A while ago I became aware that historian James A. Anderson has published a new book entitled The Dong World and Imperial China’s Southwest Silk Road: Trade, Security, and State Formation (UW Press), which is available in open access form here (https://library.oapen.org/handle/20.500.12657/92971).
I promptly downloaded it, and saved it in a file somewhere that I labeled something like “Things to Read.” From the start, I had the plan of reading it and introducing it to others on the blog.
I read some of it, but then I got busy with other things. Today I tried to resume my reading, but I couldn’t find any folder anywhere labeled “Things to Read” or anything like that, so I downloaded it again, and put it in a new folder called “Anderson.”
This is the normal flow of events in my life, and what it ultimately means is that I will end up downloading the book maybe 14 times over a period of many months, by which point I’ll be creating a folder with a name like “000000 READ NOW” so that it will appear at the top of a drive, but when I go to read it, I’ll realize that it’s on a drive on a laptop that’s somewhere else, etc.
In other words, I’ll never finish reading it, because these days, I tend to not read things unless I am actively involved in researching about a topic that a book or article relates to, and at the moment, that’s not the case for me with this book.
So, I thought to myself, maybe this is where AI can be helpful. So, I uploaded the book to ChatGPT and asked it some questions.
Here they are.
1) What work did Anderson do before this book that prepared him for this project?
Before The Dong World, Anderson had already spent years mapping the Sino-Vietnamese frontier’s people, places, and archives. His first monograph, The Rebel Den of Nùng Trí Cao (2007), trained him on upland leadership and Song–Đại Việt statecraft—the very actors and institutions that reappear in this book.
He then sketched the big canvas in a 2009 essay on the Southwest Silk Road, testing the trade-route frame that becomes the book’s backbone.
He drilled down on Song-era logistics in a 2014 Asia Major study of frontier trade networks (“Commissioner Li and Prefect Huang”), and widened the comparative lens by co-editing the 2015 Brill volume China’s Encounters on the South and Southwest.
Alongside the publications, he kept up field ties—returning to Cao Bằng sites he had first visited two decades earlier and working out of the Sino-Nôm Institute—grounding the book’s archival arguments in a lived sense of the terrain.
In short, a decade-and-a-half of focused writing and field familiarity supplied the sources, scale, and regional intuition that The Dong World pulls together.
Ok, so this new book is a continuation, and expansion, of a topic that Anderson has been working on for his entire career.
2) Please summarize the book.
James A. Anderson tells the story of a mountain-valley frontier he calls the “Dong World,” stretching across the uplands linking southwest China to mainland Southeast Asia between the Tang and Yuan eras. Rather than a wilderness at the edge of empires, this world is shown as a busy corridor where trade—especially in horses—stitched together diverse communities and bound them to lowland states. Upland chiefs built authority by policing passes, running markets, and taxing caravans; courts in Chang’an/Kaifeng/Hangzhou, Dali, Pagan, and Đại Việt learned that stability depended less on conquest than on working through these intermediaries.
The narrative moves from Nanzhao’s rise after the An Lushan rebellion to the multiethnic Dali kingdom, where Buddhist kingship, horse breeding, and market towns underwrote power. On the Song frontier, officials gradually replaced warlord rule with civilian administration but still relied on upland brokers and regulated horse markets to provision the army. Conflicts between the Song and Đại Việt spill into this landscape, revealing how “border wars” were really contests over access to people, pack animals, and routes. The Mongol conquest finally reorders the region, formalizing long-standing arrangements through native-official systems that bureaucratize upland mediation rather than abolish it.
Across these episodes, Anderson’s through-line is consistent: the Southwest Silk Road was an economic and political infrastructure that made states possible on both sides of the mountains. Markets legitimized chiefs; chiefs made trade legible to courts. The book closes by noting modern echoes—from railways to cross-border logistics—arguing that new corridors continue to integrate the old Dong World without erasing its enduring habit of being governed through local intermediaries.
3) How does this work fit in with or challenge existing scholarship?
Anderson positions The Dong World within borderlands history but gently shifts its center of gravity. Rather than treating the southwest as a hazy periphery explained mainly by imperial categories, he adopts—and localizes—Richard White’s “middle ground,” showing how mountain-valley polities made politics through markets, brokers, and negotiated rules. He and John Whitmore’s earlier coinage of “Dong World” becomes the book’s organizing lens, framing an upland zone whose hybridity and trade networks tied Dali, Đại Việt, and the Song together.
This move also challenges the most popular recent synthesis of the region: James C. Scott’s Zomia. Where Zomia imagines a space “beyond the state,” Anderson reconstructs a constellation of small states and chiefdoms that were very much entangled with lowland regimes—economically, ritually, and militarily. The result is a portrait of uplands not as pure zones of refuge but as arenas of bargaining in which local rulers leveraged caravans, passes, and markets to extract recognition and resources.
Finally, the book recasts the tribute system from headline to footnote on this frontier. On the ground, the Southern Song built offices to supervise horse trade at Yongzhou/Hengshan, funneled all contact through that corridor, and at times explicitly refused Dali tribute while still permitting regulated commerce—an administrative choice that reveals how strategy rode on supply chains more than ceremony.
The Mongol-Yuan conquest then bureaucratized what had long existed informally, embedding upland mediation within the “native officials” (tuguan) system rather than erasing it.
In sum, Anderson supplements borderlands scholarship by replacing romantic statelessness and tribute-centric models with a trade-centered account of multiethnic state-making—one in which markets legitimate chiefs, chiefs make exchange legible to courts, and empires learn to govern through the brokers they cannot do without.
I’m glad to see that Anderson challenges James Scott’s Zomia idea. I would argue that Scott romanticized a space beyond the state where people wished to live to be free from state control, and I agree with Anderson that “small states and chiefdoms. . . were very much entangled with lowland regimes.”
I would also argue that this was often because the leaders of those small states and chiefdoms wanted that entanglement as they benefited from it. I asked if Anderson said anything like this, and I got the following response:
Yes—Anderson says that outright. He argues that trade ties with lowland courts gave upland elites concrete advantages, letting them “accrue political leverage with their lowland neighbors,” and that markets themselves legitimized local claims to rule.
He also shows how chiefs actively sought relationships that paid off: they traveled regularly to court, were enticed into alliance by lenient policies, and accepted titles for their honorific value and status—classic instruments of benefit and recognition.
At a larger scale, polities like Nanzhao and Dali expanded their authority precisely through tribute and trade with the Tang, exploiting peripheral positioning to strengthen themselves.
And when collaboration stopped being advantageous, upland leaders shifted back to intra-upland alliances—another sign that their ties to lowland regimes were instrumental because they brought benefits.
So your reading matches his: these were not reluctant subjects pulled into empire, but pragmatic brokers who pursued (and dropped) lowland connections when those relationships increased their wealth, status, security, and leverage.
Ok, that’s great to know.
Moving on to the tribute system, I had never heard the expression “from headline to footnote,” so I asked ChatGPT and was informed that: “From headline to footnote” is a metaphor: something that used to be the main story (the headline) has been demoted to a minor detail (the footnote). In our context, I meant Anderson treats the tribute system as secondary to the real driver—regulated markets and logistics.
Ok, now I got the metaphor, but I still wasn’t entirely sure what Anderson’s point about the tribute system was. However, the answer to my next question clarified the issue.
4) What is Anderson’s most surprising finding?
Anderson’s most surprising move is to show that, on this frontier, the Song state deliberately decoupled ritual diplomacy from practical logistics. In 1133–34 the court created a Department for the Supervision of Horse Trade at Yongzhou and funneled all contact through the Hengshan corridor—while at the same time refusing to accept Dali tribute, explicitly permitting only the sale of horses. That administrative choice overturns tribute-system expectations and reframes “foreign relations” here as supply-chain management.
Ok, now I saw what he is arguing. But I felt like this is something that must have been talked about before, so I asked another question.
5) I feel like there are already a lot of scholarly works on the horse trade or the tea trade in this area. How does Anderson’s new book relate to these earlier works? Does he add something? Challenge something?
You’re right that there’s a rich literature on the “tea–horse” trade—but Anderson’s move is to treat tea and horses not as self-contained commodity stories, but as the operating system of frontier governance. He argues that markets themselves legitimized upland authority and gave chiefs leverage with lowland courts; in his frame, trade isn’t background to politics, it’s the mechanism that made rule possible across the Dong World and tied south China to mainland Southeast Asia’s shifts.
Interesting.
Then, finally, I asked the most important question of all.
6) How would Anderson answer the question: If frogs had wings, would they still keep hitting their butts on the ground?
Channeling Anderson’s lens: giving frogs wings wouldn’t abolish the ground; it would reorganize their relationship to it. In the Dong World, mobility (horses, passes, markets) changed the costs of movement but never erased the valleys that structured power. Wings are just new infrastructure.
So the frogs would smack their butts less—but they’d still have to launch and land at specific “nodes” (perches, thermals, feeding spots). Those nodes would become the new chokepoints, governed by intermediaries who tax and broker access, much like upland chiefs at mountain passes or the Song at Hengshan horse markets. Ritual (ribbits) might be decoupled from logistics (wing-beats), but gravity—and governance—remain.
In short: fewer bumps, more bargaining. The story shifts from how hard the ground hurts to who controls the ledges.
Wow!! Anderson has an answer for everything!!
In conclusion, I don’t have time to read a 300-page book that’s not related to my research right now.
If we follow the “traditional” way of doing things, then I can wait a few years for someone to write a normal 700-word review (what AI and I wrote above, minus the intro and the frog butt part is twice that length, hence, too long for many publishers who, for some unknown reason, still limit book reviews to lengths developed when those reviews were printed on paper), however, there is no guarantee that I’ll ever notice that review (as it will probably be somewhere behind a paywall), and it won’t necessarily tell me what I want to know about the book.
I suspect that I’m not alone. I want to know what this book is about and I want others to know about it too.
Through this brief AI interaction, I got some clues about issues covered in the book that I might want to focus in on should the stars align and I get the chance to dig into it more deeply.
Perhaps by me doing this, others will feel the same. I hope so.
In any case, I would like to congratulate Jamie on the publication of this book!! I think that it is particularly impressive that he has amassed a body of work that builds and expands on a single, albeit large and multifaceted, topic. That’s great!!