^ .vi^, 
 
 
CORNELL 
 
 UNIVERSITY 
 
 LIBRARY 
 
 THE 
 
 CHARLES WILLIAM WASON 
 
 COLLECTION ON CHINA 
 
 AND THE CHINESE 
 
"With compUments of 
 
 K.^aaka-nr8,, Yal@ University. 
 
 o^. 
 
 
 THE Origin of the; feudal 
 
 LAND TENURE IN JAPAN 
 
 Cornell University Library 
 HD 914.A79 
 
 The origin of the feudal land tenure in 
 
 3 1924 023 530 417 .„,..,i 
 
 By K. ASAKAWA 
 
 ItKPUNTBD FROM THE 
 
 ^mexitm pisiMat §mm 
 
 ^OL. XX„ No. I 
 
 OCTOBER, 1914 
 
fJmrhy^ 
 
 mi 
 
 ft. 
 
 
[Reprinted from Thk Amkkican JIisioricai, Review, Vol. XX., No. i, Oct., 1914.] 
 
 THE ORIGIN OF THE FEUDAL LAND TENURE IN JAPAN 
 
 IT is quite beyond the range of possibility, in an attempt to prove 
 a certain seqiience of social conditions and results, to try on a 
 nation a series of experiments on a large scale which would require 
 for their maturity circumstances of great and changing complexity 
 extending over centuries. It must then be a matter of uncommon 
 interest to the student of the history of human society to observe 
 that Western Europe and Japan,^ with no mutual relation between 
 them, independently evolved in the course of a long time those 
 peculiarly definite and exhaustive social adjustments that are 
 known as feudal, under conditions and upon principles which 
 were, if significant in their minor differences, also striking in their 
 extraordinary similarity in the main features. Human history, 
 which so usually eludes a comparative study of its great divisions, 
 becomes, under such circumstances, nearly comparable, and should 
 therefore afford invaluable data to the scientific student of social 
 evolution. 
 
 The interest of the theme is enhanced, in the case of Japan, by 
 the fact that she possesses a body of original sources sufficiently 
 large to invite a careful investigation, and larger than may be offered 
 by some of the other non-European nations that are said also to 
 have known feudalism in their history. 
 
 Following is a summary statement of some of the conclusions 
 and problems^ of a study of the origin of the Japanese feudal land 
 tenure on its purely institutional side. For it is thought that an 
 analytical view of the institutional aspects of a great social devel- 
 
 1 " Die beiden grossartigsten Beispiele des entfalteten Feudalstaates, West- 
 europa und Japan." F. Oppenheimer, Der Staat, p. 108. 
 
 2 A full discussion of the subject and of the sources for its study will be 
 presented later in a monograph, to which the present paper may serve tentatively 
 as an. introduction. 
 
 AM. HIST. REV., VOL. XX. — I. I 
 
2 K. Asakawa 
 
 opment might well precede any economic or sociological interpreta- 
 tion of it. 
 
 I. 
 
 The political society of Japan before the seventh century was 
 composed largely of patriarchal units {uji) and of groups of peo- 
 ple {be, tomo) formed after the pattern of these units. Each unit 
 had its chief, who exercised control over its members and the people 
 of the groups attached to the unit, as well as over the land on which 
 they subsisted. The emperor was in theory the head of the entire 
 ruhng tribe, but in reality his power rested chiefly on the members 
 of his own unit and the people of the groups he had specially 
 created, and on the land in possession of the unit and the groups 
 (mita, miyake). He was hardly more than the largest among the 
 patriarchs. One of the latter was rising in power so fast that it 
 seemed almost to rival the imperial authority, while' throughout 
 the realm the inevitable tendency toward decentralization appeared 
 increasingly alarming. It was the inherent dangers of this system, 
 added to the threatened peril of an invasion from China, that ne- 
 cessitated the great Reform of the seventh century.^ 
 
 The Reform was an adaptation of a system of state socialism 
 that had been elaborated in China. It established a broad division 
 of the free people into the governing and the supporting classes. 
 The governing class consisted of a new civil nobility of rank* and 
 ofHce,^ the higher ranks and offices being accompanied with defi- 
 nite grants of rice-land to be held during their tenure, and every 
 holder of rank and many a civil servant being exempt from the 
 payment of tributes and forced labor.® Thg supporting or taxable 
 class constituted the bulk of the free citizens f it was provided with 
 
 IT 
 
 3 Preliminary measures of reform may be traced back to earlier years, but 
 a radical and thorough reorganization of the state-system was begun in 645. 
 After much experimentation, the institutional foundation of the reformed state 
 was nearly completed when the Code of 701 was promulgated. 
 
 * The system of ranks was instituted in 603, and was revised five times be- 
 fore 701. The final system of 701, which continued till i86g, contained nine 
 chief ranks with thirty subdivisions. 
 
 5 Early in the eighth century, there were provided, exclusive of military 
 oflices, civil posts for more than 8300 officials in the capital and more than 3500 
 in the provinces. 
 
 6 The taxes were of three main classes : the rice tax (so) , the tributes of 
 local products {cho), and the forced labor or its equivalent payable in kind (yd). 
 The new nobility was immune from the last two, which were by far the more 
 onerous forms of taxation, but, theoretically, not from the rice tax. Some will 
 probably maintain that at least the office-land (shiki-den) was exempt also from 
 the rice-tax, but it may be demonstrated that such was not the intention of the 
 law, excepting the office-land assigned to the provincial governor (ku-ge-den) . 
 
 7 The population of this period has been estimated as four or five millions, 
 and at the beginning of the feudal regime in 1186 as seven millions or more. 
 
Feudal Land Tenure in Japan 3 
 
 equal allotments of rice-land subject to a periodical redistribution. 
 The citizen owed to the state taxes and military duties. 
 
 This is a schematic view of the social reorganization that was 
 really devised and put into operation. On closer examination, the 
 system will be found to have contained elements which would be 
 certain in the course of time to defeat its own ends. Nor did the 
 system comprehend all the land and all the people that existed. 
 
 First, as regards the people. The new nobility was indeed con- 
 ceived on a new basis, but its personnel was essentially the same as 
 that of the old patriarchal nobility.^ It may be presumed that 
 the new nobility could not so readily have outlived the old habit 
 of coveting private possession of land and men, and of regarding 
 affairs of state much in the light of private concerns. The new tax- 
 able population was also in all probability composed mostly of the 
 old free people of the pre-Reform period and their descendants.^ 
 It is true that the old units and groups as such were no longer rec- 
 ognized by law, and all the free men and women were placed under 
 the direct control of the state; but this must have meant merely 
 that the free were no longer permitted to possess and to exercise 
 public rights over one another. Beyond this, the new system was 
 not intended to do away with class distinctions among the free; 
 nor did it abolish the unfree, who in fact are found after the Re- 
 form in considerable numbers and in conditions of great diversity 
 among themselves. The unfree received small allotments of rice- 
 land, but returned no tributes ; the private itienials were hardly un- 
 der the direct rule of the state. 
 
 Next, as regards land. There is evidence to show that the law 
 of equal allotment of land was in reality largely enforced, though 
 at dififerent times in different parts of the country. But the allot- 
 ting need not be taken in every case in the sense of an actual carv- 
 ing out of rice-land, which from the very nature of the land of this 
 description was impracticable." The allotment was presumably, 
 
 8 An imperial rescript of 646, addressing itself to members of the old nobility, 
 said : " Now the manner of securing your service is to abolish the old offices, to 
 establish new offices, and to mark grades of ranks, so as to appoint you with 
 offices and ranks." Nihongi, XXV., in Koku-shi tai-kei, I. 443- 
 
 9 More precisely, the taxable class consisted of most of the old clan {uji) 
 members, some old group (tomo) people, and those of the old nobility whom 
 fortune had reduced to the common rank. 
 
 10 For three important reasons, among others. First, rice-land (fo) was 
 terraced in different levels for the purpose of irrigation, and could not always be 
 easily parcelled. Secondly, much labor had to be invested in this class of land, 
 in its first cultivation, its continual irrigation, and in the careful culture and 
 harvesting of rice. Thirdly, the rigid delimitation of rice-fields, in addition to the 
 heavy expenditure of labor required in their use, forced their cultivation to be 
 
4 K. Asakawa 
 
 in many instances, a flexible adjustment of tracts in the official reg- 
 ister.^^ Much the same thing may be said of the periodical redistri- 
 bution.12 It was also fatal to the system that the authorities permitted 
 the holder of the allotment to transfer its title by sale or mortgage. 
 Before the eighth century was over, the allotment-land was neither 
 equal nor re-allotted ; it had become an unequal, private possession 
 of people. Even allotting among the new population was all but 
 discontinued by the end of the ninth century. 
 
 A question of the utmost importance is, what was the allotment- 
 land (ku-bun den) ? It was rice-land, and did not include the 
 other kinds of tilled land" nor the vast tracts of ' arable and 
 non-arable^* land that then existed. Moreover, the allotment-land 
 probably was the rice-land that was, at the time of the Reform, not 
 only in cultivation, but also properly registered; there must have 
 existed much rice-land which was cultivated but which escaped 
 registration. It was evident that a system of state organization 
 built upon an equal division of so limited areas of land would 
 soon be upset by developments that must inevitably occur in the re- 
 mainder of the territory.^^ If to this consideration are added other 
 sources of evils latent in the large fiscal immunity of the noble and 
 the unfree, as well as in the insatiable aggrandizement of the Bud- 
 dhist Church, it is not strange that the life of the reformed state should 
 
 intensive. These conditions must have tended early in history to make the 
 rice-land an object of exclusive private possession (whether of the individual or 
 of the family) resisting arbitrary division and redistribution. 
 
 11 " Rice-land for each member is not considered in detail ; a grant is made 
 [in totality] to the house-head in accordance with the number of persons in his 
 house." Ryo no shu-ge, commentary on the Code of 701 compiled in the latter 
 half of the ninth century (1912-1913), I. 392-394. 
 
 12 Cf. Ryo no gi-ge, official commentary on the Code compiled in 833 (1900), 
 p. 102. Only once, in 729, it was decided that an actual complete redistribution 
 be made (Shoku Ni-hon gi, X., in Kohu-shi tai-kei, II. 170), but it is unknown 
 whether this was accomplished. 
 
 13 The homestead (taku-chi), including the adjoining vegetable-land, was, as 
 might be expected, in permanent private possession and alienable, and was also 
 untaxable, as was the land for mulberry and lacquer trees (,en-chi). Some dry 
 grain-land (hatake) was rented to people, and its extension was apparently unre- 
 stricted. 
 
 1* Meadows and woodlands were, in law, not for private appropriation. 
 
 10 Arable land was still extensive, especially in north Japan and in Kyushii, 
 and its opening either as wet rice-land (kona-da) or dry grain-land (hatake) was 
 inevitable ; the new fields thus created would, even if nothing else happened, more 
 and more eclipse the relative importance of the allotment-land in the economic 
 life of the people. Moreover, the prevailing system of agriculture in Japan at 
 this time was evidently that of scattered farms (Einselhof) based upon individual 
 ownership, not in an over-developed form of the system, but rather in its earlier 
 stages. It was futile to try successfully to supersede the system with another of 
 a foreign origin adopted for the convenience of the authorities. 
 
Feudal Land Tenure in Japan 5 
 
 have been so brief as it was.^° Out of its failure was evolved a 
 feudal system. 
 
 The gradual evolution of the feudal tenure which immediately- 
 followed the Reform may be analyzed in two sections : first, the 
 origin and growth of private landed estates, called shXi; secondly, 
 the rise of the warrior classes, and their gradual control of the shio. 
 
 II. 
 
 The sho was largely an illegal growth. When it made its ap- 
 pearance in the eighth century, the sho was seen to possess these 
 three characteristics : it contained, as its chief original element, a 
 recently cultivated area of rice-field; was under private possession 
 and private management ;^^ and enjoyed or claimed a degree of fis- 
 cal immunity. This species of land was destined in three or four 
 centuries to absorb the greater part of the taxable land and people 
 of Japan, while making itself almost wholly immune, and to arro- 
 gate to itself some of the sovereign functions of the state. It was 
 the sho, therefore, that overthrew the Japanese state-system recon- 
 structed in the seventh century. What was the origin of the shol 
 
 It is evident that the cultivation of new soil formed an essen- 
 tial factor in the derivation of the sho. Impelled by the need of 
 satisfying the growing luxury of the court and its devotion to the ma- 
 terial welfare of the Buddhist Church, and, above all, by the rapid 
 increase of population in central Japan in relation to the produc- 
 tion of agricultural wealth, the government, especially in the first 
 half of the eighth century, took bold, if not always wise, measures 
 to extend cultivation. These measures comprised both the tilling 
 of new land as an official enterprise — that is, under the supervision 
 and at the expense of the government — and the encouragement of 
 private cultivation. 
 
 The undertaking by the government seemed naturally to have 
 yielded the best results in sparsely populated north Japan" and 
 
 leThe review that has here been given of the Reform of 645-701, though 
 necessarily very brief, contains much that will supplement as well as modify the 
 ideas published in 1903 in my Early Institutional Life of Japan. The whole 
 subject is full of debatable points. 
 
 17 The word sho (Chinese, chuang) meant a rural house, or house attached to 
 a rural estate. Its derived meaning, the estate, rather than the house, later 
 became the chief meaning of the term. It will be observed that in either sense 
 private enjoyment of property is implied. 
 
 18 There is a highly interesting document of 722 recommending the cultiva- 
 tion of one million cho of land, or nearly two million acres. This is not a 
 place to prove by other evidence that the scheme was intended chiefly for north 
 Japan, and that it was really carried out in part. 
 
6 K. Asakawa 
 
 Kyushu/^ where arable land was abundant and the settlement of 
 people from other regions was needed and encouraged. The newly 
 tilled land was added to other unallotted land as "public land " ( ko- 
 den) or "surplus land" (zho-den), and held in reserve for future 
 allotment and other pubHc uses ; it was usually rented to the people, 
 in order to insure its maintenance and to derive a revenue from it. 
 
 Before the eighth century was half over, however, their religious 
 zeal betrayed the authorities into a dangerous policy relative to this 
 class of land. While, on the one hand, forbidding the private trans- 
 fer of land to Buddhist temples, the government, on the other hand, 
 gave them extensive pieces of land, both out of the reserved tilled 
 areas and from still uncultivated tracts; soon the government even 
 forgot its earlier intention, and encouraged donations of land to 
 temples by private citizens. And the "temple-land" {zhi-den), 
 officially recognized as such, was untaxable. Conditions were, 
 therefore, ripe for its further expansion by purchase, mortgage, 
 gift, or nominal conveyance. It is mostly in connection with Buddhist 
 institutions that we see the mention of sho in the eighth century. 
 
 The next century witnessed, not only further large grants to 
 Buddhist temples, but also the reservation and partial cultivation of 
 extensive tracts set apart for free disposal by the emperor^" {choku- 
 shi den, imperial lands). It is probable that a large part of the 
 areas officially cultivated in the preceding century as " surplus 
 land " was henceforth incorporated in the " imperial land ".^^ The 
 creation of the latter was accompanied by grants of these tracts and 
 other cultivated and uncultivated land by the emperor to members 
 of his house and other high personages {,shi-den, granted land). The 
 " imperial land " was immune from taxes, as was also the " granted 
 land", like the "temple-land". In a short time an abuse which 
 might have been anticipated became so frequent as to call forth an 
 edict, in 902, to prevent its repetition; namely, arbitrarily to claim 
 immunity of privately cultivated pieces of land by insisting either 
 that they were originally imperial or that they had been given by 
 subjects to imperial persons. The prohibition was not efficient, and 
 its meagre effects were more than neutralized from the latter part 
 of the next century, when a peculiar bicameral government, which 
 
 19 In 823 there were 10,000 cho, or about 20,000 acres, of " surplus land " in 
 Kyushu. 
 
 20 Between 828 and 844 alone, more than 7200 cho seem to have been re- 
 served as " imperial lands " 
 
 21 The exact meaning, however, of the sweeping decree of 885 declaring all 
 the wet and upland tilled and arable areas as " imperial lands " is not clear. 
 Rui-zhu koku-shi (181 6), chap. 159, leaf 20. 
 
Feudal Land Tenure in Japan 7 
 
 then began its regime, encouraged, with a view to adding to the 
 imperial revenues, the commendation of private lands to the ex- 
 emperors as " imperial land ". By that time, many great private 
 estates of land had already grown up from another source. 
 
 This other source was the land newly cultivated (kon-den, cul- 
 tivated land) by private persons independently of the imperial 
 house, as well as meadows and woodlands around the cultivated 
 areas as nuclei. Japan being at this time in that economic stage 
 in which money was poorly circulated and land was the chief form 
 of wealth, the disposition of arable land liable to be turned to culti- 
 vation should have been a matter of the greatest moment for the 
 government to consider. According to the scheme of the Reform, all 
 uncultivated land was to be under the direct control of the state, 
 and the benefit of its cultivation was, in principle, to accrue to the people 
 at large. In practice, however, when the need of extending cultiva- 
 tion was felt, the authorities met the problem with a vague and vacillat- 
 ing policy. They were forced to discover after costly experiments 
 that restrictions tended to discourage new cultivation, while its 
 encouragement produced inequalities and hardships among the peo- 
 ple. Finally, already before the middle of the eighth century, it 
 was decreed that the cultivator should have the ownership of the 
 land he opened to cultivation, while its area should be limited in 
 accordance with his rank. The intent of the law was that its bene- 
 fit should be enjoyed by the common people, rather than by local 
 magnates, court nobles, and Buddhist institutions. But it is needless 
 to say that this end could hardly be gained; that the law not only 
 failed to check the wholesale appropriation of new land by power- 
 ful men that had begun as soon as the Reform government was 
 established, but also assured the steady progress of occupation. 
 The government never succeeded in finding an effective policy re- 
 garding this difficult but all-important question ; all it could do after 
 the ninth century was to promulgate impotent decrees condemning 
 the aggrandizement of great holders of land. 
 
 It would seem essential, ih the study of the origin of the sho, 
 clearly to distinguish the two sources that have been discussed. 
 Immune sho were originally granted by the emperor, and were 
 therefore as such legitimate; while the sho of private origin 
 were born largely in defiance of the law, and, even when some of 
 them received a grudging sanction of the government as lawful pos- 
 sessions, were not in all cases immune. The " granted lands " were 
 standing models of immune estates, but they were, though many and 
 large, not indefinitely expansive. Privately "cultivated lands" 
 
8 K. Asakawa 
 
 could, on the contrary, be multiplied more readily and were in fact 
 transferred more freely. The former were a cause of the sho as 
 tax-free lands, but the latter were, it may be held, direct origins 
 of most of the sho whether immune or not. It is seen that, although 
 the sho appears from the early eighth century, it was not for at 
 least four centuries thereafter necessarily immune from all taxa- 
 tion. ^^ In fact, the study of the process of how the private sho. 
 originating as I think directly from the newly cultivated land, came 
 gradually to be assimilated by the immune tenure of the "granted 
 land " and finally to absorb the latter and nearly all the other species 
 of land^^ in the ever expanding scope of the term, constitutes one 
 of the most interesting problems of the period. 
 
 III. 
 
 During the ninth century, if many a sho was still partially tax- 
 able, the commendation of person or of land to the sho by a peasant 
 was also still largely incipient in form and condemned as illegal. In 
 the next century and a half, however, a great progress was notice- 
 able, both in the commendation of land, and in the rise of immune 
 sho; the simultaneous advance of these two was natural, for im- 
 munity and commendation would stimulate each other. We now 
 turn, therefore, to some of the causes of the progress of the sho. 
 I venture to suggest that this was due, in its institutional aspects, 
 to the parallel extension that took place, of the grant of fiscal im- 
 munity, on the one hand, and, on the other, of the private trans- 
 fer of real rights pertaining to land. Here, again, one was official 
 and legitimate, and the other was private and beyond the cognizance 
 of law. Although their progress was mutually dependent, it is 
 possible to treat them separately up to a certain point. 
 
 First, as to the extension of the grant of fiscal immunity. From 
 the tenth century one may observe an increasing prevalence, among 
 
 22 It has been suggested that the immunity of the sho ai private origin may 
 have been claimed on the pretext that it was an extension of a " house-land " 
 (taku-chi), '"sho" originally signifying a rural house, or of the land immediately 
 adjoining the house (en-chi, garden-land), as these classes of land were legally 
 private and immune. This may suggest why the word en was often used sepa- 
 rately or in conjunction with sho, in speaking of the sho as an estate, but it fails 
 to explain its immunity, for the institution was not in its early centuries invari- 
 ably immune. 
 
 23 In a list of the sho in the province of Yamato belonging to a temple which 
 probably dates from the early twelfth century, each sho is divided into " untax- 
 able " and " public " tracts, and in the former are included original " rank-lands " 
 and " office-lands ". To-dai zhi yo-roku, VI. This is an instance among many 
 that show that these species of land had become assimilated into sho, and also 
 that sho often comprised taxable areas. 
 
Feudal Land Tenure in Japan g 
 
 the holders of slw, of resort to a legal custom which had already 
 begun but had not been firmly established;^* namely, to invoke 
 charters^^ from the central and provincial authorities which in ex- 
 plicit terms recognized the estates as sho, defined their boundaries 
 and extents, and exempted them from various taxes.^° An impor- 
 tant consideration about the charters was that they usually stated 
 that, as before, the estates should be immune from some or all 
 taxes, and that provincial ofHcials should not invade them. A still 
 more important point is, however, that it was discovered at the ex- 
 aminations held after 1069, that many sho, as might have been ex- 
 pected, had no accompanying charters,^' that is, were not legally tax- 
 free. I think it may be shown, perhaps conclusively, that a charter 
 was not first issued for a sho as such, but for an " imperial land " 
 granted to a Buddhist temple or to a noble personage, which, be- 
 cause of the increasing laxity of the provincial administration after 
 the ninth century, needed to be protected against the visitation 
 of the local official.^* From this modest beginning, as the number 
 and prestige of the sho increased, many of them that were controlled 
 by imperial or noble personages or by powerful temples, succeeded 
 tTirough influence or favoritism in likewise securing charters for 
 the estates, and the practice tended to spread. 
 
 2* There were, from the eighth century, charters for the tilling of new land, 
 and of the grants of land to temples (see, e. g., Dai Ni-hon ko-mon zho, V. 597, 
 639-645, 652-655, 767, VI. 597-598; Zoku-soku gun-zho rui-zhU, XI. 121; Ko-bo 
 dai-shi zen-shu, I. 769-770; etc.), but the oldest immunity-chatteT that I have seen 
 is dated 845 (see Ko-mon zho rui-san, third ed., p. 87), and it is difficult to 
 duplicate the example from the. ninth century. 
 
 25 The full regular charters of a sho consisted of orders from the Grand 
 Secretariat, from the Department of the Affairs of the People, and from the 
 provincial and district authorities. Later there were charters less formal but 
 none the less potent emanating directly from the emperor or the " cloistered " 
 ex-emperor. In all regular procedure, a provincial oflScial visited the estate with 
 an agent of the sho, and put in stakes at its corners to mark its authorized 
 extent. After the granting of the charters, local officials were not at liberty to 
 enter the sho without special imperial sanction and without being accompanied by 
 a representative of the estate. 
 
 26 An example of a comprehensive charter occurs in 1050, which declared 
 certain holdings of a temple "untaxable lands, free from intrusion by provincial 
 envoys, and exempt from extraordinary miscellaneous services ". Ko-ya san mon- 
 zho, VII. 261-266. 
 
 27 A sho belonging to a Shinto temple was, in 938, declared to have originally 
 been a waste land which " for fifty years had been established as a sho and 
 suffered no intrusion and examination "- Unchartered sho of this kind became 
 numerous. Often a great noble would disregard the formalities of invoking 
 charters, and summarily order the meek provincial authorities to assist in the 
 founding of a new sho. Provincial charters alone were not infrequently con- 
 sidered sufficient. 
 
 28 Study the case of 84s referred to in note 24 above. 
 
JO K. Asakawa 
 
 What further confirmed the status of the sho was the attempt of 
 the emperor Go-Sanjo after 1069 to aboHsh all the sho established 
 since 1045, and even those prior to that date that had no confirm- 
 ing charters. Unchartered sho were thus largely eliminated, and hence- 
 forth the possession of charters was regarded as a necessary con- 
 comitant of the existence or establishment of a sho; or, in other 
 words, the sho was henceforth necessarily tax-immune in whole or 
 in part. As fate would have it, the bicameral regime under three 
 successive ex-emperors, which began soon after this date, happened 
 to cause the number of sho to be greatly increased all over the 
 country, marking a sudden and long progress in the creation of 
 immune tracts of land.^^ In the middle of the twelfth century, im- 
 munity and the sho were almost interchangeable terms, so that all 
 immune lands were alike called sho, and all private estates tended 
 to become sho. The " imperial " and " granted " lands, the immune 
 character of which, as I suppose, had originally furnished the model 
 for the immune j/io,^" had now been completely absorbed by the 
 latter. Not only that, but also the essentially private terminology 
 that had grown up regarding the sho had come to be applied also 
 to public administrative units of territory, even in official documents. 
 
 Secondly, as regards the private transfer of real rights concern- 
 ing the sho. Immunity was a public right ; its extension could not 
 have been so complete as it was, nor could it have become a factor 
 to bring about a feudal tenure of land, if immunity had not pro- 
 gressed, as it did, side by side with the private conveyance of cer- 
 tain real rights relative to the sho. These rights, however varied 
 and however freely transferable they became in later years, may 
 be presumed to have originally been derived from the fact of the 
 first cultivation of the land in question The elementary forms of 
 the rights were those in relation (i) to the use of the land and 
 (2) to the dues in kind and in labor payable ultimately by the actual 
 tillers of the soil. These and other real rights of the sho, when 
 they became greatly diversified in later years, were collectively 
 
 29 In an official record dated 1221, it is seen, for instance, that of the 86 
 public districts and private estates in the province of Noto, only two sho were 
 established before 1050; eight estates, including very extensive sho, were founded 
 apparently in the first half of the twelfth century; and the remaining 76 units 
 were either made sho or, though still nominally public, treated much as if they 
 also were private possessions, between 1186 and 1221. Noto no kuni den-sU 
 cho, in Shi-seki shu-ran (revised), XXVII. 71—76. 
 
 30 As a matter of fact, there were several other terms than sho that were 
 applied to more or less immune estates, but their precise legal meanings are not 
 yet clear, and probably cannot be determined in all cases ; for a certain indefinite- 
 ness would naturally surround institutions whose origins were private and in part 
 illegal. 
 
Feudal Land Tenure in Japan 1 1 
 
 called shiki. The transfer of shiki of a sho from one man to an- 
 other may first have implied a separation of the right of some use 
 of the land from its ownership. This sfparation once made, there 
 was nothing to prevent further division, in the hands of many per- 
 sons, of the various shiki, as, for example, the rights of the exploita- 
 tion and of the control over the various dues, of the same sho. 
 
 As has been pointed out by Professor Nakada,^^ a distinction 
 should be made between two* kinds of transference of shiki ac- 
 cording to the relative position in society of the persons effecting 
 the transfer. The surrender of a shiki by one person of a lower 
 station to another of a higher was termed ki-shin, which may almost 
 literally be translated as commendation ; a shAki granted by a highen 
 person to a lower, for the purpose either of the management or of 
 the economic exploitation of the sho, was often known as on-kyu, 
 meaning benevolent gift, for which we shall use the term benefice. 
 It will at once be observed that there was a wide: difference, in their 
 institutional origin, between commendation and conferring of bene- 
 fice. 
 
 I. The primary aims of commendation were to secure protec- 
 tion and insure immunity. Protection oftener than immunity seems 
 to have been the chief object of a common form of commendation 
 which appears to have gradually shaped itself in the tenth century. 
 According to this, the owner nominally gave up his sho to a stronger 
 personage, for one generation or more, in order, as said the accom- 
 panying document, " to borrow his might and influence " or " to 
 put a stop to the outrages of the provincial officials ", but in fact 
 reserved for himself his shiki as steward or bailiff of the sho. In 
 form, the commendor surrendered ownership and retained the right 
 of use; in reality, he retained possession and delivered only a part 
 of the dues from the land. On his part, the receiving person often 
 bound himself by a written document to forfeit his nominal owner- 
 ship of the commended land, should he presume to meddle with its 
 management, which rested as before in the hands of the commendor. 
 
 As immunity became more prevalent and more widely coveted, 
 it was a growing tendency among holders of sho to mount higher 
 in the ascending steps of their commendations until they should 
 reach an imperial personage, a great temple, or a noble person above 
 the third rank. It was in this manner that shiki of a sho came to 
 be diversified and to be distributed among several persons and insti- 
 tutions, always with a high immune personage at the apex {hon-sho 
 
 31 Professor K. Nakada's able paper was published in various numbers of 
 vol. XX. of the Kokka gak'-kwai zasshi. I acknowledge my indebtedness to him 
 in so far as regards the larger legal aspects of commendation and the benefice. 
 
12 K. Asakawa 
 
 or hovr-ke), who was able to override the local governor.^^ As im- 
 munity was becoming sometimes judicial and administrative as well 
 as financial, commendatiog of this kind seems to have tended to 
 prevail everywhere, and in fact, after the eleventh century, helped 
 immunity itself to extend with facility. At the end of the twelfth, 
 there could have been few sho which had not found their nominal 
 immune owners; in fact, it was the boast of Fujiwara nobles that 
 they controlled numerous sho but rijanaged none themselves.^^ It 
 should be remembered that the original commendors had usually 
 contrived to retain their management and possession. 
 
 In any instance, it will be noted that the Japanese commenda- 
 tion was essentially contractual in character, and created a freer 
 status for the commendor than was usually the case with commen- 
 dation in medieval Europe, with further characteristic differences. 
 The commendor, as has been seen, transferred to the commendee cer- 
 tain specified rights, but retained for himself other real rights of 
 the sam.e land; these reserved rights he was free to dispose of as 
 he pleased, by means of commendations to other persons; and the 
 several commendees thus created were likewise able to divide and 
 convey to others the rights they had received — the result of this 
 flexible arrangement, even aside from the conditions due to other 
 causes that encumbered the normal sho, being often a network of 
 customary rights of bewildering complexity.'* 
 
 2. The benefice,'^ pure and simple, of shiki of the sho may be said 
 to have originated with the privileges its owner granted to the men 
 
 32 In I0I7 people of a sho in Kii under the powerful noble Michinaga com- 
 plained to him about a certain action of a provincial magnate, whereupon he 
 summoned the latter to Kyoto. Mi-id kwan-paku ki, Yale MS., bk. 15. In 1070 
 priests of T6-dai-zhi invaded a district in Iga with a band of lawless men, 
 arbitrarily determined the boundaries of a sho which the temple claimed, and 
 maltreated a provincial official, robbing him of his horse and clothes ; the govern- 
 ment at Kyoto was unable to punish the culprits. T. Yoshida, Td-jo Ni-hon shi, 
 VIII. 268. 
 
 33 When, in 1185, Fujiwara Kanezane was offered by Yoritomo the steward- 
 ship of a sho belonging to an imperial consort, he expressed his chagrin in his 
 diary thus : " It would be exceedingly ugly to make me Steward of Her Lady- 
 ship's sho ; there never has been such a thing in the family." Gyoku-yo, chap. 42. 
 
 3* It requires all the analytical power one can command to follow the maze 
 of the astonishingly intricate tenures pertaining, for example, to the sho of 
 Kamino and Makuni, through documents scattered over the eight volumes ot 
 Ko-ya san mon-zho. The same may be said of any of the larger sho belonging to 
 K6-ya-san, T6-zhi, or Iwashimidzu. A full analysis of the tenures of Shimadzu 
 sho in southern Kyushu would fill a large volume. 
 
 35 Professor Nakada's treatment of this precarious tenure is specially lucid 
 (see Kokka gak'-kwai sasshi, XX. ix, 66-90), but nearly all the sources he cites 
 are dated after the thirteenth century. The general features of the tenure are 
 well shown, but there is much to be studied as to the exact nature and evolution 
 
Feudal Land Tenure in Japan 1 3 
 
 who were employed in the actual management and cultivation of 
 the land. These employees appear ever since the early part of the 
 eighth century under a great variety of names; their services were 
 rewarded with the use of pieces of land in the sho, with certain por- 
 tions of its dues, or in produce or money. Their appointment and 
 support were, though accompanied by a written form, founded on 
 no mutual agreement, and were regarded purely as an act of favor 
 on the part of the master, granted, as was explicitly stated in the 
 charter, for the .employee's earnest " prayer ". By the act of accept- 
 ing the benefice, therefore, the grantee imposed upon himself one- 
 sided, precarious obligations;*'' the relation established by the act 
 was more real than personal, and at first one rather of fact than of 
 right. When the use of land was allowed, it could in no wise be 
 looked upon as rented; and the furthest extent that the grantor's 
 favor could reach was for him tO' state that the tenure of the land 
 would be contingent upon the performance of his duties. The Jap- 
 anese on-kyu thus closely resembled the precarium of medieval 
 Europe, revocable at will; even when the practice later tended to- 
 ward a hereditary holding of the privileges, in Japan as in Europe, 
 the original precarious nature of the grant was often maintained in 
 form. 
 
 When one considers that commendation and beneficing were 
 in simultaneous practice, he may well imagine real rights created 
 by both exercised side by side within the same sho. Such 
 was indeed true. This is not to say, however, that they readily 
 merged into or materially influenced each other. As a matter of 
 fact, till far along into the feudal period, the tenures in all their com- 
 plexity of these two different origins persisted independently, though 
 in close juxtaposition.*^ The resulting sho was not a manorial or- 
 
 af the more important classes of the on-kyu. For this study, the abundant 
 materials in possession of those historic Buddhist temples in Japan which were 
 great holders of land constitute indispensable though difficult sources of in- 
 formation. 
 
 36 It is, for example, instructive to compare, through their oaths of allegiance 
 dated between 1265 and 1315, the relatively freer tenure of the Ban-gashira and 
 other officials with the more precarious condition of the So-tsui-ho shi, the 
 Ku-mon, and their agents, in the sho of Kamino and Makuni. See Ko-ya son 
 mon-zho, I. S03-518, VII. 187-197, 199-214, 216--223, 225-226. Both classes of 
 officials lived upon the sho, and many of them were original possessors of parts 
 of the land and had held their posts for generations, but the first class consisted 
 of direct representatives of landholders, and the second, of beneficed agents. 
 
 37 For instance, glance over the succinct comments attached to the numerous 
 estates controlled by the regent Michi-iye, in his testament dated 1250, and note 
 the diversity oi conditions brought about through division and transfer by gen- 
 erations of holders of lands and real rights which had originally been created by 
 cultivation, purchase, commendation, or beneficing. Ko-mon-zho rui-san (third 
 ed.), pp. 285-316. 
 
14 K. Asakawa 
 
 ganization comprising strips of arable land laid out and adminis- 
 tered by the joint intervention of the lord and the half-free tenants, 
 but a large unit^^ in a " scattered farm " system, which was an ag- 
 glomeration of fields of utmost irregularity in. form and size and of 
 great diversity in origin and actual condition; the interference of 
 the community for the division and management of all of its land had 
 little opportunity to develop in an organization where comparatively 
 free and unfree tenures persisted side by side. 
 
 Now the crucial question arises : How, then, did the feudal land- 
 tenure originate? The fief, as we find it in the Japan of subsequent 
 ages, aside from its military relation, was practically contractual, and 
 was hereditary and capable of subinfeudation,'^ while the benefice 
 was, even in its later state, essentially precarious, and non-divis- 
 ible without special sanction, though often hereditary. Could the 
 one flow out of the other in the process of normal evolution ? When 
 the armed man entered the estate, was the, fief naturally born of the 
 benefice? In Europe results of commendations may have been 
 readily assimilated to the common precarious status and have lost 
 traces of their separate origin; but in Japan the tenure of commen- 
 dation was decidedly freer, being specific, divisible, and transferable, 
 and was inclined to maintain intact its own genealogy of rights and 
 obligations. If merging did not take place on any large scale, where 
 could the fief have arisen? The study of the whole question is more 
 complex than it appears, and its definite institutional features re- 
 main still practically unexplored. I shall essay to evolve a working 
 hypothesis by examining briefly the process by which the private war- 
 rior made his appearance on the scene and the conditions under 
 which he gained control over landed institutions. 
 
 IV. 
 
 The exact institutional aspects of the rise of . the Japanese 
 warrior are not so clear as is generally assumed among scholars 
 in Japan; nor is it the pjirpose of this paper to discuss them in 
 detail. It will suffice to indicate the probable relation of the subject 
 to that of land tenure. The Japanese warrior was, like the shS, 
 
 38 The extents of the five sho established in Noto between 1136 and 1146 
 ranged between 60 and 400 acres, and one was as large as 1000 acres. See 
 reference in note 29 above. These were among the larger sho. 
 
 S9 In 1206 the Shogun's government ruled that no landed possession granted 
 by the late suzerain would be revoked, except in case of a grave offense. Adzuma- 
 kagami, XVIII., in Zoku koku-shi lai-kei, IV. 673. In the official manual of the 
 feudal government of 1232 {Jo-ei shiki-moku) , the right of the military holder 
 of land to divide it among his children and relatives is admitted (in c. XX., XXII., 
 and XXVI.), and his right of sub-granting it is implied (in t. XIX.). 
 
Feudal Land Tenure in Japan 1 5 
 
 which he later controlled, largely private in origin ; he had his illicit 
 birth in the social unrest that had resulted from the unforeseen fail- 
 ure of the reformed state established in the seventh century to secure 
 peace and order. No sooner had the Reform been instituted than the 
 nobility and clergy began to extend their control over a rapidly en- 
 larging portion of taxable land and taxable people at the cost of 
 the state.*" The latter's expenditures, on the other hand, greatly 
 increased from the eighth century. The burden of taxation was 
 felt with growing heaviness by those people who still remained 
 under the state's control and whose number was rapidly falling off. 
 Those who would desert their allotted lands, but were unable to at- 
 tach themselves to nobles and temples, drifted away as outlaws. 
 The growth of this floating population had become alarming al- 
 ready at the beginning of the eighth century, and the government's 
 efforts to restore order were unavailing, even at the capital. The 
 society was restless, police laws were defied, and official soldiers*^ 
 proved utterly ineffective against the deepening disorder. 
 
 *o In 1 197, that is, soon after the beginning of the feudal rule of Japan, the 
 proportions of land under official control in the provinces of Satsuma, Osumi, 
 and Hyuga, were, respectively, as follows : s per cent., 8 per cent., and .3 per cent. 
 See Ken-kyu dzu-den cho, in Shi-seki shu-ran (revised), XXVII. 46-70. The 
 following provinces were more fortunate : Tajima, 4 per cent. ; Bungo, 12.5 per 
 cent. ; 'Hitachi, 22.5 per cent. See T. Yoshida, Dai Ni-hon chi-mei shi-sho, intro- 
 duction, p. 77. In these cases, the remainder of the areas consisted of sho and 
 other species of more or less immune land ; nor was the " official " land entirely 
 taxable. 
 
 Earlier instances follow: an official report from Etchu in 1050 says that, 
 owing to the recent rise of sho in the province, " people for public labor are 
 scarce, and taxable land is very little" Kan-chu ki, in Etchu shi-ryo, I. 211— 212. 
 In 1 1 07 the province of Kii reported that, of its seven districts, "six have each 
 eight-or nine-tenths become sho. . . . Especially in the two districts of Ito and 
 Naka . . . one or two villages only are left." The seventh district is also honey- 
 combed with sho. Ko-ya san mon-zho, IV. 244-245. 
 
 The taxable population given in the Chinese dynastic history of Sung from a 
 Japanese source was, at the end of the tenth century, 883,329. The entire popula- 
 tion of this general period has been estimated as not less than four or five 
 millions. There is little doubt that the population was increasing, but the de- 
 crease of the taxable population recorded in the census was alarming. In 914 a 
 councillor stated that, out of the Ki-nai, Kyushii, and the two northern provinces, 
 there were only 300,000 registered taxable people, of whom, owing to the amazing 
 inaccuracy of the census, more than a half were non-existent. He cites the case 
 of a town in Bitchu, which in 647 could furnish 20,000 able-bodied, adult men, but 
 in a hundred years saw its taxable population reduced to 1900, in another century 
 to 70, and in 911 had not a soul registered. Miyoshi Kiyoyasu's memorial, in 
 Hon-cho bun-sui (1888), II. 11-12. 
 
 41 Official soldiers levied from among the people by a system of conscription 
 were abolished in 739 and 792, except in the two northernmost provinces and in 
 the islands of Sato and Kyushii; those in the last were reduced almost by one- 
 half in 814. Minor varieties of official military forces proved as useless, and all 
 gave way, so soon as the early ninth century, before the growing disorder and 
 the private warriors. 
 
1 6 K. Asakawa 
 
 As things grew worse, it became a universal custom for the 
 more competent among the country folk to provide themselves with 
 arms for protection, though the private use of weapons was unlaw- 
 ful. It should be remembered that among the free citizens in the 
 country there were many of noble and even imperial descent,*^ who 
 possessed large tracts of land*' and numbers of dependent people," 
 and who had added to their prestige by serving at the capital as 
 guardsmen or other temporary officials.*^ The government depended 
 in a large measure for peace in the provinces on these very people.*' 
 Among these the less responsible men became themselves disturbers 
 of peace, especially when they had commended their land and them- 
 selves to temples or court nobles, and were thus enabled to defy 
 provincial officials with impunity. The latter, accordingly, when they 
 were appointed to their posts, often proceeded thither, in defiance 
 of law, with their male relatives capable of bearing arms.*' These 
 would permanently settle in the provinces ; the governors themselves, 
 who should be transferred elsewhere at the end of their terms, 
 tended to hold their office by heredity. In these two classes of 
 local magnates, old and new, who often quarrelled among them- 
 selves and against one another, may be seen origins of the warriors 
 (bu-shi) so largely private and illegal in character. Another and per- 
 haps more important source must be sought in those who had seen 
 service in the frontier garrisons of Kyiishu and the northern prov- 
 inces ; especially in the latter, where the Ainos resisted Japanese en- 
 croachment with tenacity, colonization of outlaws from other parts 
 of the country was encouraged on a large scale, and great expedi- 
 tions against the turbulent tribes were repeatedly organized, during 
 
 *2 References to descendants of the pre-Reform nobility and of imperial 
 personages are innumerable in the records. Governors of districts (_kdri) mostly 
 were local celebrities and held their posts by heredity. 
 
 43 When in the early eighth century the cultivation of new land was en- 
 couraged with promises of grant of ranks, there came forth men who had opened 
 surprisingly large tracts. 
 
 **With much local variation, the remains of the census of the eighth cen- 
 tury reveal at times the existence of great numbers of the unfree in individual 
 families. 
 
 45 The number of those who had at one time or another filled metropolitan 
 posts and who had returned to the provinces as privileged men was large. In 
 the early tenth century, nearly 3000 such men yearly left the capital. 
 
 48 The official soldiery being inefficient, the authorities were compelled, from 
 the last part of the eighth century, to enlist the service of the more wealthy men 
 in the provinces who were skilled in arms. 
 
 4f The violation of law in this respect may be inferred as early as 719 (see 
 Rui-zhu san-dai kyaku, in Koku-shi tai-kei, XII. 6461!.). Explicit and repeated 
 prohibitions of this offense during the two ensuing centuries were unavailing- 
 then, as will be seen later, the practice furnished, 'from the tenth century the 
 very foundation of a feudal nobility. 
 
Feudal Land Tenure in Japan 1 7 
 
 the eighth and ninth centuries, with men from among the colonists 
 and from the neighboring regions. These men and their descend- 
 ants, as well as chiefs of the Ainos who had surrendered, became 
 professional warriors, at first half official in nature but later purely 
 private. Court nobles, who were engaged in keen rivalry among 
 themselves for political power at the capital, sought to augment their 
 revenue, as well as enhance their prestige, by controlling numer- 
 ous sho in the country, which were well protected by armed men, 
 and by otherwise allying themselves with local chieftains. Many a 
 Fujiwara noble himself sought the provincial governorship, settled 
 his male relatives, and implanted his influence, especially in regions 
 north and east of Kyoto.*^ Such was the condition about 900. 
 
 Thereafter, with the parallel extension of immunity and com- 
 mendation, we may note more and more warriors included in the 
 sho. Some had themselves been commendors of land; others had 
 been invested with shiki by the legal owners of estates." We may 
 suppose that, in the latter class of cases, the employment of warriors 
 as agents of the sho was opportune as protection against the pre- 
 vailing disorder, for the warrior in the sho would be as efficient for 
 its security as he would be dangerous out of the sho as a marauder. 
 He probably served also as a safeguard against the original com- 
 mendors, who, as the actual possessors of the land, had often shown 
 little scruple, when they were able, to terminate their agreements 
 with the nominal owners and enter into contractual relations with 
 other nobles or temples.^" It should be remembered, however, that 
 nearly all the owners of the sho under whom warriors served as stew- 
 ards and agents still belonged either to the civil nobility or to the 
 clergy, and that there as yet existed no military nobility. 
 
 48 See the vivid story told a little later of a Fujiwara chieftain in Etchii in 
 Kon-zhaku mono-gatari, vol. XXVI., no. 17. His descendants lorded over the 
 three adjoining provinces. 
 
 49 The case of the contiguous sho of Kamino, Makuni, and Sarukawa, is 
 typical. They grew around cultivated tracts that had been commended before 
 the twelfth century to a monastery by their warrior-owners. From this time to 
 the beginning of the fourteenth century, numerous documents reveal the de- 
 scendants of the commendors and other armed men as "residents" (ju-nin), 
 "land-holders" (.ji-nushi), and hyaku-sho (bearers of family-names ?), these 
 terms being often used interchangeably. Some of them acted as representatives 
 of their rank; some were appointed as agents (often faithless) of the monastery. 
 Many of the actual tillers of the soil were in a servile position. See Ko-ya san 
 mon-zho, I., doc. nos. 447-450; VII., doc. nos. 1581, 1583-1590, 1592-1600, 1604, 
 1607-1608, 1610, 1612, 1615, 1619, 1623-1624. 
 
 50 In 1 106, an agent of Kasawara sho was dismissed for an offense, and, in 
 his chagrin, commended the sho to a temple. 
 
 AM. HTST. REV., VOL. XX. -2. 
 
1 8 K. Asakawa 
 
 V. 
 
 The tenth century was just witnessing the beginning of a new 
 class of nobility which was military almost from the date of its birth 
 — ^though its military character was essentially private — and which 
 rose in influence with rapidity in the next century and a half. Certain 
 political and economic exigencies of the court impelled many mem- 
 bers of the imperial family, from the early eighth century on, to re- 
 nounce their privilege, to assume new family-names, and to become 
 nobles of inferior rank. Of these, the more ambitious at once sought 
 fortune away from the capital, after the tenth century, both as local 
 officials, whose terms they helped to make hereditary,^^ and as man- 
 agers of great sho.^'^ In the latter capacity, the position of the new 
 nobles, living as they did on the ground, was certain to become much 
 stronger than that of the absentee civil owners of the sho. In both 
 capacities — as governors of kuni and as stewards of sho — ^the new 
 nobles were armed men, for to arms the times had driven all men of 
 ambition in the provinces. During the tenth century, the military 
 nobles, though in the main belonging to the two families Minamoto 
 and Taira, were, as was revealed at the rebellion of Masakado in 
 939-940, still largely unorganized, and their alliance with the local 
 warriors, who had first grown up independently of them, was still 
 limited though increasing. The military nobility was as yet hardly 
 in a position to cope with the civil nobility. 
 
 The situation changed rapidly from the eleventh century, when 
 the increasing restlessness among all classes of the people impelled 
 the more active men to unmake and remake alliances in manners 
 that seemed to suit their interest best. The actual possessors of the 
 sho, who had once commended their land to civil nobles or temples, 
 now largely transferred similar shiki to military nobles ;°' and the 
 deed was done with ease, as the shiki were in the process of becom- 
 ing more flexible and divisible than ever. The mutual advantage of 
 the new arrangement seemed evident, for thereby the possessor, 
 
 61 The first Taira (899), the first Minamoto (first half of the tenth ■century), 
 and their sons, all filled posts in the local government, to say nothing of their 
 descendants. 
 
 52 To cite only two instances from the tenth century. Hidesato was steward 
 of the strategic Tawara sho, while being constable of Shimotsuke, and in this 
 double capacity was a powerful tool in eastern Japan in the hands of the Fujiwara 
 nobles. Mitsunaka, a provincial governor, managed Tada sho, and was a faithful 
 servant to his Fujiwara chiefs. 
 
 63 Already in 1091 the government at Kyoto forbade the people in the 
 provinces, with little effect, to make further commendations of landed rights to 
 the Minamoto chief. Hyaku-ren sho, V., in Koku-shi tai-kei, XIV. 55. From 
 the early years of the next century similar prohibitions were as frequent as 
 unavailing. 
 
Feudal Land Tenure in Japan i g 
 
 who was himself a warrior, secured a new master who was always 
 at hand, affording constant protection and manifesting kindred in- 
 terest and sympathy, and under whose growing prestige it was an 
 honor to serve ; the warrior-lord, too, was enabled by this method to 
 include more land and more men under his control. Actual warfare, 
 which was then becoming frequent, accelerated the process once so 
 naturally begun. For during the protracted campaigns in the north, 
 the men were thrown together in a close relationship; even in those 
 local warriors who had joined the expedition merely from motives 
 of opportunism was gradually engendered a feeling of personal loy- 
 alty to the chieftains ; while the latter's interest had become increas- 
 ingly bound up with that of the warrior class as a whole as against 
 the civil nobility. Landed possessions and personal following grew 
 henceforth in a rapid process of accretion; the private organization 
 of men and land gained, on the one hand, in hierarchical unity un- 
 der the impetus of the revived spirit of clannism, and, on the other, 
 in complex divisions of families and their shiki in land. 
 
 We shall not discuss here the personal side of the new relation- 
 ship — the forms of fealty and protection between lord and vassal, 
 and the question of the knight's service. What should specially be 
 noted is the effect of the social changes upon land tenure. 
 
 First, let us consider the position of the military lord serving 
 as protector and manager of a sho nominally owned by a temple or 
 a civil noble of the court. The former was an invested agent of 
 the latter, and his position was accordingly precarious in principle; 
 the shiki which the manager enjoyed were presumed to have origi- 
 nally been granted as a favor, not as a reward for service ; and the 
 charter creating his tenure and his own acknowledgment of the char- 
 ter explicitly stated that his office would terminate with a lapse from 
 duteous conduct.^* However, even the tillers and their chiefs, em- 
 ployed, as they were, for the mere economic exploitation of the soil, 
 whose servile tenure was stated as revocable when they failed in 
 duty, are seen to have often held their place, after the twelfth 
 century, by heredity. For the manager, who supervised these men, 
 there were, as a matter of course, circumstances which deprived his 
 tenure of its seeming precariousness. He was himself a noble, and, 
 as such, usually found benefice in an extensive sW'' comprising 
 
 5* Even so late as in the fourteenth and fifteenth centuries the charters and 
 acknowledgments of mUitary managerships of ecclesiastical sho were often plainly 
 precarions in expression. See, e. a-, those of 1363, in Iwashimidm mon-zho, I. 
 283 ; and of 1359 and 1462-1463, in To-zhi hyaku-go mon-zko, MSS., part ro. 
 
 55 A striking instance is that of the great sho of Shimadzu in southern 
 Kyushu. About 1025 a local official cultivated a tract in this region, and com- 
 
20 K. Asakawa 
 
 numerous tracts and people of diverse origins and conditions; his 
 authority as manager of affairs and leader of armed men was nec- 
 essarily great. When he was, in addition, an hereditary high official 
 of a province, and, as often happened, had the management of many 
 sho within its confines,°^ his position was commanding. Even in a 
 less pretentious place, he usually gathered about him relatives and 
 other dependents that, along with his sworn men, squatted on the sho- 
 all of whom he could mobilize at an instant's notice. The civil noble 
 at Kyoto would think twice before interfering with his agent in the 
 country whose post seemed so impregnable. The latter showed at 
 times a disposition to invade neighboring estates for purposes 
 of plunder at the head of hundreds of mounted warriors; might 
 not the same force perhaps be turned against the nominal master 
 at court? The position of the chief as manager, therefore, was far 
 from precarious, was often stronger than a contractual tenure, and 
 was perforce hereditary, as was also that of the many men under 
 his military control who would rise and fall with their lord. All 
 this had come about by an irresistible force of necessity, in which 
 the civil and religious owners of the sho had no choice but to ac- 
 quiesce. 
 
 Consider, next, the case of the small warrior serving under a 
 military noble who was either the manager or the owner of the 
 sho. His relation to the lord was either that of a commendor and 
 possessor of land or that of a beneficed servant. The distinction 
 between these two kinds of status was indeed maintained at least 
 till the end of the Kamakura period in the early fourteenth century. 
 During that same period, however, the growing tendency was to- 
 ward the gradual approach of the tenure of possession to the more 
 precarious tenure of the invested shiki in land. In the feudal offi- 
 cial manual of 1232, the hereditary possession of the immediate 
 
 mended it to the regent Yorimichi, reserving for himself and his descendants the 
 right of its financial management. Gradually annexing estates of different origins 
 and tenures, the sho came to include the greater part of the three provinces, 
 Satsuma, Osumi, and Hyuga, comprising a variety of possessors and beneficed 
 agents. Of these agents, a family which assumed the name Shimadzu became 
 conspicuous toward the end of the twelfth century as the managers-general of 
 the entire sko. During the troublous fifteenth and sixteenth centuries, the 
 Shimadzu succeeded in reducing their rivals within the vast sho and making 
 themselves its sole possessors. In this position they were respected and feared 
 by the Tokugawa shogunate all during its rule after 1600, and finally were 
 instrumental in accomplishing its downfall in 1867-1868. 
 
 56 As a result of subtle divisions of real rights, civil and religious owners 
 often found it convenient to bundle several sho that were near together and to 
 place over them a military manager. Not infrequently Minamoto and Taira chiefs 
 in this way had the management of all the sho in whole districts and provinces, 
 of which they were governors for generations. 
 
Feudal Land Tenure in Japan 2 1 
 
 vassal of the suzerain, but not his beneficed shiki, might be trans- 
 ferred by sale ;" from 1270 onward, neither could be so conveyed,^* 
 while both could, as before, be divided and bequeathed amotng 
 children — the law of primogeniture not having as yet been fully 
 established — and be subinfeudated.^' This trend of assimilating the 
 two tenures and bringing about a gradual evolution of a tenure less 
 free than the one and less precarious than the other, that came to a 
 marked stage in the thirteenth century and completed itself in the 
 fifteenth and sixteenth, may, it is surmised, have shown its faint be- 
 ginnings in the early period under our discussion.^" It would seem 
 probable that in the personal relation in military service between 
 manager-lord and vassal, the latter as a commendor and possessor 
 of land could not, even in this early period, be nearly as independent 
 as he had been under the absentee civil or religious master ; while, as 
 a beneficed servant, he would be likely to be treated with more con- 
 sideration by his new lord, for the former's service was deemed hon- 
 orable, was akin to that of the latter, and was directly needed in as 
 efficient a form as possible. 
 
 To sum up the foregoing discussion. The fiscal immunity of 
 " imperial ", " granted ", and " temple " lands induced the commenda- 
 tion of taxable land and people into growing private estates. The 
 state, seeing its sources of revenue and power fast dwindling away, 
 itself adopted the methods of the nobility and clergy to administer 
 public affairs much as if they were private concerns. The inveterate 
 habit of the political mind of the pre-Reform age seemed to have 
 returned in full vitality. In their disposition of the rights and facts 
 relating to land, the Japanese displayed their genius in an arrange- 
 ment of remarkable flexibility with a large capacity for compromise. 
 And this very condition made it a matter of ease for a firm hold on 
 private land to be established by the warrior, another illegal creation 
 of the age unforeseen by the Reformers. A little later arose a mili- 
 tary nobility, which, in a century and a half, made itself the central 
 power by the double method of serving, on the one hand, as the 
 political and proprietary agents of the civil nobility at Kyoto, and, 
 on the other, of assuming lordship over the warriors in the country. 
 As agent, the military noble approached the position of the possessor 
 of the land in his charge ; as lord, he was tending to make the ten- 
 
 57/0-et shiki-moku, c. XL VIII. 
 
 58 Shin-pen tsui-ka, ^. LXV., LXXVII., LXXIV., LXXV. 
 
 50 See note 39 above. 
 
 60 I have been unable to substantiate this surmise from the available con- 
 temporary sources, which relate chiefly to civil and ecclesiastical sho and rarely 
 to those of military ownership. I hold the theory as a probable supposition till I 
 may find it either confirmed or disproved by fresh evidence. 
 
2 2 K. Asakawa 
 
 ures of his commendor-vassal and of his beneficed vassal alike a 
 favor in theory, though in practice a reward, alike accompanying a 
 military service, and alike hereditary and divisible, though inalien- 
 able by sale. In short, the military noble was in the process of be- 
 coming the lord of a fief capable of subinfeudation. The process 
 was slow to mature, for the civil and religious estates still persisted, 
 though in decreasing numbers, and continued to ofiEer some re- 
 sistance to the encroachment of the new tenure ; the complete assim- 
 ilation of the land of Japan into the feudal tenure was not effected 
 until the sixteenth century, and was quickly followed by another 
 development. We suspect, however, that, when all the sources shall 
 some day be brought to light, beginnings of the long process may 
 perhaps be discerned as early as in the period ending in 1186. 
 
 VI. 
 
 We have sketched the main outlines of the evolution of landed 
 institutions before the advent of the feudal regime. It now remains 
 to touch very briefly upon the general movement of historical events 
 in the eleventh and twelfth centuries, which hastened the processes 
 that have been described and which culminated in the inception of 
 the regime in 1186. 
 
 In the eleventh century, social unrest had become more organized 
 as the military forces of the great warrior-clans were better co-or- 
 dinated, and broke out in serious insurrections, in 1028, 1052-1061, 
 and 1088^1091. Most of the leaders of the rebels were men of the 
 new nobility, and were vanquished only by other men of their class 
 commanding their own warriors. The enervated civil nobles at 
 Kyoto were Buddhists of the old type and abhorred the sight of 
 blood. They were compelled to depend on the Minamoto and Taira, 
 not only for the suppression of local disorders, but also in their 
 mnitual rivalry and intrigue for power. The military nobles were 
 not only establishing their influence in the country, but also becom- 
 ing a decisive factor in the political struggles at Kyoto; they 
 were not only owners and stewards of shS, but also indispensable 
 tools for the Fujiwara nobles and imperial aspirants for supremacy 
 at court, to whom the warlords had sold their temporary fealty. 
 
 Events since the early eleventh century resulted in driving the 
 Minamoto and Taira clans to organize themselves in larger and 
 larger groups, finally dividing military Japan between them into 
 two distinct spheres of influence, the Minamoto in the east and the 
 Taira, in the west. Presently, the political factions at the capital, 
 which employed the arms of the two clans, came twice to 
 a sharp conflict, in 11 56 and in 1160, that decided the ascendancy of 
 
Feudal Land Tenure in Japan 23 
 
 the Taira for a brief quarter of a century, to be followed by a more 
 complete and lasting supremacy of the Minamoto. In this swift 
 flight of events, the civil nobility was thrust into the background as a 
 political factor and dispossessed of many of its landed estates, the 
 power of the great temples was curbed, and the fortune of military 
 families rose and fell with tragic rapidity. The real rights of slw 
 changed hands on large scales,'^ the warriors in the meantime gain- 
 ing an extended control over them; vassals transferred their allegi- 
 ance in large numbers and with frequency ; and the immunity of the 
 sho, especially of those owned by warriors, usually amounted to 
 political autonomy. Finally, in 1186, Minamoto-no-Yoritomo com- 
 pelled the reluctant imperial government to sanction his appointing 
 his own vassals as constables in all the provinces, and as stewards 
 of all the sho, civil, religious, and military,^^ in Japan. With this 
 her feudal rule may be said to have begun. 
 
 K. ASAKAWA. 
 
 61 The Taira appointed their vassals as stewards, not only over their own 
 sho, but also over civil and religious sho (see Adsuma kagami, V., in Zoku 
 koku-shi tai-kei, IV. 197), and over the sho confiscated from their former lords. 
 The family is said to have held governorships of 36 out of the 66 provinces of 
 Japan, and controlled ishiki of] more than 500 sho, besides innumerable minor 
 estates. All these extensive shiki were confiscated from the Taira at their 
 downfall, and passed under the control of the Minamoto. 
 
 62 Beginning shortly after 1186, the newly appointed stewards (jJ-W) seem 
 to have been gradually withdrawn from some of the civil and ecclesiastical sho. 
 (See, for example, the withdrawal in 1227 from the sho belonging to a monastery, 
 in Kd-ya san mon-zho, VII. 182, 253.) The suzerain, however, continued to 
 hold, through the new constables {shu-go), the military and police power over 
 all the provinces, which comprised within them all the existent sho.