本文欲討論近來在華北、華南、日本、俄羅斯遠東地區與西伯利亞東部等地區所發現，年代屬於更新世晚期的早期陶器的年代與功能，並對它們出現和散布的模式與緣由提出新的思考。早期陶器所出現的情境，除了出現少數的陶器之外，基本上皆與更新世晚期的遺址相似。在東亞地區，最顯著的就是沒有植物栽培的證據，這無疑使得過去認為陶器的出現與擴散經常與農業相伴的看法受到了挑戰。因之，我們需要對於狩獵採集社會中陶器的發明與散布給予慎重的思考。本文首先檢視出現於華北與華南舊石器脈絡中的早期陶器。華南仙人洞遺址的發掘及測定年代顯示，該遺址的陶器出現在校正年代距今大約20,000年前的地層中，正當末次冰期盛期的期間，比過去所知早期新石器時代定居村落在中國初次出現的年代早了一萬年。湖南玉蟾岩遺址出土了校正年代距今18,300年的陶器，同時出現了從鹿骨中擷取骨髓及油脂的證據，可能代表狩獵採集遊群年度性地光顧這個遺址。在華北，出現早期陶器的遺址，例如于家溝、轉年、東湖林、李家溝及南庄頭，年代相對較晚，大約是在新仙女木事件氣溫下降的時候，晚於華南早期陶器的年代約8,000年，也晚於日本及俄羅斯遠東約4,000年。華北的遺址呈現出使用品田石瓣及（或）磨石的多樣性適應，以及採集禾本類（如小米）、橡子和根莖類等野生植物的證據。這些遺址也許代表狩獵採集者在仙女木事件時期退居到比較適於居住的環境，減少遊居，成為半定居，並密集閒採居住地周邊的資源。日本早期陶器開始於大約年代距今16,800年，相當於繩紋草創期。而俄羅斯遠東是出現於新石器初期。繩紋草創期的遺址大約與晚舊石器時代末期的遺址同時，分布在從九州南部到北海道（大正3號遺址）。80處已知遺址，提供了有關於陶器分布類型與從最早的遺址（大平山元I遺址）到全新世期間關於各種氣候變遷適應的良好證據。關於陶器附著物的分子生物與穩定同位素分析，提供了關於日本早期陶器使用的重要資料：陶器被廣泛的使用在處理海洋與淡水動物。這些資料完全不見於東亞其它地方。如同晚舊石器時代末期遺址，繩紋草創時期也可能擁有細石瓣、磨刃石斧、箭頭、雙面矛頭。素面陶器與神子柴形石器同時出現在陶器製作早期階段（大平山元I遺址、北原遺址與前回耕地遺址距今約16,500-13,500）。有裝飾的陶器（階段二）約出現在的，700年前開始的博令阿勒羅德（Bølling-Allerød）暖化階段，並且迅速地擴散到整個群島，在這個階段產生了生計與移動性重大的改變。階段一的陶器可以發生在密集資訊流動的社會網絡中，階段二陶器的變異性則發生在社會網絡變得與地點更加嵌合的情況。俄羅斯新石器早期陶器的遺址，如Khummy, Gasya，以及在Amur河下游河谷的Goncharka 1，是舊石器傳統與全新世典型新石器的轉換期，陶器與研磨石器逐漸在舊石器晚期的工具組中出現。如同在中國與日本，早期陶器的製作規模甚小，僅有數量極少的陶片在少數遺址中被發現。西伯利亞東部的早期陶器首次在Ust-Karenga 12遺址出現，大約距今13,000年前，陶器可能越過西伯利亞向西邊擴散，狩獵採集群體可能即是將陶器帶入歐洲的人。在整個東亞，早期陶器出現的數量很少，同時只出現在少數幾個遺址中，持續低頻度地使用於末次冰期盛期到全新世早期之間。我們有必要去理解何以會產生這種現象。早期陶器或許是為了某種用途而被發明與使用，例如用於飲宴中來達成多種社會政治的目地。然而，陶器雖然同時提供了實用性與經濟性的價值、所具有之耐久與小規模的使用性，以及大規模的擴散性，但是不能僅僅解釋為一種對生計及持續增加能量產出的適應。關於東亞陶器是否為單一起源或多個起源的問題仍舊存在。中國南部的遺址明顯是更為早期的，然而與之同期的日本與俄羅斯，並不能排除單一起源與擴散的可能性。因為在更新世晚期具有相同放射性碳素年代的遺址，在曆年上實際是落人數世紀之久的年代史度內。我們仍有必要對狩獵採集社會移動性之規模與形態，以及東亞更新世晚期陶器製作訊息擴散的交換網絡範圍有更多理解。

This paper discusses recent data from North and South China, Japan, and the Russian Far East and eastern Siberia on the dating and function of early pottery during the Late Pleistocene period and shows how reconsiderations are needed for the patterns and reasons for its emergence and spread. Early pottery typically appears in contexts that, except for containing small amounts of pottery, are otherwise similar to Late Paleolithic sites. There is also no evidence of plant cultivation, so clearly in eastern Asia, the old view of pottery's emergence or dispersal as only coming within agricultural societies is no longer viable. Greater consideration needs to be given to the invention and spread of pottery in hunter-gatherer societies. This paper first reviews recent finds of early pottery sites in South China and North China that now clearly show that the pottery first appears in otherwise Late Paleolithic contexts. Excavations and re-dating at Xianrendong Cave (Jiangxi) in South China show that pottery appears there in securely dated stratigraphic contexts dating to ca. 20,000 cal BP, during the Last Glacial Maximum, some ten millennia before sedentary, Early Neolithic villages first appear in China. Yuchanyan Cave (Hunan) has pottery dating to 18,300 cal BP, evidence for processing deer bones to extract marrow and grease, and perhaps evidence of seasonal visits to the site in annual rounds by mobile hunter-gatherer groups. Sites with early pottery in North China, such as Yujiagou, Zhuannian, Donghulin, Lijiagou, and Nanzhuangtou, appear relatively late, from the climatic downturn of the Younger Dryas, some eight millennia after sites in South China and four millennia after early pottery in Japan and the Russian Far East. North China sites variously feature such adaptations as microblades and/or grinding stones, as well as evidence for the exploitation of wild grasses (including millets), acorns, and tubers. These sites might represent hunter-gatherers retreating to more favorable habitats during the Younger Dryas and indicate reduced mobility and semisedentary practices with more intensified exploitation of closer resources. Early pottery finds beginning from ca. 16,800 cal BP in Japan (Incipient Jōmon) and the Russian Far East (＂Initial Neolithic ＂) are also reviewed. Incipient Jōmon sites occur contemporaneously with Final Upper Paleolithic sites, and are found from southern Kyūshū to Hokkaidō (Taishō 3 site). With over 80 known sites, Japan has a better evidence for changes in pottery distribution patterns and diverse adaptations to climatic changes from the time period of the earliest site, Ōdai Yamamoto I, to the Holocene. Molecular and stable isotope analyses of pottery adhesions provide valuable data on the use of early pottery in Japan lacking for all other regions: these indicate the widespread use of pottery for processing marine and freshwater animals. Like Final Upper Paleolithic sites, Incipient Jōmon sites also may have microblades, edge polished stone axes, arrowheads, and bifacial spear points. Undecorated pottery with Mikoshiba-tyPe lithics are found in the initial phase of pottery making (Ōdai Yamamoto I, Kitahara, and Maeda Kōji sites, dating ca. 16,500-13,500 BP). Decorated pottery (Phase 2) begins ca. 15, 700 cal BP during the Bølling-Allerød warming period and rapidly disperses across the archipelago at a time when there may have been significant changes in subsistence and mobility patterns. Phase 1 pottery might occur during a time of intensive information flow and fluidity of social networks, while diversification of pottery in Phase 2 occurs when social networks were becoming more embedded in place. Russian ＂Initial Neolithic＂ early pottery sites, such as Khummy, Gasya, and Goncharka 1 in the Lower Amur River basin. are transitional between Paleolithic traditions and typical Neolithic sites of the Holocene, with pottery and ground stone tools gradually appearing amongst Upper Paleolithic toolkits. As in China and Japan, early pottery production is at a very low scale, with only limited quantities of sherds being found at a few sites. Eastern Siberia early pottery is first present at the Ust'-Karenga 12 site ca. 13,000 cal BP. Pottery may have dispersed westerly across Siberia as forested areas expanded, perhaps resulting in the introduction of pottery into Europe by hunter-gatherer groups. Across East Asia, early pottery appears only in small amounts and at a few sites, and it persists in this episodic, low scale usage from the Last Glacial Maximum until the Early Holocene. We still need to better understand why this is the case. Early pottery may have been invented and used for special purposes, such as in feasting that was carried out to achieve various socio-political goals. While pottery also offered utilitarian or economic value, its long-lasting, low-scale use, but widespread dispersal despite this, cannot be fully accounted for only in terms of it being an adarptation tied to subsistence and increasing energy yields. Questions still remain over whether pottery was the result of a single or multiple inventions in East Asia. South China sites are clearly earlier, and the contemporaneity of Japan and Russia does not rule out singular invention and spread, as sites of the same radiocarbon date in the Late Pleistocene actually fall within a real calendrical range on a centuries-long scale. We need to better understand the scale and patterns of hunter-gatherer mobility and the extent of information exchange networks through which knowledge of pottery making could have spread widely in Late Pleistocene East Asia.