The various indigenous groups of south China were (and are) the remnants of what had been far more widespread peoples displaced by the ever-expanding imperial dynasties of China, which embarked on aggressive programmes of colonisation and acculturation in Guangdong, Hunan, Jiangxi and Fujian. By the Qing period, two regions remained broadly indigenously populated: the first was a broad zone in the southwest, encompassing Yunnan, Guizhou, southern Sichuan and western Guangxi; the second was Taiwan. Yunnan and Guizhou had been part of the indigenous Kingdom of Dali before their conquest by the Mongols in the mid-13th century, and subsequently fell under the control of the Ming. However, apart from a few fortified routes to administrative capitals and military garrisons, the Han Chinese presence in Yun-Gui had largely been quite limited. Taiwan, meanwhile, had not been part of any native Chinese dynasty, but its Dutch-held portions were conquered by Ming loyalist remnants in the 1660s. Resistance against both the Dutch and Chinese presence by the indigenous peoples had left Han Chinese settlement constrained to what are now Tainan and Taipei and a narrow strip along the coastline, and after the Qing conquered the island in 1685 an agreement was reached whereby further Han settlement onto indigenous land would be tightly restricted.

That Han penetration into these regions was limited, especially before and after the Ming, is unsurprising. The cooperation of indigenous auxiliaries from Dali had been of immense importance in the Mongol conquest of the Southern Song, and the Mongols' relative distrust of Han administrators outside China had led to them instead introducing Hui Muslims as their enforcers in the region, who remained after the Ming conquest, including one admiral by the name of Zheng He. For the Manchu Qing, allowing potentially dissident Han Chinese to strengthen their foothold on an easily-defensible island, or to establish themselves in the mountainous hinterlands of the southwest, would have been potentially disastrous.

But back to the indigenous peoples themselves. Within Guizhou, there were eight major indigenous ethnicities, as defined by Qing ethnographers (many of these categories survive today, if under different names): the Miao (who refer to themselves as Hmong), the Zhuang, the Zhongjia (a.k.a. Yiren or Buyi; it is unclear what distinguished them from the Zhuang), the Gelao, the Luoluo (now known as the Yi), the Yao, the Nong (a.k.a. Long in some accounts), and the Dongjia/Dongren (sometimes seen as a Miao subgroup). As the Miao-Hmong case shows, these are all exonyms, and in many cases they are generalisations. The label of 'Yao', for example, encompasses three separate linguistic groups, and members of the overall 'Yao' classification have over twenty different names for themselves. The extent to which these were 'tribal' again depends, but given the increasing detail in which ethnographic manuals classified the indigenous peoples of Guizhou, it does suggest that self-identification by these peoples was largely quite fragmented.

As for their lifestyles, they were settled agriculturalists rather than nomadic pastoralists, being situated in a region of rugged but not infertile terrain, and largely settling along river valleys. Any contemporary discussion of these peoples is unfortunately tainted by exoticism, with Chinese observers quite deliberately engaging in othering of what they saw as primitive peoples, but still, it can be broadly said that they differed significantly in terms of language, religion, clothing, and material culture, although different groups were deemed as having greater or lesser proximity to Chinese practices. The Zhuang in particular were often said to be relatively 'Sinicised', the Zhongjia to a somewhat lesser but still appreciable extent. The thing to say here is that there is so much variety it is more or less impossible to generalise.

(Note: from here on out, as I'm discussing Qing policy, I'll be using 'Miao' in the more generic sense to refer to all the indigenous peoples in Guizhou and Yunnan.)

As for why they could be a potent military threat, it was less due to raw numbers and more due to terrain. South China, and especially southwest China, is an incredibly rugged place, with narrow river valleys running through rocky highlands, and with much of the region (save for northern Yunnan) having an average annual precipitation greater than 1000mm, which is not as wet as the south coast but still distinctly wetter than the Yellow River region. As such, any indigenous insurrection was going to be extremely difficult to counter except by slow attrition.

Indigenous relations with the Qing underwent significant alterations over time, but I am only reasonably up to speed on developments in the 18th century. As late as 1721, chunks of Guizhou were still (literally) uncharted territory, with the Kangxi Atlas leaving blank spaces in two areas in the southern part of the Guizhou map. The next year, however, the Kangxi Emperor died and was succeeded by his fourth son, In Jen, the Yongzheng Emperor. Yongzheng's major policy project in the southwest was known as gaitu guiliu 改土歸流 – literally 'altering earth to return to the flow'. It sounds cryptic, but tu was short for tusi 土司 (lit. 'earth official', known as aiman i hafan ᠠᡳᠮᠠᠨ ᡳ ᡥᠠᡶᠠᠨ in Manchu), tribal chiefs recognised by the dynasty as legitimate authorities over their particular constituents, while the 'flow' simply refers to the model of administration as employed with the Han Chinese in most of China. Its effects were dramatic: the Mongols had appointed around 500 hereditary tusi, by Yongzheng's death there were 24. Indigenous land was forcibly seized and partitioned as tuntian, resurrecting the Ming system of establishing military colonies, where land ownership was secured by military service. Miao were also told that they would be taxed and subject to corvée labour, with quotas identical to those for the Han, and that refusal would be a capital crime. Lying behind this was the particular ideological leanings of Yongzheng, who was perhaps the only Qing emperor who genuinely bought into the Confucian idea of cultural transformation, and had the least qualms about maintaining ethnic distinctiveness (his policies towards the Manchus are very much comparable). The severity of his policies, facilitated by his key minister Ortai (who remained influential until his own death in 1745), created much resentment among the indigenous peoples for it. In 1735, they revolted en masse.

Discrimination had of course predated the Yongzheng reign, as laws promulgated under the Kangxi Emperor in 1703 had mandated disproportionate retributive justice in the region: murder of one Han by a Miao demanded two Miao lives; a Han dying in a Miao village would not be investigated, but instead be avenged by an attack on the village as a whole; any Miao who owned weapons was to be considered a rebel and executed. Still, there was, eventually, the recognition that the Qing had gone too far. Perhaps rather fortuitously, the Yongzheng Emperor died in the middle of the Miao revolt, and his own fourth son, Hung Li, ascended the throne as the Qianlong Emperor, in a position to attempt to arrest the extremes to which his father had gone. While the military crackdown had to continue if the Qing were to retain control of the region, in its wake the Qianlong Emperor exempted the Miao from tax and corvée, allowed local custom to trump the Chinese legal code in indigenous disputes, and reverted to a more accommodating stance in general. One particular manifestation of this was the encouragement of greater compilation of ethnographic material in gazetteers in order to better understand the situation and reach a peaceful compromise, rather than assuming that Sinicisation by force could actually work.

During the Qianlong period and indeed the Qing period more generally, there was much thought as to how to regard the indigenous peoples. Were they min 民 (people) alongside the Han, descending from a common ancestor, but had chosen to embrace primitivity? Were they not min at all, but in fact animals? Or, was min not a singular category, but were there instead several kinds of min, with the Han and indigenous peoples belonging to different groups? Were they simply behind along a common trajectory of human development, or by nature consigned to 'primitivity'? Many Han observers saw their lifestyles as 'pitiful or contemptible', to quote William Rowe, but there was also an emergence of a manner of thought not unlike Rousseau's idea of the 'noble savage', which saw the apparent cultural and technological 'backwardness' of the indigenous peoples as being accompanied by a certain purity and admirable simplicity – as with romanticisations of Native Americans by Europeans. As said, what we can know about indigenous lifestyles must come, unfortunately, through an exoticised lens.

One peculiar aspect of how the Qing relationship with indigenous peoples worked, however, is that the Qing never clearly designated them as a constituent part of the empire, along with Manchus, Han, Mongols, Tibetans and Muslims (and yes, that is part of the reason behind the Five Races Under One Union flag used by the early Republic). Why that is is not necessarily wholly clear, but it is perhaps because there was no way to construe the necessary ideology of monarchy, baseline cultural practices and linguistic unity that might be possible for the other groups (though to be fair, the reconciliation of Islam with Qing emperorship was never really successful, and there was no clear overally linguistic unity among the empire's Muslims, with Chinese speakers in China proper and Turkic speakers in Xinjiang).