See discussions, stats, and author profiles for this publication at: [Link]

net/publication/257779315

Northeast Asian Linguistic Ecology and the Advent of Rice Agriculture in

Korea and Japan

Article  in  Rice · December 2012

DOI: 10.1007/s12284-011-9080-0

CITATIONS READS

20 526

1 author:

John Whitman
Cornell University
70 PUBLICATIONS   871 CITATIONS   

SEE PROFILE

All content following this page was uploaded by John Whitman on 05 January 2016.

The user has requested enhancement of the downloaded file.

Rice (2011) 4:149–158
DOI 10.1007/s12284-011-9080-0

Northeast Asian Linguistic Ecology and the Advent

of Rice Agriculture in Korea and Japan
John Whitman

Received: 14 December 2011 / Accepted: 17 December 2011 / Published online: 14 January 2012
# Springer Science+Business Media, LLC 2012

Abstract The languages of Northeast Asia show evidence Liaoning . Toponyms . Founders effect . Bottleneck effect .
of dispersal from south to north, consistent with the hypoth- Language families . Chronological depth of protolanguage
esis that agriculture spread north and east from the vicinity
of Liaoning, beginning with the millets approximately 5500
BP. Wet rice agriculture in Korea and Japan results from a Introduction
later spread, also beginning in Shandong, crossing via the
Liaodong peninsula and reaching the Korean peninsula This paper examines the relationship between the linguistic
around 1500 BCE. This dispersal is associated with the ecology of Northeast Asia and the spread of rice agriculture. It
Mumun archaeological culture after 1500 BCE in the Kore- focuses on the subpart of the region where wet rice agriculture
an peninsula and the Yayoi culture after 950 BCE in the became established three and a half millenia ago, Korea and
Japanese archipelago. From a linguistic standpoint, it is Japan. “Linguistic ecology” refers to the interactions between
associated with the entry of the Japonic language family, a language and its environment, including the languages spo-
first into the Korean peninsula, subsequently into the Japa- ken around it (Haugen 1972: 325). I argue that the historical
nese archipelago. The arrival of Koreanic is associated with distribution of the Japonic and Koreanic language families is
the advent of the Korean-style bronze dagger culture and a associated with two events in the archaeological record. The
temporary hiatus in wet rice agriculture sites around 300 distribution of Japonic, first on the Korean peninsula and later
BCE. Both Koreanic and Japonic are relatively shallow in the Japanese archipelago, results from the relatively rapid
language families, with Koreanic the shallower of the two, spread of wet rice agriculture down the Korean peninsula to its
consistent with the chronology above. The gap between the southern tip. Wet rice cultivation reaches northern Kyūshū by
earliest linguistically motivated dates for these language 950 BCE, and the western end of the Inland Sea by 600 BCE.
families and the archaeological events is the result of a The distribution of Koreanic, and ultimately, the disappear-
linguistic founders effect, providing further evidence for ance of Japonic from the Korean peninsula, results from the
demic diffusion as a source for their distribution. arrival of a population which is associated with a temporary
hiatus in wet rice agriculture in the southern Korean peninsula
Keywords Northeast Asia . Altaic . Tungusic . Japonic . around 300 BCE.
Japanese archipelago . Koreanic . Korean peninsula . Yayoi . The methodology of the paper is as follows. I first report
Mumun . wet rice agriculture . Shandong . Liaodong . what I take to be important points of consensus between
archaeologists in Korea and Japan regarding the beginning
of wet rice agriculture in the region and describe briefly what
we know about the linguistic ecology of the Korean peninsula
J. Whitman (*) from the earliest Chinese sources. I then consider how three
National Institute of Japanese Language and Linguistics types of linguistic data can be reconciled with this informa-
(NINJAL)/Cornell University,
tion: historical data on the location of speakers of the lan-
10-2 Midori-cho,
Tachikawa City, Tokyo 190-8561, Japan guages, the dates of the protofamilies involved, and the nature
e-mail: jbw@[Link] of the vocabulary related to agriculture.
150 Rice (2011) 4:149–158

Fig. 1 Distribution of language

families in Northeast Asia ca.
1,800.

Background further north in the region. Among the language families in

Table 1, the Jurchens, the oldest clearly identifiable speakers
As a linguistic area, Northeast Asia is made up of ten language of a Tungusic language, grew millet along with other crops
families: Ainuic, Amuric (Nivkh/Gilyak), Japonic, Kamchu- (Franke 1990: 416). While traditional Chinese and Western
kotic, Koreanic, Mongolic, Tungusic, Turkic, Yeniseic, and historiography present “Altaic” (Tungusic, Mongolic, Turkic)-
Yukaghiric (Janhunen 1998: 196–197). The distribution of speaking populations as pastoralists or hunter/gathers originat-
these language families is shown in Fig. 1. The third, fifth, ing from the north of contemporary Han Chinese-speaking
sixth, seventh, and eighth of these families are sometimes regions, it is quite possible that the historical distribution of
related by various versions of the Altaic hypothesis, which these language families results from the expansion of earlier
asserts a genetic relation between Mongolic, Tungusic, and agriculturalists from the south. Janhunen (1998: 202) points
Turkic (Ramstedt 1952) or between these three families and out that Tungusic as well as Amuric and Kamchukotic gives
Koreanic (Poppe 1960) and Japonic (Starostin et al. 2003; evidence for having expanded relatively recently into the
Robbeets 2005). However, there is still no scholarly consensus
about the Altaic hypothesis; for critiques, see Georg et al.
(1999) and Vovin (2005), among many others.1 Table 1 Language family locations and means of food production in
Food production in Northeast Asia ranges from wet rice Northeast Asia
cultivation in Korea and Japan to steppe pastoralism in
Language Region Food production
Mongolic and Turkic language areas, to reindeer pastoralism
among speakers of Kamchukotic, Yeniseic, Yukaghiric, and Ainuic Hokkaido, Sakahlin Hunting/fishing
Tungusic, to hunting and fishing among these groups and Amuric (Nivkh/ Amur estuary, Sakhalin Hunting/fishing
Ainuic. Table 1 gives a basic correlation of language fami- Gilyak)
lies, location, and primary modes of food production. Japonic Japanese archipelago Rice farming
While Table 1 associates only Japonic and Koreanic with Kamchukotic Kamchatka Hunting/fishing,
rice agriculture, historical, linguistic, and archaeological reindeer pastoralism
evidence indicates that dry field agriculture extended much Koreanic Korean peninsula Rice farming
Mongolic Mongolia, China, Russia Pastoralism
Tungusic China, Russia Hunting/fishing,
1 reindeer pastoralism
This paper is not the place to offer yet another reassessment of the Altaic
hypothesis, but to this linguist’s eye the most persuasive argument for the Turkic Siberia, Central Asia Pastoralism
hypothesis are the formal and functional similarities between verbal Yeniseic Yenisei basin Hunting/fishing
nominalizing morphology in Mongolic, Tungusic, and Turkic (Ramstedt Yukaghiric Sakha Republic Hunting/reindeer
1952), extended to Koreanic in suggestions by Ramstedt and Poppe and pastoralism
to Japonic by Vovin (2001) and Robbeets (2007).
Rice (2011) 4:149–158 151

Fig. 2 The Shandong/

Liaodong dispersion
hypothesis, based on Miyamoto
(2009: 26).

northern part of its range. We know that Panicum (broomcorn) millennium BCE. In the Korean peninsula, it is associated with
and Setaria (foxtail) millet are associated with the Zaisanovka the Mumun (plain, patternless) ceramic culture and irrigated
culture in the Russian Primorye (Maritime) region as early as paddy cultivation. This third dispersion reaches the Japanese
4600 BP (Kuzmin 2008: 6), well to the north of the historical archipelago, where wet rice cultivation appears in northern
center of Tungusic. It is thus possible that Northeast Asian Kyūshū around 950 BCE (800 BCE according to Miyamoto
language groups not associated with agriculture in historical 2009: 28).2
times were agriculturalists at earlier periods. Such a view Miyamoto’s scenario for the third dispersion is consistent
would be consistent with the Shandong/Liaodong dispersion with the scenario for the introduction of wet rice cultivation
hypothesis presented in the next section. into the Korean peninsula provided by Ahn (2010). Ahn
discusses and dismisses claims for very early rice cultivation
in the Korean peninsula and finds inconclusive evidence for
Archaeological and historical considerations rice cultivation during the Chŭlmun (comb pattern) period (ca.
6000–1300 BCE). Previous researchers (e.g., Crawford and
The Shandong/Liaodong dispersion hypothesis Lee 2003) concur on the evidence for Chŭlmun cultivation of
Setaria and Panicum diffused from the Liaoning region, con-
Miyamoto (2009) presents a scenario where agriculture spread sistent with the Shandong/Liaodong dispersion hypothesis.
from (possibly distinct) locations in northeast China to the east Ahn argues against earlier views that wet rice cultivation was
and north (Fig. 2). Miyamoto distinguishes three major spreads introduced into the Korean peninsula from the south, and
of agriculture from this region. The first, associated with Setaria
and Panicum, spreads from the Liaoning region to the north-
western part of the Korean peninsula and thence to its eastern 2
In this article, I follow the revised dates for the beginning of the
and southern coasts. Almost simultaneously, a spread of these Yayoi period in Kyūshū established over the past decade by a team at
cultivars took place from northeast China to the south of the National Museum of Japanese History (Nishimoto ed. 2006) based
Primoriye and thence to its coastal plain (Miyamoto 2009: on accelerator mass spectrometry (AMS) radiocarbon dating. These
25–6). Miyamoto dates this first spread to the first half of the dates cluster around 3050 BP, that is, 950 BCE (Harunari and Imamura
2004). There has been some resistance from the community of Japa-
fourth millennium BCE. The second dispersion adds dry field
nese archaeologists to the revised dates, due the interpretation of
rice cultivation to the millets; it spreads from Shandong through external evidence for the beginning of Yayoi, based on the dating of
the Liaodong peninsula to the south of the Korean peninsula in bronze daggers originating in the Liaoning region and found through-
the second half of the third millennium BCE. The third out the Korean peninsula and Kyūshū. For an attempt to reconcile the
AMS dates and the external evidence, see Harunari (2006). In any case,
dispersion includes wet rice cultivation and associated tool
there is a consensus among Japanese archaeologists that the beginning
complexes. It takes the same route from Shandong to Liaodong of Yayoi should be revised to a date earlier by at least three centuries
to the Korean peninsula around the middle of the second than the traditional 500 BCE.
152 Rice (2011) 4:149–158

concludes in favor of the scenario where it enters the peninsula culture: “Agricultural settlements disappeared from the ar-
from Liaodong in the early Mumun period around or after chaeological record from the third century B.C. when the Late
1500 BCE. Thereafter, evidence for agricultural settlements Mumun culture,3 with a nomadic lifestyle, spread from the
engaged in rice farming is found throughout the peninsula Liaoning region of northeast China” (Ahn 2010: 91). It is not
with the exception of the northeast. The relative importance clear to what extent the Korean-style bronze dagger culture
of wet rice cultivation shows regional variation throughout the should be characterized as nomadic, but it is clear that it was
Mumun period: wet rice is dominant in the central and south- not, in its initial appearance, associated with wet rice cultiva-
west regions, while millet and other dry field crops are rela- tion. The advent of this culture can account for the temporary
tively more important in the southeast (Ahn 2010: 93). disappearance of wet rice farming settlements between the
third century BCE–first century CE.
Hiatus in wet rice agriculture on the Korean peninsula
Early Chinese sources on the Korean peninsula
According to Ahn (2003: 81–81; 2010: 91), rice farming and the Japanese archipelago
settlements disappear from the archaeological record during
the third century BCE, reappearing again in the first century The earliest substantial Chinese sources on populations in
CE. Ahn suggests that the specialized nature of intensive rice the central/south Korean peninsula, the so-called Dongyi
paddy farming might have been especially vulnerable to en- “Eastern barbarian” (Book of Wei) in the Wei shu section
vironmental fluctuations, but he acknowledges that there is no of the Sanguo zhi (late third century CE) and the Hou Han
evidence for major climatic changes that might have triggered shu (fifth century CE), describe three groupings of peoples,
a hiatus in rice farming during this period (Ahn 2010: 97). The the so-called Samhan “Three Han” (三韓 Sanhan): Mahan (馬
hiatus in wet rice sites coincides with the emergence of the 韓) in the west central region, Chinhan (辰韓 Chenhan) in the
Korean-style bronze dagger culture in the third century BCE. southeast, and Pyŏnhan (弁韓 Bianhan) in the south.4 Tradi-
As described by Ahn (2003), this culture spreads from the tional Korean historiography indentifies these groupings as
Kŭm river basin on the west coast of Korea. In addition to its the antecedents of the historical polities Paekche (百済), Silla
characteristic bronze daggers, this culture is characterized by (新羅), and Kaya (加耶), respectively. Inscriptional evidence
coarse patterned bronze mirrors with multiple attachment indicates that the term Han 韓 had ethnonymic significance not
loops (多鈕粗文鏡), shield- and hilt-shaped bronze imple- only for the Chinese authors of the Wei shu and Hou Han shu
ments, black long-necked earthenware, and clay-rimmed but for local Sinoxenic peoples as well. Thus, the inventory of
ceramics. Ahn emphasizes that the ceramic style, pattern of gravekeeper villages on the 414 CE stele memorializing
settlement, and burial styles of this culture indicate a break Koguryŏ king Kwangaet’o (廣開土) lists both Koguryŏ 高句
with the previous Mumun culture, although the two cultures 麗 and Han 韓 villages, without distinguishing the latter as
continue to coexist for some time. The Korean-style bronze Mahan, Chinhan, etc. Nevertheless, the Wei shu and Hou Han
dagger itself descends from the Liaoning bronze dagger pro- shu present a picture of some ethnic and linguistic diversity in
totype. The Liaoning prototype is found in the Korean penin- the Samhan region in the third century.
sula from 1300 BCE on and in Kyūshū from 800 BCE, but The Wei shu describes the language of Chinhan as “not the
Ahn (2003) dismisses the possibility that the distinctive same as that of Mahan” (其言語不與馬韓同).5 While the
Korean-style bronze dagger developed from this prototype
within the Korean peninusla. Instead, he traces the origins of
the Korean bronze dagger culture to the central Liaoning 4
Pyŏnhan is referred to in the Wei shu as Pyŏnjin 弁辰, usually inter-
region, where similar pottery, burial, and dwelling styles are preted as an amalgamation of the names for Pyŏnhan and Chinhan. The
found in association with the Liaoning bronze dagger proto- name further emphasizes the heterogeneous nature of this grouping.
5
type. According to Ahn, this Liaoning culture split into two The passage goes on to compare the language of Chinhan with that of
the Chinese Qin 秦 state. It cites what it identifies as the Qin words for
branches, one of which remained in Liaodong and the Chang- “country” (邦), “bow” (弧), “bandit” (冦), and “drinking game” (行觴),
baishan region. The second branch brought the Korean-style and claims that the Chinhan words are the same. It also states
bronze dagger culture to the Kŭm river area. Ahn writes of this that the Chinhan people identify themselves with the inhabitants of the
Chinese Lelang commandery in northwest Korea: “They say that the
people of Lelang were originally the remnants of their people” (謂樂浪人
3
Ahn (2010) uses the term “Late Mumun culture” to refer to the 本其殘餘人). While scholars have generally dismissed this apparent
culture designated as the Korean-style bronze dagger culture attempt to claim a Chinese ethnicity, it may in fact reflect the relatively
(Hanguk-sik tongkŏm munhwa 韓国式銅剣文化). recent arrival of the Chinhan population from the northwest.
Rice (2011) 4:149–158 153

people of Mahan are described as not knowing how to use Chinhan polysyllabic settlement name.7 The second is identi-
cows and horses as means of transport (不知乗牛馬), those of cal to the transcription given in the Wei zhi for the first two
Chinhan use them. The latter are described as planting the five syllables for the contemporary *Jamaʔdə “Yamato” grouping
crops with the addition of rice (宜種五穀及稻), suggesting in the Japanese archipelago. Scholars generally intepret these as
that the latter may have been something of an add-on. involving a morpheme cognate with proto-Japonic *jama
The Wei shu describes the Pyŏnhan and Chinhan popula- “mountain” (Bentley 2008: 14). A virtually identical spelling
tions as “living intermingled together” (弁辰與辰韓雑居). It *jama 邪麻 occurs as an independent word in the Wa toponyms
describes their clothing and dwellings as the same, and their on the Japanese archipelago given in the Wei shu.
languages and customs as similar (言語法俗相似). The In light of the discussion in the “Hiatus in wet rice
Hou Han shu begins with the same phrase about intermingled agriculture on the Korean peninsula” section, a simplistic
living, then states “enclosed towns and clothing are all the interpretation of the Wei shu and Hou Han shu descriptions
same” (城郭衣服皆同) but “languages and customs have might be that the three Han groupings correspond to three
differences” (言語風俗有異). This section of the Hou Han distinct but related ethnicities. In fact, the texts indicate a
shu, as the later of the two texts, often cites the Wei shu, but the more complex (and plausible) interrelationship between lan-
Hou Han shu drew on other sources as well. The different guage, ethnicity, and protopolitical grouping. Mahan, the
descriptions of linguistic and cultural distinctions may larger, better established grouping, occupies the area where
reflect differing responses to a situation of ethnic com- the Korean-style bronze dagger culture emerged some five
plexity. Similar complexity is suggested by the descrip- centuries earlier. Chinhan represents a population more recent-
tions of physical type. The Wei shu states that “Chinhan ly arrived from the northwest, as indicated by its oral traditions
men and women are close to Wa (男女近倭),” the and its mastery of animal husbandry. The Chinhan population
ethyonym for the contemporary inhabitants of the Japa- lives intermixed with Pyŏnhan; the Chinese reporters struggle
nese archipelago, and like the Wa tattoo their bodies. to decribe the resultant demographic complexity. Their lan-
The Hou Han shu identifies this as a feature of Pyŏn- guages may be similar, or different; some resemble the Wa,
han, stating that “their country is close to Wa, therefore some tattoo their bodies. While Wa-like toponyms are more
they frequently have tattoos.” frequent in the Pyŏnhan grouping, one such toponym is iden-
Linguistic differences are confirmed by the Wei shu top- tified with Chinhan. This is exactly the kind of complexity we
onyms for the Samhan. The Wei shu gives phonogrammatic might expect to be associated with the situation described by
spellings for 54 Mahan settlement names. As pointed out by Ahn, where a population associated with Mumun wet rice
Toh (2008: 234–5), most are disyllabic: 34 are transcribed growing culture lives alongside more recently arrived mem-
with two syllables, 10 more with two syllables and a bers of the Korean-style bronze dagger culture.
suffix, one of which is identifiable as *-pieliai 卑離,
usually related to the Paekche word puri <夫里> “town,”
itself typically compared to Late Middle Korean -βɨr “town.”6 Linguistics
These suffixes do not occur in the Chinhan and Pyŏnhan
settlement names. The Wei shu lists the latter together in no Toponymic evidence on language locations
particular order. It prefixes the Pyŏnhan names with Pyŏnjin
弁辰. However, the Chinhan and Pyŏnhan toponyms Clear evidence for the presence of a Japonic language or
also appear to draw on two distinct linguistic traditions. languages on the Korean peninsula is provided by the so-
All but one of the 12 Chinhan names are disyllabic called Koguryŏ placenames recorded in the gazetteer chapters
(like the Mahan names). Five of the 12 Pyŏnhan names 35 and 37 of the twelfth century Korean history Samguk sagi
have three or more syllables, and three of these appear (三國史記 Record of the Three Kingdoms). The crucial data
to involve a suffix, *-mietoŋ 彌凍and *-jamaʔ 邪馬. have been known since Shinmura (1916). It consists of entries
The first of these suffixes also occurs in the only where a Silla toponym is paired with the original Koguryŏ

7
Interpreted as a place name suffix, *-mietoŋ 彌凍is comparable to
6
I follow Bentley (2008) in using Schuessler’s (2007) reconstruction Late Middle Korean mith “base, bottom” and proto-Japonic *mətə id.,
of Later Han Chinese to interpret Wei zhi transcriptions. Sinographic asserted by Martin (1966) to be cognate. The latter is a common second
transcriptions devised in Korean receive their traditional Korean element in toponyms. The comparison would have to assume that the
interpretation. first syllable vowel assimilated to the second in pJ.
154 Rice (2011) 4:149–158

name for a locality that came under Silla control after the name that is phonogrammatic. A subset of the phonogram-
Koguryŏ defeat in 668. Some of the Koguryŏ toponyms matically transcribed names appears to be related to the mean-
renamed by Silla are phonogrammatic, or have an alternate ing of the later Koguryŏ or Silla names.8 For example,

1.

(1a) gives the Silla toponym, obviously shortened from the relatable to Japonic and words related to Koreanic. Koguryŏ
Koguryŏ name. (1b) gives the Koguryŏ name cited in (1a), plus used phonograms to transcribe indigenous names from lan-
an alternate phonogrammatic name. The phonogrammatic guages other than their own. They also devised standard
name is a good fit with “seven layer,” if the former is read as Chinese binomic names for some localities that came under
something resembling nan’jɨn (cf. proto-Japonic nana “seven”) their control; for such localities, the two names coexisted.
and pjet (cf. pJ pe “layer,” Late Middle Korean pʌr id.)9 From the standpoint of this paper, the important takeaway
There are two broad interpretations of the Koguryŏ pho- lesson from the Koguryŏ toponymic data is that a language
nogrammatic material. One takes them to represent the cognate to Japonic was spoken on the Korean peninsula.
Koguryŏ language. This interpretation is adopted in earlier This is a point of consensus for all major scholars who have
Japanese scholarship, by the Korean scholar Lee Ki-moon worked on this material. The range of the Koguryŏ top-
(see Lee and Ramsey 2011) and by Christopher Beckwith onymns is confined to the region of historical Koguryŏ
(see Beckwith 2007). It has been influential among anthro- control, so they provide no information about the southern
pologists, e.g., Hudson (1999). The second interpretation, tip of the peninsula, but the northern range of phonogram-
associated with the Japanese scholar Kōno Rokurō and the matic toponyms with widely accepted Japonic interpreta-
Korean linguist Kim Bang-han (Kim 1983), claims that the tions extends as far as modern North Hwanghae province,
Koguryŏ phonogrammatic material transcribes the topo- south of the later Koguryŏ capital at P’yŏngyang.
nyms of linguistically distinct, non-Koguryŏ peoples.
Scholars adopting the first view have arrived at diametri-
cally opposed conclusions about the nature of the Koguryŏ Chronological depth of language families
language. Thus, Lee and Ramsey (2011) emphasize the lexical
material in the Koguryŏ language relatable to Korean,10 while Both Japonic and Koreanic are relatively shallow language
Beckwith (2007) considers Koguryŏ to be a continental rela- families. Comparative phonological evidence shows proto-
tive of Japanese. In contrast, the second view explains why the Japonic to be somewhat older that the oldest extensive
Koguryŏ phonogrammatic material transcribes words textual attestations of Western Old Japanese in the eighth
century. Proto-Ryūkyūan maintains the distinction between
8
For a recent discussion of the format and interpretation of the Samguk sagi
proto-Japonic *e and *i and *o and *u in wider range of
toponyms, see Lee and Ramsey (2011). For a detailed recent discussion of environments than does Western Old Japanese, indicating
the toponyms with Japonic interpretations, see Beckwith (2007). that the ancestor of pR diverged from a parent older than
9
For the purposes of this paper, I interpret the Koguryŏ phonograms WOJ (Hattori 1977–1979). Phonological information like
following the Middle Chinese system of Baxter and Sagart (n.d.).
this provides a ceiling but not a floor for the date of the
Korean scholars typically interpret them by their Sino-Korean values,
but Beckwith (2007) is surely correct to argue that Sino-Koguryŏ protofamily; however, a radically earlier date would lead us
represents a distinct sinoxenic tradition since Sino-Korean is generally to expect a greater degree of phonological divergence. Hat-
dated to the late Tang period. tori’s (1953) glottochronological study estimates a date of
10
Lee and Ramsey acknowledge that the four numerals attested in the
Koguryŏ phonogrammatic tradition “all look remarkably like Japanese”
500 CE for the divergence of the ancestors of Early Middle
(2011: 43). They then state “At the same time, however, the vocabulary Japanese and Shuri Ryūkyūan. Hattori arrives at this date by
found in the Koguryŏ place names includes even more elements that adjusting the logarithmic decay function proposed by Swadesh
relate solidly to Middle Korean and thus to the mainstream development to fit the facts of several known cases of divergence.
of the Korean language” (ibid). They give no statistics to support this
assessment. They also do not explain how a language whose lexicon
A standard criticism of glottochronology is that it assumes a
preponderantly relates “solidly to Korean” should come to have all four of constant rate of vocabulary replacement across languages. In a
its attested basic numerals remarkably similar to Japanese. recent paper, Lee and Hasegawa (2011) attempt to overcome
Rice (2011) 4:149–158 155

this and other defects of glottochronological approaches using a 900-year lag between the archaeological event and the lin-
Bayesian phylogenetic analysis based on lexical data from 59 guistic evidence for dispersal.
Japonic varieties. This model assumes a single rate of vocabu- The remarkable non-diversity of Japonic and Koreanic can
lary substitution across varieties, but the rate is calibrated on the be explained by two factors. The first is a founder’s effect, the
basis of known historical dates (in this case, those for Western phenomenon by which genetic diversity is reduced when a
Old Japanese and Early Middle Japanese). The phylogeny small population settles a new area. The claim that a serial
selected by Hasegawa and Lee is problematic in its shallower founder’s effect is discernible in linguistic variation has been
branches, which represent all non-Ryūkyūan branches as made by Atkinson (2011), among others. In the case of Japonic,
descended from EMJ, but it is not clear that this affects their we might expect founder’s effects to have occurred as a result
overall results. Lee and Hasegawa estimate a date of 2182 BP of the movement of relatively small populations from the
for the ancestor of proto-Japonic. This result is important be- Korean peninsula to Kyūshū, and again from Kyūshū to the
cause it disconfirms the possibility of Kofun period (third to rest of the archipelago. Crudely put, the effect can be concep-
sixth century CE) date for pJ, something not completely dis- tualized as a local reduction in linguistic diversity compared to
allowed by Hattori’s results. However, a Kofun period date for the home population. The same effect would be anticipated in
pJ would also be inconsistent with the toponymic evidence for the establishment of Koreanic in the south central peninsula,
Japonic on the Korean peninsula, unless the toponyms some- and again as it expanded throughout the peninsula.
how resulted from a later historical movement of Japonic The second factor is archaeohistorical. Both the Yayoi
speakers to the continent. expansion in Japan and the spread of the Korean bronze
Phonological evidence indicates that proto-Koreanic is curved dagger culture were subject to bottleneck effects, to
even shallower than proto-Japanese. The evidence is similar: borrow another term from evolutionary science. Kobayashi
data from the Cheju variety show a broader distribution of the (2007) accounts for the relatively slow spread of Yayoi culture
back central unrounded vowel/$/than is found in fifteenth to the east in terms of the “walls” (壁 kabe) put up by the
century Late Middle Korean texts. Once again, this gives us progressively more robust Jōmon cultures to the east in the
a ceiling for divergence somewhat earlier than the fifteenth archipelago. In the case of Korea, the Chinese commanderies
century; once again, if the protolanguage was radically older, in the north of the peninsula imposed a bottleneck until their
we might expect greater phonological divergence. demise in the early fourth century CE. Release of each bottle-
Neither the phonological evidence nor the statistical evi- neck results in a new dispersal and founders effect. These
dence (in the case of Japanese) is consistent with a date of effects leave phonological traces; thus, the categories of lex-
protolanguage divergence older than the dates for the begin- ical accent are less complex in eastern Japonic varieties, while
ning of wet rice agriculture, as pointed out by Hudson (1999) Koreanic varieties in regions to the north and east, as well as
and Lee and Hasegawa (2011). This fact alone does not rule Cheju, lack lexical pitch accent altogether.
out the possibility that proto-Japanese descends from a pre- Note that the founders effect scenario presupposes demic
Yayoi Jōmon language, or proto-Korean from a pre-Mumun diffusion. Thus, the relative nondiversity of Koreanic and
Chulmun language. In either case, it is a prima facie possibil- Japonic provides further support for the view that the speak-
ity that such a language, indigenous to the region prior to the ers of the protolanguages arrived from elsewhere.
arrival of wet rice agriculture, expanded and replaced previ-
ously existing indigenous languages as a result of the demo-
Rice and related agricultural vocabularies
graphic expansion associated with the new agricultural
technology. In the case of Japonic, however, once again, such
Vovin (1998) discusses possible external cognates for the ten
a scenario would have a difficult time explaining the topo-
Japanese terms related to rice agriculture in (2). The proto-
nymic evidence for Japonic on the Korean peninsula.
Japonic reconstructions I cite are slightly different from Vovin’s.
The scenarios whereby Japonic arrived in the archipelago
and dispersed as a result of the Yayoi expansion, and Kore- 2. (a) *jinaC 2.4 “riceplant”11
anic arrived in the peninsula and dispersed as a result of the (b) *mə/omi 2.1 unhulled rice
advent of the Korean bronze curved dagger culture, are
11
consistent with the farming/language dispersal model Vovin, following a proposal of Unger (1977), reconstructs pJ *zinaCi
for “riceplant,” on the basis of attestations such as arasine “unhulled
(Bellwood and Renfrew 2002). The dates of these two events,
rice” < ara “rough” + (s)ine, misine “riceplant,” mi- honorific + (s)ine.
950 BCE and 300 BCE, respectively, are also consistent with Unger’s hypothesis was that *z was lost in initial position but retained
the gap between the chronological ceilings for dispersal of the medially as WOJ /s/. Because the evidence for a voiced obstruent series
two families, before 700 CE for Japonic and before 1,450 CE in proto-Japanese is weak, I have not followed Unger and Vovin in
reconstructing *z, but instead exploited the independently motivated
for Koreanic. In both instances, we know that the actual date
glide *j. The reconstruction posits glide strengthening in medial position,
of dispersal must be earlier, but we do not know how much. which may seem counterintuitive, but in fact, strengthening is limited to
Even Lee and Hasegawa’s date, first century BCE, leaves a initial position after a compound boundary.
156 Rice (2011) 4:149–158

(c) *jənaC 2.1 “hulled rice” terms for “unhulled rice” and “cooked rice.” Similarly, (2f)
(d) *kəmə/aC 2.3 “(hulled) rice” and (2i) are not specialized rice-related terms. The lack of
(e) *ipi 2.3 “cooked rice” specialized vocabulary specifically dedicated to rice is more
(f) *po 1.3a “ear of grain” visible in Korean, where the terms in the semantic role of
(g) *ta 1.3a “ricefield” (2c-d) psʌr H < *pʌsʌr and (2e) pap L both designate hulled
(h) *nuka ?2.3 “rice bran” and cooked grains, respectively, of any type. The Korean
(i) *ko “flour, powder” (and Altaic) cognates that Vovin suggests for (2e), (2g), (2i),
(j) *nəri “starch, rice glue” and (2j) are all unspecialized: they mean “eat,” “field,
plain,” “flour, powder,” and “malt.” Corresponding to (2),
Vovin suggests cognates for four of these, (2e), (2g), (2i), the only semantic category with a Korean term specialized for
and (2j) from Koreanic, with cognates for (2e) in Tungusic and rice is (2a) Late Middle Korean pjə H “riceplant.”
Turkic, for (2g) in Mongolic and Turkic, and for (2i) in Tun- These facts are consistent with the Shandong/Liaodong
gusic as well. He finds no external etymologies for (2b), (2c), dispersion hypothesis outlined in “The Shandong/Liaodong
and (2h). Vovin specifically rejects Austronesian cognates pro- dispersion hypothesis” section. If the language families
posed in earlier research, but he suggests Austroasiatic cog- commonly grouped together as Altaic are related to Japonic,
nates for (2a), (2d), and (2f). Sagart (this issue) has proposed an they presumably dispersed from Shandong prior to Japonic
alternative Sino-Tibetan-Austronesian etymology for (2d). since their historical ranges are more remote. Even in the
As this discussion suggests, it is not a straightforward case of Koreanic, if Ahn’s hypothesis that the Korean-style
matter to identify cognates in this lexical domain in Northeast bronze dagger culture entered the peninsula from central
Asia, and it is not straightforward to distinguish inherited Liaoning is correct, Koreanic may represent an earlier, pre-
cognates from loans. In this section, I will confine myself to rice cultivation dispersion from Shandong. Any rice culti-
some general observations about rice-related vocabularies in vators left behind in the greater Shandong region after
Korean and Japanese and possible relations between them as Miyamoto’s third dispersion were absorbed by the expan-
they relate to the Shandong–Liaodong dispersal hypothesis. sion of Sinitic, so no trace of their languages remain there.
As Vovin observes, some of the items in (2) are the Cognate vocabulary between the surviving languages dis-
products of internal semantic specialization, such as (2b) persed from Shandong precedes rice cultivation.
*mə/omi “unhulled rice” < *mə/om- “pound” + *-i nomi- This interpretation is supported by the semantics of cog-
nalizer, and (2e) *ipi “cooked rice,” which Vovin derives in nate agricultural vocabulary in Korean and Japanese, as
a similar way from a verb *ip- “eat.” These derivations raise illustrated in (3).
the possibility that the ancestor language lacked specialized

(3)

(3a–c) are excellent semantic and phonological fits, but they are “Chinese water chestnut” (Eleocharis dulcis), which is other-
often rejected as loans (e.g., Vovin 2010) on the assumption that wise unetymologized.
agricultural vocabulary is too recent to be inherited. But none of Summing up the results of this section, Japonic gives some
these terms are dedicated to rice agriculture. Given the antiquity evidence for agricultural vocabulary cognate with other lan-
of the first Shandong/Liaodong dispersion (5500 BP), these guages in Northeaast Asia, but none of this vocabulary is
terms may represent a shared inheritance as old as five millen- dedicated to rice. Koreanic shows relatively little vocabulary
nia. (3d) is a rice-related term in Koreanic, but if the Japonic dedicated to rice at all. These facts are consistent with a
item is cognate, the original meaning was not specialized for dispersal of some languages, including Koreanic, from Shan-
rice agriculture. (3e) also represents a semantic shift, and an dong prior to the advent of wet field rice cultivation in that
item unrelated to rice. The Japanese term must be quite old area. The cognate agricultural vocabulary shared by
since it provides an etymology for Chinese bíqí 荸薺 < pidzej Koreanic and Japonic precedes wet rice agriculture.
Rice (2011) 4:149–158 157

Conclusion Baxter, W, Sagart, L. (n.d.) Baxter-Sagart Old Chinese reconstruction

(Version 1.00). Online at [Link]
id=1217. Accessed 18 Dec 2011.
In this paper, I have sketched a specific historical scenario that Beckwith C. Koguryo, the language of Japan’s continental relatives: an
attempts to explain the linguistic ecology of the non-Sinitic introduction to the historical-comparative study of the Japanese-
language families in Northeast Asia associated with wet rice Koguryoic languages with a preliminary description of Archaic
Northeastern Middle Chinese. Leiden: Brill; 2007.
agriculture, Japonic and Koreanic. This scenario is couched
Bellwood P, Renfrew C. Examining the farming/language dispersal
within the general hypothesis of a diffusion of agriculture hypothesis. Cambridge: McDonald Institute for Archaeological
from the area around the Shandong peninsula to the north Research; 2002.
and east. According to the scenario, Japonic arrives in the Bentley J. The search for the language of Yamatai. Jpn Lang Lit.
2008;42:1–43.
Korean peninsula around 1500 BCE and is brought to the Crawford GW, Lee G-A. Agricultural origins in the Korean Peninsula.
Japanese archipelago by the Yayoi expansion around Antiquity. 2003;77:87–95.
950 BCE. On this view, the language family associated Georg S, Michalove PA, Manaster Ramer A, Sidwell PJ. Telling
with both Mumun and Yayoi culture is Japonic, although general linguists about Altaic. J Linguist. 1999;35:65–98.
Franke H. The forest peoples of Manchuria: Kitans and Jurchens. In:
the association of a culture in the archaeological sense
Sinor D, editor. The Cambridge history of Inner Asia. Cambridge:
with a single language family is almost certainly an Cambridge University Press; 1990. p. 400–23.
oversimplification.12 Haugen, E. The ecology of language; essays by Einar Haugen. Select-
Koreanic arrives in the south-central part of the Korean ed and introduced by Anwar S. Dil. Stanford, CA: Stanford
University Press; 1972.
peninsula around 300 BCE with the advent of the Korean-
Harunari H. Yayoi jidai no nendai mondai [The problem of the dating
style bronze dagger culture. Its speakers coexist with the of the Yayoi period]. In: Nishimoto T, editor. Yayoi jidai no shin
descendants of Mumun cultivators, and thus with Japonic, nendai (Shin Yayoi jidai no hajimari 1) [The new dating of the
well into the common era. Each of these demic diffusions, as Yayoi period (The beginning of the Yayoi period revisited)].
Tokyo: Yuzankaku; 2006. p. 65–89.
well as the later dispersions of Koreanic and Japonic, result Harunari H, Imamura M. Yayoi jidai no jitsu nendai [The actual date of
in founder effects which diminish the internal variety of the the Yayoi period]. Tokyo: Gakuseisha; 2004.
language family. Japonic and Koreanic, as well as possibly Hattori S. On the method of glottochronology and the time depth of
other Northeast Asian languages, share some agricultural Proto-Japanese. Gengo kenkyū. 1953;22(23):29–77.
Hattori S. Nihon sogo ni tsuite [Regarding proto-Japonic] 1–22. Gek-
vocabulary, but this shared vocacbulary precedes rice
kan gengo. 1977–1979;7(1)–7(3), 7(6)–8(12).
farming. Hudson M. The ruins of identity. Honolulu: University of Hawai’i
Press; 1999.
Janhunen, J. Ethnicity and language in prehistoric Northeast Asia. In:
References Blench R, Spriggs M, editors. Archaeology and language II.
London: Routledge; 1998. p. 195–208.
Kim B-H. Hangukŏ ŭi kyet’ong. Seoul: Minŭmsa; 1983.
Ahn S-M. Kogohak ŭro pon Han minjok ŭi kyet’ong [The lineage of Kobayashi A. Jōmon kara Yayoi e no tenkan [The transition from
the Korean people from the viewpoint of archaeology]. Hanguk sa Jōmon to Yayoi]. In: Hirose K, editor. Yayoi jidai wa dō kawaru
simin kangchwa [The Citizens’ Forum on Korean History] ka [How does the Yayoi period change]. Tokyo: Gakuseisha;
2003;32:98–102. Ilchogak, Seoul. 2007. p. 135–57.
Ahn S-M. The emergence of rice agriculture in Korea: archaeobotan- Kuzmin Y. Geoarchaeology of prehistoric cultural complexes in the
ical perspectives. In: Hosoya L, Sato Y, Fuller D, editors. Archae- Russian Far East: Recent progress and problems. Bulletin of the
ological and anthropological sciences 2 (Special issue, The Indo-Pacific Prehistory Association 28; 2008. p. 3–10.
Archaeobotany of Asian Rice). 2010. p. 89–98. Lee K-M, Ramsey R. A history of the Korean language. Cambridge:
Atkinson Q. Phonemic diversity supports a serial founder effect model Cambridge University Press; 2011.
of language expansion from Africa. Science. 2011;332(6027):346–9. Lee S, Hasegawa T. Bayesian phylogenetic analysis supports an agri-
cultural origin of Japonic languages. Proceedings of the Royal
Society B. 2011. doi:10.1098/rspb.2011.0518. [Link]
[Link]/content/early/2011/05/04/rspb.2011.0518.
full. Retrieved 5 May 2011.
12
There are points of contact between the hypothesis advanced in this Martin SE. Lexical evidence relating Korean to Japanese. Language.
paper and the proposals of Unger (2009) and Vovin (2011, to appear). 1966;42(2):185–251.
Both Unger and Vovin associate proto-Japonic with demic diffusion Miyamoto K. Nōkō no kigen o saguru: Ine no kita michi [Searching for
from the Korean peninsula to the Japanese archipelago at the beginning the origins of agriculture: the road by which rice came]. Tokyo:
of the Yayoi period. Unger however argues that rice agriculture reached Yoshikawa kōbunkan; 2009.
the Korean peninsula from the south. Vovin holds, as I have suggested Nishimoto T, editor. Yayoi jidai no shin nendai [The new dating of the
here, that Koreanic entered the peninsula from the north and displaced Yayoi period]. Tokyo: Yuzankaku; 2006.
the previous Japonic-speaking population, but he associates this event Poppe N. Vergleichende Grammatik der altaischen Sprachen. Teil I.
with a horseriding culture equipped with iron weapons. In contrast, I Vergleichende Lautlehre [Comparative Grammar of the Altaic
have related the diffusion of Japonic and Koreanic with two specific Languages, Part 1: Comparative Phonology]. Wiesbaden: Otto
archaeological events: the Mumun–Yayoi diffusion through the penin- Harrassowitz; 1960.
sula into the archipelago from 1500–950 BCE and the diffusion of the Ramstedt GJ. Einführung in die altaische Sprachwissenschaft II.
Korean-style bronze dagger culture from 300 BCE. Formenlehre. In: Aalto P, editor. Introduction to Altaic linguistics,
158 Rice (2011) 4:149–158

volume 2: morphology. Helsinki: Suomalais-Ugrilainen Seura; Unger JM. The role of contact in the origins of the Japanese and
1952. Korean languages. Honolulu: University of Hawai’i Press; 2009.
Robbeets M. Is Japanese related to Korean, Tungusic, Mongolic and Vovin A. Japanese rice agriculture terminology. In: Blench R, editor.
Turkic? Wiesbaden: Otto Harrassowitz; 2005. Archaeology and language II. London: Routledge; 1998. p. 366–78.
Robbeets M. How the actional suffix chain connects Japanese to Vovin A. Japanese, Korean, and Tungusic: evidence for genetic rela-
Altaic. Turk Lang. 2007;11(1):3–58. tionship from verbal morphology. In: Honey DB, Wright DC,
Schuessler A. ABC etymological dictionary of Old Chinese. Honolulu: editors. Altaic Affinities, Proceedings of the 40th Meeting of
University of Hawai’i Press; 2007. PIAC. Bloomington: Research Institute for Inner Asian Studies,
Shinmura I. Kokugo oyobi chōsengo no sūshi ni tsuite [Regarding University of Indiana; 2001. p. 83–202.
numerals in Japanese and Korean]. Geibun. 1916;7(2):4. Vovin A. The end of the Altaic controversy. Cent Asiat J. 2005;49
Starostin SA, Dybo AV, Mudrak OA. Etymological dictionary of the (1):71–132.
Altaic languages. Leiden: Brill; 2003. Vovin A. Koreo-Japonica: a re-evaluation of a common genetic origin.
Toh SH. Samhan ŏ yŏngu [Research on Samhan language]. Seoul: Honolulu: University of Hawai‘i Press; 2010.
Che-i-aen-ssi; 2008. Vovin A. From Koguryŏ to T’amna—slowly riding to the South with
Unger JM. Studies in early Japanese morphophonemics. 2nd ed. speakers of Proto-Korean—Ms. University of Hawai’i at Mānoa;
Bloomington: Indiana University Linguistics Club; 1977. 1993. 2011. To appear in Korean Linguistics 15.

View publication stats