountain paths and heritagized stone stairs
• The scanning of writing supports and landmarks

First trials in Inner Mongolia

When Lia first contacted Gao Chao (PhD candidate at the Beijing University of Civil Engineering and Architecture) through historian and architect Ren Congcong, he was on fieldwork in Xinjiang documenting vernacular dwellings, together with the researchers from his team. Prior to accepting the challenge and joining us on fieldwork, Gao Chao tested the drone scanning and reconstruction of a mountain near his hometown in inner Mongolia, in the Yingshan mountain range, famous for its rock art. He sent us a tentative model of the mountain he climbed – he later said the climb was really hard because the slope was so abrupt! – and we were seduced by the result.

We agreed that he would define together with us a capture area (fanwei) around each mountain – an issue that we raised already when building our cardboard models in Brussels, and when discussing fieldwork protocole at the doctoral school in Gent earlier in June (problems such as: the location of the Tianzhushan ming and eastern Wei Buddhist cave in TZS, the Northern mountain gate Beishanmen in DJS, the Japanese stele forest in YFS and just the general problem of tree coverage, the relationship between the inscribed and non-inscribed – jiguanshan rooseter crest peak – in LLS, where most of the interesting rock formations are subterranean, too).

Then, Gao Chao would also try to approach some of the most clearly/explicitely adressed rock formations in the inscriptions (location and content): the stone chamber and 秋遷口 in TZS, the peaks selected as altars (including the southern and northern mountain gates) in DJS, the left and right que pillars in YFS, the very shape of the mountain (and the Baiyuntang cave) in LLS. Some of these rock formations, we also scanned with our smartphones.

Finally, Gao Chao also gradually integrated our research questions by thinking the perception, description and inscription of the landscape … he adapted his choice of domain and rock formation according to the new findings (gorge adjacent to the stone chamber in TZS and stupa base next to the Eastern Wei cave also in TZS) and research questions (including modern additions such as the stele forests, or non-inscribed locations visible on the way up to or from the top of LLS…). He also delved into the content of the 6th century inscriptions (reading aloud to us the Guan haitong shi while we were measuring and scanning it, or communiating with the spirit of the Qing dynasty officials who drew invisible lines on mountain crests like in DJS…), ultimately building a role of his own in the team play…

Gao Chao’s journal

1 Sept – first day in TZS: in dialogue with Paula Sumera, defining the perimeter to be captured (the main mountain slope where all 6th century inscriptions are located and the side path where an earlier (186 CE) inscription stands, the 6th century Buddhist cave on its western foot, and the modern stele forest at the eastern end of the area).

2 Sept – second day in TZS: Gao Chao reconstructed the Western zone of the mountain (pushing the caption area to include the Eastern Wei dynasty cave and adjacent stupa, where the inscription by Zheng Daozhao’s son – the Tianzhushan ming by Zheng Shuzu – also used to stand) and the area around the Han dynasty “stele”. Later in the day, he also flew in the eastern zone again to reconstruct the stone chamber and the adjacent gorge in a more refined resolution, to be able to later zoom in these specific landmarks.

3 Sept – first day in DJS: Gao Chao went to the top of Xuanlinggong, the northern altar, as soon as we arrived in the valley, together with Lia, MF and Hanwen. — TT and Lia brought lunch to Gao Chao, MF and Hanwen in Xuanlinggong — PM Lia brought Gao Chao to Baiyuntang, he flew from there all afternoon and she scanned a first time all the Baiyuntang inscriptions. Both stopped on their way to charge Gao Chao’s remote control by an ice cream guy in the valley.

4 Sept – second day in DJS: Day of rest for all students but TT and Francesca. We decided to document the hardest to access inscriptions: Zhuyangtai, Qinyansi and Beishan men, together with Dir. Li and Gao Chao, who needed to complete that “quarter” of the valley. Gao Chao stopped on a summit just below Qinyansi, and while he flew the drone we climbed to the radio? — Tower on top of Qinyansi, the highest crest of the valley. Then Lia fetched Gao Chao (also because she did not document the path between Qinyansi and Zhuyangtai), and we all walked down by the path created by the radio tower workers. — in the afternoon Gao Chao went back to Xuanlinggong

5 Sept – third day in DJS: rest for Gao Chao (bad weather: he wrote his thesis at the hotel in the mean time)

6 Sept- first day in YFS: Gao Chao flew from the top all morning, focusing on the left and right que pillars, the peaks located on both sides of the summit. He could not do much because of the tree coverage. We had to supplement his work with accurate scans (iPhone powered) of the inscriptions’ writing supports and surrounding boulders.

7 Sept – second day in YFS AM : Gao Chao joined Group 1 (Killian, Kento, Gao Chao and Roxane) in documenting the inscriptions on top: Zuoque, Chongdeng Yunfengshan ji, Shanmen, Shizuo. — PM: Lia went back to the two groups, Group 1 documented the second large hutouyan/yongfeixianshi boulder and the guan haitong shi pavilion.

8 Sept – last day in LLS: Gao Chao, MF+Hanwen and Lia climbed LLS. Ma Longteng was waiting for us at the site, wanting to see the inscriptions, this was his first time in LLS. At first we used his car to get to the nongjiale (our driver did not dare to go up the road), then we departed from there, thus walking towards LLS from the north. We passed the alley with modern moya reproductions of calligraphic works (when were these created? Should we document them and with which accuracy? – including the 龍王鼻子。。。), and started the climb, staying on the widest path and resisting the temptation to try them all. When Ma realized he was with the kamikaze team, he stopped right away and went down the stairs all the way back to his car and to Pingdu… a two hours drive away. We continued the climb up to the crest, where the path is divided in two branches, one towards jiguanshan (rooster crest peak) on our right hand side (west), the other one towards LLS on our left hand side (east). Once passed the tree line, the path consisted in large limestone slabs, so fine that they seemed man-made – with Gao Chao we discussed the Chinese love for building stairs on top of mountains, associated it with the construction of the great wall, and wondered if ZDZ found the stone slabs as convenient as we did, and in which period they were modified to ease the climb. The whole landscape around felt like the bottom of the ocean, with layers of sedimented eons, and an overall smooth sensation, very far from the granite boulder we had seen since the beginning of the trip. Apparently Sakata described the landscape around like “resembling contour lines” (Kento). Gao Chao went up the peak next to the summit and stayed there all day, in the sun, without food. He ended up with an insolation but finished his job in one day. This was his last day with us on the field

Gao Chao was generally the first on top and the last to leave, with no break but to recharge his remote control (he brought a heavy bag with 8 batteries so that his drone could fly 8 hours straight), often seeing the inscriptions only on his way down, or on the second or last day when he was done with scanning. We thus had to bring him on top first, as quickly as possible and to define the perimeter of scanning without him knowing exactly where the inscriptions were located, this could constitute a problem. His work had to be complemented with close range scanning with our smartphones and Paula’s more accurate GPS locations, without us knowing exactly how these three kinds of data will be combined in the end.

There were difficulties in processing the data in terms of equipment and gaps between the GPS locations taken during different flights of the drone. We also had to think about the various ways in which one can display such models: cutting a video out of Gao Chao’s various proposals, and adding subtitles (further on we shall take into account the DISTAM summer school insights to achieve a cleaner annotation process), seemed the best option.

(2) Smartphone scanning

We were equipped with several Iphones with LIDAR (an Iphone 13, an Iphone 15 Pro and an Iphone 15 Pro Max), and tried two free softwares, Scaniverse and Polycam.

On day one in TZS, Paula experimented photogrammetry with Gao Chao and Xue Lei on this inscription, devising a way to adopt different angles around each character to capture its depth and the shape of the engraved stroke. Later on we will realize that if the movement of the phone is too exaggerated, the scanning software renders too much relief (wholes and protruding parts that not exist on the original cliff)

(3) Cartography

Since the early stages of our project, Paula started thinking about how we can represent the territory surrounding the inscribed mountains at the core of our research. Information retrieved from archaeological atlases and reports, as well as basic geomorphological data available as DEMs or inn other formats formed the base of her investigation. The questions of how to determine a catchment area around the inscribed mountain, how to document the paths and later infrastructure added to the sites, as well as the landmarks of interest within the site have been lingering for a few months in which Paula experimented with the available materials, tools and documents.

Now that she was on the field, a whole new range of possiblity arouse: beyond the models constructed by Gao Chao, she could directly walk or climb the paths connecting one inscription to the next, plot them on her app – Locusmap – but more importantly, have a direct, physical sense of the presence of the rock formations selected to carve inscriptions in. With Roxane, her companion in dismay, she ended up looking at each and every lithic candidate to find potential spots for inscriptions. She also found, on her way, the peaks re-carved by the “madman” Ren Zuoguan, and felt compelled to write about it. Hard to reach, barely visible inscirptions on five locations spread out in a large valley: if it weren’t for the Japanese calligraphy amateurs who pilgrimed to the sites, who would have noticed them after more than a millenium?

The questions remains: “How does one document the silhouette of a mountain without losing its soul? What happens when something else in standing in lieu of the inscription you have been looking for (is that another take on the “lieu-dit” 題名?)”. More importantly, the question that lead Paula’s enquiry, after months of looking at images of the peaks with different and unsatisfactory angles, was “How do I share the experience of this site with people who have not been there?”

---

The following text is an excerpt from the diaries of the participants to the September 2024 Altergraphy survey, which were edited and displayed at the exhibition “Sculpter les nuages, la montagne enlacée”.

Paula Suméra, Les arpenteurs du Mont Tianzhu¹

Les premières fois où Lia nous a parlé de ces quatre sites inscrits par Zheng Daozhao, elle peinait à nous dire la montagne. Nous voyions les estampages de ces inscriptions sur pierre, nous regardions les différentes prises de vues du site, à différentes échelles, prises à des périodes distinctes, mais il nous était toujours difficile de comprendre comment ces lieux fonctionnaient, quelle était la relation entre ces différentes localisations. Très tôt dans le projet, j’ai eu à coeur de me focaliser sur cet aspect spatial. Comment rendre compte des lieux à ceux qui n’y sont pas allés ?

J’ai passé une grande partie de cette année à me poser cette question, à essayer de comprendre les lieux sans les avoir vus, multipliant les observations de cartes, de photographies prises à des périodes différentes, questionnant sans cesse Lia. Par ses inscriptions, Zheng Daozhao semblait lier les lieux les uns aux autres, se créant une géographie propre. Ici, l’on évoque le mont que l’on voit au loin ; là, on dit se situer sur le pic central. Comment alors se contenter seulement des reproductions de ces inscriptions, dénaturées sans leur contexte spatial ?

Le mont Tianzhu me paraissait, avant le terrain, être le plus facile à appréhender. Une seule partie de la montagne est inscrite, le versant nord étant trop abrupt pour s’y aventurer. Une seule voie possible du pied de la montagne jusqu’à l’inscription la plus haute ; des marches de bitume qui dénaturent le paysage, qui forcent le chemin. C’est aussi le premier site que nous avons parcouru lors de ce séjour. Une montagne qui se découpe, isolée, dans une région de plaines au milieu des vignes, comme si elle avait subitement émergé. Le premier jour, je gravis avec Gao Chao jusqu’à l’inscription centrale « Ceci est le mont Tianzhu 此天柱之山 ». Juchés sur la plateforme de pierre devant le pavillon censé protéger l’inscription, on se met au travail. Le drone doit balayer toute la zone que nous voulons modéliser. Gao Chao m’explique comment il va s’y prendre ; penchée sur une carte affichée sur sa tablette, je lui montre les endroits que nous voulons voir apparaître sur le futur modèle. Ensemble, nous traçons un rectangle qui sera l’emprise du scan effectué par le drone.

Le drone s’envole. Maintenant, il n’y a plus grand-chose à faire. Attendre, et surveiller de loin la course de l’engin dans le ciel. Il fait très chaud, le soleil nous étouffe. Un moment passe ; une brise, je respire enfin. En face de moi, la réaction de Gao Chao est tout autre : oh non, c’est pas bon, ça. Le vent est plus fort, là-haut dans le ciel ; il pourrait dévier la trajectoire du drone, l’empêcher de voler correctement. Désormais, le ciel s’assombrit. On s’inquiète à nouveau, il faut finir avant que l’averse ne vienne, le drone n’apprécie pas l’eau non plus. Gao Chao ne finit ce premier balayage qu’en milieu d’après-midi. Le lendemain, il faudra encore faire voler le drone sur le versant ouest de la montagne, où se trouvent une grotte bouddhique et un stupa. Cette fois-ci, Gao Chao doit commander manuellement le drone. Après le traitement de ces photographies aériennes, il nous propose très vite – deux semaines après la fin du terrain – une modélisation complète du mont Tianzhu. Tout y est. Une montagne miniature. Seulement, la résolution n’est pas suffisante pour proposer des vues détaillées des inscriptions. Surtout, le drone permet seulement une vue à vol d’oiseau, certains lieux disparaissent. Alors, pendant le terrain, on complète ces modélisations avec des détails, avec ce que l’on ne peut pas voir depuis le ciel. Munis de nos téléphones, on fait des photogrammétries, balayages vidéo qui permettent d’effectuer des modélisations 3D.

On s’intéresse aux inscriptions et à leurs alentours proches, on recrée certaines conformations rocheuses que l’on juge pertinentes. On documente l’intérieur des grottes, des chambres de pierre. On recrée les sentiers. On veut tout documenter, donner un moyen de laisser voir aux autres cette montagne qu’ils n’ont pas parcourue.
Il y a toujours ces marches de béton qui nous gênent et qui nous forcent, comme si une seule déambulation était possible. Elles donnent l’impression que le passage d’une inscription à l’autre a toujours été aisé. Tchou-Tchou s’y refuse, part chercher de nouvelles voies dans la montagne. Roxanne documente son itinéraire sur le chemin officiel en laissant son traceur GPS activé sur son téléphone, via l’application Locusmap, que j’utilise également pour récolter les localisations précises des inscriptions.

Lia effectue un scan 3D complet du chemin. Elle enregistre d’autres voies possibles. Ici, on aperçoit quelques marches de pierres, désormais condamnées, mais qui semblent être un ancien chemin. Là, gravées à même la roche, des cavités de la taille d’un pied pour que le promeneur y place ses pas. Moi, je tente de rassembler toutes ces informations dans une seule carte, tant bien que mal. Une carte, c’est traître. Elle laisse croire qu’on peut suivre du bout des doigts un lieu, mais elle réduit ce même espace à la seule expérience de la vision, et à un seul angle de vue.

1. Extrait de texte présenté dans l’exposition. ↩︎

---

---