#1
                           1. ORDER OF THE HEAD OF GO-DAI IN, 1135
                                 (SK, I; originally a Gon-Shuin doc. now lost.)
THIS is the oldest document extant in which the place Iriki appears. Poor as the copy in SK is, the
document affords grounds for inference regarding certain important phases of the institutional his-
tory of this locality prior to the advent of the Shibuya warriors.
  The order was issued by the head(in-su) of the Buddhist house, Go-dai in, which was closely
related to the Shinto temple, Niita Hachiman, both situated by the river Sendai not far from its
mouth. As it was customary in those days for Shinto temples to have near them Buddhist houses
which were associated with them and which transacted their worldly business for them, so the
affairs of Niita Hachiman were managed by Go-dai in.
  Niita Hachiman is supposed to have been dedicated to the spirit of Ninigi-no-Mikoto, the mythi-
cal grandfather of the first emperor of Japan, Zhimmu. Ninigi is said to have sojourned for a time
at Kasasa (often identified with the present Kaseda, further south in Satsuma) on his descent from
"High Heavenly Plain," and them moved to the site of this temple; his supposed tomb is situated
directly behind the temple. The latter was established at an ancient date at the side of a hill known
as E-no-yama, and, after the fire of 1173, rebuilt on its summit, where it stands to this day, enjoy-
ing the high reverence it has always received from all classes. The temple was once a great institu-
tion, with a considerable number of Shinto, Buddhist, and lay persons devoted to its service,1 who
lived on both sides of the straight road leading to the river and in the immediate vicinity of E-no-
yama. During the period of the civil war, the fortune of the temple, like that of many another reli-
gious institution in Japan, declined greatly, but its prestige was such that, in the sixteenth century,
the Shibuya lords, who then dominated this region, repaired the edifices and provided for their
maintenance. On the surrender of the Shibuya to the Shimadzu, the latter bestowed upon the insti-
tution a solicitous care, regarding it as the first temple in Satuma.1
  Go-dai in was built at an unknown time a mile and a half west of the temple; and, as the chief
Buddhist house connected with Niita Hachiman, wielded great powers during the first feudal ages.
Then it decayed, probably during the period of the civil war, leaving merely the foundation stones
of its edifices. The annual rites for the five Buddhist deities that were originally revered at Go-dai
in, however, continued to be performed at the temple throughout the ages,2 till Shinto was stripped
of its Buddhist associations immediately after the downfall of the feudal regime of Japan.
  Returning to our document, we should infer from it that Masanobu resided, in 1135, near Go-dai
in, while the title of the latter's in-su was held by a priest at the Buddhist church connected with
another greater Hachiman temple established far away at Iwa-shimidzu, near the imperial capital,
Kyoto. This great temple was then considered chief among the numerous temples in Japan which
the people inaccurately thought to be all dedicated to the same deity Hachiman, who had come to
be identified with the ancient emperor O-zhin. So were the Niita temple and Go-dai in subordinate
in an ill-defined manner to Iwa-shimidzu, receiving its orders and following its jurisdiction. Cf. No.
3 below.
  The first Hachiman temple in Kyu-shu, and indeed perhaps the oldest in Japan, was the one at
Usa, in Buzen, some one hundred miles in a straight line to the northeast of Niita; the Buddhist
church connected with the temple at Usa was Mi-roku zhi. Mi-roku zhi stood to Usa Hachiman in
a relation similar to that which obtained between Go-dai in and Niita Hachiman. At the time of
this document, the latter two are found to have been in some respects subject to the former.3 Some
of Niita's domains were, therefore, also Usa's, or, in other words, the domains of Go-dai in also
belonged to Mi-roku zhi, as both drew revenues from them (see No.9).
  Our document reveals the fact that Mi-roku zhi and Go-dai in, or, Usa and Niita Hachiman,
held domains in the fertile basin of the lower course of the river Sendai; this is precisely the country
which more than a century hence fell into the hands of the Shibuya after their descent from Kama-
P90
kura. Since we shall be concerned throughout this volume with this and neighboring regions, it
is well to take note of its early condition as seen in this document.
  It will be observed that the financial control, at least, of these domains had still been poorly
organized. From the brief statement in the text, one may infer that people had been allowed to
settle freely on the land, and to cultivate it into vegetable or grain fields or to irrigate it and turn it
to rice paddies; there had been little system in the assessment of the taxes, and naturally their
evasion had been common. Such a state of things will not surprise us, if we recall how peculiarly
free the condition of arable land had been in Satsuma for many a century. How much greater
financial control resulted from the present order may not be known, but we presume that the gen-
eral laxity must have continued more or less, and facilitated in this locality the establishment of
warriors' power that was soon to follow, if it had not already begun.
  It now remains for us to consider who Masanobu, the manager of the Go-dai in domains and
recipient of this order, might be. We suspect that he was of the Tomo family, which had long been
settled in this neighborhood, and of which we shall bear often, beginning with the next document.
This surmise receives some support from the genealogy of the Miyasato family (cf. No.11), in
which one Tomo Masanobu appears as successor to Ki Masato and father of Ki Masa-ie. Masa-ie
is conspicuous in the cadaster of 1197 (cf. Nos.9 and 11).

"ORDERED to Masanobu, the man-dokoro4 of Go-dai In,
  "That, in accordance with the order, the man-dokoro shall at once assign5 for culti-
vation the ta6 and hata6 held by the Zhi,7
"[which lands are] situated at
"Togo, 8 of Taki;8 Chugo,8 of the same; Iriki in,9 Satsuma kori9 and Miyasato go,9
  -among [the places given] in lieu of Ada kori.10
"The aforesaid ta6 and hata6 are in Spring cultivated at will by individual [persons],
without letting the officials11 know; the crops in Autumn are unexamined, and the
[payment of the] fixed dues is evaded. These are exceedingly perverse acts. Hereafter,
inasmuch as the term of office of the in-su12 is limited, while the man-dokoro Ma-
sanobu is a perpetual man,13 the man-dokoro is commanded at once to assign5 for
cultivation the aforesaid ta and hata held by the Zhi.7
  "Of these, as regards [the lands at] Iriki kori,9 though [they are accompanied by]
legal documents defining their divisions in detail, heretofore it has been neglected to
make report of them to the officials; which is no light offence.
  "It is ordered that the command shall be immediately executed in accordance with
this order. No error or omission shall be made. For these purposes, the order is hereby
issued.
  "Ho-en 1y. 19m. 25d.14 [1 December 1135],14
"The In-su, Iwashimidzu gon zhi-su,15 [of the rank] dai hosshi."16




  1 San-goku mei-sho dzu-ye, XIII, 10-11. A report dated 1331 of the functionaries connected with
the temple enumerates 319 persons; Gon Shuin mon-zho, VII.
  2 San-goku mei-sho dzu-ye, XIII, 20.
  3 In an undated list, also, of the holdings of Ki-ta in, of Mi-roku zhi (the Iwa-shimidzu mon-zho,
II, 146-147), which probably is about a hundred years later than the present document, appear
Niita sho[?] and Go-dai in. Cf. also No.9.
  4 Man-dokoro. This term is ordinarily used in the sense of the administrative office of a high
personage; the first sho-gun, Yoritomo, used the term in designating his executive office. Here,
however, man-dokoro refers, not to an office, but to the official of a managing office; it means the
manager of Go-dai in and its domains. This usage of the term is uncommon.
  5 To whom is not stated; probably to cultivators who had settled here and had been cultivating
P91
the lands in question. The man-dokoro was now commanded, it would seem, to charge each culti-
vator, in definite terms, to be responsible for the cultivation of the land he was exploiting and for
the payment of the dues from it.
  6 Ta are wet fields, level and terraced, and flooded for the cultivation of rice. Hata are upland
fields, either for grains other than rice, or beans, or for vegetables; sometimes gardens planted with
fruit-trees are called hata. Ta and hata are terms that continually recur n our texts.
  7 Zhi and in are Buddhist churches or monasteries, in being sometimes but not always sub-
ordinate to zhi. Here Mi-roku zhi, at the Shinto temple Hachiman at Usa , in Buzen, is referred to.
Cf. No.9.
  8 Togo and Taki, both situated further north than Niita Hachiman on the same side of the river
Sendai, are usually regarded as two distinct places. But here Togo (originally meaning "east go")
and Chugo ("middle go") would seem then to have been districts within the kori of Taki. For the
units kori and go, see the following note.
  9 The largest administrative territorial unit was kuni, often translated as "province," of which
there were more than sixty in Japan in this period. Each kuni was composed of several kori, their
numbers in the kuni and their sizes varying considerably. The in was an equivalent of kori; the two
terms were sometimes used interchangeably. In its origin, the in and the proper name by which it
was designated had been largely unofficial; and it occurs but rarely out of southern Kyu-shu. The
go was a unit within the kori or in, consisting of a number of mura, hamlets. Since these terms
frequently occur, their original forms are preferred to arbitrary translations.
  10 This phrase shows that formerly the temple held domains in Ada kori, further south, due
probably to the tradition of the earlier residence of the deity Ninigi in that locality. These domains
had apparently been exchanged for those in the north, some of which are mentioned here.
  11 Sa-da nin, officials for the domains acting under the man-dokoro.
  12 Head of the in, the Buddhist house. This in, though written in the same character, should not
be confused with the territorial division of the same name. See notes 7 and 9.
  13 It is probably meant that the place was held for life or by heredity, while the tenure of the
in-su office was temporary.
  14 In reducing the dates of the lunar calendar to those of the solar (Julian) calendar, Mr.
Ernest W. Clement's Japanese Chronological Tables, a highly valuable work, has been followed. The
San-sei so-ran has also been consulted.
  Throughout this volume, the abbreviations y, m, and d, will be used for, respectively, year,
month, and day; thus, Ho-en 1y. 10m. 25d. stands for the twenty-fifth day of the tenth lunar
month of the first year of the year-period Ho-en.
  15 Buddhist ecclesiastical office.
  16 Buddhist ecclesiastical rank.