DATE
6/10/25
TIME
9:22 AM
LOCATION
Oakland, CA
Hunanese As a Ethnicity
何为湖南人?
0-1:
我原先想写“中国人”这个概念的历史,可惜这个话题太大,完全不知道从何开始。先不说“中国”的疆域一直在改变,各政权的民族、习俗完全不同,甚至各文化可能从完全不同的路线、从非洲板块迁入今天意义上的“中国”的地盘。而关于这些智人、能人是否是从非洲板块迁入的,目前学术界都还没有定论。我只好把话题缩小,从湖南人的历史来看。究竟湖南人的ethnicity是什么意思,需要往上追溯多久,以及我到底是所谓的湖南人吗?我很怀疑。
湖南地区最早的化石证据似乎是永州道县的福岩洞人,但具体该化石是什么年代的,争议激烈。发掘的团队,刘武,认为可能是距今8–12万年的早期现代人。而另一团队,李辉团队,则认为或仅为9000年前。
化石在2010–2011年间,被发掘于湖南省永州市道县乐福堂乡塘碑村福岩洞,由中国科学院古脊椎动物与古人类研究所刘武、吴秀杰等主导发掘。出土物包括47颗人类牙齿化石,以及丰富的哺乳动物骨骼和石制工具。在这篇2015年发表于Nature的文章 《The earliest unequivocally modern humans in southern China》,他们主张:他们发现的47颗人类牙齿,测年显示最早可达距今12万年,形态上属于完全现代人的解剖学特征。也就是说,南中国可能比中东与欧洲早了整整三万到七万年拥有“真正的人”。这个发现不仅挑战了“非洲单一起源”的晚期扩散论,这好像是主流pop culture认同的论调,在《Sapiens: A Brief History of Humankind》 里也是这么说的。
但刘武的这个“发现”,如果是靠谱的,会完全推翻这个主流看法。这意味着湖南这片土地,可能是世界最早“现代人群落”的所在地之一。
0-2:
亚洲南部在更新世早期的古人类化石记录十分稀缺,尤其是能明确归属于现代人类 Homo sapiens、且测年明确的材料。而在中国湖南道县的福岩洞发现的这 47 颗人类牙齿,用他们的铀系测年显示这些化石距今至少8万年,最久可达12万年。牙齿的形态特征高度现代,明显属于解剖学上的现代人类。
他们使用的铀系测年(U-series dating),对牙齿上方的钟乳石进行测定,而不是牙齿本身。铀元素会放射性衰变成钍。这个衰变过程是稳定可预测的,半衰期为大约75,000年。通过分析 Th-230 / U-234 的比值,可以推算材料的形成年代。
虽然我们并不熟悉每个元素的特性,但化学元素表大家都还是背过的。有这个元素周期表的基础,很多资料可以自己去查到。比如这里使用的铀,读作you2。由来的由一个读音,英语是Uranium,也是《Oppreheimer》里面制作原子的主要原料之一。以下维基百科:
“已知的铀同位素都不稳定,其中以最长寿的铀-238(半衰期44.7亿年)和铀-235(半衰期7.04亿年)在自然界中最为普遍。铀是在地球上大量存在的太初元素中原子序最高的,原子序大于铀的超铀元素由于半衰期较短,从地球诞生至今早已衰变殆尽,且现今自然界中也缺乏形成它们的途径或机制,因此都是以人工合成的方法发现的,仅有錼和钸等原子序较小的超铀元素被发现在铀矿中痕量生成。
自然界中的铀以三种同位素的形式存在:铀-238(占天然铀的99.2739至99.2752%)、铀-235(占0.7198至0.7202%)、和微量的铀-234(占0.0050至0.0059%)。天然铀在衰变时会释放出α粒子。[5]由于天然铀同位素的半衰期极长,因此它们被用于估算地球的年龄。
铀独特的核子特性有很大的实用价值。铀-235是唯一易分裂的天然铀同位素,可被慢中子撞击而裂变,如果其质量超过临界质量,就都能够维持核连锁反应,在核反应过程中的微小质量损失会转化成巨大的能量。这一特性使它广泛被用于核能发电以及生产核武器。
然而,其在大自然存在的浓度很低,必须经过浓缩方可使用。铀金属具有相当的反应性,在空气中表面会形成深灰色氧化层。天然的泥土、岩石和水中含有百万分之一至百万分之十左右的铀。采矿工业从沥青铀矿等矿物中提取出铀元素。”
而铀系测年主要依赖两个放射性衰变链,U-238 → Th-230和U-234 → Th-230。在自然界中,U-238 和 U-234 是可溶的,能被水带入钟乳石或骨骼,但 Th-230 几乎不溶于水,只能通过衰变产生。因此,当你发现某个样本中有 Th-230,它的来源只能是由 U-234 衰变而来。
而,由于U-234 的半衰期约 245,000 年,而Th-230 的半衰期:约 75,000 年,因此根据它们的比值(Th-230 / U-234),可以反推出碳酸钙沉积物的形成时间。
在铀系测年中,最常用的就是 钍-230。通过铀-234衰变到钍-230,和Th-230 的半衰期为约 75,000 年,通过计算 Th-230 / U-234 的比例,推算材料形成时间。
“Thorium” 这个名字来自北欧神话中的雷神 Thor。命名者是瑞典化学家 Berzelius,1828 年首次发现。钍读作“土”,tu3,属于弱放射性元素。在自然界中广泛存在,常见于钍矿、独居石等矿物中。曾被研究作为核反应堆燃料的替代品,因为比铀更丰富、更安全,但技术未大规模应用。钍的硬度一般,具顺磁性,为亮银色的放射性金属元素。纯钍的延展性相当好, 与一般金属一样,可被冷轧、挤锻及拉制。
0-3:
需要注意的是,这里铀系测年测量的并不是测量的被发现的牙齿的年份,而是牙齿上覆盖的钟乳石推定牙齿“最早”在那个时间之前就已存在。这叫 “下限测年”。
“下限测年”(minimum age dating)是指当人类学家无法直接测出一个东西确切有多老,但能确定它至少有多老。也就是说我们不能说这个化石是10万年,但可以说“它不可能比8万年晚”。刘武团队没测牙齿本身,而是测了牙齿上方覆盖的钟乳石,也就是钟乳石是以牙齿为基础,形成的。钟乳石是8万年前形成的,那么下面的牙齿不可能比这更新,否则钟乳石就不会长在它上面。所以我们可以说这颗牙齿至少比8万年老。覆盖物的年代可以告诉我们底下那个在就在那里,但无法确定具体在那里呆了多久。
牙齿本身不容易直接测出准确的年代,尤其是老得离谱的那种。首先,牙齿本身不是一个封闭系统,放射性测年要求样本内部的元素是封闭系统,不能被水或空气交换。牙齿暴露在土壤里几万年,可能进了水、漏了矿物质,测出来的年代就会不准。同时,牙釉质(enamel)硬度极高,基本不溶、难以吸收铀,但测年方法往往需要有一定量的铀元素作为起始点,如果牙齿几万年都没吸够铀,根本没法算出可靠的衰变链,就没法测。
铀系测年,靠的是一开始有铀,然后看它衰变成钍的过程。如果一开始铀含量太低,或者根本没有铀,那后面也就没什么可衰变的东西,等于这个时间钟表根本没上发条。牙齿是生物组织,不是矿物,它一开始形成时,不含什么铀元素。
所以科学家指望的是牙齿埋进土里之后,地下水里的铀慢慢渗进牙齿,牙齿才算正式启动衰变计时器。如果土壤贫瘠 ,没有足够铀,或者牙齿位置太干、太深,水无法渗入,或者吸铀过程断断续续、时间跨度不一,衰变链乱套,没法建模。或者吸得太晚 ,测出来的年代比真实年代还晚几万年。
在福岩洞牙齿的争议中,刘武团队使用的就是这种测定牙齿上方钟乳石的铀系测年法,得出“牙齿至少8万年前”这个结论。但刘武发完article之后,李辉等人重新测量了牙齿的年代,并得出牙齿不过9000年的说法。
0-4:
2015年,刘武团队在 Nature 上发表论文,主张福岩洞47颗牙齿距今约8–12万年,属早期现代人。随后几年,李辉团队通过正式申请、合作研究或取样授权,取得了部分原始牙齿样本或副本进行独立测年分析。2021年,李辉团队在《PNAS》上发文《Ancient DNA and multimethod dating confirm the late arrival of anatomically modern humans in southern China》,运用多重测年法,重新评估这些牙齿的年代,得出约9000年的结论。
李辉的团队使用了电子顺磁共振测年法(Electron Spin Resonance)来测量牙齿,是一种测量物质中电子被困能态的“年轮”方法。牙齿埋进土里后,会不断受到环境中天然辐射铀、钍、钾等的轰击。前面也说了,铀会衰变成钍,用铀的半衰期和牙齿上包裹的钟乳石的铀钍比例可以推算出年份。这里他们用了同样的放射元素为基础进行测量,但是不同的方式。
这里是直接使用牙齿里面,被这些辐射打出原位的电子进行测量。这些电子通常被打出原位后,被卡进比如牙釉质中的羟基磷灰石晶格里。电子就这样被困在牙齿内部,年复一年地积累下来。羟基磷是牙齿和骨头的骨架,牙釉质里96% 是羟基磷灰石,而骨骼里65–70% 是羟基磷灰石。
牙釉质(enamel)是我们牙齿最外面那一层半透明的“白壳”,硬、无血管、不再生、富矿物。人们补牙就是因为这个外壳无法再生,只能靠补的,身体本身不能再长出来了,除非你研究出什么干细胞之类的我猜。这种羟基磷灰石提供极高硬度,保护牙齿不被磨损,同时也构成骨头的无机部分,负责支撑和钙储存。
而Electron Spin Resonance是1944年被苏联物理学家叶夫根尼·扎沃伊斯基发现,是属于自旋1/2粒子的电子在静磁场下发生的磁共振现象。这个自旋 1/2粒子大一化学课学过,虽然细节不了解,但是知道有这么个东西。由于分子中的电子多数是成对存在,每个电子对中的两个电子必为一个自旋向上,另一个自旋向下,所以磁性互相抵消。因此只有拥有不成对电子存在的粒子,才能表现磁共振。也就是这些被打出原位,卡在羟基磷灰石晶格里的电子。
Electron Spin Resonance测年用微波激发这些电子,测出有多少电子被困,形成剂量值。同时测周围土壤每年辐射剂量率,也就是土壤每年释放多少铀?这样电子被困数除以剂量率,就是牙齿可能的年龄。
而测量周围土壤的年辐射剂量率,李辉的团队使用了Optically Stimulated Luminescence,中文叫做光释光测年。这个方法我不是很理解,似乎需要保证土壤不见光才有效。这些牙齿被挖掘,保存,转运,又到刘武手上,我很难相信这过程中完全没有见光。因此我对这个测量结果保持怀疑。两边比较来看,可能还是刘武的结果比较靠谱,虽然二者的测量方式都很容易产生偏差的样子。
0-1
I originally wanted to write about the history of the concept of “Chinese people,” but the topic turned out to be far too vast—I had no idea where to begin. Not to mention that the territory of “China” has constantly shifted over time, with completely different ethnic groups and customs under various regimes. Some of these cultures might have even entered what we now call “China” by entirely different routes, possibly migrating from the African tectonic plate. And whether these Homo sapiens or Homo erectus actually migrated from Africa is still an unresolved debate in the academic world. So I had to narrow the scope and look instead at the history of Hunan people. What exactly does “Hunanese ethnicity” mean? How far back do we have to trace it? And am I really what they call a Hunanese? I seriously doubt it.
The earliest fossil evidence in the Hunan region seems to be the Fuyan Cave Man from Daoxian, Yongzhou, but there’s intense debate about how old these fossils actually are. The team that excavated them, led by Liu Wu, suggested that they might be early modern humans from 80,000 to 120,000 years ago. Another group, led by Li Hui, argued that they might be only about 9,000 years old.
The fossils were unearthed between 2010 and 2011 at Fuyan Cave in Tangbei Village, Lefutang Township, Daoxian County, Yongzhou, Hunan Province, in a dig led by Liu Wu and Wu Xiujie from the Institute of Vertebrate Paleontology and Paleoanthropology (IVPP). They discovered 47 human teeth, along with many mammal bones and stone tools. In their 2015 paper published in Nature, The earliest unequivocally modern humans in southern China, they claimed these 47 teeth, dated using uranium-series methods, could be as old as 120,000 years and morphologically belonged to anatomically modern humans. That would mean southern China had “real humans” 30,000 to 70,000 years earlier than the Middle East and Europe. This discovery challenges the late dispersal theory of the “Out of Africa” model—something that mainstream pop culture tends to accept, and which is also echoed in Sapiens: A Brief History of Humankind.
If Liu Wu’s “discovery” holds up, it would completely overturn the mainstream consensus. It would mean that this piece of land—Hunan—could be one of the earliest known homes of modern human communities on Earth.
0-2
Early Pleistocene fossil records of hominins in southern Asia are extremely rare—especially those that can be clearly attributed to Homo sapiens and have reliable dating. The 47 human teeth found in Fuyan Cave, Daoxian, Hunan, were dated using uranium-series dating, which suggested the fossils were at least 80,000 years old, possibly up to 120,000 years. The morphology of the teeth is highly modern, clearly consistent with anatomically modern humans.
The team didn’t date the teeth themselves but rather the stalagmites above the teeth, using uranium-series (U-series) dating. Uranium decays radioactively into thorium. This decay process is stable and predictable, with a half-life of around 75,000 years. By analyzing the Th-230 / U-234 ratio, researchers can infer the age of the carbonate material.
Even if we’re not familiar with the specific properties of every element, most of us memorized the periodic table in school. With that foundation, you can look up a lot of this information yourself. For example, the uranium used here is pronounced you2 in Chinese. The word was phonetically borrowed; in English, it’s “uranium”—also the core material used to build the atomic bomb in Oppenheimer. From Wikipedia:
“All known isotopes of uranium are unstable. The most long-lived are U-238 (half-life 4.47 billion years) and U-235 (half-life 704 million years), which are the most common in nature. Uranium is the heaviest primordial element still found in large quantities on Earth. Elements heavier than uranium (transuranic elements) have short half-lives and have mostly decayed since Earth’s formation. They are not formed naturally today and are only created artificially, although trace amounts of some like neptunium and plutonium have been found in uranium ores.
Naturally occurring uranium consists of three isotopes: U-238 (99.2739% to 99.2752%), U-235 (0.7198% to 0.7202%), and trace amounts of U-234 (0.0050% to 0.0059%). Natural uranium emits alpha particles as it decays. Due to their extremely long half-lives, these isotopes are used to estimate the age of the Earth.
Uranium’s unique nuclear properties have high practical value. U-235 is the only naturally fissile uranium isotope, meaning it can undergo fission when hit by a slow neutron. If its mass exceeds the critical threshold, it can sustain a chain reaction. Even a tiny loss in mass during the reaction can release massive amounts of energy, which makes it ideal for both nuclear power generation and nuclear weapons.
However, its natural concentration is very low and it must be enriched before use. Uranium metal is highly reactive and forms a dark gray oxide layer when exposed to air. Natural soils, rocks, and water contain about 1 to 10 parts per million of uranium. The mining industry extracts uranium from ores such as pitchblende.”
U-series dating relies primarily on two radioactive decay chains: U-238 → Th-230 and U-234 → Th-230. In nature, U-238 and U-234 are water-soluble and can be transported into stalagmites or bones, but Th-230 is nearly insoluble in water and only forms via decay. So if you find Th-230 in a sample, it must have come from the decay of U-234.
Because U-234 has a half-life of about 245,000 years and Th-230 about 75,000 years, their ratio (Th-230 / U-234) can be used to infer the age of calcium carbonate deposits.
Thorium-230 is the most commonly used isotope in U-series dating. By calculating the Th-230 / U-234 ratio and considering Th-230’s half-life of 75,000 years, researchers can estimate when the material formed.
The name “Thorium” comes from the Norse god Thor. It was first discovered in 1828 by the Swedish chemist Berzelius. Thorium is weakly radioactive and pronounced tu3 in Chinese. It is relatively common in nature, found in thorite and monazite. It was once studied as a safer and more abundant alternative to uranium in nuclear reactors, although it has not seen widespread use. Thorium is a soft, silvery, paramagnetic metal. Pure thorium is quite ductile and, like most metals, can be cold-rolled, extruded, and drawn into wire.
0-3
It’s important to note that what uranium-series dating measures here is not the age of the teeth themselves, but rather the age of the stalagmites covering the teeth, which is then used to infer that the teeth must have existed before that point in time. This is called “minimum age dating.”
Minimum age dating refers to cases where anthropologists can’t directly determine how old something is, but they can determine how old it must at least be. In other words, we can’t say this fossil is definitely 100,000 years old, but we can say: “It can’t possibly be younger than 80,000 years.” Liu Wu’s team didn’t measure the teeth directly, but instead measured the stalagmites that had formed over them. Since stalagmites are formed by calcium carbonate deposits accumulating over time—on top of something—the logic is simple: if the stalagmites are 80,000 years old, then the teeth must have been there already for the stalagmites to form on top of them. So the conclusion is: the teeth must be at least 80,000 years old. However, the age of the covering material tells us when the “roof” formed—it doesn’t tell us how long the teeth had already been there before that.
Directly dating the teeth is tricky, especially when we’re talking about extremely old specimens. First, teeth are not closed systems. Radiometric dating requires the sample to be a closed system—meaning no exchange of material with the environment. But teeth, buried in soil for tens of thousands of years, can absorb water or lose minerals, throwing off the dating results.
To complicate things, tooth enamel is extremely hard and chemically stable—basically insoluble—so it doesn’t absorb uranium easily. But dating methods often require a certain amount of uranium to kickstart the decay chain. If the tooth hasn’t absorbed enough uranium over thousands of years, there’s no meaningful decay chain to measure—and no date to calculate.
U-series dating depends on there being uranium to begin with, and then watching it decay into thorium over time. If the initial uranium content is too low—or nonexistent—then there’s nothing to decay. It’s like trying to read a clock that was never wound up.
Teeth, being biological rather than mineral, contain almost no uranium when they form. So scientists hope that after a tooth is buried in soil, groundwater carrying dissolved uranium will gradually seep in. That’s when the “decay clock” starts ticking. But if the soil is poor in uranium, or the burial environment is too dry or too deep for water to reach the tooth, or if uranium seeped in inconsistently over time, then the decay model falls apart. Worse, if uranium absorption occurred very late, the calculated age might be tens of thousands of years younger than the actual burial time.
In the Fuyan Cave controversy, Liu Wu’s team used exactly this kind of “stalagmite-over-tooth” U-series dating method and concluded that the teeth must be at least 80,000 years old. But after Liu’s article was published, Li Hui and others re-measured the teeth themselves and concluded that they were only about 9,000 years old.
0-4
In 2015, Liu Wu’s team published a paper in Nature, asserting that the 47 human teeth from Fuyan Cave were 80,000 to 120,000 years old and belonged to early anatomically modern humans. In the following years, Li Hui’s team formally applied for sample access, conducted collaborative research, and obtained some of the original teeth or replicas for independent dating. In 2021, they published a paper in PNAS titled Ancient DNA and multimethod dating confirm the late arrival of anatomically modern humans in southern China, using multi-method dating to reassess the age of the teeth and concluding they were only about 9,000 years old.
Li Hui’s team used Electron Spin Resonance (ESR) dating to measure the teeth—a technique that reads “annual rings” of radiation damage based on trapped electrons. After being buried, the teeth are bombarded with natural background radiation from uranium, thorium, potassium, and other elements. As mentioned earlier, uranium decays into thorium. While Liu Wu’s team calculated dates based on uranium and thorium ratios in overlying stalagmites, here the researchers used the same radioactive elements, but a completely different method.
In this case, they directly measured the electrons that were knocked out of position and got trapped inside the hydroxyapatite lattice of the tooth enamel. These electrons accumulate year after year within the crystalline structure. Hydroxyapatite forms the structural framework of both teeth and bones: tooth enamel is 96% hydroxyapatite; bones are around 65–70%.
Tooth enamel is the translucent white shell on the outside of our teeth—hard, avascular, non-regenerative, and mineral-rich. That’s why people get fillings: because once enamel is damaged, it doesn’t grow back—unless you discover some kind of stem cell-based regeneration method, I suppose. Hydroxyapatite provides that extreme hardness to resist wear, and also constitutes the inorganic backbone of bones, supporting structure and calcium storage.
Electron Spin Resonance was discovered in 1944 by Soviet physicist Yevgeny Zavoisky. It involves the magnetic resonance of spin-1/2 particles under a static magnetic field. (You probably learned about spin-1/2 electrons in general chemistry—maybe didn’t understand the details, but you know the term.) Since most electrons in molecules exist in pairs with opposing spins (up and down), their magnetic effects cancel each other out. Only unpaired electrons can exhibit magnetic resonance—such as the ones knocked out of place and trapped in the hydroxyapatite lattice.
In ESR dating, scientists use microwaves to excite these trapped electrons, measuring how many are present to get the total “accumulated dose.” They then measure the annual radiation dose rate of the surrounding soil—essentially asking, “how much uranium is this soil giving off each year?” Dividing total trapped dose by annual dose rate gives the estimated age of the tooth.
To measure the annual dose rate in the surrounding sediment, Li Hui’s team used Optically Stimulated Luminescence (OSL)—a method I don’t fully understand. Apparently, it requires that the soil hasn’t been exposed to light, otherwise the signal resets. But these teeth were excavated, stored, transported, and eventually handed over to Liu Wu—I find it hard to believe they were never exposed to light during all of that. So I’m skeptical about that part of the data.
Comparing the two, it seems Liu Wu’s result might actually be more reliable—though honestly, both dating methods seem prone to uncertainty in different ways.